<%BANNER%>

A New Model for Analyzing Sociolinguistic Variation

Permanent Link: http://ufdc.ufl.edu/UFE0022375/00001

Material Information

Title: A New Model for Analyzing Sociolinguistic Variation The Interaction of Social and Linguistic Constraints
Physical Description: 1 online resource (335 p.)
Language: english
Creator: Habib, Rania
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: algorithm, arabic, change, choice, colloquial, constraints, effect, frequency, gradual, grammatical, learning, linguistic, migrants, optimality, psychological, quantitative, reality, rural, selection, social, sociolinguistics, syrian, theory, universality, urban, variation
Linguistics -- Dissertations, Academic -- UF
Genre: Linguistics thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: In this study, I present a new model for analyzing sociolinguistic variation within the framework of Optimality Theory (OT) and the Gradual Learning Algorithm (GLA). This model contributes to the advancement of sociolinguistic methodology as well as to OT and the GLA, unifying both linguistic and social factors. I propose a number of social constraints and incorporate them with linguistic constraints in the GLA. I show that incorporating social constraints yields variant-usage percentages that match real life occurrences. I follow the evaluation process of stochastic grammar in which social constraints are treated like linguistic constraints on a continuous scale of ranking strictness. Their ranking values may differ from one speaker or group of speakers to another. These differences in the ranking values may result in intra- and inter-speaker, intra- and inter-group, intra- and inter-community, or intra- and inter-dialect variation. The number and type of social constraints involved influence the percentage of occurrence of each variant and affect the ranking values of other constraints and the grammar chosen by a speaker or a group of speakers at a particular time and place. This model enables us to show the effect of linguistic and social constraints simultaneously. It allows us to investigate individual as well as group grammars and shows that those grammars may act independently from each other. It also enables us to project expected percentages of variation by manipulating either the ranking values of constraints or the pair distribution of the output. Evidence comes from the study of the naturally occurring speech of migrant rural speakers of Colloquial Arabic to the city of Hims, Syria. These speakers show different degrees of variation, particularly regarding the use of the two variants q and ?, based on various social factors, such as age, gender, residential area, and social class. The aim of the study is to show how rural migrants adopt the new phonological system of the urban dialect to appear prestigious. The study will focus on the sound change of q to ? and the four stable variants t and s and d and z. It will show that variation may be influenced by different factors. The variable use of q and ? is attributed to prestige and social factors. The use of t and s and d and z is not social in nature; it has developed historically from the Standard Arabic (SA) voiceless interdental fricative and ? respectively as a response to markedness constraints. Today, this variation is stable and each of the four variants is used in specific lexical items. Faithfulness constraints play a major role in maintaining the pronunciation of the input as t and s and d and z in the output. This stable phenomenon is further explained in terms of the two opposing effects of frequency (Bybee 2001). In a situation where prestige plays a role in adopting a new variant, as in the case of rural migrants to the city of Hims, social constraints are viewed as the motivating force behind changing a grammar at a particular time and place. However, one should take into account that the activation of social constraints depends on the speaker?s selection or choice to activate them or not. Hence, these constraints may rank very high in the speech of a speaker who is highly aware of the social values attached to a certain sound. On the other hand, they may rank very low in the speech of a person who, for example, does not care to adopt a new form. A theory that can take into account all the factors that lead to variation is a better theory than one that takes into account social factors in isolation and ignores grammatical factors or vice versa. Integrating social constraints into formal theory and showing that social constraints can have equal weight or even more weight than grammatical constraints in conditioning variation and change is essential in this study. Including social constraints in the computation provides explanation of the observed sociolinguistic variation between q and ? among members of the same social group or among different speakers or social groups. Interacting social constraints with linguistic constraints in the same framework is a simple comprehensive method to depict and explain the mental process of a speaker at a certain time and/or place. Feeding the GLA with the right output distribution gives the specific ranking values or grammar of each speaker or group of speakers. In other words, the model takes advantage of the stochastic grammar embedded in the GLA to generate grammars that match real life output percentages without trying all the possible rankings of constraints and counting the times those rankings give a certain output. Furthermore, where a statistical analysis fails to indicate the interaction between one social factor and others, the specificity implemented in the GLA by dividing a social factor into a number of social constraints enables us to see interaction among the same social constraints that emerged as insignificant in the statistical analysis. In this sense, implementing social factors as constraints and accounting for sociolinguistic variation within the framework of OT and the GLA has an advantage over other theories.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Rania Habib.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: McLaughlin, Fiona.
Local: Co-adviser: Wiltshire, Caroline R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022375:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022375/00001

Material Information

Title: A New Model for Analyzing Sociolinguistic Variation The Interaction of Social and Linguistic Constraints
Physical Description: 1 online resource (335 p.)
Language: english
Creator: Habib, Rania
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: algorithm, arabic, change, choice, colloquial, constraints, effect, frequency, gradual, grammatical, learning, linguistic, migrants, optimality, psychological, quantitative, reality, rural, selection, social, sociolinguistics, syrian, theory, universality, urban, variation
Linguistics -- Dissertations, Academic -- UF
Genre: Linguistics thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: In this study, I present a new model for analyzing sociolinguistic variation within the framework of Optimality Theory (OT) and the Gradual Learning Algorithm (GLA). This model contributes to the advancement of sociolinguistic methodology as well as to OT and the GLA, unifying both linguistic and social factors. I propose a number of social constraints and incorporate them with linguistic constraints in the GLA. I show that incorporating social constraints yields variant-usage percentages that match real life occurrences. I follow the evaluation process of stochastic grammar in which social constraints are treated like linguistic constraints on a continuous scale of ranking strictness. Their ranking values may differ from one speaker or group of speakers to another. These differences in the ranking values may result in intra- and inter-speaker, intra- and inter-group, intra- and inter-community, or intra- and inter-dialect variation. The number and type of social constraints involved influence the percentage of occurrence of each variant and affect the ranking values of other constraints and the grammar chosen by a speaker or a group of speakers at a particular time and place. This model enables us to show the effect of linguistic and social constraints simultaneously. It allows us to investigate individual as well as group grammars and shows that those grammars may act independently from each other. It also enables us to project expected percentages of variation by manipulating either the ranking values of constraints or the pair distribution of the output. Evidence comes from the study of the naturally occurring speech of migrant rural speakers of Colloquial Arabic to the city of Hims, Syria. These speakers show different degrees of variation, particularly regarding the use of the two variants q and ?, based on various social factors, such as age, gender, residential area, and social class. The aim of the study is to show how rural migrants adopt the new phonological system of the urban dialect to appear prestigious. The study will focus on the sound change of q to ? and the four stable variants t and s and d and z. It will show that variation may be influenced by different factors. The variable use of q and ? is attributed to prestige and social factors. The use of t and s and d and z is not social in nature; it has developed historically from the Standard Arabic (SA) voiceless interdental fricative and ? respectively as a response to markedness constraints. Today, this variation is stable and each of the four variants is used in specific lexical items. Faithfulness constraints play a major role in maintaining the pronunciation of the input as t and s and d and z in the output. This stable phenomenon is further explained in terms of the two opposing effects of frequency (Bybee 2001). In a situation where prestige plays a role in adopting a new variant, as in the case of rural migrants to the city of Hims, social constraints are viewed as the motivating force behind changing a grammar at a particular time and place. However, one should take into account that the activation of social constraints depends on the speaker?s selection or choice to activate them or not. Hence, these constraints may rank very high in the speech of a speaker who is highly aware of the social values attached to a certain sound. On the other hand, they may rank very low in the speech of a person who, for example, does not care to adopt a new form. A theory that can take into account all the factors that lead to variation is a better theory than one that takes into account social factors in isolation and ignores grammatical factors or vice versa. Integrating social constraints into formal theory and showing that social constraints can have equal weight or even more weight than grammatical constraints in conditioning variation and change is essential in this study. Including social constraints in the computation provides explanation of the observed sociolinguistic variation between q and ? among members of the same social group or among different speakers or social groups. Interacting social constraints with linguistic constraints in the same framework is a simple comprehensive method to depict and explain the mental process of a speaker at a certain time and/or place. Feeding the GLA with the right output distribution gives the specific ranking values or grammar of each speaker or group of speakers. In other words, the model takes advantage of the stochastic grammar embedded in the GLA to generate grammars that match real life output percentages without trying all the possible rankings of constraints and counting the times those rankings give a certain output. Furthermore, where a statistical analysis fails to indicate the interaction between one social factor and others, the specificity implemented in the GLA by dividing a social factor into a number of social constraints enables us to see interaction among the same social constraints that emerged as insignificant in the statistical analysis. In this sense, implementing social factors as constraints and accounting for sociolinguistic variation within the framework of OT and the GLA has an advantage over other theories.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Rania Habib.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: McLaughlin, Fiona.
Local: Co-adviser: Wiltshire, Caroline R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022375:00001


This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101221_AAAAAK INGEST_TIME 2010-12-21T09:36:37Z PACKAGE UFE0022375_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 17692 DFID F20101221_AAALNY ORIGIN DEPOSITOR PATH habib_r_Page_290.QC.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
8ed176eb60cb04a0fd53a54842b6b5e5
SHA-1
6a12db6bd19ddf7baf99218e36986ef0934c59ed
35633 F20101221_AAALAN habib_r_Page_046.QC.jpg
e0e50d50677a203a6c2c5cfb9b27818f
bec0373c058af9c9783bfeb3a9afd51bb2a0cb6e
25271604 F20101221_AAAJZE habib_r_Page_139.tif
b28e3f35afcae2768416c5df8bf154e6
1a3bb20df4ea11d9babe25df4e978828c1410fc2
F20101221_AAAKDU habib_r_Page_307.tif
72e6c1cf54407822246bcddbdffc642f
1b132b55e208b00d4097f10cf11df8f678cffba1
1051976 F20101221_AAAIPA habib_r_Page_100.jp2
7b021081ca7da315df986f90c37d857a
b3283430cdb83cd4e67fdceab0753ab3eb857925
5648 F20101221_AAALNZ habib_r_Page_290thm.jpg
33f865872d258734953dec52cd3459f8
000462bf72a185fa991d1d4c8bab248b5b1cced6
8787 F20101221_AAALAO habib_r_Page_046thm.jpg
8172052af087993fd16941bd5fe72770
19b493edf3d80c2ebbfcff2371850b7653d45304
F20101221_AAAJZF habib_r_Page_140.tif
57e58359229b4c31942fb64f8d3e9916
dc3197c14e373042d24064af0e942c9d1ed52ae5
F20101221_AAAKDV habib_r_Page_309.tif
bbb12c65d352cd26c14dfe666083fddc
9a2f40c2183ebc7552979becfbd063edf1690ec6
F20101221_AAAIPB habib_r_Page_062.tif
7c27960b8583e0bc2aa14539249abd3b
468b27a4fcd47627c02da1970394718d324e1735
38091 F20101221_AAALAP habib_r_Page_048.QC.jpg
e35f454e11ffb7a2ec845d38941fbfb9
125f96321ea7465e39b89c29b6a73bc195e8b98d
F20101221_AAAJZG habib_r_Page_141.tif
772b31cdcbd640963f53170a0247ce0d
d1cb62410c55503e03debaa54b7cf491f9e726d9
F20101221_AAAKDW habib_r_Page_310.tif
598515ff11d96a96aa0b993ea87ca448
a3ae002a0183881b9a1fc29563c9c07bbe3ba66d
F20101221_AAAIPC habib_r_Page_019.tif
0d827588b0b1fb38786e5d666f420304
4c04be4d051b64aac208cbde3aec08e10671890f
9085 F20101221_AAALAQ habib_r_Page_048thm.jpg
7c451e97600c262efcf0c4523fb69e6d
f298e5c086f3f64a6fc3cb48a3616e9a403aa24f
562 F20101221_AAAKWA habib_r_Page_270.txt
dc3ffe0197b83b3e43457f15541174a3
6feb5924dc8e845d254ba8cb6a9f5bc4c46ae276
F20101221_AAAJZH habib_r_Page_143.tif
2153f180dd6bee70fbf52d95f59e1349
b169d7bf55288c7e6cb7cacd1eef839845907088
F20101221_AAAKDX habib_r_Page_311.tif
c4d0ac4bf771a02864866655f56b1202
f2cca36a4c8463de0b75834e3fe148a9e4eff4fa
1051971 F20101221_AAAIPD habib_r_Page_229.jp2
cb93b6ee534e7b052c226f668b063fc5
7a6875a0ef522d33d7849d8addc3d24baf5caf19
39089 F20101221_AAALAR habib_r_Page_049.QC.jpg
6a824ece3682517246f810f263f43238
dd74ad36d05dee0a043cbf996b55237a855c119e
1546 F20101221_AAAKWB habib_r_Page_272.txt
5a346eb24510143fe0473c76c5a32cc7
5294f6c44bde639eaeacde7b7e9f8fa27af54405
F20101221_AAAJZI habib_r_Page_144.tif
611d6bbb6a8366825b6ae7e199e1aa09
91ddd117188468d65c76552691072a2b042efc10
F20101221_AAAKDY habib_r_Page_312.tif
1fe625487efc959f3c6d7e5e55a23abd
edaa9ab663dfa133d36c1d8b33c9601d0ced29ac
34736 F20101221_AAAIPE habib_r_Page_038.QC.jpg
6d1889784cd4777ef36eef958ec1e923
990958f86ee68a692597ee44832c20ba4811e2b8
9301 F20101221_AAALAS habib_r_Page_049thm.jpg
2e41526d5beb18e9d36dee7ad675ad97
ea1210c782272d2f023cff837d9354faebdb391e
893 F20101221_AAAKWC habib_r_Page_273.txt
3e83ab23889b05a73834129664ce8f44
0a7a73cf5abc7eb801714a7033d34c71e862c9b5
F20101221_AAAJZJ habib_r_Page_145.tif
386bbee87482670842a061e6afee7179
16e0de9516e67608f659608c5b7eb7709b59a382
F20101221_AAAKDZ habib_r_Page_313.tif
821a82734fb4e108cdff55714b5b3597
1ad91ca8ec426cc52e22d3eca056a2f26444cbf7
F20101221_AAAIPF habib_r_Page_030.tif
0424131a37b5d97a5c130722004a2937
cc030b12681d5c53d979341dd655b6b9e125434c
18466 F20101221_AAALAT habib_r_Page_050.QC.jpg
4655ec7ebf4ef8eb17f083724a891852
037d9a858a1afab260debe2bccae37f2c02e7de6
1820 F20101221_AAAKWD habib_r_Page_274.txt
e9adf2525f5f0ea50a7a57f626a3dd9a
b1199d0ff517d9c1ab1df03f2e59a0313ecd9414
F20101221_AAAJZK habib_r_Page_147.tif
24ff27ec356f2130068b069aadae94c2
b0c4100bbeaa841dabc4322e5e86a850a5e704dc
47971 F20101221_AAAIPG habib_r_Page_173.pro
ea3bb7a87123ee0f873c6a2f5fb8c3c6
94827d7f5bdd985af5bf546f10c8b3e39f09a030
5239 F20101221_AAALAU habib_r_Page_050thm.jpg
8827b61230f2366dccc467709f6e19af
046776970eb8adbb4ce9ccc37c3cde3d935728b3
905 F20101221_AAAKWE habib_r_Page_276.txt
1245aa1e813b7b7a3d9b50be998766d7
1cd6f5df4c51de0308167272206ba52c1aea2805
F20101221_AAAJZL habib_r_Page_148.tif
6d103f05dbc0e2d886aadb2da3cc9c0a
390ba9a62008e25eca3cd48ca4059dedb5689352
898 F20101221_AAAIPH habib_r_Page_196.txt
29638f38fa48504c995d4133794d400a
47bac2b5255bbaed2a7b52cd329f743217eb7f67
1051941 F20101221_AAAJMA habib_r_Page_024.jp2
6e57ea28ada758bafa22e4cdf6d2b36f
caeeb56ce0510ba5b39d65d037eecd40163d8da8
37407 F20101221_AAALAV habib_r_Page_052.QC.jpg
7a7a5870f5bac7622fe277ad6e6bc6ce
68d4481aff2de6a1673eb514f84a696e2797f776
646 F20101221_AAAKWF habib_r_Page_277.txt
15c3258e8ed181affc5af4537bff3176
28b75b9b71adb6d2452089a1f26c08cfb8c0ddac
F20101221_AAAJZM habib_r_Page_150.tif
b4b13c4e0bcb1c9c9b6bb7fe96be3d35
d3e5da437c6c1e38d4045c2a6bda254564a4f5c2
113944 F20101221_AAAIPI habib_r_Page_201.jpg
e792674f55e426e2c7b747a8978c8e79
8c02d52086851dbdc2c16cb9934f81b375934af8
1051960 F20101221_AAAJMB habib_r_Page_025.jp2
497cd78b1c1871ed50ebd079fce6b994
d2fd7cd28d817e88751d817fa698e8a54556f8b5
9235 F20101221_AAALAW habib_r_Page_052thm.jpg
1f70db9f2e28e4c4705878ad47fd620b
40df00ec077e315bbd04a510d92f9d06b2dc3cc1
653 F20101221_AAAKWG habib_r_Page_278.txt
ad8ddaa0683804c9456b58c45b658f84
5736fbced639b6351a09cea03ba9bfbc2929478b
F20101221_AAAJZN habib_r_Page_151.tif
74eb4a78eedb12c0d7f589f200b91f46
27dbc6e0b9b333859e504ae862761b3c980e9f9c
F20101221_AAAIPJ habib_r_Page_296.tif
b318a650f0e1edd8b94e656e6a577f6d
a59ea37653f0e70877af9b1bacfad05748d8eb54
1051982 F20101221_AAAJMC habib_r_Page_026.jp2
b738bc2c137fde3fdc08a0e6c525ba6a
3905946f817287bdfe282e1069d816940cbc53b8
36099 F20101221_AAALAX habib_r_Page_053.QC.jpg
f6f7b2d3d76e70defe6fe3129908d1ba
8ac050a1d4bf1950378630d42dc1a607344c9a37
1421 F20101221_AAAKWH habib_r_Page_281.txt
d8d12dabdeae441ee990bbd94db53e37
5c0c1c4b1056405841abe870addd89003b20ca4b
F20101221_AAAJZO habib_r_Page_152.tif
0deab49b05b9d9ee1c4b2c2b3e434fac
6574d7533cc167d7bd7f708051ca0601d9127773
66544 F20101221_AAAIPK habib_r_Page_085.jpg
3cb59a4777a13bda23260c4a68fe0945
4b6f527b6d42c72cf6f1a1fe6dec355a6b43c7ed
1051983 F20101221_AAAJMD habib_r_Page_027.jp2
946617c4de04c0fb2058a5e96e37ec5b
6072f52ccf1f9a06a079a259651686d6bc64332d
35317 F20101221_AAALAY habib_r_Page_054.QC.jpg
934e914ab19e82ae660fc59bfe19dfc3
e9029aacb2d43d5e086cf5dbedbf4ae35df89c7f
665 F20101221_AAAKWI habib_r_Page_283.txt
1bf647b7ca8e9767aeffc3969d2c6462
7428c66759ed98ee3dee0b0e65e93c26475ee63b
F20101221_AAAJZP habib_r_Page_153.tif
ef58cc0ee6999b10da2b3b073d49bc4e
ee36fa47dc97495722118b3fc042b8fa8ca0cfeb
1051966 F20101221_AAAJME habib_r_Page_028.jp2
400c4f393af5a214f1fbc6d07999bd9e
d459702385538fbb3239973d79d93741baf1d22e
1819 F20101221_AAAKWJ habib_r_Page_284.txt
7426db4c73a2e4526769c20161e296e9
33d28a6d6b526d8314da4b36ccdc2d96c6d79a6a
F20101221_AAAJZQ habib_r_Page_154.tif
3891e9a6e3e832e93e74b0c54180323d
ff13fe689f22f6ffc784dff14f94f7d8a418be23
F20101221_AAAIPL habib_r_Page_245.tif
5148168571b60255d6237dd703768a43
12f04cba4fe937f88d4686e7991aa6044555aa93
1051957 F20101221_AAAJMF habib_r_Page_029.jp2
943029fe4c4ed47555ae99f3f7c9e5ee
e35b22321a2268823f3f5c943b85e5b54b53c3b7
8632 F20101221_AAALAZ habib_r_Page_054thm.jpg
7b7a115792e6558eb5c20fe9396b27d1
014684f7cc39a093a275d49a2ec2b60583295396
1393 F20101221_AAAKWK habib_r_Page_285.txt
d7133a4ead18b5c27f5b1fc7a3aa9c07
639d9838beec342b58e3b459b5f9e0c6731f98f9
F20101221_AAAJZR habib_r_Page_155.tif
12da609569a63e4a1a0cbe407025ef39
bc9289b1a6231e69ebb445be1e16d6c409863650
32834 F20101221_AAAIPM habib_r_Page_145.QC.jpg
3276d90d009a45dd14ced978ad1c34aa
b9e5e51eafbc41615cbfe52b8f612385722c3e4f
1051926 F20101221_AAAJMG habib_r_Page_030.jp2
6de2e22d6beb58b5fd0f6fde4013a3b9
5a5765c5da6262d3aab5068d8f90e30217fb401a
1636 F20101221_AAAKWL habib_r_Page_286.txt
89c2c758da01db07a598d2b364f55f4d
107a849e6628db9e24602102867ee9bf3d1bf46f
F20101221_AAAJZS habib_r_Page_157.tif
d122c6d2efdf15bef4d73048e552776b
e6303bce176934565e0b1892336f7428655554df
16793 F20101221_AAAIPN habib_r_Page_309.QC.jpg
c3647261326207e5be40a749da6d97e5
c93cfab867a8cf6997489c581d2fbb0d3f46bf6b
52487 F20101221_AAAKJA habib_r_Page_152.pro
da6f64be549f07c20a107a1b61270999
690c26d7cc08bae9246f7e26aa2dd9c75e37ed9a
1215 F20101221_AAAKWM habib_r_Page_288.txt
4a4baae824fb249c860b6f69a31d2263
5e2ce395b525c5b4b76cdc0c11fec5e0c01b9b90
F20101221_AAAJZT habib_r_Page_159.tif
ff9244523750febd13e18de65a9bee2d
edc9fc9a6ea890d256d37c9837c76a32722d11f1
22946 F20101221_AAAIPO habib_r_Page_148.QC.jpg
36f9493bc77f5b9de167a86eb22e2545
fdfd2fea2a384471a0069806086b19a7768d0de4
1051981 F20101221_AAAJMH habib_r_Page_031.jp2
d8c5937add99db69f2aadaa8d0936b54
47090d66ac0b3868ff085d3b962fbd7f8d64710c
53396 F20101221_AAAKJB habib_r_Page_153.pro
76c3b471dcef9b582a37bb3a8cb1139e
80b8c6a31bf1396629e10bffc8302eda94acda1c
1878 F20101221_AAAKWN habib_r_Page_290.txt
302be7750550e79f4530258e331289b6
5ca367274f40fd2672b38903194a8f78fe35878a
53724 F20101221_AAAIPP habib_r_Page_078.pro
e36a177410c04e6322ac072c2acdf86d
3dccd4b5bbcc47f23066ca34c45095082100b7e8
1051881 F20101221_AAAJMI habib_r_Page_032.jp2
91f70ef98cdf975d0d2886e7caf31c82
c1d994368950f399b0a4e110d7c514c6ef981bf8
48383 F20101221_AAAKJC habib_r_Page_154.pro
47804c115354e8b14dd3efe0fe33f432
e6077ad5990290bd63995e7673ba9f5f145e03c1
1824 F20101221_AAAKWO habib_r_Page_292.txt
2e9ab49dbc6c3d04302321c7e92f29e7
6768cd49d4d6b4b4cfb38528cffc4a275083f1c1
F20101221_AAAJZU habib_r_Page_162.tif
2f59b2cd07773ffb39fe409faf0982c8
9232100cc50abf15f4f55ae608c496b396404fc7
2110 F20101221_AAAIPQ habib_r_Page_234.txt
e690fe5a1596471b6a72a25d8f75478f
0328fccc73f3e96080a7902a297424c8c80bfc44
F20101221_AAAJMJ habib_r_Page_033.jp2
aac87122f1495b0c7a692f559a79eedf
d99419588bc10ba1365e229b85a04c0ee57ee5e4
F20101221_AAAKWP habib_r_Page_294.txt
6fabf0eefa365bd0c66d26eda19c66b6
ec4c81334a7e2e8691e9d5b6403a73b183ba8557
F20101221_AAAJZV habib_r_Page_163.tif
7b978821f26a3f3cf45d1f8c8628b9c7
fcfa362380d6013eaf489f8fff078c5c504a1648
2388 F20101221_AAAIPR habib_r_Page_165.txt
01078b1e34e9319147ef0ab18c866d4c
16275f55a8831c808c9257913fd7559685e3a362
1051980 F20101221_AAAJMK habib_r_Page_034.jp2
fdcdf772a21a42e250daa49f4109cd97
81a3de1262306db2eb7b1798f85a2184b7e64beb
52074 F20101221_AAAKJD habib_r_Page_155.pro
68d07fec4f2b5f7c9935b41f50fca0fa
931e5f72272bfcabc5fdd9d766191a9f73489ec8
F20101221_AAAJZW habib_r_Page_164.tif
5597b03ade7250dc21c0238de0bc7b13
6b6a2dd7f6fb5e56d91eda5fb417436e85b109d0
9241 F20101221_AAAIPS habib_r_Page_160thm.jpg
6b6a5ed7a939d09a48d0a7c11eb07bff
7f77913a71a35f45cefbf3c4da3d195b19bfe26e
1051943 F20101221_AAAJML habib_r_Page_035.jp2
9df300c07887a3c45610e1d8488bcf61
6c7e3888c9f73e768a2c7f4bbce88e6630550fb6
33013 F20101221_AAAKJE habib_r_Page_157.pro
ed645258c94caaa7873249e3f356956f
551d29af5d3d712136776ba1a7cbf6e8a03a65ce
1260 F20101221_AAAKWQ habib_r_Page_296.txt
983e47d1c9fcc914c4e05786a849cebf
42ffb5f409340e4fc62b885b0992edad7ea1e97f
F20101221_AAAJZX habib_r_Page_166.tif
93804320599bfac3e8a4a50afcb0b1ea
0280bcccad6c374a5c19e5f95089fdc27ca7185b
37477 F20101221_AAAIPT habib_r_Page_218.QC.jpg
f224ea77fe66b83c0c4d2c7baad61113
72a265e0495f09383c3441503026c64d90412a49
1051955 F20101221_AAAJMM habib_r_Page_036.jp2
994a1e9aa9c4ae51f5df06b499bc3da6
a715d5f968e70ba1d9ecc83b6005755d094cc9db
43641 F20101221_AAAKJF habib_r_Page_158.pro
3be3003f5349887b15c0eed9c143c405
9feacd9bbb2f5cbb984d4eaa421efb842a886586
1434 F20101221_AAAKWR habib_r_Page_297.txt
4d1cb98c41eea08fd49788487ef08192
e94ac48482332edb5fb7abb58d5e608f2388b5f4
F20101221_AAAJZY habib_r_Page_168.tif
83a9e8fabd1ae8c1d8d294661fb6cd21
ea9aac2c575675dd02f42a035b1e207fb599498c
F20101221_AAAIPU habib_r_Page_109.tif
32a243294b54666d2a5feb4fbf430036
7e007c0ab1ac4cf47870d9d2763752a15e825a77
1051986 F20101221_AAAJMN habib_r_Page_037.jp2
81afda43d353ee55ddc31582048ad607
9aebad6166b0cea9b3547d1f22b758d46e4e1382
57660 F20101221_AAAKJG habib_r_Page_159.pro
5ad477c08f16b0b6c750ef8c557e2aae
d1f10367172f2d40078cbdda4f2f8710f2ba916d
8311 F20101221_AAALGA habib_r_Page_140thm.jpg
5e5e959d8e3a7797d0c18c23472b1976
ec30144732e6130ff32aecf02ed303d59e88b0ee
1306 F20101221_AAAKWS habib_r_Page_298.txt
0e800b8b8a6481abf08f5bcfd9dd935c
876027c55beb116462970edc6fe8f824217ca166
F20101221_AAAJZZ habib_r_Page_169.tif
936dc5cda0ebf9cfe29e0d7d1eed1aa3
b815ab3a8e2b351001097aa6ada73e05dcec4d6e
17251 F20101221_AAAIPV habib_r_Page_259.QC.jpg
191e6d93d88a3b2b6cef51f9bcdc6138
7a2070f83a3cabe247890ead92cca1cc36fce650
F20101221_AAAJMO habib_r_Page_039.jp2
eb61b85ee517c9ac71e2f63cd55dafe1
b127a98acb0e3e5a481d2864802a0fe8182233cc
58608 F20101221_AAAKJH habib_r_Page_161.pro
6ad27eac60431fe5a398722c2568cea3
e557b42bad19c3ae1c1d66c6bbe196682723f86b
27892 F20101221_AAALGB habib_r_Page_141.QC.jpg
62e499f6009a0ed76bf1b98ce15d1ac4
4e997cf6fd9a84eb8c4ac9b1f0a0f8e11f908be1
2224 F20101221_AAAKWT habib_r_Page_299.txt
fb6baa88d4a8672ec683da32a037dd33
c51834eaf8da98e2442ea1af8a7a674298decdfa
7595 F20101221_AAAIPW habib_r_Page_315thm.jpg
7466081084d68a5f636d22b5aa39d24e
1772b3568ca174c48dea506b89492a97be3faf33
F20101221_AAAJMP habib_r_Page_040.jp2
14eae2d6f89a6e25720583755dc97a93
1f721f85efc278b493046131ea2f250029c4abfb
56299 F20101221_AAAKJI habib_r_Page_163.pro
ac0844fde9378a1b3811a574cf08bb9b
fd4860a9fe42928f64d7133ee6219dcea7325aa4
7294 F20101221_AAALGC habib_r_Page_141thm.jpg
d415e3f3ab68ca75b081cc146d9c6f8a
c06e99f16dd7a284b30841b00ca3d3685466612e
1535 F20101221_AAAKWU habib_r_Page_300.txt
1447f6ccd5dd0de4c21b286f97b593a7
372d07d9d3cf5fd58db4db80168666a7959b640e
8802 F20101221_AAAIPX habib_r_Page_154thm.jpg
99ec0931b12f44ed7eda24141b0b08fa
f7b04876427214d398ba2595b8fd96396ea2ac0f
1051885 F20101221_AAAJMQ habib_r_Page_041.jp2
73d9cf6e544839eadae0f5eba1612d6e
7d6e0a71bdf080d977c13e720e906d4abd126ee5
52724 F20101221_AAAKJJ habib_r_Page_164.pro
673378bf856e2ea7d4508ba25a5d536a
21d9705c436e1ef1eccb2e84758177b22969f52e
31730 F20101221_AAALGD habib_r_Page_143.QC.jpg
c3f708920b64c3ddaccb926cc15fdecb
99f12ed0cbedea1555bd8d01c383cc254062b9be
1610 F20101221_AAAKWV habib_r_Page_301.txt
5e916ebb696345ad5309c68b0060b53b
d5fa9571e33bacf80c00d66eb569084a5086a128
1051958 F20101221_AAAIPY habib_r_Page_096.jp2
b9da0f3e36506d13c209d047eec70248
b449326bd83ded464b0a2a294de79c5bd85ac110
1051888 F20101221_AAAJMR habib_r_Page_042.jp2
8586fad8ac5c0696b2ee603305a0bceb
21fb2648a00009d2a99d042b189f7b0f4ba8bbf1
51576 F20101221_AAAKJK habib_r_Page_166.pro
2c24c6a59a71ebc1548cf931285814d9
048e77f88a7e218476ecd8ac84fd4641caad4283
8181 F20101221_AAALGE habib_r_Page_143thm.jpg
98a0638f90a4dcf7136169b290a52a5d
de2f29b080d6e8878f1e48884e789ada49da7f6b
1766 F20101221_AAAKWW habib_r_Page_303.txt
d852d3405eb18f25b9deac30f8fb20c9
78ef3405a036918840559f33602effa69a129bde
33162 F20101221_AAAIPZ habib_r_Page_126.QC.jpg
e01c94ecdf20e2ba9105c4fba8c5f81d
404cbdcd7b4ec94432ff2110bda190d07a260946
F20101221_AAAJMS habib_r_Page_043.jp2
ec4fb3b6dc1432b6e32db62eac7764a6
47cd2b4cfcbbac9157ae983f2fee130061e2ce19
51999 F20101221_AAAKJL habib_r_Page_167.pro
7891f932226dac131bdfb9c3c3615e2d
ab8810054ff9500027dc804045a1ebffc077555d
30193 F20101221_AAALGF habib_r_Page_144.QC.jpg
c07f903ddafb5bcc492047ac3245f28b
34c90380107d06b47df1e278b753a42755931c0e
1520 F20101221_AAAKWX habib_r_Page_304.txt
8fec7effd72403623d8539a230a1d085
f056cea64fcbd2e9f201cfff97c75eb3f48f3018
1051868 F20101221_AAAJMT habib_r_Page_045.jp2
6c9421b15c8bbe1c2a8ecaceb459c2a2
34f63a7ac6d77d09a809a2b7403a384d75216139
54452 F20101221_AAAKJM habib_r_Page_168.pro
13656b84f4fe64a3740a0f35de515a77
ec5c657baeba2abfdad3d9e4e98f47e60faf6d7b
7421 F20101221_AAALGG habib_r_Page_144thm.jpg
5b7ae2fe0e9ab00c77a99ea34bd346e0
6ce7c95c7b576c6daae69d1a2a2d1e59a0d449e9
1946 F20101221_AAAKWY habib_r_Page_305.txt
dafd8b358a6dcaad24fd78d7366e2ac9
447de74f8bf4b89cda1f1de0a3f3d7d995fbdd21
1051938 F20101221_AAAJMU habib_r_Page_046.jp2
e2171019cf6bd08531ceb72664675111
8d67bcba0c92b3bb8e2e2f7834b68bcd91077e3e
52675 F20101221_AAAKJN habib_r_Page_170.pro
dbb5a8cb94f93afceec9b2b7868ec909
ef120134b3e5a563a72e8bc5421402b15e6be09d
37200 F20101221_AAALGH habib_r_Page_146.QC.jpg
16bd983e80f46cfe6a0931ede17454f0
066c1ff68f0241bcda9d4b936a105c69d3d1fb77
717 F20101221_AAAKWZ habib_r_Page_306.txt
873ac85f0b39ceaf63dc8ec278c5b439
143054c7530e6888cd39ed6c6c1353f8e034c89c
1051939 F20101221_AAAJMV habib_r_Page_047.jp2
a9f415791ac11c5fca0e8cc6064d8076
dca38fd9110a9ffe6bb30dcdadc1d52d6cb93d53
58241 F20101221_AAAKJO habib_r_Page_171.pro
973adeb810bb0f04ed8f8ef331c1117b
d719dd7a48be72a0918cc251c4d6bede6be94a14
36724 F20101221_AAALGI habib_r_Page_147.QC.jpg
6e08fa4c29d82dec02758bf5e241f35a
c22140ba61ab7b07c1d063d53a6d7327f3e4182f
F20101221_AAAJMW habib_r_Page_048.jp2
b6e4b8ba0a63023b1f8f71ba4dfb1a88
dfd24aacc3ec064f1cdb89daaf1a658f1ccdfa77
52463 F20101221_AAAKJP habib_r_Page_172.pro
d5b721fdac138df5b6a94ad6ef17ef10
1c9c99a7ae85fe32670b89d970ae0b44f6495e0b
9148 F20101221_AAALGJ habib_r_Page_147thm.jpg
2e450d279582e26cf3f480163c9fac47
14bea60f77ef0e59cb66032fcc44a609399310f9
1051984 F20101221_AAAJMX habib_r_Page_049.jp2
6e7a7cd67529c27f3f8b591dffce8e06
c402c77eaebe765f098ca4d81bdeb744e40acaeb
47112 F20101221_AAAKJQ habib_r_Page_174.pro
e30c03b33a76dbcb2ec1b0020d58f892
489c1299bdbd2dee0bacb3217a8abf37e262ea90
6803 F20101221_AAALGK habib_r_Page_148thm.jpg
7ebdaac9964a843611683ff96419287f
72d439a56c260bcba95df11f009c2c94c37f4854
541072 F20101221_AAAJMY habib_r_Page_050.jp2
fb826782be9461f8b8dfe5ddcc4e7e41
c9219b7d051611f47ab9ec05b3c2ca1e52db4130
55507 F20101221_AAAKJR habib_r_Page_175.pro
6cc438c2c2d989f61c3aa9f30821c186
b3cdef01ae6555f8809bdaa9d3186e945db58973
19983 F20101221_AAALGL habib_r_Page_149.QC.jpg
c6c5eb5ba237f921f310fdd3dc343a93
e7f4d8d6cf0712d65525a1215588105c1c3ece40
1004736 F20101221_AAAJMZ habib_r_Page_051.jp2
f3ffbff2b3c50ff2ee1076dd96b1f1e5
e28112c17f7cb1c09b02faa4401c233a75528c68
55651 F20101221_AAAKJS habib_r_Page_176.pro
a5affa591154957fc3889f84402fc0a4
22ac914ecf0d3c890f034ce43575dd52df61b920
5662 F20101221_AAALGM habib_r_Page_149thm.jpg
071772da8729fe5987d8adb8b2459c1b
a13108146983cf82ddccd5f39152fc7601bc1592
59119 F20101221_AAAKJT habib_r_Page_178.pro
cd27a00984d84e761dfbebbd2090c3cf
70713fda9dfb6c7d3a05104be08b7ed8fffcf57f
19978 F20101221_AAALGN habib_r_Page_150.QC.jpg
5727d0c68592fc6ef454d29f6acd497c
17d844cb02d045043dd31044337bd0bc1999b941
38348 F20101221_AAAIVA habib_r_Page_056.QC.jpg
fc80702ab11c64a9aa0dcaca3fb97902
b7698678c6becff05862cd39258c6ef71cad8396
47968 F20101221_AAAKJU habib_r_Page_181.pro
2d2998533144bf58ba029f3ec4da1ed7
c6e47bfb4d3d0cc305e0bef80f194939c8f9b145
5729 F20101221_AAALGO habib_r_Page_150thm.jpg
45aecbbfdba2354a4db7adf79d3e1cdd
35fb8ecc201ae5d50daf1f09549b81a987238f01
11301 F20101221_AAAIVB habib_r_Page_213.QC.jpg
6f47f9744d94f82ee00268974f488d8e
c6fdeb98405701505cd2489baa31d923a3fb5a28
50501 F20101221_AAAKJV habib_r_Page_182.pro
76f5ece1d5bcfde9f9ddd4ed551c0d2f
ea9931a17fe229dcaaf0d471c35fcba285efaae5
7228 F20101221_AAALGP habib_r_Page_151thm.jpg
b6b157e556a371eb5c063931bb06d315
7e3f63d860144ccc149a1061cd2822f728cdfcf9
34863 F20101221_AAAIVC habib_r_Page_227.QC.jpg
45e7d3a53bb4e33e6e95866e25a043dd
269c0e8f314de27baa09d0de21bb4f4e2e60e066
49422 F20101221_AAAKJW habib_r_Page_183.pro
4558180182ccd42a702b33b15e97cff9
2dee228dfe47e3ad2e749743a8fcccf0db284697
34718 F20101221_AAALGQ habib_r_Page_152.QC.jpg
652388c86ba80aff98e657e1623d2a3b
65b08994533e31a13a70e484eccb88feae32f73e
5268 F20101221_AAAIVD habib_r_Page_309thm.jpg
e046d8085a8d2fc2687c68c782918303
b70d83f217a0ecc11560c95c9d782d432d41eb94
53320 F20101221_AAAKJX habib_r_Page_184.pro
a5c445011e6b5e5b6acfed2b11c1c914
d7aedc5c3048dc2140283c80c383d28ce7b9473e
34354 F20101221_AAALGR habib_r_Page_153.QC.jpg
ffb51e655661637e887fca4414961fb3
d5072004a038f8e3cf3b4c34c0eb5f2e12e1932c
F20101221_AAAIVE habib_r_Page_209.jp2
76c53e037b10cddcac16f890b19c4dfc
724a76abd0b213f1b2e49478ea39a61ce685e52e
55007 F20101221_AAAKJY habib_r_Page_187.pro
1e8d093454a51393bad697dbb10cefab
c6b1490e5caa8167c988d5ec0a47c94ac8e135b7
37664 F20101221_AAALGS habib_r_Page_154.QC.jpg
70ef80d927491902282914dfc19c6f45
bc743536fa66cb76dcafeaf0840cbe09747daf44
F20101221_AAAIVF habib_r_Page_167.tif
96b7fe7ed634ce462fd5a4788fa39d2a
8548845115aba866071a234af8d779760e85069d
52642 F20101221_AAAKJZ habib_r_Page_188.pro
14646c06840bf31f4e139be81f752e52
240549e71f1bb2a71b770088ae02a04c56ce4098
36144 F20101221_AAALGT habib_r_Page_155.QC.jpg
762768fa9932000a199706f3f397f655
670f5bd7f4defb4f52788d7ddcf17f4105d5898d
107664 F20101221_AAAIVG habib_r_Page_133.jpg
142bfcb313511bbd1ea3f22c9f29cc3a
f7a5d8a5921e4f830936cb2ce254deed12759347
9144 F20101221_AAALGU habib_r_Page_156thm.jpg
528866432205e8928c155ac42983a7ca
72dfba06ea0f02a8d50665b687f46791053689a1
3651 F20101221_AAAIVH habib_r_Page_278thm.jpg
454fd2d9c1ffa3d71e63bf9183162b01
8d3f710a36bb20f4f23319bf78c5e7c5728f089b
F20101221_AAAJSA habib_r_Page_225.jp2
dff09aed7d965bd2d17c44c16e34f708
20d2835525ebeab4137cfd12df89878da420f8b1
24769 F20101221_AAALGV habib_r_Page_157.QC.jpg
ce0ad300e8021a675868e6354bc77379
df2010c24c51f3aebc21c526a974bac85ab50bba
1034534 F20101221_AAAIVI habib_r_Page_232.jp2
125d4002e81c83fca7e15486ee5976c6
2a21e846b91e76b74ff3098f06ee8858b443aa40
1051972 F20101221_AAAJSB habib_r_Page_226.jp2
2b9e4aa71633c941fb37d9f3c7e9ab5e
7a493e89e176dd6477aa73613aab9f4c2a89be4f
28169 F20101221_AAALGW habib_r_Page_158.QC.jpg
310b4cfc95ecf7c9b2bd190eb126c6fc
efd9fd909221fa3dd077408e404e708a96dfdac3
110754 F20101221_AAAIVJ habib_r_Page_062.jpg
0818021e18fad2a7684d7c87206088f3
2e4e6874cd103532f33f6af40891ab59d87aaee8
F20101221_AAAJSC habib_r_Page_227.jp2
71678c1ebf08502fd265dcdb7f1a6bba
4bf35f0edaf5959336dadc5c48dc8ab22dc5eb4c
7766 F20101221_AAALGX habib_r_Page_158thm.jpg
72e0696f8a27cecb6d55983dd956ccc5
8ccf047852b5d424ea35e63ecf4c0caf319b4480
53700 F20101221_AAAIVK habib_r_Page_113.pro
f126b8f54ef46407b1b7fdd306ad2e45
9978867c8d6eee73d59d858d8cc946e31e9b33c8
1051969 F20101221_AAAJSD habib_r_Page_228.jp2
7f580aa3006f65e78bc580b05480a65f
595fcfa4563ee74bc04dcfe9133bd60d330454b8
39287 F20101221_AAALGY habib_r_Page_159.QC.jpg
633a459505acc29d291b0cf0ccde613a
c3735b850df809663a4873e59aa095f42d5a7a1f
8160 F20101221_AAAIVL habib_r_Page_255.pro
c5e552ae7467c63a2fb84d9e25205758
4d791d8bd4257e04764337a6aa561cd70ca91c06
1051933 F20101221_AAAJSE habib_r_Page_230.jp2
30c5615657537d7c3aef0cc9616ceaf6
24956762fe7d0285e8de1fe82bf603f869adbb01
65870 F20101221_AAAIIA habib_r_Page_103.jpg
a381b1b934cd391279eebd0fe855494e
1caf897008eec491efddf361191df3962cb18757
9138 F20101221_AAALGZ habib_r_Page_159thm.jpg
11bf086443ee51d46b1b5ce842e09b5c
acdc82bd3b08acfb22f85b93fb7c1f07d7ef0472
8012 F20101221_AAAIVM habib_r_Page_003.jpg
2b2d98c4d5a533a75f8e0b0866aa77a7
36a2bd0c88a344209eb4f1a81dc92820221e02da
1051946 F20101221_AAAJSF habib_r_Page_233.jp2
76e03a91c3409d359769c7a05b971e48
cb2fd17198fcac827f3b24c0074c763ef098d53a
4026 F20101221_AAAIIB habib_r_Page_205thm.jpg
96b79f9e1b058ca052fe054d92e38086
3228b36bd225262d00c4486eee33cae142610c48
7455 F20101221_AAAIVN habib_r_Page_006thm.jpg
2e1f3be0123a0940c1a099e94f6cbc97
b78c2a7024e03a710d4a298512357bc0aaf8e5cf
F20101221_AAAJSG habib_r_Page_237.jp2
9c92c8579897d4b96284bc85a220b33a
373339739a7db39981c2de9f86a78c3c1a354ba1
8748 F20101221_AAAIIC habib_r_Page_235thm.jpg
489c73944fbc5a745db654d9109f191a
bae886e50c03a6efbd6bd4ab83b7921131532609
2195 F20101221_AAAIVO habib_r_Page_081.txt
b0db294fbfed3a653d0e7ac8f029ff40
648125086549075a8641ac153e99fa1cc498ff43
1051961 F20101221_AAAJSH habib_r_Page_238.jp2
f267cadbcf5460f07d128a167cf8079c
26d91ac89f1e9ecff658dd816ddbae7794a8e4c4
2250 F20101221_AAAKPA habib_r_Page_030.txt
7eea2252e66d213e70184f142988b1d9
2cf9c0709b565533abd875e1afacd251fe21e150
53603 F20101221_AAAIID habib_r_Page_245.jpg
e061c72640d2cabcc876a122184014fb
48bb4b4d0977ad12712f5f9b43a6f4b6e0f902f1
2271 F20101221_AAAIVP habib_r_Page_073.txt
74eeb0aa7a4afb464532a8b3426e089d
a5de39ee313d8a4bf61f836c86e7674c132b767a
992315 F20101221_AAAJSI habib_r_Page_239.jp2
8e62b62edd5ae27688d5d55269546c12
df057887f789b7ca99a0c2a79a85fb54917e7c20
2206 F20101221_AAAKPB habib_r_Page_031.txt
4ae6e862ec157a3ea78111b2cc8fd1bc
f23e1fea50141cc3d475bc430d811ca1ae94c8a3
26818 F20101221_AAAIVQ habib_r_Page_294.pro
c8be7fa3f9d595388c5fb0f8f41f15d9
1eb4db2957f0def7a4b95dfa7724318631bcef33
1051978 F20101221_AAAJSJ habib_r_Page_241.jp2
a158cfb2c7e0a3ca01c97bb8f5a6ac7a
a843b815be9484ff7db2ce11f3f5f8b8c1502690
2181 F20101221_AAAKPC habib_r_Page_032.txt
9d2f74f1ccb7059beceba3893de6174c
f174ac9d588030afef9325f87c7aaf53622710cc
F20101221_AAAIIE habib_r_Page_052.tif
32f9ca22900a7f62cb93af8f0376673b
151e1593cde18d0924da8d29c28e567003b1d7af
1051970 F20101221_AAAJSK habib_r_Page_242.jp2
411cc5f8facdc3816449938a9153325c
a7758b7b1e9907f6aca8c3398f696330903c3f98
2138 F20101221_AAAKPD habib_r_Page_035.txt
217ff22ba62be4e4dfa901438ee48c3d
abc08cfb1668847f1c83a58b75fd222a6193f836
2231 F20101221_AAAIIF habib_r_Page_055.txt
38b159cc65d4c716495fb3c6bf870af8
cf6e40a14d8bb3ad7829042aebed5ef6628ccc9b
F20101221_AAAIVR habib_r_Page_146thm.jpg
44628a84c2ef2e26412404dfa41a7187
7ec0e8bc886fe9b3df6f459ed9d5bfb04c3c9f60
1051962 F20101221_AAAJSL habib_r_Page_243.jp2
1366fe889847f50d1e0874ca0c087604
af74530f3cc793076bc2ffc97bf0e708f1e0c821
2331 F20101221_AAAKPE habib_r_Page_037.txt
4a3a7078fddb9817059be45090c11724
8658dcfcbbcf0b4fd3469b59ab81afe4c91ec3fd
F20101221_AAAIIG habib_r_Page_103.tif
aee6a55a36044b66eb1dbd84406f49cc
a6e8491f9fef507e811aa47d59d960e96b4703a2
2219 F20101221_AAAIVS habib_r_Page_047.txt
c616894b35c4205c13a1cc960315a2ea
ea8b195e573840b9658e73a3152b93e6f3e210fa
1020037 F20101221_AAAJSM habib_r_Page_244.jp2
df61207c89ddeeeff36c668f377f88e0
39b38983d42bdf8b4be036448f1307ca25ecb5bb
2227 F20101221_AAAKPF habib_r_Page_039.txt
ca94e8748f25da5bdb8d92c14608b38c
375bf9b4a9527b8a0386b7206599153535a68b8a
48842 F20101221_AAAIIH habib_r_Page_123.pro
bc944c365a9de1343db0f1314eb7cd3c
e647355100cd868ff9c9b591e92bd0847358b5bb
109574 F20101221_AAAJFA habib_r_Page_107.jpg
9828108a0ad5a410648da9c26e46b902
e34c99226ad8d10c1cbe3fd8a845dd4be626116f
1896 F20101221_AAAIVT habib_r_Page_240.txt
6de381601f170386d4d54e280631ab30
c0d6d52fcbee1cb4372f3f9c05fda716e23105bf
2176 F20101221_AAAKPG habib_r_Page_040.txt
272e61d97f531ee808439979d85c44df
0347022d0f822facad7e7d815c14b320930ba58b
56702 F20101221_AAAIII habib_r_Page_160.pro
48c8375eb005c6aef1717234f283583a
884427723a0d2c1d2a26a4a5bee3c63b60bc5ab4
112814 F20101221_AAAJFB habib_r_Page_108.jpg
1142ceac18e6986d4c9eb29d57e69f2d
ae44c75fb099195ebed6e04bf3224d895a24768d
10938 F20101221_AAALMA habib_r_Page_255.QC.jpg
a3da17704452792473bd80f7c734d0e4
6cfc4efceeaf988c3be1e6d6dbf0a54bcb33be55
57069 F20101221_AAAIVU habib_r_Page_055.pro
39d6e906520d26b950560a73e2c9508b
183f0dfe1c7fc2af54cf75d429d0ccb3bb83a1c0
467404 F20101221_AAAJSN habib_r_Page_246.jp2
a9571d04b96e9f5c122c4361669b87c6
f4006f7e4061c5c69297d53a5f28b18f98dc5fbf
2233 F20101221_AAAKPH habib_r_Page_042.txt
1a1f9dc366106dfd55c8600a9c595ae5
b8a7fa06354bcb37efe8f9d45899bd95762b24cf
108162 F20101221_AAAIIJ habib_r_Page_229.jpg
9df0288af02e53ff7e5a99185e1fa228
ba5ea753cab2bc77e3bef4e6565abb80abf29b9a
105211 F20101221_AAAJFC habib_r_Page_109.jpg
6fa03f1dfeb3db26eaf33a43deb0e29d
8695a34b2cae7936a58317351b7033722a9e426b
3712 F20101221_AAALMB habib_r_Page_255thm.jpg
efe1d31abc2acf1808fbc8a034989750
7af3432776f7ec19aeab0562721c2a680c5a9872
8597 F20101221_AAAIVV habib_r_Page_138thm.jpg
72156f645850008b9812de191fc8253c
41acb15651efadc4502fb54f05a3763be34364c3
501267 F20101221_AAAJSO habib_r_Page_247.jp2
f96f3cdc84d4b40cd8ed04f6c81be23b
819d78dfc5c2c35ef769fb3f95b2c84d213b67f2
2260 F20101221_AAAKPI habib_r_Page_043.txt
52c430cb4edd74634eb48f292e18c51c
e4033ef0baba245bd2fd51ff66f5ea6e3fc2ed8e
1919 F20101221_AAAIIK habib_r_Page_095.txt
8957a3914f805d33ed94b2cada4ff8bc
e51001e88519cfc360e55f3ed1f5c3afc04e883c
92179 F20101221_AAAJFD habib_r_Page_110.jpg
16ab7549ea56f1326032100c511073c1
2bd0529aa6333b52f85dced233be291544166eae
9623 F20101221_AAALMC habib_r_Page_256.QC.jpg
aec4fa4586e1fd359b55ac4ac62630bd
e26101baa269da01d3af1d66db0e497ae27df44b
F20101221_AAAIVW habib_r_Page_288.tif
eab93eee4728bb258db8e2c73cac599e
da640193798ba07bbf918ac5a7c11d492c3eeac2
465461 F20101221_AAAJSP habib_r_Page_248.jp2
2fc7d89f073a0fd8e495c465cd6c2044
07c0b6ee811af9cf6e360e0039883867484e82e6
8610 F20101221_AAAIIL habib_r_Page_139thm.jpg
8f2896c9e733569e26044aef3edd7661
85b16c28273fa100ea4c2acd4bb002241ef970c2
103060 F20101221_AAAJFE habib_r_Page_111.jpg
0c62c6eb5e5dae869afcdcf7f2531a82
417262d88c4fa3101907c3a4bd609b91c84e82e0
3636 F20101221_AAALMD habib_r_Page_256thm.jpg
9a0a5507a7468d2e82ab82481fd3b9f2
392f7e032e2c885bc0b320a08e79944cd6811c77
9225 F20101221_AAAIVX habib_r_Page_034thm.jpg
86bf88d7ea9245ac6a6f14a1e3590f3f
0639ed4aa142a0be796f7e65ba6daf31b9f38ded
513867 F20101221_AAAJSQ habib_r_Page_250.jp2
bf7ae7a3761cd9729de75d966575499e
febd81476cb1981c1e5b03f882317f47dc51f470
2188 F20101221_AAAKPJ habib_r_Page_044.txt
136f26f29b24b63f4dcb4f1a671b9b64
d3a93b57adc62de10ff4b844577a0293f201be40
36010 F20101221_AAAIIM habib_r_Page_184.QC.jpg
9f9fe2af8a3abc3c065f85cb12e4d97e
052b2c3f7d0823c700cf7c51c47d568932a3200a
106074 F20101221_AAAJFF habib_r_Page_112.jpg
ecf66394561bc90a52021d4f865d0dd6
e7a89f16ece85186c4a5e5c0338f5971b38d1c35
18073 F20101221_AAALME habib_r_Page_257.QC.jpg
eaf800cfdb11a33ddf08f4192e911e35
9a19dea3eddffb82bc2b0183dc04caa22568ed81
65521 F20101221_AAAIVY habib_r_Page_317.jpg
986da471cea5f55e7396a50dfce3a7f6
6f1c38681e16fe0705fc2b70b228c071fc03ac4f
387032 F20101221_AAAJSR habib_r_Page_251.jp2
7949dceb0ce42df3b578a8d53e138641
0f0786002d1277f6f9b12426c1d5c756661d7284
2474 F20101221_AAAKPK habib_r_Page_045.txt
bd2c1d1fd3fa211b577760d6e1010638
6e94d8b8066ca164abd4003cd698773e5b848548
F20101221_AAAIIN habib_r_Page_234.jp2
f31f7b0687fa487ef2872ba4d1c887cb
33045b354e57b5960be85f5a6ff7e58b5b38a3b7
109927 F20101221_AAAJFG habib_r_Page_115.jpg
78fe482792ebe4c836e5626e67b15d37
3dec93e41981a25f6ffbfec092773d82a019fc9f
8376 F20101221_AAAIVZ habib_r_Page_220thm.jpg
8b4ea60356b15e7cdc6c57bfe280224e
34ce03aae4eeab0a2cff87d5cff01b1f769c65ac
F20101221_AAAKCA habib_r_Page_252.tif
3a898cbf783ce04ccf5be6dd8dfc8928
130893ebed759e39ec7fe8de0e8c66199e3e7cc8
434695 F20101221_AAAJSS habib_r_Page_252.jp2
bae8c6962d6d5c38dd9dba0dc4fbc3dc
bc88a10f778a9bbf3b7bb45f64ca5742fa5942bb
2153 F20101221_AAAKPL habib_r_Page_046.txt
fd6e532a653de73703e5d726dd8bb7b9
7f75ddfce14cbd56db12b508571ff210ab492371
738292 F20101221_AAAIIO habib_r_Page_157.jp2
5c332c42168bcec9507eb1a652ec9839
1336afe521a707b228f2bb507fd6f2bfa924fbae
101900 F20101221_AAAJFH habib_r_Page_116.jpg
ea4edc58c511c4dbd75eb9c603ed7e32
a9dff2a728003250108e55047c0ba5a3d593b0ea
13717 F20101221_AAALMF habib_r_Page_258.QC.jpg
d8e061ddb281ec4f71d4cae83f098d66
d474ae0035ea1cc073892e94dce6fa1d8b0e429e
F20101221_AAAKCB habib_r_Page_253.tif
9842564d304e5ae49bd5b351fb86f8d3
672a05bf0cecbae21568492066da0cc92a256d00
470998 F20101221_AAAJST habib_r_Page_253.jp2
9883579fc0d37ecee0ca2f45e2b7139a
7f1f346222ca3b96cb3048bd0f3cc0a3e7702572
2245 F20101221_AAAKPM habib_r_Page_048.txt
fa59d44fdd270df4d16f72b5bdef82a6
a424d0dc132d68a25e45010a9f6f2596b8406744
4376 F20101221_AAAIIP habib_r_Page_261thm.jpg
a1dd34e0db63cd92808afef5e894a070
4a85205e1c48473f616572dc64a91c1d0f91c8c1
94277 F20101221_AAAJFI habib_r_Page_117.jpg
0c7327df4c087ee3d205f1aa10c0c8f6
027b948abe55eb83d3db7a45259cc8c31b8a9c60
4291 F20101221_AAALMG habib_r_Page_258thm.jpg
613c59a28ab16641137ad49e9c087aa6
589c4ccba0a40dddd00f8b96b8aa115d558e8b0b
F20101221_AAAKCC habib_r_Page_254.tif
7c37915f40774739724fc9be229d4068
4e82c8f261bb6ac596ebc8d0b2ae254383880b9f
227012 F20101221_AAAJSU habib_r_Page_254.jp2
79afb92b29199e545dd4be7f6faab6ea
3622e12e91b561f903e7df73e267089b9e925fea
2305 F20101221_AAAKPN habib_r_Page_049.txt
9f428a18f200d193fc63caf3dfb7c296
426862875bface21c24929b43749bd095edf6efb
1465 F20101221_AAAIIQ habib_r_Page_295.txt
a0c7788de0a4e320f9a9d5c07f460e5e
968cb2fb3f8b30da3060125c330a9b15b535341e
95559 F20101221_AAAJFJ habib_r_Page_119.jpg
b2b5800064d7a4ba16ece53f69ef6069
b44e1db11dd3c7c51448d111a84436461ad4f5be
5717 F20101221_AAALMH habib_r_Page_259thm.jpg
2af2de3c76671f26dc2573d26aa285d1
804704f229d66c552c392d940598ed8829c1022e
F20101221_AAAKCD habib_r_Page_255.tif
b49ca1724c4661f4fedb4bcda2b429ec
778e313a04c61c242ae3bdb33e07c5f97bcdcb30
217106 F20101221_AAAJSV habib_r_Page_255.jp2
41bd7736967a36a07e9b199aca691074
41c5124def5da795a06391c9d60e03da322d0f0e
1081 F20101221_AAAKPO habib_r_Page_050.txt
0c22c9514ecfac0ee7b2ef805d706100
18661b6c286e16b538ef7fe6536431f40427c11d
1992 F20101221_AAAIIR habib_r_Page_315.txt
5490a5c7d1893e76f84e6634176e882d
769b0880b2a4a2da0f45277c80a56273474bc7ff
94268 F20101221_AAAJFK habib_r_Page_120.jpg
cec5cae23f0e3d915501d6d5a4dd962c
e2953904c75bc1b7f269f8feb003b253bdf225cc
12888 F20101221_AAALMI habib_r_Page_260.QC.jpg
a2eed054fe21794b934adecb9a0087d2
620c5f0ee5100e7101077d3e87df969958871d69
F20101221_AAAKCE habib_r_Page_256.tif
274334fef6688f7798de2c699e0780cc
0219ed07a7221ed05b54a98812bdc97fe9372b95
460880 F20101221_AAAJSW habib_r_Page_257.jp2
61dc72f7688129b3a6a80c07c68d9478
c8430a8a20d2e1832a5286219bcb93752aab9acf
2136 F20101221_AAAKPP habib_r_Page_053.txt
83a436ec28c72178cf0bb84905f60bad
c9517ea8e8d35b6cfcd71f4b5c45c633ed8981bb
1051900 F20101221_AAAIIS habib_r_Page_334.jp2
1b34278de1b510a6b08cf21850a07df9
20eae67e5bab405ff9b92ce25073f926808aa930
111092 F20101221_AAAJFL habib_r_Page_122.jpg
8e764c80839c866e3744da91896b6d44
4975df3db93edea6575c11467ced146871501767
4029 F20101221_AAALMJ habib_r_Page_260thm.jpg
e88b09822bb3e0fcc70e509133605adc
2f9286294eca0106ba25230a35dedc2b3900c817
F20101221_AAAKCF habib_r_Page_257.tif
f026ff51b54fafcfc415c3aaba244b0b
5a849cac557e0b757800a9390f8179dfac3d5c9c
366884 F20101221_AAAJSX habib_r_Page_258.jp2
b6fad9234a24ffe729b65bd68adafc2c
44205fc91d9f7615f3da0f0a05d680b9dc43edba
F20101221_AAAKPQ habib_r_Page_056.txt
e606907b62f6d4e3d9879e15534ac862
1cc99611e34272c43e63f5bc112c38cddb0c4a7d
34191 F20101221_AAAIIT habib_r_Page_192.QC.jpg
5dda21db0097bd18919e4a28d72ec231
bc7576de80e8f0cb3ea4e4c2bfdb54c6373492c0
98333 F20101221_AAAJFM habib_r_Page_123.jpg
d759349ffb104304b0d6f60d6b3c1e00
3185e8c4cbf9f74b0a7ac8470069635d273b943c
3588 F20101221_AAALMK habib_r_Page_262thm.jpg
fb0ed07f766c4d259e1abe35637d60b8
abcdd4c6d13cc979d7a4713088c81750adfd1651
F20101221_AAAKCG habib_r_Page_258.tif
ea83454b3203a2abd68616c013283c86
135ef2f3e2897af379b18dca0dd6c7915912f95e
425219 F20101221_AAAJSY habib_r_Page_259.jp2
a0bebd296b579b3fe813899e4bdcdeaf
39373ce502c90a9912279b0d2d0f57c4c58ef714
2302 F20101221_AAAKPR habib_r_Page_058.txt
cb4e6eb78def53013f4c2438f76ef7dc
4f92d001b5524d22763b5baf8d817be7b02ce97e
F20101221_AAAIIU habib_r_Page_043.tif
1e08fafe1b0dd85a2de8cdad71ccb613
a1c8ef934617322bc7d0c805760f3f3ec7f57d97
99298 F20101221_AAAJFN habib_r_Page_124.jpg
22c7e4976e355263c2fd04b9fb69e866
e68d567a1dbcd3e5d9490a2048889c0fc7863d8b
3382 F20101221_AAALML habib_r_Page_263thm.jpg
18c4819eca0af3767cb1d6eba85517ec
39eeede1c2addf8c5c10964557a4fde19adc8738
F20101221_AAAKCH habib_r_Page_259.tif
a2ca6d17dc90ca87e191156db5e2677c
34a1a10acffd49b5222061a807c72b73827c8347
298489 F20101221_AAAJSZ habib_r_Page_261.jp2
173a54f49feacb564890b3ff88c9232d
4e3b44cf121639b5d040ea632adbe870360aaa32
1728 F20101221_AAAKPS habib_r_Page_061.txt
76f322fd4ad2a0bb436bca930b4dbadd
c46f019d5f9fcc6daa59e450a4bf0b0be3918168
318667 F20101221_AAAIIV habib_r_Page_199.jp2
fedf24fa5bedd491c91d52107840ba60
75a6a4e7672a07e76f0b26c0a99bba3b77e6747e
104703 F20101221_AAAJFO habib_r_Page_125.jpg
2407c9e856bd4d32c3da95fe528302fe
acd718defa03646720782fa2ce49007c237a8b10
12783 F20101221_AAALMM habib_r_Page_264.QC.jpg
fb2de85ac0f03c5a2e5ce66200fd69e4
6c4e31eea705284c5af03a4026578e2688a5aebf
F20101221_AAAKCI habib_r_Page_260.tif
cfb3a728310d220e5a8ff44a4a3c02ae
c61b61a0c893e7553b4fa1f03a288bf78a54cbf6
2203 F20101221_AAAKPT habib_r_Page_062.txt
9c62da30a19b4755ca8c62b0b6b068cb
d579a804c76aff85058922b5b188bdc91e5ba4ac
14093 F20101221_AAAIIW habib_r_Page_249.QC.jpg
893f32cdeb11b87a29f7aaf894eccdd7
724bc2d666f43ebc7df1b0999a2a9763b9dcec65
104827 F20101221_AAAJFP habib_r_Page_126.jpg
fb95c9debdeb342c4b3c06316bd2db96
a17e2815cf634a9288cae36f2e48ffbb9fd68ec8
4328 F20101221_AAALMN habib_r_Page_264thm.jpg
8d92f58d77044d4f118f9692c2ce9461
8cce9adff8d173ec4615644368c36d09ac2e9f35
F20101221_AAAKCJ habib_r_Page_261.tif
a7fdca1b52ec3d49fd82e5032006d166
2973f4895677be2fbc51218d362e14bad366bd96
1027 F20101221_AAAKPU habib_r_Page_064.txt
51ba154c09a9292e0684e22b5a3a2541
f0ee140181fa32ecd6194beddc7d130469f00d43
F20101221_AAAIIX habib_r_Page_246.tif
0497c6fc91e9b4bc84b06aba8b3418a4
284ff319116ac2463f8c9bcf7170c323d61b330e
107625 F20101221_AAAJFQ habib_r_Page_128.jpg
b919f872a79976ad6fa33ff4976415ef
b498d8f992b3fea8e7e07ed76edea8eb1b109ded
16928 F20101221_AAALMO habib_r_Page_266.QC.jpg
ae16f8b776c140227799a57af1065703
bfd7076b3100dce7ec8095b850cce06d2bcf25b3
F20101221_AAAKCK habib_r_Page_262.tif
e3f6ee0aa78602f0e4c128d0a38e5735
8b9903a589f7aaf627818c1bc757d4a8d21036b4
2239 F20101221_AAAKPV habib_r_Page_065.txt
9d4567ec1289b338942a1cedcfb89d81
df85df55da0e617f3b0097447551cadc6104a9db
F20101221_AAAIIY habib_r_Page_203.tif
99eef3968245c3e286e19285472ac939
e11035fe6543b9817250912d68ceed37a89e16d4
127293 F20101221_AAAJFR habib_r_Page_129.jpg
3f81fa90fae9081d01088202bbfda8c3
3c157f12951f478005200253eba2fb5e337c7d15
5120 F20101221_AAALMP habib_r_Page_266thm.jpg
5fca7e3e202244cc3e8604f8b7256e9b
98b6e667bca35d093d382ce971baea1756c81c08
F20101221_AAAKCL habib_r_Page_263.tif
fc1d9b02fa6fde49ad24eb54db78433e
31f5472c0a64ec33ac79b8d6877ba31494424157
2162 F20101221_AAAKPW habib_r_Page_066.txt
d9e7c7e5ce4c43fbe1534ab2b805b43b
febe33d0aa34e450ba0f827135b7156ee512bdea
105246 F20101221_AAAIIZ habib_r_Page_182.jpg
21dbd2af9e641bf6a4f855be780833a7
f5e19f409ece4c710d85293677aaa8dbe439f973
100450 F20101221_AAAJFS habib_r_Page_130.jpg
4b282da583078f063742cfd932bfd491
640f6a5efec2d31eb47cda66b2d884cb43c3f83a
13578 F20101221_AAALMQ habib_r_Page_267.QC.jpg
7f67690289247ae91d7dab29f34f6ae5
30c675079a5e17892dcda2df726e08ba85ddbc18
F20101221_AAAKCM habib_r_Page_264.tif
c6460e4df8141f70de8783adfcf5a638
bdc6422cf1b0a925659a934116341864950a26c8
1165 F20101221_AAAKPX habib_r_Page_067.txt
eab0d88da83ccc4222d4cd8e922111e0
db4fda486394c3334c6fc24475d439e21432ef71
115220 F20101221_AAAJFT habib_r_Page_131.jpg
68ced7ef90dbf691449c77ec07163a9d
bc21152a0d044e20b199ac2547e674b467d8b0e5
14364 F20101221_AAALMR habib_r_Page_268.QC.jpg
274d308e97a253c6498c85017ca46eb3
c118a6c165c6701f6838b01287449b8b342a271a
F20101221_AAAKCN habib_r_Page_265.tif
1def757af56aebfbbc7b6f9f80132711
4c59392a7d84fe908a9dd1c00c15168d02bc437e
2118 F20101221_AAAKPY habib_r_Page_069.txt
344274ccea79f26ede6480d60577d396
96b4352f77656b7e9739da07afa789a93bc5098a
105689 F20101221_AAAJFU habib_r_Page_132.jpg
5d81e5e13ba0ed87850fbf5c670ed1dc
dbb87535160b2b6ba2a2d4360c335131ae8de3a4
4363 F20101221_AAALMS habib_r_Page_268thm.jpg
d55c44d628020c2700e6df7fe11a7c50
ad9814ee2a15360a23efc4b97720e188acb7ea97
F20101221_AAAKCO habib_r_Page_266.tif
7558d23db4fb560b8d794c3b4afb0c8a
e9563a3d6100cec7c680c67e4747489e070b49cc
2242 F20101221_AAAKPZ habib_r_Page_070.txt
89c67a71c479687cf552c4ea82f1dff4
f5420de561b72b0c179573c5039c823c69331938
106569 F20101221_AAAJFV habib_r_Page_136.jpg
3f284424d59b5527171ff3082712f17b
69d8e0626e1d39c6e8f338cd8874561e24399061
10491 F20101221_AAALMT habib_r_Page_269.QC.jpg
d62c3980d9c694bc2e136c806e2db387
4b4637f08e944e675f6daff3acc9c2c2aea6681c
F20101221_AAAKCP habib_r_Page_267.tif
02b7753e10814066a0c48ac7bc8bcad0
d4fd9ae3134ff3ffb9ca1e50259c2fa2c4dff2a8
102311 F20101221_AAAJFW habib_r_Page_138.jpg
2df2ee8ad426e5b3dbd1ddd9d533052d
cf6fa7e72ed2cb8204716325958985afb64d60ab
3582 F20101221_AAALMU habib_r_Page_269thm.jpg
bd33a2119f96a0d27b901ba93d7d6ba2
0485eb660cfe9e102d190453923761db6b6c5e06
F20101221_AAAJYA habib_r_Page_107.tif
372d9abab2ef5fa712f99a564b8a7901
2baad49cb8b6e3b1ac7e72a1eaf01ec624103ce8
F20101221_AAAKCQ habib_r_Page_268.tif
483ba9f4b1dfd241c7dac635b7a67d62
86ec7510dcb57b593f65a903336d4e7af7c432dc
105386 F20101221_AAAJFX habib_r_Page_139.jpg
2ff5cff5a4fd482d4436c17afccad245
97537117c414122f7470c6484ab13f0f80b07870
4038 F20101221_AAALMV habib_r_Page_270thm.jpg
83fe0acf92bee821cca0d6530139bcbd
2a88b16f39fe084b71bf3559fc386fe77b19f812
F20101221_AAAJYB habib_r_Page_108.tif
14f7f1a7a5d9d84c6baa02f98487f62d
09c1f52a31f003f45c251394022076fa291f9e60
F20101221_AAAKCR habib_r_Page_269.tif
95ee716d8a0b21e93cb212186c3749d8
89760dbbb116cb6a4c2176645d46710211c0d79e
111053 F20101221_AAAJFY habib_r_Page_142.jpg
a2980fe4f57a90e87f4a45eb0fa2aa9e
18c547b1bfc4366ed470bb97cfb9c50f29f7052d
9673 F20101221_AAALMW habib_r_Page_271.QC.jpg
77891d4e12ceb738b4b87ef199243814
1e39a40181bfe281e2fd899fdeeedd891bb0a4d8
F20101221_AAAJYC habib_r_Page_110.tif
5af4af435b31ecbdbbaf9c40c07cb104
ba758955baf2bcc4bc815e98a4ce7f44cffc30f7
F20101221_AAAKCS habib_r_Page_271.tif
afcd5619788909200227dc148f215648
691e6b9f7b4a8f247a30c2f92eae171d81fcc953
100192 F20101221_AAAJFZ habib_r_Page_143.jpg
98efe8486535c3f63a3a2af4300fd1f4
dfbbab5fd6a73fa7a82bed7206f83e6c4b2905f2
18335 F20101221_AAALMX habib_r_Page_272.QC.jpg
5407eaafc6f90721d05629f1eec296f5
82d07a3863a0477bf510038099e5f94bdbe5b61d
F20101221_AAAJYD habib_r_Page_111.tif
c13176150c3d138c61a7a62085d9296d
388ea7e703b546ff612001e2f2f87c631823eeea
F20101221_AAAKCT habib_r_Page_272.tif
343256beb3ea36bf0c8fb765e29c7307
6c1ac577ad44d3d7564d9af2bb582d3e8a4b0b57
4445 F20101221_AAALMY habib_r_Page_273thm.jpg
68368f87e04de1abbc240249ca3239b1
4219be4be558dc46d6bfe314453b4cf7e1c73295
F20101221_AAAJYE habib_r_Page_112.tif
d054c7eaa8b4d47c6b88d53bd9625b3c
8a4295b4567877d896f91747b0a4901eed9611bf
F20101221_AAAKCU habib_r_Page_273.tif
b2d1a65a5c10bcdfb4c333182348619f
c6d157f350fad97bb34550b611d0a42726ecdc22
55674 F20101221_AAAIOA habib_r_Page_133.pro
7947aec5f1d7ef0028807b4eb470ec23
1951ba2f95ce04572b7275f9bcb8286aa021f233
17221 F20101221_AAALMZ habib_r_Page_274.QC.jpg
3a894ea1b84059e837da669b9fdc4437
23315075a33ab58bfb699f6a93e50b2477d51b78
F20101221_AAAJYF habib_r_Page_113.tif
c9069f88aa7540385417eba1b6b45c17
2ca03f141303ed0f353362ecfab4516cefcf8f77
F20101221_AAAKCV habib_r_Page_274.tif
79933f0310d8ebe441291e46258fed0f
e49eaa5ba9cd8a4e0ab8952c804564a665d9af63
6015 F20101221_AAAIOB habib_r_Page_247thm.jpg
1c6f136d4ba2654ec95ebfa0d7cf65b4
985089157a35ece8bb82cae497bf58a1fe442d49
F20101221_AAAJYG habib_r_Page_114.tif
eff15230e6d7d83e521ce715f771aaaf
385a322116fcc1ecfdaab3922f08a490b402e8a3
F20101221_AAAKCW habib_r_Page_275.tif
fd59e9fba77483b50f2577b40ac8fc1f
3e7b45cf8da5fc6422b3990f85338d31704cb9f6
60919 F20101221_AAAIOC habib_r_Page_165.pro
ad97d9482f288cd987288daf9b790761
9dc46aad08e2cbeccb682850db9c2bd97d1ed7e8
1767 F20101221_AAAKVA habib_r_Page_239.txt
5eb1e4fde8d6f0508dcb1b6fe0ddf2fa
eef42cced7e33dd43de9cb33bc52aa990817ed2a
F20101221_AAAJYH habib_r_Page_115.tif
c506673cd3c3590e2a546d4017343b4d
4f91aac8cb2ee081cde986f1b52f39b8dba062d9
F20101221_AAAKCX habib_r_Page_277.tif
ba57cbfce4335f3cca0436774675209d
8f6fcc36ef7186a381e35b881ea3e85a3263b06c
32727 F20101221_AAAIOD habib_r_Page_093.QC.jpg
14a513b0bb16d0c778089d0eef089281
a7390652694ff083a7a25d057b5d322f419b593b
2159 F20101221_AAAKVB habib_r_Page_241.txt
f8e37a016193202cd05496f36cbb14bb
e5296362cae6af9f38becdf4f1b294850848d454
F20101221_AAAJYI habib_r_Page_116.tif
66fa3fca14b6464f5ec1f9634ba26a8b
4867e3a386300fd502f804c177319b62b025f036
F20101221_AAAKCY habib_r_Page_278.tif
843ad6b428fa76ba31ec93f38a35b5a6
a6d1fc273f5207c1e4f84202c9fa6f9dffacb5d6
1051965 F20101221_AAAIOE habib_r_Page_095.jp2
b96afc8e9f91d69a91c67a8efa0c1bcd
df1a8e3dbd0259a011adf5697db0d66109680294
2145 F20101221_AAAKVC habib_r_Page_242.txt
a6a0d35bea76d02efd7e3272e4a4c974
dff0c7f8f399c0d16f8c478d26758084f732750a
F20101221_AAAJYJ habib_r_Page_117.tif
436e6efbd6e4a08152f26f65d117bb22
715afd63f5c0a496f0462567036018ca09259cfc
F20101221_AAAKCZ habib_r_Page_279.tif
1e94e263ac5473636b4dc98d25ae8af2
4e2046e65e4f76d2149ffa24f9fc87222fc025f0
8545 F20101221_AAAIOF habib_r_Page_133thm.jpg
44e1bb1de34861222d3db8cf402f4a1f
b25bb90e5a8507f784076251cd26dd1269ee9e07
1956 F20101221_AAAKVD habib_r_Page_243.txt
42b117367031e9c2f0ca0a9174ea4153
b386e8fa5be09b4480b49eee156412215f461390
F20101221_AAAJYK habib_r_Page_118.tif
e02b00c1a855d8b084070e5d9cf5ef53
6009c4d19e4f89c4ca46a0c675d815cf755ef968
37685 F20101221_AAAIOG habib_r_Page_029.QC.jpg
c48fad349a3766a51c2285ebb179577b
35f9235debe6667c013d665bada0b4973c22b398
1855 F20101221_AAAKVE habib_r_Page_244.txt
84a34458e85ee490470c090eeeaff614
cef7e79abeb34755611d04473a4b31a6530fe465
F20101221_AAAJYL habib_r_Page_119.tif
154999c5ce6a5739f9ba55c9eab781f9
302b7de511e21f969e88f146363a3b216b88bceb
49469 F20101221_AAAIOH habib_r_Page_111.pro
3fd3431a71c5012637643eb86bccaeaa
9a0336172236e5cfbde5b3fa7de2b6f6ef5112d7
112488 F20101221_AAAJLA habib_r_Page_321.jpg
e6c628700f27c87b501eefa9c16ddab7
61811daa3714ff7b31a2a78be793f8fa2030b341
1353 F20101221_AAAKVF habib_r_Page_245.txt
127321bd01a0571dad9d1b5c6207e35a
9d381b71c85c7bc1d3fff0d77151c6e4630e5a8c
F20101221_AAAJYM habib_r_Page_120.tif
de2ccf08f4a13e243a4eef7884546d70
b4787efdd7dab86803371bce444d2a68655c3510
36802 F20101221_AAAIOI habib_r_Page_074.QC.jpg
7facc22b7a69db145fccc9d027351138
b871a0a574b2894e52a3daa38fcf2f9bdcb8b29c
129846 F20101221_AAAJLB habib_r_Page_322.jpg
ea2f89880051be93e61028d2579419d5
fac9375fdaf81ad2a3d7f987e8a29f44a205f22d
1523 F20101221_AAAKVG habib_r_Page_246.txt
aa000d58d276a7af0b488677bd4c2cbf
28f6b8dae605e1397e9dc417611b7d791314c820
F20101221_AAAJYN habib_r_Page_121.tif
dd00d8ac4721b7a9147e39d09055f953
671bf78db142f0ed33e307dc874d397d67f538fd
105683 F20101221_AAAIOJ habib_r_Page_113.jpg
20043b4e1f454c717950483a83752c7e
56a38ca0c99bf837f63f9f4721dcebe5f9429796
125874 F20101221_AAAJLC habib_r_Page_324.jpg
c971524c97640490a8fedc0a4fa2f214
2117b2b9ac150257311b579590cb8bd705a310fc
1779 F20101221_AAAKVH habib_r_Page_247.txt
590b595c3f53e5f2efd73c2721512372
8a99c10f5927d510ddf576c18fe3018952adf308
F20101221_AAAJYO habib_r_Page_122.tif
e4d31d6b15337e33fddc13a2fa88056d
43308d4765dcc019323a2e5309ca46a98fc5e5f6
119400 F20101221_AAAJLD habib_r_Page_325.jpg
896650b0b201dc39e4e7a07516141317
8cf1617dcd5e6ac9c86343c68fa74a3028af06e3
1450 F20101221_AAAKVI habib_r_Page_248.txt
42159d8273e9d8558602436f88761b97
274c645a4b5165e8a612d0781d937520a312d25c
F20101221_AAAJYP habib_r_Page_123.tif
07cd4dbc1283acadc0f40eaca18c6610
62513b3ba3048bc2858ef6993515fa6fba9fe418
1051977 F20101221_AAAIOK habib_r_Page_007.jp2
114dbd88d94f85d64a6848ec46db3f1d
95ba00ef8314e83a3724690f9b2c8cc1e0c2f291
121364 F20101221_AAAJLE habib_r_Page_326.jpg
5dddcb6617a7a89e666c4847885acfad
45233ff08bf9db7726cb90dbedc819eaa452718f
1107 F20101221_AAAKVJ habib_r_Page_249.txt
70bdb4f3ef2d91b0acdcd06043d26dcd
2ed3b899e36d0f8ea749453e33467d2e9ffc98e5
F20101221_AAAJYQ habib_r_Page_124.tif
88de70bc80aa12636788f17922a9c5b4
7af3d5e0680ae995bb41a27fe9b7de5cb8188b8f
8162 F20101221_AAAIOL habib_r_Page_093thm.jpg
e3c0fd781c1967ecb2d9d622899f4f01
42552b2552b12dbf339d08ea51f5eb42de2d09a8
123266 F20101221_AAAJLF habib_r_Page_328.jpg
4a6dcb5f083f2ffbb9043a1751a490db
e24503a829c82e68a6510f517591327f26ba0b44
1426 F20101221_AAAKVK habib_r_Page_250.txt
425a46ba1005e953c48b2d0204e2d6cd
6d5a0ddf4c536e2c0bba325a61234fbe4e144eca
F20101221_AAAJYR habib_r_Page_125.tif
3ed99b5d423a5a47f409279c1534542c
224a6971a97f2a0d1fc022aa36539153517fdbfe
57233 F20101221_AAAIOM habib_r_Page_326.pro
f51f1faa3e41eccbe43a5767c37acb24
e30f264807fa5a7ade1578450abb148213d07ebb
2922 F20101221_AAAKVL habib_r_Page_251.txt
6f497e6376dafc552ceaa35f143c3d32
945063e4cff783a955708d043521dadd62c176f0
F20101221_AAAJYS habib_r_Page_126.tif
81cbaec666de2e1580c74eccd1b825b5
6afb581b3ae3a2147d1ef74f5076c0b67880acd3
506 F20101221_AAAION habib_r_Page_001.txt
21edfa4290cc2ea694ef99773ad66d7c
c410f5fcd7f2e949da7bf92791817e0ab4efc2b2
122959 F20101221_AAAJLG habib_r_Page_329.jpg
fa060cdf7a360d1812eb2015133235b7
718efdf5aedbceb8570b37c25c05c9f8eaef9da2
49402 F20101221_AAAKIA habib_r_Page_116.pro
b7e0382abc56231ea380dc904bf30f66
0b8e26cc957e7f79969c90362439783560016164
2606 F20101221_AAAKVM habib_r_Page_252.txt
4601d1fcc8859e9fc6189831f121a709
25c3f319ea7db15933696fa95fa2498490cd2119
55168 F20101221_AAAIOO habib_r_Page_146.pro
98cf4beef0da3d022f0333d354faa119
a805a44a54727e024df8357e933a18e8fcfebd22
125506 F20101221_AAAJLH habib_r_Page_330.jpg
11e65daffb6b8d09e47feed5d0a20817
18c0024ca4630c46a7fb1eb7f56537649edbe18e
45447 F20101221_AAAKIB habib_r_Page_117.pro
1dcff8fa9af1a9663239894d6173e0d5
b652871d3d55fc04d5e5e18d654d10364c8a518a
1522 F20101221_AAAKVN habib_r_Page_253.txt
1a3e42767f7fb49a3b32aebdd7b02bf8
5306a1093fa261ba47c9d38deb03fb72ee49a34e
F20101221_AAAJYT habib_r_Page_127.tif
adaf83131e08034bd856b342df631539
2485073b8605006cfe1137bbf7b35cb87c83add8
56478 F20101221_AAAIOP habib_r_Page_031.pro
61f0b807a314b1b4cec8b935711e85b9
b1073f99828219b49a3e509542fb5794fc9a050a
104853 F20101221_AAAJLI habib_r_Page_332.jpg
0b44d94ea762b10b4fa3e3d69e0549d2
ceabd62e68af300d6b04763982c9797b283c5f71
542 F20101221_AAAKVO habib_r_Page_254.txt
3710175ae915527e935edab4ea2b72ee
25ba52fe8e4f2b7ee21fc07d86554a826be957b1
F20101221_AAAJYU habib_r_Page_128.tif
839c5090b078653a6bfd1b6dda20f5b2
ea4e502b81863a2342187b5cadae7ff60844da53
31610 F20101221_AAAIOQ habib_r_Page_118.QC.jpg
2363b1acbb6dc2f9b98389542245ca82
d644b5952bd98b25cd8334129531512a5bec9e56
77667 F20101221_AAAJLJ habib_r_Page_335.jpg
cff922b9a2211ce0e72ff1580a2ee8c4
e4bacf4969ccb331306920a81892488018f7a918
46174 F20101221_AAAKIC habib_r_Page_119.pro
53e1e9dc47ee749f532b9d2a1d875da0
4f220b21f7e781b3f7f21f358186c19f32a87b64
F20101221_AAAJYV habib_r_Page_129.tif
ffdaf4e9edf54ab7dbe0bda466940281
05cd7696514b78daf06bd1dffe375944d13bb192
51232 F20101221_AAAIOR habib_r_Page_212.pro
00acd79b1cb1727f95d1bda42af1080d
424ef087a9619374a797cda2e56643aed24f9965
23042 F20101221_AAAJLK habib_r_Page_002.jp2
c26c45ed30d2df572f78195184dc85db
676aafd7fc1b95173193e9da1c53c7a628a63314
45875 F20101221_AAAKID habib_r_Page_120.pro
1734b1f9d825956fc729a21171657b03
5a64b0ab54a14f4a81e3038299b526007e4c3ab5
433 F20101221_AAAKVP habib_r_Page_256.txt
f7f22d9a477a8f1abb5286305dca2026
6e2fc71ab7ee2e7a76a71cef431a651f843985b9
F20101221_AAAJYW habib_r_Page_130.tif
61b7856d96ab23f9b98e8ee6a890722b
577ff1265d516d8ce1bbc69b35e2b136aa89f417
F20101221_AAAIOS habib_r_Page_186.tif
39929f5c211add7052b7df1e2de9b598
04320cc63b8b084c00af143c8a71db9407159847
1051919 F20101221_AAAJLL habib_r_Page_004.jp2
d4b2e610acaf3537fb64582ae3b4c92e
50ca34d031779c0b9de9ff26279e99ef04f6d7d4
39448 F20101221_AAAKIE habib_r_Page_121.pro
4bc97a635a676db17662055db5ab16a0
3f3c37b50b195062240639e6863acb269ac7f48f
1826 F20101221_AAAKVQ habib_r_Page_257.txt
4cf603a575164f91116684ef709e5f5c
bc68e2f4d86e33a38b3e83dc6acc3c5b3f693d22
F20101221_AAAJYX habib_r_Page_131.tif
86996a3bec73fb1cb69eb91774fc0b7e
1f03c441e0d530ae4441b5478ae992fc330bdc38
F20101221_AAAIOT habib_r_Page_142.tif
dad999a4e3bb996c00edaf956f0422ed
3728908ecea3a113ecb76eb3c89c8684bde165a1
819235 F20101221_AAAJLM habib_r_Page_005.jp2
67149749b480c6a6b77f8b90d3f9e605
0f36c8cfcdd996b6e9d41d1a748f9c1c68ab2406
56241 F20101221_AAAKIF habib_r_Page_122.pro
4850a9e8e7daee930007331a30705266
b6bd7ec12c3345c89c5f3947a713ddd5abb3f878
2772 F20101221_AAAKVR habib_r_Page_258.txt
08a862e9b8bfd3241ecc21514b76f428
9d21713530693a96e66314f6db9874a32612404a
F20101221_AAAJYY habib_r_Page_132.tif
3d281baeedaf2d068caf24eecc71accf
ac03740170335062c06676427b4fd802e0c5cb7c
9032 F20101221_AAAIOU habib_r_Page_321thm.jpg
60a2d7a7be92b666550f09ba4b292a65
8155120444d40435fd2c844f3b3245150ed9f8ab
F20101221_AAAJLN habib_r_Page_006.jp2
0c5e53dda2e3c1509d85afe16e6746f8
9090d981306cc775ffa2669d79bcd6eda7cd71ae
49271 F20101221_AAAKIG habib_r_Page_124.pro
58e0e93c56a6773310f0956b585aa7a6
a31bbfa757f821c43e67922cf7664152643090b8
33334 F20101221_AAALFA habib_r_Page_124.QC.jpg
67762953016375b278141f7560110067
be9673c1c8aaabc033269a790667dfb984eed4c5
1932 F20101221_AAAKVS habib_r_Page_259.txt
cd9c38eff040b0ccf0944889b8871585
6c4273efa229c8b7036809db88d3f1f80f38cf2a
F20101221_AAAJYZ habib_r_Page_133.tif
3b65a077a297cb2f792e3e2335768985
04347ea827ecc1b99304ed1bd10af047bbd20b23
1494 F20101221_AAAIOV habib_r_Page_307.txt
053398eba6653c1d145d40cfa9189c1b
d76bf754d9839242eb57acd4760fd534810af9c4
1051948 F20101221_AAAJLO habib_r_Page_008.jp2
6367c479461b54d917b3b4b22098de18
6b2e0bf93f7b63cca500e08cb28de90afa6a5e5e
49732 F20101221_AAAKIH habib_r_Page_125.pro
118faf27dd150618011d18d064ad2167
1237c24cf460a7f02ee9f3390e7b5ea1e550978f
8325 F20101221_AAALFB habib_r_Page_124thm.jpg
7b4d951e594d0c93f0927aa0a23075d6
63a661cea273d80d7594f63c941f1ec3ce197e59
2547 F20101221_AAAKVT habib_r_Page_260.txt
f7682ebb05b626aad928caf5a1177b36
fbbb9f80b9fc89b06f24fe3d93851a99965c303b
113608 F20101221_AAAIOW habib_r_Page_334.jpg
c7570fb6227f521ba25681b22fbb8844
39b967549def315f6caa06f82ad15b3b6cca7599
F20101221_AAAJLP habib_r_Page_010.jp2
cd8e9b852e6c9d5bbfe4d776cd674918
e88cc4248e1dbdd0254ccabadf7cc686678e9cf9
51727 F20101221_AAAKII habib_r_Page_126.pro
a6add2490c4f30a6ad51c5914ecd99dc
3f8d79e7d3ea3908ebb988c7cf54e15b1feeb955
34257 F20101221_AAALFC habib_r_Page_125.QC.jpg
fcfe54dd41d641625ce382c71ac8e7b3
054dded899628a057642420a669bd83313db64be
732 F20101221_AAAKVU habib_r_Page_261.txt
02eeb04fb85f3967ff41298004ddbd8d
e1cadfb24948e98fa556c59d960b986c5c195dfe
106239 F20101221_AAAIOX habib_r_Page_114.jpg
1145030a4e6268881463c12004e77440
3a2039b9779f6c543d6e473bc3ac8dc329b4c4fe
163138 F20101221_AAAJLQ habib_r_Page_011.jp2
3fbfa1088307dab5ac8e8cef9f3fb5d7
4ff59d3a4ef8034db2482b4312b9f06f1603092e
53385 F20101221_AAAKIJ habib_r_Page_128.pro
2b6d2264eb4f86d2977fad9ff6ac76c5
06953d1ce2ce98373aebc51ad31748322dfcc276
8256 F20101221_AAALFD habib_r_Page_125thm.jpg
bcf04f6d84e4251c320eefe3ab820c4d
71fa43461e5c92887ee539507eaa79a6ac6128c6
512 F20101221_AAAKVV habib_r_Page_262.txt
6e2179230d305b94a9f4884ba11e59e0
f3af7f6ee02c7a892df5cbd7d94aea94fdfebedc
107098 F20101221_AAAIOY habib_r_Page_190.jpg
23dc22eebcec46305a1235e7f75e2bc6
3687482b021b439f2981a3fc68fa1d23d79b77fc
1051985 F20101221_AAAJLR habib_r_Page_012.jp2
d6bdee5d08e39ad3fef3de13cc7a8fc6
03e8f57c037881df22a992496f01149a10631071
73461 F20101221_AAAKIK habib_r_Page_129.pro
120bb6ceb6717c76d88d1d6b650e7c8c
1ed2fd45f143f821b02d1e0cfd3fb6535d415a63
8923 F20101221_AAALFE habib_r_Page_126thm.jpg
c89d98327e493d310fb12119ef1357b8
a4a0d6f973e67b1e3c8324b660ec7eb76e8e2cdc
1049 F20101221_AAAKVW habib_r_Page_264.txt
10745464692fb4bcbef759c86ac042ab
3d78120d5b96357401cf61d8d84590fac610c753
47768 F20101221_AAAIOZ habib_r_Page_266.jpg
0fe8b99aeae3bd5cb94343a823ee9ec0
847bb369b8004b626a48aa798f1ed44d274e2c83
F20101221_AAAJLS habib_r_Page_013.jp2
68f619a0706066c74724bc58dea83f18
9d5f65c3336e03df9f77ed5edd3a3b06b0eb48e7
34309 F20101221_AAAKIL habib_r_Page_134.pro
9d0fbf2918853f541df256c59c0d3102
48f4fb9d121293a5b7e82def3e4ec2ad6f9427ce
36109 F20101221_AAALFF habib_r_Page_127.QC.jpg
2c179fc6430978dda4b0a94f1c61d242
fe43cc9f629aba82f0a9e01b599915406d235676
2552 F20101221_AAAKVX habib_r_Page_266.txt
3f0ce4f4350a8c91e6daef51a9b469aa
8d4e6068a4c6d7b54ec9988b2c57d8fe6a23fe6d
F20101221_AAAJLT habib_r_Page_014.jp2
aba8427b8ab104192c07ae1b69937513
56cd8909074d4bf0bb4c474f237e7f311153176c
51346 F20101221_AAAKIM habib_r_Page_135.pro
3feb1e7e761aa24434b7d9dbe4fc8196
5f578eafa2afd7b6661e3745257520a0d0fbe6a4
8118 F20101221_AAALFG habib_r_Page_127thm.jpg
f20fa886466d76972b8c698c8599de54
1cf02a29fe7d4d20251dbf50e362acb6dea177a5
1407 F20101221_AAAKVY habib_r_Page_268.txt
5e56924664dcc820f4411816b4cfe8f8
0fe55f914912e21171bd70fe1298235ab88b0d28
620291 F20101221_AAAJLU habib_r_Page_015.jp2
d254db494496954e8cdeba623101acda
df74d40fc644db30ce81a9f52040dabfdbeb27b7
53129 F20101221_AAAKIN habib_r_Page_136.pro
18b8e22f348c5c341f0d865dddc42038
bc3be74a4a65002e10168bcb48d764c209ed7b04
34879 F20101221_AAALFH habib_r_Page_128.QC.jpg
0e9efead9c3eb40b0434b30913b5a089
14aeac66a28149f0b5f10d349380581abde3f469
683 F20101221_AAAKVZ habib_r_Page_269.txt
df42f710c02a56ff24bbeeb7a0176162
35645f9be14edec444285b7ec0e647f90ebf19b9
F20101221_AAAJLV habib_r_Page_016.jp2
c23209469a4f3a5c41e64433fb680c66
f245f644ede60c355ad2d3f34fdcc4f5b1e1f5e8
46667 F20101221_AAAKIO habib_r_Page_137.pro
67657f976c1e4bcae1d27bd3ab4d9a90
9dbdac5b9efcda21ef02f9a07d0a7d554d1bf1c8
8981 F20101221_AAALFI habib_r_Page_128thm.jpg
2968730d96cfaf049d406823263f1594
5ed5d9d0e4fa5211a993a5fbb49d3d096b932105
1050852 F20101221_AAAJLW habib_r_Page_018.jp2
932557503e11e5c07e50309f04c55d69
6a28283dcb91acce590843660da4e5d043c8fe70
46938 F20101221_AAAKIP habib_r_Page_138.pro
a083bd0caf673c2adfa73b3472d98c73
7174246e88cdb953a049fa1f63c4ba60cc12f373
37268 F20101221_AAALFJ habib_r_Page_129.QC.jpg
4dbec6e95113611436c08681e50b7682
0bedd417f9e4a42d22f66d0b4bab5fe243f731e3
F20101221_AAAJLX habib_r_Page_019.jp2
b96e2ff015a70029e1c5ca3ba3519af2
0658302e52652c867d120954ea1197bf3ea84dda
52985 F20101221_AAAKIQ habib_r_Page_139.pro
79dd3bd11101ac1655c6f3a8afdb48b8
083906ca4938726c6c43c07ef533b41466a59ed9
30323 F20101221_AAALFK habib_r_Page_130.QC.jpg
cdac4d91335fc0bdbec810dcbef58e1c
5b7f1edc17a77fd75ca17f0d0ee270ff722ff786
F20101221_AAAJLY habib_r_Page_022.jp2
4583885f6c9cc2585f7c9f09535fd5ca
85987386ca63f2b1c0f375326aebf3051ec1860e
39814 F20101221_AAAKIR habib_r_Page_141.pro
7c8ec8cc92d3e9611654a73986739068
ab7f24c16573788122b04b7edad6973b8dba3ab9
8562 F20101221_AAALFL habib_r_Page_130thm.jpg
8dcc19fb7a3f9516b161d97da6da495c
9dd4cffdc63814110470a93d56dad4dddb615f04
1051956 F20101221_AAAJLZ habib_r_Page_023.jp2
429c3d94949f8e045cc7dc254c3cc3a2
2c465c5aac2857e59b329d3309e03460df7de6b7
54358 F20101221_AAAKIS habib_r_Page_142.pro
137c949e37b2319d8432da5a4308aa6b
312765776d0141a33d2eb286d023acb3505c0fd8
41969 F20101221_AAALFM habib_r_Page_131.QC.jpg
1b0ef1044ca699caf484efc17d5d17c7
5e7057d6966a96bdf76537990a6cba80e53b5c69
49250 F20101221_AAAKIT habib_r_Page_143.pro
ef370522f37cc6e41a287677516a760f
82d20277cf5c2704e39cd8e57bc2522e30ced083
9080 F20101221_AAALFN habib_r_Page_131thm.jpg
84b6e01f354d10e613892c593551a992
222ef2708ae9c35260b73adae6728974274fc03b
F20101221_AAAIUA habib_r_Page_317.tif
1ae2282f4314ff9d8533b80f9adeb3e8
39ce99e847bc51c1e66d558836cc5e3691c8f8cc
44086 F20101221_AAAKIU habib_r_Page_144.pro
bd7c902ebc8f51ea79491fdfa72992bb
514679e70a2c992016785daaac94e8ee3e25ebca
34651 F20101221_AAALFO habib_r_Page_132.QC.jpg
d13fbf95dd3304601e17c5e97905c1bd
4d94bf51501dc07ec38ac280b10005de5750af06
55910 F20101221_AAAIUB habib_r_Page_149.jpg
750d56ce21b95f0592a700a8059fa8de
18752d990da0c3f5a1f58a2f948c31f5564768ab
50862 F20101221_AAAKIV habib_r_Page_145.pro
d8e2a97edbd66fa4649ff1118957fc73
46373ad154b75fcf567110f30fafe176f444a971
34127 F20101221_AAALFP habib_r_Page_133.QC.jpg
02f00acdafa4c2014d1fabe8924e88bf
e8ed3c772fe8c7dd51c309380daa5a8c0ba74557
109928 F20101221_AAAIUC habib_r_Page_068.jpg
044f103c29c17c4649cf5fa096aa77a8
004d32b2125bafe34e72a625483d25fbe607bfff
57103 F20101221_AAAKIW habib_r_Page_147.pro
e01079f875b7883202cc17383f867ba8
41d03367310d3c5feaeb3d9824b9f7594acbc824
17814 F20101221_AAALFQ habib_r_Page_134.QC.jpg
e574ac0b84d4e69c2bf32e1219af9640
24ae70eab03880d98d556914a80404a14b8e7ef8
F20101221_AAAIUD habib_r_Page_173.tif
69c9155bdbaf9db97c9107108aa89bed
6fb6a0f27ed4000d549cde461a49bca976b10912
23549 F20101221_AAAKIX habib_r_Page_149.pro
d1da3ad42612bc9fe8c1008a0e7bc1ef
849b39e0b3583b06062463f9b6d885736118f34d
5061 F20101221_AAALFR habib_r_Page_134thm.jpg
298a2f5c8ac3da88266350326863c379
2c5d3e9c54b5d4b34ac7dd37a979950a99a5915c
5622 F20101221_AAAIUE habib_r_Page_274thm.jpg
25742e7838991bd109b8c4fbc929eb6b
28b10ebee6946f2fc1e524985df3e0aeda1671d4
23119 F20101221_AAAKIY habib_r_Page_150.pro
6b9c6f6f46d1ca6cbd783524015333fd
884346ba4106d20dc575f859d60b6a139147be4f
31323 F20101221_AAALFS habib_r_Page_135.QC.jpg
02610925764133a8b4e82aa8f1103917
70f773fc033d318fbc069e4bae63da811b3cd736
680664 F20101221_AAAIUF habib_r_Page_202.jp2
f47696d567b838c3bec4b4db6e3f55e1
25d0e57692646108e576a99b116280c286fe2556
32446 F20101221_AAAKIZ habib_r_Page_151.pro
db80e18bbc687556513d71ade749aa8e
3d545b2a47e0e3e7eea73447bdab8ad90abbe58d
8346 F20101221_AAALFT habib_r_Page_135thm.jpg
e7007d0836c6b8c339de11c917ecbbf5
d001e9529c09063a087dfd35fc08d2a7206f4324
F20101221_AAAIUG habib_r_Page_196.tif
e3b487fac66b42fa241eb366a625a9f9
92caf78637f50f0054bbeed424e30ed071abf3bf
34713 F20101221_AAALFU habib_r_Page_136.QC.jpg
5268e2c76c34146f8782f4b60c42648d
1834a1346cd6d71c4e2119f2fb081f1c8e76d52d
1942 F20101221_AAAIUH habib_r_Page_099.txt
30c345c3cfe5b5aad2365158c6bfd17f
5c98cfe76fdf20e7067144d8ec1a2f1ccbbd0c31
F20101221_AAAJRA habib_r_Page_189.jp2
e9d233fac32467a693d76b691d94bfcc
30a247b6f317d16a8e75bab6b9e9fa4d349591f8
8576 F20101221_AAALFV habib_r_Page_136thm.jpg
11c942308cb4f3c1fec4c77653c7039f
06b3d62b736938cc7624d9acaf26b0f2d63616f5
8480 F20101221_AAAIUI habib_r_Page_021thm.jpg
0de1be2260d2c01d2afa196a55dddcd7
ada923b08d681d4e808863ee1de8fb7a08f1751b
1051954 F20101221_AAAJRB habib_r_Page_190.jp2
74b122507fdebdb1f86f44e894e87e59
5a1abe22d86059c6fdb3a52e0a653671f3f6df0f
31623 F20101221_AAALFW habib_r_Page_137.QC.jpg
ce522a61ab95e0a450c90ae7c1da2f38
c947bcc3c8ca45126425a5770a5c616790db376f
9524 F20101221_AAAIUJ habib_r_Page_330thm.jpg
5d9c0f7f36fefde1a65ce9a7a5fd0fe9
fe87120967419d8471b206e85d4754c18d68becd
475568 F20101221_AAAJRC habib_r_Page_193.jp2
c3fbf718c921a4d43f3fcf593f738eb6
a058949fb09ec1f2571469d50e0afd7315103505
7891 F20101221_AAALFX habib_r_Page_137thm.jpg
a7fb15211ed0497618fb92200c79e95f
42f7146a9d1e86e38bc388e41f8135e21ceac6ca
35395 F20101221_AAAIUK habib_r_Page_222.QC.jpg
21df9cbe7a30ad5b9442d5926ea4940e
5980a3797df422f3bb3960fb5b371fba63a92bac
684740 F20101221_AAAJRD habib_r_Page_195.jp2
469a4630b38f9d0337bf25777622dda9
74ef7bcdeacf1e64f710d863ef171c9c504e3467
33789 F20101221_AAALFY habib_r_Page_139.QC.jpg
d64bb493af2e895ff0ad377efe3f3d21
b431bce9c4a717c5d8ec716eba1dc93861b0ad55
F20101221_AAAIUL habib_r_Page_022.tif
0a3c40c28d0b712cb3bb0fd9b3e57366
0fe08902bab176f107b7502acdfb973d1f312e3c
460774 F20101221_AAAJRE habib_r_Page_196.jp2
7c0c02fba2d4d2d1bd82de38885e7287
1b0ab7ec8222183a7e644117cf1754d31b018d09
15366 F20101221_AAAIHA habib_r_Page_313.QC.jpg
33b14d8ff7c1f15942389bede1ffa674
df734a1d98770a7c3902cf2475f34368b085b813
34275 F20101221_AAALFZ habib_r_Page_140.QC.jpg
5df780d895f296eb5bcf8d4d717de6f2
e06f1ddb2d080c73bf9f54aa11e906b25bcd5c96
55230 F20101221_AAAIUM habib_r_Page_040.pro
ece87c9bbec0366ba9ec02eee4bdb46b
4958bac4deb073d900a6f348e989bea9b87fb846
490386 F20101221_AAAJRF habib_r_Page_197.jp2
f0c9d8ec801a4c7d5d7e5725df18856c
eee6677e55a569e016e3a6a61182aaf4aadfefcf
F20101221_AAAIHB habib_r_Page_078.tif
bc25f66c930c53618a18caca67f06adc
abdf2fcac29c90eb47855639f6fb9fa9222869b9
122818 F20101221_AAAIUN habib_r_Page_165.jpg
81d690d7a93ff6ab1663886bde82d8d7
6b687ab25bbc219d095ac66cde67aee25760e192
1051975 F20101221_AAAJRG habib_r_Page_201.jp2
c4a9c0244730d9fbf5a9e827d2554db7
31ad1472d52e8f433e608bcdc37e2c7cb26bde40
8425 F20101221_AAAIHC habib_r_Page_219thm.jpg
30be7f332e1a7cfba7628e39ba49cd0a
869bf2362647177cc095409855b4d10790e8c568
35589 F20101221_AAAIUO habib_r_Page_223.pro
20fc1b8263223ad007213cf5d97c858c
4a4453eb1953dbc13e09c488d64fafbbf66da2f9
F20101221_AAAJRH habib_r_Page_204.jp2
39701535650e26137954c52d54cf1f6d
5e290e12fd8819e8ccdd51d4569ee18ff5b3be21
60564 F20101221_AAAKOA habib_r_Page_330.pro
3ca33682bf113153941102d8f2f71829
0279e5123ad6728bf4d877bafad3f4be50495972
F20101221_AAAIUP habib_r_Page_149.tif
523ee798bcdb3c6b34a2aa727d288abf
cc021a64536a55d7ffae99c2f4335988f0af2e00
206992 F20101221_AAAJRI habib_r_Page_205.jp2
77504327589b34759f8dc721834c72b5
4ae8ca9e575ab2ff3ddf439bc74460d35dcc6a07
50666 F20101221_AAAKOB habib_r_Page_332.pro
c6efc423a26adb5477b0c5f97b012b05
cc23db2a9a1ac4070f0ff33d6f72499579561fbd
2217 F20101221_AAAIHD habib_r_Page_282.txt
f433f626c1c96f0fc9b75f1fa7d33c08
81fde57985b7699411b60371038f5099430b6bae
F20101221_AAAJRJ habib_r_Page_206.jp2
0f779e2bd9d78063dc1982548d4d5526
c10eea39b78fd25993635a5b94bf153a67d0358d
55924 F20101221_AAAKOC habib_r_Page_333.pro
4fe3c5784bc2dbd2e14cd078220f0002
16daad46bd855e7c589a4d9b8abfebd6723215e9
43153 F20101221_AAAIHE habib_r_Page_180.pro
b8627167f46804b4cc6b3298d713d31e
3d1563d59c8f2ed533b1d50ac3654afe05eefece
1051847 F20101221_AAAIUQ habib_r_Page_164.jp2
98370b2e6f0e5ead3242cd02c004a22b
27b8ec6622c30e92efa3c8694b9261bb631f1a4a
F20101221_AAAJRK habib_r_Page_207.jp2
737b128292be56c6aa5e05ee974e2f7a
8446399f66d0996fbd9b4bf307d99c1a6c4e0a25
55831 F20101221_AAAKOD habib_r_Page_334.pro
b0d988c2e89badcecd2cfd45481f8868
2bef6790476739f084b6625b71b64e4c2bdc6215
48052 F20101221_AAAIHF habib_r_Page_100.pro
b500405226977c98d24c9de645dd2031
276208497b5df660034c31d56fbf1efcc19a3936
1816 F20101221_AAAIUR habib_r_Page_051.txt
73c6c78cfe7ce8898c22f5c750dd03c2
3cde9715ce41a7319d93c1a7c2e2b54c91340495
1022619 F20101221_AAAJRL habib_r_Page_208.jp2
f3fca21941259700fd9916ddcc420c02
0ded5cd2b57eff7196b54f138821848b0bdb2b53
36752 F20101221_AAAKOE habib_r_Page_335.pro
96d73a93fd8aaa848307d51ed547b69f
135d0cecaae8c11f6d693759900d94c0d936a027
2221 F20101221_AAAIHG habib_r_Page_033.txt
ff82b8b7d8d687b49b462c23ae093825
193446489f3a01ad87085d3ec1b839c43cd7895c
115573 F20101221_AAAJEA habib_r_Page_070.jpg
7a52cf25dd6b42adae9010f16a211d77
5a43e4bf3635694fd4523f8ee36a51ed5a5f7f10
8727 F20101221_AAAIUS habib_r_Page_164thm.jpg
f8d55b7513d6bd5abfc8f35a717bd66e
d2acfa17cdf2d6fd0b6ea4bc8e389b191adb08a2
2198 F20101221_AAAKOF habib_r_Page_004.txt
95794f34ba1630acc355cbbc7ff2376d
fbf0b65d7ef4303ceb74775d38e859a6de07736b
2299 F20101221_AAAIHH habib_r_Page_186.txt
04692bfee9043eeb5271eedafc45d0a1
11cabfee0fea839026f517a461ba67b164201bbd
6759 F20101221_AAAIUT habib_r_Page_318.pro
855e1eaf0e05253c0eae5eac9ec8e632
33218afa6f1b12aef9c0a456cd7a4d2a4afa6b84
724362 F20101221_AAAJRM habib_r_Page_210.jp2
7393ade1eaf38af251d1fbaa3ec21022
dc450686f1784df7eb125832f6f2980d9da7b442
1470 F20101221_AAAKOG habib_r_Page_005.txt
ada50071e915704ca6f661982b4406dc
8edab683df6c15118d7e10a26a312e68544d48cd
37219 F20101221_AAAIHI habib_r_Page_024.QC.jpg
2eaac3032d2716508e767befc25f5eca
5ec586545b6bdb7a63f1d7e45d6432e6f1a0f6c5
106600 F20101221_AAAJEB habib_r_Page_071.jpg
82737eae15a45f81f111b72d8dae45a7
8080f5390ebeac32cc6df58f3d4bc70dd6a28517
34349 F20101221_AAALLA habib_r_Page_237.QC.jpg
a2e27a014224bd72a84ca3d3e445e9fd
61944c1469be2a98c2403c40f77e56fa5e1d0228
F20101221_AAAIUU habib_r_Page_156.tif
6a02bd0234e102809e099c25d7c92a87
bebb2166a277f543c6de882af2188037b46491a1
F20101221_AAAJRN habib_r_Page_211.jp2
e1e1c7cadef6d1cf3f4e72c1836b6e64
80256a14c1b52df4e458e4f35d002ebc0e014e98
3361 F20101221_AAAKOH habib_r_Page_006.txt
c0d0fa9a29638ec38175d606603617fc
47902a87dfa9f5cce8c85e0f9860f0e6f9468cb9
48241 F20101221_AAAIHJ habib_r_Page_092.pro
2e0ddb74d6eea3687ccc0e964f1cb316
24fd766a7eb68162f5d40f63993ee84bf8b8564e
117816 F20101221_AAAJEC habib_r_Page_072.jpg
b3e51b9b5b969960e38ab1827df0b3b3
9e2d3af53e76ec4540840666a6b38e23a786ea46
8275 F20101221_AAALLB habib_r_Page_237thm.jpg
84250d15e82340d67225cd4cfce02dfd
1577c751bbcd460ea445082b48ae2df2374d2efb
5711 F20101221_AAAIUV habib_r_Page_303thm.jpg
6a9aba9c65d803b76fc408d60ff3d3c7
99d362901f9ace54995279ef8f1a79ec5bdd6856
1051927 F20101221_AAAJRO habib_r_Page_212.jp2
5052a16ce22ff9ca04c2029123520b1c
a3292c7c0d056c7f708f9f6521d43e8cd3ccfda2
9330 F20101221_AAAIHK habib_r_Page_318.QC.jpg
f0cb73105f4831c9d7a1a7f26af711b7
ffeb35fa763e0f7e390eddf0336a556c060c1230
114369 F20101221_AAAJED habib_r_Page_073.jpg
fca92a07a450477aae16d631bfddff09
3e8c5f80cc28cf1eb351bcb64d6a2f7c6f989ff6
30555 F20101221_AAALLC habib_r_Page_239.QC.jpg
a56962e509c1187cdc7cacfacf3d652b
b8545832251c3d13870b4d383443a6cd194d7ef5
F20101221_AAAIUW habib_r_Page_074.jp2
5a6b21f1a2695f1abe9334d74de028a3
8651b8a778dbdccad175e80bfea26a88ecf37487
211771 F20101221_AAAJRP habib_r_Page_213.jp2
2318561e01e51660c12f48f9f9c07eae
92e291a3dad9f3242f1bd14af2e8dbe7fba7233e
3979 F20101221_AAAKOI habib_r_Page_007.txt
79f03f23efff88ca3a793fd70545a496
53e29f69427348f761d6c682efe2252ea1cbfcab
31315 F20101221_AAAIHL habib_r_Page_120.QC.jpg
df1c1bb09777bfe12a4bada2e216456c
bc5d918f33b0229392d969de8f9132bdbdf293ba
91184 F20101221_AAAJEE habib_r_Page_075.jpg
89fe9f31d81cbd4d11b53647cd69887c
7f623c19d86783f48ff50e4bef79699d8005574e
7648 F20101221_AAALLD habib_r_Page_239thm.jpg
3f10273427afcda4def6474919d5abf7
0f7c880b5800354e380c871a1e4f8328d8c10c88
F20101221_AAAIUX habib_r_Page_124.jp2
72ea7d6e41b695f516db556599ef998a
31e89473330744869eed486d246db795bd18a8be
1051979 F20101221_AAAJRQ habib_r_Page_214.jp2
e7e6a6a0c4072d8fc9d40f743ef507a9
0d52356ea274ca5d8032c23a0d585f2b72e3e902
3555 F20101221_AAAKOJ habib_r_Page_009.txt
ea389674e4ebaacb6e0e77a3c9391463
e90fa1ff5381363832971c4a5c7b613984739dd8
834 F20101221_AAAIHM habib_r_Page_289.txt
5ecfb04801a71eca215ab4498fe8d17f
a8f51f9b865ba1160515e1c936430da026eb9c14
115338 F20101221_AAAJEF habib_r_Page_076.jpg
ebc3a58ab6bbb84e5c77a6a052548721
c4791c71248ceef4f3ca760df00e988ae2f11b2d
1051974 F20101221_AAAIUY habib_r_Page_122.jp2
68f2f14591e1e7dcf7ae03f2bb37d147
8d22f61009a65ef4bd1d04fa2913f5750384b898
1051931 F20101221_AAAJRR habib_r_Page_215.jp2
f3201b03ab9b02c2f5f5f57eb36d7984
4a32854fd1faf43875b1327b6ad27a1b2499a4a4
3306 F20101221_AAAKOK habib_r_Page_010.txt
6cfdf9dbe017013595f3d3304af32ec4
54b5467e4225143baab588ce31341953d0304aac
9017 F20101221_AAAIHN habib_r_Page_155thm.jpg
b03dba1ea4e521155ed6c19e43e62650
3bed3fee8e1e33db113d6ed9d3d1258492304b96
107532 F20101221_AAAJEG habib_r_Page_078.jpg
ccd5f188a54f1560c553b383d45fa1bd
c94f2ed25bf1bf547ad9934d216c94fa8ac99bf0
30955 F20101221_AAALLE habib_r_Page_240.QC.jpg
63387b4d5e989e7a456bbd7394077d70
7a28b4daeed7157843844f2de7ee6db4c4d7158d
49196 F20101221_AAAIUZ habib_r_Page_257.jpg
a2cc06c60282bde55662ece5474754dd
6d7d2a251f696bcfe8e74ada26ff424c2461d49b
F20101221_AAAKBA habib_r_Page_213.tif
1a680e6ced0305d994dd87baa39cc82d
80498e1a88efd6b410da187862f1913b48d0db3f
989593 F20101221_AAAJRS habib_r_Page_216.jp2
1217cd48aff4aea44b9176ae7a6bfb7f
862d5ec05fa496f1abd679ef3079a9712fdb6a2c
363 F20101221_AAAKOL habib_r_Page_011.txt
3582b49a7d58e413f4084415468e7999
b97d2e1d17dee4033fcf0d5b0ba3aeb02360832d
68155 F20101221_AAAIHO habib_r_Page_195.jpg
1453e7a8dfffeb6baf69def9728c6358
8960d0ceafe818be025a60aef380ec8dcbade962
104090 F20101221_AAAJEH habib_r_Page_079.jpg
daf68c48039e348004822dcd5b136119
a8b3189f67b36d2abf51327e1a33fc7e40d40d2d
7935 F20101221_AAALLF habib_r_Page_240thm.jpg
175425018ed0792cf19c6ceb99a70888
f6beef0750e7ba784a6d340a28bbbedcec653545
F20101221_AAAKBB habib_r_Page_214.tif
1741a2dea6c6eaf4f6490c95a6ca3fbd
8455cb305513e1382760de9dbc11388d8a6af92e
F20101221_AAAJRT habib_r_Page_217.jp2
d553c44ebf00fc6dc87e812c48c182f0
9016ac0603783b1b9079e57104f5dbb4d0f2e2a2
2523 F20101221_AAAKOM habib_r_Page_012.txt
0482e4ff6cd6b198929958c7f7a997bc
30222bbd8517bc348f41c0a2a46b6be834837282
55707 F20101221_AAAIHP habib_r_Page_131.pro
fa3b070dff444e34bc7dcbf929b22d3c
a963966339d1cdc0ee783c6d7dda21d20792386c
112387 F20101221_AAAJEI habib_r_Page_081.jpg
983f46db331037f201ce97cce4d9ca8a
e7d699672b2bdd9c8ebae55a301734159576ba65
8844 F20101221_AAALLG habib_r_Page_242thm.jpg
2cec2a078dc0d8561bea7f06c9224f8b
7f881f92f0922fbb10f998681b1cf6e6612736c8
F20101221_AAAKBC habib_r_Page_216.tif
8537f9a13578fc12b7d0733f2462848e
db9bb31f7d599eb0cbb792df6ae6629bb0de3169
F20101221_AAAJRU habib_r_Page_218.jp2
709391394bbde966da1b9da88eb56776
0afa13db196ad09cfb6c36c89b1995665413d9a4
2945 F20101221_AAAKON habib_r_Page_013.txt
5b29a6dfcfd485dbe991b7a99d51e226
0a8f62521fd6963869406aef067911ec08dc5c5c
104519 F20101221_AAAIHQ habib_r_Page_096.jpg
d71f1348366e1735b13d14b4183399c1
012ceaf141b18f2376d166ddc09d2a64b406304d
129515 F20101221_AAAJEJ habib_r_Page_082.jpg
bbe6785f618d958d38f5d8194e0b5322
9b1aa0a6046cdfbb453113893f5f6497261df3be
31560 F20101221_AAALLH habib_r_Page_243.QC.jpg
56bbd89b8059d1cebdabdac81cdc15dc
c977a704db07869d0cc417515a780cf144496823
F20101221_AAAKBD habib_r_Page_217.tif
5803cf520d98801e0acc68caff6f4a7f
17fb674a7524f5117a8bb478e9be2de98e98f623
F20101221_AAAJRV habib_r_Page_219.jp2
93134abc8cfc3603b0771c37d69f3bc8
0ed145a9877a93a50badf48156001cefee6b2d0f
2768 F20101221_AAAKOO habib_r_Page_014.txt
ca0ac6efafb03e5472aa2c5a2bd8c3ed
8dd594017eb19bc0a277755c0c7f7bdef2bff5a8
F20101221_AAAIHR habib_r_Page_243.tif
806f730eeceb968f580cebdd4ff84f81
ddd7e0226be651a1a0fdf1af053265435062112f
103958 F20101221_AAAJEK habib_r_Page_083.jpg
363446e248e2bfd96e55112eb2dca914
0a7e1174d453da2d51c503674c8035d5d8dc510c
8460 F20101221_AAALLI habib_r_Page_243thm.jpg
30f193137551eb12fc62950d2d23379c
10abca4316dce02a6f1017d9edcd640380f27914
F20101221_AAAKBE habib_r_Page_218.tif
1c7f53be8ea102fcaa31ab070fa8ea71
096ba15c663a6bacbd8091e2788f7c78fb253690
1051968 F20101221_AAAJRW habib_r_Page_220.jp2
fd76b417e7dc7a72cbca392ad817c1a4
18604a6d1c001901d259168dc0e55d4352c61007
618 F20101221_AAAKOP habib_r_Page_015.txt
c62a00cbf7fb27009729c431583688d9
6272455081bd521e1f8f85fe637910e8469e2558
19292 F20101221_AAAIHS habib_r_Page_317.pro
14b3af5d500ebad6749819364ffe6bae
01b2e01da7ba6cf82e0321486cc5be0a96a84ea2
116011 F20101221_AAAJEL habib_r_Page_084.jpg
66908bc75d07e8c0f25509b541839369
a17e0639119486e5bef5fbeba50bcbd06073a478
28082 F20101221_AAALLJ habib_r_Page_244.QC.jpg
bbf09691ecf2592dcb8b7ba64bf37348
a3de14a694814b8b782a4cd8f1ee84ed60bd2705
F20101221_AAAKBF habib_r_Page_219.tif
031e04651f701af49132250f9f114a53
556a9d3f877dffca11c48bc067c6ab1b85ae8219
F20101221_AAAJRX habib_r_Page_221.jp2
dedb06d22a6b590441a5e6e1c702844c
3a1e78818b89423637ce66e39aa506f477326794
2282 F20101221_AAAKOQ habib_r_Page_016.txt
f659cf4596d7bffc41292e91f4b7e298
70c8df641628370380e3afaff3f3c4aaa50fd7bf
5250 F20101221_AAAIHT habib_r_Page_196thm.jpg
4f67d664bbe8c6124533720352c8d03e
4f4fc51de7eec560d0a7f39dc4370f78b65a515b
65721 F20101221_AAAJEM habib_r_Page_086.jpg
541f55f49aa8fe3b3999b0d011f16171
ca7574cc0c304e6b5781e61d44435bb6064dfb2d
7558 F20101221_AAALLK habib_r_Page_244thm.jpg
a0424fa1f06b656386cf80d037a192a8
09120d2995fa34550eadf059be1360b3b8c27d24
F20101221_AAAKBG habib_r_Page_220.tif
2d378a0c09fc6727e45160b8b150e8ea
9539e04cd3a7d46651258d194b229149ba4d53cf
F20101221_AAAJRY habib_r_Page_222.jp2
770666c4e66f44001dc52ec80cec16a9
b28f34573218a454dc24ca7ed898d5cb2bc9365b
783 F20101221_AAAKOR habib_r_Page_017.txt
b3fb4fe5f9735ad9779bdc516000814b
04613470537674b556c6eed63b5c5e1057922516
970381 F20101221_AAAIHU habib_r_Page_102.jp2
e4cae97263491cf3fba2ca7174e4c4b7
c73e8bd5bb9aea861a73395670b272577bb2b794
101978 F20101221_AAAJEN habib_r_Page_087.jpg
204b39ecdb5bd806f991b3d59121d92a
51720b4e1548cb2c3ac2c0d86fa94d1f5fccd06b
19188 F20101221_AAALLL habib_r_Page_245.QC.jpg
2fc59e8d2959066c0549dc30a2fe55c0
ca15a95a80e9351f65f6ee1c90ffc8018ef158a1
F20101221_AAAKBH habib_r_Page_221.tif
7a788f2427c46e1bb82b768e994ab3a6
d98d510c2d9c554f2254b013cd6661c3c8da1024
F20101221_AAAJRZ habib_r_Page_224.jp2
f26f3980c0ee4b9f98362819fcb3f755
d1b2dfd91c1b523098785098b48ed1ae0b372084
2035 F20101221_AAAKOS habib_r_Page_018.txt
decf8d32b8d1aed67609aaef6b4fb762
174e1368257bf382954a2b1441e4079ce33659d6
165116 F20101221_AAAIHV habib_r_Page_007.jpg
4a40167af7a4c5bc490ea4ac30166a50
089d392c5e54abeec542454ec6490d8c29ab8e5f
104367 F20101221_AAAJEO habib_r_Page_089.jpg
c603987337d5d29034dc95ec85df097b
186ab1699c7323766ce87d1e80e713ade1089253
18300 F20101221_AAALLM habib_r_Page_246.QC.jpg
5fcf919566bf4927efff063cd68344d6
f60ae37582160f90bfade98f108bc3b77836546f
F20101221_AAAKBI habib_r_Page_222.tif
9f72dae92565905ec90a387542742746
73fd567fce78185e81e1e17935d79bb2aa4dee53
2223 F20101221_AAAKOT habib_r_Page_019.txt
bd5fd3cf76a44c908a8f06baba1b482a
9b6c9e8d9831f2553bb71849fec22f3f32f809b2
F20101221_AAAIHW habib_r_Page_095.tif
59ff5e21209bab4574a407014076d729
1c45caee1dbaa2ccc651dbb1271214781828dde2
91763 F20101221_AAAJEP habib_r_Page_090.jpg
e04b71a24673126c6cacb1bd32b33e02
2c89a68ced4798a9970ef966a91941eb35ada09c
5663 F20101221_AAALLN habib_r_Page_246thm.jpg
cfd5ae9345a2b34c9bc96afce55efdfd
44e9e1c2070c8a3df246e483543c5d9b3a88ba3b
F20101221_AAAKBJ habib_r_Page_225.tif
8c9d1bff02a28bc68ddfdc040968c89c
e2bbf1e6df2111d999580278b8db3b5b70012922
F20101221_AAAKOU habib_r_Page_021.txt
8904fe1b39fe083f21e6166e71509b16
a5c82c8f81658c03d24017da4f719bca342d92b4
F20101221_AAAIHX habib_r_Page_135.jp2
cdd3b0ac75ac1c8a43c98647c7c44018
adf3dc8072d69a7f850e4098517ed2150cc3497c
101211 F20101221_AAAJEQ habib_r_Page_091.jpg
789d8135669f496bf174d8d3d9e8afe4
e2ec97036ce2b6d06a334b5d85091707c52237f8
18922 F20101221_AAALLO habib_r_Page_247.QC.jpg
76a8edd4bc5f37a211d597ce8ab7d5cf
44b2570ad343dccc2679fbc370195c122cc2924e
F20101221_AAAKBK habib_r_Page_226.tif
7c7d6617cbe16e9112f299fc56c3bc6f
27c7870d0afd151bbe7fe5f98a1ffefdd9ba9fce
2235 F20101221_AAAKOV habib_r_Page_022.txt
a12c35c480b54472f3f284a3b18e66cb
08fcd2d2e02b9f75d6564920921b482e39be9f67
23320 F20101221_AAAIHY habib_r_Page_317.QC.jpg
5d1668a6470abe73efc6d6b07edfee52
f87d96aa0eb5182b98f2686fc5a5d9654238677a
98569 F20101221_AAAJER habib_r_Page_092.jpg
ce461d881a50d83234806d25a8cfd071
0b0d6539ebfebdc3ca8907b46cf3e11180d409e5
18027 F20101221_AAALLP habib_r_Page_248.QC.jpg
6190a429c799ecef6d434c17d8502da0
96d4ecf22d4d25b7d2e8eff7a2be30a2727180ba
F20101221_AAAKBL habib_r_Page_228.tif
339a44aa95c4ec8fef7dbb8e460554ac
6f2672d23a41668fa5cefdfbcdfdde0632dbe69f
2470 F20101221_AAAKOW habib_r_Page_023.txt
f501d28099922255c21ca4c95d4b3c74
0cbeda51b3090997c4b3dc12de8c7ae03016549a
9056 F20101221_AAAIHZ habib_r_Page_014thm.jpg
c3f906450bcb0c85ade7ca75eee1d48a
6c454555962a8dc5cc5eb0c4b712bc57a459ff4f
91274 F20101221_AAAJES habib_r_Page_094.jpg
234f2f2685afe5f292fddc8a5e1204fe
444891dd11a57a435f24824539b08d54e59cc50a
5411 F20101221_AAALLQ habib_r_Page_248thm.jpg
1e5cb6286915cddb0aca4a5affe48b94
939236754bd0bdd343209228ae4e53bb09ee3400
F20101221_AAAKBM habib_r_Page_229.tif
73d8ab13bf1604e8267a4034169b3159
3c42c52fda6f10f969aaf35ac9fb448318bc2c2a
2266 F20101221_AAAKOX habib_r_Page_024.txt
7c9d67dccb3f563031b2ae22075d39c9
69409f9ec4450f3712d5a4096b14ea28dcd46a0b
98551 F20101221_AAAJET habib_r_Page_095.jpg
aa8bcf55bc042ed7e3016710b56a8387
2cbc22ec51f606b08696e5a68728ce6c30a4b75d
4339 F20101221_AAALLR habib_r_Page_249thm.jpg
4ced113be9766235212eb233fb77e16a
b0199326e404de7f2c157fdd51c0ac776186ac66
F20101221_AAAKBN habib_r_Page_230.tif
2da71afb849848e701a1770105aba166
51a42fb1cb4ca81bec247c3f2453819391bbf7d2
2135 F20101221_AAAKOY habib_r_Page_026.txt
ffa2427c89ef71e120c835d09b0ab9db
fa03e91dfa32651b639929a5dd189571bb909de1
81895 F20101221_AAAJEU habib_r_Page_098.jpg
10071224785188120ede22e9bbaab072
f083fb98fc392674182daf8b400bcba68587a7d7
19737 F20101221_AAALLS habib_r_Page_250.QC.jpg
d2e88d0640ee9ad42eb3cdc95608b6b5
0bfb54d8d80105b4fb92bb88a42610d82605882d
F20101221_AAAKBO habib_r_Page_231.tif
9bf95b2e5f642229ddc72bebfe5b5932
f242e09a4c228e9c3fd8d1c445561e0768dca3aa
2274 F20101221_AAAKOZ habib_r_Page_027.txt
26ecb6db4b0214ad7a088f61475b2f96
421db11da92c400760d1f5f37b636b81b05f565e
97048 F20101221_AAAJEV habib_r_Page_099.jpg
f2da02125564a864757e63b8786a6aec
efc3e46290e194a6010f5c799935bdb4c9bda9b8
6066 F20101221_AAALLT habib_r_Page_250thm.jpg
69d7be137a4e5d789ad5643663cea228
2ecfbd36e226f753f72e8a4fe0732f4cabb9494e
F20101221_AAAKBP habib_r_Page_233.tif
64a95b50e0c86ca956d88204708fe72f
0af3faf38fac65907774f88ef3b7303a9a106968
111500 F20101221_AAAJEW habib_r_Page_101.jpg
8d53a0f647d6120d1ae9b2f4889ffa84
39aed26e546486e4f830fb346585a682459b4552
14422 F20101221_AAALLU habib_r_Page_251.QC.jpg
5fab478f146b6d77926b2cb17420ca86
0a6a91035e1f462485879bc2a2edd5127496b361
F20101221_AAAJXA habib_r_Page_073.tif
99464ca08d92cd42a79c5f325a5bf9f5
0802670b48f9e1196c19ba20bc377ae319dc07db
F20101221_AAAKBQ habib_r_Page_237.tif
c73677970a27e83d9a4d04d897a8250e
6507e07ffebc4ea6f92c23d1c805fb3d03c9e97e
92440 F20101221_AAAJEX habib_r_Page_102.jpg
30d6a0339ea1b05af16b804a3b36138b
ba0d901da8803af533adee54d3583220e399f063
4581 F20101221_AAALLV habib_r_Page_251thm.jpg
ee9e5b794bfe9ddb2a492be2fdddbd6e
70cf33d892092eda159f3508b3e6596d0b8a8d1c
F20101221_AAAJXB habib_r_Page_074.tif
b3a82e942f446c1da336b71499aa1663
f68b631a37b02b27b3c0224d79705d7ceff15dd0
F20101221_AAAKBR habib_r_Page_238.tif
248a1302a800523ac2420226bc9667e7
f7e934e6606fd60b2197f85f1937778ee7007200
114610 F20101221_AAAJEY habib_r_Page_105.jpg
561499da2137dd8081fc9e5b56405691
8b3af8461e863fa8f4e415443dfa4f92b25206f4
5395 F20101221_AAALLW habib_r_Page_252thm.jpg
ade9ec931619f45bcf08a1e51369aa31
4739ce66783928743ca83e0fcfa82660ffa40dba
F20101221_AAAJXC habib_r_Page_075.tif
25ce5afc4e4f6a1b6d490abcc35d3443
2cf2dabdf3e141b3da5db4365da1cd79f75fa1fc
F20101221_AAAKBS habib_r_Page_239.tif
932390b8d6bd3e7a8848fe031cd5daa6
b1382c3dd0ff70c08e51132df1546db559f99613
100686 F20101221_AAAJEZ habib_r_Page_106.jpg
a55e13bd59ba58881212eadcb40e3c93
795bc55bf99b27442eb59554ada041f4513be1af
17893 F20101221_AAALLX habib_r_Page_253.QC.jpg
69fad3cd71494932a5088ed0f259c86d
0d506ddffcadbfbaf3fcddbc529f369bc4064c59
F20101221_AAAJXD habib_r_Page_076.tif
e2725a96680e351009aaa887a6341ec9
26ddb6588b418eeadb1838d3f76a979445da7589
F20101221_AAAKBT habib_r_Page_240.tif
4dca3ae84a1b7f6bfda1d25fdb73c962
ea5d81a47172eae6394411c535cd65722182b4d4
5613 F20101221_AAALLY habib_r_Page_253thm.jpg
be1f5bcb91d9f36a9d674e6e8623fb6d
08898dac72d97faec9fd03aae0e1c8813e6f9ff9
F20101221_AAAJXE habib_r_Page_079.tif
bbf546442a04ad761b83a2d358886543
bd24e30b002153070feac9bc87b7e4367806e89e
F20101221_AAAKBU habib_r_Page_242.tif
37667edc12f8b7efdab800d0c4a9eb78
167b0d5e7a8b3bbefa62386d5d1e167ee2d2fbfd
40757 F20101221_AAAINA habib_r_Page_251.jpg
ecfede52c38f1a6e2a2a3c71e6544296
dc8f12422c24c0119626f177954104c0d24b85d3
3486 F20101221_AAALLZ habib_r_Page_254thm.jpg
de26783b70130a32cd0431436e0f6895
58d2d60d334a1193848a51682da2a985a8da06ae
F20101221_AAAJXF habib_r_Page_080.tif
9074181fe6bb9d17bd9066cdda67c0c5
1351111d6595aea67ded105db14713630ddf8120
F20101221_AAAKBV habib_r_Page_244.tif
33a8075917512d2aedf2206c88d1a809
5ef114c12fbe0ed4a6f800650f716df9309902c6
54990 F20101221_AAAINB habib_r_Page_241.pro
5e126cdf7f63eab1f674d745129d501b
95f2baffe8cc3d581e8c1539883222c072381bbe
F20101221_AAAJXG habib_r_Page_081.tif
4afd0a67f3604c2cf154bd9ed13bdacf
575a1b1d4102e2140052ff25f89ba2ac8a86cd0d
F20101221_AAAKBW habib_r_Page_248.tif
eacc90e067145c2288821d4a03b6c269
dd710668f88eedf0cf313dd9c73351ea1d7de4b4
111723 F20101221_AAAINC habib_r_Page_074.jpg
547d7b8f6eb8e5c93bce413aa5ac36c0
fa470a4fcf7963bb4a91f20c5c791550c46d0317
2169 F20101221_AAAKUA habib_r_Page_206.txt
a8783d179d1a02a7e57c9e81411b96e6
a4e5b33004cb4f2416647c3fad19023e5ca8837d
F20101221_AAAJXH habib_r_Page_082.tif
002827e26bf627922f834168ab542d5d
465d3b775d8e5b4d0c967115329920f788112e95
F20101221_AAAKBX habib_r_Page_249.tif
131cb8a32c3cfe864129ae9a142fdd91
10711a10c33307cd4d859979994adac468c4c8c2
53024 F20101221_AAAIND habib_r_Page_021.pro
a454d5a12def1931cd9ec915798d9863
81e8b42e59f68b3d0d8ddc8e0a58f4a893250733
2262 F20101221_AAAKUB habib_r_Page_207.txt
09746bda9d329265b8dc4bef624a178a
935f0f9f6112a7b6cefaaf245cff4e447442a605
F20101221_AAAJXI habib_r_Page_085.tif
1ad106bc906cd445bd188f30bd437840
00e5e8d08216a5548808ba5ae8bd5246ccd7a663
F20101221_AAAKBY habib_r_Page_250.tif
9d715b1070a99f65e98c9df998e1b268
cc444d38094f2b103b455696430add9124fdb296
F20101221_AAAINE habib_r_Page_247.tif
648c599fbcfea74fbea3a9ad5bbaf3af
5961712f255ef7952a700545a0f6b1345ef1051d
1746 F20101221_AAAKUC habib_r_Page_208.txt
bb9823248b47d2a88c20f39b33ddbaf3
a0de68dcf734905df44c563dedbd30844a7cca00
25265604 F20101221_AAAJXJ habib_r_Page_086.tif
dcda85dbbc52c6ae184b7ef87bc07f2a
93e29f39dd041685bf850527ba3a9d3db38b020f
F20101221_AAAKBZ habib_r_Page_251.tif
2ef14511669dbef3b72dc261e8976601
83c6186ee01167bb99fc391d966724d2f06802d2
45639 F20101221_AAAINF habib_r_Page_179.pro
276e74105da2adbf87f2fb42c37773a7
49ffb486a8fa8849303c5f5349b0fc85eea37679
2155 F20101221_AAAKUD habib_r_Page_209.txt
8acd379bf32e85a6ec19df461b22b355
c75abbedbdc1869e85e51e6048265b127c07016f
F20101221_AAAJXK habib_r_Page_087.tif
b477c5b19c02fb807d50d8c622921a64
3b655e17f2b0f8a1033587f79eed170afefcfb09
1486 F20101221_AAAING habib_r_Page_291.txt
06cb6eae99d0bfab6800f66ba5cc92cd
62c203ef13da25f9ac6822a5491cbd3fc6d30936
1456 F20101221_AAAKUE habib_r_Page_210.txt
b6c5a1e8a7eaa8f41432268e036eaab3
4921243b9c8f5ebcae19931bbcb3c6b7c835fb9f
F20101221_AAAJXL habib_r_Page_088.tif
d9ddcad55f074717e5e4c37285af8639
e30a07dbdd663c3439ead5365da2ebe822a00f65
8822 F20101221_AAAINH habib_r_Page_142thm.jpg
e928aa8152d7af86401496c06107719a
0463a24c1275ca9ed7129e2db446730ea793d30b
50632 F20101221_AAAJKA habib_r_Page_290.jpg
b290b6bd4b89d9419ed5637821feb7f2
f441aa78b5422c816e5b1475a7173edef2974639
2196 F20101221_AAAKUF habib_r_Page_211.txt
b7d59a054808742cc98197a04544f46f
788b69d5f32df2661aa986c1bb9d377760ba0d2a
F20101221_AAAJXM habib_r_Page_089.tif
689d601b55c205339ecf1121dbc3921a
48d2266e266bc152a810d29283573cca2cadc50a
F20101221_AAAINI habib_r_Page_171.jp2
33bfe0156f8d4245f9b603c27c9e18f6
d9592e2c3cd70ec29abcc914e4dc93425ad90bb4
55454 F20101221_AAAJKB habib_r_Page_291.jpg
f6da25836cd72e713008b81e3ec73df5
3e3b6767f465df89393dacbb756a19aad9baa38c
2008 F20101221_AAAKUG habib_r_Page_212.txt
10b26d57dc4c4390b3d81b7e56830113
860cafdbf00eb6a8d0928dd00b2cc3f0c5ff66db
F20101221_AAAJXN habib_r_Page_090.tif
5eb50a81d77eada3e59fe12047b33ae4
759b2dcf1bbb5e9ff177603deb1cd357e0a73ca8
63699 F20101221_AAAJKC habib_r_Page_292.jpg
77e38a1be5d8bf628d71f6422a0a4706
83b3870febe7ce351aae3b2ffc08b0e7225944dc
412 F20101221_AAAKUH habib_r_Page_213.txt
66e6fae49f0b9b4ba38be683a198fe79
2e497db3b6b3be2e70365af21dc53445d5122ab5
F20101221_AAAJXO habib_r_Page_092.tif
3a8731111abbd1251d72b08fed8bdaa3
58c67e9555c9f77545712afc5dd64fd507691cdd
F20101221_AAAINJ habib_r_Page_225.txt
cb12c4a3a3ad2bbc5778ef7f38e245ef
103190dace7236c87d5e8aface36baf9e4700eea
31803 F20101221_AAAJKD habib_r_Page_293.jpg
c851e65f289660eee9820ff16847febf
f86821b5130c96239be40b18792bc36df17d8935
2161 F20101221_AAAKUI habib_r_Page_214.txt
e9178c869a47cb62da38fa9cdbb466d3
bd92a00369a3fa56e68e5afb4865d3dc503fc932
F20101221_AAAJXP habib_r_Page_093.tif
9d6b34f9b2891035dac9b92f76e39a8d
1b39c8ae145bea597ddb06b7da84033190fdb11e
27934 F20101221_AAAINK habib_r_Page_148.pro
0751dc74b41bf5acb5021c47cc311eb2
99e3374d7979871a5eb3234141607a7faca2e905
51188 F20101221_AAAJKE habib_r_Page_294.jpg
822088e4ba25b7e98361030af044ad44
6123014ad03558965e8f2a3ecc68c63995afadb4
F20101221_AAAKUJ habib_r_Page_215.txt
f3d78d744ed547eac43d40ae32d0243f
0b990aad092451275156992625b1ace7995e5df7
F20101221_AAAJXQ habib_r_Page_094.tif
35d922c868870867f01fccb1767a2954
aff9f397716da69bef298e46191016de27ca2976
F20101221_AAAINL habib_r_Page_184.tif
9553fd4c853a41f04f39828217906ae8
7c3ed0df862a29ad3f6efb07448ed5303eefea6e
1807 F20101221_AAAKUK habib_r_Page_216.txt
c1dafc68a120706ac93b1064ef215355
3d2fb63745346f10f80765b860b03736574f75f5
F20101221_AAAJXR habib_r_Page_096.tif
f9935add2df0b789773cb9e873e6cfd5
6fee93cff81d44d1862ccb3f51e947af33565850
36889 F20101221_AAAINM habib_r_Page_209.QC.jpg
4508db5dc7fe8507ce6e24e9988ecd5f
0a6af81a1468bc736c6b467d52bda4ffe4ab6ff5
53492 F20101221_AAAJKF habib_r_Page_295.jpg
b1baa31ad7c1593dbca5ec01e6a0d277
4d4d498fdefe5f60228e4b4474115a90203958b5
F20101221_AAAKUL habib_r_Page_217.txt
2e473071c7680eb38f11ec23cdf0628b
963bcb2b7312d06f15dea99c78bc89fa26329ee5
44084 F20101221_AAAKHA habib_r_Page_080.pro
b8b955d307667edad373a090e35fff47
7d9bfc4b9a137660055d0d2174aad9ca8e6e5c03
16359 F20101221_AAAINN habib_r_Page_265.QC.jpg
e40c904382b7797c3d21e8835d3df154
5171cc14ca75c3477b6b75033a31a92eb8c5ae8f
46054 F20101221_AAAJKG habib_r_Page_297.jpg
80f902ef0674df418abf0bca4904453f
237f3a47305a37ccf48f77e451ce13763fa11a75
2258 F20101221_AAAKUM habib_r_Page_218.txt
5493de4c1f5c84edcf07d9cdde2e9c40
3d875787319cf72f12a96b6aaa30474e3a040f70
F20101221_AAAJXS habib_r_Page_097.tif
24e087026d5e5722fe078f4150f66d24
bfaefa877ba4d086eb518dafe2f397270d163853
103051 F20101221_AAAINO habib_r_Page_183.jpg
13d18a4d0f1cfa4a1825d5fd542c4f52
c0efe8ee48284d8f33aa6122c05687b616ab6818
67872 F20101221_AAAJKH habib_r_Page_299.jpg
cc1b3d84e2e3a1ecf8b3062ef4228628
4c0e0e0889561fbc0931e459ad48ae651f7dd833
2096 F20101221_AAAKUN habib_r_Page_219.txt
4596af2ddbe43887e435787c6167bbac
bd6ade18854913a5e6234b4ae3b4a08f5ac9cf2b
55976 F20101221_AAAKHB habib_r_Page_081.pro
e705c7f7bd7100c8d379fb69d4f10cae
43fd2286d1bfbfaf14637a32f9b0e54b85971ddc
F20101221_AAAJXT habib_r_Page_098.tif
9c060eda6101facb2222381b16f8f08f
eda6fdbe26e796b72f4e99944b1dc4502389c670
108305 F20101221_AAAINP habib_r_Page_026.jpg
015ba7de75fa125216edda6d01dfcfd8
dd69f574c015078f755c07618fb3287a7a917edd
54262 F20101221_AAAJKI habib_r_Page_300.jpg
6777fc48ef21f5546cbf0c631943b591
1623adc47cc165b434b81bfe3e55265be13b7329
77410 F20101221_AAAKHC habib_r_Page_082.pro
c949c0bed440fc0c13e79e4343ec5d5e
144af1ecc8240628f864f71220840ea4c8d67e38
F20101221_AAAJXU habib_r_Page_100.tif
b7079eb8a4958ecfb8a06b3f99b3d34e
31c1a2c5824ec089646733b05988fa04999aaf00
35079 F20101221_AAAINQ habib_r_Page_040.QC.jpg
ea6eb34dc9521d36b62c95c193944deb
0f3df4283f5ec016ca2d5c4b83449ac9141d100a
38820 F20101221_AAAJKJ habib_r_Page_302.jpg
9c0683bd8153d9b37e78f74120b52f84
16b9f36996bc53aca580a0adda1c266f22c761a4
2180 F20101221_AAAKUO habib_r_Page_220.txt
901324ba9908a9123578d6211ed23627
2eab871d407c584c9dcbc5e1fa4d31194918047c
F20101221_AAAJXV habib_r_Page_101.tif
4ca5433523aaf3c18c9962ef1b1c3cea
42c9f9d6dbda02e5430bbcbf76fc40fc4b208cd5
F20101221_AAAINR habib_r_Page_065.tif
7383c71ce874ae94796d6daf992063d1
c144441bcaf01d8e867f44a1f2b030d44b502a3d
50742 F20101221_AAAJKK habib_r_Page_303.jpg
4e1508ea024a6da93d0aa27f915590af
61d0d8e4e55e6e3d5cc09d385bfa9a08d8c1868c
53457 F20101221_AAAKHD habib_r_Page_083.pro
b7042de3c0517f030faa8e7c00962882
a49828abed975f531f8df22f881001cdc59be627
2115 F20101221_AAAKUP habib_r_Page_224.txt
cadd7f6b55b335b900a847654fd7afc4
9399b6873255668b64b6037f95641dc191f2dca4
F20101221_AAAJXW habib_r_Page_102.tif
9a0883dfef7193fb534d7d30ccfe2354
debf32718a7d9693b8c9f2193e60b977791960ce
F20101221_AAAINS habib_r_Page_072.jp2
34aaa085dfa2f7fde2a266629e2e0cd2
b06d8908922b164c33693442f57b91ae6f57b3ca
55748 F20101221_AAAJKL habib_r_Page_304.jpg
19624fb2b70577c42831fde7a2bff723
feeba979fa0d6dfa44febdd6758493dd6a1dc65b
58124 F20101221_AAAKHE habib_r_Page_084.pro
afcb4d35b1c5b11e3ff178845d3cb4cf
62995b68a5c8bda4259d26638d6f5eceb3616802
F20101221_AAAKUQ habib_r_Page_227.txt
292eeefd31d6595bc11a2910260d6bc7
b0b968d48c887c1ed9d984f239ad47e656a97f82
F20101221_AAAJXX habib_r_Page_104.tif
25a5eb5795ed74bf2b35f2f8de68c787
8715408094b446733e9aef857454a606298feb26
101704 F20101221_AAAINT habib_r_Page_135.jpg
e9d249985d0a079f1afa10b1c7cb2fce
06a8e099fb49f1d771fc80b9fb4c1d3c8042439e
63450 F20101221_AAAJKM habib_r_Page_305.jpg
c8aa9269981a8f56e171a8e1faf389ef
cdb0663c5332209a70c321e8b046eae19e605666
9395 F20101221_AAAKHF habib_r_Page_086.pro
be434df7c74df99fa695f9a1bf525ccd
1d52e672c99c2c63b9ce2dd1bcbd69f20cbe9f81
2074 F20101221_AAAKUR habib_r_Page_228.txt
14b127f338a6af2147472430fe13c01e
34324d63d6d78c2384a07eb716334919c0e74319
F20101221_AAAJXY habib_r_Page_105.tif
9293e557eb66e6fb72e782d922fee10c
5e6e9ea165b5206d894992216a9206a27665f202
99676 F20101221_AAAINU habib_r_Page_093.jpg
4e3eff3cab7248ed60a03cc599c63c45
11c00089559e492dcef56a99912b36ec8d3ea8fa
36396 F20101221_AAAJKN habib_r_Page_306.jpg
4b1794cd5a738c659f70de939bd23b82
91724e3f221fca43f9963bbda77370b5e421184b
51591 F20101221_AAAKHG habib_r_Page_087.pro
07487a0198bcb7ec7d2b7a3ae03c9ac1
0d07da4a232c288fdf93662fb8484046d1c8e340
8156 F20101221_AAALEA habib_r_Page_107thm.jpg
fc6a4ee70565ac60e920a96f5c7d4eb9
cdad1e4ae2163f3d6c7d550dc72e9f7047c70af1
2142 F20101221_AAAKUS habib_r_Page_229.txt
a440cebe43243da05cdf2dbef5e83819
6a51a979c568c63cf13f666e501a2aa910388a9f
F20101221_AAAJXZ habib_r_Page_106.tif
e3379951c18098d781f9ae4081efc00d
d706d54ba7df89aa45d20372dde1c97d969cfb82
F20101221_AAAINV habib_r_Page_021.jp2
144b72d89f76f03639bebd6d011430c4
c073c9d3e76d3b121ec3188bb85ac7d9a8648887
49958 F20101221_AAAJKO habib_r_Page_307.jpg
f7a1be83553079749269b89bf8b63d88
bab9db71754df50343549069dbaefb5a34aa3f0f
49827 F20101221_AAAKHH habib_r_Page_088.pro
b8db542ce818ab6c00045204d01112cf
b442979d5770315410f40fd8dc7f7e796731d86f
36240 F20101221_AAALEB habib_r_Page_108.QC.jpg
b820d0b5e476334be72b65b6cfb63a60
a9efdbc0df61140169aab753b782b755f4eacea6
2054 F20101221_AAAKUT habib_r_Page_230.txt
9f4d376b0111109e4fdaf64105a1b7f4
99025a726994933dc8c2ed79c4af3a5e535d9d1e
F20101221_AAAINW habib_r_Page_099.tif
303aa2edf12405f37a6261315cf2a512
c49411a55f713b9863ef6f77671b016904363ae8
45844 F20101221_AAAJKP habib_r_Page_308.jpg
f5cc5a6fc15f856d52c14d7c942c5141
d39ad4b097b8231b3b5cc01b35592a5326c95d5c
53374 F20101221_AAAKHI habib_r_Page_089.pro
594620829cdc0db49c918ca35ee535c9
d8580520352a95ccd4b38a997b214b3673b78470
36703 F20101221_AAALEC habib_r_Page_109.QC.jpg
0f17fcc424a8bdfbf94d71031eb54be4
4940895e26f43258bf926df5e09c6ff5a61792c4
2214 F20101221_AAAKUU habib_r_Page_231.txt
47339959f9323493b6b9152d3165a5a9
5ecab538770022c6e43d052308ac33958a3a85e4
1034665 F20101221_AAAINX habib_r_Page_020.jp2
8c6e5b9e0b6458fc425ad58407dc2409
776fcbe573d0b071afed98cade11fda9d9d44c34
48467 F20101221_AAAJKQ habib_r_Page_309.jpg
b69e72f2c29f20005de8fe2ba5e6afde
1d8edeaf734ddc987420a2def46affada1dc87f6
42646 F20101221_AAAKHJ habib_r_Page_090.pro
ce13abdc75601811d899bbf82a8cc274
0ad4ffef9bf96bb2aca3e50e3f413fe2db48ec8e
8427 F20101221_AAALED habib_r_Page_109thm.jpg
0cfd8d4e7df0c5c1e691a8bf335777e9
8b58d8638be83289e84edcaf20a23942ccbfce60
2089 F20101221_AAAKUV habib_r_Page_233.txt
8001ad2bcb2f19c7257571222c57e93e
abadd372bd58580aafbe71cae2767c018185f562
38066 F20101221_AAAINY habib_r_Page_098.pro
f55276cf0fe3996d7b54388a071cd3d9
9cedf969d5032d5707067d8a458e5186bbc64c78
46396 F20101221_AAAJKR habib_r_Page_310.jpg
545e1f047a72476f0c8c6b65f8d2c2dd
04a3f1b247aa7bbbe0c6901b183d0a1564c56172
47445 F20101221_AAAKHK habib_r_Page_093.pro
d7af931420aa3a637207b5ed5f65ac8e
553702e90e7c71775db140bc546f7dc06c4a7b3f
29067 F20101221_AAALEE habib_r_Page_110.QC.jpg
e5500bcecd3d3d588d83422b4034732e
ddbb4bee213fe7bb7532b7cb5f6729b5c6bbe221
2171 F20101221_AAAKUW habib_r_Page_235.txt
c6503592bcd1e90fd4b79e5fcfd6b8bc
39b9c2f8446744046e4c277309271df7e49f3cf1
53477 F20101221_AAAINZ habib_r_Page_247.jpg
fde98e575b2785d9c03b087cd808073c
0c66398bda2bf256d512bdf18dc6e13224002cad
59927 F20101221_AAAJKS habib_r_Page_311.jpg
5f94811f694a4177c4867efc3f738e69
f7f17da433a9510d67efba800175dd167b1c3886
42585 F20101221_AAAKHL habib_r_Page_094.pro
50aa10580d27d7e03d6691105e29f736
a1715e0cdc7be70500c9a9091a2be6662484358d
7656 F20101221_AAALEF habib_r_Page_110thm.jpg
8e1f6f60c388ed04f3e201e7e1e20005
fd32c8da825a4309e4c951ef22091e90e198ec7e
F20101221_AAAKUX habib_r_Page_236.txt
c3fca4adb4b8a53442a841d3bdcf3df0
9da0fa62d177748d4d0b3470a8709110c41333b2
42875 F20101221_AAAJKT habib_r_Page_313.jpg
8428876bec8662ee34cd282db5653a4f
3926d2e4fbb6b59257f316a0492cb8809d7d4bed
48697 F20101221_AAAKHM habib_r_Page_095.pro
a9ac1334e6bee81b271a98d3329ea08b
0a320d518b160d332692a36547f938700946e869
31243 F20101221_AAALEG habib_r_Page_111.QC.jpg
bb6284502cf83768d563a6cfbb995eef
a35db4fde0605125d6f5ad3907fe0ad057affb3c
1978 F20101221_AAAKUY habib_r_Page_237.txt
d31aba534241b091393714e2ac6a1f30
54ab59258c00ae2d2ee4a93cfd1ad7982c78bace
54347 F20101221_AAAJKU habib_r_Page_314.jpg
d4ebccdbfa1b8ab734ec781f64bf1653
1f8a5b9d798105e5d624d2d2f1bb451f7388ef95
52055 F20101221_AAAKHN habib_r_Page_096.pro
fe6f2ce412205bb88703a81237b0ab4b
ccf7efda3d0ab17266d9e6f4a9ebf65f00494012
34852 F20101221_AAALEH habib_r_Page_112.QC.jpg
9a8423cfadf2fc9f09dcfca2de2bcf8e
41b77670577c26fab672fa2861040364574e67cb
F20101221_AAAKUZ habib_r_Page_238.txt
1aa4bbddd8812d07fc2da3aa13005ad6
1e6def8d3e582cf012c988dfd1db69062ec383f6
74257 F20101221_AAAJKV habib_r_Page_315.jpg
c8359845e3cf2588c6fb52ab056f1a7d
f95e1e96c4cd3deeb9ecb473da38deb4b4baca04
50604 F20101221_AAAKHO habib_r_Page_097.pro
aa9a57ae020bc66e572ea829d1be3c9f
e10a210f8d3cde0934ee66425bb50ba0b6cc94bc
8546 F20101221_AAALEI habib_r_Page_112thm.jpg
2cfe53a2abf7ac0c01e5412e95ee7628
0422b4c116e911303b554b7373d3bf2f685271c9
26765 F20101221_AAAJKW habib_r_Page_316.jpg
c75f2df1db226e8e3515bf03af3875f9
df26000c8aad3a965ba6bde4ab709c7c7ed6a08b
47509 F20101221_AAAKHP habib_r_Page_099.pro
c58ad198969550474a4a3acf1211360c
a2625df9ddddb34f4da9658e77e28c23377a8589
34985 F20101221_AAALEJ habib_r_Page_113.QC.jpg
3ee234b73603796afa05176c7438c83c
ce55470e778b2c34a60c78e2c46d3fc1ec605f4d
25121 F20101221_AAAJKX habib_r_Page_318.jpg
bdcb1fa0836c67327171b257bf8b2310
34fb442dc67d65ad2dbf4bb7d0852c62ac7a77e1
55768 F20101221_AAAKHQ habib_r_Page_101.pro
5ea1ac7d97d0c9f85ce1fafb9c15091a
c054ef648fe7f8fdf24ba0fed84c24a4796f98e2
8555 F20101221_AAALEK habib_r_Page_113thm.jpg
e0a3686b5a669aba8cca0111f5014533
5f28aeba43c4a4d9282bec4db26fcde1732b2157
119482 F20101221_AAAJKY habib_r_Page_319.jpg
f6bbb20b78213df4c79a6637ddb0fee5
e457b3a56b09d0f5a6d51f4e0dd3d31da66afe76
30885 F20101221_AAAKHR habib_r_Page_103.pro
941c1d96e8398b16ff48ec21460d72ff
ca99e6c0a2c605418478901ca14275a610955759
32859 F20101221_AAALEL habib_r_Page_114.QC.jpg
aa1a4eb1e835b85128179447e8ed032e
a8742dc82099769b2b110fa448634b0179fea1b1
121042 F20101221_AAAJKZ habib_r_Page_320.jpg
a55bc3a37fae86b29a91f786c20fe539
851f9ca4f637d288733b2aaa3097ff6fa21b0a9f
45349 F20101221_AAAKHS habib_r_Page_104.pro
3451511477bb7ffebc4a5398617d212a
24aad60e16272072bca8f2b6d48984cf0c32fdcb
40093 F20101221_AAALEM habib_r_Page_115.QC.jpg
24364be2456bc66983fac8db9a6c9f2f
c76be182b036571600040ae310f932e1e2c69c06
51342 F20101221_AAAKHT habib_r_Page_106.pro
b8817ea1a2958d1044165ee15234faa7
479becc5ba9beebc3b6d7736c9ab0389190692b1
9158 F20101221_AAALEN habib_r_Page_115thm.jpg
ba0e7c9cc3d875e015a53e02b0675818
064dc5c75c9dbde730f7d71ac061f299b838914b
2716 F20101221_AAAITA habib_r_Page_072.txt
5030fadd4f7dad83a7bccae92e3c0ccb
e5cf616d0eb3b5d431625439e8590ce137e56354
55629 F20101221_AAAKHU habib_r_Page_108.pro
12327c8b1cc257c5e7e323932921bcc0
eda6dad3d4d4615da4adb5c25fedc981e86b4414
33213 F20101221_AAALEO habib_r_Page_116.QC.jpg
d9a521f9b2c7cf039cad8489e9f5763f
58e2c0aceb30a81dee8c94c3a869e9f57fc9b673
19033 F20101221_AAAITB habib_r_Page_298.QC.jpg
a39b736ffa42ec0ea33533d508099144
740684607e8c1010aec8bb61c75d8d310ed3b296
49551 F20101221_AAAKHV habib_r_Page_109.pro
84efeb7ece6c1af4e3890542833d3ee3
36db159b911e0eec109c51919d74b2825f924097
F20101221_AAALEP habib_r_Page_116thm.jpg
c67beb0e52d5e6b793798a40e438e981
537221f5a2cf0ffd94b8cfc1aa5827f530c77555
9263 F20101221_AAAITC habib_r_Page_217thm.jpg
5fc3718eb1a6f427856ec42e058f5626
0829de84a2b977c699e838f2235ea8ca34e00dd1
41946 F20101221_AAAKHW habib_r_Page_110.pro
f5e8ffcae982f53e3be44bbe87323ac8
952f6d76d572011cf108b4d719933c5babf3af8a
29296 F20101221_AAALEQ habib_r_Page_117.QC.jpg
50578d7b2ca88d560ed3f8d02994a069
33b0acd59c575232b271065e35352bdf10147427
8500 F20101221_AAAITD habib_r_Page_076thm.jpg
a1b7b40c8e4d4821e277659fae08a383
e01c8342c3542859e71946c20ffc67a98a3543af
52420 F20101221_AAAKHX habib_r_Page_112.pro
d99716096c4e0f31c502ecfd7cab8bd4
390304bedf064ac7ac6b35953afd46371a0ec3f3
7990 F20101221_AAALER habib_r_Page_117thm.jpg
b9b1018a90e67c2fa2f1d34d8d13c4c6
025105d65b621165841b60c9cd5e4cf7b68061aa
537585 F20101221_AAAITE habib_r_Page_150.jp2
75f013f1fa9f958e93ea2fdf96f749cc
15d2e296a440f2f099e53a6ff7374585506c6003
52209 F20101221_AAAKHY habib_r_Page_114.pro
a496f04baa9927648f7b84445f27c7b7
979877d67e0f1ad2eeda07143da9fea1209f29a6
8305 F20101221_AAALES habib_r_Page_118thm.jpg
51793482019f2bb04907e035c66e40bd
70623b6d8c2e6e755fb21bdf4b69c33d568ccfe7
25756 F20101221_AAAITF habib_r_Page_064.pro
0153d5d50214e1818050bfee59708370
30424f85cca4fb0e2d98d42aecef1e42ba5dc80e
55404 F20101221_AAAKHZ habib_r_Page_115.pro
3d2b88da2c57a9e63908f6409c2c5ba7
15d49c0f13aba84fdaa8531dc3f47331d94584ec
33115 F20101221_AAALET habib_r_Page_119.QC.jpg
9782b30f100be20455fe4110e9b6e4b7
70b7838526abad771a44d0d848404dbae9500982
9216 F20101221_AAAITG habib_r_Page_073thm.jpg
a379b0a3d698040afc7c0ad95ed426b0
9eb540c6f3fa2e45e8d178d4d64f4cd25de140db
7738 F20101221_AAALEU habib_r_Page_120thm.jpg
d82542caab259ec8561c384ca5e45761
60a8bcb3d83e5f93694e88dfda4a3d72a794c2b6
111152 F20101221_AAAITH habib_r_Page_022.jpg
9878654731baeadf8a41e234de8dabf0
09d42d063fd70e8f95585785e8014dccf704306f
1051963 F20101221_AAAJQA habib_r_Page_160.jp2
97f9e83d28bba09825e3cf0bee33556b
d4dafd9c5e361df4937f8ceb6a5c1ec1988a190c
27233 F20101221_AAALEV habib_r_Page_121.QC.jpg
94c3817ae4ff59e5959f9345d075981d
f5158e0b2a870e433c583c14eeb4106915e033f5
F20101221_AAAITI habib_r_Page_270.tif
9c9ac70b3795897cafd2c992c6abd67b
bc8b6cfce09cab542c4651dc399c4ca195cb6926
F20101221_AAAJQB habib_r_Page_161.jp2
540401ef418a13a41bce7b19b7f26f43
9c2a58af4fed0009370d3ec5ba03d8fdcc73bddb
7037 F20101221_AAALEW habib_r_Page_121thm.jpg
026dd2d711813c4b9e5b73ace05900fc
d3fcdd9703d3631b30f8e63a4d0f36dd1ef7dcef
8743 F20101221_AAAITJ habib_r_Page_095thm.jpg
a090dfc64d8fd5399557b972bdc9cf0c
5fb343923ac82488b4cf202067f6903148f44b1a
F20101221_AAAJQC habib_r_Page_163.jp2
032eb5af3bcee8e3f9f34911596a9362
9212fbe78cacb78cd72ba4dd869cadeda2fe37cf
35608 F20101221_AAALEX habib_r_Page_122.QC.jpg
8dd7208ceff3990503312753e144e3ad
33e19fd91cb6fa2764f732b852054b88b01dcc44
35123 F20101221_AAAITK habib_r_Page_242.QC.jpg
bc6b45506871b50ca582b33d516ae0d6
217375d6e56f1c3744a8871bdab30db37db74100
F20101221_AAAJQD habib_r_Page_165.jp2
59213df68c5c380c31a88fe4c5e7dbb8
4964a0b1abd3fd4bfc16f1bd206a343568e913ee
32501 F20101221_AAALEY habib_r_Page_123.QC.jpg
1db31a1c9164f0e8ddfa0adbd444ee3a
c6787f0bc1d5a929d0a379cea6d2f86c9042f1ce
F20101221_AAAITL habib_r_Page_040.tif
65a654ceb7955600628dcd2be37c2314
6193f360b17f4d7203c419b5d59dd9f51296399c
F20101221_AAAJQE habib_r_Page_166.jp2
af5c8614d481c947f762c1c773b705e2
002502e7bdce18a1bcdfd421d443aef54817e359
37781 F20101221_AAAIGA habib_r_Page_163.QC.jpg
84483601bd46718a3d8b55e789e36d52
76ed05a64bec5471148013b896b592d7162b57e2
8602 F20101221_AAALEZ habib_r_Page_123thm.jpg
e67e5a4976315f5ca9a65583a4293da6
8ab5076e5c2d0c0bf5fd3f323fd4a16582258ef8
40677 F20101221_AAAITM habib_r_Page_008.QC.jpg
ed9471a7fb5e3c174ddbf203e634955f
89b8027347d0d4fbff23e71948953caea84e6b63
1051952 F20101221_AAAJQF habib_r_Page_167.jp2
b4d02024ecd216410c472da1088c0d45
14db1b6ac47543950e92be812d6d45a22ad8c857
F20101221_AAAIGB habib_r_Page_234.tif
5c2c70a82eb5c5ed1576846787ac5d14
5dbf7208ea47ae62b8ceaa0fdd524508e5072683
F20101221_AAAITN habib_r_Page_331.tif
83f14299b37ee64a513c9b3c9ff1d795
14639f94426a03b10f9eb35fc9c1c2ff1f80f4b7
F20101221_AAAJQG habib_r_Page_168.jp2
6862ce808999eb20e960c073e185f566
96ae07539ddcb59804cd59f361fb4d169037e014
1340 F20101221_AAAITO habib_r_Page_308.txt
c98da55ec553e990f9377bc2b0d8bd63
951c3becf8ee4267d83428e63e71a564869def0c
1051911 F20101221_AAAJQH habib_r_Page_169.jp2
1b24ed7913056cdf8e7500cff72006f1
6d1f4ef53180aaa1b10ae0c727aabc1d24a4380f
18700 F20101221_AAAKNA habib_r_Page_288.pro
db5c3bc5627123c39ca97ada48311d17
74c104089bbad1523ff8dd1b52c672d1abb6c294
F20101221_AAAIGC habib_r_Page_322.tif
ed8b4ece6516adf6c7c8730b786a2308
a6ea10d4b83bab4299849afe15637ca098a839e9
1051928 F20101221_AAAJQI habib_r_Page_170.jp2
eadd1c18b415503d5b0a6fbf185ff4f0
b4c84897a034f778cffd51d23dbd2d4ba75f65de
14748 F20101221_AAAKNB habib_r_Page_289.pro
e2bda2037893a54cba8418deb4a8a8f7
a939d24764e9f3e359e56aa4a5bbdf494698bf91
353 F20101221_AAAIGD habib_r_Page_279.txt
1311261a58c0efa35578443bfc2d3a98
cd11ffd2b151bd68d41bb6c09854c82b57ca4069
35457 F20101221_AAAITP habib_r_Page_063.QC.jpg
89443d98c113e48291c2c1ac3d0022c7
a3a64a817523cdeac961391cf474fdcec619c6bf
F20101221_AAAJQJ habib_r_Page_172.jp2
a31fe3bbfd9049d47a445dba78737601
7a720baae632e9a3d9b6bef67e31c6545f9ba1a0
30075 F20101221_AAAKNC habib_r_Page_291.pro
788966b6cedb437c3b6ad751f6e8f6e4
1d2b9dfdc391393cb37e196b86feda4b15d20d36
58847 F20101221_AAAIGE habib_r_Page_037.pro
2beaf938b31541f95128442dd88d47a2
30c88789f568e735bdb3dde3bf45b1f97753330a
F20101221_AAAITQ habib_r_Page_165.tif
9a335478dd91702954a2c17e18b0f324
8575e43ebf9e4185aaa60fa13cbc806e17f3a091
1044705 F20101221_AAAJQK habib_r_Page_173.jp2
d08f22661b977d1f72b4c9090e8c5d5e
a6b5a3d352a0b3b71b1efdb11d13fee2b463c56a
14362 F20101221_AAAKND habib_r_Page_293.pro
c6617caa59b9fbd918171210a438ca84
6a8a5331985938fa82bfb0ee8bd71a2f6036726f
14948 F20101221_AAAIGF habib_r_Page_017.QC.jpg
ad31fe063dfeace5a47ef638abb4d67a
60035e56466ce6b0817a6d9436bbc6dface74eb1
105736 F20101221_AAAITR habib_r_Page_230.jpg
156bc9deff6d68d3e833d764fa792559
60140cd88e1f0c59b335caa09d713c4ba92e1e3b
28326 F20101221_AAAKNE habib_r_Page_295.pro
2ac8b499eb0e053ea7c53ec32db3675a
704ed818b006328e8e07ad432c415bf25fc3dbe0
945 F20101221_AAAIGG habib_r_Page_302.txt
24a92d833930e6c763692e9c1f889da6
144a27d64aff29310c44665147b67908dbdbe5ac
110457 F20101221_AAAJDA habib_r_Page_039.jpg
7ecc5e2202c79826849a49dd9663c8b1
687b69c31345344d09697858967f05b81efdb205
11038 F20101221_AAAITS habib_r_Page_270.QC.jpg
1c1a6a9903fdfe807c1277267957da5a
34b56553b37d62b5a2f9119ae0a5c3b5e30a167e
F20101221_AAAJQL habib_r_Page_174.jp2
7de96eb06b76ffb48538bdc1c70e6b79
feace503e29f226ce56fbd3763673e06985f8d99
23143 F20101221_AAAKNF habib_r_Page_297.pro
ce02f174b4170390a626eb561f9169e8
13c5a142e0e4f6eace6fb7b3fde378e1c98126e0
972715 F20101221_AAAIGH habib_r_Page_090.jp2
1024d0810c4816de039f1e8da2f2efc1
6783f44db0ecf516e9941ac5ec1bfc8c424cd632
108536 F20101221_AAAJDB habib_r_Page_040.jpg
0990463616be03d3df0c4bbcde3e6510
b63de0a0ca0904d24372dc5918b47943930618ee
8676 F20101221_AAAITT habib_r_Page_325thm.jpg
e9f19176df1496b726c1b0ad572d17f2
23e559ea4aca2651b754d1ca65f671425a7e6dd1
F20101221_AAAJQM habib_r_Page_175.jp2
4f24391ea1b9cefdf5686696ba3088e6
b13e6f11c0c633f4cb4898e9760413ae507ba5e4
41464 F20101221_AAAKNG habib_r_Page_299.pro
5950bc05a08bf712bb7625d597521e40
98d874cb337712b9633780acebfbc0848439976f
84695 F20101221_AAAIGI habib_r_Page_141.jpg
feebc40019610d7843ca32a6331644ad
0605c410fe94ae99e9a9d31f6979e7e65ab63678
9232 F20101221_AAALKA habib_r_Page_218thm.jpg
360bc10ca0468e5dfedfde39b694c839
2c22ff222aef6292e30770fb27bcd36839560f2f
110109 F20101221_AAAJDC habib_r_Page_041.jpg
8815745d4851b6d4da2880fc8c0b2fc7
155fb9a18a903b67bd0ebf8e03f325128e623270
7267 F20101221_AAAITU habib_r_Page_282thm.jpg
e9eb275ac436884696ce78443defc69e
158493204c23337602410d7e0f80eee76d50cb2f
F20101221_AAAJQN habib_r_Page_176.jp2
3641141cb187dd34bc754d8f7705a952
8c28489800c1062730999651fbb0b93fb27dee04
1051951 F20101221_AAAIGJ habib_r_Page_044.jp2
d8f234a2b40508efaa4043c0b2ddde62
f39d3c8984f7f5b9831ef66f50345aa7e575c314
32593 F20101221_AAALKB habib_r_Page_219.QC.jpg
f661efa0713b9433a7d4dd4e2fe2b0ae
567bd67b91fb9ff28bc18b65ea7b3c1a2f205eb3
114759 F20101221_AAAJDD habib_r_Page_042.jpg
83b0ada8090a5b574ddce0780cbcfeb9
e6299af1aa224ddef0790fff986c9eecbd44a483
24463 F20101221_AAAITV habib_r_Page_313.pro
aa6b140f65cb13a67b6d14477f312731
340adfd63f017eb4a176f7bd67a51384f28b6a49
F20101221_AAAJQO habib_r_Page_177.jp2
e5b75c2d93184ff7d3c2d29a7fab8bfa
7bced62368c20eb37bee6614c545cd6ea7d2cae9
22832 F20101221_AAAKNH habib_r_Page_300.pro
b6664510f4699addfb2e1200e7d20c62
b13ad3187db3812223649049fe0491f296e7b4a1
2590 F20101221_AAAIGK habib_r_Page_324.txt
a77841370f19cbb78aac947512ba4191
ca26a6d65d55d1242412cf31eb56a22077adecd0
38698 F20101221_AAALKC habib_r_Page_221.QC.jpg
becabe9d43b6699918c5ccbe5c72f5b5
a5dd599d072c0597b8f55388e270805aa1e5ee37
115517 F20101221_AAAJDE habib_r_Page_043.jpg
368e674d9fe5e82c7c28255a05da8696
66c56432879591e88eb0da9efe18a95af31bd649
496862 F20101221_AAAITW habib_r_Page_287.jp2
59d26e6d52b77780bfc401b78bce70e3
478c789d33d014feb60b32f43684c7f0beba6769
F20101221_AAAJQP habib_r_Page_178.jp2
adf3a7126fd53230b75ff6b4b128c8d5
6b83138727f60c3d967a7144a1a83f93c70e250d
23157 F20101221_AAAKNI habib_r_Page_301.pro
dbd03a1dfd1d8bb892dbb6998e335cde
35017c21565c36ad81da12f82eb0f3a9934be235
F20101221_AAAIGL habib_r_Page_282.tif
4a505a3eb771262dc34ad0eb2e02bf16
8e1054f91dd11b422594ff5bb6108ff02a143014
110850 F20101221_AAAJDF habib_r_Page_044.jpg
3a621d3a048d33be486f61edf355f86e
2870a5b3c622a56779955d5f92e6c3719ae968eb
114528 F20101221_AAAITX habib_r_Page_048.jpg
f8fbdccb1686f2838485a61a7bc51589
a3662886886b790fdf94991ce71606c52e878a0b
1020972 F20101221_AAAJQQ habib_r_Page_179.jp2
206f69d6ab2ade52d95c911b4d94de05
7a0116e9299ae2f12a5e46cbdd866ef9b25d5777
26267 F20101221_AAAKNJ habib_r_Page_303.pro
d42c875ef4d7171e0b3a161b26bc5ddd
5ae1ff31eafaaf286b37bee676dc3ecde00e3dd1
15703 F20101221_AAAIGM habib_r_Page_279.jpg
22977674edc16e0a642240e0cbfc9e69
1955cba91137975cc620633a6b65969d76f3ded6
9156 F20101221_AAALKD habib_r_Page_221thm.jpg
209d522839f81270e62e40d8885f1a43
9bf010bd02f3b162e049d10c9a3441dd2bfc68e3
115556 F20101221_AAAJDG habib_r_Page_045.jpg
8f2db3daa58c8fe7b3e22455d95ebdc2
49fa05b02882627b8b05eba154be93305415b438
106770 F20101221_AAAITY habib_r_Page_166.jpg
a7183f00698f69ca43f86df8548710b9
431aa4573ba69b4be434fe0a1831b2ae879a4f23
1031048 F20101221_AAAJQR habib_r_Page_180.jp2
2fe36efbdb16f2f3f680d7231e41fcf5
d377335823cf4f80f2ede27dabc058a0616a23bf
30915 F20101221_AAAKNK habib_r_Page_304.pro
44b71a220edbe06364045a09a0736598
4f2e5366ae51cfbacd45cdcc69efb9fac1d2e5ba
F20101221_AAAIGN habib_r_Page_056.tif
3e70352927d1f9c717df197cde96635f
a7300911a3165d24532d10dc0ed1b4598b1fb8fa
8768 F20101221_AAALKE habib_r_Page_222thm.jpg
a9e9b97a8c0649cb6a1a9353cccabbc0
957cd9db70c5785870f14c86f7b41e3e62077fcd
111307 F20101221_AAAJDH habib_r_Page_046.jpg
e17ad4cfac91a1f125f8f210e4faa31c
26e8c299985d731048e5d4687aabaeccf0444ba0
298056 F20101221_AAAITZ habib_r_Page_293.jp2
808a7bffa45e7cab5bed59e995637e36
9429827ab7d74c29816a29d5f01a9c72acffd96d
F20101221_AAAKAA habib_r_Page_170.tif
6197054defeadc6fb5f259d01f6388f9
59b5ec178c9a49f0579076cf6b7a915ac9815b18
F20101221_AAAJQS habib_r_Page_181.jp2
2627fdd577928e144cf297ed2e9091dd
385f67938689aacfaaf3b35d91765fbc8462b568
22809 F20101221_AAAKNL habib_r_Page_307.pro
3ff901acac10746064c98a5cabc9bf1c
bcf959d376480cc2886d8e2f1bf78a86c6af338e
50873 F20101221_AAAIGO habib_r_Page_079.pro
3559b4f7a5934f5c903c9f4a021c2ab0
6b63d8d74c204b087569b70d4b733cdaf1272c89
25405 F20101221_AAALKF habib_r_Page_223.QC.jpg
a5cda566b531141d0a638500a3bee451
28c7aad32111bcd42b03e1bf947bb365401bfa97
112991 F20101221_AAAJDI habib_r_Page_047.jpg
761faf8fe7f11c5813cfd65d732d20e3
281406a1c9327f299f8538558253e16477d6b6fb
F20101221_AAAKAB habib_r_Page_172.tif
5cae82ab7035a46941bce0c61221829d
cde77ec96c51df2f6f223491a55cfd9f1d192fe2
1051964 F20101221_AAAJQT habib_r_Page_182.jp2
f593d48023feed85834e9cf3ef05e037
a37b8d97ef5274ec9316c4b8648a96a3637074f2
24950 F20101221_AAAKNM habib_r_Page_308.pro
e8d573822d6d0d061cb3a16e8853e288
448c5f29f045bcbb0b8ade86aedc503dc851b68d
8875 F20101221_AAAIGP habib_r_Page_068thm.jpg
0fd6121fcaf478368d06577abea95763
6b18037221305119a12b6efd49fecf25ba055b94
6204 F20101221_AAALKG habib_r_Page_223thm.jpg
38bc4a78f8578a16fbad7e08d163ac92
63554e42d64daf6181beb11147e301e5b993136d
117655 F20101221_AAAJDJ habib_r_Page_049.jpg
49b8f6a8f5cb9913b97433f6d878e877
64b590e3ab7be9bf6721ae95bca3b0f10b939ad3
F20101221_AAAKAC habib_r_Page_174.tif
895eb7f794671ece5cf88b5e0602355b
da2ff6e9b02c919c6f50d895752ecc5819cc913b
F20101221_AAAJQU habib_r_Page_183.jp2
821636d0dfef7b7d9415509ce816e531
d94921bf4e4fee4d46341ada5d0bbbb345f6ed0a
23437 F20101221_AAAKNN habib_r_Page_309.pro
2d169153f0465986aca566178e1214e8
df321461a9af87b94c6c87326f07f378bb3af2f9
1977 F20101221_AAAIGQ habib_r_Page_130.txt
24af4adaec1b4f476da6f1b4e318797f
cef25c93ccd99955892940f0b0b1a082b4e2b962
31972 F20101221_AAALKH habib_r_Page_224.QC.jpg
6f684c2d83ae85196a6f6a1a3ba573bf
e18d7417760a573d5b8b3ae89b3f9302e61c7ae5
53871 F20101221_AAAJDK habib_r_Page_050.jpg
52ae5e64838ab7ceab5cb1ea6d4b03fb
c6fe8b4b13d022155656d4e4c5c9da27f14f8969
F20101221_AAAKAD habib_r_Page_175.tif
1545aa3ded00fe082ce2163b4c324aa0
7d2efd46ad03d43b558a7fd932d1d0857fed87ab
F20101221_AAAJQV habib_r_Page_184.jp2
ff7e2e9401f919cde178fc4419f52817
4a09c429c9aafb562199dc572551f0718d4a77bb
19196 F20101221_AAAKNO habib_r_Page_310.pro
75c2dee9ef69acad050002083a8cc6c8
d31765bf02a1959d1e899641d56461d0b9abb99e
F20101221_AAAIGR habib_r_Page_188.tif
4d84c41509a82093e861cb28a6148aad
0896c1244b5b4edf4c511290cbcfebd95ed7abb7
36915 F20101221_AAALKI habib_r_Page_225.QC.jpg
25598afe2f17c20afaceef9258a38d3c
dfc6f27aabce260abf2aaad6c5f3a703e09b9a6e
93692 F20101221_AAAJDL habib_r_Page_051.jpg
5f7d3420ca56724473c45d64ecc47e86
3651b4a52136f8d13ef7bf839a6da949f1ec3402
F20101221_AAAKAE habib_r_Page_176.tif
e9090993090b6b6e50db6585dba8184b
a31fafb210d1bd19c9e5a2f9f6893b6df1a04fc0
F20101221_AAAJQW habib_r_Page_185.jp2
22dcb4cad5f69ab264e5489249468647
eb77c03f1cfe305d461bbf017008ab613ceecde3
30540 F20101221_AAAKNP habib_r_Page_311.pro
5af72cadc624e346a5faa149c60b5ee3
7cab6bbc154636fc705e00107150dca3dca6fdfd
8526 F20101221_AAAIGS habib_r_Page_224thm.jpg
1c9e49b9fdff7e5332aa5f6a0ede19ac
1d7475be38e8c7443922be008941a10a7e88cd87
9002 F20101221_AAALKJ habib_r_Page_225thm.jpg
45bb6c84511b2eb2bae724bbf317e31e
30bdd9e45100141db5355264251ae2cfb5f20e3f
114344 F20101221_AAAJDM habib_r_Page_052.jpg
84fccec45fe401862e9439d2e6b0f3da
a808ee96a97191017b290bd3a4ce800de99f43b7
F20101221_AAAKAF habib_r_Page_179.tif
732b7bb28481a9365485f1d166fb4bee
93a5674815d3aeb65de93a9aed73f1ab8dc5832b
F20101221_AAAJQX habib_r_Page_186.jp2
a55534b05eea78e0b8a03883619ad99c
93f401da00e3794559a6ee65803dcad73895814c
41383 F20101221_AAAKNQ habib_r_Page_312.pro
de23ccb981a490a5b3c0114c3a928fb2
b32a75d066b84f015118e06a93593f61271e0f82
F20101221_AAAIGT habib_r_Page_161.tif
6a96ad7c43d7e56f500e1e0ba0c108cb
d4e638800275709c279a18ed0164470181771cd5
33051 F20101221_AAALKK habib_r_Page_226.QC.jpg
d3397b485177249a0196da59d3351058
d1f3a67379bb37a1ac079c8ab795725cd057183a
F20101221_AAAKAG habib_r_Page_180.tif
612092ad838263836f50176ef69beb5e
73badf2cc4546ae02e8f51d87b45a822e6da154a
F20101221_AAAJQY habib_r_Page_187.jp2
72e682a6bd861846d778049211b48ccd
1cb9b6f84f0dacfe4a1d65ca83a09d90d3936bfa
34118 F20101221_AAAKNR habib_r_Page_315.pro
7ce09f7cad32358be86970edfa003223
9e270526def1c5bc22c4c09b961ccdcf5ec0d83e
F20101221_AAAIGU habib_r_Page_296.pro
97bed8ac99ed8ad67570ce1ba80e7ba3
c631aeeecc8e55d901594123d09130920f2f2578
110819 F20101221_AAAJDN habib_r_Page_053.jpg
a08b0e54d59c3a8bfe1e57716f0f198b
5e5f445b49b25054c9c02a63a2c47851644e11ff
8730 F20101221_AAALKL habib_r_Page_226thm.jpg
c46ab48231fcb9e047db877bfa157687
dad542ec753e4d1ccd43c4de6637f32677185d18
F20101221_AAAKAH habib_r_Page_181.tif
64a61a7267a77b2e1c7eac7df0cfb2c1
ad66caaee4d1a838ce734c055e6205e3079cbfbd
F20101221_AAAJQZ habib_r_Page_188.jp2
ea282b6a601f5d6b89fa99f10b43f207
ff516a68eb1e93ce213a0157d3cfb5706713d6bd
12765 F20101221_AAAKNS habib_r_Page_316.pro
d16babc56e826c9bd72431865db70bdb
2606f477ecd26f31a3aa8f458df55ad879399a26
2232 F20101221_AAAIGV habib_r_Page_029.txt
96c47155c896834e836f8436f2bc1b5d
d8ff41a620ac9b897e5a3ec90302ceb9670a463c
106911 F20101221_AAAJDO habib_r_Page_054.jpg
df767282a8d89121a5d0d6abc8c10054
5816acea3d726c59e7df0a9d8a57e51e0c83e30b
8736 F20101221_AAALKM habib_r_Page_227thm.jpg
85ecac9dd39c38d14f0796ef5cb8a235
654ae33850ef77ad7cbffed83c3a23adb058f27e
F20101221_AAAKAI habib_r_Page_182.tif
3cb9e03932ba0d9f6dfdc6a379cb9373
c10ade7cf37a74621f01f7a1c589eccf68715253
57981 F20101221_AAAKNT habib_r_Page_319.pro
d0ead4154ecc7bee9cdd3ba9e6971c4e
3399d01ddc04249fabec354ccf66af14ba87a623
F20101221_AAAIGW habib_r_Page_295.tif
f8a54f4c5083ef59361bc63f4c571496
57256809f1a89debb135b793b8287e6754153f50
116121 F20101221_AAAJDP habib_r_Page_055.jpg
6dfb02ce206bd4e4e78309aa22c23ffe
9eb68e4993afb31907fe2dab59d46f61202e98f3
34829 F20101221_AAALKN habib_r_Page_228.QC.jpg
cd3bda9b3d64820ccbfa7de1a7836ef4
e54701ad0b2dade1b7ebdb226cd67e2e57800d1e
F20101221_AAAKAJ habib_r_Page_183.tif
20db478c467e98beddbffb293e7f10a6
1ed62707b8bee599becfe68b9b7298cddb38c60b
57902 F20101221_AAAKNU habib_r_Page_320.pro
fb3ec36708ab74128426d4c62f6fcb96
82bd37c94339c73fdb14f32dea801cdfec954a3a
2103 F20101221_AAAIGX habib_r_Page_078.txt
e98a3d5ee210a07b2f21d8072768a4fa
5becefefa4ef671ee5f961277238792e57560ae3
33081 F20101221_AAAIZA habib_r_Page_095.QC.jpg
2bfd7e03d888084ed7d19f4ce4beedb2
993970aa13c94a041fa93dd61d2c2b5b783aac20
112673 F20101221_AAAJDQ habib_r_Page_056.jpg
d010e59a0befd2e70c08d965a37b78f1
a1274cfa8794d18b49f9b582c7372708040f226d
8944 F20101221_AAALKO habib_r_Page_228thm.jpg
5ad5eb8409d24d99b9c20505e7e103af
58f3a04632d9c75a34e6f5e15488c3cf2d765d89
F20101221_AAAKAK habib_r_Page_185.tif
712408e166aab5ffc71998ddc16d3e70
63a8bd0a2303e2e7d1504dc1f1d68ba24ae6809c
56501 F20101221_AAAKNV habib_r_Page_321.pro
1d1f5a88b76ef84a72c8d29a81ca508c
96402d3de741fa4be85eb59995ed0ac432f973b1
35416 F20101221_AAAIGY habib_r_Page_187.QC.jpg
2c61506de002f96702cbf22ab54b16a7
0e2e675dc0eb1e40f9810fdb377bd1ebd70b070d
F20101221_AAAIZB habib_r_Page_191.jp2
147249bc2abb9aa197e5eb0223fc833a
e60cee4465ba8b7a9017d89ab2ffc917f648dc73
114800 F20101221_AAAJDR habib_r_Page_057.jpg
59bc7c041978bc5dfc97dd3ae4a34a88
8c96c379a00419cc2667377871ce5e1023dd8860
35071 F20101221_AAALKP habib_r_Page_229.QC.jpg
e1cc704f865cdecf4e98f2a37c05bccb
17e27de6f1c4dbe9f7e3e9e9e428588ad79163f3
F20101221_AAAKAL habib_r_Page_187.tif
8ac0f9159184caa08765b3d3d6120ccb
e496480cde7a8c6b2953129c8351ada4ef3b929b
64397 F20101221_AAAKNW habib_r_Page_322.pro
399849c33df4de1bd8545b573778d13d
33212c339233a931037ae5a77521944faa2caf72
562324 F20101221_AAAIGZ habib_r_Page_317.jp2
3d11cca9870fb5c6e3a31c80720b6016
ec5013b433764359b1465799935ff7fbfa1fb312
23266 F20101221_AAAIZC habib_r_Page_210.QC.jpg
f5a0355658340d5d098bd592f6365ecd
70129cfb07f17f2d6c3b06acb23cc94f64d0a9fd
116980 F20101221_AAAJDS habib_r_Page_058.jpg
41189d9683fe52c3e7c5c9be8bee2469
6fcb32142b2ce967e550505270ea7f2a1c33d17d
8818 F20101221_AAALKQ habib_r_Page_229thm.jpg
d70e25143598b5ad308e044e7abd95ee
ab7e8acd1453f64471d2f10b701e196a616861b0
F20101221_AAAKAM habib_r_Page_191.tif
4a0f152165a6e3d545808e2f6b2057eb
12e4454cc91dea194884329fc558de07c564c6a2
63255 F20101221_AAAKNX habib_r_Page_324.pro
e1752d095bcda1418c845b0344a73bb4
51f9b482eb6cc3199a6d5c3f96fe00d16860952a
54505 F20101221_AAAIZD habib_r_Page_107.pro
52f353f03a382766c264fbb8d696dcb5
1222a5c0a20790876a93d278983805e523aed4bb
105144 F20101221_AAAJDT habib_r_Page_059.jpg
cebe4b79d0c97080ef27ce57a7a95afa
24ba55cef6e76488c5e31da56a5020e564fbe4ec
8630 F20101221_AAALKR habib_r_Page_230thm.jpg
818ea9fcdd7a73f56c544b98094c4aa4
8f805f91c489631c1e1bf0e6925f34d6951da608
F20101221_AAAKAN habib_r_Page_192.tif
7b687483d074c8714f7128a376fa3575
f8eb278b2bd45f1f96baee597eb1637a6665a381
54593 F20101221_AAAKNY habib_r_Page_327.pro
8f86465b11bfd3a0f1097461f588d294
bc46159b841a3dcb323d2e449c088932d4956383
44066 F20101221_AAAIZE habib_r_Page_102.pro
c68393fb7e2c33c8705d63ec9b1603f2
9a1299af0a0dee42c83e9256036bb71ca486d3c4
115194 F20101221_AAAJDU habib_r_Page_060.jpg
8a4fb1aed088adc6a87dc3e6f3504e9d
1d9a9f5f21eed159e5af3b82d4802397bf0d8574
38937 F20101221_AAALKS habib_r_Page_231.QC.jpg
15a047c9d6677058e53ff70f03304611
1e228b9832bd440563233d32cadffde9b98eff74
F20101221_AAAKAO habib_r_Page_193.tif
ea0060a18a26680a2956f279357f7889
6958ea01ec02d87dc8489fc52d83f957a380d2cc
59072 F20101221_AAAKNZ habib_r_Page_329.pro
dc574666da7e5e15262d21fbf73dc20f
000286c3948c2460eaaf222688b5b2f4d94d21ce
37793 F20101221_AAAIZF habib_r_Page_156.QC.jpg
f80ab65ba6f980c14631595762228673
9ced918ab31635e981ca5f520c4a66ce47c4126e
82529 F20101221_AAAJDV habib_r_Page_061.jpg
bd6195985560e2b5d1beab507876aecc
283a43fe42b5ebf0df3f0dcdec68858a1bed6211
8883 F20101221_AAALKT habib_r_Page_231thm.jpg
241971e198881e6d096992e64d3051a0
f5ca553e69e2bc7c09b8bb9813be680b58341dd3
F20101221_AAAKAP habib_r_Page_194.tif
bde3e8f1464e5b17cc378629d0ca0763
bf504c527a1dfcb7767a436a4b4a5ac816fba0a9
3595 F20101221_AAAIZG habib_r_Page_271thm.jpg
a79dda32575e67a0c318a180caf28183
266a0f34d98df735d3162f06a2a35bd66fed3b1d
111085 F20101221_AAAJDW habib_r_Page_063.jpg
866b4d8a2351d5e159dcc4ae9a454abe
b635ee84dee4922a0c27e4fe1cb3c9b2168b3756
30251 F20101221_AAALKU habib_r_Page_232.QC.jpg
7d0d3729f8437ade76b03b151011802d
3c904e5526a2e2cfb6779d3adf2f41b2ac4ca84b
F20101221_AAAJWA habib_r_Page_035.tif
d80cb38bb7803f9de0ca76b32d1f562a
591cc7bb478c5e37c4efb91374b87fb8fa150798
F20101221_AAAKAQ habib_r_Page_195.tif
945d7be6eaf764a18d6948096820d00f
59959089361aa8559b735524a053d9d4d9d195c2
434904 F20101221_AAAIZH habib_r_Page_017.jp2
2f7321d087fa2594b37079e44186e2b1
03587d489f8fd942e0cda7231792306374d5b240
55684 F20101221_AAAJDX habib_r_Page_064.jpg
03aa58c1d4bf41538547234aa22bb748
e6d99245f32b2566c55e28c1cafdab425bd4bd6c
32531 F20101221_AAALKV habib_r_Page_233.QC.jpg
a5bf8881ca06c0a1894428fb2d11f442
0d19b214bb80be5e05f8ad2fa6b7a63609f1ef46
F20101221_AAAJWB habib_r_Page_036.tif
5eaa41d0429cce5b9d0bf4873d4b2836
0248ecd0b5da1245ac125bb1ca730d8357388ff5
F20101221_AAAKAR habib_r_Page_197.tif
390f2b3a2296a5cb5c4467c9c3461d5b
8a566b54d46bfd4747869811bfe9285eefc82f6b
9131 F20101221_AAAIZI habib_r_Page_186thm.jpg
4e8af9a05b24ddabb50d9b76decd85b1
58ed4959f4000dc9a45de8d32c27c024f62729c0
111859 F20101221_AAAJDY habib_r_Page_066.jpg
b846bbf508c9edbcfa403f0934133a53
ce1bafbd22e2cbe068e9ab178ba46103c52f50ad
32591 F20101221_AAALKW habib_r_Page_234.QC.jpg
b92c54f32d79a50cc5345bc378c09e62
3c0949d5c211b5cee604d689cd0ad6892a378244
F20101221_AAAJWC habib_r_Page_038.tif
02d25ae1c8e8a6d22f804e78934874df
6d3cdf6d91f7609bead1e2dc5e803aa6cc22188d
F20101221_AAAKAS habib_r_Page_198.tif
80f5ada26df66b7071076062fa3067a9
782dda319ce45def469d5eab289b729bf8970976
F20101221_AAAIZJ habib_r_Page_202.tif
c94eb843d5b6283a3a8518784bc218a3
79199a128ea3549c3b9882b8ee2a63f6a76f33d3
106064 F20101221_AAAJDZ habib_r_Page_069.jpg
afd240f185a208fd0045582acdd3a831
b35e0456bbc81a542757dfa412e456dbe6e49221
8469 F20101221_AAALKX habib_r_Page_234thm.jpg
77022349a0814c23ac2ee83328232cf8
4f945306df6286081d1ebf40d1432fc5e60ef0d7
F20101221_AAAJWD habib_r_Page_039.tif
987dcaaf3e4041f2f469a6df178474cf
4fc45e65c7e748a75f31119e267109d36b8ebebc
F20101221_AAAKAT habib_r_Page_201.tif
5e07ef25aaac3ad09444266aed557785
8492d1846a724c24cecfd6718a59e46a790c6a45
12788 F20101221_AAAIZK habib_r_Page_261.QC.jpg
bbbc98de51446a8fbe1d8b9e9d84641e
fe4875ca0daf94d4fa8b68e42e27d1da76912298
35390 F20101221_AAALKY habib_r_Page_235.QC.jpg
12dbba30b02fa8f9bb77276ec46b167e
3cb7d2fde25b37ebfdcc7347dbb829e3dc112b96
F20101221_AAAJWE habib_r_Page_041.tif
46ffd4d1ef7f0d3edba7397a34c581ee
7f8aafb7d9b647a52086be04691b503738276e2a
F20101221_AAAKAU habib_r_Page_204.tif
bc041f69003402c6a734814980dbdbc2
513a124ccdcc63b735e2b6461b342f7e8c03ea88
46343 F20101221_AAAIMA habib_r_Page_067.jpg
5da750d7d77b1eb8f69c4fa7341f66fe
f22077319fd19ddcc2756aaeca606da3206024c4
118401 F20101221_AAAIZL habib_r_Page_034.jpg
b858495626d68646615707deac445446
3a51c1cf70d58854aeedc41a920c0bdc72e25f5f
36071 F20101221_AAALKZ habib_r_Page_236.QC.jpg
90a61a317c28b03b3ba51a76c7494428
e7c0f9260833533350e87846215cb9133eeb2e56
F20101221_AAAJWF habib_r_Page_042.tif
7b988220bc2531cbdea740f6bf3c76b2
9a9e57801551d29c801b6af7db643ab294801061
F20101221_AAAKAV habib_r_Page_205.tif
bf47d246ce38de7d80404b467b764a29
2c2a08c7b1daefa94300d05d9f381742aee4ce15
2165 F20101221_AAAIMB habib_r_Page_068.txt
b40770aeb2fb008f3125eb04bc3ce679
e8ac2ab613425f8adabaa7379bf6f7670b770436
456550 F20101221_AAAIZM habib_r_Page_296.jp2
6c272320645564467fcd6566391ef4c6
4a1d99af4dbb9f04b1f3c616dcbc819b1fee44a1
F20101221_AAAJWG habib_r_Page_044.tif
3c57edcc337d918e9df45cae223236e8
0322ad172c370aa9e39916ee5d9a705869c5fb2d
F20101221_AAAKAW habib_r_Page_207.tif
fd65b7ab706f386a15153915c60d9c3c
9c3f27224a136a93205f10a4d67c2aa038f020ac
7414 F20101221_AAAIMC habib_r_Page_194thm.jpg
b9d50e85e1770bb8f690bbf58804b277
9d337e8119cf45d12239aed5fb4c20b2b18c690d
31679 F20101221_AAAIZN habib_r_Page_283.jpg
777c2456bd9b11112f050927e5374f6d
605d54c04e114b6e094c35730ffd109efcda4098
1924 F20101221_AAAKTA habib_r_Page_173.txt
c2822b6d517945bb3b105346c196eff3
04447021e868b01867089d8a972e78987b2384bf
F20101221_AAAJWH habib_r_Page_045.tif
38ccfcc8604f3fc0f5402f1f986b1748
8ae692f9a71f354a3fc91476c6f23d0b539e1635
F20101221_AAAKAX habib_r_Page_208.tif
a63b0fbde0527b57db5d05721f490854
7a18123c13357cd581938b02fc9dc6ef67286ce7
18971 F20101221_AAAIMD habib_r_Page_281.QC.jpg
38db45835cb897fdd068c49b8cf6eb80
72da72660330892b05ea39fd30143cea51501925
788883 F20101221_AAAIZO habib_r_Page_223.jp2
cd9aaf8f7004c3be237db3fb9e0ada63
67e3674821c34af529717f7201e8637452937d2c
1888 F20101221_AAAKTB habib_r_Page_174.txt
ee7cfc47ae3857f9e70ddb562fb19b47
173d7fe89b1bfbc029f9d4b4a85ba7fd735514ce
F20101221_AAAJWI habib_r_Page_046.tif
269bdc68ffa1197bc02ef3962462b2c3
2116df0d6fc12e9e11fc51cb1d0f0047a122ecec
F20101221_AAAKAY habib_r_Page_210.tif
001b2a35b317a2d5786f0717f74fba4a
07fe0436d1524c1bf6f40dc51a45ee5eae7d39da
15547 F20101221_AAAIME habib_r_Page_276.pro
161b04a8eba5b838cc3e97520310e704
4808133eb2f8cc6344ece3ef84531e1d1b125d38
58306 F20101221_AAAIZP habib_r_Page_286.jpg
6dda559b6044e2baf9d1096da837e441
34cd218c1ce83ffbc6f6793aa10e28f4d5cbaf2a
2172 F20101221_AAAKTC habib_r_Page_175.txt
1f1ca1daa92d6dd479ee60d32dc185ce
a2d125df57ef3b251f62033ad885c77cdea5f9de
F20101221_AAAJWJ habib_r_Page_047.tif
f0603ef7d1c04261c217e55724a26b5f
e89aada93ab133ead5318482f5e3147d2fe467c5
F20101221_AAAKAZ habib_r_Page_211.tif
307690a5859ed44cd7c32b379c36ad11
25ad5d7ce5aa3a19afc69096eda0cc5501a96826
2265 F20101221_AAAIMF habib_r_Page_060.txt
327f126ed786c8ad64bc5a96495e263e
45aea17ba32ca93ac1f4f4f38aeece99d9a9a44b
64176 F20101221_AAAIZQ habib_r_Page_045.pro
3dc0c63b2a21d5f26aac355a7e5a8035
fcf46cc398bce16b80af0b748f266d5af525ed56
2184 F20101221_AAAKTD habib_r_Page_176.txt
8fa71d273008966ab3c4bece4202488d
8f0a8225d2e2f78d412f42678d8b09555f187a83
F20101221_AAAJWK habib_r_Page_048.tif
4a3a396c6557bb69a0a108fc09c3896d
e4649dccdc6f7fc2c6bb47e50ee939ddb488e300
65257 F20101221_AAAIMG habib_r_Page_003.jp2
606e160deb4a18324eff1537215621d4
1481e5ee85891ebef73c9557f6b3a020d96e7567
4095 F20101221_AAAIZR habib_r_Page_008.txt
e92d65ce87e74b62040c6399aeee95d6
fb3e6c14896f0a50a922297aa209832d585a5445
1904 F20101221_AAAKTE habib_r_Page_177.txt
e7dcdaad3d2b3c012d320519cefb6bea
b49a65e59aad951f18b1c03d206c07ccde08b618
F20101221_AAAJWL habib_r_Page_049.tif
9db1b2359da1cc4cf13df28a8fcd5388
800b0ecf6fe1e35f4ed2c5e007da60ad28184b20
231005 F20101221_AAAIMH habib_r_Page_269.jp2
d7f07872186efb8a2a7d8a1cba087b2b
4590dbc68f33566771b2e963e5bf34011997ce0a
27920 F20101221_AAAJJA habib_r_Page_254.jpg
944638e72c2637ee37d806aa5f382dad
a7f07ae661ab7a7929a626b0e8475e4b8b037ca5
8222 F20101221_AAAIZS habib_r_Page_167thm.jpg
c4e6a80fd8ccae98e52ffdc038dca709
c427053441f974be3dc0dffc3c9ebcdc989a202b
2345 F20101221_AAAKTF habib_r_Page_178.txt
a6f770197c7c2c8e12d085eb2a9df409
d3b74d91d7de7824c87209a5fe094ca66c88938e
F20101221_AAAJWM habib_r_Page_050.tif
42315a46e2b86090b59bdb946b8dc06e
a4b978b9285de17a20157e9ec5257077d10c91b9
27978 F20101221_AAAJJB habib_r_Page_255.jpg
195a74410f235eeaa34b26a3004db856
7748cb9583a2b9cd5ff03284a92daf0c9910583b
396873 F20101221_AAAIZT habib_r_Page_249.jp2
8c0bca48d60ef854faf21748412ce720
0a215761c16a20df64efe2f12899242efc9214c5
1789 F20101221_AAAKTG habib_r_Page_179.txt
1ae6c0c6bb4ff251c0d478f3303ddd92
8e612c18d81493d396c43e8ca53c169d7fd00dd8
F20101221_AAAJWN habib_r_Page_051.tif
de291499365ddb6d6da82dbc44ec8cfc
aeb59604c80f4126fb99cf9898e66af5be99fd3b
6676 F20101221_AAAIMI habib_r_Page_157thm.jpg
16861d58dbcb86415005376b6be611e0
180a5a503ca1efd7645e15415e2bad672ac4f56f
45124 F20101221_AAAJJC habib_r_Page_259.jpg
6229c76c40ec56620eb1d6208f6f37ac
dc918b4d5739ebe8a8958016d97b18546e61dd22
33895 F20101221_AAAIZU habib_r_Page_071.QC.jpg
6fae6d22329bafb267824022edb0b545
39a0c017e5b817d32a568339a9bf418353760adf
34617 F20101221_AAALQA habib_r_Page_328.QC.jpg
921bea14a940197da64ad50c1b2bd2b2
2a6f9852d73983c893cec61016e8258abd574817
1772 F20101221_AAAKTH habib_r_Page_180.txt
b56dc4cdb3b834ba1f6f1815b7f08bda
b4913a9c1cfb2d5da02c903a67510cf1c0e1656b
F20101221_AAAJWO habib_r_Page_054.tif
9cfcfe57910654c1b4ad822de654f6d1
1704a9175c1c3dbb0befcac1d744d13c51181519
30569 F20101221_AAAIMJ habib_r_Page_090.QC.jpg
c3f51ff223f3d4cae39e6d2d141c5c1d
bed9657df92008de8a882fceb2bef460b1802f6e
36236 F20101221_AAAJJD habib_r_Page_260.jpg
3a3b9f125feb7ac42b4b6702c9352d78
0e8456fff3b89ca169496635f29739dce355eded
8745 F20101221_AAALQB habib_r_Page_328thm.jpg
3d2be51e671a3ff6a263ac183a016562
946ec339bdb6e5de89b7d963451cefc92b184c24
1940 F20101221_AAAKTI habib_r_Page_181.txt
d79fdeab5af269780c4f1eb919ca0dc2
080025b9d32d52b48024b7b8b4f3848e61737b65
F20101221_AAAJWP habib_r_Page_055.tif
a1b15a1abdee7c436070087660ae6143
b0304a466647b710c00bf79512599dbba879b7ac
58054 F20101221_AAAIMK habib_r_Page_215.pro
3f0309614b9b8f3ddcd503d56705c68c
ffdf69e01c09e0824b5989ecaf9107a9b4fed936
5296 F20101221_AAAIZV habib_r_Page_086thm.jpg
1794fe155e10484bfdbecb20d4485cba
0ada99534a7429fd8c8695a40bf04235f9550baa
35521 F20101221_AAALQC habib_r_Page_329.QC.jpg
95df8907e33ee49860b7ee7777a8ec51
cf24f46c4f08c4d039248333ab9b48f82342beb2
2042 F20101221_AAAKTJ habib_r_Page_183.txt
0eb51fe22138c0da2468d89c3691c167
8e317087e28ea4b0cd6f3e5b081ee02c869a6386
F20101221_AAAJWQ habib_r_Page_057.tif
1716983feeb4e63ca4a9b714d35cf1b6
92716f6b98d9e3920228e597076576e6911c7195
58317 F20101221_AAAIML habib_r_Page_105.pro
b342947e90dd0179ca9a3b6c97cc38c7
737dcc4474fb88e164d1079b4daef870f205e5dd
35234 F20101221_AAAJJE habib_r_Page_261.jpg
aefd9ec84c551786454664db62516732
7fe01adb1f40fb538db5639e8cb580ffde6bbf3b
F20101221_AAAIZW habib_r_Page_008.tif
a64ea42156326ff4adb600bcdf15b26d
e9aefbc1c401aa25186453a32905c3db45f42525
9000 F20101221_AAALQD habib_r_Page_329thm.jpg
1db98f3664a89c827ab613e5750cc1d9
b683ab9b3fd921c306054ab1b961abb95c3d8950
2090 F20101221_AAAKTK habib_r_Page_184.txt
39de2806c4c0213d314eae98baa7dca3
243854e2ba6b6f4a16155de00876dddae83aabb6
F20101221_AAAIMM habib_r_Page_222.txt
b075202b6de8e34ff23db2cd6fc7b225
3e7b1604c5959bfadf7c56af8e789dc68cae0fde
25893 F20101221_AAAJJF habib_r_Page_263.jpg
84832e9843adee9bf91d5ba5c71c06ce
ec0bc6ba422774ca56bacf6828d6e4693347071d
5644 F20101221_AAAIZX habib_r_Page_272thm.jpg
60c94100b951f0d0490089f6927325de
e571192a5fe1da9ec518de484b701b3334ee8c17
36507 F20101221_AAALQE habib_r_Page_330.QC.jpg
8e40b5d80b0faa52772ef4bf73d810a5
5cf4e178ab598e57e2cb7d97d1f414f4b1a5dee7
2081 F20101221_AAAKTL habib_r_Page_185.txt
f4a28a3874cf4ef66781fa86ca19aed0
a5e2bad5fd29a2efe2754ad391c091adb7dd95ed
F20101221_AAAJWR habib_r_Page_058.tif
cfc0fae9a5390a7495487d0fe6821314
efd1504deb0b08f07a8405149dd2b2742971c8b6
F20101221_AAAIMN habib_r_Page_202.txt
d98fc293c8b3cdbd2f5d5dc80da4889a
5254af7e6d7505c2e8eaacc16c86f8aecb82d654
33023 F20101221_AAAJJG habib_r_Page_264.jpg
8add5e05a51c95f0cf071ad695834262
d1ca28db9390266a4aa8679b5d5e98455bd1b9f2
451389 F20101221_AAAIZY habib_r_Page_274.jp2
cdd86e3e14b63fc9350fdaca2493ecc9
a056f5079d17ab56db7d9910577af1b12acf5dcb
29922 F20101221_AAALQF habib_r_Page_332.QC.jpg
19dee33c325c55fd3789763d7e65b460
1e23b30c57a7160571050ed51d827e0058eeded8
2152 F20101221_AAAKTM habib_r_Page_187.txt
5d60c2a0bc7b434f2799a5661cf6a7c7
7aff4e5c2d0714addfa32bf8a94dca4e198bf023
59058 F20101221_AAAKGA habib_r_Page_049.pro
b29579228c004b9109703c8b949ffb5a
38cf9d3dc902888835d901eb79563a6339479f7e
F20101221_AAAJWS habib_r_Page_060.tif
7f755032bf1e524c32ce1b40b6ed8c3c
64c91d8fcc6144f8f65138191d85dd4364f51987
51951 F20101221_AAAIMO habib_r_Page_169.pro
d420f4aaabc279749d96192708d669f0
76a4a74d990f384e2aeb7f51bd78ec581b2a1ec3
45856 F20101221_AAAJJH habib_r_Page_265.jpg
af4c7b8eacb8c441c0a3ce14b656a48b
5b85d6183f99d11463d9e9852f9fc0b338213e60
9356 F20101221_AAAIZZ habib_r_Page_057thm.jpg
7a216f321a36a0e72db6bb432d1f8f06
1dcc9841ba04c2591eca2a1c2eb521d4a181183a
7303 F20101221_AAALQG habib_r_Page_332thm.jpg
791afc46bac098db4c0e2ef15b2d7ec5
1230fe8e40f7eba9100280d17d6754ecdfca40cb
23913 F20101221_AAAKGB habib_r_Page_050.pro
180fa7a8a6a8f58719b06ae520e2c981
ef5384d45a521f7332bb4382eb340d1d4776d3e9
F20101221_AAAJWT habib_r_Page_063.tif
28187a154178aafd5d4ec264c901e464
225803fe3062d684180b20f1c3504be7e9dfe09c
8436 F20101221_AAAIMP habib_r_Page_166thm.jpg
148ac0068bbde58bf9925f740a0f44e3
25ed1bc5eb857e140f257008b7dc777ca4a59237
36049 F20101221_AAAJJI habib_r_Page_267.jpg
5913e133dc044773a38693be2840eaad
3dffa272577842af250280c761ba7f0de0dce84e
37154 F20101221_AAALQH habib_r_Page_333.QC.jpg
2a6e6742beb61f574174df80da56fa3f
6d7533ea2f4f144d0990d88e0cbb673d2e77d80f
2086 F20101221_AAAKTN habib_r_Page_188.txt
42b17755ef5683e67cbaa868d3e8c65e
5baf908bafc7d5275e8fedc2349700e9af12d096
45084 F20101221_AAAKGC habib_r_Page_051.pro
11e7672fc60fbc55b062126eb847707e
f74194ecc8cda3d83f3f751df7a5dc1bb6741adf
F20101221_AAAJWU habib_r_Page_064.tif
d7516a79580377a7bdcf584a46ea3cd5
ab1e4aa142ae8fffd282a05295ddc57417e30849
1272 F20101221_AAAIMQ habib_r_Page_287.txt
a24293547ff6a1074b7f3a1c89a0acbc
847b068717a5d40beff586792eaa8d211aac9055
28821 F20101221_AAAJJJ habib_r_Page_269.jpg
fc5d047c1522bcbef79321ee8fc4956c
939a72e64317007cc91e1b8bf9c309993b4089d3
8732 F20101221_AAALQI habib_r_Page_333thm.jpg
4b49f2bb66a7127a6946104882673ee2
c5d2123c84e62bd486f5959389944a1b8d367c89
2146 F20101221_AAAKTO habib_r_Page_189.txt
d0ab8eb2ef38d173a4cfbe4d4a31c37b
426353cc2e7827d3eca9a708ce82f092148e17c0
57711 F20101221_AAAKGD habib_r_Page_052.pro
8dceef152c08ecc7c23afe2b44d466e1
4d37007c5fd09fa66f5157cd1b0776f7ef5572c5
F20101221_AAAJWV habib_r_Page_066.tif
412d0ff4682526a308cee63fd27e9c70
a26c60019125c9d1476776fef0ff9af18894d0cd
F20101221_AAAIMR habib_r_Page_146.tif
7d8752300a0c9207b2eac137c0efb4ad
ec8d7c61da3ba5333c80a3f454ca0cd6ee542355
25147 F20101221_AAAJJK habib_r_Page_271.jpg
a62f656e035bc7d7b9ee94b8f9694de2
6e366cadfee55c0c5907ee8fefd7d70216e71e4e
F20101221_AAAKTP habib_r_Page_190.txt
499ac444376d351c8baa1220c775d182
f998fb2ee007cea88bdbbcf4aad4ac4819b44a02
54392 F20101221_AAAKGE habib_r_Page_053.pro
bf96e6203f9850051c0c3be0f3e44a43
60ec562fe7b12b682755ad53db58c91536594d3c
F20101221_AAAJWW habib_r_Page_067.tif
bfdd05ffcc6c3a3dcf1380d2b5f38ad9
35c6a006f6ba41a3096995db8ea9c641b3752896
15876 F20101221_AAAIMS habib_r_Page_302.pro
b8a92d623175799fc72d72c45e9bd8ac
33ec11e6c5b26c61e4cc560c29309a177febb0e5
51772 F20101221_AAAJJL habib_r_Page_272.jpg
e389553306e9efd6a496035ecca065da
46f277c3ef2673ee628f8526c65c0e3b5ae89798
37057 F20101221_AAALQJ habib_r_Page_334.QC.jpg
56f5b08ca11caa8f1b237f059613a26c
ee16a349eea1954ac38717ca641ef84ec7fddfd1
2209 F20101221_AAAKTQ habib_r_Page_191.txt
200d8b9abf408966426af0a656b0578a
a38f1968aa2527697f17ee665930f3208935d947
52489 F20101221_AAAKGF habib_r_Page_054.pro
1c80314be846f9c8cbd92e57b5b8591c
e8a8235f0ad8ec36c79b7aca38913b13ae0c93c3
F20101221_AAAJWX habib_r_Page_068.tif
db30262c2dab380b455155a35ca8ddbb
50140951e68520779bb1fb967af6ce04eb6b2d68
25787 F20101221_AAAIMT habib_r_Page_315.QC.jpg
76204f652bd369f7db85b2fb00e7bd2a
b1c11e5a702e130cbffafc3ead53c117e3ff1f91
37968 F20101221_AAAJJM habib_r_Page_273.jpg
1e1a95c9f36b5b8dd6919be31b110512
d0ebe6fd67bb2869d09ecc83ec2d066634147227
9201 F20101221_AAALQK habib_r_Page_334thm.jpg
e080d54af8fa8028e00e3f8830cd5675
57f1693ec48f01f67995db31d50604dbe7e521c1
2200 F20101221_AAAKTR habib_r_Page_192.txt
4fddfe4c039766a02ce26c4732cc9b9b
1d3241efb58687c02fdb43288a0c4f8617524a6f
55959 F20101221_AAAKGG habib_r_Page_056.pro
70d6072b012ecfa8bee7d1c79c31a63a
b5b712f32c7538eb10f88f9e871796ea8bae648c
F20101221_AAAJWY habib_r_Page_070.tif
32d65de0d6e88d15496d8359b1726cfb
3d1aaf876917c0667bf5fe734e86ccd4815eb41f
87 F20101221_AAAIMU habib_r_Page_002.txt
bdf39e7bf12b2739771fe2021d684e93
aa03d7dba47f5b170598369cf30e00801b82debe
47185 F20101221_AAAJJN habib_r_Page_274.jpg
feb22aeb419083b10b6a69220841a374
349636e696d4b039a4604ebdb1083e48b1c8a66d
26524 F20101221_AAALQL habib_r_Page_335.QC.jpg
caad78f4497dce5fcf35750ad5abc4a1
758c91d4fee389bdf01a5f6885a452c4381c262b
31733 F20101221_AAALDA habib_r_Page_089.QC.jpg
13d65f3dd73b3e7c83c4f98a5b84a43a
04d16b13b4893a65d167a3652d52bc9c24f0d486
856 F20101221_AAAKTS habib_r_Page_193.txt
d2c550d09cc7bcb2952ff114489b0dd0
1d024eb2d3f87742c64a4da6cca0cbdfd0bffc30
57564 F20101221_AAAKGH habib_r_Page_057.pro
948fd0380f6d2e42f804b56b25bd26f2
5746c29cb8b100cdcf973b87f36e69bfb06a75b6
F20101221_AAAJWZ habib_r_Page_071.tif
99895610d58d076876953e00fa3d7439
03de1a5217d95046a0bd4529924d148b6c5ff94b
23598 F20101221_AAAIMV habib_r_Page_282.QC.jpg
03d474d8f08be958efe8305b658cade4
717c5a93c4ddf123d5407c6dcb206e3be77acabd
42051 F20101221_AAAJJO habib_r_Page_275.jpg
6d90c734d0c8e0127e243bff3e28993d
99c0d7d613ae9d673914ee32b39927425487bee9
381810 F20101221_AAALQM UFE0022375_00001.mets FULL
dd5385a316108d280ba724e6fb3984c2
3f86d287272e6589123cc8b7430d0ba3e9235198
8685 F20101221_AAALDB habib_r_Page_089thm.jpg
16a2d1b4fbcc3ca5a4cf348521073b23
94c80329a0382a898609135006a376e4584dc89d
1436 F20101221_AAAKTT habib_r_Page_194.txt
d18a731ce0b5b4698fcd1e4905f2579a
a8a0d51ac9312ea9c3e4c8331ff573ae2cb52520
58945 F20101221_AAAKGI habib_r_Page_058.pro
99128a80d00b084a28eb58ffc546f1ea
bebdddbb2f530632c53305dc41b4ed59ae567150
839187 F20101221_AAAIMW habib_r_Page_194.jp2
4a6986e96c0b999ee7d1553d8da6be76
faaedfebc66f1fecae1699bcddaef4c98db34855
37113 F20101221_AAAJJP habib_r_Page_276.jpg
103a495ffcbd334ab96cf043db80f9ac
5807e3d411c32f2d4cc2d408b91d169ce5f51a38
7605 F20101221_AAALDC habib_r_Page_090thm.jpg
593bd599fc3a07d419981fcc1585cac2
66f1c9fffc58a45a0939000560ea65804931b52b
780 F20101221_AAAKTU habib_r_Page_197.txt
03d05f8508a480d1e0a4c322711c6093
49e59ac5ff9c16ae388e7f5e9988a828ba36a942
52089 F20101221_AAAKGJ habib_r_Page_059.pro
0857b865056bdd0afbe398fe90357783
10a8d08f66d527b63aedf202e7388678e0790c2a
7671 F20101221_AAAIMX habib_r_Page_216thm.jpg
c8a4222923897a5d40d5d6d45b9c26d3
da1c232e5eebd459bc4ca6880896a086dcf52218
27662 F20101221_AAAJJQ habib_r_Page_277.jpg
697c3ed496dc17d391cfa9f569bbd07a
7057fccd94c56403f6173837618094174149a81a
30143 F20101221_AAALDD habib_r_Page_091.QC.jpg
7af8a50e4d505a6152bb85f7227d3910
265103c35fdcfd2714a6ebc4e557fb1ee329e0e2
692 F20101221_AAAKTV habib_r_Page_199.txt
5f20e20d5ab6114fb80b3a31c37a5d5e
bfbb42d382e10f39af5b295476f149906318d85d
57904 F20101221_AAAKGK habib_r_Page_060.pro
bf8ab232f5b54c8646790bade7ba331f
0db8e31f825b6ac5ca02a5e411d49c0932498706
F20101221_AAAIMY habib_r_Page_038.jp2
c2b305025e88b8c518806ce534dcb2c9
8d30b708cc29aa54430ddf5146a8cc72121c6160
28579 F20101221_AAAJJR habib_r_Page_278.jpg
fa6e0034eed60b5ab66d207779c48b7b
22732798a4e1615e8a8ecbe807521ac0b4680e5f
8470 F20101221_AAALDE habib_r_Page_091thm.jpg
dcb3ceead2b578f9af88463d92df47d2
37ce7a0c92b7181b2158720f959fc9017c97b5b0
F20101221_AAAKTW habib_r_Page_200.txt
68daa32c8bd4444f6ef76d8518bed9b5
9c7ec9f62212eb71cb167df9da340490bfca569f
41118 F20101221_AAAKGL habib_r_Page_061.pro
6d86603594fa626cc69a9f5a5e42d4fd
8c492d25c5cf50f6938680b1a0ec32a61b083d81
8857 F20101221_AAAIMZ habib_r_Page_132thm.jpg
13062b40d4a1992284b2b214a581b381
bb7d7a0629fbe664a00835e1c36c649630a00053
48386 F20101221_AAAJJS habib_r_Page_280.jpg
cfb9db503cd3f665136a7b1b393d9ada
1e8dc2009fb268c075fa4bbb38d415484ed58bb0
8277 F20101221_AAALDF habib_r_Page_092thm.jpg
1183e43d4608050182266e70d176dfa1
ef02ddbb4fb1a8c3ec187fd05474f8bd382f13c8
2518 F20101221_AAAKTX habib_r_Page_201.txt
486c021e351faba0ad69f3b26fb30a5c
de8e72ad8b5fa3754edd28158e0eadb4cced1939
56171 F20101221_AAAKGM habib_r_Page_062.pro
45f534f3d160558be4c1cd0832df1470
74cc1d4a099491fda243a26aeb71c0857c1922f1
54716 F20101221_AAAJJT habib_r_Page_281.jpg
f74101ccd1468dd8a4b75c83b59e5d1d
2a25a200754cd42545543971d8245992f3c124a5
27266 F20101221_AAALDG habib_r_Page_094.QC.jpg
e49ed1027efcad7e6bc326e9ca4c4cc8
66cb86b116a59198d21b52e6b4b878877bf06930
1886 F20101221_AAAKTY habib_r_Page_204.txt
129669b738d307f94d070d37af45795f
a877cdef7be5a19141440f902fe712cc664d8b26
54685 F20101221_AAAKGN habib_r_Page_063.pro
38edb69b6a8d2dfa2e82c3cf1f12a5bf
81cc2a07bfc5fff2acc86ec01eba13919b5b1932
68480 F20101221_AAAJJU habib_r_Page_282.jpg
8cbc9e3824f3e7f6d60792b165360eeb
e123c31eb67380ed30e946b5e559d0a4a75c05c7
8115 F20101221_AAALDH habib_r_Page_094thm.jpg
289ba9126ea36e6d3a16cad28b3e0b27
b70533bd1abfe6f044d8e8fca702ae0e27837987
579 F20101221_AAAKTZ habib_r_Page_205.txt
59449d59ccb12ff7a6ea55e37988f4e3
b23ec7ade70e538fcef9a11accb056385c88bf3a
54384 F20101221_AAAKGO habib_r_Page_065.pro
d745e715613122ceaf97ceda6b3b5729
038218ec25d9a1c00cb1f712e362d412f66c5a76
49959 F20101221_AAAJJV habib_r_Page_284.jpg
8760e91705e7ce8d28fa1bbbb2e79b2c
9dcbce90b51a727a05acd50b6a6923de18214e6f
9043 F20101221_AAALDI habib_r_Page_096thm.jpg
b47c531f927318222108c9bd572b8d56
adca4e0eb507a393b7327445acc43d94cf6f9d47
48361 F20101221_AAAJJW habib_r_Page_285.jpg
a43cf7e6bd2357ef8660393960fb70a3
78eb2707f4a360ce0efb6894b1fcd48219e4f598
55197 F20101221_AAAKGP habib_r_Page_066.pro
341da9a7d9c52dc3581ac6843e91a15e
c9c596704f23e6d739b2c1f69799d0796ed87203
8989 F20101221_AAALDJ habib_r_Page_097thm.jpg
9e742d6760787d96db2611d9f92186c7
47365a07c439d13576d270907821eec0a7ceb734
53536 F20101221_AAAJJX habib_r_Page_287.jpg
b3f22692f7c3ee5e16b8f37633732cd3
1c1a6be7964b2ce5c340fd0c1763cc0757d9bdc0
16182 F20101221_AAAKGQ habib_r_Page_067.pro
a3e4e94bc2bb126d00410d35e570e4c1
0fc7d6b32c3f4d3fc09dd7a211f9c7c4bdf004e8
27637 F20101221_AAALDK habib_r_Page_098.QC.jpg
505680908d731e0ba9118b60d30e02bd
ad10063bc3f57416fa22a979057b0913adf649ff
43119 F20101221_AAAJJY habib_r_Page_288.jpg
150f4cdb8645db521a6270d7d9851338
643af2c1f7ef9401b55d495705d991ffa0a2b937
55002 F20101221_AAAKGR habib_r_Page_068.pro
9269f3620ba95800928c3cc97f8c2fa5
8cc02654ffa4ca51933a1f98b1bf07ab8298c970
32877 F20101221_AAALDL habib_r_Page_099.QC.jpg
5f80fd470e8aa1f96b725c3448c6a0e2
2d9afff5df48c8ce8c69e4d31e219bdad12cc115
35816 F20101221_AAAJJZ habib_r_Page_289.jpg
858ac45f312a5fd79576a7c7ae8d8bc7
d5e0066d427535221b094f06d3607e56496eb3eb
52983 F20101221_AAAKGS habib_r_Page_069.pro
c52f8655bea1f1f7fbda551dab7c7c44
9374b1d84d24b97cb85b727dcf0c39ecfd8b6d7d
33454 F20101221_AAALDM habib_r_Page_100.QC.jpg
6704d4125d5d1d47d9ed74913042d360
fb2f798c4c2e996e4ed98c1c2c900db0c44769e6
57311 F20101221_AAAKGT habib_r_Page_070.pro
6ae5d42079d569f42d7a9eff211d3cf8
56e537c960ca04ed2b98c565cc645923698deda5
8367 F20101221_AAALDN habib_r_Page_100thm.jpg
d06fc7e64ace9c75e891f5c3e5c8a229
92b2586ce5d7571c2514fa9ce26a3e63db579b5c
1417 F20101221_AAAISA habib_r_Page_223.txt
47914d2d18165482392f423ac459cc79
6adbe13a467b58963e11bb3ce25dd2b3354fb09f
52366 F20101221_AAAKGU habib_r_Page_071.pro
061d284b42da8e56e894a2f2c1cb09db
41f319cbe8a90b0118b2344f12e9f42e1204563e
35920 F20101221_AAALDO habib_r_Page_101.QC.jpg
df4e77ebabd1deea509ea5596d4330da
c6f2c9d6a510597e3e77f8280283f6fe373db2f8
89218 F20101221_AAAISB habib_r_Page_104.jpg
30ee2ba3e2d26e18f5fbd4444c4c2976
875a7df9f021e8979d4b44f9318c0db3f07b097d
67692 F20101221_AAAKGV habib_r_Page_072.pro
2ad216debff5810643f99408b6bb93d1
0ec881b78a80b79ac8d1cb0b6e0a973e2623fd71
9168 F20101221_AAALDP habib_r_Page_101thm.jpg
dd8669b86143e5ad3fcc4398aa3f0d70
84bd81e58d42af0b07b090363f5d6ede99ccb2fb
174 F20101221_AAAISC habib_r_Page_003.txt
84169a498640139b0379808c585f7ab7
8d56f8c29a44bc1ee7c169ceb6b4afa893b9b5fa
58129 F20101221_AAAKGW habib_r_Page_073.pro
caa1f3915edc45424becc4d53baef32d
bf4ed9950f601a0b9f736e2d7b53855f3c9defc5
34986 F20101221_AAALDQ habib_r_Page_102.QC.jpg
13383ed6984b4dbeeb0745bfb56ed7b4
04c981a0bd71c3f938a978ff5e5e3ad0cacc70e9
9171 F20101221_AAAKZA habib_r_Page_019thm.jpg
e872d68d383e97a69b3d35c60fd36dbf
6e1a251089433307d1b73df2ae69fc3e5ac75f13
43110 F20101221_AAAISD habib_r_Page_075.pro
2ec1d958bbc138140fa4998df47c5e64
c59f3e3cde2f826e94b5fe2708ef0227228f1ab2
56229 F20101221_AAAKGX habib_r_Page_074.pro
58a8a6c2b2deb481d2b79662728688e8
d00901d3bd06750adf0d8942ff6b177459d2ab81
7764 F20101221_AAALDR habib_r_Page_102thm.jpg
a63025f9877b078e5de04074c310af66
cbee30b8ed5baa0e9341174e2bb6c9fc2d5e08ae
30137 F20101221_AAAKZB habib_r_Page_020.QC.jpg
141a94f95ad046752c7fababaeb75e54
b333ebba5069e3ca6b57761635f62133874a22f1
F20101221_AAAISE habib_r_Page_070thm.jpg
8ef8ef99679b103b19a0c657199b8d8c
50ff5d76a8e5a759805edf86ac4056c51a706401
64757 F20101221_AAAKGY habib_r_Page_076.pro
a189bc7865dbabef98cb40ea7e6ad28b
f40da62c6fc5a1f38a0b0dff6a4d766f137f3a3c
22261 F20101221_AAALDS habib_r_Page_103.QC.jpg
939caa50f61d7757b54c0e045fb4d057
050a979b5a8848e4de47f8a09c2513adeb922174
7440 F20101221_AAAKZC habib_r_Page_020thm.jpg
5c1731a04e395cd3a0632b21b1bca3f6
d6472d7669fd606b519140ae09b5c916187090cf
27166 F20101221_AAAISF habib_r_Page_314.pro
4dad848cca5dc1f6c9971d6c7c3b5652
4d6bdd6dc51eeb147ddc59c7c61c174464745fd3
51890 F20101221_AAAKGZ habib_r_Page_077.pro
7abde78f6501b08457bcd96a4eac0266
acf1f0e939ff4aa4a3d9361bac92befabcb9ef6e
5586 F20101221_AAALDT habib_r_Page_103thm.jpg
91a82f087f060557584782726c5893f8
8c5d6c23054d63d973a27e107a7825da91d86838
35285 F20101221_AAAKZD habib_r_Page_021.QC.jpg
48d0f2d6d628fd97f9f029e5396858b5
38d18121a6666fe95a0a3d45e14823e71b7eb59f
33737 F20101221_AAAISG habib_r_Page_325.QC.jpg
49008cec06315e9b96860913c5013391
abd0c1733a3599dec3cfe06e6a9f43e82be9e216
29494 F20101221_AAALDU habib_r_Page_104.QC.jpg
0f7d5e9235b191714da2e51480a86030
ee3f45a54fb1696a555cd8da3fde532cc2879ec1
9160 F20101221_AAAKZE habib_r_Page_022thm.jpg
7b9498a3d86c8ffa422a6defef4ce05f
18f5ea973f619dc32a632e056a97bec1e265156d
119700 F20101221_AAAISH habib_r_Page_323.jpg
35ebe79caa9fb9b005ead4e4b2c2f3a3
54ea6a114eb16e868575b76e52f8d26a217e84d8
F20101221_AAAJPA habib_r_Page_123.jp2
bc55a7d9a432ebbcb9a069e294298b9d
722132326305f24d515cfdfa8ede4cf212646151
7510 F20101221_AAALDV habib_r_Page_104thm.jpg
6e4bf72bc9d64d61c10a69bbb5bc55da
944b1a8184837c4bdfafb6bc4cf32e71e7a2447a
9134 F20101221_AAAKZF habib_r_Page_023thm.jpg
9ba3161fd7a3b2eefa9c2d507ce65b96
48a580357a92920428bb1377cc2d7500f017cae5
37469 F20101221_AAAISI habib_r_Page_211.QC.jpg
7e3777a32594bd087cd271020e467d42
40d82a4477fa7c60bc9edfc01c5967b3e8827570
F20101221_AAAJPB habib_r_Page_125.jp2
1121ee05f3dc51f75ddb6de6eeb931c9
ec21c25e1d53ede378b44896095ec8b02f8b35be
37495 F20101221_AAALDW habib_r_Page_105.QC.jpg
dfb2fc05480554fd08122a02d3605c56
67cf4b293de290a9ac19bbfd14e73c7c390e5807
9173 F20101221_AAAKZG habib_r_Page_024thm.jpg
4663d206c2ce198a1cfcbae42eeeb44e
4d9568a156e176acb4fd2b8ac1e985d296872625
16869 F20101221_AAAISJ habib_r_Page_296.QC.jpg
4c72fc2ecb8650a65f8304707df5a9c5
cb7723f2a2beacc752f6da77e8a6a99bdcc99b46
F20101221_AAAJPC habib_r_Page_127.jp2
6ed3dcc6132013bf3ddfe8df948319b3
2bc6d7c01d58aafe36ed4a2ebcf8145b65e8e98c
9247 F20101221_AAALDX habib_r_Page_105thm.jpg
25609a11d43459fa339165fcf3643370
e525915ab07dab5294caf9af8f5d5440c904ff56
38302 F20101221_AAAKZH habib_r_Page_025.QC.jpg
ecb712d8ab72042e3c8be54d0a5ad6c0
5e97cf6aa8715a2bd996e7e50932d6899d375a48
39120 F20101221_AAAISK habib_r_Page_034.QC.jpg
2bd78fbb0c998a88706ff69aacaa5642
c6e46fe3b9132b553961d7218771c4d6188687df
1051973 F20101221_AAAJPD habib_r_Page_128.jp2
e8cff28591c2f037e85eeb06dd39d854
540e075a6f7f68bf04a300d540fa1bd1a889651e
8383 F20101221_AAALDY habib_r_Page_106thm.jpg
773e8312a4f062dd5c209b63964d15b2
f9bd89df9d994a50a8fae66441adf8c1533e985d
9202 F20101221_AAAKZI habib_r_Page_025thm.jpg
6011b57de509e22c23fbfc44337a14c2
b5f0fa73b57b1bc0dff4911435a8dd9bf4ed1718
52654 F20101221_AAAISL habib_r_Page_235.pro
5a911a2fd6118921efd1149602b9bf88
a0e84cac0bf1262d0d2915777b9d9e808845cf16
F20101221_AAAJPE habib_r_Page_129.jp2
3c1218d545efb8e263705158f4716861
0fa7b24bfa318288efd12650bf993b0233708c20
1051925 F20101221_AAAIFA habib_r_Page_076.jp2
24a8487a3f7af00afba0dd8442765c13
28553c23b834ee27bda7ca5167185db61d6455ab
35368 F20101221_AAALDZ habib_r_Page_107.QC.jpg
4221d1a2a048426fbe2a0e305886a37a
6e43abdc079170132d50d8d1b26d0c12bc6f75fe
37320 F20101221_AAAKZJ habib_r_Page_026.QC.jpg
b8afebf0b5992c99a409e82a962fefb7
fe7b35cfffcc900b24644e714d055c3654637a2e
9930 F20101221_AAAISM habib_r_Page_263.QC.jpg
50340c78c258ca7fca3068aad9f648c1
3be825383bdc8e2d0f3cf6e91571ff35a7c6935e
F20101221_AAAJPF habib_r_Page_130.jp2
6105cbdbc17eafa4c16ed57508cb17cd
a32536faef595ae4aa29f75119f9660ceb91a618
38380 F20101221_AAAKZK habib_r_Page_027.QC.jpg
cdf5533392b682ed7c1a51815c402c79
e16e4e2c776da25b18f8142e4c492ceefbc077fd
2079 F20101221_AAAISN habib_r_Page_226.txt
3c4ffca38dcfdc81fa43b5b0a2fb66c7
9fa34bc38b334dbbadf866db2e3adb3c8b598814
F20101221_AAAJPG habib_r_Page_131.jp2
5602181fb50b8d0c5d9684a720916b65
186358ef47c73e6bcfb26a6581f28f8bded7be20
F20101221_AAAIFB habib_r_Page_034.tif
90bc51319a8f1ed591061499ff974ebe
2be17cb7f9c3db865a9471bd88129dbf2304baf4
9166 F20101221_AAAKZL habib_r_Page_027thm.jpg
0136d22fc7af5bdd020339f4ad7f27ac
2891215100fe5ebe39321c548f5a67f52aa1ff99
F20101221_AAAJPH habib_r_Page_132.jp2
7e8f14ea07d8693f7ee447b796e16332
f952f8332b84e1dd9a1dfbde5f6f9b4aa6b296d0
24381 F20101221_AAAKMA habib_r_Page_257.pro
e99c5d3e4b1585ba4f34edb0c5d8faec
c2357108e0b05427f851029e413a2ea9d9d7624d
F20101221_AAAIFC habib_r_Page_180.QC.jpg
69792a4d61352d8a76fceb9cd1756846
22f4938e1de6602c4892402459c1207ad6426e70
37568 F20101221_AAAKZM habib_r_Page_028.QC.jpg
6cee09c332940aa9b2e58abd84165b98
889048f48c465e3aacd3ee2eaaf14d2d9ccf0b98
2375 F20101221_AAAISO habib_r_Page_325.txt
b477543b31b642eaa706f3860b4dbaa1
763a2231a4ebbda68156823d663ff042523ad1b7
1051949 F20101221_AAAJPI habib_r_Page_133.jp2
aba33f8181662011289dead7af80444f
1038207e3ad863c23d0a5f964ff6e345f847118f
33768 F20101221_AAAKMB habib_r_Page_258.pro
b8ec21ab1c61bfe0eb65fe3ed962be6e
c3b1bba5db03ea2723ec5f709050c49733b754b1
6409 F20101221_AAAIFD habib_r_Page_300thm.jpg
d0dbe5f04eb6fdd936fbc59f23fc153e
eac047e37fc42a1d54d41fcff66168aadd624bba
9343 F20101221_AAAKZN habib_r_Page_028thm.jpg
294ac6294e124ce268b2eaa26716e46b
b65b04cc8959968950604aefe522934d2a504db6
29784 F20101221_AAAISP habib_r_Page_092.QC.jpg
d3a31265aa62128170bbae3ada57e925
46518c0a6dba045c51b18bc4d90da5cb1e664be2
603801 F20101221_AAAJPJ habib_r_Page_134.jp2
8562cae3322cea4a2a4255c31992fded
6482d544ea8b4f879e51f2c398e3f5a3314bd755
23849 F20101221_AAAKMC habib_r_Page_259.pro
c49241fbd49fbf36e6d07629ca59f2e9
c9239869f15cd5d3990e923c9e7bff0c4227f268
F20101221_AAAIFE habib_r_Page_053.tif
799845c564cf4b4653e604dff8af5857
a0b3249316ab8bab4d304d4db03761fd82d1da88
8987 F20101221_AAAKZO habib_r_Page_029thm.jpg
1a24c5b44bd06d8d1deeca85ff946ac5
d11e31e0819f1872cac42ab28c14bcea8fef2861
F20101221_AAAISQ habib_r_Page_084.tif
f4b877b3c5a3f6fd3b7de42d4de49737
c6b0773ab9a49d575b6148326dd2c1f3ccc52853
28724 F20101221_AAAKMD habib_r_Page_260.pro
23077733028123c52a92e2a48a171a70
1954ff0238372f91247a2c9f948e822a2ae05739
2275 F20101221_AAAIFF habib_r_Page_028.txt
04d0ba0d484bda8bb26c7fe137bbc24a
6c68f03e0e89823a53ee016c27ab866f90b71cc1
37506 F20101221_AAAKZP habib_r_Page_030.QC.jpg
05eb5874804567fbab3d9dd89735c819
7f93add415de0e376488269bc585058cef11eb20
2281 F20101221_AAAISR habib_r_Page_041.txt
3818631ffaa7515484b2f4c12cde7ed0
c7e791c7c9c5fff6385f4fbd684e62246eb29edd
F20101221_AAAJPK habib_r_Page_136.jp2
e1d36b18594ce4c6d92e83e4860610aa
0e50b4cfe87d510d1f629ada9d3116d0b53ccc4c
11840 F20101221_AAAKME habib_r_Page_261.pro
4b547df814190ec927cf04320db4565c
af66d2223f3f270d352de28e90f768897e4df8dd
34927 F20101221_AAAIFG habib_r_Page_097.QC.jpg
1a06664580cc7a786bf036f4a023bfbb
d22945111a281a7b2a15aaaf44944e412e9c6c7e
9199 F20101221_AAAKZQ habib_r_Page_030thm.jpg
840cb4ce4b304c65b5f97eed9b5a6465
e0c564f21c406d4d077158ee421b04cd27252fdd
76138 F20101221_AAAJCA habib_r_Page_005.jpg
f2dad873be0a99d5771261037e4a0cee
35f431804b0f3544644d48a0f846cb44b74fa42d
536 F20101221_AAAISS habib_r_Page_263.txt
ea4cba652f3cab41465d8727a89d9382
ac6a01a2d7357e1a60f244be79006f75d1dbffb9
1044223 F20101221_AAAJPL habib_r_Page_137.jp2
e5377ed094ce4f84bf9786a4edcff540
704bd8e0b73c989c0fc8dd73fbf4164eb1c66735
9574 F20101221_AAAKMF habib_r_Page_263.pro
b733f1a4e30eb21f0383b7cb715a3462
984176e3d477f8c5c02d69349b1de132441ba1b2
F20101221_AAAIFH habib_r_Page_061.tif
1c9de86cd25b83b2ddd2e918b7ccc84a
27ebce7a6ba00f03c7b15963d7fffa005f343d7b
35287 F20101221_AAAKZR habib_r_Page_031.QC.jpg
2f93915761bde3d510a91bf89ab29c3f
e8a60da4b3f1a42c4d506a3ebc6ae0d3ffebaf02
130430 F20101221_AAAJCB habib_r_Page_006.jpg
448fdd2e2bd849da9f67143559b34101
ef0b75ece4e05381cca112f095185ec55de4b323
53958 F20101221_AAAIST habib_r_Page_301.jpg
fb156cd032bc07aee225601fddf6f576
1c04fd492f79379fab996f6c4f226df80dea0a27
1051907 F20101221_AAAJPM habib_r_Page_138.jp2
c7de04a1fc334c530b1bcf2856ef1dbc
62d15d86a8f189febcf2d4bfbb70952f14e82fde
2728 F20101221_AAAIFI habib_r_Page_275.txt
19a60128deb64f93cd2bf12473c47d3a
1d9de111c6c39aa478dbd51e8f658bb9bb79b239
21433 F20101221_AAALJA habib_r_Page_198.QC.jpg
d7577b5d90b3aefd95ad3b5af9f772da
afbfb7ab7cd38842a2e3e9281ec8e637bff3587c
38101 F20101221_AAAKZS habib_r_Page_032.QC.jpg
dd026dc5870acc01245e3221d4141a81
3675c1431e0b958b9e552f124dcb9ebebafa6c48
169000 F20101221_AAAJCC habib_r_Page_008.jpg
151b6a49e70f8ece399631e03507d7e9
ab9aa70d6c9d7a21776c426adb5b7149a263e005
81307 F20101221_AAAISU habib_r_Page_194.jpg
8b026288b31e9cc16dcf41321a2ec020
2fcbcdec28d1d78daf12d613e352d71e52ccc048
F20101221_AAAJPN habib_r_Page_139.jp2
6d627e471da9321ad8bc6abb245b587a
324d16e398469cf8e4d1629b67f49b55bf066068
13064 F20101221_AAAKMG habib_r_Page_264.pro
63b11d3851fd94d5f47fc37e1d635915
e56f9ba288292b16463c5e29f8ce404dc5717c19
9252 F20101221_AAAIFJ habib_r_Page_163thm.jpg
e190743a7c5be6d4ee8a7ec1fc2f871e
e80895d1182030b7e2dd5e8395b351365f6af97f
6001 F20101221_AAALJB habib_r_Page_198thm.jpg
94f83612f88e705efc86201cc925b7bc
d7626c3aa2bd83bab61b258f3940ae6cd574b129
144260 F20101221_AAAJCD habib_r_Page_009.jpg
50d53320ae9b86a7667aadc82bfcedab
13e3d4d7f21bf848b57d5e552985aac2e8622b08
27626 F20101221_AAAISV habib_r_Page_270.jpg
5d60d9efb033c48bc0bb7328b377247a
8de5e8e490b219bdfb0f78d79d69b44ad9073121
877795 F20101221_AAAJPO habib_r_Page_141.jp2
5541c17a6c84c91c17871f91a1b7ffa7
c4ea0337be9964174ee3a8cc43721dd56e395eef
31950 F20101221_AAAKMH habib_r_Page_266.pro
cc8bd1afbb6a8a9f77f321339b905632
7c61b68863fed2e4ed1987b9157f18c353671448
F20101221_AAAIFK habib_r_Page_013.tif
085a5dcf6f7ae2d5d152ee99fd1b1d29
dd04e82ae2b5c2234ff2cedfa875832fb94feb2b
8861 F20101221_AAAKZT habib_r_Page_033thm.jpg
8ded48a7f6793a4d1043216e732d0a72
210096eb542756b71ec848699050c8e13928ed60
167108 F20101221_AAAJCE habib_r_Page_010.jpg
81a1271059fe81fbe696bf4f199a72c6
df2b09eaac44283c5437dd7400a6ed271d0ec776
9045 F20101221_AAAISW habib_r_Page_161thm.jpg
caf78a02608489d1768c840d1e73bd7d
1356e6d852da0a8eee6f4a990af5b5e90b8778ea
F20101221_AAAJPP habib_r_Page_142.jp2
2dd779288ac346ee66ade46e385690d4
a05e3a48a1fade461632336904c25bac2923a244
17500 F20101221_AAAKMI habib_r_Page_267.pro
ac9e7c614291f94e0b7cd96eb2f6b399
d0b156a5116351a45ddb3ac04a173170064f2677
32792 F20101221_AAAIFL habib_r_Page_230.QC.jpg
d09cd44e5e16d523744838aa48beb9a6
fc2240e4e3863b0698514c4c92e9d2dfae96b3d3
14204 F20101221_AAALJC habib_r_Page_199.QC.jpg
f8372fcbbe407d1d9b5dfcd99d28ab49
23f96a47ab2ff6c0c1d5dda7c3e5615a37c0c2bd
35865 F20101221_AAAKZU habib_r_Page_035.QC.jpg
81343761d420b5187cfe08339d3e8fa7
bbec7567e9e2eb12797891759cfd0211eda5ca0a
12915 F20101221_AAAJCF habib_r_Page_011.jpg
b9375b3402874f064e4be6c3f24a6ab1
ff5c6d655b156bbdcdf7cb07be29e217ac080662
F20101221_AAAISX habib_r_Page_215.tif
02761801d3e21a27e3b8630985f8ff84
c4335804f4c5db142c7c07ff24af5121cdfeeab7
1051967 F20101221_AAAJPQ habib_r_Page_143.jp2
ba9fdd2020085e243b57650f440bfec0
1301b78345bdb9155436e12a913d86284e5f2dfb
21830 F20101221_AAAKMJ habib_r_Page_268.pro
9604fd6817932b80659fff8b138a6a1c
858b0be3ed137d6e22fbfa5384a62f37aff6751f
9266 F20101221_AAAIFM habib_r_Page_047thm.jpg
ae630a13db3d55a48e91867a19c0c0d9
5ca31207a938d90002e923f6f16cf771e962f99e
4105 F20101221_AAALJD habib_r_Page_199thm.jpg
227d458bbd086b30697b55a80111269e
dac841afc576112ecb1d31eecdcf2b837f409eed
8933 F20101221_AAAKZV habib_r_Page_035thm.jpg
24426502ce3e077a5b69485c038ced29
897ed05a3c0bcb106b5868b85b60b64b98cd7f0d
134375 F20101221_AAAJCG habib_r_Page_014.jpg
19cbf21afece316d28877a2d3eb0926b
b2e080be43311647325bcf5fa70194fd1dd76507
F20101221_AAAISY habib_r_Page_010.tif
20dedef36b10e44caa8e6f88e1a1d14f
60db30839baa793a40d9dc41a2f86d86da057b4a
965583 F20101221_AAAJPR habib_r_Page_144.jp2
3a17eb8b178b4d2365f97e2341b75e8d
28d9f7bb9cad362ff1885e4b43382c8e68f966c7
10596 F20101221_AAAKMK habib_r_Page_269.pro
9108091339087defe5674f8982961d78
218f134efb15d8d9da41777aad67cc9617af18c8
2229 F20101221_AAAIFN habib_r_Page_036.txt
352e8502baef1589eaa5937d7e490bbf
06b45bb1bd03725ae554cc0a57c09b4b53a2ac66
35332 F20101221_AAALJE habib_r_Page_200.QC.jpg
17fe0ab9ad057dbbbf552dac7576fcbe
dcaa6fead8d2954c6ac9d49fea3e8c7d27837aa2
37128 F20101221_AAAKZW habib_r_Page_036.QC.jpg
36d1a877abcc54707aacf05cbea21f9b
54e777389819fe07562db6898ac5896ebfc10e1b
107739 F20101221_AAAJCH habib_r_Page_016.jpg
115eaf31e99c214e138e96eed41cb6e9
2eb5b529e7e9d6ebd2e95e474f1d7c4f0e0eec29
56144 F20101221_AAAISZ habib_r_Page_134.jpg
6215d076a06d824e358f7a3a2633f1e4
baef5484be573927547cca2fdd8c24e1e0637649
F20101221_AAAJPS habib_r_Page_145.jp2
06a451eacfbfc95f09da50cec20de5cd
1d5446e52b727deed967c0f607231b205638b693
F20101221_AAAKML habib_r_Page_270.pro
37decb918c1a644b615a5af3b8af79ac
64e25ce24135401118a8321fa6797be8f60bf07f
38492 F20101221_AAAIFO habib_r_Page_055.QC.jpg
53b498c59f18b7268f336e8e5222645d
bd916a6d5e6ea84421c4b615bdae5f5d60610326
8594 F20101221_AAALJF habib_r_Page_200thm.jpg
484130a3dfc39d9f626694b641ef183b
88e0d22165d35532e3eb62ccee4b4f317ad4ab72
9237 F20101221_AAAKZX habib_r_Page_036thm.jpg
893bcb775ec32be98646e860221008cc
f97010a04e8573e48c821b553199d053b989a6b9
41824 F20101221_AAAJCI habib_r_Page_017.jpg
17c6eaf23d22e125367d41fa7303c360
062e793b946c4596e9cf0efa872c8700ef642f9b
F20101221_AAAJPT habib_r_Page_147.jp2
70dde7340a3f6aebb3c9af385adf8cf1
dff07b37da54185fb5c68b185a152f57bec0b585
7389 F20101221_AAAKMM habib_r_Page_271.pro
d344711f104268f322f9862e15249513
8268a07eb2ea9b25358f45aa572e90705281cb8f
F20101221_AAAIFP habib_r_Page_200.tif
fc096aa7c4e6c6cea258b02a443aebe2
c54ddc2716060014dfb88833ded0c87815b62bf2
35637 F20101221_AAALJG habib_r_Page_201.QC.jpg
c177df48d00986269a9a4970b46958f7
383e63017b961588fbdbdbf097813b721b4a5b57
37693 F20101221_AAAKZY habib_r_Page_037.QC.jpg
26162c80d9fd3731621556d688886587
d275eb777d32071d70e9a61d60655d5dcba9ead2
97269 F20101221_AAAJCJ habib_r_Page_018.jpg
c5783727a28dfde7a7288b7e602f2fb1
b7a595b6feb96d7bd371c9ba0fa4da3dadfa2dc6
647844 F20101221_AAAJPU habib_r_Page_148.jp2
fc969f989524a8b5a1663b49ef80c84b
267da0e99181d2b2b888b60f592953ad3ce5b20e
24854 F20101221_AAAKMN habib_r_Page_272.pro
1024839a258ba9bf73e50f11b980474a
ef0d620e432eb40e66a5011c5d90fb88fffb0b76
2295 F20101221_AAAIFQ habib_r_Page_161.txt
0ccd07608209f079b9b1c3da6f18a164
02d8fd9e33dd49e607ad288862b51ab4cf58b8a1
8767 F20101221_AAALJH habib_r_Page_201thm.jpg
7feda875bd4b5f3b967f3d59b391e3d8
6a09c463b8c6d0819067f26f4372fbe501dabee2
8901 F20101221_AAAKZZ habib_r_Page_037thm.jpg
9c2d3e7a1dab2a4ba099529f1f2b294c
b810af6466ccdbcfc1f2e4adf32d4576587fffb8
113630 F20101221_AAAJCK habib_r_Page_019.jpg
48f76cba477b15717b674b5f1a197235
586691481c16ce24173e101c289e2c00a6e0f59c
552355 F20101221_AAAJPV habib_r_Page_149.jp2
59a30044c402b1a169cf6bd6cdca0dd8
ac610d25414ab95b9097bd55d0eb24700bf12fb1
16609 F20101221_AAAKMO habib_r_Page_273.pro
1c2398831e5f9f52b68caaa1fd3b8176
5559841faed6815eb6a25d1ef006cddce25f4e62
8136 F20101221_AAAIFR habib_r_Page_204thm.jpg
73fae0cc1e6861f3b76976e7e61e5465
1db0cdfc4bd3306093169adcdf2affee01974e57
9061 F20101221_AAALJI habib_r_Page_203thm.jpg
8fe6f5833386e1650c7e9f7d342464b5
51dd86eef6dd63251eaacdb9cf979fe878dee474
91591 F20101221_AAAJCL habib_r_Page_020.jpg
686d15860126a3e75bf43dda9cd5459f
d895b6a07d6cefbdfcd9631cce0d29e913a93079
741830 F20101221_AAAJPW habib_r_Page_151.jp2
3fc3154a91655ee15ce9035083f86866
7ac5df3048afd809f386082bb97e5e2d93128050
22296 F20101221_AAAKMP habib_r_Page_274.pro
e6e91a4ca91eb3b3299cdb9c7a728c2f
c103a84f8ef7f25a3e9247a997ad8f00e3165177
F20101221_AAAIFS habib_r_Page_066.jp2
c1fd15b1a77f2a3d603ac0af715239ac
09b663f2b3b677b50450f0d5475052583b4a1516
32604 F20101221_AAALJJ habib_r_Page_204.QC.jpg
ff0f7fb826f18e121a3a70e020a8eae1
8a93c8845c8266851bd996ee5874daeb6318a132
107462 F20101221_AAAJCM habib_r_Page_021.jpg
19099c463c993d258d4198bbd5d9ae06
c92b43d8e47d97c0edcc0423280f8cddc08dee53
1051922 F20101221_AAAJPX habib_r_Page_153.jp2
56694a649f3ee45eed70f35f27af9289
77aa26628627551a920f158a09df63ca56c79881
32327 F20101221_AAAKMQ habib_r_Page_275.pro
aa381bc4d087c18c4dbce0abd3e6381c
8701b35d1b859c13eb557309acac41f8e3e7e4da
58063 F20101221_AAAIFT habib_r_Page_196.jpg
c494ef15d29d0f653aa2d224f67b949b
407f4d62295245ddbcdf0c0b887f3c36c37b9844
10761 F20101221_AAALJK habib_r_Page_205.QC.jpg
8b0b053edaee9b7187eee06708042572
f90e36f5e4fd28870259c0e83ec13532168bcac2
123414 F20101221_AAAJCN habib_r_Page_023.jpg
5984f84c11b0d23c3e1374bf81eddf62
7404752e9130195c65eb70123b3d4723a365f51d
941951 F20101221_AAAJPY habib_r_Page_158.jp2
a4498d3a9515242ef66af835d5561870
32f049f660bf61f10804fc15e99d57425a42bfbb
10606 F20101221_AAAKMR habib_r_Page_277.pro
7d3b72d51ba75aa81b9ff6d6ec12027f
ba0c19a4be173916d76f78f94fd9698227292ba6
2199 F20101221_AAAIFU habib_r_Page_038.txt
1103f8a440cd91a4973338c30daca159
0ccea88fa88237c87ab630257785a83f3212f3fb
35479 F20101221_AAALJL habib_r_Page_206.QC.jpg
a0114a359806b917676d781264aa5543
b21fc3548a08859f734ff83317a84ce66169a42f
114888 F20101221_AAAJCO habib_r_Page_024.jpg
7267a708ddd66795d20c8c16c079a16b
86c0b1b60562f6e60a8187dd8fb3b520c561abca
F20101221_AAAJPZ habib_r_Page_159.jp2
ea8fdaba388e2faca2fc5e10c9d07fe3
74d2e1ae3005acb81a1a29b0a00b0af163adce8d
10043 F20101221_AAAKMS habib_r_Page_278.pro
8f32a5a89b7928c974899b07a66cea4c
8e03026ae71dba2b6d0438b933d69177ca331639
463466 F20101221_AAAIFV habib_r_Page_280.jp2
5a8f267b5c8cce61531c751e6fe73298
47848ec62e6d9891d37bd5a49e1d704753406275
37597 F20101221_AAALJM habib_r_Page_207.QC.jpg
fd202cfb0bd8695e5c2a381dbf49211e
251ad1c4c3d3f1c8d9a99313769d3841fa50ccf8
113557 F20101221_AAAJCP habib_r_Page_027.jpg
ea027d9dcf720407e9eeb99faed187b2
2fca6ec092420b2e63883b521db324aa0ccba7eb
28564 F20101221_AAAKMT habib_r_Page_281.pro
9fc74c8c063cc526bb279fe757a9027d
43c38f0602776f5d91c6ec509a0f8b69df26f7ba
2056 F20101221_AAAIFW habib_r_Page_167.txt
2a7b2b59a8a2bd9343552e7315d15a81
71e36d360551e9de3c1600051c297ce3c2beeca4
9313 F20101221_AAALJN habib_r_Page_207thm.jpg
55d3a8592513d85e860d1929f368e747
7258f81a070f64130d342331f12eab892a5e8fa9
48532 F20101221_AAAIYA habib_r_Page_296.jpg
543d4b56d817bb1d9f2f8f2f604a01a6
e5d5f75c408b04f4883bdcec7b0da25238d73a9d
115719 F20101221_AAAJCQ habib_r_Page_028.jpg
fc0db7104f1000fef46846bf60f82acb
e9e52016185e3cd1ea2289f0f67b9fa34802d46c
40729 F20101221_AAAKMU habib_r_Page_282.pro
99baa1a5bd6e017de4ec7546ee2059e3
872d098794a4adcf01b81c7570790ff6dd369c0d
99087 F20101221_AAAIFX habib_r_Page_088.jpg
a7bb431c3de58bbbff8cb7213c01bc3c
a7dc6da4d31a5459b3ce1cbff54cce424edd34a5
30426 F20101221_AAALJO habib_r_Page_208.QC.jpg
10c3358e7f5fc76ea0b89744dd708005
652401124e44ca42fcfe2e5c441b10eb8082e8e2
38550 F20101221_AAAIYB habib_r_Page_014.QC.jpg
cf418f84ac6f6c62982c3e7fd004fff7
2c73f466db93498e450b70f6e7f000cafe864148
112097 F20101221_AAAJCR habib_r_Page_029.jpg
89e2e7aa250d332db3dcc61ac70215db
a65d2b1b3bbf7c0fa0a4d71aabad962c73cfe635
12583 F20101221_AAAKMV habib_r_Page_283.pro
accaec7e39fb1659d9a3efe267eaab63
ecedb8b2e1231e608d5a7080d8731077bb6c0d9a
5321 F20101221_AAAIFY habib_r_Page_265thm.jpg
871b032978b54bfcbed34d46cb814f3d
5d082418227fbd3739d1f4979ec4985455fd540d
7777 F20101221_AAALJP habib_r_Page_208thm.jpg
40a05d0d1e44f5efcc6a74a9a5ffc0e8
bcc03f2faaf60e3e96e7bb733d5858c8be84d54d
57920 F20101221_AAAIYC habib_r_Page_162.pro
b1a6ae0125201f6e0be1fc57c80b9795
46b2feda6ed087fbd7e46de1b8c614f9eb2ebdea
114861 F20101221_AAAJCS habib_r_Page_030.jpg
ac80a2a386bdd05663ef0f8e6482c5be
5bc8f845fbc111aab5840ce7b6b4f90f0707edc8
26940 F20101221_AAAKMW habib_r_Page_284.pro
d4ceb4a0d74bb998b251670c704f6079
990ab1d34f959fbf9d318eb3dbaac92f09cd07d8
F20101221_AAAIFZ habib_r_Page_209.tif
84c83151d56cf203772ab768b3dbf8f4
37f51bf223d483523e58b5887a7c4c6bd282bcf4
8817 F20101221_AAALJQ habib_r_Page_209thm.jpg
cce84c59a9faaa0e2b186adc2e8d60d3
9a6cc161220910790faa67703e4fde11dfd2c707
F20101221_AAAIYD habib_r_Page_224.tif
f7c8b8e305f1494e08295e49c7e0a0c4
2f87d6e58a18fd5906de253391f5d77444a47aab
111639 F20101221_AAAJCT habib_r_Page_031.jpg
5f1b471593779588dd8729e310b5bd84
7d44b8abc2d2ab1e4e65f3c588da31846ebccaaf
26438 F20101221_AAAKMX habib_r_Page_285.pro
7eaf24363c255adbc36c827af0a19c27
aca50e873660bb4bc41728e593fce4148af0bed4
6079 F20101221_AAALJR habib_r_Page_210thm.jpg
9ad406fad508f3aefea02d0930de82ac
fa3a29dcd9da0622feb4b3ee37876ab36ae66f1c
159379 F20101221_AAAIYE habib_r_Page_013.jpg
9835e6dd25bb6b6fc946ecf82c5836ec
9872c50e4d3c8139706b12e1f42893d81c2cbc42
112789 F20101221_AAAJCU habib_r_Page_032.jpg
8ea7ca8d5c1aa401842de926ba5349d7
e1f55e683390bde04f2162971a9614dba2fccb48
30156 F20101221_AAAKMY habib_r_Page_286.pro
fd1d0232f3e59ca893625290889a6d0d
7909a8e55475340845fcc76e2d08d1ae949aa839
F20101221_AAALJS habib_r_Page_211thm.jpg
6a32f8a8ecef22f3b9b3b7170e3d1e66
6f266393208b13ffa25aa9c7834572eb3aff63d7
113920 F20101221_AAAIYF habib_r_Page_333.jpg
3c5240b7300d84765c31243ffcc1ffd5
8fcc995dd79e207b9a2422bf448ef690adcbdf73
114836 F20101221_AAAJCV habib_r_Page_033.jpg
6c2b9d59b278fc4c5b637036c358ceda
dbab6d739c01680dd7a2b2a8637d1f1767bd3c88
21592 F20101221_AAAKMZ habib_r_Page_287.pro
80135e4aedb526df0a9d8043b084aa2e
139d60a059faa2bc07aa96cf2f8c434b62321706
8626 F20101221_AAALJT habib_r_Page_212thm.jpg
d0006afa11700d9539b1781178c3fbef
b0972530cc64fb2c38864137189246fe9a24f102
386237 F20101221_AAAIYG habib_r_Page_275.jp2
4598188272b20799b81c5b59773d4e37
73ef72440b91ae02819f5cc752819e87784d6643
112085 F20101221_AAAJCW habib_r_Page_035.jpg
e39ed89bb37c84c4d7c5f60e08781b92
7aaa8189eabd5afd8222681e0035afe896dcc002
3942 F20101221_AAALJU habib_r_Page_213thm.jpg
685e8aea521e8c907c014e7edeaefdad
29678d7e416996b8d9bf5c843dc678ce3b42b517
114693 F20101221_AAAJCX habib_r_Page_036.jpg
35031aee9b94e88543898beef9ded3bb
fa2856cd00e859fae994498db3278d0ed6241f58
F20101221_AAAJVA habib_r_Page_332.jp2
f18f937bbf7ed227166c4a9c9e2b1370
8082f633f167c28be1b7d8a97698e6ab7a81b0fc
8849 F20101221_AAAIYH habib_r_Page_122thm.jpg
8db3ab511b235a31fbca273328fea313
661165f1a9183438d2b426fe643e886afc982bc1
34434 F20101221_AAALJV habib_r_Page_214.QC.jpg
1b7401cad5ee9319c90ce7f2008cf7b5
f5c66a4ffb78c1d708bcc385213b57be95f073f3
117660 F20101221_AAAJCY habib_r_Page_037.jpg
4943ca1c2deabd98afc6a9fcc8bd417e
d6123c15db07b1370892c0ce8fd1fcdf64456de9
F20101221_AAAJVB habib_r_Page_333.jp2
2072e7ee47357f4a6e07af10069cc07a
e48bda98ab0837ed20304dbb9c69cb2e5ffffe78
8979 F20101221_AAAIYI habib_r_Page_241thm.jpg
09c2cfe42e304d23ed0effdba8ed0c3a
f022f8e217152ac740ad59d444d73feee7da1fc6
8780 F20101221_AAALJW habib_r_Page_214thm.jpg
97fa0e0630c0cd647927902152733125
a434ef267e3cec10c7b6ab53a15efcc9d2eb0b19
842002 F20101221_AAAJVC habib_r_Page_335.jp2
0357aea30acb4bd9205a093cdde273c9
154c6e8c0437e0453743dfd21155518345764753
1044435 F20101221_AAAIYJ habib_r_Page_240.jp2
ad7ada6e2b9b4f3c57fb8543add95edf
94a42a3d1abb4a0394916ec5503a23c7378125cb
111668 F20101221_AAAJCZ habib_r_Page_038.jpg
86a6c8d439312cd0bd9d6e7aba194b19
46ae5a68985985c79b1ccc81cabb6bb684093922
38863 F20101221_AAALJX habib_r_Page_215.QC.jpg
15fcbb6cb0a52167986f8489fec3355d
ae7694b754a8bab92c261821e93ed4b5a8ba3961
F20101221_AAAJVD habib_r_Page_001.tif
900610b6250476dbde15a968c9669bae
1957627c407f62cabc36fee191017f9c8397c1f5
57897 F20101221_AAAIYK habib_r_Page_331.pro
96c3e057d967978c0e382c05dda11f64
569ea6144607c4f6f84bbfc30906df7499e131c9
9309 F20101221_AAALJY habib_r_Page_215thm.jpg
b4e717a3aa1e1bbe16cd8a7dfe7bac9b
18a757fd13ded2968c767372ea5827fcfe5a4eb8
F20101221_AAAJVE habib_r_Page_002.tif
2fd485015b5810527c751c0ace08daa7
dee410525a7abb4d09bf451ea91ac864b8ae7643
110505 F20101221_AAAILA habib_r_Page_327.jpg
389b14ed7343c38192bbbdf846eac891
5a6955bcf76469dab7843f40eed7df5ab62bf61c
F20101221_AAAIYL habib_r_Page_101.jp2
688f42b42051b3d1d21560b1b4949308
60781bd61efe5e8c2e085289cfc31ea8e7fc557b
39592 F20101221_AAALJZ habib_r_Page_217.QC.jpg
e44a1cda7ea048889626aa7a60a119ea
0c9c151ca93efe5f95d4d12a070bdfed2fe6718e
F20101221_AAAJVF habib_r_Page_003.tif
5237caeb647a987c5b33e6d718e24c21
2f505dc02ca86f59be12353e120a4132dbc77879
F20101221_AAAILB habib_r_Page_156.jp2
6a469f355a1c36f0b85b4212e2acbfe9
93bfc41f628f3cce35e6b615593966425f754799
8406 F20101221_AAAIYM habib_r_Page_111thm.jpg
d990b71135190cd50e14e1b0cee2843e
a261799d80934cb9daa28f3663edc98f6f8eb993
F20101221_AAAJVG habib_r_Page_004.tif
8061bfb56d7a481f5fb663734082a37e
dc2b38389e590fda9ef6bea707340185f44ffea4
37096 F20101221_AAAILC habib_r_Page_022.QC.jpg
7fc2be023967e7d6b4c82759d7e3e6db
6a2246396dda82464552c9f4c807384201510da4
9052 F20101221_AAAIYN habib_r_Page_053thm.jpg
3f0308b365ead6b1b3f5e7934dfc741f
e378f8971913613e33d8542ac8643c0322271586
1634 F20101221_AAAKSA habib_r_Page_141.txt
8959db0287eebeaf473ec6177583a8b5
15e703c2201af93606532ac41fba1bacd0c1b099
F20101221_AAAJVH habib_r_Page_005.tif
511959914030fdf2e3ff2bf286d95b5c
a90e7e84bc13388bc03fbdc2bdb5e2a73d06044b
F20101221_AAAILD habib_r_Page_072.tif
f632844ed4b94220b4ec8c9234d1a958
8a18e5002d518bfb758072d2fd41c45485eb14ea
F20101221_AAAIYO habib_r_Page_314.tif
bd38296a42911ad5fe1ff9da2c955165
4d1998e40ea80980f73a7e2b4233cd53edf6ff23
F20101221_AAAKSB habib_r_Page_142.txt
c2f200093a3aa2d4242e6e3e57834162
a15007de3ea068f2b497040137df730dfddbfb53
F20101221_AAAJVI habib_r_Page_006.tif
91c8ae46614eccf1adbe358472364a9e
8db0274c7b0589b0ea9c7c10ba4373a7ccaffc37
4512 F20101221_AAAILE habib_r_Page_267thm.jpg
6639281687b58aec6aaab0b7900f565d
44d222d53841d65262bef0b11367ecb2f5c4269c
75772 F20101221_AAAIYP habib_r_Page_223.jpg
c66b8432e7e05d026d9eee0760afaf6e
cbc88eb3f7a87c155c5c84c4b8e68cf7aaf3f16d
2269 F20101221_AAAKSC habib_r_Page_143.txt
17e73e680bcf9fd66229bad01a74aa9c
d9147954f72f7c03dfdfb9f20801fee399113647
F20101221_AAAJVJ habib_r_Page_009.tif
d962e510f3732821fb5563592f260bd8
34ec6006a952a56a3ff49b0339f690796d382d09
F20101221_AAAILF habib_r_Page_079.jp2
b4df70f8d91923a8b2c7f87d2a513552
7e0dada792b7abe545969a5ddb2782d0be6fc098
F20101221_AAAIYQ habib_r_Page_154.jp2
393580e6f86f5564c0c198def337cb53
9374bf61b57b04bd6341fae92bdf78e3651e611a
1749 F20101221_AAAKSD habib_r_Page_144.txt
2fb3b317ca174893b4cabbfed5a26239
a102690507dac91a57a66cca3eeec7b131abddfe
F20101221_AAAJVK habib_r_Page_012.tif
e6e840089cb5395ffd8395edd7fcc57c
6c33bb93259eeafed46a4c30fc379a3a239b4fe4
2141 F20101221_AAAILG habib_r_Page_063.txt
1373d42334455c5ba8086839e65c2e99
ff4e7c2d3a8dd8d9757a95ba335f0d220256f23c
9038 F20101221_AAAIYR habib_r_Page_043thm.jpg
b572d18698b86a500e3beed12f18c3ee
5bbfaebeab9f9a837f6a35d9130a26133a6b0d87
F20101221_AAAKSE habib_r_Page_145.txt
6ec5ff6ffca45f1e81a6545b93ba7443
bf83ad6ae6ae2d9edc30bdedeee3fac1325ce89e
F20101221_AAAJVL habib_r_Page_014.tif
46069760de18320d8abdcaef58824eb4
99b80d8b10a5c0f446018058a0ec1608e51f0446
116010 F20101221_AAAJIA habib_r_Page_217.jpg
63bb946831cb69502f2c68119f7ec8b2
9fa4854ac3f772904968831116573099950fb2b1
2174 F20101221_AAAIYS habib_r_Page_115.txt
b83d9ed83217ed148070be9c552793bd
b2fb18134f50b14c13040d2309c44da4a8172be5
2166 F20101221_AAAKSF habib_r_Page_146.txt
b5f71b99857cea5fed10960d3c5a5c14
f4cab68f6060083e6a4a389bf52904d2abec920e
F20101221_AAAJVM habib_r_Page_015.tif
1ed37e49497dc7e6922ead3163b214dd
9ef30a4274ad90c68d15d775477fcfefe65cdb4b
F20101221_AAAILH habib_r_Page_223.tif
c70e82024373933f79400d5621e5dbd4
9a774e405b8566ab0347ba1f8cf50a0efd725640
115486 F20101221_AAAJIB habib_r_Page_218.jpg
99188fdcb7dba3f558fb11342df351ff
b586925baec13628a66458bc2680572532f0b943
261205 F20101221_AAAIYT habib_r_Page_001.jp2
6b60c9e29617cc13390bd2ce67276559
a02d006e6d56287d3ac6394c5125ef9ec340c892
2244 F20101221_AAAKSG habib_r_Page_147.txt
25aaf8272f5eb1298561204f1b8897cf
aaf2801470cd2a6c2255cb65b4f92d148f06ccfd
F20101221_AAAJVN habib_r_Page_016.tif
d6f92a75def371bd27215e02610e1dd9
730d8b8d2194d907b5592ec47f27c5ab3b9a968c
35388 F20101221_AAAILI habib_r_Page_238.QC.jpg
6397b25bc87bac5234bf0c7f463b96a8
919d274f25097a2543f58e979e552df74f9cacda
103121 F20101221_AAAJIC habib_r_Page_219.jpg
e496d754b64c8297fcead420209da114
33ff10357026419c2a5b5af040c98b7d72e79696
15751 F20101221_AAALPA habib_r_Page_308.QC.jpg
506ce03f995a4ffa4a422b303ec10caf
0629a0f0d76a63f99ac7c0d1e2141a10951acd1d
1217 F20101221_AAAKSH habib_r_Page_149.txt
61aeb2a950653a79bf3e23726307c302
4960b4a0997eb313fd9359f9d71aa8e335ace902
F20101221_AAAJVO habib_r_Page_017.tif
44235cc5536e4e9b7b9ea1bb22509611
ed63624ba93e417fee1b1af2d499284db4d75fe3
F20101221_AAAILJ habib_r_Page_069.tif
0fe8817ac09b82dd0adb4274bfcb5689
1aa9458aeee40e702b7ba9c7c133e3eb832e59ec
1284 F20101221_AAAIYU habib_r_Page_195.txt
688f1790c0cbe66c414403c689c5977b
83a8ea35bdce19a4948d01036ab0fe53ef72774b
4519 F20101221_AAALPB habib_r_Page_308thm.jpg
6e956882be5a8a103606a41734917fef
c956f7b5c599870943fbfa0744480e91f79fe1bd
1281 F20101221_AAAKSI habib_r_Page_150.txt
a42e417a7532dbf5e7783317fc6a6974
2def14a98804eb9415dbb1d98e8013f65be8901e
F20101221_AAAJVP habib_r_Page_018.tif
59cf243fcda34eeba17f639d1fed2daf
d329145f81e44b1593376afcc0465d665d523a74
30074 F20101221_AAAILK habib_r_Page_051.QC.jpg
61b5c7dcae3e4966f2175e2ba7504337
9ca77b88edb1f024658924b88b0f758496e2d30a
107341 F20101221_AAAJID habib_r_Page_220.jpg
31c878ab54e198b830652f2e8ecde36a
542bb6b0ed2eb02f6394446a3df8fc4e20cb5195
9180 F20101221_AAAIYV habib_r_Page_236thm.jpg
6ce063ac72a734caf8feb6e7603f79ff
56b52f5248419308c47e134b544f093efa7acd1a
17326 F20101221_AAALPC habib_r_Page_310.QC.jpg
1bab62d351f3a49fa233404c4c2be211
b6900a552f05b9a374b914c8af7ca304c1ae686a
1513 F20101221_AAAKSJ habib_r_Page_151.txt
6f19128f86d9e464a163c42783635bb7
151bd142e28d98546a1f82a040df7288fa53b04c
2126 F20101221_AAAILL habib_r_Page_153.txt
9969e57b8d003047c1e9db0cb306a029
2704e27604052d7db6be4c6dca98ccc7b3daa99a
121427 F20101221_AAAJIE habib_r_Page_221.jpg
6265d972a275841be12cf3f05c8d6101
5aa7f9991d0a77fcc2d7d3e802c616270c6e11c9
33908 F20101221_AAAIYW habib_r_Page_327.QC.jpg
7d04ae7755c3eb2a4228b411231db489
b2c214d79b38106e4e2b1a995fbc9294268fb523
5341 F20101221_AAALPD habib_r_Page_310thm.jpg
79e7c463a72904ab1945c7797fb6c43e
c4f6d4c840da47d1f7719cdf7892eccf10ae879f
2067 F20101221_AAAKSK habib_r_Page_152.txt
7db14e993b61017f7610981ed0f07e2b
ecfe620bfe42ff087cf6adabc308cc44ce388ab9
F20101221_AAAJVQ habib_r_Page_020.tif
c1276f4e097f848225f5463c2580dd3c
bafbe6982c67cda10b6f28af9ece9a46d024ac47
F20101221_AAAILM habib_r_Page_146.jp2
1c4d27a330230ce76de0166cdacd4c5a
be36b33c8d145533665ef64ff3545360f325445e
116749 F20101221_AAAJIF habib_r_Page_222.jpg
8b8d696de09af8a3e5c9fef3a90cb5e0
5bd1daa541d3667740283dd49b70d3fdde692832
19172 F20101221_AAAIYX habib_r_Page_301.QC.jpg
fc781fd4677135f40f0deee58a885033
ac41c29e9857a48fae6855ae5132bd0047bd2d65
20132 F20101221_AAALPE habib_r_Page_311.QC.jpg
d301be729fc3a057f98412420debe2ec
59783df4a6eb92cfb1f0641636373348b57f2452
1962 F20101221_AAAKSL habib_r_Page_154.txt
5cd2d425c9247fc55a88c994dd1b1695
725875c60bd3a4b1c507e6d62de6c5419d5080dc
F20101221_AAAJVR habib_r_Page_021.tif
529ae09e9eb41021bc8bb7b675fc91b7
c11bca4a29f21c24d80d133210bc3d9eb0f4aa6c
100134 F20101221_AAAILN habib_r_Page_100.jpg
3ce19250b8f601109b8c78df724136b8
5f971f468881223b04b67ba7463f7a63758006ab
100369 F20101221_AAAJIG habib_r_Page_224.jpg
6902525c6768877b33c438d508b3309e
ac970a6fdf33909d5481beb3a75a6dcc177fe448
F20101221_AAAIYY habib_r_Page_160.tif
382f6dcd4f64e06fc4c9a5c5b33fee65
8a17c503ec734915c8d2ac868173a6e770455bab
5862 F20101221_AAALPF habib_r_Page_311thm.jpg
0f16cef9317fa85303282b6d30e065c1
0b16de565a72092bab26a2976778594fca79eb35
68304 F20101221_AAAKFA habib_r_Page_014.pro
41c2eaa5524e44ad43740cc049cad3a1
3e107eb1bd22d496d7e8f784179651857fcf8312
F20101221_AAAJVS habib_r_Page_023.tif
6ef21bc6110e80fd753b75ac6e75f020
f01b3572ba4b2fd75acd4385c4d88e2c08900764
682752 F20101221_AAAILO habib_r_Page_103.jp2
37686e1098776f53e2ccdda96b25e5bb
3863721b7bfcc2d7417b9aebcdca7b6a3143182c
110341 F20101221_AAAJIH habib_r_Page_225.jpg
97e9490488faa58bb736434cc6a921a8
49a2ab825b1786131053dd5021f433b6edc1a3b8
8891 F20101221_AAAIYZ habib_r_Page_153thm.jpg
fec8679c477bf993cf2184945f3a26f2
5ff798aa569bdfcff7935214057a345a1554c220
27813 F20101221_AAALPG habib_r_Page_312.QC.jpg
9bef21be41bc95962767aa679d7f683d
a02e333af91852e1082aa8136946b88532881bcb
2102 F20101221_AAAKSM habib_r_Page_155.txt
48287c78e863e53e18e7d71c60dcb7db
71a899be199e86cc59efd7d50b6c1a682f00ddff
17481 F20101221_AAAKFB habib_r_Page_017.pro
50d1f725925f2aa2b866106d5ad8e5c6
f3916c7ee10e759e16bda2c2b677a868d58334bb
F20101221_AAAJVT habib_r_Page_024.tif
a46dbb5c7cc242dd7d3a36def741bf87
e9f1ad3c58547f6f1caf1eb738d685c680c078f7
1836 F20101221_AAAILP habib_r_Page_020.txt
6a574a8788fdf0b3c781213bc56fab53
5a8090de28e9f6d4d6c294c09ce16a3102dc9745
106640 F20101221_AAAJII habib_r_Page_226.jpg
5de15c998423f09a0c8284700e269b76
aa1043ed27280feafa54d28e7a56e553c66aecb3
7890 F20101221_AAALPH habib_r_Page_312thm.jpg
abc89e69ad8ef0403ed354ef652ea6a4
fec2a37726893bdc47ba2e6dbf10dde0e0821dd6
F20101221_AAAKSN habib_r_Page_156.txt
222453f100ce1083bce4b77d55ee85dd
232585aa00b2c50309aa216cb9b482dccc14fe83
46638 F20101221_AAAKFC habib_r_Page_018.pro
0dc2123981398c16e0c8e2722f40d8e7
7917ff44cf67f4487ceb707d48f3f4996cd7ad4f
F20101221_AAAJVU habib_r_Page_025.tif
9d61e17da7371f5579dd124d3cd95d43
798f6e75191bfe922006f99d2681df90367e1a55
F20101221_AAAILQ habib_r_Page_192.jp2
012122a9a46c509a7167307ecc0f8ecb
06e76693dfc22a55b347484b6bd1bda22b32c6d7
100076 F20101221_AAAJIJ habib_r_Page_228.jpg
f71164e9918e28f7f2fec06d07689717
07c035502267026bfd1b43a5032e1b2a4fea60b8
1967 F20101221_AAAKSO habib_r_Page_158.txt
e67e7e839c33efb9ba4e8ba4aeb9d035
46eac824af1dbd86b1d9db0a67339de4f28c8c87
56778 F20101221_AAAKFD habib_r_Page_019.pro
afc2ef5481c23c9a9d1385343ad781de
c836c2676128f6a9078b944f29d92f73cdaea1fc
F20101221_AAAJVV habib_r_Page_026.tif
3e0d3e33f2ee583f99e6ce957bb3ff41
5d335ba6d00dc645b57a690f48f3f3c9cc378178
57082 F20101221_AAAILR habib_r_Page_036.pro
cd02d9a1fbe18f75228b3ee514c71a10
62a1ca8999cff0daeb9d4a286550166c8703b864
113697 F20101221_AAAJIK habib_r_Page_231.jpg
0247e0141fd829e903121b31b5125ccf
04ef389b4ae7a9b3a255e3bd414788fbee135b6a
4999 F20101221_AAALPI habib_r_Page_313thm.jpg
c28e2423343d72755de4962cd19ffdb5
6a4a3646e1ca00a1b1211c65fa0b63574a3114ab
2259 F20101221_AAAKSP habib_r_Page_159.txt
d95e64ca089737a035cfbe01df8d629a
a11902f18e6a7c4b0568da523c02015c6fabc4ca
46286 F20101221_AAAKFE habib_r_Page_020.pro
e1378f3a8c1dff1833d8a9e0963acc7b
9e6223f3c9c11fd988ef9a45bcaf0058633683d7
F20101221_AAAJVW habib_r_Page_027.tif
4f0141c01f1ab33c923a77ebf029a662
1c37c6cacc7dd7401b9962a97f58a3b97469118a
1051880 F20101221_AAAILS habib_r_Page_231.jp2
d58e07bf5a64bbac4954be262d85874e
d64b110ac41683e89b722b38c8e78658c9931298
95614 F20101221_AAAJIL habib_r_Page_232.jpg
fc2e063cdadcf781c30a1e8c9edbb116
b2acf7198b14cd712fbceeefa6b9336b546babbf
19302 F20101221_AAALPJ habib_r_Page_314.QC.jpg
0096f006f726b23d4243f5378af8d034
71e24573aa3f59ab2662b9f3c022334197678757
F20101221_AAAKSQ habib_r_Page_160.txt
fc6135c87d77fff1b39345424c4ccfe9
de8524b3452c9dec0b0caf02e0edd9296660641c
57114 F20101221_AAAKFF habib_r_Page_022.pro
21d312606579c541d963ddaf4bd5b95a
bce3cd4fc3d96199a832aee9194b84b0a9862faa
F20101221_AAAJVX habib_r_Page_028.tif
b21b46c92c1852e4a0bf530beb66707c
371383b0b4d7c95bec111a4d4b294ca390f7d2c9
7492 F20101221_AAAILT habib_r_Page_051thm.jpg
bfef0f77af859724c74c0de4e4819e70
5a3e8ef1f41626a6244e24da7fac3916ad63066b
102613 F20101221_AAAJIM habib_r_Page_234.jpg
0d5965b52c3c480d0e97b2edaca025d2
08da63bfb7e8cc573300e3cf1e0778055626f2a4
5562 F20101221_AAALPK habib_r_Page_314thm.jpg
4d95daf22f03269a8fcb05cefef1a09e
5c05719523023c02eeebfcdc8d0bfd6753387076
F20101221_AAAKSR habib_r_Page_162.txt
2d695a960259b4bf9dc0ee740b3c5ff3
f13150bb26977374e134abd0e74bad09c9da1a44
62208 F20101221_AAAKFG habib_r_Page_023.pro
26d15314c8da9a372d46a7fde7cf0f66
42a3ba5ca4623edb0791cc0cf6def6ff606a8921
F20101221_AAAJVY habib_r_Page_032.tif
34ec2e676a6fbad24f6cb70fb831bf37
c91bb453d4b74a5f007bcf97f377063fb406f5ac
47908 F20101221_AAAILU habib_r_Page_177.pro
943e64cc5974fd1f456c612b1c2b06fc
fbe68b75bd02325d23feaba7b3628130d5e55109
104596 F20101221_AAAJIN habib_r_Page_235.jpg
5a84f76f8ef8f79c3b26308f83b1e06d
3b064349f618744e34bfb45acc3b67361d7456f0
2693 F20101221_AAALPL habib_r_Page_316thm.jpg
23891a3e5a81f1c8da006915cd62b7b9
cdcd8d883175f4592f0e66502812222a527faa5b
9226 F20101221_AAALCA habib_r_Page_074thm.jpg
722f10c193614910e12c92402932c57d
fd98cb56775f44a7830342cc387cdf9c6fbe11ce
2202 F20101221_AAAKSS habib_r_Page_163.txt
48f902684348f19e6101e82f4049d533
85fccc38ba4073045426ddb4b6ae1d0378102738
57899 F20101221_AAAKFH habib_r_Page_024.pro
6d6ff987d970ec70de17485e73f16f6d
ba245e0da3cb3a960b9e24ce2a94df889358392b
F20101221_AAAJVZ habib_r_Page_033.tif
9335389fcaf3e4626c5408a058498aac
7c5c4315f5d7df534d085694b6c239e7fc4c1a85
F20101221_AAAILV habib_r_Page_059.tif
595ac24c0201a03f4338f4c2ac5d5ad9
4500ee5fa3015aaa131f68b3703d4490a2a80cc7
110186 F20101221_AAAJIO habib_r_Page_236.jpg
daaac51eff8e69431785f933571e1e90
4b96868127b3b7478ae296541c49602fe5b224e0
2766 F20101221_AAALPM habib_r_Page_318thm.jpg
c1b7de02dff51b0f8ba50a451f790bb8
ea95192ae22832507200d8f2077a6031cd493bf1
26804 F20101221_AAALCB habib_r_Page_075.QC.jpg
918b49a60ba219c908297b4c89d250bd
3e5cbd0c31c4dc69ed1ca2202c6bbad298673657
F20101221_AAAKST habib_r_Page_164.txt
7e85521509d1374b3feb1800ce4463c5
714f2ef6ad4ce3fa78a0c1249d59fa8291790692
54353 F20101221_AAAKFI habib_r_Page_026.pro
625ffca925fac88015f76cd60da6c55a
bd3f10122f30b200d0c08d288b794a2d8002d50e
1048 F20101221_AAAILW habib_r_Page_198.txt
11183ad35c67d90cb50010ff32612afa
444d2d7face1bfca746feaca707e3fd64524944e
99603 F20101221_AAAJIP habib_r_Page_237.jpg
ee143da6c1ca84ff35ca18b874b2fc76
c068cf6311691a1c5291fb600f5ab03a2b3ab5b6
33419 F20101221_AAALPN habib_r_Page_319.QC.jpg
8c04cc26e22672430748003e425b609d
7abfd6fb2a191337a58590d7c0a47d68370be062
7676 F20101221_AAALCC habib_r_Page_075thm.jpg
a6fc8c63b2ee07a35c6ee85e892f3d62
9340a5f67aa2c6d9ad02916c476d6943cb7bff4b
2027 F20101221_AAAKSU habib_r_Page_166.txt
99980ccdc683fc4138dab40e1c48e3c7
7225a07b89e311c947c0530a579138f339020204
58144 F20101221_AAAKFJ habib_r_Page_027.pro
26f9eb4e2b31f4161a6d5c5c208b4989
ad075f0b5485c0288efac8bb269cd4591a037c46
F20101221_AAAILX habib_r_Page_199.tif
22d1cdc4b7126092d693ad90c6901080
b95088048abe7f34a6f1e30de2acb03d2406c95c
107516 F20101221_AAAJIQ habib_r_Page_238.jpg
44fe4843dbefa2694ae91bdb90bcde0a
177db9d0450f7809499eb4686b628d5348836d26
8663 F20101221_AAALPO habib_r_Page_319thm.jpg
ab4ba69ad751e39fb87f47e8cedbe80d
672516b48de4c0810e1daf09bc984a24dc83393b
32719 F20101221_AAALCD habib_r_Page_076.QC.jpg
19576695a17c80f5139a7aeb2b0579df
1018c7c10b8c664a9f19915527daf30b2e857120
2139 F20101221_AAAKSV habib_r_Page_168.txt
f7fa56105efe08bd776e0c3175a696a8
c26fc649d21a1eae691ef837147430f44b9b8512
58314 F20101221_AAAKFK habib_r_Page_028.pro
4e9368ec66f1eb30c7b07ac817d4bbab
a135340904d297034cab2c203f9027fc2f3d5ae5
105545 F20101221_AAAILY habib_r_Page_200.jpg
71670e35575e7af25fd3a3b425726e45
4fbf3cf1c880650f8d91f4aaebf57c3ced15b90e
90282 F20101221_AAAJIR habib_r_Page_239.jpg
5e8e682a233530b8fe3c50425d4ff220
56cbd2294e4bfdba7276e48f60ef39c9aeaac05f
36091 F20101221_AAALPP habib_r_Page_320.QC.jpg
3256c2d1e13efdcf8b263870a473d87e
51405dac76a3ddd462af83cebb966e57c0b588ca
35678 F20101221_AAALCE habib_r_Page_077.QC.jpg
8b95819db046aa2993e19d9ccd54592b
bfeafdb73b27378536ea92aff82520b512971ff7
2046 F20101221_AAAKSW habib_r_Page_169.txt
1c967af791d7db3a6fe66e5333698ada
602462a436704772d3ef81f74e278879777f9fcc
57041 F20101221_AAAKFL habib_r_Page_029.pro
dc71d58aa699c9762f4e9fb135434928
cec0b2b4b040dd1bd50f3454426fff80759535d9
1007806 F20101221_AAAILZ habib_r_Page_117.jp2
c342f31d38fe23ccd8dce9200f741032
21ec5d343cd0a841307f7aa4ff4d05dad5b34f1a
109014 F20101221_AAAJIS habib_r_Page_241.jpg
8ef14b5dbc059f6edab5df43a11429ab
4e1e7e6acd58a871067e233f955c9d24fed10a7a
9293 F20101221_AAALPQ habib_r_Page_320thm.jpg
d3f549cd9e171362cac6e1b7d9d01f53
ba3f84ab342442f9f95aaf2f6babfea016f90502
8984 F20101221_AAALCF habib_r_Page_077thm.jpg
1c37e88f66b34ef453503128882eb389
d5517df3245e7c329cf5cd6d7417d488c5692d2d
2080 F20101221_AAAKSX habib_r_Page_170.txt
8c6b897ce606c4cd4715daff69d3e3da
16ad00062bc0271c493e7caa75fbc81b0f000c13
56870 F20101221_AAAKFM habib_r_Page_030.pro
34d71d5633f700ca758e6f59be0b6f26
7a7212e1a1d8fbca7fb7b85ad8f861fc3817633f
107983 F20101221_AAAJIT habib_r_Page_242.jpg
9acc28704c32c67eab7c5bcd6b34a5c4
6e8f9891f403fa34565e883610a317cf3e84aaa2
33318 F20101221_AAALPR habib_r_Page_321.QC.jpg
2ccf7f351d06ffe43a2f5893bc8111a1
0307d106f3100dea1b87de17026a12229e105f82
31254 F20101221_AAALCG habib_r_Page_078.QC.jpg
5903a16ce23c95fc8b19f8e8a3aa1003
9e2de6b1741790e03590a6adf4d3442b39d0f435
2285 F20101221_AAAKSY habib_r_Page_171.txt
8bf3dff78a4180c98b8c05f48d73638f
bffb454b07b127cb65e4384aa8e77c3dd7f77b44
55588 F20101221_AAAKFN habib_r_Page_032.pro
fd062a85d5cac409939c769964a08866
b13b63a7e51401dcbc3cd19a47c6d32c741ed813
94166 F20101221_AAAJIU habib_r_Page_244.jpg
b01380b0c7c5f2cb5ce9b70de21e38bc
7e8430e4e76c04e533f4ac22dec2d3cb2c7fb2c9
F20101221_AAALPS habib_r_Page_322thm.jpg
2453a23ee7ba0cc656c75529a6e7b297
45836e2eb26a14ff5842bd61b1413e575eafe25b
9063 F20101221_AAALCH habib_r_Page_078thm.jpg
899b9e23c85b43cda8423b4c7d15a979
c82cf7252b1a0460fd4446e0b710d2ca2088c64b
2061 F20101221_AAAKSZ habib_r_Page_172.txt
f89be6ee5fb29417a9f50385aaa50a69
3964133b1fb7282c66d02d4d47eeadc1ae9930f9
56832 F20101221_AAAKFO habib_r_Page_033.pro
78be25bd2d6ba914580b7f5c52ff4628
204304e148ad68ffdc0d28320006786d90c02d72
50136 F20101221_AAAJIV habib_r_Page_246.jpg
349136f2158c581a796738b9b21e57ed
97d2f5381e02204db13e2846153e44e2980cce49
34958 F20101221_AAALPT habib_r_Page_323.QC.jpg
5ad1153cd57972147b474238dda35472
a69874c1287653eb3dbd25606dd4054cad087b95
35622 F20101221_AAALCI habib_r_Page_079.QC.jpg
317329e5fd3e370cb6ceb786731585ce
fd0b27313f09f7fc494fdcc3bfe96c858f96e785
58690 F20101221_AAAKFP habib_r_Page_034.pro
ae7aa95e309ed814c3f0188fd073dbba
ca0d251295686ffe18973a114f68067f8c889596
50320 F20101221_AAAJIW habib_r_Page_248.jpg
99f903a04cf1d9b961e6737afc414cf5
3447027c195ac3a1b1dc6981d05b49d6d2821c00
9108 F20101221_AAALPU habib_r_Page_323thm.jpg
6d13ba6cb4eb31498ad30a60b9bfd23f
6e3d0c448426937c63597c15c52f00cbceda7265
8977 F20101221_AAALCJ habib_r_Page_079thm.jpg
782daec1714ded6d9cef355c46931be3
0fab2dcabdfb3e3ddd7d0f0d92f76f07e8750116
54644 F20101221_AAAKFQ habib_r_Page_035.pro
6a9ea2fd3d00e522a7404db4611a09a8
e68bc4ae4000c01a013e3b3f1b565a07bbddefad
40501 F20101221_AAAJIX habib_r_Page_249.jpg
5fb712fb52485740d0c2a964a6654a73
55552dff51ae07b81f8609b58ce61f30d9e7ebc4
35293 F20101221_AAALPV habib_r_Page_324.QC.jpg
9800da91efd8ba2391763df9ccbb47c4
0f86a8de663ad8571a456b0bb78178ceb25dbf1e
29036 F20101221_AAALCK habib_r_Page_080.QC.jpg
8682f173b634c7f9c7a7631b51975965
fb8235d985a650705c9cef5c0f99beb547dd7ea3
56180 F20101221_AAAKFR habib_r_Page_038.pro
3adf2c356382f90f4413833c9dc8b214
df6be5704b1cc249a85307cda3ec6dc81b9272f2
53077 F20101221_AAAJIY habib_r_Page_250.jpg
c8b8401d7101bfb39d343cf85cdd91fe
eb4bd11531e43b0bce3fc7470dabea6ff66d30ee
9103 F20101221_AAALPW habib_r_Page_324thm.jpg
bbf103fc5423430c6f00f67b982f1eba
44a4646ef2dd4503a6232f06ba3e6d81690dbed6
7407 F20101221_AAALCL habib_r_Page_080thm.jpg
50baa7f60298df8bb98073cd58ac5563
b63e4eac762b52cadb1eac6b56d9ffa826734eac
56457 F20101221_AAAKFS habib_r_Page_039.pro
28f2434e77028fdd49cf106dc63a007f
bf2f2f618463a63762bf1b3c53820d3e851d2ec7
45101 F20101221_AAAJIZ habib_r_Page_252.jpg
9eaded938dbfc28f4fcc1b81c55ba24b
842a460d822e6a6e16bea1c564a38c2f62fd49d4
35276 F20101221_AAALPX habib_r_Page_326.QC.jpg
7584bfc48e3becf7f4d16206b109b5fa
fce40b047baeeb8f1b70e9a835e22ce025e8cbc0
37266 F20101221_AAALCM habib_r_Page_081.QC.jpg
8a01c0a1682fba786abeaf551ad2f5f9
1ecfdf0117d6ef01f0c2b63af81bd8188927394b
57357 F20101221_AAAKFT habib_r_Page_041.pro
be13a7b7826d17c3e8983e42f3ebdff9
639f9da477d67c927b1992ef9ef2eac8d246f2ac
9270 F20101221_AAALPY habib_r_Page_326thm.jpg
ccc5b1957ac125eba881cd8dadcb8afa
6555f3f9b59c697846d7bd08fd493bfc03dcfbe4
9218 F20101221_AAALCN habib_r_Page_081thm.jpg
3c894e81937bd6f3eb4899567afe0b74
66bd72b288cde7d5eeddb4306e33f570552cc9b5
57034 F20101221_AAAKFU habib_r_Page_042.pro
c0ba52dff24fbcdf072b48e9ffa28834
0583c9ad8e1f6ff23a4e9267587f39eba375cfd5
22959 F20101221_AAAIRA habib_r_Page_265.pro
f8a33f12473840df5c4e890a95b2e2e6
f3cd093bb5f14ff56694d3f5c5bb2ac6b4497932
8616 F20101221_AAALPZ habib_r_Page_327thm.jpg
b61868093468be233b95f99cefea45c1
8c84664b195816c7c4475823468d232f7f408db6
39176 F20101221_AAALCO habib_r_Page_082.QC.jpg
630cce237de725504aec4919cd8ab29b
c3f6713913d96e35f494887e702fa6f215d782a4
57256 F20101221_AAAKFV habib_r_Page_043.pro
5243341905f7456874c186f0390a0fd8
bec105e13aac9526e3d98774c716736d3b70caa9
F20101221_AAAIRB habib_r_Page_025.txt
925e9363f7ed8e8f20377b4b9007e914
60456f51a470a65fca17363341e1cb9d6d7c038a
9068 F20101221_AAALCP habib_r_Page_082thm.jpg
7503fbb09ac9c8e419aae19f693dd3e7
3962c53d1f1ebca9ac3250118097f54a5e896de4
55781 F20101221_AAAKFW habib_r_Page_044.pro
96fee273618b2569700f48ef4b3f73b6
39107d028b2c2ac1c84ec314a1867f3d3c342d96
9132 F20101221_AAAIRC habib_r_Page_032thm.jpg
56890ce2342b61c539153b933c6054f9
fc8d3756f3fe00e31d1e29d7ad07c03a86fd5ff3
8645 F20101221_AAALCQ habib_r_Page_083thm.jpg
a59e08a64aba926afaae4018f3150dc2
4da80974b003ee335b21eb4315ea1d875d8c7ecd
2540 F20101221_AAAKYA habib_r_Page_003.QC.jpg
f33d75098bb91b47e1a4f224c41ec71e
8bbe9ce19555f4ca1aaa2506b3e1648bcbc54ed6
54791 F20101221_AAAKFX habib_r_Page_046.pro
1c257759ca64273ef8d5a66b10a92357
206d48f94b7cda90a14a38096673de90224b9038
F20101221_AAAIRD habib_r_Page_162.jp2
b616cdf5c488a9e37a73e25f11f2cf29
22647a637b2388dc432e4d3209c154d9a496773a
37582 F20101221_AAALCR habib_r_Page_084.QC.jpg
7938afeee43b51b753728f0b08b6ac82
8b1c9de40fb06ec81bdcc485f7002b4104ba1a37
835 F20101221_AAAKYB habib_r_Page_003thm.jpg
4b78351492d0cc30964183c2896b6f2a
ec251cc8604312becd09c4740121a23c7c93d31b
56883 F20101221_AAAKFY habib_r_Page_047.pro
45cc76e2bc2521b80a06a20fa21b1f9c
af8cb3cf480738c37b58e3d406de2bb828095d3e
F20101221_AAAIRE habib_r_Page_082.jp2
710f765875540c4a04a108ded88bdf4b
e479089edb30002f754bf8edcde94c725cb0f2d4
9323 F20101221_AAALCS habib_r_Page_084thm.jpg
bd28bbac50029783686b1b12e80562b0
4c5307f7fa26536a80a9442f9545229c128ba113
36758 F20101221_AAAKYC habib_r_Page_004.QC.jpg
2c7850af7d6c3552ac270551ef58cd17
fed996cf8bbbb5b46fd793489c465247719ef0e8
57481 F20101221_AAAKFZ habib_r_Page_048.pro
be686c180124e84c6afdb8e39e794ab4
cf910af97b1948a7cf7dc6f7dda17c2dd6b1a73d
8588 F20101221_AAAIRF habib_r_Page_183thm.jpg
3d5f7b8799b078c06ab53d29d949276c
b8878623de969f295bf40c07ae4f690c5b70f2c9
21871 F20101221_AAALCT habib_r_Page_085.QC.jpg
8a417f2c50ebafbf2927205cd791d3f8
d6d2aba34e2fb0e8e6cda705107f4da580aa6a8c
9027 F20101221_AAAKYD habib_r_Page_004thm.jpg
2b64cc034819aab2dadcd4a5bf64cae6
8d76c1af7b955d1263db3f806065fdb9c28efa79
5372 F20101221_AAAIRG habib_r_Page_257thm.jpg
287faecf0744bd4f61dc88f946c027c7
be47ef8c1e447a1dbcda2e8b530932725f2f545c
5451 F20101221_AAALCU habib_r_Page_085thm.jpg
f58ea37ab253b5c2096896717176a7c2
dee16d1795fa366829baa85f0e119b59f6f4cf3b
25305 F20101221_AAAKYE habib_r_Page_005.QC.jpg
8b721c31a37f0925c0eb4e3125b91a02
1c9420a3155b918e28a25aa8b8a6f5c0976e7733
92346 F20101221_AAAIRH habib_r_Page_007.pro
cd9b60abbccc021e8d0f49d22f9bb74d
54ab8fb1d869ad23e3726f2c9bf771d263f2bb53
F20101221_AAAJOA habib_r_Page_084.jp2
11c3e766629e205f9f70011e0f7e5e82
ebeab19b7f31cb608e9269633210ff5499bea427
18271 F20101221_AAALCV habib_r_Page_086.QC.jpg
8438a26e0d4750201f64b69fda5600af
0706d4f7c99fe1d2f8c262766c58cda87ed09e9f
6732 F20101221_AAAKYF habib_r_Page_005thm.jpg
c93670d4cff7abdb1733fe02354642d9
1d95a83f0e736e180852b83f1a64ed9c21293ebe
52067 F20101221_AAAIRI habib_r_Page_230.pro
436f15c37ada04c79ab32aa994c162ba
8d5429abd9784990b07442c36ef19468e9e2d9a3
1051934 F20101221_AAAJOB habib_r_Page_086.jp2
92b5c37ec85dbc0e48c4e7ee631854e3
2fd722ac14648399c52d877199b9c258cd1a24f6
32222 F20101221_AAALCW habib_r_Page_087.QC.jpg
1c345d1650e47d7620218cff0e3e7220
1ae14529290d5b23427eed252cb3d79fb55a3289
30007 F20101221_AAAKYG habib_r_Page_006.QC.jpg
9b1e9ac18cee32f429628f1d1b0e5ec6
f567412cbea5ee7209dc9d275b0fa3c31af2cbf5
54433 F20101221_AAAIRJ habib_r_Page_242.pro
6643951d74974339a43a91e7b77e38bb
4133dd4f90fafedfd78a391f5d2f6ff6af09f964
1051913 F20101221_AAAJOC habib_r_Page_087.jp2
339589b2cff7ade44f6bdd6d372ce7a0
664894b85fc85e63e16bc49e8bee0142c22240f7
8534 F20101221_AAALCX habib_r_Page_087thm.jpg
81cfd5614f267ba6593b1fe55d7812cb
fb28e043dd2d52fbabedf3c960a26f078ba36902
38860 F20101221_AAAKYH habib_r_Page_007.QC.jpg
1c2bd082bca61d18071d50b063b40232
92ab197e1ce9f41a6cbea776a0515cffd5104c20
115010 F20101221_AAAIRK habib_r_Page_207.jpg
c60eec1cec6796a76a3c13d53ca28c7b
560ea16e43812e72803ef68dfc14f3bc2cb311ff
F20101221_AAAJOD habib_r_Page_088.jp2
a4a0e0d24bc11991650c14ebb1a37a9c
bd91febccbae59365b127ae5900388417a8432fc
32287 F20101221_AAALCY habib_r_Page_088.QC.jpg
4bc7269ca0903a3b8e24f23903ab8482
e5ef7288954e68aa3b71521c5f5d2614a1f2ab79
F20101221_AAAKYI habib_r_Page_007thm.jpg
14daa111e9e64f681575b497fe66b02b
db6845a36f2df00bdf8ab24728a89d137fbdd04f
F20101221_AAAIRL habib_r_Page_276.tif
bb39f6041656a5a575cb9e6208e44396
7a8358914246e144128a74a7f0ac6be24a857977
F20101221_AAAJOE habib_r_Page_089.jp2
9b044f2781143d7262de48418421dae7
953a41af2547b689b27f67d5bde920be7fd45d52
8235 F20101221_AAALCZ habib_r_Page_088thm.jpg
266a6d59725801464e077b303e1e2c34
703d2ff432f01a74dd7bcc971063a32e708bc56a
9209 F20101221_AAAKYJ habib_r_Page_008thm.jpg
64faf2a4cc78ae053f4d30e2fb334d4b
925fb74c9bc81d9e4cf80fa60404fc37f320147e
45869 F20101221_AAAIRM habib_r_Page_193.jpg
ae51de76c44b4d44a75aea3a04905d6c
08470cffc06429e033e01fa42f05f03768c5430f
1051950 F20101221_AAAJOF habib_r_Page_091.jp2
bda2ec5fe9d0f2936887e5371ee74f9a
9dcdb8a247a7978198bb88bb849590c819dee9ec
30787 F20101221_AAAKYK habib_r_Page_009.QC.jpg
3da23c54bfbcadbf33c257312d3467ae
8d784a3f09962a9510a2d47835904db02966efb9
F20101221_AAAJOG habib_r_Page_092.jp2
9b6a659d955c2e0e7fad766ac73bd2fb
f3fde2636c9d61cde2e06c25a0b23e1dcc6af94c
9457 F20101221_AAAKYL habib_r_Page_010thm.jpg
a0fc3ace60c34aaaa47c6ef5df3f6a4a
37d62999dcd877fc60706799cee2346bb28cf43b
1401 F20101221_AAAIRN habib_r_Page_265.txt
bf8839af4b984bc7b293b98c3f862834
6d53aa08db94664f229cb89b7bb95d911b1d4646
1012357 F20101221_AAAJOH habib_r_Page_093.jp2
b04470ee7fa03bffa4532eaacb0afd02
89e84ee02084cebe49fece2f51fe388e10d26b52
53377 F20101221_AAAKLA habib_r_Page_219.pro
24d08c232f21bdba98d7f16ce484daad
c29d173eff186e91eb0077fb8a2c2de06590a102
3703 F20101221_AAAKYM habib_r_Page_011.QC.jpg
9afbd1575ddc2c402d5055252040b3fb
f0f97493aab390b647236b155637e5d27dd3582b
13710 F20101221_AAAIRO habib_r_Page_276.QC.jpg
d05a8fc0b77c87c24eedf83a935549df
95b294b6e9b35c241da115705b457bc7f159f4dd
951750 F20101221_AAAJOI habib_r_Page_094.jp2
1f97b50085ef5ad26c1ab238849ed68f
1305d7e0e14e9533ab79e079d801704a36cfe22b
55262 F20101221_AAAKLB habib_r_Page_220.pro
7ebeaabc708c2a456f19b1a9b40c12d6
4e539b032b8b981364ea880b1190028505df7df2
26005 F20101221_AAAIED habib_r_Page_305.pro
1de750e0acbc7ad0ebdbcd704d1b0c22
3622db518516cba2c082f74f697d4efae9bae2dd
1125 F20101221_AAAKYN habib_r_Page_011thm.jpg
cbf9913cbf5b5e641e02c5b5ccb7bef3
6a2ecbbaaff2d46e4ce78d7656b32813f372ebb2
5691 F20101221_AAAIRP habib_r_Page_202thm.jpg
90cdd22972fe6e59b383ad9e6b4988dc
cf19876843602079e5a60ce367c72b9951091a12
61457 F20101221_AAAKLC habib_r_Page_221.pro
2395d67f5b4cb81596fd9f0c9de9e7ce
48d1c76cc46a9660d5875d159612c4743da4e696
2392 F20101221_AAAIEE habib_r_Page_319.txt
9b3f72c0b7d033421a3173432e76da4a
8e142ed7484e50ebb3acf40f3fb4b911e67add80
35693 F20101221_AAAKYO habib_r_Page_012.QC.jpg
b6d405d9e0a06316a10cf309d1f59a80
3342785d726c00ffb890d4feef614747af69d1e0
703728 F20101221_AAAIRQ habib_r_Page_085.jp2
a84728de76b12e2f55b130d6e368c264
ccd7405aca6d65e5569d866fdcc40cbcfdb37c16
1051959 F20101221_AAAJOJ habib_r_Page_097.jp2
d858a0cad8bfeabc2ceca32e82466b1b
256bf6ec908aed77285f6f59d6e5c61748b94a78
57148 F20101221_AAAKLD habib_r_Page_222.pro
bbada49b87441afbbcf025f2784b8554
16e0b814ab8c73d68fd22b63fc1218490b0175bc
695 F20101221_AAAIEF habib_r_Page_293.txt
c7f136bea769e473d166900cda6b9e50
7e53a104e031dbd800f9a2848d07915430431c20
8978 F20101221_AAAKYP habib_r_Page_012thm.jpg
a6f55684acb003d7b2b58a51f76f31df
567f3d31c1aa0180e9f4e818964737fc4b9a609f
59003 F20101221_AAAIRR habib_r_Page_186.pro
fcbc59f236f4ae4f5d00b0652d285b58
41eb4b29d10af302e968611b52f090fba3ae539a
875624 F20101221_AAAJOK habib_r_Page_098.jp2
8206db3ff4764d155ad00a5bcd56c0b0
1bb9991315a4f3875fa6ab4d4edc98c12c6aedbd
50059 F20101221_AAAKLE habib_r_Page_224.pro
b8a009b8269023e41ec10f599cc7ad9f
68433ab1eadf52bef9c269784b1c1e7baa5fe7c1
F20101221_AAAIEG habib_r_Page_126.jp2
5e553312a020d686e790d68ff48f9acb
af243bfd1708bef5c68fb24b29683670271b56e4
43496 F20101221_AAAKYQ habib_r_Page_013.QC.jpg
129c34b3a2b0852bf042c0106e4e17ea
ebf23abb8a81ae6dc074e9c7678cafb2cd02fa20
8198 F20101221_AAAJBA habib_r_Page_262.pro
4f999a1b48d8f739be4745cbb1e3111d
21c9e286fee7163131b0a8b5dbfa5f6b2ad459b9
2111 F20101221_AAAIRS habib_r_Page_128.txt
08aa3289e91aca05d13324e065ddae36
88ce0ef3026957a636a0755a47b396cf2262f1d7
1039632 F20101221_AAAJOL habib_r_Page_099.jp2
9b1deab2418b86b51ad216edcae2c80a
fcd2ad5430ce8cb408585a0ac7d315a4beb01de0
36370 F20101221_AAAIEH habib_r_Page_066.QC.jpg
4c7dd3cc1fe41a034ed5ecb045785eaa
45b62fc6de5265a5ed22dbe7f3652c8f5b4d8fa4
10304 F20101221_AAAKYR habib_r_Page_013thm.jpg
fc0b37369e6216c2f87d646883743337
a147b83fe98fdd16f26815c1790c684f27814daa
2099 F20101221_AAAJBB habib_r_Page_118.txt
cd5bc9709395a2f236deac1d5fe473b3
9f8673d78be73ea8beb74748cab112f89e666484
97305 F20101221_AAAIRT habib_r_Page_137.jpg
d6c7d048f41a8fa68e8675986575366e
5140ff62a39e2676362e59db36b5d23194eb1458
965629 F20101221_AAAJOM habib_r_Page_104.jp2
eedd29373be34171921a30ae72afb2a7
e24ce9586fd1acadad75b6d696ddc48b74caea95
54693 F20101221_AAAKLF habib_r_Page_225.pro
e60e3f31d97787f38a9ae06c68cad043
278ef0af85762a96d1755a5eedd84802fb3c6742
8615 F20101221_AAAIEI habib_r_Page_181thm.jpg
4188061310c276418b5d93d3a738a0ae
22a2b137a1500e15021a253a2e7f0a8ef4f47823
F20101221_AAALIA habib_r_Page_178thm.jpg
c3e9b6baae9209729921d4080be4f75c
b03a48418a9a0b9729a9aa2599be8333e49dbc96
36713 F20101221_AAAJBC habib_r_Page_142.QC.jpg
2f6696febda40f9610739a3440d4247d
041d37cae79fe8171a5cdc80cedbec716a479fa4
F20101221_AAAIRU habib_r_Page_291.tif
e49a67cdb9ac86ea6082cf0233586ecd
6a40022852d78608b06467cb6b175a3c17d5dc92
1051914 F20101221_AAAJON habib_r_Page_105.jp2
199f1db1af82d1dce7021d58e1fd2b46
983e409639b68864db7fe17b509c4fbfabf095c9
52222 F20101221_AAAKLG habib_r_Page_226.pro
ce3fb44a24bfedce962b0a0a22866d0f
fba48716113af40c51dfbf284850bdbe97107ee3
4574 F20101221_AAAIEJ habib_r_Page_279.pro
e5bea528e8f6a0705df874de43bd73e7
9fd11ca50bd684a79882c43ec4d6763d8003d35a
10458 F20101221_AAAKYS habib_r_Page_015.QC.jpg
a237ecf96510ba449a024cf42d437784
ff49d6e39754c84083316e4f001bceab946080ce
422941 F20101221_AAAJBD habib_r_Page_313.jp2
7277b5cf3a1db7b00f600b970df93d94
7dfc90b3cba8475d8bcbe435dec866eb763af109
14627 F20101221_AAAIRV habib_r_Page_015.pro
f931b53ff5cf5cfdbb7f615f34381d21
45792318bccc06edf4dcee74a9a5d68d4450bcf9
F20101221_AAAJOO habib_r_Page_106.jp2
38c9cfb2c3723c234024aeb0cd0d8db4
dcc782e7d6be2e7e024d861c6613a812a9e1bfb4
51846 F20101221_AAAKLH habib_r_Page_228.pro
7cca5d2caede775f8b3e24043386eb36
f9820d6cd3e7b0c87dc27945c8fbc7395dfb5cf2
F20101221_AAAIEK habib_r_Page_077.tif
302771dd7548e92cddcb568784017de7
655107c77cf0adf3d6e565949924e6b49cbbedca
32263 F20101221_AAALIB habib_r_Page_179.QC.jpg
c6d295c427938a6c53e5b4e2334d6d3c
a383438cb4fd4c4a6bcecb4239c2dcf37971ce3b
2585 F20101221_AAAKYT habib_r_Page_015thm.jpg
ce2801588eb483ab10273f460d0b3bcc
94675cc4ab9fea9213ea2948925389d79b58757c
93973 F20101221_AAAJBE habib_r_Page_240.jpg
5f12b091468ceb885050796bc8087f61
be4cf1da790cd31d46bc11834aac5a854aeb9b7c
34539 F20101221_AAAIRW habib_r_Page_083.QC.jpg
43d3b346edd64fe32343fa6e56f898b2
81c7b459d036556e20113fd2f29fecea15356cb4
F20101221_AAAJOP habib_r_Page_107.jp2
01e7c0ab58bdd55ad8d33d6bc1bba44c
0f63b1190d8ea47d39f1916c0ad92129a9601c11
53949 F20101221_AAAKLI habib_r_Page_229.pro
02544bd62f987f6f6522dc56009d7c22
f1ba5e96a480ee534b8701b2394f3f33bf5b5653
35759 F20101221_AAAIEL habib_r_Page_062.QC.jpg
a9c55c0d05035ab102a5f2e0e9175613
e9e43bf380e40c64e8ccb941ee5d74e127e71068
8349 F20101221_AAALIC habib_r_Page_179thm.jpg
9cddafe61e0132cf45a6fe1df59da404
73ef71b22476c3e0f941598d80bb5f65dda5e23b
30706 F20101221_AAAKYU habib_r_Page_016.QC.jpg
290bb1fd6ae63237c0fee7b003ef51bc
37bef7e92a377b2e65bef846e9950e5d07148b2a
52579 F20101221_AAAJBF habib_r_Page_140.pro
a28d8b44f4f83c016d599eadf6a6b812
5fa7667bde083aa92de7f7dce557b81a91f0652e
24849 F20101221_AAAIRX habib_r_Page_262.jpg
fbbddf32185a2d24a340a839289ab6b5
f286ea3500b069def397e422fbe5f5bc437710bb
F20101221_AAAJOQ habib_r_Page_108.jp2
c68bb42a08d30e4d03194dd51d9ff567
990fd9250b5c35e437d42384fef4c23c32f8c639
56537 F20101221_AAAKLJ habib_r_Page_231.pro
04023b1e5f2bf5df1521625cebff842a
53eaefb3a615ab409da382311d7a64c393b5b710
21358 F20101221_AAAIEM habib_r_Page_202.QC.jpg
7bac80e5b5f412c16ffcc8f10397a437
94f116da56a87c3f4ae5ce3bdf242a81b3caa86b
8310 F20101221_AAALID habib_r_Page_180thm.jpg
09a771dd83614176817e9a1d42587c9e
615c926d8908611892c8b9464bb7dbde103727e2
7398 F20101221_AAAKYV habib_r_Page_016thm.jpg
8df2e0074b2cc8fbd7db93bb6c0e72f2
cc9d1990a6c1e42b6f592b59621a5f209a19f5d6
F20101221_AAAJBG habib_r_Page_136.tif
6e91c56283bafaed256636332460e491
a7a86d8b865943af7e9afc3d622d0a7effb0d408
47542 F20101221_AAAIRY habib_r_Page_127.pro
7ae9ea20d3f838a029826bf0e24b4f92
1326b95378eb23510eacf78f80a292818ef439b2
952396 F20101221_AAAJOR habib_r_Page_110.jp2
aa7988cb43dfdb09fc490aebe99284f1
b9ee137dea143be84e9d081d166ac897d876d72a
51806 F20101221_AAAKLK habib_r_Page_234.pro
c52e9d8ca40ff83eebaf6f91fff6d913
7693c404bc65d8aac66e22d1eb86f9537c5ccaed
98704 F20101221_AAAIEN habib_r_Page_118.jpg
b72c600107b89613f3bbfba6c50308f9
549d1c4621720aa2550d4fb57009f54515066a16
32970 F20101221_AAALIE habib_r_Page_181.QC.jpg
404dbbf65dbd98236b8e923c7ceba704
8a8ac3b653b191107e0650fa81093da801502b5e
4066 F20101221_AAAKYW habib_r_Page_017thm.jpg
d5699ca36d4981a64841a71ad72b6b2d
b7e95b81adefc39e7f291c6cc98d7442f959e175
99034 F20101221_AAAJBH habib_r_Page_127.jpg
b308d3ee3f7d087bf08d5e6e9cd65997
511a2117e80f2d72b8a07963031e52c07ce5d89c
6189 F20101221_AAAIRZ habib_r_Page_197thm.jpg
36c34c4be1e08fe4a868c80cd5f4760e
145384d389bd1f746306141a9072c1f3ebd9c229
1051896 F20101221_AAAJOS habib_r_Page_111.jp2
5bca5ae0a39eb6e7abd0f3104edc0bf8
3f0d15b41f354a4959e3d6043b9ffe4859532e28
55294 F20101221_AAAKLL habib_r_Page_236.pro
af012c6148c80de2cb70115a726d79cf
ae8df6c0e3513f316a8008563b1a00b622d752bf
49932 F20101221_AAAIEO habib_r_Page_130.pro
22807790d5b4ea244001c35c378d0215
4b59716be736f103758552d84a9f75f859cfa0b1
31666 F20101221_AAALIF habib_r_Page_182.QC.jpg
62dd5718538f5e4ba54447fb9c92975b
3e1aeead096231a5e9a21c7b2e929271f01b8a03
29660 F20101221_AAAKYX habib_r_Page_018.QC.jpg
168ede0c1a8aa624a1fa96d8db606796
3c4166f49e82d2fca79edc9e9aa61f430b944968
1623 F20101221_AAAJBI habib_r_Page_157.txt
5f90e052827a79b3a9977deebd31f474
b72ef5e1334cf4cd7741e56e1b356c5d9f0ba37a
F20101221_AAAJOT habib_r_Page_112.jp2
60b49df68326d2dc593f0bdaa2a9f7b0
0865b08f8e0c9382c587d486cbfe9236b3c87858
49209 F20101221_AAAKLM habib_r_Page_237.pro
7e737255c53bc6347c5f0d37ac667b4a
50c9373ee172fdf62a541975005ed9e7e921125d
37978 F20101221_AAAIEP habib_r_Page_203.QC.jpg
5e9e8003b35d1c0dc5e4dcd398aa68a0
53077d08c07b6a9a661be4b752afa1ca51ac454b
F20101221_AAALIG habib_r_Page_182thm.jpg
36bdaaa7ea332f201be0219d2645fbb4
7acd86cb25e6b1c54937dfaadd36d8021f646e7d
7619 F20101221_AAAKYY habib_r_Page_018thm.jpg
07acba9a1a1ecce31274415e3a8a26fa
5bce2cca37b9ed7b8980a1076aeb1ec8803dd151
F20101221_AAAJBJ habib_r_Page_203.jp2
9282261f2db01b5da43085f49782b29e
33c56a04813f304719d8f2220c7eb5732fa9721d
1051945 F20101221_AAAJOU habib_r_Page_113.jp2
29ffe09f7fb99bf99344c624024da410
1f3b716d7ad44ed5b922b1f3cb5c4640fa221503
54312 F20101221_AAAKLN habib_r_Page_238.pro
ce4fd6ff9c88cbff521b89a183423edf
238b9008169adc6092bf45f32e03ddeddf4bf564
47189 F20101221_AAAIEQ habib_r_Page_232.pro
65bcb2990ea7f42a6ca1e3083b79be66
78a08e4b0892b0ebc9156270d2d37b43279ae0d5
31948 F20101221_AAALIH habib_r_Page_185.QC.jpg
cc1c411bfd2fc9c5ba6416d0f1342ec2
f8e89aa9ed37f9aaebed7312515ffb5fdf208ce6
38273 F20101221_AAAKYZ habib_r_Page_019.QC.jpg
5f87f3a95319a1743f04f1e7332c4e29
faee377235dff66b65086e88a9d0da72870ec6ee
110803 F20101221_AAAJBK habib_r_Page_065.jpg
f71cd94af754576528f3d619f8884387
540d8d8475f0076fb6177b315a03d58c3217b169
1051836 F20101221_AAAJOV habib_r_Page_114.jp2
94b80a241d1330225b4843fa3e6c80a2
5a53de723b63987d98c8a6718a2d79ef10605642
46899 F20101221_AAAKLO habib_r_Page_240.pro
5259dd2ccca26e9da7ceda12ccd9cf4f
ec58cc296b5adf26a8fcbad3774a13db45e83ccb
35267 F20101221_AAAIER habib_r_Page_220.QC.jpg
36fc67ffb270fbff0f272560f77b2809
eb6cf5fad64178e0387bd45e5f05a26fea73226f
9022 F20101221_AAALII habib_r_Page_185thm.jpg
c7050aa97a4a269b184131222f46967f
e53e622939782034eb2c262c5228ba49b7c0426b
9465 F20101221_AAAJBL habib_r_Page_129thm.jpg
0376a4f8e299fdc0b04fb9eaa074d52b
5abfcadddf4f5524d982f114c4ce2ad3ac5d96a1
F20101221_AAAJOW habib_r_Page_115.jp2
07299126552c550f5875d03768b8bd85
b1f5c627f3503b036f62620da76cd804b5785d8d
49665 F20101221_AAAKLP habib_r_Page_243.pro
71b5c0d6a72216f3d1d10afdcc61eafc
7367f679f68b43953789c5236dfa0527504cdeb9
334821 F20101221_AAAIES habib_r_Page_306.jp2
74c9915630d1c6c47b85c1cff64b2912
0bd54ca5be8503cafc7eae2604dfd92dfc1c6e4c
36490 F20101221_AAALIJ habib_r_Page_186.QC.jpg
96257ea00b259d91da23a55805f8062a
e756464411360395677b238d595be6a6f4eea38e
120167 F20101221_AAAJBM habib_r_Page_331.jpg
23399705cdff8af8fc511c34bec7bbb8
f05d5810493c6c1df1087fc4203215b3cb00dd9c
F20101221_AAAJOX habib_r_Page_116.jp2
2e46a1560baf50ff4bc956101647eb05
701555a70eda16369dc4245d10804d3d2b159d37
45667 F20101221_AAAKLQ habib_r_Page_244.pro
2c1328e4549e4316488ad4364c4d3160
7cce26e8aec804ef1f354a9c073b80e321f20027
32136 F20101221_AAAIET habib_r_Page_252.pro
5a73b4658abd877dccafcf3b04124698
fe1957d504fe354e1b20adb1ff2f24863c8fcd9a
9082 F20101221_AAALIK habib_r_Page_187thm.jpg
3fd38c091c3d2972a102581d89e86ca1
9b1f1f0bb3831a50137256e391781af011188ace
7967 F20101221_AAAJBN habib_r_Page_119thm.jpg
5ac3ff1c7788a654e134015e5509306f
875cb2780d584851f814464560ac512e320c079c
1034162 F20101221_AAAJOY habib_r_Page_119.jp2
c38f35f8a49d51d3b00da793c62892f0
cc480340ddbf4d6fd479fa69dbad1f706c632aee
22472 F20101221_AAAKLR habib_r_Page_245.pro
18bc95606452fd8eb38ed45dc1350fe2
920bcb91f37ac4124f89ee35429d4f925961b375
F20101221_AAAIEU habib_r_Page_083.tif
65d6b0a8e063b7bf02c01fd53798d2af
31b68dca53e44b16058387ea0e35cb35d3491d12
35420 F20101221_AAALIL habib_r_Page_188.QC.jpg
0d749c705a32cb4c20d6c8290f7d9de3
d22788e511d4a63d0399dd8298bfb286f00c037d
2179 F20101221_AAAJBO habib_r_Page_136.txt
03095260390a8ad6f834b0b14e1f24e8
1ba4fa9fea245bcddae9545844ead8f8516fde7b
893836 F20101221_AAAJOZ habib_r_Page_121.jp2
00b964556f8c3d2ebe1a29a9c2e95bd4
7bd2f6d93ef649733355bca59bf9549f6a29898b
23697 F20101221_AAAKLS habib_r_Page_246.pro
c8c96f0b11309b43c2987e3ddb5bf7f2
c34baca435e7972a6bb7f913d5c1c994e64f4fe6
2263 F20101221_AAAIEV habib_r_Page_052.txt
42fba2a4ebbac138e40e4c394bacee9e
69abbba04400bd1f1b6d1239029811030e6097c0
8466 F20101221_AAALIM habib_r_Page_188thm.jpg
8d40135c00ce32943acae122659aeb3b
32cd3dd3e61663119bf4f15c404ae4d768d4d261
37420 F20101221_AAAJBP habib_r_Page_070.QC.jpg
a59f76b6b4ae3eb285529610ecf10f8d
f517c6f795d7e9e33eaa58434c43ad18f150fbb4
27768 F20101221_AAAKLT habib_r_Page_247.pro
ef97b2d1b9a49bf35ab8f5018e62a9c9
e394d5572524fe2247544fe8a1c8a7bf82dce2d0
27346 F20101221_AAAIEW habib_r_Page_290.pro
6ea96ca4b597d6ff9c24f3c7baccba49
07fe01499453d1778b0591ba585d3bbdfbf56162
35891 F20101221_AAALIN habib_r_Page_189.QC.jpg
90ecce162b0f3c495edc2439c9e0a93c
be35598b0b2913cc96240506f87156c27a985769
531674 F20101221_AAAIXA habib_r_Page_314.jp2
d007f61f2c9fb5e40790ba9e3b050c7e
259abd17d988514328caf9ca5a95f9ae098ac638
8631 F20101221_AAAJBQ habib_r_Page_071thm.jpg
52539e5c4aa8c223eb3f8f62cb6c6d86
56521c46d25c08b4b5679c8f8cacf3b5b2158f6a
22910 F20101221_AAAKLU habib_r_Page_248.pro
4c4d4101b0472e02891f2fda8db8922e
86c020b78eaf8e2d3d844f3bf321d442393ee7b0
106603 F20101221_AAAIEX habib_r_Page_077.jpg
3ed73cd000468ea6dca31a13ea23e54f
5980e1d88f943ef14004e1a80438749636b88a05
F20101221_AAALIO habib_r_Page_190.QC.jpg
2a1c2b655f0592dc37016d730a6b9f96
9c135a586a00c78f8526e1a7739d846f78309a8b
32224 F20101221_AAAIXB habib_r_Page_085.pro
786ba728ddcb107ff6e3b8ccf0c73dd6
e7e00200cfadfefa331b77a76ee15bd7df85b7fa
23405 F20101221_AAAJBR habib_r_Page_280.pro
05cd346a994fe3072efefc4ec33db7ad
9ea4b9253ad90198fb5009afcbf9e44f49f0214d
19765 F20101221_AAAKLV habib_r_Page_249.pro
b2cb1863a0f8e05c1264d9a61acbb123
d10f99f1ad6a1fba79e89dce1b3cafc5b99697a3
F20101221_AAAIEY habib_r_Page_235.jp2
609200fcfa7ecda447cc865833595f82
0a50dd7eb0c5572ccc4cee76d60515c4f4f13a7b
9312 F20101221_AAALIP habib_r_Page_190thm.jpg
59f835d1eb3fa05ec1d1ccbab8448601
75eb662a547f361858b4cdbc83e123a3baead5e8
10095 F20101221_AAAIXC habib_r_Page_254.QC.jpg
dc67e48e7142abcd99f61718e22a7662
395ab209276b74c208895f5c47fa2c153d43483a
F20101221_AAAJBS habib_r_Page_007.tif
26507cd42ee402f67f2a3844eed5c25f
298f18792ef1aa1d854c9ed88c4a00151ce1d3cd
23686 F20101221_AAAKLW habib_r_Page_250.pro
d59d91e267251a12ed9c534f32645d17
559a066f4fcbeecce60a7be28d765915d582bc97
36081 F20101221_AAAIEZ habib_r_Page_176.QC.jpg
a06d8c92803a6df6bf44eadc23cfcc7e
4a1c501608656bd4ea87421d6b00c14cea4fa681
36527 F20101221_AAALIQ habib_r_Page_191.QC.jpg
f9f2406e55a13aa77625333f51cb9f34
e62f9f2acbbf4662af55a222e1ba9893929f0912
36495 F20101221_AAAIXD habib_r_Page_241.QC.jpg
f0878eae19cac66b4b0ad4cf3fa9582e
a469b9fbffa74d41d9becf5f87718ed20febe69f
F20101221_AAAJBT habib_r_Page_177.tif
b6fb4087ca73459ed6e84e359199a7bf
cb7690ec7141cee7ac5d4f6040d9a3ebf1a47cf7
F20101221_AAAKLX habib_r_Page_253.pro
7f52a89fd52a983d06115e10373490d8
43a57ec680325a3156f83938546985432c253e21
9155 F20101221_AAALIR habib_r_Page_191thm.jpg
98a9c822c51c63410d20e7104233cad2
59594533793d2ecf7127ae1b7a46fc6122b5d58c
F20101221_AAAIXE habib_r_Page_118.jp2
816b816e33252911dd39cc7dbda7f1ad
2a274de692cd9e078c4a0ad8fc37b6665ac1f46b
3646 F20101221_AAAJBU habib_r_Page_002.jpg
0b257bfa63fc62823eba9cd565151c48
196615e129bd490c68872d9762ca2122e220334f
9366 F20101221_AAAKLY habib_r_Page_254.pro
1fc82d9045f8c68f89954388e5867eec
10da2939b5eb0ad3b99e63787adc00e33ebeb59b
8934 F20101221_AAALIS habib_r_Page_192thm.jpg
7ab1542d6d4247246ba81dfabef97de8
b39aa7422c649d7830d57f5dfe99a3a467e9c809
2085 F20101221_AAAIXF habib_r_Page_059.txt
6bbddbb6aa5108727ef51e250cd82675
1d62122e68e62a976d2473b75ddbaa468efff01c
494558 F20101221_AAAJBV UFE0022375_00001.xml
8e4db5b877d0467e926c0049ee2fe6da
8205438e47781a6150ed121749b40a23a3fedabb
7431 F20101221_AAAKLZ habib_r_Page_256.pro
96ea0a6a27073d3afe5a5cbd7885d8f5
f333f95d3484db4881f1750e4eb6a826ca86bd23
15304 F20101221_AAALIT habib_r_Page_193.QC.jpg
16a16d136a8927d5c305231c3d949164
f5392bcd8f8d24788d581be56c281a6f722b9de2
F20101221_AAAIXG habib_r_Page_189thm.jpg
4f8964db42b606ce98c3fbc9fc949020
8be2ba7fd83332609a25d793dc44bc8039afb726
3708 F20101221_AAALIU habib_r_Page_193thm.jpg
a1b4c15a80219bf79e812e1bc6a66682
0970851174a80f4ca4abd788f09c2f9f8245eb5a
108997 F20101221_AAAIXH habib_r_Page_168.jpg
2b1a94b09fe882e7ff3ab390a849d8c1
eb95356620bbf6aa3b3eb2728914db6460901472
473310 F20101221_AAAJUA habib_r_Page_294.jp2
1f303cd87c76b776b2156238038286a7
5949dcc97f4978fa8a344cd456fc83099b83d5be
25718 F20101221_AAALIV habib_r_Page_194.QC.jpg
c2b7ef162b48e1024480d0a6fb0388c5
f8837663232da6506c5e32e62b5aee7ec6bf5633
6761 F20101221_AAAIXI habib_r_Page_317thm.jpg
7abb23d5be2f1f3c383fdf4ec5601493
7bb74d2207d562d4c5b0ea4d6c619b9ce34f0d2f
27689 F20101221_AAAJBY habib_r_Page_001.jpg
24c7599033477dc847a3bf72251c89cf
7eb373e812b6b8bf046238958a5800bdfd894cec
521318 F20101221_AAAJUB habib_r_Page_295.jp2
431a2968d181b30f0e2b98ee08cb9b06
3164b3449d3e86220b1ee0adb06bcb4c0d27e51c
21605 F20101221_AAALIW habib_r_Page_195.QC.jpg
a636bd66e7fe9ca893e4d257cdea56ba
4a03857086c0ec19cce7134ed2bad95c0e984495
193354 F20101221_AAAIXJ habib_r_Page_256.jp2
2469afec5e815c0e06052439b0c45853
ded85ac3f920ea38d3ba199050dcd7956a05aa5f
113305 F20101221_AAAJBZ habib_r_Page_004.jpg
56932251f34ef8a1b471193eadc9977f
2bcd9dca5960b4bd0a944979923057ff5bbefa66
434133 F20101221_AAAJUC habib_r_Page_297.jp2
67456fde23d6d04eb6d74de89393734d
513a464f264514607b2cf592ce8cba8c99642330
6207 F20101221_AAALIX habib_r_Page_195thm.jpg
1177c9a93ae6320f6a21e2a413da80f4
1006a59f391a667d7737d5b4b3624a75af2ddec1
F20101221_AAAIXK habib_r_Page_325.jp2
997ca9942a32342b0a6b742db881313c
819424211e015a668de56d659bf03276660e7d40
551335 F20101221_AAAJUD habib_r_Page_298.jp2
d59ed2d7c380f32e82a9ec847509f508
2882df1566336aad11ab8e704a47d7f181139c33
19880 F20101221_AAALIY habib_r_Page_196.QC.jpg
ae574560e585120114c9124cb8c1670f
d8503730e7fd5b43d7817b1def47bc439de9e97c
F20101221_AAAIXL habib_r_Page_171.tif
eefdc28809737e0a474d3a23ed181375
d88d633e35afa1bfbe107c0197031bbff7a1b238
645910 F20101221_AAAJUE habib_r_Page_299.jp2
4ffde0201b249f2107cbf4dca0aac074
30d031eeef412f5a657bced8c3a7368d7266cad9
6712 F20101221_AAAIKA habib_r_Page_335thm.jpg
c3fb5a5e393c4c1d3cf22e02490b3513
274ecd2346e291b4409f4f68984c54cbafdf519b
21894 F20101221_AAALIZ habib_r_Page_197.QC.jpg
99af48c981e6ce115f78c229ed3b0c06
446bd018abf645ef2748528c172178dd8f8496cf
34708 F20101221_AAAIXM habib_r_Page_015.jpg
0e91a6cdb320b8462ef875b82d0a9420
6722a3a5c14897309107fa4fbac2831470f19b4e
509508 F20101221_AAAJUF habib_r_Page_300.jp2
7e88f0fc83818e293cc0997bfd51fc3f
08a69a839b31241114f8b70c666840feebe18de5
71166 F20101221_AAAIKB habib_r_Page_210.jpg
55828bef3055e67ecfe03407b6c1ff1e
fdf4896c864077c65204cf56c9b686ac1d66f9c9
F20101221_AAAIXN habib_r_Page_079.txt
979e83bdab1cab612ec9efbe56dc0573
bb79ff4c6df87766b4167532b2674e30e4db99ae
510789 F20101221_AAAJUG habib_r_Page_301.jp2
94c6b466f30595be60cf35f831af9a53
ce508c922f94a2810fb0ec615e10e35227fc512e
54492 F20101221_AAAIKC habib_r_Page_016.pro
b74c28ee680dd6a9e8638e719061caa8
f7261b9225aad3e4152d1b98e8376c6e585fc281
110446 F20101221_AAAIXO habib_r_Page_097.jpg
e26bc337d9e63eb3a802b57cf884df7b
cf71dda85573d1d13ddce58c05ac525be2451ec0
354659 F20101221_AAAJUH habib_r_Page_302.jp2
1ba7978e8a6d5a5eb77b7846cb7324c8
fcbabe976844157882ad30013d70343952a8b9d5
F20101221_AAAKRA habib_r_Page_108.txt
3f334fa773049d6616e4cd2077e4fd9d
5fab7d4dd399330602c62ee4127acf0734da6e23
5883 F20101221_AAAIKD habib_r_Page_245thm.jpg
deb712fe8e4afbc9fdff72282cd19da0
ea11a14abc46441998c8d94251813fce05406469
2429 F20101221_AAAIXP habib_r_Page_221.txt
99411de92904a4c90b4d45f70f99e591
085dd0843afe51baa1a7f6576104b20877372035
578713 F20101221_AAAJUI habib_r_Page_305.jp2
b3aa42861beb861ff70e202f0ac3a1ae
ac0e9787583bdc2564e31fa83bf5a54a32ccfb83
2062 F20101221_AAAKRB habib_r_Page_109.txt
2e5963f4af69ef231447931fc4d15aa0
7d5831df6a6120965bdadc6ed815e203310da920
F20101221_AAAIKE habib_r_Page_104.txt
ee5584001b884df51cc416aa4c4fa38a
5a69d8d360fb3adf2e65d11d7b64fbb925a956b5
F20101221_AAAIXQ habib_r_Page_158.tif
23d32e420f8728678d0326ce2acc945c
255dde046e40ee877a21c38d1dc0643cc9864b20
474325 F20101221_AAAJUJ habib_r_Page_307.jp2
01216f22c29c501d38d03a64037568bd
6b6a2d01de92c735a7605fea9d937bc33921099b
1873 F20101221_AAAKRC habib_r_Page_110.txt
b8045ee9d253cd36d9a804856c539893
38f449893465469b960e83197d8f2b4a525efbe3
38186 F20101221_AAAIKF habib_r_Page_023.QC.jpg
761ded76e68587e1224cd6876e502bb0
94636f38edcc95ef0e53b522e714293693a2589e
F20101221_AAAIXR habib_r_Page_212.tif
a2c4311a6f090a217ead1551d59f7e49
41c72b4a4edb0e9f987aa3e647beb72388cac241
449311 F20101221_AAAJUK habib_r_Page_308.jp2
c197e1fc75b600f356a14c47c865d468
c27d286bf5a43480316ed38d37eeef8a018949dd
2044 F20101221_AAAKRD habib_r_Page_111.txt
5d188c7ab35ebc8155abed3d357498fc
60ba67021cc0a92fb6b655f4fdefb2faac1778e4
16384 F20101221_AAAIXS habib_r_Page_252.QC.jpg
9e4bf74064c67050faf5306237f98745
176f3a6ad537a9d54b398957b26a91addd16d6c0
452974 F20101221_AAAJUL habib_r_Page_309.jp2
aea329d57df03a1e954dda9008f53673
016f06885589024985f0e19e2850923b53bdbbf0
2063 F20101221_AAAKRE habib_r_Page_112.txt
2d91473f6d30a3b55a2b0e2cf8b85b05
cad90bef434af338bf740dd60007136c7d1aaa2b
54953 F20101221_AAAIKG habib_r_Page_298.jpg
6ea03b914a706cb7661cddb6e19616a6
d3e784189cad78833325720360d5f1af5479bcc8
117793 F20101221_AAAJHA habib_r_Page_178.jpg
8f116034192c5df68aba4efb0346c95d
cd726e5f012ea1e2ab2750455cdbc0e2cc3e1a84
451581 F20101221_AAAJUM habib_r_Page_310.jp2
b8248ca383f18392df6081e82bad9876
be9f03332ebdf7ecd8c360a187d60a8f5896b8a8
2119 F20101221_AAAKRF habib_r_Page_113.txt
4f6a2b9e96a33b121f2d7eebd5859682
5a8adf963984b2c5153cb53990028746955b79b2
F20101221_AAAIKH habib_r_Page_031.tif
3190a8dbd4a03dad99c038dad7a21f8b
7f99a417c1a21bda9a5be9e5b0da6f451b2fff64
97215 F20101221_AAAJHB habib_r_Page_179.jpg
5856962fbffd261403f59c078113d537
fcdfa28ccdb3c1ddd563c9353ae343d600a248f9
589675 F20101221_AAAJUN habib_r_Page_311.jp2
0646b432ee4b7ef12985250ef6b324df
bb04c3be9c0963cab0bcccc652e707858c3df15c
2194 F20101221_AAAKRG habib_r_Page_114.txt
b3321a43369e93b0f4999e9b6c7c102c
684898f716b9dc408a7e7ae24c903d2badce8d81
90234 F20101221_AAAIKI habib_r_Page_080.jpg
30242a6f130a1ea34bc335751e730c56
14155ce3d60d536e0962a2811327369322ce407b
34485 F20101221_AAAIXT habib_r_Page_331.QC.jpg
c7265b278cee3773741b5695b71452aa
8457135f9ddbde215f51e83a448793b625351c2b
17974 F20101221_AAALOA habib_r_Page_291.QC.jpg
525f20a74caf149ef5ad6974398b0646
e91233786654cdb07fe573a5965aac360577fb20
741378 F20101221_AAAJUO habib_r_Page_312.jp2
dec00837a67215ef4b0e09122db199f2
f454167da0587de1bf09ded487c7d8ab31be3114
2071 F20101221_AAAKRH habib_r_Page_116.txt
7368ee7dfe9b0bdde55746d917517355
f9c3ef9139b3ed48ccce82877e54e01d685ebd30
63228 F20101221_AAAIKJ habib_r_Page_012.pro
fd2aeabb8144bdad2bd9aa698391f215
a921e69d1760f7797dc3309a1d6037c895c52976
98559 F20101221_AAAJHC habib_r_Page_180.jpg
e14896bc033e0b0213f58cbc933f830a
61b8d24aa218a73cf52c34d50a4a79bd5b157d29
27116 F20101221_AAAIXU habib_r_Page_298.pro
3c5f664dd9a86ca949b88b6e760f232f
08ac9e1825702d3b9e7ed32158d402cbd4f597b3
5346 F20101221_AAALOB habib_r_Page_291thm.jpg
5b934e7def8071f8be76e4d23e468c98
aaf8b08476a00dab9bcb230361856fd0b5f2a210
1959 F20101221_AAAKRI habib_r_Page_117.txt
496dee311845928fd760d9d151eb0e61
4efecd572bb49c7b0f65691511332de3f8255a97
F20101221_AAAIKK habib_r_Page_232.tif
4951ca7c35887e1472467b1894e80f85
2c484142f54846055a6020960d796bc5399d7300
100328 F20101221_AAAJHD habib_r_Page_181.jpg
2af496b2c5f824262629b53101fb21f4
b47f81b23dd8d1ed7173bfe09e725f9466ad9da0
2280 F20101221_AAAIXV habib_r_Page_105.txt
ec12c8ca40d72dfe1a4e6dcc09b14abf
aa40953ffd9490fbfecee2d747bacec36360e617
21725 F20101221_AAALOC habib_r_Page_292.QC.jpg
a82655d28b3c136f7d3731d64c99233a
3e44964057ca08b8dac74ad812715a9e064db687
714574 F20101221_AAAJUP habib_r_Page_315.jp2
936b06da79091c012d6150ab56b17321
247ed9abdbd51cf2054f86dc4604dc5991d7a9e9
1998 F20101221_AAAKRJ habib_r_Page_119.txt
cc558197066753aa2ea7c0bdcfdbcac1
ce9b5db88345d4c6d9dcb9f430a9c25a7c81d2c8
F20101221_AAAIKL habib_r_Page_190.tif
09ceb0e503a1d0c0cbfd14e4bf60b5a3
a932565c67c5cbf770b682bc1085750e8e588515
107528 F20101221_AAAJHE habib_r_Page_184.jpg
f155149db5365347934fa4fdee967fa4
4aed6498430a0690be89f28006d83d6e5d833956
40314 F20101221_AAAIXW habib_r_Page_165.QC.jpg
a4566edb6655eb58fd65826d55ccc6d6
8e9e2b7fa96cc56c2c97df5b98f5d00503205952
7109 F20101221_AAALOD habib_r_Page_292thm.jpg
8a7556abf38c13c05b3d6270c73fdaae
d305a51830b8d344d9294f089ecbf2dbd4186317
232831 F20101221_AAAJUQ habib_r_Page_316.jp2
b68afb652aeebfbb2aefe3514d280db8
ce444343e85fd3a295c3dda8461538621d94e1da
1986 F20101221_AAAKRK habib_r_Page_120.txt
c9fd8f7d5412aa43052acf207c5ece21
e13396478c5414b73d41b5320a2082803ee49e88
615 F20101221_AAAIKM habib_r_Page_255.txt
f7600109d877d150c4ff2fc429df5b68
b995f9473fdb41a227bac4b5dbd6170b560709d0
107180 F20101221_AAAJHF habib_r_Page_185.jpg
996cabb9cb2c25004a4ba76a04ab68d9
820a132c4c1ee2de3a83b94c0d04f00441e9f988
460 F20101221_AAAIXX habib_r_Page_271.txt
870e4701514d0096aeb35553f69b2ed4
a9fc71f87878a17d08eede267e51cef589217bf8
11155 F20101221_AAALOE habib_r_Page_293.QC.jpg
b4187479db6610e1807682eb3d690f91
968ba18ef6470d50d7fbdef3edf372ebe765f931
192347 F20101221_AAAJUR habib_r_Page_318.jp2
db92c1e2e12d2778c30ebd36367d46dd
952d7842e97799c0f824cd65c13bd1b3b05f5469
F20101221_AAAIKN habib_r_Page_321.jp2
106953e5158aad95acb0e664eee8c947
b9e332fa06558177e6a1d2a20020ec5e145b05fe
115700 F20101221_AAAJHG habib_r_Page_186.jpg
fa50c639b382bb2bb26c3bb8e15f7835
8a609a3ee91c262ec89e1e503828a986bbded26a
2201 F20101221_AAAIXY habib_r_Page_203.txt
5b2cb26f884fa40250008b55f9162c38
42529d6d5fbab7cd03b97bc5e940f338f12f339c
3999 F20101221_AAALOF habib_r_Page_293thm.jpg
79c30c980d5966bc496d314f04ea744e
04fbb55097d948942442121c1da94b6949fecc93
1922 F20101221_AAAKRL habib_r_Page_121.txt
66b8b76845834d91ca2ad740d9f65b0b
869d3ff892304448d53025527848be566767981e
F20101221_AAAKEA habib_r_Page_315.tif
f6e1e14e68146a35a934198b45cb060e
95dd7c7b66445a278143bcf115bb753ae25edc8b
F20101221_AAAJUS habib_r_Page_319.jp2
c76fee83a85e6a13fedab2ae82b12200
e64f94d762fcb0016f31328d8c8914265d001549
26055 F20101221_AAAIKO habib_r_Page_256.jpg
967a389f1d4990790696d13d12658d0c
30351be08113011acccab817f39024c93131f846
109869 F20101221_AAAJHH habib_r_Page_187.jpg
a4f46c01210914002a3edb7533364d1a
04094c3a6966299d57d8bbdf533306bae9837853
50979 F20101221_AAAIXZ habib_r_Page_091.pro
36baf47dc5fa71cd1151810e4c86bc6a
b8589af23d4e8afb1b15f78cd29a9273039ce8ba
17812 F20101221_AAALOG habib_r_Page_294.QC.jpg
f17ebb528db154ae59c9963c96ace224
266a98ad689c24d901991fde07faac61a482f891
F20101221_AAAKRM habib_r_Page_122.txt
70f3fc06abeb0fc1c6392b4353dd7cf2
c7b0e1dd0484cd29501aa5e6a346477954e3d8c6
F20101221_AAAKEB habib_r_Page_318.tif
e8ec51be28ff5e7df2158ed9ec29ce0a
253208c1c4fedfaf68f7a51023f123ae87211a7f
F20101221_AAAJUT habib_r_Page_320.jp2
3df07a38ffad51dfeccc092c8896ded9
b6b0d82b7317b81dc7e2391fbf54809c608844b9
9262 F20101221_AAAIKP habib_r_Page_184thm.jpg
5385e2323715b91b15a2a015121a7a34
95f14f4ede17e36041c9b96ceab557f4130a3d2e
105184 F20101221_AAAJHI habib_r_Page_188.jpg
04d28e078483f27f8a601956fa2a1485
86159da626d51b98a2923af9977d97cd6cf86273
1927 F20101221_AAAKRN habib_r_Page_123.txt
4a71c57673bca6de9a6d6a2dcbca4589
d547018cf184f594b50cdbefafd1548c8207d55d
F20101221_AAAKEC habib_r_Page_319.tif
3600baf449246296c51a2e01d8db30a4
c30488a412f24f2fce05480243c0e74bc67e4a1e
F20101221_AAAJUU habib_r_Page_322.jp2
679485f143847059ee654e41ef404201
f0ef1d09cfa370ecf392c61a849064aa3ca190b2
F20101221_AAAIKQ habib_r_Page_054.txt
d07b922c55a23362e8a3464ddf42815e
e9154a6923d8a5c47372932ffafbe561075ef31e
106426 F20101221_AAAJHJ habib_r_Page_189.jpg
b45a8eec1faef98481637842f74cad47
63b2eff45f7c06863c5bf60d1cb801ae780093be
5690 F20101221_AAALOH habib_r_Page_294thm.jpg
de5cfd2c31f1f3a5d2069cd91c9e5d6a
c2bf7e26cc2d07d8f6a2994fd6ef03b6f1d5413e
2105 F20101221_AAAKRO habib_r_Page_124.txt
c2bce3436f7a2acfc79d1b4112937f6c
060f4dcbb321090637d7fddda6431002306c0325
F20101221_AAAKED habib_r_Page_320.tif
fa32a15017b93c03cf3d10e71b01b490
32986eb97b060e8ba99cd2bfcb9357242d622ef8
F20101221_AAAJUV habib_r_Page_327.jp2
4912749527ce227a75481dd1f68b4a46
2472fe79fe164598514476b2f0b8f7cec8f8433f
F20101221_AAAIKR habib_r_Page_287.tif
3e43db8c29fbc70f0c0e4c236f3f3a8a
a928f7d5c07483d7f872e4efc89715550b166842
113208 F20101221_AAAJHK habib_r_Page_191.jpg
409077f5ef2308acb94bd323b5c0b209
e782316e71502732047011c6ddd89e303139490c
18097 F20101221_AAALOI habib_r_Page_295.QC.jpg
251e8e54ed0f1d4fe23b2c01a4b90ef0
afa31074b282980791374592ef7a3fa2b6e2b760
2030 F20101221_AAAKRP habib_r_Page_126.txt
f7e6b792f867329f3fe7549bde0ff4c2
fe4d0405e3aed098eb560da8cedc05d343daba1e
F20101221_AAAKEE habib_r_Page_321.tif
b28a7bac2736be1b4f49a156a731958c
c9bb22f7edbe1618d1667392e32330785cd5e649
F20101221_AAAJUW habib_r_Page_328.jp2
060e668de2463e483bebf101185b7b67
11a5ba75e271375b8c647d81800db2964f83d292
F20101221_AAAIKS habib_r_Page_238thm.jpg
c36e465f7d1f1c3155d04a2be5d9ff87
4eaf1ad1111991865ce193b7d467e194278d77c7
109631 F20101221_AAAJHL habib_r_Page_192.jpg
bee0336de98633e3310058c6f08185f5
5e3249b0f541184b6213b63555e50e75503a4575
5037 F20101221_AAALOJ habib_r_Page_295thm.jpg
e26b08dfe2f00aa18d6fe855230b9722
febcf7169837df1796e4996f6ef6a8f1b7d011b7
2058 F20101221_AAAKRQ habib_r_Page_127.txt
3170c13a985c82c6d246ba6e0990ec67
c487f75e592134933f7844aa410e098f19126133
F20101221_AAAKEF habib_r_Page_323.tif
de103609e72660032acd15c3dd0bc20c
85c1c609465aefe66ab96d866ed78c08ac688a22
F20101221_AAAJUX habib_r_Page_329.jp2
d7a7fb53e7413017879e2e7e62059d66
34067226df6c527a901e3f037a4acd5ca51bd0cc
557897 F20101221_AAAIKT habib_r_Page_304.jp2
6dbbe6990c7ac0ab8002b8f47036f336
32918e2901f7a8c523ded3b64166269dc1688489
64166 F20101221_AAAJHM habib_r_Page_197.jpg
78de9fabd0f3eb492904dcdf6f897c61
70536317f272882413729da8e1c3c82a61a8ecbc
17267 F20101221_AAALOK habib_r_Page_297.QC.jpg
8b19bf80acac16594e79441d27f28846
e9adabb2227bbb705b481fbabdd5a08b390b47d7
2820 F20101221_AAAKRR habib_r_Page_129.txt
7c494983c141f68d2d95d6b4ecf69702
f361a9f967fdf1549c6c03533e2f4af574e90367
F20101221_AAAKEG habib_r_Page_324.tif
d8531f367a01572e616c891d64136b4f
677265c9bfe6f55d9e269973544a3dffba7fda4e
F20101221_AAAJUY habib_r_Page_330.jp2
680bb1b04565bf8f3f5459a55325554d
586fedbbf500766077fbf85d26e921ba6bdd0e0d
113710 F20101221_AAAIKU habib_r_Page_025.jpg
069a67a8058e56fdfb4033506f564c1e
0596d640f7f0b56551c3413bd485b245b22325c4
61284 F20101221_AAAJHN habib_r_Page_198.jpg
3f7d0fe5e31c2d7a2c48f539a370bd1b
13d0119bdd5c4b35aba56bbbef74b5516e198d17
5303 F20101221_AAALOL habib_r_Page_297thm.jpg
57b6a47332fc6a67e906fc0f71b17771
9826a90feabb6a433e615e3cde11df0eb0c1c484
8963 F20101221_AAALBA habib_r_Page_055thm.jpg
bd9bebaf5a93cb8845d10decfcbcd2a9
df201e5ff101758a25b8699638659b25a6a0f349
F20101221_AAAKRS habib_r_Page_131.txt
b0e97e149b493e84c6a7fe235f7f9934
244ea621ece18bbaec3d92a4d0467492bb79f328
F20101221_AAAKEH habib_r_Page_325.tif
4673df5943f3c24bde7288824dd7d921
eb179d34c37da5110b2f22b2a505fdee0eea9d3a
F20101221_AAAJUZ habib_r_Page_331.jp2
663ef6b05aba1030263a741ee78b2010
72543b64ab938f54c2fbac6bbe51d0e2b8b3cfaf
F20101221_AAAIKV habib_r_Page_308.tif
8540052605ecc530ee7403bbcafa0122
d67a49257adea6ac96a0c3622152937b537dad2e
41569 F20101221_AAAJHO habib_r_Page_199.jpg
3f5b6dc9404b74f5d6038e75b66ad2f8
e0ae04ce7ea5054b6d7ae2bb152d8077d3ddbaa6
5557 F20101221_AAALOM habib_r_Page_298thm.jpg
a8c6e3ba21026571a0cc981bd7701b8e
94044062e4a430db6321cde711d0ee5f15d87c1f
F20101221_AAALBB habib_r_Page_056thm.jpg
0ae03fa2191eac4eb19b1e9e5870f49d
528dc405cec417c9ac5e490ff903ca43630b7ed9
F20101221_AAAKRT habib_r_Page_132.txt
c22aa35f4d53c1cb83e9fcdd1949fdc1
40dcfeca9050d064a99ee9e64bb854661ab5af2d
F20101221_AAAKEI habib_r_Page_326.tif
dcebc2cd4a60abd2a9695efdb7b09c37
02f8b3400be1c5098f61517575bf71bb56880391
52123 F20101221_AAAIKW habib_r_Page_132.pro
c7b953871be18899fbe67227f791da71
81a9ec8d8db0c1705627ceb81fd50dc200a495b4
67291 F20101221_AAAJHP habib_r_Page_202.jpg
abf97e6877770acb4fbb93530c110e47
d9fd57873f421f25dc1e666d62bbfc9d52c282ee
24670 F20101221_AAALON habib_r_Page_299.QC.jpg
e6cca87686fc861f6785e4b62278cc17
781f2fa540bb4905afc14c60889b39ce7c531c0e
37309 F20101221_AAALBC habib_r_Page_057.QC.jpg
e694488746924a990c5a6bbe7f010c71
b5d6ccbdc0141d9af858604b0192361febbfcbb7
2241 F20101221_AAAKRU habib_r_Page_134.txt
7b0610407375ce42c0988001cb4a8da9
2d53b07d087d6c46362743bb08b2088744f5b793
F20101221_AAAKEJ habib_r_Page_327.tif
b8c7ea640287d88ebf76d996767f1128
22353bac6a3692136a86f136a6e35ef6a26799ac
15357 F20101221_AAAIKX habib_r_Page_306.pro
cf5c6d4be41ebbf16bfafe07cfc8b56c
2bb1249c3cdc0c9062717b52767e8bf5a14793f1
113744 F20101221_AAAJHQ habib_r_Page_203.jpg
1108338ae15cda0891d454d42a60cca5
06e0a2f3102d418b9b2c6ee078f92c2398c3b719
19713 F20101221_AAALOO habib_r_Page_300.QC.jpg
41e30cfbb03c21ac7b4b484561edb260
00492a114a02ee26877e37cdd86b27f2210797e6
38375 F20101221_AAALBD habib_r_Page_058.QC.jpg
3327b8561ba9902e03d235636075d734
b7400cdeb2efaaaa4331641aca39a9a1b72477bf
2019 F20101221_AAAKRV habib_r_Page_135.txt
1f9e3931cce0afeb45ccd169e9994694
b67d5c618f5e8d8c7c4a53b2e76194ea3df21dd2
F20101221_AAAKEK habib_r_Page_328.tif
1427c760cd6bea9fb72dbbfd73783430
f152cd789534a31f06067b845b8db77d00220837
F20101221_AAAIKY habib_r_Page_034.txt
adbb1ac609a214188feb0b3dc6127b11
24626b5788cd3ed629ae601290a1c74898858bd7
27078 F20101221_AAAJHR habib_r_Page_205.jpg
9030b28b8b465604aa489366c14887b3
b6932211e5c7d09bc959dd05a64d9c1b651c3bf3
6365 F20101221_AAALOP habib_r_Page_301thm.jpg
0507ace72a19d57b6809e57589455481
e1f289b196435eb1fd64e81e1bc04013b11156b7
9227 F20101221_AAALBE habib_r_Page_058thm.jpg
01cc6b3ccfe9bf5105cb01a09acf3f76
184affc0a8ce10e1c9d6a7be8eb085c82cc603df
F20101221_AAAKRW habib_r_Page_137.txt
ccad0957f0e74c9b62ed4a003e06906d
8fe6b72affb14d4d12b640bd52ccc8c1c15e6374
F20101221_AAAKEL habib_r_Page_329.tif
14424b01691f04991698b0dd32ca57d6
6849164fc14e226636601ae184d6dfe7be666837
8798 F20101221_AAAIKZ habib_r_Page_026thm.jpg
1a2717c39e561b1f62ea8940fc8b2fec
660818375aebae2cfa35d380fca3edf4826f5029
102190 F20101221_AAAJHS habib_r_Page_206.jpg
948ba67dfc7c1e926adf36ae375527fd
abcd51b734fb249f745c260ef8bbfef3c6cc4299
13473 F20101221_AAALOQ habib_r_Page_302.QC.jpg
f2f1905e41b2444d38869eb061af501f
74797f29d1b16e8a084e18f15d62d9e487c8fc9d
32962 F20101221_AAALBF habib_r_Page_059.QC.jpg
a1925628094a091d04384f0226980c34
735e7679093b615848dec92de629b2ba10bdb97e
2078 F20101221_AAAKRX habib_r_Page_138.txt
22a5ec38b468f6d4de510e5994f26418
7b8cace52c22ee75160213c93b036e8b6800cd36
F20101221_AAAKEM habib_r_Page_330.tif
845b581858626202c3ad8a79dc25d84b
a9bfcc1f220574c8f4b11b1d97b7c0167fc95b98
95826 F20101221_AAAJHT habib_r_Page_208.jpg
c9b2fe49071ba353044dbd87795adc0a
04a558b1219362dbdfc419ef54e7c6fca6cea610
17766 F20101221_AAALOR habib_r_Page_303.QC.jpg
e1baf0f5130bf8bed5ce2d20ef114d58
1e1564fe47e0b9a27d4d300ea80ed267ee9144fc
8479 F20101221_AAALBG habib_r_Page_059thm.jpg
629541511656b10fc900e7690c2a48c6
0ecd0ae9bc0af20ae9b541967f0a5b75a1f7dac0
2186 F20101221_AAAKRY habib_r_Page_139.txt
5d9c0dcb920538de8575a4d2fc7972d4
fed2715fc1339f5f0c858d93414ac382d3bd5133
F20101221_AAAKEN habib_r_Page_332.tif
31f59931b146a1f857fdbfa9a403f8a9
c5b9cab1b532df71396c8e8a4d1bc92f7b51f884
115183 F20101221_AAAJHU habib_r_Page_211.jpg
8bc1d0e5cc03263cec785852489d81d5
c3e25aea71968864d757e55add0edafac9c4b8c1
18043 F20101221_AAALOS habib_r_Page_304.QC.jpg
2b758a4882c95a8df3a614bdba74feaf
ec8d23d568911df7088626a573263dd7a7254b50
36953 F20101221_AAALBH habib_r_Page_060.QC.jpg
bb83944245c21465429debcde9aeeca7
652505b9de3bd9c40d60f83444a2bec43e4ae62b
F20101221_AAAKRZ habib_r_Page_140.txt
be314af3cc7c3d6fb3e0e88a5e2b4666
162bfd2bb9a123727bdbd9f8c95fe5d1e2b24616
F20101221_AAAKEO habib_r_Page_333.tif
41a7ba586a11caa58638574bd0422bd3
17ecdf683e70064483882b41d5d4e8ab54c5962e
106674 F20101221_AAAJHV habib_r_Page_212.jpg
088b68146a288175505efef53bf19fb0
0450446ad3c3f4a9a945a368dd3834ff8b67a3db
5431 F20101221_AAALOT habib_r_Page_304thm.jpg
48d098bbd41a27bebc39c9521ef0b8c2
f7ffef13ae10d8e3d8aa21b5b1603e8f2aa6c161
9243 F20101221_AAALBI habib_r_Page_060thm.jpg
7e17512a01bbb379dca3c8181fee292d
8bb6f067410c7f288c70c8054f967d259c30e11e
F20101221_AAAKEP habib_r_Page_334.tif
90a036549f89fa1a339a133b7f058149
0d7480132bb0c38d7698ddc9f580d549d94a52d1
27827 F20101221_AAAJHW habib_r_Page_213.jpg
74f8f3998dd194e69919b9435d44233f
d0bb2cdb26835d4cb1d758a1bcfb7889eae8d2ea
21633 F20101221_AAALOU habib_r_Page_305.QC.jpg
2b1f298ce4d41e2fa5acf0a5c7874820
333034914e1f18da9a7eefd006902924a67ca1b1
28919 F20101221_AAALBJ habib_r_Page_061.QC.jpg
1f786ec11a42c212d08a2fea89e46d22
e88c5f80bd2cf7a1632d94bb024adcef5c322867
F20101221_AAAKEQ habib_r_Page_335.tif
bbb6e416a41a857fdee6a396b6201dd5
46df6cda0a3aa201a343c0bdf4082224f01c83f5
102963 F20101221_AAAJHX habib_r_Page_214.jpg
b87a8157acb1003ba03fecb9dd156093
21f2ad626095b53236d24a19bb6b9bf2c1b75938
7065 F20101221_AAALOV habib_r_Page_305thm.jpg
7f0d705019806657c05dc758cbc59fe5
e86fd267d7164c9bf8b314fb506a62b1f883e8c9
7133 F20101221_AAALBK habib_r_Page_061thm.jpg
fd0608f6b2583ab8effacb05666af327
3847aa9562a18f90cfbbd71d6edf1f9afe6c6759
F20101221_AAAKER habib_r_Page_001.pro
c78e67688ee04ab43703e8dd5ff4e91c
59ea5edc43baf2ff1ebe905e76fbcb142c863cf8
115942 F20101221_AAAJHY habib_r_Page_215.jpg
74c369544693f7c06eff8cdcc69bfbf2
4742f63ddecff91f3706b4ac3db091b24409cec2
12722 F20101221_AAALOW habib_r_Page_306.QC.jpg
f2bf822231537e01aeaea448d5542cb9
c0ee3992eae31a0d560efbe2266ca495018c6a33
8961 F20101221_AAALBL habib_r_Page_062thm.jpg
cd7d6324010f96778ac0d94f74b25743
5d021205204f344557a283ba44784f01e7e102e3
751 F20101221_AAAKES habib_r_Page_002.pro
b2a7d6647827f72ee5568fe3710d24c7
2c85080e02e7bd83239bdf62617f9d5d700f1a10
94722 F20101221_AAAJHZ habib_r_Page_216.jpg
0de121e838cf79a0c8735cb68cc4b62c
2657f8a1b905295161f1824d6a780c7401e222ba
F20101221_AAALOX habib_r_Page_306thm.jpg
da62971509d5c56b80b0620d44dbe4c7
a9f115f3e514a4a6c88c56fe01c1e97450d6f80a
9021 F20101221_AAALBM habib_r_Page_063thm.jpg
def75db76be79c651253f43030a69b62
8344ad6b9c951891559533a1d172cde37baa0a1c
2876 F20101221_AAAKET habib_r_Page_003.pro
82cf15c5e08826f7b1737af78918bedd
bab026b438b75e91e7b1d91a696f8ebff60b74e9
17904 F20101221_AAALOY habib_r_Page_307.QC.jpg
16fbf1cf4332a685bffa327b4d1ac2da
63f7d578e54feca3c05ac8cd9d12c78a1f9e4086
18092 F20101221_AAALBN habib_r_Page_064.QC.jpg
5363b0be3c7aa90552aa46ff616f7c5e
dd411efb7ed06567de4bf5ee5bbfc08e576b918a
55016 F20101221_AAAKEU habib_r_Page_004.pro
bf3e2bc2e0576917b33cf75a7253ec90
6557883d03a42ce1914b04bd8174f9a6026c3b9d
473579 F20101221_AAAIQA habib_r_Page_303.jp2
bd8a08fafefacf2c2668a74882d5bede
578d0e2bc221a10443ba8632084f028a0aa7cdb8
5686 F20101221_AAALOZ habib_r_Page_307thm.jpg
9a0917564d2403d309b1558bcd3258ee
1666d9a3e5490c744e252c470bbed0a075257956
4648 F20101221_AAALBO habib_r_Page_064thm.jpg
29db702cc0fcba92a715dd87a4a74d81
abd648953078d9341b7af9a1c728faa3ef398ae3
36610 F20101221_AAAKEV habib_r_Page_005.pro
80eddfcd9fc3d93c39644207cdd3f9dd
3a3036fe7e30203f22c054b3b3ec53745b2109fe
57563 F20101221_AAAIQB habib_r_Page_323.pro
0a464e5cbafdf5319534f6d0115175a4
d5079c80d8c2352d073831174a969fc5dfb7c39c
36076 F20101221_AAALBP habib_r_Page_065.QC.jpg
e2b6de495fa2221dfad2e1855d30ffff
bc20064e127006f9beae3e621cd57fc8841b3483
93814 F20101221_AAAKEW habib_r_Page_008.pro
d5003d71ff4fc42a87e3c5fcd4498342
9578366f385e66da52f7313cde4e6efcec4b0dac
F20101221_AAAIQC habib_r_Page_037.tif
75b165cde7010395c50de2fe9b45fced
5e5c1f8fd00e2cea61929a4179977033ba659362
F20101221_AAALBQ habib_r_Page_065thm.jpg
d4a0613ec645e9534489b5aa3ea61e78
ad5d04c6fd4844b4666ea0abd6485db33416e4cb
1263 F20101221_AAAKXA habib_r_Page_309.txt
4bda240f9ba70d2f0522ccce3e0fe63f
4ae652dfc4aa83746c417f2565d389f7833688a9
83223 F20101221_AAAKEX habib_r_Page_009.pro
6201f4a21eb0ff63f22fb6cfd83f0409
88a552f637e2bc41551d85af1045a354c7c5db05
86135 F20101221_AAAIQD habib_r_Page_121.jpg
fc6ccdb65c00ade361fe27da85c507e5
30870528296399feb636e84c1c419739b742fdf8
8903 F20101221_AAALBR habib_r_Page_066thm.jpg
abb38165033b68df71fd9666d857d415
3d434d57866605230dd2b631638042f4697f5d1b
1106 F20101221_AAAKXB habib_r_Page_310.txt
6ae1a4e6531a9e7698ab795c7629468b
b125b877a3cb79ae7a0c78819d55522bb307a5c6
80370 F20101221_AAAKEY habib_r_Page_010.pro
54fd55dd792aaa9f590471e5197ce9dd
9bc83e54856a460f6fc63dc12734e8bf8afcc887
1511 F20101221_AAAIQE habib_r_Page_148.txt
265e57c2a2164764b096e8819a572a33
5f4b0e0a2d76d8d9dce7454c88756c60df0a89be
15704 F20101221_AAALBS habib_r_Page_067.QC.jpg
9f0529ea35ba2838e899f686ddeebbcb
3e9693c90af8e5e5e5c4d06271c0524d0f110bd0
F20101221_AAAKXC habib_r_Page_311.txt
5241dbb04a7c6a352af5dc23b51d4b45
010f6664e2de7435a67dea0afaa4bdf86c76907a
72817 F20101221_AAAKEZ habib_r_Page_013.pro
713aaaa294e4cb23afb23c0dea5f7553
4bced3ddd6f6bcfdde5caf15ebc22226d0e565d7
1051920 F20101221_AAAIQF habib_r_Page_140.jp2
df096c724109d0826cdbc504a40b8d20
97bcf8b09d1765f3a9984d1adcd9cd6dcb291067
5072 F20101221_AAALBT habib_r_Page_067thm.jpg
1a0186da744eb3725ac63d9621bbbbd8
1dc24134da2e983b48e4092e87b94b8ded0b1006
2211 F20101221_AAAKXD habib_r_Page_312.txt
47b0aa2ba659f634e1dee50b337c7ac9
9a336013ea3725e6357407080ad8b37e95544411
8324 F20101221_AAAIQG habib_r_Page_152thm.jpg
c00d62c32d3230dcd20b53d867b46993
9a32ff9abc39592c5445dfbf07c96a6f5af00ff1
37391 F20101221_AAALBU habib_r_Page_068.QC.jpg
0a29eebd270355e37545bcffaaaa1b96
a45cc2a67a3ce8a4a5781a8509b2b13ce9a53827
1587 F20101221_AAAKXE habib_r_Page_313.txt
cddc9ec5d6ab8bb1ae86e28ef922ccb5
fe5d01709698f527d6bd5815b7c134f1824049ac
F20101221_AAAIQH habib_r_Page_109.jp2
552b49fcf90ed79cf3948afc2b72fd34
f12a61ba24c37b20608bd1427a48f30701dd118c
F20101221_AAAJNA habib_r_Page_052.jp2
a73a353c9c1a11cfe3476b36dc834d80
20c15137bfe21cc996197f8ef35294cbf6350c55
34841 F20101221_AAALBV habib_r_Page_069.QC.jpg
7d663a86a468a893c3181833f5731fae
a47de41821c67ff47c03bd6632fba08e993e38a2
1463 F20101221_AAAKXF habib_r_Page_314.txt
ededf94c3e208426562935b9bfbc53f3
a5bcfb0caf9efbb1def2cad9ed0c73038428920d
8276 F20101221_AAAIQI habib_r_Page_232thm.jpg
b5589c042a6dbc1f68e58e0465b5918c
9d8f7cf0c1af469fbde8c3b4299823f46bf2fd6b
F20101221_AAAJNB habib_r_Page_053.jp2
d42973e2ba4d603dea922bf91e8f886f
9caff6ea51f7a792b216d289f53406d2775be952
8853 F20101221_AAALBW habib_r_Page_069thm.jpg
7fbad1f1ae85daa299bd67b20a58dafd
c10fba90388840b7b0cce725a37eecb127c468e7
585 F20101221_AAAKXG habib_r_Page_316.txt
eec84f5c3e593d9fa52be2497e436985
1050ee5723e560da7ef5816de87f5505afd90374
100538 F20101221_AAAIQJ habib_r_Page_145.jpg
a2abfdaba661bac12d6822329e45c562
af0ed7f5764db72db2cbe1997d4a7e9fa7dda750
F20101221_AAAJNC habib_r_Page_054.jp2
2b2d9c739e985dc970743ab924cd22e5
67630672ac837e94b9c5d7af121471109b2f4e0e
34022 F20101221_AAALBX habib_r_Page_072.QC.jpg
5d3449f728112ff7217dc1eeb097ba0c
bf732ba678d9f12e5aaf1d491c0c5ec8da9a65b6
975 F20101221_AAAKXH habib_r_Page_317.txt
a6d0aa0f6d199551678781ecc934b146
4040f05f3267b2c62a0d4b8377448576461fca4b
F20101221_AAAIQK habib_r_Page_125.txt
a29ebdc893583ad99bb09344bffb8983
0546ed18dea82d7bcc4b2b0cf2ecfdc6e5c67a9c
1051915 F20101221_AAAJND habib_r_Page_055.jp2
045f54584ea336e400b1147fa4dcb3c2
3d61e5cf57069cbfb755daf8785940576d1e5819
9023 F20101221_AAALBY habib_r_Page_072thm.jpg
37e877bf165266eae9efc2d284152ba5
18049d1712bf838da89c88dcd696d36b8c13a642
392 F20101221_AAAKXI habib_r_Page_318.txt
247b921a7ffe7b70ec750e0a57af356b
20cdd1bf1286f53ff3c8e11013f925f6dd1043e0
105557 F20101221_AAAIQL habib_r_Page_140.jpg
12dcb9dcf9dca33859564b25f1a601a8
6c1f3355d47471f2ee4e3156e046a43e8c0e73bc
F20101221_AAAJNE habib_r_Page_056.jp2
f827710e09cfe8407eb69081a70c528b
c53ca81bb6c3cdb4d99fe508d478d03cbf7221d6
36148 F20101221_AAALBZ habib_r_Page_073.QC.jpg
c61d37cbc2cdd719015d37b3dfea2a19
6daa42e780ddf4a499ff06ef6da139798664ab75
2349 F20101221_AAAKXJ habib_r_Page_320.txt
edfe70155cc4ac240e89a14308934f27
0394c62417695ad2068a306aa629261b49536fb4
F20101221_AAAJNF habib_r_Page_057.jp2
550501522b7e520b9a2ad688839af534
648c413169d8daeff2ce9a2917e4cb9359fcb3da
2314 F20101221_AAAKXK habib_r_Page_321.txt
b43c0e670ae7ceab188d307e87155028
7c5a6166e92e18b2712268ff457e5697e163ab0a
494081 F20101221_AAAIQM habib_r_Page_245.jp2
b8abc4c4bb58fccf9a5d3a6f6bf179d7
cafd4db1b2b929f120f0bde7e4029b5c2f6ef7bf
F20101221_AAAJNG habib_r_Page_058.jp2
9aa651e65bf81186e1dda423f3d30493
dcc78ed879d9578d5d9892eb2192180e17e19086
2627 F20101221_AAAKXL habib_r_Page_322.txt
79d536b24e0d9b25a7b54c2109134ddf
34cd893c14b95fd0f3091ba76729bba5421815b1
103361 F20101221_AAAIQN habib_r_Page_233.jpg
6207f83e2304a29bcd9b6263ad9cbf5a
918e452504d1444f82e25b02abefffa6e10ce368
F20101221_AAAJNH habib_r_Page_059.jp2
68206436fd1b0edeca4f1614c6c088e8
76b5423d839a61e54bb45a362bb2690be511466b
54638 F20101221_AAAKKA habib_r_Page_189.pro
5c4a1685347c0c90f47e524b0fdbb992
7877e44564caf4aedfe12b9127a1f90245175fe0
2350 F20101221_AAAKXM habib_r_Page_323.txt
392d87f8b9bd281ca525f3d3f906a7d6
0419c80121dd567202b2f86bfb187632e17124d3
97399 F20101221_AAAIQO habib_r_Page_243.jpg
b5c51eb67a3fdad14c1e5ae2426fe0a6
70c0d544f51cb4b1a9a171e88eeb4f96443649c6
53515 F20101221_AAAKKB habib_r_Page_190.pro
8fe730e9bb926f2d4ad1414a20e7c90d
29c0f51aef1b6fb732d69124e48fdbf4ebce0048
2333 F20101221_AAAKXN habib_r_Page_326.txt
a1d87429c18b29b46e19e5ee45b2f7e2
cff7c4bb77e86a725277b690923287f2f3d1eef1
40408 F20101221_AAAIQP habib_r_Page_010.QC.jpg
3f9d3e5abd612eee8acba6a6536c6160
3bf15c6eee9acdf2586dbef3f58642720a13a63b
F20101221_AAAJNI habib_r_Page_060.jp2
eead0f063e8dfb4ecc4dadba5f271860
22d2766e96f680ed777d1ac1d72f015232dc0551
56411 F20101221_AAAKKC habib_r_Page_191.pro
8b42cee805b8bd130edbabd41bc71832
2ad2a1333ec5cada096889091930f06c731760a9
2225 F20101221_AAAKXO habib_r_Page_327.txt
67a1d991b9ca6da2a4afce1c5ffe25c1
498ac1256e65ad1e490032a8a13c6736f5791108
7048 F20101221_AAAIQQ habib_r_Page_098thm.jpg
64b980a238df65df9ef33758b48ab16c
0f08c3dbbd0d653e6e9d54862886d7ddc4e83a66
908444 F20101221_AAAJNJ habib_r_Page_061.jp2
5ee5b88ddfe42175afba3c2655a91e79
2f954db99b99f2b9f4b73fc47e52917ac49aa0e7
55300 F20101221_AAAKKD habib_r_Page_192.pro
ef226b8b96e1f0dcfc4d774bac5114fe
8244de7c23b48a39d8be37d8a2b6724878f155c7
2448 F20101221_AAAKXP habib_r_Page_328.txt
288a8bf5a1b21315858fd1dc31591b1e
200bc92854c78be9d210f1e619207d0de507d504
F20101221_AAAIQR habib_r_Page_235.tif
b55e941b7c774693a22205456debfa31
521fdd5752a66e63f5587e8ade159a2003b4d9b7
F20101221_AAAJNK habib_r_Page_062.jp2
7bf8df151a3fc19f92f42c5cb0bc3000
44f5922566533c727168d8edd20a757c2929d0bf
2427 F20101221_AAAKXQ habib_r_Page_329.txt
979ebfdf4bbcf8d81a0ad56e454975f1
6b104e666277b3a71bf489acffe363344547231c
74251 F20101221_AAAJAA habib_r_Page_006.pro
0c598852d97802ff1f6280030fc0e9ec
d4d4e60614ad63501aea4c459179c1ccc09f0640
F20101221_AAAIQS habib_r_Page_241.tif
e823c1470a4147ca42c805d646b7ef26
6a4715b64452b3d014487f21865dedf95366931d
F20101221_AAAJNL habib_r_Page_063.jp2
8f04467677f86e48ea3d586f23ef1e82
8ebaea411e7ad7b2f4fb9001eb0331a3653dd1f5
21541 F20101221_AAAKKE habib_r_Page_193.pro
a7c24d385d06881ddaa51c1bcf9efa78
5383437745d03ef570fafb235b2cb79283845f00
8671 F20101221_AAAJAB habib_r_Page_145thm.jpg
9be50abaf20343851144e4b8bcd3971b
3cbd714a2f121311c6324229c8601171654442d2
37762 F20101221_AAAIQT habib_r_Page_033.QC.jpg
549b93ddea38f54fcb3fba6efec6f4fc
d0f2816ce63bc85576f56a82c787410f1acd5acc
568811 F20101221_AAAJNM habib_r_Page_064.jp2
d7405b72cb0c1ba2f9fe0633c98e94c3
2a240bc83c02a7c2e5574ff66a4ef93e890b17bd
34427 F20101221_AAAKKF habib_r_Page_194.pro
f0f2d99b3341261e891928ee5db95a2c
dc361546d3f1fc0abdc79f2954449b117a2b9f43
2466 F20101221_AAAKXR habib_r_Page_330.txt
08d84ec87a4435b3011cd1f5d94df82e
09f072dc4ef5b448360a436f5a1ee0a0ed16b296
F20101221_AAAJAC habib_r_Page_009.jp2
3555b8b6e9e45d4af7f7c9cfc8e03ebe
23234bb26d670d867b08861eeb36b68fb36b71f1
35614 F20101221_AAAIQU habib_r_Page_096.QC.jpg
f1ac0693e2c174180b67b5ce6a12cbdf
90e2df8acd69bda790cc5c5d7edbefc29b7877fa
F20101221_AAAJNN habib_r_Page_065.jp2
4c7678de7f88a8b9cd6dc9af5a3a3f89
7794c1e91d719ff6f92d8a73a62ea060343fda25
30976 F20101221_AAAKKG habib_r_Page_195.pro
6c5fef84842b68cd36dafe1a7c24b38f
ebad08d0e4b2a6d1e850fb8498959d9ec070b2f6
37350 F20101221_AAALHA habib_r_Page_160.QC.jpg
bdd2af873e2ad1734cfc7dfa39c871ab
84b250ef4fe4aacbfea91290ea2a26151f75bd39
2381 F20101221_AAAKXS habib_r_Page_331.txt
d7b917eb45a1410d0b5e0c4a15ba1fd8
d67fac70e2c96b0d7830bb913aaf1abedf1ea35c
9009 F20101221_AAAJAD habib_r_Page_175thm.jpg
76c074f8502bee9d6ed68d40f8d2a16a
cf3a4447e8e51e3c2d568d224b03b948f6a95bec
5024 F20101221_AAAIQV habib_r_Page_011.pro
f719d007d6161088f887872a2f5d9cf2
10e63961e835c4f13e686e5415632188043b5287
F20101221_AAAJNO habib_r_Page_067.jp2
c939058b78f55e5840148c4a3a0bc190
9641dad16f7953013d39d40cf886ba1fb8db18c7
13937 F20101221_AAAKKH habib_r_Page_197.pro
3d41c13d50a8dfb80762b9faa97a8a5c
ee5089e60d20431454c6c3380ccf5a30ca712840
38694 F20101221_AAALHB habib_r_Page_161.QC.jpg
f44722ea2874f9b1b6623c90aa82bbb2
1d6e3bff2720657af3c98c535c3a220360642642
F20101221_AAAKXT habib_r_Page_332.txt
eacd5d5c796d62cc1401384938d05b6e
081cc70ac0a458f62f30c5ef98c6180d2642b6b4
F20101221_AAAJAE habib_r_Page_326.jp2
3ab5e550999dfeabc76cf899458c3e96
3ea4aed824f41c4fc79ce0f14a4ccec456ea7494
15093 F20101221_AAAIQW habib_r_Page_196.pro
061dc099ef0f2d2a304df713b3d2fe08
ab32c8895bdfd5691fdd06e8f39bef0df39cb24f
F20101221_AAAJNP habib_r_Page_068.jp2
8bf539c441bd2e4d3854b1a284f47256
d494bd3cf680e68b5e65f0a679ecda71365245b3
14650 F20101221_AAAKKI habib_r_Page_198.pro
e980ba26c8f0a32db105111c2fa8778c
296a9d1671594646972f3d46856d7839a51ce8fd
37274 F20101221_AAALHC habib_r_Page_162.QC.jpg
62fd2c88aedeaf60306bd34b244a8fca
ad5efb48e1b80df5ce6a4c4a7ad66d1d7fac53b4
F20101221_AAAKXU habib_r_Page_333.txt
7a55a49dd7488f71640b92b14ed03ba6
3b3958024ebbd8d9ddf08f8ca92af2b7d2744c63
F20101221_AAAJAF habib_r_Page_325.pro
4c6bbd0b956729113b6bd794b9333841
b9ef39ec864b6addfd6ddce9b33a0ddb5352360a
114776 F20101221_AAAIQX habib_r_Page_162.jpg
c83b925083d259e0d01069ebab382d34
83abdb8c7efcbc6cb534a1c0e62b10aeeda6d3a1
F20101221_AAAJNQ habib_r_Page_069.jp2
06cff000d4a2aa93bdd50cabc0e36246
58c3d785537f3992d442eeed8647af48c4f3aca7
10542 F20101221_AAAKKJ habib_r_Page_199.pro
644bf33c1848627377d33a7efad4e8e7
089b964d02698308496da9b24285f67dfa94f342
F20101221_AAALHD habib_r_Page_162thm.jpg
b27b2fbf5b0fcb772bcff8f9a5abe52a
759b9d7b5678e9f190979bb03b435017227ebeed
2193 F20101221_AAAKXV habib_r_Page_334.txt
5e5940afd94375ed90f7d689f7d8ab0e
2393b28dbe3ef2001f76079fd6bfe505d4af04b1
36161 F20101221_AAAJAG habib_r_Page_322.QC.jpg
56cef0959391d07ecc93c28085ff5e16
58fa0a24ac02f8ec71cf1cc08bc0d2145640a696
F20101221_AAAIQY habib_r_Page_206.tif
f010948fdb4e5dc3f62330ad1ee15e22
4633f794ca7730da4e1b06c85086357864ff6e7b
1051944 F20101221_AAAJNR habib_r_Page_070.jp2
0680f1e5ae952e875c4856d589a43755
4b388da5470a8a9cac74a54d52cc2d609e9fd1ce
52090 F20101221_AAAKKK habib_r_Page_200.pro
ead3c3158947e7c991a68858ab8c7838
4d79415865c589dc6ef46fa2b1a2a66d7717b201
33767 F20101221_AAALHE habib_r_Page_164.QC.jpg
52cc9abf0a6b54bc6f8ffad6a04bfcfa
063fdb32ffdadc541ccb6ae2acd0df7e4c0087fe
1464 F20101221_AAAKXW habib_r_Page_335.txt
a22420e6b6408fa99fdc664208a4030e
42923d4af084efcf09aa0bbab98a6c98e163f170
40628 F20101221_AAAJAH habib_r_Page_268.jpg
ab66ea37ab8e8dc03af61773775eb587
f426ed322d509cef73f9795152a59362f1b44a67
104988 F20101221_AAAIQZ habib_r_Page_227.jpg
10aa9717f0a74a07e50eb8ea749e997a
6defd5004718666dfbbf3796621a833cb1eaa236
F20101221_AAAJNS habib_r_Page_071.jp2
59f5275911f6857f74fa17d162920bdd
24861dada3979f8c29b335b4f747b3314fd5a1df
64814 F20101221_AAAKKL habib_r_Page_201.pro
d4d6d7f06d3e8dffc10d99213a555cdf
ba28ddb8a3911307b67e6a76a23ea9e8ef466a22
9496 F20101221_AAALHF habib_r_Page_165thm.jpg
0f5d1377dcd54c4670766a7b0d26c3e5
673e99cd61c1f7f0b3c2d59ebc8cf447832f0636
2562883 F20101221_AAAKXX habib_r.pdf
08c8a9315ea31586d9c849e5a02adbc2
bbe4110977a6416998fa526cdaaade94d3d0a4c0
34614 F20101221_AAAJAI habib_r_Page_212.QC.jpg
f9013e07a41841910db7e3e2734588ef
2365968580f5b31e48b05eeea775d37a8ee8dcc1
F20101221_AAAJNT habib_r_Page_073.jp2
4e277e6682cb0bfc56413e1bf566eba1
6556458f569516652b19b187093dfbffe6e42841
56131 F20101221_AAAKKM habib_r_Page_203.pro
42b04bad79689f0df4e35356847a7196
3e719fbb4fb4650245fee5e655653026b5dd9b04
35297 F20101221_AAALHG habib_r_Page_166.QC.jpg
16f3b6445c5d21f7dbc3f11f06d7b095
9f19574190fe04630e124aa07c3015e7cee0c7d5
1103 F20101221_AAAKXY habib_r_Page_002.QC.jpg
4ed327c5a24d2369f60e32fb11c48217
9d59576e01d5f099f25d425e46a62dce4154b25b
8806 F20101221_AAAJAJ habib_r_Page_206thm.jpg
1c2a8a23d67d02f8a4ad46c7276f44d3
13a926659da7b567f0b3b266e86e86b3a31f3c53
959866 F20101221_AAAJNU habib_r_Page_075.jp2
cbd104f807783759dadcd321f6af6ae3
25062d9fe4146ab77362acea36182aaf8524454a
47719 F20101221_AAAKKN habib_r_Page_204.pro
bf4358cc0b4bfcabf0074f80dfb9d6a3
fb56bcff182207d2b5e4cf28d9dd8887713cb9f2
34750 F20101221_AAALHH habib_r_Page_167.QC.jpg
7e30b9baac53f97de335a42f1c4e10d8
369a73bdaa699dae6d0e9a5e94b00fbdd91a9f40
495 F20101221_AAAKXZ habib_r_Page_002thm.jpg
3c27a495d16cdb890183dd65e5db789f
66a526bca865ae6622bd89b47249c54551e7e8d8
2041 F20101221_AAAJAK habib_r_Page_001thm.jpg
026e504042283ebe9a367eb3bd238f2f
11498d68c7489b9631b79e538634d6a80f0e1656
F20101221_AAAJNV habib_r_Page_077.jp2
e097b67bad177fd4ec106cdb9846eeec
702aff7242f27ec44c1cdfa7a20607625209ed97
6912 F20101221_AAAKKO habib_r_Page_205.pro
41b5f7205dd78998101bf5afe7435c96
7783bb9a217ec44de8d6b7b9a514bc2a3ef00743
37108 F20101221_AAALHI habib_r_Page_168.QC.jpg
57a8de0297320e933fd2179501a681f2
a9610e8af260bd0b76df83988d340224a87bae0f
F20101221_AAAJAL habib_r_Page_011.tif
fc64b28b8ae6538a609b3b36c3f70c7c
acdde8673c5601a58f6d2dbbf2a15e45e2f46737
F20101221_AAAJNW habib_r_Page_078.jp2
b2007c142d9b63fbcec796a5a684536f
294ddad9ca6bd3dbd42eb1345f111a7f9ed4e5e3
50097 F20101221_AAAKKP habib_r_Page_206.pro
1e7b760f97989a07aa049eaaaa5bce18
8e15ee796af668ae72fd5474a18b24ba817168fa
8835 F20101221_AAALHJ habib_r_Page_168thm.jpg
dbca16f8463341e8310818510367c8ac
821a851b4a47872dcde65bb6964fd4eb0082fee7
1549 F20101221_AAAJAM habib_r_Page_098.txt
4bceb4aee1ad1c653b7c8100b06574b0
0fcfc360241687fecffa04aea3bb785c6271024a
982409 F20101221_AAAJNX habib_r_Page_080.jp2
97fa6a8d2901660acce293aea2db8e8a
5d50e60d0a55cb02a52f48e805738f6f430299e4
57876 F20101221_AAAKKQ habib_r_Page_207.pro
e359d64de1acff32cb6a2a485db921eb
fa254c7ab8f97f8ea8dd6011580417f88e4fdc8d
35892 F20101221_AAALHK habib_r_Page_169.QC.jpg
c29d71c13cb53ea7880ba0b588f77145
42c2fede0703dda2bd87dbcab3c7dbf41d9bb347
1316 F20101221_AAAJAN habib_r_Page_267.txt
c59f9fe14a08f3ae2524f4e5b0d7b740
d245282c0196a40934afd2c4cd9a4f5d3a84fd32
F20101221_AAAJNY habib_r_Page_081.jp2
9a301d212bee172e236afb5885688042
b39f63a2113a2dbe2c360db24b5064b80b2830a3
44345 F20101221_AAAKKR habib_r_Page_208.pro
8d63aa0f482af9817b3a1fb9edde6cdb
644e55e74d9502b840d556eae9b3486be247dc8b
9142 F20101221_AAALHL habib_r_Page_169thm.jpg
5730b3df9326a7c09bc239444e8a004e
250762e2a2602fefc9daa64957805af48136cdf0
1006120 F20101221_AAAJAO habib_r_Page_120.jp2
2eab658103fd7c6e7466caa3e17d523d
463883852d0d80a1a799c962df7102380b41ef0c
F20101221_AAAJNZ habib_r_Page_083.jp2
e87ca5dea2c1873917dac683d3a5c921
e6c998dbc297dbd49b7e3644d145980b57dc9095
54374 F20101221_AAAKKS habib_r_Page_209.pro
863b425529015a0043664be769012f30
d54e02e5f61db285ab4c3d5bb148cfc20ba48df7
33330 F20101221_AAALHM habib_r_Page_170.QC.jpg
2fa5fd690a607f17c0b6b3c8bb9c0290
85b70fe6273b7047db050527e072eb14d8b0e4b4
F20101221_AAAJAP habib_r_Page_091.tif
4990488a2b3cc4af20b58202bc5e3f2c
34896328e01d1c4a190f36b87232a1a48a66a38d
34951 F20101221_AAAKKT habib_r_Page_210.pro
702e40c4b087ac9673a7b4f316b43256
f3171c59da171ddb58c1c9c0f6d686e2105750df
8582 F20101221_AAALHN habib_r_Page_170thm.jpg
1fd27558f9685a181e4c71ea8c5c0a91
3767ced8f9ac72863b8eff22b68f70d1a9ed5345
313496 F20101221_AAAIWA habib_r_Page_260.jp2
d2336b0e2370571e72bd80f6c5e83a47
e702fbfc07eb158de4d99085058db49b304ae1b6
51230 F20101221_AAAJAQ habib_r_Page_233.pro
e8684308da17b36eeb9e3dd09381b8d6
cda5649cb7ead1cad57325bc8aa4530d9d9a27ee
55801 F20101221_AAAKKU habib_r_Page_211.pro
a916ee652bd0ced17b84659d6fc9496e
30bacdf79db927c29a5ec6262de06dceb99ad39b
36602 F20101221_AAALHO habib_r_Page_171.QC.jpg
c9c64cacf76d1da2214d2db04b926b8e
4c23cc98f0f4f30ab006f43bff4d16ee91a25e15
F20101221_AAAIWB habib_r_Page_323.jp2
b7ee186d2b1b52bb6f8d1ffd51ebbf0d
88b98ce5be8e7901e1ad8c83b55a0fedbd149754
43569 F20101221_AAAJAR habib_r_Page_239.pro
13915365160b9ae90e5d4025d82cac9d
8c15bfc4318e5fcbcdeeb27d6e1bc7811c531602
6675 F20101221_AAAKKV habib_r_Page_213.pro
896ae77240701742d2d219d3f8aa1715
7e326f0ff643da858e53590d7e20ba6f4e04eaeb
9264 F20101221_AAALHP habib_r_Page_171thm.jpg
9545dbd8f8f1fa6dedf59716a0aefb4d
445ca8bd939676067b1c46ee3ba63f1f79bf0e6f
33871 F20101221_AAAIWC habib_r_Page_106.QC.jpg
39dd03e3292c10461c6fe38b14f21d97
243d95952232792ff6a4ce3d727bd1e60e196df4
F20101221_AAAJAS habib_r_Page_324.jp2
18da09a0ff3625c9c817b24c0d67326c
e39a59c0c7e701ad41615261338737453b8e4215
49943 F20101221_AAAKKW habib_r_Page_214.pro
b1d3726eafa3131451afdf4f31e3f24a
b81aa80366e64cd4fca9aaf621b45ea02900ec56
34905 F20101221_AAALHQ habib_r_Page_172.QC.jpg
6da2eafec50eba909113f60e2f784fd5
3585a4c250746b38f5b0170f1e860aaa1b6b1c40
F20101221_AAAIWD habib_r_Page_316.QC.jpg
d81c65edbeba73345bd3d2bd886dba56
13056cfea8e9281f9a93c37c145315ad06916adb
F20101221_AAAJAT habib_r_Page_236.jp2
a6c248ae05610d355d01d7edb80be065
7fa067008ba9e229e746d5c11ef26d640730340c
43512 F20101221_AAAKKX habib_r_Page_216.pro
583e98ea3597a4ea25819f35be69a98d
18286f0afe7deb46a8e5755eb899b7e4dc6361aa
9012 F20101221_AAALHR habib_r_Page_172thm.jpg
b343e43bae19e1109a3b8664aebbcd3c
0fed9667492bfe6798e7d732c35d701c2a24e62f
38213 F20101221_AAAIWE habib_r_Page_258.jpg
91c0c8b6d13e3cd041706f6e3d17a5c2
ccbcbb47275bb3c7b57096b799c347844e94c8f9
F20101221_AAAJAU habib_r_Page_189.tif
e02995b157c6f8cd2013b14814ee6c97
72845becaa507dd95a7d8fd9043bc63dc46f7690
57378 F20101221_AAAKKY habib_r_Page_217.pro
c8f8cb919cc254b158489db06aa20979
d4d0a8d76e21717107b753120b68bee4ca22a116
30686 F20101221_AAALHS habib_r_Page_173.QC.jpg
1579449e415ad78134b049aa53bfd76c
a2d187d1ff67820629ea70aac708eef97784f1bd
F20101221_AAAIWF habib_r_Page_178.tif
64f6e89fe73744a25129113f50530937
9c791874a751249286d8e9e900b8958914241a29
30211 F20101221_AAAJAV habib_r_Page_216.QC.jpg
13caf93d3772fd08f05f802009cf8489
951efdfcb00b87549009b20ef792c1d2ed6f150f
57725 F20101221_AAAKKZ habib_r_Page_218.pro
302d990eba0b09370c9f5ec79ac2dac6
4aac501a6bdbbd8e0b1e8ba74b7792281cfaa23e
8531 F20101221_AAALHT habib_r_Page_173thm.jpg
4261cf4c4dfb568836c65a476c977491
052c3fd7b05ccd721414141534027dec6a417234
8298 F20101221_AAAIWG habib_r_Page_233thm.jpg
8e3e61bdba6d750cd37201a0204c5ff4
3f8a7e04b881196039dd3e88fdb5516137a60d36
7594 F20101221_AAAJAW habib_r_Page_009thm.jpg
a32ac37fe650283edc5ad8dad4d0c695
96bd6507e0daaac3e367d8a33d4d0f1672176ffe
8439 F20101221_AAALHU habib_r_Page_174thm.jpg
0f2366d6ef75484a3f0832d9a9616bff
f4410a74eec5c0aa91541e009ad8d9f30be3983c
2252 F20101221_AAAIWH habib_r_Page_057.txt
5c4df0afaf8983c5d7d31351eaf8bd3f
be9e8e3c6d9c0645ad6d7695d50d8de965045c0b
50693 F20101221_AAAJAX habib_r_Page_253.jpg
b5c9f376caa9acd601e247be03c139e5
cf9eb436d52c0cc58c1120d6a88eb917218218dc
191225 F20101221_AAAJTA habib_r_Page_262.jp2
fe235304b7aac2e9247cd1fb8eaa4acd
3818322764e508e76317a1596b446360ba17c892
35437 F20101221_AAALHV habib_r_Page_175.QC.jpg
af6ff1a1b7e60dcfb45b1783c8fd117a
a357f0527fa661e29de1d3d8ca2e0a49e776629d
20851 F20101221_AAAIWI habib_r_Page_286.QC.jpg
84a59ccdfd5235bf0041c3b21af3568f
d2a4cf3363d2559d9ecdebfd2040511cb1318b91
108046 F20101221_AAAJAY habib_r_Page_209.jpg
6bd79297c0d8e41427843499fb8ac1ac
77a20600dd371fbde64e1ef360c49dce6f5dd471
209399 F20101221_AAAJTB habib_r_Page_263.jp2
addb327d79d5ddf46cfd1dad4e8e18f5
f69940c0f57ef84b4014f207b20512e71a450fa9
F20101221_AAALHW habib_r_Page_176thm.jpg
bed4f932fcc5f09d0d04f377cf176b44
8f268782b54db063ec240cd4df13e75baaf87f66
472411 F20101221_AAAIWJ habib_r_Page_198.jp2
ca4e4e6f0a89af56c0fd8b692c6810f7
ecd763fdbd2a95c4fe1b3f13a69efcce6686277c
2057 F20101221_AAAJAZ habib_r_Page_182.txt
cdceac6e094bd81c10fc1d60e5a73f51
7eb55bcc05b93ec92e292ffeebc19776814cec77
265029 F20101221_AAAJTC habib_r_Page_264.jp2
9fcd7831fff9c74e48f8ad0828c58eca
005ee871fa885a83ddb2146a16625d4112f78bb2
32577 F20101221_AAALHX habib_r_Page_177.QC.jpg
7ca803fbd60190c8b2a8dab7052b5e27
94093bc8e9a5fd54e0d999c9abd14be25ad1d168
49470 F20101221_AAAIWK habib_r_Page_118.pro
0ba644ece2eccb475e049824b83e6e38
257fd2e4fce945d96754502b64b4f819b766157a
457994 F20101221_AAAJTD habib_r_Page_265.jp2
a2e71ca21ab180bcf4f8f56578dd78d8
07a83dd6852a41e49a74d6d04176c8c29862fa72
8881 F20101221_AAALHY habib_r_Page_177thm.jpg
f57ad03cd76f5c815555007b0a03a75c
afc9adef9ec1391f9458345a2ca33b0edac57ac2
13191 F20101221_AAAIWL habib_r_Page_273.QC.jpg
46e354e9b7130b765517c66f8cb2111b
f4fa93acf1b4b261b8378b8cb7a08d47739a03a3
451719 F20101221_AAAJTE habib_r_Page_266.jp2
e3575331596bb7dde8cf27dba8601c3f
7a0b05eade181f6978b6c35911ea3c00efdef447
52963 F20101221_AAAIJA habib_r_Page_185.pro
990d1966edc15a4bf7096b90bdf8c911
2cc0435c515d7e08629898932c153b0882ac1266
36757 F20101221_AAALHZ habib_r_Page_178.QC.jpg
5b409aae81700db594443e3557c12026
a13e19ca755846ef1e6a224f6d778ffa7d64907c
77666 F20101221_AAAIWM habib_r_Page_312.jpg
bef7746e17acdbd7791ea56ca796531a
d296052ec43b36cf351febbe6af69b3c72ebc71c
296884 F20101221_AAAJTF habib_r_Page_267.jp2
fd074b89a19ec8f15d9e3453044f3eab
a4e4afaae28fc3125d8c47e2d7ae8b9d8196e4a4
57865 F20101221_AAAIJB habib_r_Page_025.pro
e54fde664c6534ab9f023e35ac9be4f5
3303b532642ea0496e5f6a40b17ab58d4b7a5880
2191 F20101221_AAAIWN habib_r_Page_133.txt
aa2856362bfd519ffd68fe03c404fb36
ef83a7606fa82f2082f36dac6728987a3330c83b
388936 F20101221_AAAJTG habib_r_Page_268.jp2
b42fcab8f742cbca262ec59a83933814
396dfab00ab7f53405b7f06bf6a7d2d4617111b4
33157 F20101221_AAAIJC habib_r_Page_138.QC.jpg
7eb145e9b59dc9a9036d093bab74e251
47ef5ee004b9b138ff47e7bed4f5e15e03fd99f9
8753 F20101221_AAAIWO habib_r_Page_001.QC.jpg
824b7a927f9e1d5f544ce220537c9ebc
36bc74f94eb97523d37eb5ec979bfc5dd3eb6689
217498 F20101221_AAAJTH habib_r_Page_270.jp2
5b9631ff10f817e36815dfdde9197a62
3e48b4b232b8a6d4666ee102f95f5f49fc8e246d
2055 F20101221_AAAKQA habib_r_Page_071.txt
77d1f3a8521f8af688e146d18ad3e6f2
144f9bcafcf7e1a8af885bcb8c8af601b51fec71
F20101221_AAAIJD habib_r_Page_114thm.jpg
55e2a7a95b88166e51849c020f1284e7
c7c76d5653f0ad667a8f05fbb528a375d7c168a1
34201 F20101221_AAAIWP habib_r_Page_174.QC.jpg
9f309c92e4a8b133bc2671748a935f79
d27a810134b9192789001ffbb5e5b31a4d0864e7
193863 F20101221_AAAJTI habib_r_Page_271.jp2
2635a6b0312a5447a10b0d474cfcfc30
6c17f48d2d79cb3c314356948b409bae0ff56951
F20101221_AAAKQB habib_r_Page_074.txt
673f9eead2758a47bfbfd03b9092a753
d934f12290d1cdc727db93a1a6bd9521a561f201
34643 F20101221_AAAIJE habib_r_Page_251.pro
2b075579db307087ad82334923b2f1b0
31d1fbbfa16d86619ee2f2fe9401e34f5ae976b0
9552 F20101221_AAAIWQ habib_r_Page_262.QC.jpg
f80ed37ab4805f74dafc0985d0058c22
938edab65534e2e50f968c0e0fc836f4b361aa70
497401 F20101221_AAAJTJ habib_r_Page_272.jp2
ecbf90ea75c5e5889363b49f134edb24
1c91d47d72e96a2f65740b483434b5d713d58c7a
1808 F20101221_AAAKQC habib_r_Page_075.txt
c2bc8848a26ad00c2c9957456ee33821
001d7675c42bcaa95d0e5b9bedb6c05538f3fd26
107335 F20101221_AAAIWR habib_r_Page_167.jpg
75cfac26eb1d5258e1fddfc5289cc190
c3f168683d3b39af778917d21d2bc7fba48759ae
375652 F20101221_AAAJTK habib_r_Page_273.jp2
df1f13108580f3b110a05e7b3e22eea5
c92d8c3b94a0d84e3617c7c120dbb3ef1e429f2f
2505 F20101221_AAAKQD habib_r_Page_076.txt
3bd4a36dca559aa52605e8860a1ec2a3
a2de7f64d1fc528d0845b01a0fa64512195216b0
4942 F20101221_AAAIJF habib_r_Page_302thm.jpg
b05827d84400a39075e0f66b21e2e677
2091823bb85401e37188ad325d7c8f08a0e90e76
330144 F20101221_AAAJTL habib_r_Page_276.jp2
05b5dade1a03895ed2aa4b880545cac4
10aeb355035ea6b9404719706ecce1e2a8be28c0
1814 F20101221_AAAKQE habib_r_Page_080.txt
03de66ea0a6b29cc71e394eada1370fc
0f50173b36656dd4b30d56c9f569cbcfd3bc3481
32475 F20101221_AAAIJG habib_r_Page_202.pro
2087e73e1c262fa8b4501dd56dc6e9cf
8e7b5b1f45e8d7a4ae7d84f6ad880e4621b96da3
89178 F20101221_AAAJGA habib_r_Page_144.jpg
cfd629c322709db1b853e24a790be531
e385077e3f47282e539d15ce2b6454c32c1e830d
7481 F20101221_AAAIWS habib_r_Page_299thm.jpg
a863c03528a89af955a371c0523ba1f0
c38e6a19ca9e888aefcedcd46f37f17be4b90e2c
216461 F20101221_AAAJTM habib_r_Page_277.jp2
5c7c522125405abf99cfe6abfe3256da
2aca9fe5be3d520d6cbf0e049dc45f4ff25efa3c
2957 F20101221_AAAKQF habib_r_Page_082.txt
32349fb73bf50ffb3ed2d6adaedb2f9f
125c445e22c605cd442d7afb5ca9a2d94f1341a3
100095 F20101221_AAAIJH habib_r_Page_204.jpg
787b86ae73233f46a2a51fbd00ca6980
20613cdee79a8747bbf0ff7d7337b5683e4f3100
52950 F20101221_AAAIWT habib_r_Page_227.pro
2219b3c5eb7cd3422b91c8d8d61118b9
cb32c6d130c14c20e2738baad2c8b8c053bd118a
236513 F20101221_AAAJTN habib_r_Page_278.jp2
a4a8cb5462b3b08c83dc8e300a390b52
738704648c554ae69217b039dbe41ec0d63af68b
F20101221_AAAKQG habib_r_Page_083.txt
7e3bd493f1eaec8f7e9fb0b5bdbc5891
af3af83af095ba800599f7d44d0f5a41da93ea59
1383 F20101221_AAAIJI habib_r_Page_280.txt
edae8796612072a1633243621c7d5dd9
574921ea6f267d7ba9b4e6eb6abbab891e993c8e
109848 F20101221_AAAJGB habib_r_Page_146.jpg
197c58f3c02bc397aabee4cde0f72243
5f4bf6fee592dbf8985d245f16173a95ef168e72
13888 F20101221_AAALNA habib_r_Page_275.QC.jpg
fa0e89ec69289c8eb0d92ab4e91f27ae
329261507f68a62efb2a2f45693acfe495a1dab6
F20101221_AAAIWU habib_r_Page_200.jp2
8adf9a7f04d439a23d56f186dcae90aa
d33bc8e6922f4ce3a11aa6ae36c9be10598fada2
2273 F20101221_AAAKQH habib_r_Page_084.txt
576f3536cd492cd2995a969e04829e44
c362d287fa5454d6650ce2bedbb205e171332bce
8873 F20101221_AAAIJJ habib_r_Page_108thm.jpg
232031afbfcd42081bd6c1ed248c12a5
b87b53f03022f2ff4e5a6857660ef86be74f9dde
113343 F20101221_AAAJGC habib_r_Page_147.jpg
7cacdc0eaf5f78458c8ccb8301c17eda
ef586ec52947a8575cdc56adbb74c4d6b4ac997a
4063 F20101221_AAALNB habib_r_Page_275thm.jpg
98262be36d6080ebe513a30598696c69
2db24e57529a34867a385f921a4f9d9f7ab688a6
F20101221_AAAIWV habib_r_Page_236.tif
154faf53f88d2b00dc8ea47bd82239a8
e9e1fad7d41d20e35295979c85e338dbc142a995
110976 F20101221_AAAJTO habib_r_Page_279.jp2
dd0071f336a45dbb2a9be5f7fb3dbff5
4b925e7dc3e178897abe88201d53acfdb9b0ea88
1283 F20101221_AAAKQI habib_r_Page_085.txt
83204bca93114be4bf8295d642ebd1e8
2fed43dacf88acfb3aae2fc28e8cb4659175736d
107863 F20101221_AAAIJK habib_r_Page_172.jpg
46a15d28e589b2b1e7563205c51e0f8a
cbc54506f5c6070849981cfe26e3f010fd609b73
66273 F20101221_AAAJGD habib_r_Page_148.jpg
ce5ec8b0c995f1c542e214f47b23b673
345793808bcbf0eb94959d0eb7b22645ceada56c
4405 F20101221_AAALNC habib_r_Page_276thm.jpg
f2152c483f086c8050a8e55e25e3dcc4
705e2e9155b7ac9c06084c8da8204673cf24c452
59666 F20101221_AAAIWW habib_r_Page_328.pro
a75e0508c8d178ddc2126c5d3cfbdc82
6e11390d4d98d8ec55db5b1af436a46e204ba9be
556545 F20101221_AAAJTP habib_r_Page_281.jp2
a6a3da411da9d33d3d70380d7f8ad68a
747430ce3d214a849fbda7632f7da113fd296e71
435 F20101221_AAAKQJ habib_r_Page_086.txt
71f1d86d0d9e6bbb577153977b5b7a67
cf8ddc628224489b5692fdf6c3889174b032e442
F20101221_AAAIJL habib_r_Page_077.txt
1a7d7e8a5a3c01c95471b9d56ef7d085
49bfd8d158524a973220e4f2e18199317700aa15
56953 F20101221_AAAJGE habib_r_Page_150.jpg
ed8bf4f5b935596e981bd5d8e4111850
d75d2db9294148bf6f3ad4b8376d22b4ee6a5fc6
10153 F20101221_AAALND habib_r_Page_277.QC.jpg
12af2ce305e4539c76cad2a74ee57414
b945b1b8d2a5ea13e7804fa41b02ae7490555a74
F20101221_AAAIWX habib_r_Page_031thm.jpg
b6132054209c5afce22a29c0cadd38df
d148874080bc85e4c9fa535425f53a3c24438179
630114 F20101221_AAAJTQ habib_r_Page_282.jp2
ef5973f4cf481f38688c6e10aa4feb53
7786736d0197068f16c6ec0628cff69227db76d8
31205 F20101221_AAAIJM habib_r_Page_183.QC.jpg
0a99c6e828c719c6e5882d99c2bee805
5d93649308d7596198e7a9a78c4485bfc5fef9b8
75059 F20101221_AAAJGF habib_r_Page_151.jpg
0f9818e3d55956e91144388c86c201a6
5b7a7464408ceee2988ea48b7e7587f4bdbf4b4a
3821 F20101221_AAALNE habib_r_Page_277thm.jpg
2c99f606b5a1d59cd4f1e67c09ba01f8
04ca0b59e342ac50120b3052c22cfd8f34baa97d
F20101221_AAAIWY habib_r_Page_227.tif
77e4ee2fb8b787045d7352f1f33a7172
2869a0e5fe150715efb1a046f5b7952bccf2b833
283626 F20101221_AAAJTR habib_r_Page_283.jp2
1e38307db0da1c17341e2326ec1a1197
6dd2059a0b253f4bc86a63b570da43a4a28489b9
2125 F20101221_AAAKQK habib_r_Page_087.txt
a0ebf7cd6931af046b2fbfac1693bdaa
75fd1e0f2d665545f9978137cd21d130ba0634ea
F20101221_AAAIJN habib_r_Page_283.tif
d05915e9fe4627f4d4acd69c09113475
7c1e00b4e06f09d23b03337a575f55dfe534dd70
104856 F20101221_AAAJGG habib_r_Page_152.jpg
90c87f97950a6cf7e9ffd5db8e68d26f
793b704a2080224e870f111c938de1929cd5a36a
F20101221_AAALNF habib_r_Page_278.QC.jpg
6c2bd70ca593ef98b927354e3d698850
ec22c047d333d039996374e864afcc02c9f5af12
F20101221_AAAIWZ habib_r_Page_155.jp2
062ff4963e046b79f2262ab485adfa56
fe8d717d94100a802ab5ce7febd050d2441f6a78
F20101221_AAAKDA habib_r_Page_280.tif
0cb2563ed804c97c2999717f6d5bae3f
5cf745ec4b4dd68fcb8c0a3d7eef769d2cbb5595
473018 F20101221_AAAJTS habib_r_Page_284.jp2
adb14d8c869a127da4ccb9d7932fa627
c7a49a61de6a369e33cefadcee14063ab095848d
1989 F20101221_AAAKQL habib_r_Page_088.txt
67cd05b482715f8a56c39324f38f39d3
6c33ce377a60311f1c00cd529e0ec9f026d5d027
F20101221_AAAIJO habib_r_Page_029.tif
ad6f2d6d8a6d8e67f847aa1e99f6840f
1406a77e3d991af046002f8a2eb413f7337c8957
105755 F20101221_AAAJGH habib_r_Page_153.jpg
c0a6ae64c146bcbd422cc6ad1e8fe4d8
09fa2f950322ca5855d58d8521690cfdb3ce4d68
F20101221_AAAKDB habib_r_Page_281.tif
318792696bf114f6e134645872aad16f
3c88ad116d2dbe039b6c39cc6f3a669a0d94b68d
466427 F20101221_AAAJTT habib_r_Page_285.jp2
66f13059c12d851bdb4f1d9fe9f37c9e
b652acc82aa135f3281be8348a2913536d32f7f4
F20101221_AAAKQM habib_r_Page_089.txt
fcedef5f4efc2b0e89a467cd00da21b9
86027054724837f654761dfc4ac6b6f1c1e8fb60
F20101221_AAAIJP habib_r_Page_296thm.jpg
5f83035e6b6f4bfa162552eabd26159b
1dcd3ab942affb49481809c1798be4b8b29ba137
120499 F20101221_AAAJGI habib_r_Page_154.jpg
a9f3072ced583b7f315e56444f277965
6cf574b93edce754a05829f8c9d787353a8c07ec
6086 F20101221_AAALNG habib_r_Page_279.QC.jpg
b68f85eab1f18ccb497a4cc3f5e20a98
af5a2cd80501c5202ea5c50c860451677dc65271
F20101221_AAAKDC habib_r_Page_284.tif
d1ffd5d2d54f5cc2de3bdc26da2198f7
c493f3409b3b81ed723e20c565e6a53badd149d9
555062 F20101221_AAAJTU habib_r_Page_286.jp2
2353dc5cc2357dee77197f7b2278a4cc
5098ba192b91fe679d28942abe92440e73bac7a0
1769 F20101221_AAAKQN habib_r_Page_090.txt
ad4b8b789acc46f45945c2c212f5f38e
afc97d49b1aabefc73fd9f1b81d9362502b7d3cc
F20101221_AAAIJQ habib_r_Page_152.jp2
6885ac80cf7d284c4be1a1c9576f5344
7f43a2b49e84d77166bf53a73ce0e24c68aa114c
111304 F20101221_AAAJGJ habib_r_Page_155.jpg
9f56b9ee1f0693058f9dc683b3fe53b5
99a3774eca46b6bb5ebcce6c94f1070950a3d28f
2387 F20101221_AAALNH habib_r_Page_279thm.jpg
22978d031385afc00533abadde07aa2e
046b2c959aebb50dec58ad949af6ec897742bd5d
F20101221_AAAKDD habib_r_Page_285.tif
e30911d3e02b3130b570b5df6b592fb1
49fe560b6c4a51e5c6b25371ca2400946f152aee
415335 F20101221_AAAJTV habib_r_Page_288.jp2
4be6b8d9b8b70e8191e042e1aa77c04c
30d83d69cadf43bffb4f60e89835caf9d6324ea9
2000 F20101221_AAAKQO habib_r_Page_091.txt
854bce72b0474eee9339a9bf9bcaf1b5
320185b4aa316bb6c40e6813f076c706f6a320c0
9249 F20101221_AAAIJR habib_r_Page_331thm.jpg
49671175820308d0a56909082c539e64
0010deeb36b0f1c8e07e3554fa92f4205036d704
113897 F20101221_AAAJGK habib_r_Page_156.jpg
ac93f20633279cb327c1c8465fe25488
b87c04046bc55c6c2fc3b2532703db0173d97314
16809 F20101221_AAALNI habib_r_Page_280.QC.jpg
167897544365824e13245a24eb335292
ea28639cdd330e757e628ac963ab1cfdabe4e87b
F20101221_AAAKDE habib_r_Page_286.tif
bda7c1b888aaf1ff13dd16c042dc6d04
1e21b729a9298f1a3e482b30e478e7707e5217c9
314527 F20101221_AAAJTW habib_r_Page_289.jp2
89d5e1cd296e31afaad3df4b4375fb2b
3c89b603a42be98b28cc1caf638d96b83d3a9eb8
1954 F20101221_AAAKQP habib_r_Page_092.txt
d4505521c080900e31b93311acaab2eb
de388a753579d7f34c997c58eaf066da01ae744f
25299 F20101221_AAAIJS habib_r_Page_151.QC.jpg
76f9b8d8a317a26ffd83fc6c3a9c1aa9
804c7e9083f3005d5f2af5f13d67562a992d7c36
73023 F20101221_AAAJGL habib_r_Page_157.jpg
27c5c949dc27092aae88a5032686d9d9
60276239c887ef6be88aa43a2239276e57dc454a
5462 F20101221_AAALNJ habib_r_Page_280thm.jpg
4263176371a4accfc3aaa15d3ec7934a
2768f2c296ce009e52f4aa54d6fed85ae99361b3
F20101221_AAAKDF habib_r_Page_289.tif
ecb5a25cf2fa41ba3bafaf5e90305732
60d6717fb740751b1aea147711120d68edd89620
474233 F20101221_AAAJTX habib_r_Page_290.jp2
70d320236f142ad80a0d8bae25fceab1
99604b7ce382dade0e840c84ef6163a496412d0e
1970 F20101221_AAAKQQ habib_r_Page_093.txt
30fc0587bcb6315220302f79179a151a
73fd5d89ab4095939a780af9715c2ba6689526d1
36267 F20101221_AAAIJT habib_r_Page_047.QC.jpg
e20f56312bbf7bf1dbc229328b553626
f25d51f46f0952f4b62eb0c028a32f3b6671239e
87954 F20101221_AAAJGM habib_r_Page_158.jpg
96d372105306cfd35adfd765f9c80865
4fcac357a2dafd348546d9d91562b16e56779daa
5465 F20101221_AAALNK habib_r_Page_281thm.jpg
d57e39727d667b7232f4e2a16fc33ca7
56e2943fcdf6b2c4d771c8f950a7cc8cd657c5f8
F20101221_AAAKDG habib_r_Page_290.tif
a9f3f341269e33ea09604bab4f2f041c
e025ddda1ea012c0a2f2a0419f418e0d2ebfdbe7
556004 F20101221_AAAJTY habib_r_Page_291.jp2
5762ccb2360387cdd46fd7b515860133
c622bba484400e0f435b60e4bd7a127dde7b9f1f
1743 F20101221_AAAKQR habib_r_Page_094.txt
b5fb9695ac358b9c62172c0e9324606c
77d58bb99ff76a16e9c80cbe29c51e38ef627eab
F20101221_AAAIJU habib_r_Page_316.tif
33347d5756eb0d114cd89336ef4cf038
ad32edc2057d0d43912193dcdb44c2c38578d3c2
115569 F20101221_AAAJGN habib_r_Page_159.jpg
415aed4da2453c9e195c6781fefeda8f
1c854b461d4674f9685bc88cf3a9485c08cbd0c4
10907 F20101221_AAALNL habib_r_Page_283.QC.jpg
b00eea852a4f0c0d411c9f4afbdeb87a
d86a71a2c975dfce0149380a68fa0725ac54f291
8967 F20101221_AAALAA habib_r_Page_038thm.jpg
06249465fee19192cbe3dd93b9cbd366
2a0c58edcada01df24d48fc49bc09abbaa34a84b
F20101221_AAAKDH habib_r_Page_292.tif
57d181c1d4966e806b586c0063eab2c1
7c4f02bf34d880d1481d8d8b5ab6c5cc4d4e8024
583435 F20101221_AAAJTZ habib_r_Page_292.jp2
8681a2a72fbb6b78758e771294640085
dd8a2548484a48d1a724bcf4176023188f99c4e6
2050 F20101221_AAAKQS habib_r_Page_096.txt
eaeada75b6f3f2b01ceade11b51beb45
6e977239babc87c95ebaa5b6ce8e21a16774877b
24567 F20101221_AAAIJV habib_r_Page_292.pro
9785ff329a78c7ab2e8c24ffcf84c190
47ad00100093c9bfa0f17a1f09117484e3af28d3
113246 F20101221_AAAJGO habib_r_Page_160.jpg
3bcd620bfe5642cfc890546215fe7048
05d9df20f20708d34ae00ce5b051813a9c675a11
4146 F20101221_AAALNM habib_r_Page_283thm.jpg
f923dc4e7fba4da87ac02243affaac6c
cd7178cd0919498092ad27025fec81f52f002d29
36128 F20101221_AAALAB habib_r_Page_039.QC.jpg
65612a571e46d1466e842d256af4ddce
c6accb5c09652ca3baf47eef12cc711014d89ef9
F20101221_AAAKDI habib_r_Page_293.tif
ab7be2e3c8c942e2de4d010f919a490d
ac37ae760b0f1108dde8ba2ec019e66c71dfb6e5
1981 F20101221_AAAKQT habib_r_Page_097.txt
e25f04e00d8734c021efdeeea90c0c9e
28e831ab20f8d6d5355b35686945a7d860b8fa25
2038 F20101221_AAAIJW habib_r_Page_232.txt
4a8e3215af5833646e68fd05ac3e1b99
ebad8464b3636204123b1d46ff0b7b0b55d08f35
117300 F20101221_AAAJGP habib_r_Page_161.jpg
60cc14463468668c2a5cbd88ac335987
a331772e9962e0d55d2c52db2c89b7c86ac2596d
17479 F20101221_AAALNN habib_r_Page_284.QC.jpg
433791b9091ff04871d542bfd6dfd0e8
ec29d7b36e22040c7202f784c7f4ac013741d739
8966 F20101221_AAALAC habib_r_Page_039thm.jpg
a617ae20ca04fc719f86b7a33467fb68
63c99e810d52a087a9c0ab1174bbdd859122a823
F20101221_AAAKDJ habib_r_Page_294.tif
be0688fcb65dff3be8f3a41431c49cee
5c3a564310d1a16b3b71a47313c756910d9dc745
1928 F20101221_AAAKQU habib_r_Page_100.txt
fc5581facf62cae67b694c517e8dd076
4ebfb38299fe565500e3ef843a697d57bb49eb52
57877 F20101221_AAAIJX habib_r_Page_156.pro
d4d7461acf004ce1beb7752633c3e77f
a41b13304cdf4fc84a18025f9d0520062f0c2223
113146 F20101221_AAAJGQ habib_r_Page_163.jpg
5eb5240a48e664d38ecf87cb9f9fd83e
3cdc8c0becc889d5918a1f7487be37d9c70517f8
5647 F20101221_AAALNO habib_r_Page_284thm.jpg
2f3040c321f9a72ce6beb105eb40769a
12566f268ee3a8cb9288d3a2b72c4a623622b456
8819 F20101221_AAALAD habib_r_Page_040thm.jpg
ac0f25f760090850caab556fde87a101
c74bf5c09fb76cf797e853b1004dfb9286bdd49e
F20101221_AAAKDK habib_r_Page_297.tif
a0a988b5943111e85736eb94f0a22183
c13e5f0d0ac8059209d0447ddf3cc715725c2b1b
2192 F20101221_AAAKQV habib_r_Page_101.txt
e80ce6295b0c9cbe128b84405944b3e5
a2139febbf4bb6c7b1a21977775514759da72cbb
132661 F20101221_AAAIJY habib_r_Page_012.jpg
e7c5b1478e0ab507a10eed6477a7356b
65d4160d299840b72c27e27c41c96158008bc4bc
104630 F20101221_AAAJGR habib_r_Page_164.jpg
bc6c37fa029aae1ae5fd450e02053705
bff8040558912bf2f1ab89ab164041c48c63b544
16084 F20101221_AAALNP habib_r_Page_285.QC.jpg
c4c3f9f843de4fffb364602080dcfd24
d42b547bcebb6b4bb1ae5b0536fbf505b16109c7
35016 F20101221_AAALAE habib_r_Page_041.QC.jpg
930bbcaf83e63131a54aae886c0643ea
6065de89326c835f42bc6d6c396a4524c860153a
F20101221_AAAKDL habib_r_Page_298.tif
f6acb7e537c89818546bb7a96496d6bc
c695ea59e5bd38607b3f9a20d06c3b52fc14f7ce
F20101221_AAAKQW habib_r_Page_102.txt
4ee803fd6352d060d255c3f4043901b4
0435cc633c751bd204c6e347d784558c1bb780b3
8077 F20101221_AAAIJZ habib_r_Page_099thm.jpg
067aa045ea48bcaa66dc3a621906c06a
c29c87cba8808f49cc9b54396bfec74c1ac0ffda
105290 F20101221_AAAJGS habib_r_Page_169.jpg
2f92bcf5062f49d43bc267368e6e47c1
af50958f3cf78d27c4de231bc104575a2d403767
4860 F20101221_AAALNQ habib_r_Page_285thm.jpg
6ece7dc10ec39f45e42b8ffda677da94
8f6506521c29a34b05b0c7dcd9b591a0e65957ae
8474 F20101221_AAALAF habib_r_Page_041thm.jpg
8c735eaba99543b7f10080a49ae0c1e3
efa4fee05a575d16cc0dc1a45ff88f030041af32
F20101221_AAAKDM habib_r_Page_299.tif
5dc8bb1282ce19184fd2d6611cf62d93
4f212f9950ef164575b69a878992e470fb20c988
1324 F20101221_AAAKQX habib_r_Page_103.txt
80435bdb5526c70b1c9d886ed8de7a8d
41dfd1aef9028e69c232002dc2fee218a802d670
104034 F20101221_AAAJGT habib_r_Page_170.jpg
97e8932e1a843000ce5a573294a1ff05
c2456d1637c150e3493fb24a22cae7dd98d0ff76
6326 F20101221_AAALNR habib_r_Page_286thm.jpg
74e1f2c650572b36d2ae28208ffd64d6
c9abf0df1f53a59bfe49d4f5f967823d86e4edb3
37302 F20101221_AAALAG habib_r_Page_042.QC.jpg
edca1c62603623c89c50d261c7b0476f
fabd68751115241544b91528e72fac554c5fe565
F20101221_AAAKDN habib_r_Page_300.tif
04dfaca8ce3df9e1ec982b9a046f34e6
dc9e826157c837fc94656e1f30fedab5681c61a4
F20101221_AAAKQY habib_r_Page_106.txt
3b9f0bb02a63de8c167d9366b542deea
14e1a798093d74a5bf8d88f7760feaf757cf5c36
115921 F20101221_AAAJGU habib_r_Page_171.jpg
e3661915abdd39f99fcd8c03ba9ae0a5
455002992a9261c6dcf3a410f742bd30d74df2d6
19200 F20101221_AAALNS habib_r_Page_287.QC.jpg
4c4904a6870331da6ae8307813e7cf67
6685133bf8997ade38ca77b22e6031de52cc4ead
F20101221_AAALAH habib_r_Page_042thm.jpg
6988ca458a0d0bd48b4d83190a2237c7
c034a28782d115633a0d85b2405bf5c282475b33
F20101221_AAAKDO habib_r_Page_301.tif
dc314bfabcad7f401f15962187ffbf99
315b2eaebb442f76d441b07ef28ea63934f82744
2290 F20101221_AAAKQZ habib_r_Page_107.txt
4b08de2222a8b4c3f7702dafd34679ec
3ae39ed8e5aec49d140f3648010141efc1ff290c
95473 F20101221_AAAJGV habib_r_Page_173.jpg
5d548ea662c984fe0a1ca4024bb60dd4
d8067a90a6174f9bca72aa14c576f89638d50b46
6261 F20101221_AAALNT habib_r_Page_287thm.jpg
c90e881363bd49d6af2e5c6ba2cdd431
00e253a47ce233fa6b4a78c0f5ba5f40f0fb4cd8
38260 F20101221_AAALAI habib_r_Page_043.QC.jpg
305bb096712460516cecd63000a79eca
cb650e3dfe536fd2983ec0bc5af46b7aab98c676
F20101221_AAAKDP habib_r_Page_302.tif
fceba6c4365b2880464a57b3f78f2e4b
07d2708c25e0021081762b0b2651c2ba70023341
98685 F20101221_AAAJGW habib_r_Page_174.jpg
f29f3bd664accee2a46b2b8af496dfce
ee507421cca51c06057f0cd9ec69598c3dd09c22
16361 F20101221_AAALNU habib_r_Page_288.QC.jpg
b0ba5c78dae37c140b84353942411201
34b58a74c5240e8458d5f25a772e97062f8620bb
37390 F20101221_AAALAJ habib_r_Page_044.QC.jpg
f8336798915e7d69c6852f8993e16859
13ba7288151a75528c9f6de5348cd43d1d950d9b
F20101221_AAAJZA habib_r_Page_134.tif
fa43d850dd784fca0b68da32996ab418
231d01871942984e27a576e294a3fcd5e17be45c
F20101221_AAAKDQ habib_r_Page_303.tif
82b321ece3d634539c165ee1d588046e
a910a04321027eccf4529110b7899663923c3f1e
112945 F20101221_AAAJGX habib_r_Page_175.jpg
8c987b7931fe7e4d1f5de4ce1ac23697
44a3944a6ba68af16060beaf2bc67584d1e95b37
5515 F20101221_AAALNV habib_r_Page_288thm.jpg
04acae55f493d0268da04c7a87d42823
efb9d0e73b43e8407a7187601c7cc37b2c05d75e
8918 F20101221_AAALAK habib_r_Page_044thm.jpg
36945c90adff3069778ef45fd987822d
00f0c0f6eb5a6d19a7c10655ce982feaa0beaffc
F20101221_AAAJZB habib_r_Page_135.tif
edfa67136d9f1657b8a76bfec68175e2
5cd8290203567cd4fb04364faa0ee217f5b66264
F20101221_AAAKDR habib_r_Page_304.tif
05c75385aa5383cb52df53114c8813db
9d0f65c1801b79b5512325726ccc8e55b9578c70
111832 F20101221_AAAJGY habib_r_Page_176.jpg
61e694baec3b75c9ce1671c020d57eb3
eaaf39d6f39bcc302712c2405d06b0269553fdfd
13108 F20101221_AAALNW habib_r_Page_289.QC.jpg
e76afbbc3e99a8bf937828f59f09e394
b203897f3bfb5447fba8a26e96a4279b9e3856ca
34861 F20101221_AAALAL habib_r_Page_045.QC.jpg
396344c27d5f9915320b8431f2a05261
2498b9026bfa0f8042639882238205a373cfb080
F20101221_AAAJZC habib_r_Page_137.tif
a494d0ac7e27e8a6ceebc28e70317084
4a2294bdbd4f7b50d40f5f3c5803072b82255c64
F20101221_AAAKDS habib_r_Page_305.tif
70dbcd490a26a05f1c8510d597508a0f
76dfc4bbb92b0cfd8be3fd5e727f2c9fe25f3d95
101307 F20101221_AAAJGZ habib_r_Page_177.jpg
bd6e1194c89249726e1bf3e7e26f5571
86106ffa4ac813db3a6a9025ea0d93980a6a6331
4451 F20101221_AAALNX habib_r_Page_289thm.jpg
d51f48efeca1446b714d84b3dae4cd9d
dfca0dfe9ad08f597fff2041725217fa56821b7d
8668 F20101221_AAALAM habib_r_Page_045thm.jpg
b51b204e7c94e6dfad96c1c02c00047a
5167c967badc89190a436b35d20e667aea0a3a7f
F20101221_AAAJZD habib_r_Page_138.tif
f69433beda0d3f141cdb54115f1639ea
d9ef52866ce4278be0a3e148291f3a4cb7e734d0
F20101221_AAAKDT habib_r_Page_306.tif
a403c858ab75c7e0e035a9ab8bd25daa
7570235d75707b6bdcfc62063ba0f1689c088ab4







NEW MODEL FOR ANALYZING SOCIOLINGUISTIC VARIATION: THE INTERACTION
OF SOCIAL AND LINGUISTIC CONSTRAINTS




















By

RANIA HABIB


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA INT PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008



































O 2008 Rania Habib





























To my parents: Ibrahim and Amira
To my sister: Suzi
To my brothers: Husam and Faraj
I love you ...









ACKNOWLEDGMENT S

This study owes a great deal to my adviser, Professor Fiona McLaughlin. Although I did

not take a course with her, I had a very nice experience working with her as a Research Assistant

in "The Proj ect on the Languages of Urban Africa." I admired her eloquence and personality

from the time I met her and when she attended one of our Research Methods class as a visiting

professor. In that class, Dr. McLaughlin shared her experience with us about collecting data in

Senegal in Africa as part of an introduction to Field Methods. She has been very kind and

listened closely whenever I felt hesitant towards making a decision. She has been supportive in

my job search and promoting my research and me among colleagues.

This study also owes a great deal to Professor Caroline Wiltshire who has helped me with

the Gradual Learning Algorithm (GLA). My interest in GLA started when I was taking Issues in

Phonology with her. Then, I wrote a paper for that class, using the idea of the GLA. This idea

extended to my study in greater depth. She has been caring and supportive from the time I came

to UF as a Fulbright student. I do not forget the time when I was very sick and she offered to take

me to the doctor and the time she sent me a wonderful report about my performance in my

classes and the professors' evaluation of me.

I would like to thank Professor Atiqa Hachimi for being on my committee and for the

support and comments she has given me. I would also like to thank Professor Jessi Aaron for

agreeing to be on my committee as an external member at a short notice. Her comments have

been valuable. I would also like to thank Professor Andre Khuri for his help in statistics.

Unfortunately, uncontrollable circumstances prevented him from staying on my committee.

I would like to extend my thanks to all the faculty members at the Linguistics Department

who were part of my training in many fields and a window for widening my horizons. Many of

them have been very helpful and willing to discuss some questions regarding my study or were










very interested in my work and progress and supportive of my career development as well as the

promotion of my work among colleagues, such as Diana Boxer, Edith Kaan, Brent Henderson,

Wind Cowles, Ratree Wayland, and Eric Potsdam. I would also like to pass a note of thanks to

all my graduate friends and colleagues at the Linguistics Department, wishing them all the best

in their careers.

I would like to thank my international adviser, Debra Anderson, who has been like a

mother to me in the absence of my real mother. She has been a great companion when I needed

one. She listened and advised whenever she could. Thanks go to all my friends and those who

opened their hearts and homes to me here in Gainesville, FL and in the U.S. Everyone has

contributed in some way to my life.

My greatest appreciation goes to my parents who brought me to this world and have shown

support throughout my life. Their support has made me who I am. I hope you will always be

proud of me. To my sister, Suzi, and two brothers, Faraj and Husam, who have also shown me

great support, I want to say I am proud of you and of being your sister. I love you all so much.

Finally, I want to extend special thanks to all those who participated in this study. Without

you, it would not have been possible. I appreciate your enthusiasm to help and be part of my

work. I am grateful to you for the rest of my life.











TABLE OF CONTENTS


page

ACKNOWLEDGMENTS .............. ...............4.....

LIST OF TABLES ................ ...............12........... ....

LIST OF FIGURES .............. ...............16....

LIST OF ABBREVIATIONS ................. ...............17........... ....

AB S TRAC T ........._. ............ ..............._ 18...

CHAPTER

1 INTRODUCTION ................. ...............21.......... ......


1.1 Proposal .............. .. .. ........ ... ... .. ..................2
1.2 Correlation between Linguistic Variables and Social Factors in the Course of the
Development of Sociolinguistic Methodology .............. ...............30....
1.2.1 Quantitative M ethods ................. .......... ...............3
1.2.2 Long Term Participant Observation Methods .............. ......... .. ............3
1.2.2. 1 Participant observation (ethnography of communication) and
quantitative analysis............... .. ....................3
1.2.2.2 Participant observation and qualitative analysis .............. ........ ................3
1.2.2.3 Participant observation and combining quantitative and qualitative
analyses .............. .. ..... .... .. ......._..._... ...........3
1.2.3 Summary of the above Sociolinguistic Methods ................. .........................3 9
1.3 Introduction to Optimality Theory (OT)............... .. ............ ...............4
1.3.1 Constraint Demotion Algorithm (CDA), Floating Constraints (FCs), and
Stratified Grammars (SG) ............. ... ....._.._...... ...............45.
1.3.1.1 Constraint Demotion Algorithm (CDA) .............. ...............46....
1.3.1.2 Floating Constraints (FCs) .............. ...............46....
1.3.1.3 Stratified Grammar (SG) ........._.......... ......_.._. ......... ..... ...........4
1.3.2 Working Mechanism and Advantages of the GLA over CDA, FCs, and SG........49
1.3.2. 1 Pure linguistic applications of the GLA ......___ ........... ........._.....54
1.3.2.2 Example of a sociolinguistic application of GLA............_.._ .........._._...55
1.3.2.3 Concluding remarks about the GLA .............. ...............58....
1.4 Conclusion ........._._ ...... __ ...............59...
1.5 Research Questions............... ...............6
1.6 Structure of the Study .............. ...............61....

2 METHODOLOGY .............. ...............65....


2.1 Setting: The City of Hims............... ...............65.
2.2 Speech Sample.................. ...............6
2.3 Variables under Investigation ........._._.._......_.. ...............75....











2.3.1 Linguistic Variables............... ...............7
2.3.1.1 The variable (q) .............. ...............76....
2.3.1.2 The variable (6) ................. ...............77.......... ...

2.3.1.3 The variable (a) .............. ...............79....
2.3.2 Social Variables............... ....... ... ........8
2.3.2. 1 Overview of Al-Hameeddieh ...._..._._ ....._._._......_... ................81
2.3.2.2 Overview of Akrama ................. ...._ ...............82. ...
2.4 A analysis .............. ...............83....

3 THEORETICAL ANALYSIS .............. ...............87....

3.1 Introduction to the Theoretical Analysis of the Data. ......____ ............. ..............87
3.2 Modeling Variation between [q] and [7] .............. ...............88....

3.2.1 Linguistic Constraints Pertinent to the Variation between [q] and [7] ...................89
3.2. 1. 1 Grammar of RCA ............ ..... ._ ...............91
3.2. 1.2 Grammar of HCA .............. .. ......__ .... ...__ ..........9
3.2. 1.3 Acquisition of the HCA form by an RCA speaker. .............. ................94
3.2.2 Social Constraints Pertinent to the Variation between [q] and [7] .......................101
3.2.3 Praat Analysis of the Variation between [q] and [7] .............. ....................10

3.2.4 Concluding Remarks on the Sociolinguistic Variation between [q] and [1].128
3.3 Modeling Variation between [t] and [s] and between [d] and [z] ................. ................130
3.3.1 Linguistic Constraints Pertinent to the Variation between [t] and [s] and [d]
and [z] .............. .. ...... .. .... ... ... ... ..............13
3.3.2 Faithfulness and the Variants [t] and [s] and [d] and [z] .............. ... .... ............... 136
3.3.3 Social Constraints Pertinent to the Variants [t] and [s] and [d] and [z] ...............140
3.3.4 Will Social Constraints Affect the Previous Results? .............. .. ..... .................4
3.3.2.1 Concluding remarks on the stable variation among [t] and [s] and [d]
and [z] .............. ...............144....


4 QUANTITATIVE ANALYSES OF [q] VS. [7] .............. ...............145....

4. 1 Introducti on ............... .... .... ......... ....... ..... .. .. .. ...... ..........4
4.2 Tests of Association of Social Class with Education, Income, Residential Area, and
O occupation .............. .... .. ......... .. ... .... ..........4
4.2. 1 Test of Association between Social Class and Education ................. .................147
4.2.2 Test of Association between Social Class and Income ................. ............... ...148
4.2.3 Test of Association between Social Class and Occupation ................. ...............149
4.2.4 Test of Association between Social Class and Residential Area .................. .......150
4.2.5 Concluding Remarks on the Tests of Association between Social Class and
Education, Income, Occupation, and Residential Area .............. ......................5
4.3. Speakers' Distribution of [q] and [7] ................. ........... .. ...........5
4.3.1 Exploring the Type of Statistical Distribution among the Various Groups .........155
4.3.2 Bivariate Correlations Procedure between the Two Dependent Variants [q]
and [7] .............. ...............158....











4.3.3 Generalized Linear Models (GZLM) Procedure for the Effect of the Four
Independent Variables on the Two Dependent Variants [q] and [7] .........................159
4.3.3.1 Negative binomial regression showing the main effects of the four
independent variables, age, gender, residential area, and social class, on the
variable use of the dependent variant [q] ........._........... ....._._.... ........_..........161
4.3.3.2 Negative binomial regression showing the effect of the interaction of
the independent variables, age, gender, and residential area, on the variable
use of the dependent variant [q] .........._.... ... ....._.... ........ ._._ ...........16
4.3.3.3 Negative binomial regression showing the effect of the interaction of
the two independent variables, age and gender, on the variable use of the
dependent variant [q] ................ .... ....... .... ... .........6
4.3.3.4 Negative binomial regression showing the main effects of the four
independent variables, age, gender, residential area, and social class, on the
variable use of the dependent variant [7] .........._........ ......_... ........_._._.....168
4.3.3.5 Negative binomial regression showing the effect of the interaction of
the independent variables, age, gender, and residential area, on the variable
use of the dependent variant [7] ........._...... .. ....___.. .......___. ..........17
4.3.3.6 Negative binomial regression showing the effect of the interaction of
the two independent variables, age and gender, on the variable use of the
dependent variant [7] .............. ...............172....
4.4 Discussion of the Statistical Results ...._.. ...._._._.. .......__... ..........17
4.4.1 A ge .............. ...............173....
4.4.2 Gender .............. ...............177....
4.4.3 Residential area .............. ...............180....
4.4.4 Social class ....................... ... .. .. .. .. .. ... .. .. .................18
4.4.5 Relationship of the Statistical Findings to the Theoretical Analysis. ........._._.......185
4.5 Frequency Effects on the Acquisition of [?] ...._._._.. .... .._.... ......__.. .......18
4.6 Conclusion .........._.... ...............192.__..........


5 QUANTITATIVE ANALYSIS OF THE VARIANTS [t] AND [s] AND [d] AND [z]......200

5 .1 Introducti on ............... .... ....._ ..... ...........20
5.2 Analysis of the Use of [t] and [s] ............... .......... ... ... ...._ ............20
5.3 Discussion of the Frequency Analysis of the Stable Variants [t] and [s] .......................206
5.4 Analysis of the Use of [d] and [z] ............... .......... ... ... .... ._ .... ........0
5.5 Discussion of the Frequency Analysis of the Stable Variants [d] and [z] ......................214
5.6 General Discussion ........... ..... .._ ...............216..
5.7 Conclusion ........... ..... .._ ...............222...

6 CONCLUSION................ ..............22

6. 1 Introducti on .............. ....... ... .. ... ..... .. .. ... .... ............. 2
6.2 Use of OT and the GLA to Account for Switching between Two Different Dialects
and Inter- and Intra-speaker Variation ................ ...............224........... ...
6.3 Incorporating Social Factors into the GLA .............. ...............226....
6.4 Role Played by Social Factors .............. ...............227....












6.5 Consistency of Variation ................. ........... ........... .......... .. ............. .....228
6.6 Markedness Constraints versus Social Constraints .............. ...............229....
6.7 Advantages of and Caveats about the New Model ................ ................. ..........232
6.7. 1 Advantages of the New Model .............. .... ............. ......... ........ ..........23
6.7.2 Caveats about the New Model and Other Limitations of the Study ................... ..23 5
6.8 Frequency effects .................. ........... ..... ....... .... .... .............23
6.9 Implications, Directions and Recommendations for Future Research ...........................238
6. 10 Conclusion .........._.... ...............244.__._. .....


APPENDIX


A TESTS OF ASSOCIATION FOR SOCIAL CLASS WITH EDUCATION, INCOME,
OCCUPATION, AND RESIDENTIAL AREA ...._._._.. ..... ..__... ......__.. ...........4


Cross Tabulations (Contingency Tables)............... ...............245
Social Class Education .............. ...............245....
Social Class Income .............. ...............246....
Social Class Occupation ................ ...............247...............
Social Class Residential Area............... ...............248.


B DESCRIPTIVE STATISTICS EXAMINING TYPE OF DISTRIBUTION IN THE
D A TA .............. ...............250....


Gender ................... ........... ...............250......

qaf: Gender Histograms ................. ...............254...............
qaf: Gender Boxplots ................. ...............255................
hamza: Gender Histograms .............. ...............255....
hamza: Gender Boxplots .............. ...............256....
A ge ................... ........... ...............257......
qaf: Age Histograms ................. ...............261................
qaf: Age Boxplots............... ...............26
hamza: Age Histograms .............. ...............263....
hamza: Age Boxplots .............. ...............264....
A rea................... .. ..............26
qaf: Area Histograms............... ..............26
qaf: Area Boxplots ................. ...............270................
hamza: Area Histograms .............. ...............270....
hamza: Area Boxplots .............. ...............271....
Social Class...................... .............27

qaf: Social Class Histograms............... ..............27
qaf: Social Class Boxplots ................. ...............277...............
hamza: Social Class Histograms .............. ...............278....
hamza: Social Class Boxplots .............. ...............279....

C GENERALIZED LINEAR MODELS RESULTS FOR THE DEPENDENT
VARIABLE [q] .............. ...............280....










Poisson Loglinear Regression Showing the Main Effects of the Independent Variables,
Age, Gender, Residential Area, and Social Class, on the Use of the Dependent
V ariant [q] .............. ... .... ... .. ... .. ... .. .. .. .. ................28
Negative Binomial Regression to Show the Main Effects of Gender, Age, Residential
Area, And Social Class on the Use of [q] .............. ...............283....
Estimated Marginal Means 1: Age ................. ...............286..............
Estimated Marginal Means 2: Gender ................. ...............287........... ...
Estimated Marginal Means 3: Area ................. ...............287..............
Estimated Marginal Means 4: Social Class ................. ............ ... ....... ......... ......28
Negative Binomial Regression Showing the Effect of the Interaction of Three
Independent Variables, Age, Gender, and Residential Area, on the Use of the
Dependent Variant [q] ................... ...... ........... ......... ........ .. .. .. .. ........8
Negative Binomial Regression to Show the Effect of the Interaction of Gender and Age
on the Use of [q] ................. ...............293........... ...

D GENERALIZED LINEAR MODELS RESULTS FOR THE DEPENDENT
VARIABLE [7] .........____...... .____ ...............297....

Negative Binomial Regression to Show the Effect of Gender, Age, Residential Area,
and Social Class on the Use of [7] .............. ...............297....
Estimated Marginal Means 1: Age ........._._._......._.. ....____ ..........30
Estimated Marginal Means 2: Gender ...._.. ...._._._._ .........__. ............0
Estimated Marginal Means 3: Area ...._.. ...._._._._ .........__. ..........30
Estimated Marginal Means 4: Social Class .........._.... ..... ...._... ..... .... ._._.........30
Negative Binomial Regression Showing the Effect of the Interaction of the Three
Independent Variables, Age, Gender, and Residential Area on the Use of the
Dependent Variant [7] ............... ... ... .. ....... ... .... .. ........0
Negative Binomial Regression Showing the Effect of the Interaction of the Two
Independent Variables, Age and Gender, on the Use of the Dependent Variant [7] ........306

E GENERALIZED LINEAR MODELS THE DEPENDENT VARIABLES [q] AND [?]....310

Negative Binomial Regression Showing the Main Effects of Education, Income, and
Occupation on the Variable Use of [q] .............. ....... .. ........... .. ..........1
Negative Binomial Regression Showing the Main Effects of Education, Income, and
Occupation on the Variable Use of [7] ....._____ .... ....___ ...............313.

F INTERNATIONAL PHONETIC ALPHABET (IPA) SYMBOLS FOR STANDARD
ARABIC, HIMSI COLLOQUIAL ARABIC, AND RURAL COLLOQUIAL ARABIC...317

Standard Arabic Phonemic System in IPA ....._ .....___ ........._ ............1
Himsi Colloquial Arabic Phonemic System in IPA .............. ...............317....
Rural Colloquial Arabic Phonemic System in IPA ..........._ ..... ..__ ..........__.......1











LIST OF REFERENCES ................. ...............3.. 19......... ...

BIOGRAPHICAL SKETCH .............. ...............333....


































































11










LIST OF TABLES


Table page

2-1. Distribution of study participants .............. ...............72....

2-2. Variants of the variables (q), (6), and (a) in the speech of non-migrant rural speakers,
native Himsi speakers, and migrant rural speakers ................. ....___ ................ ..75

3-1. Variants of the variable (q) in SA, HCA, and RCA ....._._._ .... ......... ..............8

3-2. Phonemic difference between HCA and RCA speakers .............. ...............90....

3-3. Ranking of RCA: Underlying form /qaraabi/ ................. ...............93......___..

3-4. Ranking of RCA: Underlying form /7alem/ ................. ...............93........... ..

3 -5. Ranking of HCA: Underlying form /7araabi/ ................. ...............94......___ .

3-6. Acquisition of HCA form by RCA speaker: Underlying form /qaraabi/ .............. ..... ........._.97

3-7. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR) .............. ..................107

3-8. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR), manipulating the
output distribution to 100% use of [q] and [7] and 0% use of [?] and [q] when the
input has /q/ and /7/ respectively ................. ...............109..............

3-9. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR), manipulating the
values of the pair distribution to 18% [7] and 82% [q] ................. .........................1 10

3-10. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR), manipulating the
values of the pair distribution to 4% [7] and 96% [q] ................. .......... ...............1 10

3-11. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR), manipulating the
values of the pair distribution to 98% [7] and 2% [q] ................. .......... ...............11 1

3-12. Ranking values and disharmonies of *7, *q, *F[q], and IDENT-IO (RTR), without
manipulation of the values of the pair distribution ................. ....._.. ............... 1 14

3-13. Ranking values and disharmonies of *7, *q, *F[q], and IDENT-IO (RTR),
manipulating the values of the pair distribution to 65% [7] and 3 5% [q] ................... .....1 16

3-14. Ranking values and disharmonies of *7, *q, *F[q], *OLD[?], and IDENT-IO (RTR),
without manipulation of the values of the pair distribution .......___........... ........ .......1 17










3-15. Ranking values and disharmonies of *7, *q, *F[q], *OLD[?], and IDENT-IO (RTR),
manipulating the values of the pair distribution to 37% [7] and 63% [q] ................... .....1 18

3-16. Ranking values and disharmonies of *7, *q, *F[q], *OLD[?], *AKR[?], and IDENT-IO
(RTR), manipulating the values of the pair distribution to 18% [7] and 82% [q] ...........1 19

3-17. Ranking values and disharmonies of *7, *q, *F[q], *OLD[?], *AKR[?], and IDENT-IO
(RTR), without manipulation of the values of the pair distribution .............. .... ...........120

3-18. Ranking values and disharmonies of *7, *q, *F[7], *OLD[?], *AKR[?], and IDENT-IO
(RTR), without manipulation of the values of the pair distribution .............. .... ...........121

3-19. Ranking values and disharmonies of *7, *q, *F[7], *OLD[?], *AKR[?], and IDENT-IO
(RTR), manipulating the values of the pair distribution tol8% [7] and 82% [q] ............121

3-20. Ranking values and disharmonies of *7, *q, *M[7], *OLD[?], *AKR[?], and IDENT-IO
(RTR), manipulating of the values of the pair distribution to 4% [7] and 96% [q] .........124

3-21. RCA input: Ranking values and disharmonies of *7, *q, *F[q], *YOUNG[q], *HAM[q],
and IDENT-IO (RTR), with manipulation of the values of pair distribution to 98%
[7] and 2% [q] .............. ...............125....

3-22. HCA input: Ranking values and disharmonies of *7, *q, *F[q], *YOUNG[q], *HAM[q],
and IDENT-IO (RTR), with manipulation of the values of pair distribution to 98%
[7] and 2% [q] .............. ...............127....

3-23. The variants of the variables (6) and (a) in SA, HCA, and RCA ................. ................. .133

3-24. List of environments for [t] and [s] and [d] and [z] ................. ...............134...........

3-25. Ranking values and disharmonies of *s, *t, and IDENT-IO (Manner), without
manipulation of the values of the pair distribution ................ ................ ......... .137

3-26. Ranking values and disharmonies of *s, *t, and IDENT-IO (Manner) when ranked at
90, 90, 100 respectively .............. ...............137....

3-27. Ranking of *s, *t, and IDENT-IO (Manner): Underlying form /ktiir/ ........._.._... .............138

3-28. Ranking of *s, *t, and IDENT-IO (Manner): Underlying form /masalan/.........................138

3-29. Ranking values and disharmonies of *s, *t, and IDENT-IO (Manner), manipulating the
pair distribution to 100% use of [ktiir] and 100% use of [masalan] ............. ..............138

3-30. Ranking values and disharmonies of *s, *t, and IDENT-IO (Manner) with
manipulation of the pair distribution to 100% use of [ktiir] ..........._... ........_.._........139










3-31. Ranking values and disharmonies of *s, *t, IDENT-IO (Manner), *F[t], and *F[s]
without manipulation of the values of the pair distribution ................. ............. .......143

3-32. Ranking values and disharmonies of *s, *t, IDENT-IO (Manner), *F [t], and *F [s],
manipulating the values of the pair distribution to 100% use of both [ktiir] and
[ m asalan] ................ ...............143..............

4-1. Distribution of [q] and [7] in the speech of each speaker ........................... ...............154

4-2. Bivariate correlations between the dependent scale variants [q] and [7] .............................158

4-3. Distribution of [q] and [7] according to age group ................. ...............174...........

4-4. Distribution of [q] and [7] according to gender ................. ...............177........... .

4-5. Variability in the speech of Speaker-20 .............. ...............179....

4-6. Distribution of [q] and [7] according to age and gender ................. .......... ...............180

4-7. Distribution of [q] and [7] according to age and gender without speakers 29, 30, & 31 .....180

4-8. Distribution of [q] and [7] according to residential area .............. ...............181....

4-9. Distribution of [q] and [7] according to age and residential area ................ ................ ...182

4-10. Distribution of [q] and [7] according to social class ................. ...............183...........

4-11. Most frequent words produced with [q] and [7] ................ ...............194...........

4-12. Percentage of the occurrence of frequently occurring words in the speech of varying
speakers ................. ...............196................

5-1. Most frequent words produced with [t] and [s] .............. ...............202....

5-2. Contingency table showing the frequency use of [t] and [s] words and their conditional
distribution in all speakers .............. ...............204....

5-3. Two tailed One-Sample T-Test comparing the significance of the difference in the
frequency of [t]-words and [s] words............... ...............206.

5-4. Similarity between the vowel system of observed words with [s] with their counterparts
in SA .............. ...............208....

5-5. Assimilated words with [s] to the vowel system of the dialects ................. ............... .....208

5-6. Most frequent words produced with [d] and [z] .............. ...............210....










5-7. Contingency table showing the frequency use of [d] and [z] words in all speakers ............212

5-8. Two tailed One-Sample T-Test comparing the significance of the difference in the
frequency of [d]-words and [z]-words .............. ...............214....

5-9. Similarity between the vowel system of observed words with [z] with their counterparts
in SA .............. ...............2 16...

5-10. Assimilated words with [s] to the vowel system of the dialects ................. ................ ...216











LIST OF FIGURES


FiMr page

1-1. Categorical Ranking .............. ...............50....

1-2. Free Ranking............... ...............50

1-3. C2 > 1............... ...............50...

1-4. C 1 >> C2............... ...............51...

2-1. Map of Syria and neighboring countries. ............. ...............67.....

2-2. City Plan of Hims. Adopted from the Homs City Council (2008). I point to the areas of
concern in this study and give their name in English, as a point of reference. .................. 86

3-1. Ranking of RCA ................ ...............98...............

3-2. Ranking of HCA form .............. ...............99....

3-3. Ranking of the three constraints IDENT-IO (RTR), *7 and *q in an oscillating speaker......100

4-1. Bar chart for the distribution of the six educational groups between lower-middle and
upper-middle classes............... ...............148

4-2. Bar chart for the distribution of the three income groups between lower-middle and
upper-middle classes............... ...............149

4-3. Bar chart for the distribution of the four occupational groups between lower-middle and
upper-middle classes............... ...............150

4-4. Bar chart for the distribution of the lower-middle and upper-middle classes in the two
residential areas: Akrama and Al-Hameeddieh .............. ...............151....

4-5. Histogram showing the skewness to the left in the females' use of [q] .............. ................157

5-1. Distribution of [t]-words between high and low frequency .............. .....................0

5-2. Distribution of [s]-words between high and low frequency ........._._.... ....._.__............205

5-3. The distribution of [d]-words between high and low frequency .............. .....................1

5-4. The distribution of [z]-words between high and low frequency .............. .....................1









LIST OF ABBREVIATIONS


AKR

ECDA/CDA

F

FC(s)

GLA

HAM

HCA

L1

L2

LM(C)

LWC

M

MGLA

NR

OT

PF

PR

RCA

SA

SF

SG

UG

UM~(C)

UWC


Akrama

Error-driven Constraint Demotion Algorithm

Female

Floating constraints)

Gradual Learning Algorithm

Al-Hameeddieh

Himsi Colloquial Arabic

Mother/first language

Second language.

Lower middle (class)

Lower working class

Male

Minimal Gradual Learning Algorithm

Non-prestigious area

Optimality Theory

Prestigious form

Prestigious area

Rural Colloquial Arabic

Standard Arabic

Stigmatized form

Stratified Grammars or partially-ordered grammar

Universal Grammar

Upper middle (class)

Upper working class









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

NEW MODEL FOR ANALYZING SOCIOLINGUISTIC VARIATION: THE INTERACTION
OF SOCIAL AND LINGUISTIC CONSTRAINTS

By

Rania Habib

August 2008

Chair: Fiona Mc Laughlin
Cochair: Caroline Wiltshire
Major: Linguistics

In this study, I present a new model for analyzing sociolinguistic variation within the

framework of Optimality Theory (OT) and the Gradual Learning Algorithm (GLA). This model

contributes to the advancement of sociolinguistic methodology as well as to OT and the GLA,

unifying both linguistic and social factors. I propose a number of social constraints and

incorporate them with linguistic constraints in the GLA. I show that incorporating social

constraints yields variant-usage percentages that match real life occurrences. I follow the

evaluation process of stochastic grammar in which social constraints are treated like linguistic

constraints on a continuous scale of ranking strictness. Their ranking values may differ from one

speaker or group of speakers to another. These differences in the ranking values may result in

intra- and inter-speaker, intra- and inter-group, intra- and inter-community, or intra- and inter-

dialect variation. The number and type of social constraints involved influence the percentage of

occurrence of each variant and affect the ranking values of other constraints and the grammar

chosen by a speaker or a group of speakers at a particular time and place.

This model enables us to show the effect of linguistic and social constraints

simultaneously. It allows us to investigate individual as well as group grammars and shows that









those grammars may act independently from each other. It also enables us to proj ect expected

percentages of variation by manipulating either the ranking values of constraints or the pair

distribution of the output. Evidence comes from the study of the naturally occurring speech of

migrant rural speakers of Colloquial Arabic to the city of Hims, Syria. These speakers show

different degrees of variation, particularly regarding the use of the two variants [q] and [7], based

on various social factors, such as age, gender, residential area, and social class.

The aim of the study is to show how rural migrants adopt the new phonological system of

the urban dialect to appear prestigious. The study will focus on the sound change of [q] to [7] and

the four stable variants [t] and [s] and [d] and [z]. It will show that variation may be influenced

by different factors. The variable use of [q] and [7] is attributed to prestige and social factors.

The use of [t] and [s] and [d] and [z] is not social in nature; it has developed historically from the

Standard Arabic (SA) [6] and [a] respectively as a response to markedness constraints. Today,

this variation is stable and each of the four variants is used in specific lexical items. Faithfulness

constraints play a maj or role in maintaining the pronunciation of the input as [t] and [s] and [d]

and [z] in the output. This stable phenomenon is further explained in terms of the two opposing

effects of frequency (Bybee 2001).

In a situation where prestige plays a role in adopting a new variant, as in the case of rural

migrants to the city of Hims, social constraints are viewed as the motivating force behind

changing a grammar at a particular time and place. However, one should take into account that

the activation of social constraints depends on the speaker' s selection or choice to activate them

or not. Hence, these constraints may rank very high in the speech of a speaker who is highly

aware of the social values attached to a certain sound. On the other hand, they may rank very low

in the speech of a person who, for example, does not care to adopt a new form.









A theory that can take into account all the factors that lead to variation is a better theory

than one that takes into account social factors in isolation and ignores grammatical factors or

vice versa. Integrating social constraints into formal theory and showing that social constraints

can have equal weight or even more weight than grammatical constraints in conditioning

variation and change is essential in this study. Including social constraints in the computation

provides explanation of the observed sociolinguistic variation between [q] and [7] among

members of the same social group or among different speakers or social groups. Interacting

social constraints with linguistic constraints in the same framework is a simple comprehensive

method to depict and explain the mental process of a speaker at a certain time and/or place.

Feeding the GLA with the right output distribution gives the specific ranking values or grammar

of each speaker or group of speakers. In other words, the model takes advantage of the stochastic

grammar embedded in the GLA to generate grammars that match real life output percentages

without trying all the possible rankings of constraints and counting the times those rankings give

a certain output. Furthermore, where a statistical analysis fails to indicate the interaction between

one social factor and others, the specificity implemented in the GLA by dividing a social factor

into a number of social constraints enables us to see interaction among the same social

constraints that emerged as insignificant in the statistical analysis. In this sense, implementing

social factors as constraints and accounting for sociolinguistic variation within the framework of

OT and the GLA has an advantage over other theories.









CHAPTER 1
INTTRODUCTION

1.1 Proposal

In this study, I present a new model for analyzing sociolinguistic variation, employing

Optimality Theory (OT) (Prince and Smolensky 2004 [1993]) and the Gradual Learning

Algorithm (GLA) (Boersma and Hayes 2001). This model contributes to the advancement of

sociolinguistic methodology as well as to OT and the GLA, unifying both linguistic and social

factors. I propose a number of social constraints and incorporate them into the GLA, where they

intersect with linguistic constraints, and show that incorporating social constraints with linguistic

constraints in the GLA yields results that reflect real life performances. Introducing social

constraints, for the first time, as integral part of linguistic theory could be considered an

advancement to OT and the GLA, which so far dealt only with linguistic constraints. As the GLA

is based on continuous ranking and stochastic constraints, these social constraints will be treated

as continuous constraints whose ranking values differ from one speaker to another, from one

group of speakers to another, and from one community to another. These differences in the

ranking values may result in inter-speaker, intra- and inter-group, intra- and inter-community, or

intra- and inter-dialect variation. These ranking values may even differ within the same speaker

yielding intra-speaker variation. The number and type of social constraints involved influence

the percentage of occurrence of each variant and affect the weight of other constraints and the

grammar chosen by a speaker or a group of speakers at a particular time and place.

This model enables us to show the effect of linguistic and social constraints

simultaneously. It gives a mental representation of the grammatical process that takes place in

the speaker' s mind as a result of the integration of social and linguistic constraints. It allows us to

investigate individual as well as group grammars and shows that those grammars may act










independently from each other. In this sense, the model allows us to see that certain social

constraints, for example, may be activated by a certain speaker, but not others, and by one group

of speakers or more, but not others. Thus, in the same way that linguistic constraints have their

language specific ranking, social constraints maintain their relativity and specificity in the

particularity of the individual grammars and constraint rankings of individual speakers or groups

of speakers. The model also enables us to proj ect expected percentages of variation by

manipulating the weight of the various constraints or the output pair distribution. Interestingly

enough, when constraints are set at an initial equal ranking value, the addition of one or more

social constraints) changes the ranking value of other constraints and gives us predictions on

what variation will look like if those social constraints are involved. Certain social constraints

give us expectations on what the speech of a particular speaker or group of speakers should

sound like. Furthermore, feeding the GLA with the right output distribution generates grammars

that match real life output percentages of each speaker or group of speakers without trying all the

possible rankings of constraints and counting the number of rankings that give each output.

Evidence comes from the study of the naturally occurring speech of migrant rural speakers of

Colloquial Arabic to the city of Hims, Syria. These speakers show different degrees of variation,

particularly regarding the use of the two variants [q] and [7], based on various social factors,

such as age, gender, residential area, and social class. Their variation may also depend on the

interlocutor: a family member or a stranger.

The ability of the model to give a mental representation of what goes on in the speaker' s

mind and the conscious choice of a ranking value at a certain time and place endow the model

and the proposed linguistic and social constraints with psychological reality. The psychological

reality of grammars is a broad issue that is hard to determine or support completely. The idea of










the psychological reality of rules and representations goes back to Chomsky. However, many

expressed skepticism about Chomsky's (1980) argument that rules and representations are part of

a grammar and a mental process (i.e., rules are psychologically real). Devitt (2006a), for

example, is skeptic of rules and principles and believes that they are not psychologically real.

Devitt (2006b) believes that there is only a "linguistic reality" that is mainly associated with

what is produced (whether spoken, signed, or written) (i.e., "the real subj ect matter of grammars"

(Slezak 2007)). Devitt' s (2006a) objection to Chomsky's view of the psychological reality of

grammars as mental representations was critiqued by Slezak (2007) who observed that the

difference between Devitt' s view and Chomsky's is merely terminological. Furthermore, Aske

(1990), in testing the psychological reality of Spanish stress rules, arrived at results that

contradict the traditional idea of the mental representation of rules (i.e., rules are not

psychologically real). Rather speakers access their lexicon when assigning stress to new words

they never heard before. They assign stress by analogy to existent patterns in the brain/lexicon.

Thus, patterns, not rules, are mentally represented. Similarly, Bybee and McClelland (2005)

argue against rule-based systems and the mental representation of rules in support of exemplar

models and connectionist models. They adopt the view that "specific experiences" have an

impact "on the mental organization and representation of language" (p. 381). For them,

language use has a maj or impact on language structure. The experience that users
have with language shapes cognitive representations, which are built up through the
application of general principles of human cognition to linguistic input. The structure
that appears to underlie language use reflects the operation of these principles as they
shape how individual speakers and hearers represent form and meaning and adapt
these forms and meanings as they speak. (p. 382)

Bybee and McClelland's view does not eliminate a cognitive representation of some kind

of form and meaning that are influenced by language use rather than by innate universal rules. In

this sense, language use may influence the mental representation of language. This view is not










very far apart from the influence that social factors could exert on the mental representation of

language in a sociolinguistic variation situation. In a sociolinguistic contact situation, speakers

are most likely aware of a linguistic change around them. For example, the Casablancan dialect,

which developed because of the extensive contact between different regional dialects of migrants

from rural and urban areas, was described towards the end of the twentieth century as an

"interdialect" that has psychological reality for Moroccans (Moumine 1990, cited in Hachimi

2007). Furthermore, Milroy (2006:151) espouses the view that "linguistic change is a process"

and some aspects of this process are "related to social factors". For Milroy, language cannot

change on its own. It is a social phenomenon that involves a speaker and a listener. There is a

distinction between speaker innovation and linguistic change. Social factors constitute part of

any type of linguistic change because even internal changes involve production on the part of the

speaker and perception on the part of the listener. If we take Milroy's argument into

consideration, we can conclude that in a sociolinguistic variation situation, people are aware of

the linguistic changes surrounding them and are most likely involved in them.

Returning to my initial assumption that the model and constrains proposed in this study are

psychologically real, I add that the speaker' s awareness of the social significance of a sound and

the change of the ranking values of constraints to sound prestigious involves learnability of new

ranking values. Learnability is usually associated with some mental grammatical processes. This

idea may contradict with Rumelhart and McClelland's (1987) idea that learnability and language

use do not involve rules because of the probabilistic nature of natural language and the

variability in language use. However, I argue that variability in language use could be the result

of external factors that influence the choice of one ranking value over another. Choice is related

to consciousness (of grammatical processes and the likes), and thus has some mental realization










and psychological understanding of what is going on in a social situation in which a speaker

switches from one form to another and is noticed by listeners. In support of this view, I present

Cutillas-Espinosa' s (2004) view that a speaker can control the ranking values of his/her grammar

consciously (Section 1.3.2.2). For him, variation is not mechanical or automatic; rather, it is

based on personal conscious choices. In this way, Cutillas-Espinosa opposes the Neogrammarian

view that change is mechanical and automatic. Many researchers have argued, however, that

repetition and frequency may lead to the automatization of words and sequences of words (e.g.

Bybee 2001, 2003). That is, words or sequences of words form chunks in the brain and become

entrenched in their own representation and thus easier to access and use. This automatization is

applied also to social identity and information about the interlocutor and social settings. In other

words, speakers access this stored information along with the form or structure suitable for the

social setting or situation. Labov (1994:604), on the other hand, supports the Neogrammarian

view that language change is mechanical and automatic when it is "outside the range of

conscious recognition and choice." In this way, any conscious attempt to change language is

subj ect to "higher-level stylistic options" or adherence to social factors and pressures. In this

sense, sociolinguistic change is mainly conscious and involves choice. Whether this change

becomes automatic or not later on in life is beyond the scope of this study and is worth

investigating in further research. However, it is noteworthy that for a sound change to be

considered automatic, it should not be the "goal of anyone at any time" (Boersma 1997b,:2).

Since the change from [q] to [7] is driven by prestige, it is teleological, and thus follows a


conscious process at least in the initial stages of learnability of the new prestigious form [?]. In

addition, the stigma that is associated with the rural form [q] encourages its users to abandon it

and adopt the prestigious form instead. Stereotype variables such as (q) which are avoided by









their own native speakers are indicative of the social awareness of them and their stigma as well

as of the prestige of the form those speakers switch to, in our case [?].

Moreover, the social awareness that leads to learnability of new ranking values of

constraints provides evidence of the speaker' s consciousness of those constraints and the process

that accompanies them, not only in him/herself but also in other speakers or interlocutors who

tend to switch to a different sound or dialect or maintain his/her native sounds or dialect. A rural

migrant speaker to the city of Hims, for example, is aware of the accommodation to the Himsi

prestigious form, [7]. This awareness is expressed by many speakers in the recorded

conversations. Those speakers notice their own shift as well as the shift of others towards the use

of [7] with strangers or non-family members. A number of speakers mentioned their observations

of other speakers changing their speech and using [7] instead of [q] outside the family sphere

(Habib 2005, Sections 5.2 and 5.3). This awareness reflects on the psychological reality of the

different grammars and ranking values of constraints in each individual or group of individuals.

Awareness of such grammatical change in the other is a further reflection on the psychological

reality of the model and constraints. What makes this model interesting is that it does not

separate the social reality of language from the psychological reality of language. It integrates

both realities in the same framework, presenting the external influence on language as part of a

speaker' s internal process to sound externally prestigious.

The aim of the study is to show how rural migrants adopt the new phonological system of

the urban dialect to appear prestigious. The study will focus on the sound change of [q] to [7]; [a]


to [d], [z], or [a]; and [6] to [t], [s], or [6]. It will show that variation may be influenced by

different factors. Variation in the use of [q] and [7] is attributed to prestige and social factors.









Variation in the use of [d] and [z] and [t] and [s] is not social in nature; it has probably developed

historically as a response to markedness constraints. Today, this variation is stable and each of

the four variants is used in specific lexical items. Consequently, the underlying form of these

four variants is /d/ and /z/ and /t/ and /s/ respectively. Faithfulness constraints play the maj or role

in maintaining the pronunciation of the input as [d] and [z] and [t] and [s] in the output to

preserve the specificity observed in lexical items, in that each variant is designated a particular

phoneme with its correspondent allophone. Thus, the historical development led to the

disappearance of /a/ and /6/ from the phonemic system of speakers; they rarely appear as [a] and


[6] in colloquial speech. If they do, they appear as lexical borrowings from SA. Even then, they

may appear as [z] and [s] respectively. This is further attributed to markedness.

In a situation where prestige plays a role in adopting a new variant, as in the case of rural

migrants to the city of Hims, I view social factors as constraints that were inactive when the

speaker was living in his/her hometown. When this speaker moves to the city, s/he starts

activating those social constraints in addition to linguistic constraints based on positive evidence

or negative comments that s/he may hear from city people. The situation is analogous to a child

who is born with no previous knowledge of his/her mother tongue and based on positive

evidence, s/he starts activating the appropriate constraints and changing the ranking of those

constraints as s/he receives more input (e.g., Demuth 1995; Gnanadesikan 1995; Smolensky

1996). The difference between a child and an adult is that the child is not yet influenced by

society and social factors. In the case of the adult, social factors, such as prestige, societal

attitude, age, gender, social class, residential area, education, occupation, and social networks

potentially influence linguistic performance. Thus, I view social constraints as the motivating

force behind changing a grammar at a particular time and place. In this sense, the two types of









constraints linguistic and social interact to arrive at the desired result. However, one should

take into account that the activation of social constraints depends on the speaker' s selection or

choice to activate them or not. If the speaker chooses not to activate them or to activate just some

of them, s/he may continue to use his/her mother dialect or rather vary his/her speech, resulting

in intra- and inter-speaker variation. Many researchers emphasize the role of social factors in

affecting a change (e.g., Labov, 1963, 1966, 1972a, 2001; Trudgill 1974; Milroy 1980; Eckert

1991a; Haeri 1996). For example, Labov (2001:498) emphasizes that social factors are "the

forces that move and motivate change, and are responsible for incrementation and transmission

across generations." Since social constraints can be the impetus for grammatical change, then we

can assume that they can play a maj or role in intra- and inter-speaker variation. Hence, these

constraints may rank very high in the speech of a speaker who is highly aware of the social

values attached to a certain sound. On the other hand, they may rank very low in the speech of a

person who, for example, does not care to adopt a new form or is too old to think about changing

his/her speech or his/her identity for this matter. Cases such as these have been observed in some

studies, such as Labov' s study of Martha' s Vineyard (1963). In Martha' s Vineyard, the Chilmark

fishermen, for example, showed strong defiance to forms from the mainland of New England;

they clung to the island's old ways of pronouncing (ay) and (aw). That is, they maintained the

island's centralized features. The younger age group also showed more centralization than most

age groups to show strong identification with the island and to distinguish themselves from the

summer tourists who come from the mainland. In addition, situations in which linguistic

behavior differs with age are very common. For instance, Miller (2005) found that there is a

great difference of accommodation to the Cairene forms between the first generation migrants to

Cairo and the second generation migrants who were born in Cairo. The second generation shows










complete accommodation to the Cairene forms, whereas the first generation migrants show

variation in their accommodation to the new forms. Furthermore, since those social constraints

will be considered on a continuous scale, they may lead to variation even within the same person,

leading to oscillation in the choice between two variants in the same conversation regardless of

the interlocutor. In this case, the speaker is trying to adjust his/her social constraints to fit in the

community. At the same time, s/he is also trying to adjust her/his grammatical constraints to fit

the social requirement. The result is that the speaker is probably confused and is unable to have a

good grip on one particular grammar and s/he will continue to oscillate in selecting different

ranking values and rankings for the various constraints.

Consequently, a theory that can take into account all the factors that lead to variation is a

better theory than one that takes into account social factors in isolation and ignores grammatical

factors or vice versa. Integrating social constraints into formal theory and showing that social

constraints can have equal weight or even more weight than grammatical constraints in

conditioning variation and change is an important development in this study. Including social

constraints in the computation provides explanation of the observed sociolinguistic variation

between [q] and [7] among members of the same social group or among different speakers or

groups of speakers. Interacting social constraints with linguistic constraints in the same

framework is a simple comprehensive method to depict and explain the grammatical mental

process that takes place in the mind of a speaker at a certain time and/or place. Furthermore,

where a statistical analysis fails to indicate the interaction between one social factor and others,

the specificity implemented in the GLA by dividing a social factor into a number of social

constraints enables us to see interaction among the same social constraints that emerged as

insignificant in the statistical analysis. In this sense, combining OT and the GLA and including









social constraints in the computation have an advantage over other theories in accounting for

sociolinguistic variation and change. Traditional generative phonology manipulated grammatical

rules to deal with variation, which have their problems (Section 1.3). Traditional OT only

manipulated the rankings of linguistic constraints to analyze variation (e.g., Kochetov 1998 on

variation among four Polish dialects; Morris 2000 on variation among three Spanish varieties).

Sociolinguistic methods of analyzing linguistic variation focused on correlations between social

factors and linguistic variables with more emphasis on the development of methods of data

collection. In Section 1.2, I will give a brief background on the sequential development of

sociolinguistic methodology before presenting my model to allow the reader to compare my

model with previous ones.

1.2 Correlation between Linguistic Variables and Social Factors in the Course of the
Development of Sociolinguistic Methodology

Since the mid 1960s and with the foundation of sociolinguistics by William Labov, social

factors started to gain importance in the field of linguistics and to play an important role in

analyzing and modeling speech. In Sociolinguistic Patterns (1972a: 163), Laboy stresses that

"[t]he process of sound change is not an autonomous movement within the confines of a

linguistic system, but rather a complex response to many aspects of human behavior." Further,

in Principles of Linguistic Change (1994: 1), Laboy asserts that "[t]he separation of 'internal'

from 'external,' 'linguistic factors' from 'social factors' may not seem practical to those who

view language as a unified whole where tout se tient, or those who believe that every feature of

language has a social aspect." Throughout decades of sociolinguistic studies of variation, a great

shift in views took place, a shift from viewing language as reflection of the social to viewing

language as creator of the social (Rickford and Eckert 2001). The shift starts with Labov' s (1966,

1972a; Trudgill 1974) view of style variation as different levels of attention paid to speech and









ends with the constructivists' view of the use of style to proj ect a self-image and to construct

identity and social meaning (e.g., Eckert 1991b; Coupland 1985, 2001; Shilling-Estes 1999,

2002; Cameron 1998). Other views were also formed, such as Bell's (1984, 1991) model of

'audience design' in which a speaker' s style is seen as a response to an audience. Bell viewed

style as a reflection of social variation, whereas Finegan and Biber (1994) viewed social

variation as a reflection of style. This is not to mention accommodation theory (Giles and

Powesland 1975; Giles, Coupland and Coupland 1991) which draws on the speaker' s orientation

and attitude to the interlocutor and on the role of identity (Coupland 1980) in determining

speakers' style and their perception of style. In recent formal models, stylistic variation is

starting to be viewed as gradual, not abrupt (Boersma and Hayes 2001); it involves optionality

and learnability.

1.2.1 Quantitative Methods

Linguistic and stylistic variation was mainly measured by correlating various social factors

with the production of a certain linguistic variant in a particular way. Being the founding rock of

sociolinguistics, Labov's quantitative methods of data analyses and modeling have influenced

many Arabic sociolinguists (e.g., Abd-el-Jawad 1981; Holes 1986; Walters 1989; Haeri 1991;

Abu-Haidar 1992; Daher 1998; Habib 2005). In his study of New York City (r) (1966, 1972a),

Laboy correlated a number of social factors age, sex, social class with the pronunciation of

postvocalic r as (r-1) or (r-0) using various statistical methods. To examine the social

stratification of (r) in the speech of New Yorkers, Laboy conducted a rapid and anonymous

survey to avoid the observer's paradox. He selected three department stores in three different

locations in Manhattan, which are socially stratified: Saks Fifth Avenue (a high-status store with


S(r-1) stands for r-full or the pronunciation of the variable (r) as [r]; (r-0) refers to the deletion of the variable (r) in
pronunciation, that is it is realized as null or not realized at all, [0].










high prices), Macy's (middle-class with middle prices), and Klein's (a store with cheaper prices

for less affluent customers). In order to show that his hypothesis is general, he chose one

occupational group, salespeople, to show that social differences occur even within the same

occupational group, based on the status of the store and the customers. His prediction that the use

of(r-1) is socially stratified was confirmed. Saks' employees used (r-1) sixty percent; Macy's

employees fifty one percent; and Klein's employees twenty percent. However, in the more

deliberate answer, all groups showed an increase in (r-1). What was striking is that the middle-

class store showed the highest degree of increase, which was attributed to hypercorrection among

the middle class. The study showed minor differences between men's and women's linguistic

behavior. There was also no strong correlation with age. While the younger generation in Saks

showed higher use of (r-1), the same was not observed in Macy's and Klein's. Rather, the

middle-aged group of the lower-middle class showed hypercorrection; it used more (r-1) than the

upper class, which is a sign of 'linguistic insecurity' (Labov 1972a, 2001). This survey was the

starting point for a larger proj ect, expanding to the Lower East Side of New York, in which the

findings were no different.

In the Lower East Side of New York study (1972a), Laboy sought to gather a more

representative sample of the city than its salespeople, using extensive interviews. He used a ten-

point scale to classify the various social classes. Laboy assigned 0-1 for lower working-class

(LWC), 2-4 for upper working-class (UWC), 5-8 for lower middle-class (LMC), and 9 for upper

middle-class (UMC). In Labov's study, the division among social classes was most evident

concerning the variable (th) that is realized as [6], [t8], and [t]. The latter is the most stigmatized


form and the most common among the LWC, whereas [6] emerged as the most common among


the UMC. This is evident in the increase in the use of [6] with the shift of style from casual









speech to word lists. The (r) variable showed similar results, but slightly smaller division among

social classes. There was also increase in the use of (r-1) in the speech of participants along the

style shift towards the most formal or minimal pairs. In this study, Labov used the "matched

guise" technique that was developed by Wallace Lambert (Lambert, Hodgson, and Fillenbaum

1960), which is a subjective reaction test, in order to see if all social classes evaluated the (r-1) in

the same way and whether New York City constitutes a speech community. Participants were

asked to listen to a tape, containing 24 sentences from five female readers. The same reader

pronounced the same sentence twice: once with [r] and once without it, but listeners were not

aware that the same speaker had said both utterances because sentences were randomized.

Participants were asked to judge the readers on a scale of occupational suitability (i.e., will the

speaker be acceptable as a television personality, secretary, factory worker, etc). All participants

age 18-39 positively evaluated (r-1). Based on the uniformity of those subj ective evaluations,

Laboy concluded that New York City forms a speech community and that (r-1) is the prestige

marker of the city.

1.2.2 Long Term Participant Observation Methods

Rapid surveys and individual interviews are not the only techniques advocated by Labov;

rather, in later work, Labov (1972b) conducted with the help of fieldworkers a long-term

participant observation of adolescent gangs in Harlem, an African American neighborhood in

New York City. He also applied this method in his 'neighborhood studies' in Philadelphia

(starting in the 1970s). Such neighborhood studies aimed to obtain a large amount of social and

linguistic data, treating individual neighborhoods as social units, maintaining the sociolinguistic

interview, but abandoning random sampling. Regardless of the method of data collection,

consequent work proceeded with correlating linguistic variables with social constraints (e.g.,

Milroy 1980; Gal 1979; Eckert 1989, 1991b). While studies like Milroy's (1980) employed










quantitative methods of analyses, studies like Gal's employed qualitative methods of analyses.

Other studies such as Eckert's incorporated both qualitative and quantitative methods of

analyses. The combination of the two types of analyses may be preferable for a better

understanding of a community in ethnography-of-communi cation studies (Hymes 1972) that

involve becoming an in-group member, gaining acceptance and integrating into the community.

1.2.2.1 Participant observation (ethnography of communication) and quantitative analysis

The ethnography of communication approach to data collection was developed by Milroy

and Milroy (1985, 1992) in their investigations of speech in Belfast. Their concern was

vernacular maintenance and their hypothesis held that the use of vernacular forms is related to

the speaker' s integration into the community's social network. The two maj or concepts in the

model of social networks (Milroy 1980) were density and multiplexity. High-density and

multiplex networks exhibit stronger relationships among their members in contrast to low-

density and uniplex networks. Usually dense and multiplex networks are characteristic of rural

villages and working-class areas. Milroy (1980) carried out her fieldwork between 1975-6 in

three working-class areas in Belfast: Ballymacarrett, the Clonard, and the Hammer, all of which

were characterized as dense and multiplex networks. Milroy's concern was how to access the

most natural speech from these social networks. She successfully entered the community as the

friend of a friend, claiming connections with students from the area. She was able to meet more

informants through her original contacts. Her use of the participant observation strategy of data

collection prompted her to become part of the community. Consequently, Milroy was able to

collect a variety of natural speech styles in different situational contexts. Rather than conducting

the formal sociolinguistic interviews that were implemented by Labov (1972a), she recorded

natural conversations among informants. Milroy applied quantitative analysis to the data

collected from forty-six participants. The analysis showed that a number of phonological









variables were stratified according to gender in the three working-class areas. The effect of

gender was particularly evident in the use of the variable (th); the vernacular was characterized

by deletion of (th) intervocalically (e.g., M~other becomes [moor]). In both age groups (18-25 and

40-50), women used fewer vernacular forms than men, particularly younger women. This

difference between men and women within the same community was attributed to their social

networks, strong ties with the local people, and degree of integration into the community. To

calculate the degree of integration into the community, Milroy (1980) developed a six-point scale

from 0-5 to measure the degree of density and multiplexity of each individual network. This

method is called Network Strength Scale or NSS. The analysis showed that male network scores

were higher than those of female networks, indicating that men had stronger ties with the local

community than women and that their use of the vernacular is a way of showing solidarity. The

Milroys (1985, 1992) found that language use is affected by status and solidarity and that the use

of standard language is associated with high social status, whereas the use of the vernacular, and

thus covert prestige, is an instantiation of solidarity with the local community's linguistic and

social norms and customs. The norms do not have to be prestigious; they are the dominant ones

in the community.

1.2.2.2 Participant observation and qualitative analysis

Like Milroy's study in Belfast, Gal (1979) conducted a one-year participant observation in

Oberwart, a town in eastern Austria, near the Hungarian borders. She also had to become part of

the community, observing people's behavior and recording examples of language use. In contrast

to Milroy's use of quantitative analysis, she used qualitative analysis of the data. The maj ority of

the inhabitants of Oberwart were Hungarians who were bilingual in German. Language,

according to Gal became an indicator of social status: Hungarian was associated with










"peasantness", as Gal terms it, and German was associated with higher status: the language of

prestige, money, modernity, and economic prosperity. From closely studying the linguistic habits

of individuals and groups, she was able to identify the motivating factors for the shift from

Hungarian to German. The informants came from eight households and their visitors. 68

speakers constituted her sample: 37 women and 31 men. Gal observed when they used

Hungarian, when they used German, and when they used both. Initially, she used the traditional

sociolinguistic interview devised by Labov, but she was only able to elicit a narrow range of

styles. She used the interviews to elicit information about informants' language use in different

contexts and their daily contacts with others. After spending more time in the community, she

was able to record naturally occurring conversations among her informants in a variety of

settings. Using an ethnographic approach, Gal observed and tried to make sense of naturally

occurring linguistic behavior of the participants. For this reason, her study contrasts with survey

studies and those based on the formal interview. Gal found that the two factors that strongly

correlated with the choice of a language were age and peasantness. The latter reflects the

individual's social status: a peasant or a waged worker. Age played a role in the choice between

German and Hungarian: the younger generation preferred the use of German, whereas older

speakers preferred the use of Hungarian. However, the choice of language also depended on the

person spoken to: peasant or Austrian/urban. Younger speakers are associated with urbanization

and since German is associated with urbanism, it is the language chosen when speaking to

persons associated with 'urbanization' or 'Austria'. Gal also found that social networks played a

role in language choice and noticed the great difference in men's and women's behavior with

respect to such choices. Women in general were faster to adopt the German language as a sign of

their rej section of peasant life. They preferred marriage to a wage laborer than to a peasant









because they "do not want to be peasants; they do not present themselves as peasants in speech"

(Gal 1978:13). This forced peasant men to look for marriage outside Oberwart where peasant life

is less stigmatized. These marriages accelerated language shift because the outsider women were

usually monolingual German speakers and children of such marriages became monolingual

German speakers. This behavior is reflected in the younger generation's preference for German.

Thus, both local young women and incoming women prompted the language shift from

Hungarian to German in the community.

1.2.2.3 Participant observation and combining quantitative and qualitative analyses

Other works that used the participant observation technique are Eckert' s (1989, 1991b)

studies of high-school adolescents in Detroit (i.e., the two adolescent groups: the Jocks and the

Burnouts). Rather than assuming a class continuum, like Labov, she applied previous social

grouping in the community under investigation. Eckert (1991b:213) believed that the best way to

understand membership in a social group and the influence of such membership on the linguistic

behavior of the members of that social group was through long-term observation and

ethnographic methodology in studying variation:

The use of ethnography in the study of variation allows the researcher to discover the
social groups, categories, and divisions particular to the community in question, and to
explore their relation to linguistic form. It is in these small-scale studies, ultimately, that
we can directly observe social process at work in linguistic variation and change.
(1991b:213)

Eckert spent two years observing a graduating class of a Detroit suburban high school. The

ages of those participants did not exceed 20. The differential linguistic behavior of these two

social groups is manifested not only quantitatively but also qualitatively to show the opposed

directions in variant choice between the two groups in a situation of sound change in progress.

One would expect the adolescents' identity and variation to correlate with their parents'

socioeconomic status; rather, their identity and variation correlated with "adolescent social group









affiliation", as Eckert commented (1991b: 214) "[t]he main business of adolescence is the

accomplishment of separation from the family and the development of a social identity defined

in terms of the larger society". The Jocks are representative of the social category whose lives

revolve around the school and who gain status by participating in the school's activities and

extracurricular activities. In contrast, Burnouts reject the school's and their parents' domination

over their lives. The two categories differ in every aspect of life: appearance, the music they

listen to, clothing, behavior, etc. Because of this social and identity polarization, one would

expect difference in their linguistic behavior.

One of the variables that was under investigation was (e) which is realized as two backed

variants [A] or [a], in addition to the less conservative form [e] and the two front variants [ne] and


[I]. The Burnouts exhibited more backing of (e) to [A] than the Jocks. On the other hand, the

Jocks led in the lowering of (e) to [ne]. The Burnouts' urban contacts and orientation made them

adopt urban forms, whereas the Jocks' suburban association prevented them from adopting such

forms. Hence, they were seen by the Burnouts as conservative and "talk just like their parents"

(1991:220). As a result, the Jocks decided to develop their own innovative form, [ne], in order not

to appear conservative and to establish a signaling identity. None of the groups rej ected each

other' s variant; rather they developed their own as a signal of their identity and preservation of

the split and distinction between the two groups. The centralized variant and the fronted one also

corresponded to each group's demeanor. The Jocks were known for their juvenile nature; they

were always smiling and making a point in greeting and smiling to people in the halls. On the

other hand, the Burnouts were always somber and always looked at the Jocks as privileged.

Because of their parents' low socioeconomic status, they perceived themselves and were

perceived by others as people with problems. Eckert commented:









The open-faced, smiling demeanor of the Jocks and the more somber aspect of the
burnouts correspond to their choice of variants: the fronted, spread [ne] variant for Jocks
and the central, unspread [A] for burnouts. (1991b:230)

The VARBRUL analysis showed that sex differences were insignificant. Eckert's

conclusion was that gender should not be separated from social class; rather, gender and social

status interacted as a Jock or a Burnout and urban-suburban orientation. For example, the

Burnouts led slightly in backing of (e) to [A] before /1/, but the Jock girls showed a slight lead in

backing in the same environment. In other environments, the Burnout girls highly exceeded the

backing of the Jock girls, which is a strong indication that this variant is related to Burnout

identity. On the other hand, Jocks led in lowering to [ne] more than the Burnouts led in backing.

Girls led in the use of this variant, which meant that Burnout girls also showed use of the variant

[ne], though their use is less than that of the Jock girls. One can, thus, observe the "powerful

social symbolic role for variation" (1991: 227) in the existing correlation between social groups,

their sex, and the linguistic variants.

1.2.3 Summary of the above Sociolinguistic Methods

In the last three reviewed studies, the sample might have been smaller than those presented

in survey and individual-interview studies. Nonetheless, this does not necessarily jeopardize the

validity and reliability of the results because the researchers took upon themselves the task of

becoming part of the community to have a better understanding of the social and linguistic

behavior of those communities and to be able to arrive at carefully investigated conclusions.

However, the three studies differed in their choice of data analysis methods. Milroy (1980) took

a quantitative approach to measure the density and multiplexity of the working-class networks

that she studied and based her conclusions on it. In contrast, Gal (1979) adopted a qualitative

approach in her evaluation of the language shift observed in Oberwart from Hungarian to










German. Eckert (1989, 1991b), on the other hand, combined both approaches, which I view as

essential for a more comprehensive understanding of the interaction of the social and the

linguistic. When the qualitative analysis and the quantitative analysis complement each other, we

have better confirmation of the results and stronger observations to report.

One observation of all the reviewed studies and other similar studies is that they only focus

on the social aspect of variation and the correlation of linguistic variants with social factors.

They do not take into account how the human mind works and the grammars that are

interchanged and exchanged as those social factors exert pressure and play a role in initiating,

implementing, or advocating a change. Language change without doubt involves the social but

one has to look behind the scenes to observe the grammatical changes that take place as a result

of the social as well as the internal/linguistic aspect of language change. Mufwene (2005), for

example, argues that both the internal and the external are involved in language change. For him,

the interaction of "linguistic ecology" and "social ecology" should reveal the causes of linguistic

change and its spread. Mufwene compares linguistic species to viral species whose change

occurs as a result of the social practices of their host. In this sense, internal selection of one

variant over another is related to the speaker' s relationship with his/her social environment, the

speaker' s personality, and the social group that the speaker would like to associate and fit in

with. This implies that internal change is affected by external or social factors. At the same time,

internal change is constrained by the relationship of one variant with other variants with which it

coexi sts.

The studies presented so far are representative of the sequential methodological

developments in sociolinguistics. As we have seen, each method of data collection and analysis

has its advantages and disadvantages. While surveys enable us to gather a large sample in a short










period of time and in a cost effective manner, the data transcription might suffer from

unconscious bias. Interviews, on the other hand, allow us to gather a sizable sample, but it may

suffer from superficiality unless one is an in-group participant or becomes one. While

quantitative analyses require a large sample, qualitative analyses are less demanding in this

respect but may suffer from generalizability. However, qualitative analyses are recommended in

cases where one needs to understand the cultural and social relations and processes in a certain

community. Thus, the quantitative and the qualitative could complement each other and provide

a comprehensive understanding of language variation and change. The choice among these

methods of data collection and analyses depends on the purpose of the study and what the

researcher is trying to establish and account for. In this study, my intention is to depict how the

human mind deals with social and linguistic constraints simultaneously. I would like to show the

interaction of both types of constraints within one theoretical framework and observe the degree

of influence that each one of them exerts on the degree of language variation.

Many researchers have stressed the role of both external and internal factors in language

change (e.g., Labov 1972a, 1994, 2001; Bell 1984; Eckert 1991a; Finegan and Biber 1994:316;

Mufwene 2005) and the importance of taking both of these components into consideration when

choosing a method of data analysis. For example, Eckert indicates that

Any theory of language must account for this association, and any theory of linguistic
change must account for the social meanings involved in the patterns of variation that
constitute change. (1991a:xii)

She adds stressing her previous conviction:

The study of sound change, then, needs not only a theory of linguistic constraints, but a
social theory that deals with the limits of the symbolic function of linguistic variation.
Indeed, what we think of as internal (linguistic) and external (social) constraints in
linguistic change may not be as separable as conventional practice suggests. (1991a:xiii)









Eckert' s strong convictions show the importance of combing the social and the linguistic in one

comprehensive theory. While some phonologists rej ect the idea of incorporating external factors

into phonological theory (e.g., Anttila 2002), others encourage future research that incorporates

sociolinguistic or external factors in phonological theory (e.g., Gess 2003, Holt 2003). Reynolds

(1994) also advocates the idea that sociolinguistics should make use of phonological theory and

vice versa. Recently, many linguists call for unifying sociolinguistics with other linguistic

theories; they would like to see more research focusing on the intersection of various linguistic

subfields with sociolinguistics. For example, Hume and Nagy (2008) call for such unification

with phonological theory; Johnson and Niedzielski (2008) call for unification with phonetics as

well as working at the phonetic-phonology interface; and Gundel and Sankoff (2008) call for

unification with pragmatics. However, in order to combine both the social and the phonological,

it is important to find a model that can integrate both of them and can reveal the inner workings

of the various constraints: linguistic and social. The choice of an approach to data analysis

should be geared towards a method that instantiates the real, gradual and natural life learning

processes, a method that could integrate the social and the linguistic.

For this reason, I adopt a formal model, OT and the GLA, to account for sociolinguistic

variation and change to delve into the inner works of our linguistic system in pursuit of a better

understanding of the interaction between the social and the linguistic. I believe that a linguistic

theory should be comprehensive enough to be able to account for the phonological processes

accompanying language variation and change, which are influenced by social factors. This

prompts us to review the working mechanisms of OT and the GLA as well as to review the

developments that have been made towards accounting for sociolinguistic variation in

phonological theory (Reynolds 1994; Nagy and Reynolds 1996 applying FCs; Cutillas-Espinosa









2004 applying GLA). The advantages of GLA over other accounts or formal models will be

made clear within the course of this development.

1.3 Introduction to Optimality Theory (OT)

OT (Prince and Smolensky 2004 [1993]) is a non-derivation model that depends on the

application of markedness and faithfulness constraints to an input. The ranking of these

constraints determines the most optimal output among a set of generated output candidates.

Constraints are universal, but the ranking of these constraints is language specific and determines

the grammar of that language or dialect. With the emergence of this non-derivational model, OT,

researchers began to point out its many advantages over rule-based models (e.g., Bermutdez-

Otero and Hogg 2003). First, it avoids the conspiracy effect that is attributed to rules (Kisseberth

1970) that have no output goal (Kager 1999): a rule determines the structural change based on a

structural condition. On the other hand, OT exhibits unification of interaction among constraints

in one parallel step, aiming at a particular output goal. Second, in rule-based models,

intermediate levels take place in the derivation. Rules are based on serial/linear ordering; each

step results in an output that becomes the input for the next rule and so on. Thus, there are many

mappings from various inputs to various outputs. These intermediate levels do not exist in OT;

there is a single mapping from input to output, which is more economical. The role of the

interaction of markedness and faithfulness constraints is to shape the output in a manner

satisfactory to Universal Grammar (UG) and to the input. A further advantage of OT over

derivational models, in which the underlying form should be as underspecified as possible

because of the principle of lexical minimality, is that the input can be fully specified without

recourse to lexical minimality or underspecification (e.g., McMahon 2000; Bermutdez-Otero and

Hogg 2003; Holt 2003). Lexical minimality and underspecification "presuppose an overpowerful

learner" (Bermutdez-Otero and Hogg 2003:93). Thus, in a rule-base model, a learner has to figure









out the rules and work his/her way through to arrive at the fully specified and correct output. In

OT, what is important is the relationship between the input and the output and how harmonious

is the output with the input. The learner hypothesizes the input and tries to adjust the ranking of

his/her constraints to match the output with the input. In this sense, it does not matter how

specified or underspecified for features an input is. For example, in the case of a child's

acquisition of L1, the surface form of the adult is fully specified. It is this surface form that is

considered the L1 input, and from this input, the child develops an underlying form that mirrors

the adult surface form (Smolensky 1996).

One example of the OT superiority over a rule-based model was explored by Lombardi

(2003), who investigated the acquisition of the English interdental fricatives, which are marked

and uncommon sounds, by speakers of Thai and Russian who used [t] instead of [6] and German


and Japanese who used [s] in place of [6]. Lombardi pointed out that in the absence of the [6]

sound from the inventory of those speakers, it is impossible for them to apply or come up with

rules to account for the replacement of [6] with [t] or [s], putting OT at an advantage with its


dependence on constraints and positive evidence. In OT, the change of [6] into [t] or [s] is

ascribed to the interaction of faithfulness and markedness constraints, and "some L1 phonology

has forced reranking, making this an effect of L1 transfer" (2003:225).

However, traditional OT (Prince and Smolensky 2004 [1993]) ran into problems when

accounting for opacity, absolute ungrammaticality, free variation, positional faithfulness, and

allomorphy vs. underlying representations. Some of these problems, such as free variation and

absolute ungrammaticality, are due to its strict ranking principle. For these reasons, researchers

started proposing various improvements to OT, which could provide better accounts of

learnability and variation. One approach that has been gaining momentum in the last few years is










the Maximal Gradual Learning Algorithm (GLA). Since the GLA must be coupled with OT

(Section 1.3.2), it gains the same advantages that OT has over rules, and since the GLA can

account for free variation without the addition of rules, loss of rules or inversion of rules, as were

necessary in traditional rule-based theory, it certainly excels in this aspect over rules. This is not

to mention the idea of Lexicon optimization (e.g., Stampe 1972 cited in Kager 1999; Prince and

Smolensky 2004 [1993]) and restructuring (e.g., Boersma 1997a,b, 1998, 2000; Boersma and

Hayes 2001) of the input based on positive evidence. The GLA not only can learn and account

for categorical outputs but also can account for free variation, optionality, and gradient well-

formedneSS2 (Boersma and Hayes 2001).

1.3.1 Constraint Demotion Algorithm (CDA), Floating Constraints (FCs), and Stratified
Grammars (SG)

Before exploring in detail the working mechanism of the GLA and its applications and

advantages over other approaches, I will briefly review in the next three sections the operating

mechanism of the following approaches that were developed to account for learnability or

variation within the framework of OT: Error-driven Constraint Demotion Algorithm (ECDA) or

(CDA) (Tesar 1995; Tesar and Smolensky 1993, 1996, 1995,1998, 2000), Floating Constraints

(FCs) (Reynolds 1994; Nagy and Reynolds 1996, 1997; Morris 1998), and Stratified Grammars

(SG) or partially-ordered grammar (Anttila 1997a,b, 2002; Anttila and Cho 1998).






2 Gradient or intermediate well-formedness is used to describe forms that occur in speech, which cannot be judged
as totally ill-formed or totally well-formed. These forms are neither perfect nor impossible; hence, judging them is
intermediate or gradient. These forms are different from speech errors that may occur rarely in speech. They may be
rare but they occur more frequently than speech errors and may have an effect on the learning process of a certain
form in addition to the well-formed forms because they can also be considered positive evidence. In other words, the
GLA can learn intermediate well-formedness judgments based on the frequencies of those forms, because those
judgments are "the result of frequency effects in the learning data" (Boersma and Hayes 2001:76). In this sense, the
GLA learns those judgments by learning those frequencies.









1.3.1.1 Constraint Demotion Algorithm (CDA)

Tesar and Smolensky (1993, 1995, 1996, 1998, 2000) proposed the CDA to account for

learnability. In the initial state of acquisition in the CDA, ranking is random, and the process of

language acquisition of rankings follows incrementally from positive evidence. However, one

observable fact in real life is that the task of acquisition may not be complete and some degree of

randomness is carried out in adult grammars. This creates one problem for the CDA. Hayes

(2000) pointed out that the CDA is very successful in accounting for language learning but not

free variation. The CDA's main principle of acquisition is constraint demotion, never promotion:

demote the constraints violated by the intended winner below the highest constraint violated by

the intended loser. On the next evaluation, the CDA makes sure that the learner will learn the

correct form. Thus, the learning process is drastic, although in the real world a child does not

change his/her constraint ranking unless s/he has heard sufficient amount of data to do so. For

example, hearing [dl] that is an illegal onset in English once or twice does not suffice to change

the child's constraint ranking. In addition, the CDA cannot account for free variation because a

demotion of one constraint to generate a variant may lead to the generation of another unattested

output that will lead to another demotion and so on. The constant constraint demotion where one

demotion may lead to the demotion of other constraints causes irreparable damage to the

grammar or leads to what is referred to as entering into an eternal loop (Kager 1999).

1.3.1.2 Floating Constraints (FCs)

To account for variation, a second approach, FCs (Reynolds 1994; Nagy and Reynolds

1996, 1997; Morris 1998), was developed. It holds that one constraint can be ranked freely with

respect to all the other constraints. Nagy and Reynolds (1996: 151) used FCs to account for "the

rates of inter- and intra-speaker variation" in "Faeter, a branch of Francoprovengal spoken in two









farming villages in southern Italy", taking into account the influence of social factors (age,

speaker, and sex) on the occurrence of a certain form more frequently than others. In Faeter,

variation results from deleting a syllable or more from the word. To elicit a controlled number of

tokens, they recorded natural speech by showing a picture book to 40 speakers. They first asked

the speakers to name the obj ects that are at the bottom of the page to see how they pronounced

them in isolation. Then, they asked them to give a description of the picture to see how they used

the same word in context. They considered the following phonological variants: "(1) full form

surfaced, (2) final schwa didn't surface, (3) final syllable didn't surface, (4) final syllable plus

preceding schwa didn't surface, and (5) more than schwa + onset + schwa didn't surface"

(1996: 153). The problem with their use of FCs is that you could have in their case twelve

possible rankings, where you only have two to four optimal outputs. Thus, the model may predict

some unattested outputs to be optimal when they are not. Though they showed that the number

of rankings that give a particular output corresponds with the observed number of its occurrence

in real speech, their data showed exceptions to this observation. For example, /kut.'te.j a/ 'knife'


is predicted to surface more as [kut.'te], but in reality [kut.'te.je] is used more. They attributed

this to analogy to similar forms such as /ka'di:j a/ 'spoon'. A further explanation offered by them


is that each of the surface forms, [kut.'te.j e] and [kut.'tei] (this is the Italian cognate), has its own

underlying form, /kut.'te.j e/ and /kut'tei/ respectively, and surfaces one hundred percent of the

time. Thus, some speakers, particularly young females, use the Italian cognates (the two syllable

forms) more than the Faeter forms (the trisyllabic forms). They found that social factors exhibit

influence on "the relative likelihood of the various rankings possible for a Floating constraint"

(1996: 158). For example, the FC might be closer to the high end of a set of constraints in the









older age group while closer to the lower end of a set of constraints in the younger age group.

They themselves emphasized that much work has to be done to give weight to the various social

factors in order to "restrict where they float" (1996: 158). Thus, FCs could not give a complete

account of their Faeter data.

1.3.1.3 Stratified Grammar (SG)

FCs is one instance of a model that accommodates multiple grammars. Another instance of

such a model is the third approach I discuss here, SG or partially-ordered grammar (Anttila

1997a,b, 2002; Anttila and Cho 1998). This approach is based on the traditional view of

Weinreich, Labov, and Herzog (1968) and Kroch (1989, 1994) that variation is the result of

multiple grammars in the individual or society, and that there is the possibility of the existence of

an intermediary dialect between the two dialects (Anttila and Cho 1998). According to this

approach, the grammar consists of sets of constraints; each set constitutes a stratum and each

stratum has a fixed ranking with respect to the other strata, but the constraints are freely ranked

with respect to each other within the same stratum. If a stratum contains 5 constraints, then there

are 120 possible ranking, calculated using factorial typology as follows: 5 x 4 x 3 x 2 x 1 = 120.

According to this approach, the speaker has a pool of grammars from which s/he selects a

grammar and the role of social factors is limited to the choice among grammars. In this sense,

Anttila adopts the modular approach to grammar, which views internal factors as separate

entities from external factors. This is evident in his remark "it is not the business of grammatical

theory to explain the effects of sex, age, style, register and social class" (2002:212). For Anttila

(2002:231), SG constitutes of one grammar rather than multiple grammars. His conclusion is

based on his adoption of Liberman' s (1994) definition of grammar "as a set of input/output pairs

where for every input there is some fixed output" (Anttila 2002:220) and on his regard of SG "as

a relation, i.e. as a set of ordered pairs of constraints" (2002:231). While Anttila' s (1997a,b)









strata model may account for free variation and predict accurately frequencies of outputs, it will

face problems if there is a large disparity among output frequencies, as Boersma and Hayes

(2001) showed in their reanalysis of Anttila' s (1997a,b) Finnish genitives. The strata model

assumes fixed rankings among strata and free ranking within a stratum. If the frequency disparity

is 99 to 1, then within one stratum, 99 constraints should favor one output and only one favors

the other. If the stratum has five constraints, then "only one of the 120 possible total rankings

gives rise to the rare outcome" (Boersma and Hayes 2001:72).

1.3.2 Working Mechanism and Advantages of the GLA over CDA, FCs, and SG

Now we direct our attention to the GLA and its advantages over the mentioned approaches.

Some of these advantages are: the GLA "can learn free variation, deal effectively with noisy

learning data, and account for gradient well-formedness judgments" (2001:45). In other words,

the GLA can handle optionalityy"; it is also "robust"; and it is capable of accounting for

"intermediate well-formedness" (2001:46). The GLA must be coupled with OT because it is a

"constraint-ranking algorithm" whose purpose is "learning optimality-theoretic grammars"

(Boersma and Hayes 2001:45). It is based on the frequency of an output, which affects the

ranking values of constraints. The two main components of the GLA are a continuous scale of

ranking strictness and a stochastic grammar, that is "a small noise component" is added at every

evaluation of the candidate set, "so that the grammar can produce variable outputs if some

constraint rankings are close to each other" (2001:45-46). The GLA is error driven, in that it only

alters its ranking if it is faced with a mismatch (i.e., the input conflicts with the output. In other

words, the optimal output generated by the constraint ranking conflicts with the real life or actual

output a learner is learning from). The gradual learning of the GLA involves learning a

categorical ranking at every step, which also makes it able to account for non-variable data. The

"ranking value" of the constraint is in the center of the range, but "the value used at evaluation










time" is called the "selection point" (2001:47). Categorical ranking of constraints results from

non-overlapping of the ranges, as Figure 1-1 depicts:



C1 C2
I I
More strict Less strict

Figure 1-1. Categorical Ranking

However, overlapping of ranges results in free variation or variable ranking, as Figure 1-2

depicts:






More strict Less strict

Figure 1-2. Free Ranking

At evaluation time, "it is possible to choose the selection points from anywhere within the ranges

of the two constraints" (2001:48). If the selection point is chosen at the lower end of Cl and the

upper end of C2, the common result is C2>>C 1 in an overlapping situation, as Figure 1-3 shows:



C1 C2 1 C2



More strict Less strict

Figure 1-3. C2> C1

If the selection point is chosen at the upper end of Cl and the lower end of C2, the common

result is C1>>C2 in an overlapping situation, as Figure 1-4 shows:












11 ~2 ~2



More strict Less strict

Figure 1-4. C1 >> C2

Thus, each of the last two rankings may yield variable outputs. This process happens at one

single evaluation time. Boersma and Hayes (2001) use "probability distribution" (Boersma 1997,

1998; Hayes and MacEachern 1998) to depict each constraint range and to make predictions on

the relative frequencies of each output. All constraints are treated equally; they all have the same

width and standard deviation (i.e., 2.0). This value itself is the "evaluation noise" value and the

behavior of a constraint relies on "its ranking values alone" (2001:49). The width and standard

deviation are the same in all constraints because overlapping among constraints should not occur

unless their ranking values are close to each other. If one constraint is wider or has a higher

standard deviation, even when the ranking values of constraints are far apart, overlapping may

occur, and thus variation, which may not always be the case. It is this equal treatment of

constraints and ability to predict the ranking values of constraints that make the GLA more

restrictive and reliable than other less restrictive models such as strictness bands that have their

own width with parts of them designated as "fringes" (Hayes and MacEachern' s 1998) and FCs

(Reynolds 1994) that do not have control over where a constraint could fall or be ranked

(2001:50 & 73). Boersma and Hayes (2001) argued that the GLA predicts that FCs are

impossible because this would require a very large standard deviation for one of the constraints

with respect to the others. FCs are impossible because all constraints should be treated equally

regarding their standard deviation.









Boersma and Hayes (2001) explained the process of the GLA in 4 steps. In the "initial

state" of the GLA, each constraint is given a value specified by the linguist. The value could be

the same for all constraints or different; the grammar that results after learning nonetheless will

be the same but the amount of input data and computation necessary to learn it differ. The first

step is to present the GLA with a learning datum (i.e., the input) that is the surface form of the

adult, not the learner. The GLA was first used for L1 acquisition. Thus, the child has access only

to the adult surface forms from which s/he learns his/her Ll. From those surface forms, the child

learns the L1 grammar and develops an underlying form that mirrors the adult' s surface forms.

However, when using the GLA to model L2 acquisition, in our case a second dialect, the learner

has already developed an underlying form of his L1 in childhood. It is this underlying form that

is used as the input in the GLA and it is this form that undergoes change in the learning process.

This is the core principle of Prince and Smolensky's (2004 [1993]) idea of Lexicon

Optimization. A learner may or may not be able to restructure his/her underlying form from

exposure to a surface form that does not match his/her underlying form.

The second step is to generate outputs by adding a noise value to the current ranking value

of the constraint to obtain the selection point at every evaluation time. The third step constitutes

comparison. If the generated output matches the learning datum, nothing further is done. If a

mismatch between the learning datum and the generated output occurs, the grammar must be

adjusted; the algorithm takes a learning step and thus alters the grammar. The next step is "mark

cancellation" if the same constraint is violated by both candidates. The fourth step constitutes

"moderate adjustments of the ranking values" (2001:52) to obtain the right result. Since there is

no certainty on whether the constraints with uncancelled marks for the learning datum are too

high or those for the learner' s output are too low, both promotion and demotion take place. The









adjustment of the ranking values of constraints involves increasing or decreasing the constraint' s

value by the amount of the plasticity value. Pla~sticity refers to "the numerical quantity by which

the algorithm adjusts the constraints' ranking values at any given time" (2001:52). In the "final

state", the algorithm arrives at the appropriate grammar that generates the adult's surface form

after cycles of repeating steps 1-4.

Boersma (1998) and Boersma and Hayes (2001) realized the incorrectness of the

predictions of the earlier version of the GLA that was referred to as the Minimal Gradual

Learning Algorithm (MGLA), in which only one constraint, the highest violated constraint by

the intended winner, is demoted and only one constraint, the highest violated constraint by the

intended loser, is promoted. In the GLA, all constraints violated by the correct form are demoted

and all the constraints violated by the incorrect form are promoted. Although Boersma and

Hayes (2001) do not deny that other methods of evaluation, such as the CDA may work well for

non-variable data, they point out that the GLA has many advantages over other models. To test

the GLA and to show that it has advantages over the CDA, they provide empirical evidence by

reanalyzing Hayes and Abad' s (1989) data of metathesis in Ilokano and Hayes and

MacEachern' s (1998) data of English light and dark /1/. What makes the GLA advantageous over

the drastic learning process of the CDA is that it does its j ob gradually, continuously generating

more correct outputs without causing any damage to the grammar.

To test the GLA against the SG approach to free variation, Boersma and Hayes (2001)

reanalyzed Anttila' s (1997a,b) data of Finnish genitives. They found that SG runs into problems

when faced with large disparity among output frequencies. This makes the GLA more

advantageous than SG. In addition, Antilla thinks that Boersma and Hayes's (2001) model of

continuous ranking and stochastic constraints may have advantages over other models because of









its use of the GLA that enables us to arrive at the right ranking from the surface forms and the

corresponding underlying form. While the multiple grammars model and its derivatives (e.g.,

SG) rely on frequencies and grammar-counting, the continuous ranking approach moves OT in

the direction of quantification.

1.3.2.1 Pure linguistic applications of the GLA

Boersma and Hayes' s (2001) first test of the GLA was Hayes and Abad' s (1989) Ilokano

metathesis data: /7w/ optionally becomes [w?], where [w] is not an underlying segment, rather

derived from /o/. The process includes: 1) computing "the factorial typology of the constraint set

over the candidates given" and 2) then add gradually "pairwise constraint ranlangs, recomputing

the factorial typology as constrained a priori by these rankings, until the output set had shrunk to

include only the attested cases of free variation" (2001:61). The CDA could not converge on the

same data, concluding that it cannot account for free variation because, as mentioned above, this

leads to constant demotions of constraints and damaging of the grammar. In contrast, The GLA

was able to converge, did its job gradually without damaging the grammar, and continued to

generate correct outputs.

Another advantage of the GLA is that it can account for intermediate well-formedness

without recourse to different width of ranges or to fringes (Hayes 2000). Hayes and

MacEachern's (1998) model of strictness bands and fringes is "less restrictive" because "it

permits individual constraints to be affiliated with "bands", each with its own width, that specify

the range of selection points". The model also "permits parts of each band to be designated

'fringes'," which lead to intermediate well-formedness if a selection point falls within them"

(Boersma and Hayes 2001:73). Being a less powerful theory and being able to account for the

same data, the GLA is advancement to other approaches as it generates its own learning










algorithm. To account for intermediate well-formedness, Boersma and Hayes 1) "use one

equation to convert gathered gradient well-formedness judgments into conj ectured frequencies;

2) then, they feed these conj ectured frequencies into the GLA, "which will produce (if all goes

well) a grammar that closely mimics them" (2001:76) (i.e., predicted frequencies); 3) "by

feeding the predicted frequencies into the mathematical inverse of the first equation", they get

predicted judgments; and 4) finally, make comparison between the observed judgments and the

predicted judgments. These comparisons show that the grammar learned by the GLA is very

close in form to Hayes's (2000) fringes.

1.3.2.2 Example of a sociolinguistic application of GLA

After discussing the advantages of the GLA over other models, it would be of interest to

review the only application of it, to the best of my knowledge, to sociolinguistic variation and

change (Cutillas-Espinosa 2004). Like Boersma and Hayes (2001), Cutillas-Espinosa

emphasized two main points: variation is gradual and continuous (2004:171), designating a

special weight to personal choice in variation. His approach involves: (i) a continuous ranking

approach to constraints (Boersma 1997, 2000; Hayes 2000; Boersma and Hayes 2001) and (ii) a

limited number of available grammars, (2004: 171) (i.e. his model consists of three grammars to

account for sociolinguistic variation). The first grammar G1 is the standard one that is associated

with prestige, education and propriety. The third grammar G3 is the local grammar that is used by

the community. The second grammar G2 is the default grammar or intermediate grammar that

could be closer to either the first or the third grammar due to various social, personal, or

contextual factors; it is the everyday, informal, and 'in use' grammar, characterized as

"dynamic" because it changes according to the speaker's need (2004:172). Cutillas-Espinosa

suggests that the GLA could account for one of the three grammars that every person has (i.e. the

second/default grammar). In this sense, G2 is assumed to be a continuous ranking of constraints,









based on conscious choice of the speaker of the appropriate ranking for the appropriate output in

a particular context. Cutillas-Espinosa assumes that a speaker who is exposed to two different

grammars tends to build his/her own constraint ranking to accommodate the social context, to

build an identity, or to proj ect a self-image. In this sense, grammar is no longer observed as a

mechanism in which personal decisions are not taken into consideration. Rather, a speaker can

control the ranking values of his/her grammar consciously. To quote him "Grammar is no longer

seen as a fully automated mechanism; personal and meaningfud decisions are granted a place"

(2004:175).

Cutillas-Espinosa applied the GLA model to both Labov' s (1966) study of the English (r)

that can be realized as [.( or 0Z in the speech of New York City and to the Castilian Spanish (s) in

Murcia, which can be realized as either [s] or 0Z (Hernandez-Campoy and Trudgill, 2002). He

examined the case of Susan Salto from Labov' s (1966) study, who showed great stylistic shifts

for the variable (r). She used [.L 100% of the time in her most careful style, minimal pairs, and

suppressed it 98% of the time in casual speech. Starting with the fixed ranking value 80 for the

markedness constraint *CODA/r, the results showed that MAX would have the ranking value

91.31 in minimal pairs, ensuring that MAX will be higher ranked than *CODA/r, whereas it will

have the ranking value 68.69 in casual speech, guaranteeing a higher ranking of *CODA/r.

Cutillas-Espinosa also examined the case of a radio presenter at a local station in Murcia (Spain)

from Cutillas-Espinosa and Hernandez-Campoy's (2004) study of the differences in (s) deletion

in the speech of the presenter in two different settings: 1) in broadcasting, talking to non-standard

speaking audience on the phone, whose speech is characterized with suppression of (s), and 2) in

a formal interview with the researchers (Cutillas-Espinosa 2004: 176). Cutillas-Espinosa and

Hernandez-Campoy (2004) had found that the presenter used the [s], the standard variant, 89% in









broadcasting and only 1% in the interview. Cutillas-Espinosa conducted the same analysis done

on the previous study and found that the ranking value of 1VAX correlated with the style chosen

by the presenter: 83.46 in broadcasting and 73.4 in interview. Thus, he assumed fixed ranking for

markedness constraints and that faithfulness constraints are the ones that change ranking value

and move around to arrive at the desired output. For him, the continuous ranking scale overrides

all other models (FCs, CDA, and SG), in that it can feature a continuum of grammar where style

can shift gradually, rather than abruptly. His approach is more of the antimodular type because it

takes both internal and external factors into consideration in opposition to Anttila' s (1997a,b,

2002) modular approach that only takes internal factors into consideration.

Cutillas-Espinosa' s model is insightful and it opens a field of sociolinguistic investigation

within an Optimality-Theoretic approach, embellished with the numerical and functional

workings of the GLA. In this study, I will elaborate on this insightful model by taking into

account that the human mind may not be limited to three particular grammars and the observed

intra-speaker oscillation and inter-speaker differences in adopting or acquiring a new grammar. I

also introduce social factors as constraints to the GLA and interweave them with linguistic

constraints to show their immediate effect on the ranking values of linguistic constraints and the

choice of one grammar over another at a certain time and place. One further advancement is

showing which social constraints are activated or not, and to what extent, by certain speakers or

groups of speakers at a certain time and place. In addition, I would like to stress the differences

that may be observed among individual grammars within the same community or the same group

of speakers. In this study, I am also using my own original data that is based on naturally

occurring speech to see if all speakers behave in the same way regarding choices made among

grammars, and if the pattern is uniform throughout the community. A larger sample of speakers









and a comparison among the different grammars used by the speakers may give us a better

understanding of these differences. Furthermore, G1 and G3 may not be stable as well; they may

be variable and might be presented on a continuous ranking scale. In Arabic, for example, the

standard language itself is variable: not all speakers use the same level of Case marking or their

degree of reverting to the vernacular may vary within the same formal setting.

1.3.2.3 Concluding remarks about the GLA

The discussion above reveals many advantages of the GLA over other models. The last

example shows the possibility of applying it to sociolinguistic variation and change and naturally

occurring speech. It remains to be seen if one can include social factors in the computational

process of the GLA as ranked constraints. In addition, the non-teleological approach (Boersma

1997b) to the GLA may need some revision, in my opinion, because if selection and choice play

a role in determining the ranking value of constraints, then there might be some teleological

effects in the adults' acquisition of a second grammar. The non-teleological approach to the GLA

might work well in its application to L1 acquisition, not to dialectal and sociolinguistic variation

where social factors can be involved and play a selective role. Thus, social factors may influence

the ranking values and rankings of linguistic constraints. In applying the GLA to sociolinguistic

variation, one can predict the ranking values of constraints and depict the intrapersonal and

interpersonal oscillation that one may observe in the speech of those who try to change the way

they speak in attempt to sound city-like, more prestigious, or probably more or less formal. The

stochastic feature of constraints and the continuous feature of ranking should be applicable to

social factors because sociolinguistic variation is gradual and stochastic in nature. Not all

speakers behave in the same way or have the same level of sociolinguistic competence. Even the

same speaker may vary his/her speech along a continuous scale of values. Sometimes, there is a

kind of internal conflict between grammars, leading to confusion regarding the choice of one









grammar over another or a kind of stalemate situation where the speaker/learner is in an

intermediate position from which s/he cannot escape.

In the same way that social factors should be included in formal theory, sociolinguistics

should take markedness into account, because markedness, in addition to social factors, does

sometimes play a maj or role in language variation and change (Reynolds 1994). This markedness

could have played a major role historically in producing the SA [6] and [a] as [t] as [d]


respectively. The merger of [6] and [a] with [t] and [d] respectively was completed around the

14th century (Daher, 1998a, 1999). This is comparable to the use of Thai, Russian, German, and

Japanese learners of English of [t] and [s] instead of [6] because of the absence of this sound

from their inventory, which is attributed to the interaction of faithfulness and markedness

constraints (Lombardi 2003).

1.4 Conclusion

A number of studies have employed OT in one way or another to account for language

variation from various perspectives. While most researchers agree that sound systems change

continuously (Boersma 1997b; Labov 1994; Ohala 1993), even if the change only results from

"internal factors" (Boersma 1997b: 1), they may disagree on the reason behind sound change.

Sound change does not have to be teleological (Boersma 1997b; Ohala 1993), facilitative or

based on meaning preservation (Labov 1994). It could be based on historical factors (Adam

2002) or perceptual and lexicon optimization reasons (Holts 1997). However, Boersma (2000)

argues that optimization is not the main "internal factor" (Boersma 2000: 1) for sound change.

Thus, most of the language variation studies that implement OT do not deal with sociolinguistic

variation and do not refer to the social interface with the linguistic and historical factors that

commonly involve internal, perceptual, or lexicon optimization factors. Unlike other studies, this










study deals with variation from a social perspective where social factors play a maj or role in the

change. They are integrated into the mental computing system of the speaker in the form of

constraints that intersect with linguistic constraints to account for real life occurrences of

particular variants. The study reflects the intra- and inter-speaker preference for certain sounds

according to the interlocutor, although the speaker is capable of pronouncing both sounds of both

dialects perfectly. The situation is similar to knowing two languages and switching between the

two according to the interlocutor. The difference is that the switch is not done for the sake of

comprehension on the part of the interlocutor but for the sake of competing with the interlocutor

for prestige. This is the case that we will observe concerning the two variants [q] and [7].

It would thus be of interest to see how one can model this sociolinguistic variation within

the framework of OT and the GLA and how one can incorporate social constraints into the

human mind computing system. This may give us an answer to how speakers oscillate between

two different dialects and how the degree of social success that those speakers achieve depends

on the level of acquisition of the new dialect. It may also explain the intra- and inter-speaker

variation observed among speakers who use more than one variety, in an attempt to

accommodate the speech of the interlocutor. Thus, at the core of this study is the development of

a model for variation and change that could have cross-linguistic implementations. The model

will present how the intersection of linguistic constraints with social constraints can model and

predict real life variable use of variants as well as predict the direction of change when

constraints are given specific values. This model will also highlight the importance of social

constraints in determining the ranking values of other constraints as well as the chosen grammar

at a certain time and place. The model will show that variation within the speech of a community

could be driven by various factors at the same time. While one variation may be led by social










factors, another variation could have been led by markedness diachronically and be maintained

by faithfulness constraints synchronically. I will also show that this model can account for these

different types of variation within the same dialect. It can deal with these different variations

independently from each other since these variations are the result of different factors and act

independently from each other. This discussion provokes a number of research questions that

will be stated in Section 1.5.

1.5 Research Questions

The research questions that guide this study are:

1. How can Optimality Theory account for speakers' switches between two different dialects

and their intrapersonal and interpersonal variation in the use of particular sounds (e.g., the

use of [7] in place of [q]) based on their interlocutor' s background?

2. How can social factors be incorporated into the GLA?

3. To what degree do social factors play a role in building our grammars, or rather developing

new grammars?

4. How consistent is the pattern of variation and sound change within and among rural

migrant speakers to the city of Hims?

5. Which factors are stronger in leading a sound change: markedness constraints or social

constraints?

1.6 Structure of the Study

The study is organized as follows. Chapter 2 presents the methodology used in data

collection and analysis. It includes a brief description of the community under investigation. It

introduces the city of Hims, the speech sample, the linguistic and social variables, and a brief









description of data analysis, including the statistical models and other quantitative methods

employed.

Chapter 3 introduces the theoretical analysis of the study. In this chapter, the new model

proposed in this study will be explored in detail. The chapter begins with descriptions of the

Himsi and rural grammars and the change that the rural migrant speaker grammar undergoes in

his/her attempt to appear prestigious. Then, social constraints and the stochastic process that is

followed in examining the simultaneous effect of social and linguistic constraints on the choice

of a grammar at a certain time and place are introduced. This chapter shows that social

constraints should constitute an essential part of our grammar. The inclusion of social constraints

in OT and the GLA yields results that match real life occurrences. These social constraints play a

maj or role in determining the grammar chosen by certain speakers from a certain age group,

gender, residential area, and social class, particularly in the use of socially conditioned variants,

such as [q] and [7]. Including social constraints in the computation explains the variability in the


use of [q] and [7] among speakers from the same social group or different social groups. Chapter

3 also shows that certain social constraints give us expectations on what the speech of a certain

speaker or a group of speakers should sound like. The same model deals with other types of

phenomena independently from the sociolinguistic variation between [q] and [7]. It shows that

social constraints do not play a role in the stable use of [t] and [s] and [d] and [z]. The chapter

shows that each of these four sounds has an underlying form that is identical to its surface form.

Faithfulness constraints play a maj or role in maintaining the input in the output.

Chapter 4 presents the quantitative analysis regarding the use of the two sociolinguistic

variants [q] and [7]. In this chapter, I explore the association between social class and education,

income, occupation, and residential area. Income, occupation, and residential area emerge as









significant indicators of social class, whereas education emerges as insignificant. These Eindings

correspond with my predictions of how social class should be determined in Syria and other

Arab countries. I show that there is correlation between the two variants [q] and [7], using a Chi-

squared Bivariate test. I also explore the distribution of the data with respect to each of the

independent variables to determine the type of statistical tests that should be implemented. The

non-normality of the distribution and the overdispersion of the data called for the use of Negative

Binominal regression procedures. In the main effects tests on both [q] and [7], age, gender, and


residential area emerged as significant factors in determining the use of [q] and [7]. However, in


the interaction tests, residential area emerged as insignificant regarding the use of [7]. This result

added one advantage to the model proposed in Chapter 3. Making social factors more specific by

dividing them into a number of social constraints and implementing them in the GLA enables us

to see interaction between one of the constraints of a social factor and other constraints. This

specifieity shows that the whole social factor does not have to be involved in the variation, but

part of it may. Chapter 4 further examines whether frequency plays a role in acquiring certain

words before other words. Frequency emerges as a facilitative factor in the acquisition of the

prestigious form. Highly frequent words appear with [7] in the speech of varying speakers more

than less frequent words.

Chapter 5 presents a quantitative analysis of the use of [t] and [s] and [d] and [z]. Because

this stable phenomenon is not socially conditioned, frequency effects are explored against two

diachronic changes that led to the replacement of the SA [6] and [a] with [t] and [d] first and

later with [s] and [z] respectively. Today, the use of the four variants is stable due to two

opposing frequency effects (Bybee 2001). The first frequency effect led to the first diachronic









change and the merger with [t] and [d]. The second frequency effect made highly frequent words

resistant to the new creeping change: the use of [s] and [z] in place of [6] and [a] in borrowed

words from SA. That is why we observe the use of [t] and [d] in highly frequent words in

comparison to the use of [s] and [z] in less frequent words.

Chapter 6 concludes this work. It gives a brief answer to each of the five research

questions that led this research. It also highlights other findings of the study such as the role

frequency plays in both the use of [q] and [7] and [t] and [d] and [s] and [z]. It also lists the

advantages of and caveats about the new model. It gives suggestions and recommendations for

future research, showing the possibilities of expanding on the social constraints proposed in this

study. In general, the conclusion calls for exploring the possibility that social constraints could

be universal in the same way that linguistic constraints are.









CHAPTER 2
IVETHODOLOGY

2.1 Setting: The City of Hims

The city we are concerned with in this study is Hims or Homs, as some people refer to it as

a result of rounding the first vowel that is also a feature of the Himsi dialect. Hims is the third

most important city in Syria and it is strategically located in the fertile Orontes River Valley in

the central western part of Syria (Figure 2-1) on a hill approximately 450 meters above sea level.

It is a crossroad between the capital Damascus 160 km to the south and the second maj or city

Aleppo 190 km to the north and between coastal cities to the west, such as Tartus and Latakia

and the eastern cities of Syria, such as Palmyra (in Arabic Tadmur) and Dayr az Zawr. Hims,

known in Roman times as Emesa, is an ancient city dating back to the year 2300 B.C. (The

Colombia Encyclopaedia 2007). The population of the City of Hims, according to 2002 estimates

provided by the Homs City Council (2008), is 1,033,000. The city of Hims is the capital of the

Hims Governorate, which is the largest among other Syrian governorates (40,940 square

kilometer) (Syrian Arab Republic Central Bureau of Statistics 2002). The population of the Hims

Governorate according to civil registration records is 1,791,000 (Arab Republic Central Bureau

of Statistics 2004). This number includes both urban and rural areas (i.e., the city of Hims and

the surrounding villages respectively). Excluding those who live outside Syria, the number of

physically present residents of the Hims Governorate is estimated at 1,577,000.

The important strategic location and size of the Hims Governorate distinguish it from other

Syrian governorates. It is the third governorate in industry, trade, and agriculture. The city itself

is the "third largest industrial city in Syria" (Gilford 1978). The centrality of the city of Hims

makes it an attractive center to a large number of rural people from the neighboring countryside.

Those rural migrants find haven in Hims' Al-Baath University, the third maj or public university









in Syria, job market, and shopping and trading centers. From the late sixties and early seventies,

a larger influx of rural migrants to maj or cities in Syria, particularly Hims, started taking place.

Rural people started abandoning agriculture and their lands in their villages. With the

government facilitating education and making it available free to everyone, they sought higher

education to obtain governmental jobs, such as teaching, construction, and industry that includes

the refinery and other maj or phosphate and chemical plants all of which are situated in Hims.

However, Zakaria and Sibai (1989 cited in Mahayni 1990) suggest that migration was not only a

response to industrialization but also to the dire socioeconomic situation in many rural areas of

Syria. Thus, they suggest that the search for better life style and higher level of income motivates

many to move to the city, particularly those with higher levels of education. In addition to the

growth of sectors such as building and construction and social and personal services, the higher

growth in the government sector provided more employment opportunities for the educated from

the rural areas, inducing more migration (Mahayni 1990). Mahayni (1990) traced the population

growth of both urban and rural areas in the Syrian governorates between the years 1960 and 1986

based on statistics from the Syrian Arab Republic Central Bureau of Statistics (1960, 1986). He

found that the population of the city of Hims increased annually by 4.76% from 150,000 in 1960

to 502,000 in 1986, whereas the annual population growth of rural areas increased by 2.43%

from 25 1,000 in 1960 to 469,000 in 1986. The higher growth rate of the population of the city of

Hims is not ascribed to higher birth rate in the city; rather, it is the result of the migration of huge

numbers of rural dwellers to the city of Hims. The proportion of rural migrants who lived in the

city of Hims in 1970 was 25.4% of the total city population according to the Syrian Arab

Republic Central Bureau of Statistics (reported in UNCEWA 1980).














Turkey


77. ? A~r Raqqab j Pro~vince Nlame

yprus ,C------- + nayr at zaw rp Ar aqqah
Tadtus+Ha~ H-imas As suwa~yca'
Himn D Iar'a
T-admurI Dayr Az Zawrr
Wardi A -N era
(Palmyra) I I -Hamah
Leba o
SHasaka (Al Haksa)
)gmaggs g I Hims (Homs)
'0~4r~i Iraq N IdIIlib
... Lattaklia (Al Ladhiq~iyah)
/~ [) unetra
,-- 0 203 40 80 120 160RiDmsh

Jordan [ --- ] As~ia



Figure 2-1. Map of Syria and neighboring countries.









Rural migration to big urban centers, such as Hims, is not only a Syrian phenomenon but

also a phenomenon of many Arab countries that were developing and growing industrially. For

example, Casablanca in Morocco has seen a huge influx of rural migration in the second half of

the twentieth century because of industrialization (Hachimi 2005, 2007). This migration also

resulted in "social, cultural and linguistic changes" (Hachimi 2007: 97). Thus, rural migration to

urban (e.g., Hims) or urbanized (e.g., Casablanca) centers seems to be a wide spread

phenomenon in the Arab world and usually results in linguistic variation because of language

contact among various language varieties.

Hims, like other Syrian cities, is home to a diverse population of Sunnis, Alawites, and

Christians as well as Armenians and Palestinian refugees. Before the Arab conquests after the

birth of Islam in 622 A.D., Hims was mainly populated by Christians. It was taken by Muslims

in 636 A.D. and the large Christian element of the city was eliminated after the rebellion in 855

A.D. (Encyclopedia Britannica 2008). However, Hims continues to have smaller concentrations

of Christians in different residential areas and contains historically significant ancient churches

from early Christianity, such as Umm Al-Zunnar (The Virgin Mary's Girdle) and Mar Elian (St.

Elias). These churches are situated in Al-Hameeddieh, one of the residential areas from which

most of the participants (39 speakers) in this study come. Hims is also surrounded by a collection

of Christian villages called Wadi Al-Nasara 'The Valley of the Christians' from which almost all

of the study participants come (Section 2. 1.2).

The Himsi people are known for their pride in their dialect, which is characterized by the

use of [7]. For this reason, they usually stigmatize other dialects, particularly the ones that

contain the [q] sound such as many nearby rural dialects. The stigmatization of the use of [q]

leads the rural people to adopt the [7] and other speech features of the city people, so they can










integrate into and be accepted as part of the urban community. The use of [7] instead of [q] is

considered more urbane in major Syrian cities, such as Damascus (Daher 1998a, 1998b) and

Aleppo. Thus, many rural migrants start switching to the use of [7] instead of [q] and

consequently switch to other speech features.

2.2 Speech Sample

Thus, the center of investigation is the speech community of the city of Hims to which

many rural speakers have migrated at one point in their life. The study greatly depends on

naturally occurring speech from speakers who are not picked at random. According to Labov

(1966: 43), the effects of the "observer' s paradox" may be partially overcome by obtaining

samples from natural social interactions among in-group members (e.g., interacting with family

members or "peer group"). The speech sample is taken from rural migrant speakers who belong

to the Christian community at large to which I myself belong. They reside in two residential

areas in Hims: Al-Hameeddieh and Akrama (Section 2.3.2 and Figure 2-2).

One important criterion when choosing the participants was being rural migrants to the city

of Hims, representing the first generation, or the sons and daughters of those migrants who were

born in the city of Hims, representing the second generation. In this sense, the study focuses on

two generations or age groups: older and younger. Both male and female speakers were chosen

from different economic and social background. With the exception of Speakers 19 & 5, all other

participants come from the collection of the Christian villages called Wadi Al-Nasara (Figure 2-

1) and mainly from the village of Oyoun Al-Wadi (thereafter Oyoun) 'The Springs of the

Valley'. The reason for choosing the village Oyoun and other neighboring villages is that I am

originally from that village and familiar with the surrounding villages. Being an in-group

member adds more naturalness to the conversations, having the opportunity to interview people









who are relatives, friends, family members, and neighbors. Being an in-group member has been

emphasized by many researchers as the most preferable method to obtain the most naturalistic

speech data (e.g., Labov 1972a; Eckert 1991; Milroy & Milroy 1992). Hence, integrating into the

investigated speech community has been the focus of many ethnographic studies, which required

researchers to spend a long period getting to know the community (e.g., Gal 1978; Eckert 1989;

Milroy 1980). Fortunately, I have had the advantage of being part of the community under

investigation in this study. My own family is from the village of Oyoun, and I moved to Hims at

the age of two years and two months. Most of the speakers moved to the city around the same

time like my parents in the mid seventies, and we have strong social ties with them in one way or

another. At home, my family uses the village dialect, but with distant friends and acquaintances,

there is a switch to the Himsi dialect. Thus, I am acquainted with both dialects. This

phenomenon of switching between two forms or two dialects is apparent in the speech of many

rural migrants who live in the city and occasionally go back to visit the countryside.

A sample of fifty-two interviews comprises the data set (Table 2-1). The data were

collected on two separate field trips to Syria. One trip took place in 2004; the other one was in

2006. In my first trip to Syria, I recorded forty-two speakers of whom I only investigate twenty-

six speakers, those who fall within the age ranges that I am investigating in this study. During

my second trip to Syria, I recorded thirty-six participants of whom I only investigate twenty-six

speakers. Ten of the twenty-six speakers, who were recorded in 2004, were rerecorded in 2006

with the purpose of obtaining better sound quality as I was using a more advanced digital

recorder with better voice reception equipment. Their recordings from 2006 are used in this

study. The other sixteen speakers constitute new participants. In my first trip to Syria, I was able

to record more participants because I went in a time of the year May and June when it was









easy to find participants and record them; people had not yet started their summer vacations.

Most of the people were still in the city because schools and universities were still open;

university and school exams usually finish by the end of June. On my second trip, it was more

difficult to get in touch with participants because July and August are the main vacation months

for people in Syria. Most people commute a great deal more in these two months between the

city and the village as well as take trips to the beach and other excursion trips. The fifty-two

speakers are almost equally distributed between males (24) and females (28); lower middle class

(29) and upper middle class (23); and younger age group (18-35) (24) and older age group (52+)

(28). It is important to have a varied sample to examine the variable use of [q] and [7]; [a], [d],

and [z]; and [6], [t], and [s]. Most of the participants fourty participants come from the

village of Oyoun where [q] is dominant. The other twelve participants come from the

neighboring Christian villages, Wadi Al-Nasara, in which [q] is dominant. In addition, those

twelve participants are either married to someone from Oyoun or are the son or daughter of one

parent from Oyoun, and his/her other parent is from a neighboring village that has the same

speech characteristics of Oyoun. For example, Speaker-1 is from Tin Sbil and married to

Speaker-14 from Oyoun; Speaker-8 is from Juwaikhat and married to Speaker-20 from Oyoun;

Speakers 33, 34, & 43 are the two sons and daughter of Speakers 8 & 20; Speakers 9 & 26 are

from Ain Al-Ajooz and are husband and wife; Speaker-17 is from Habb Nimra; Speaker-21 is

from Treez and married to Speaker-6 from Oyoun; and Speaker-42 is the daughter of a man from

Oyoun. Speaker-19 is from Muharda; Speaker-5 is from Ilbi and married to Speaker-22 from

Oyoun. Muharda and Ilbi are not in Wadi Al-Nasara, but they are also Christian villages

characterized by the use of [q].











Table 2-1. Distribution of study participants
Speaker Sex Age Social Occupation Income Education Area
class


M 77 LM
M 67 LM
M 64 LM
M 60 LM
M 70 LM
M 67 LM
M 64 LM
M 53 LM
M 70 UM


M 69 UM
M 62 UM
M 62 UM
M 64 UM
F 75 LM
F 61 LM
F 61 LM
F 61 LM
F 59 LM
F 56 LM
F 52 LM
F 53 LM
F 67 LM
F 58 LM
F 58 UM
F 57 UM
F 61 UM
F 58 UM
F 57 UM
M 31 LM
M 25 LM
M 35 LM
M 30 LM
M 23 LM
M 19 LM
M 24 UM
M 23 UM
M 24 UM
M 36 UM
M 27 UM
F 35 LM
F 28 LM
F 24 LM
F 18 LM


Retired Gov. employee Low
Retired Gov. employee Low
Retired officer Low
Retired Gov. employee Low
Retired Gov. employee Low
Retired Gov. employee Low
Retired Gov. employee Low
Teacher Mid
Retired Director of Customs High
(also published a historical
book)


Middle
Middle
Middle
High school
Elementary
High school
High school
B.A. Economics
B.A.


Akrama
Akrama
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Akrama
Al-Hameeddieh


Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Akrama
Akrama
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Akrama
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Al-Hameeddieh
Akrama
Akrama
Akrama
Al-Hameeddieh


Unemployed
Civil Engineer
Business man
Teacher
Housewife
Housewife
Teacher
Teacher
Gov. employee
Teacher
Teacher
House wife
House wife
Teacher
Housewife
Housewife
Housewife
Teacher
Housewife
Medical doctor
Civil Engineer
Assistant engineer
Salesman
Medical student
Medical student
Medical doctor
Salesman
Dentist
Medical doctor
Dentist
Gov. employee
T.A. architecture
Agricultural engineer
Student


Mid
High
High
Mid
Low
Low
Low
Low
Low
Mid
Mid
Low
Low
Low
Mid
High
High
Mid
High
Low
Low
Low
Low
Mid
Mid
High
High
Mid
Mid
High
Low
Low
Low
Mid


Middle
B.A.
High school
High school
Elementary
Elemental
A.A.
A.A.
Middle
A.A.
B.A. Fr. Lit
Middle
Elementary
A.A.
Middle
High School
High School
A.A.
Middle
Professional
B.A.
A.A.
A.A.


Medical student Akrama
Medical student Akrama
Professional Al-Hameeddieh
High school Al-Hameeddieh
Professional Al-Hameeddieh
Professional Al-Hameeddieh
Professional Al-Hameeddieh
A.A. Al-Hameeddieh
M.A. Al-Hameeddieh
B.A. Al-Hameeddieh
High school Akrama


SLM and UM refer to lower-middle class and upper-middle class respectively.

4 'Middle', 'High', and 'Elementar in Table 2-1 refer to the type of school they have completed. 'A.A.' refers to
associate degree or some college. 'B.A.' refers to Bachelor's of Art.










Table 2-1. Continued.
Speaker Sex Age Social class Occupation Income Education Area
44 F 29 LM Gov. employee Low A.A. Akrama
45 F 28 UM Agricultural engineer Mid B.A. Al-Hameeddieh
46 F 33 UM Private sector employee High B.A. E. Lit. Al-Hameeddieh
47 F 32 UM House wife High A.A. Al-Hameeddieh
48 F 28 UM Housewife High B.A. Law Al-Hameeddieh
49 F 23 UM Civil engineer High B.A. Al-Hameeddieh
50 F 25 UM English teacher High B.A. E. Lit. Al-Hameeddieh
51 F 21 UM Pharmacy student High B.A. Al-Hameeddieh
52 F 26 UM Food Engineer High B.A. Al-Hameeddieh


Informal conversations in Colloquial Arabic, lasting between 30-45 minutes with each

individual, are audio-recorded, transcribed and analyzed. On my first trip to Syria, I used an

analog tape recorder that was placed close to the participant to capture the whole conversation as

clearly as possible. I, then, digitized the tapes, so I can listen to them on my computer. On my

second trip, I used a good quality digital recorder (marantz Professional Solid State Recorder

PMD660) and a small microphone that is pinned to the blouse or shirt of the participant to have a

better sound quality recording for the sake of phonetic transcription. The recordings took place

either in my family home in Hims or in the informants' homes, whichever was more convenient

at the time. In the interviews, I used the [7] sound with all the interviewees, some of whom were


very open to using their village dialect with me despite my use of [7]; probably, they felt

comfortable using their mother tongue because they know that I come from the same hometown.

Had someone from the city led the interviews, I would think that more variation and thus more

use of [7] would have been observed. As for the other variants, I talked naturally, so they will be


pronounced as I would usually use them in natural speech. It is difficult to specify which variable

I was using because I myself may vary in using the other variants. This is not a problem because

all speakers behave in the same way regarding the four variants, [t] and [s] and [d] and [z]; each

of the four variants is used in specific lexical items or words (Chapter 5). There is barely any









doubt or rather it is predictable which variant will occur in a certain word. No one was informed

of the exact focus of the study. When they asked about the purpose of the study, they were

informed that it deals with the social influence on sound change, but no specific details were

added. However, the participants were instructed that the interviews were intended to be as

natural and as informal as possible, so they should not put any effort into thinking about what

they should or should not say. In addition, they were informed that their identity as well as their

conversation would be highly confidential. All of the conversations were natural and did not

follow any preconceived format. Topics of the conversations varied, allowing the speaker to talk

about any subj ect that appealed to him/her (e.g., telling j okes; funny, sad, or frightening stories;

relating dreams; work, love, marriage, and relationships stories; family issues; and other personal

interests). Thus, the interviewees were free to speak about any topic they wanted. Conversations

mostly started by my asking the participant about his/her family, children and other matters of

mutual interest. After some questions and answers, if the conversation slowed down or there was

not much to say, informants were asked to tell a happy or sad story, a dream, or an experience

that they had experienced. To maintain the naturalness of the conversation, other family

members were allowed to be present during the recording of all the participants. Thus, occasional

intervention from other attendees sometimes heated the conversation and made the speaker more

oblivious of the tape recorder.

It is worth noting that some of the speakers are related to each other. Speakers 3 5, 36, 46,

47, & 48 are respectively the two sons and three daughters of Speakers 12 & 28. Speaker-52 is

the daughter of Speaker-25. Speakers 49, 50, & 51 are sisters. Speakers 45, 37, & 38 are

respectively the daughter and two sons of Speakers 27 & 13. Speakers 44, 29, 30 & 31 are

respectively the daughter and three sons of Speakers 15 & 2. Speakers 33, 34 & 43 are









respectively the two sons and daughter of Speakers 20 & 8. Speaker-41 is the daughter of

Speakers 3 & 16. Speaker-40 is the daughter of Speakers 24 & 10. Speaker-39 is the son of

Speaker-11. Speaker-32 is the son of Speakers 1 & 4. The following sets of speakers are married:

27 & 7; 26 & 9; 22 & 5; 21 & 6; and 14 & 1. Only Speakers 17 & 19 do not have relations in the

subj ect set, but they are very good friends of my mother.

2.3 Variables under Investigation

2.3.1 Linguistic Variables

This study investigates three variables, (q), (6) and (a)", whose realizations in the speech of

non-migrant rural speakers or rural colloquial Arabic (RCA), native Himsi speakers or Himsi

colloquial Arabic (HCA), and migrant rural speakers are illustrated in Table 2-2. Note that Table

2-2 includes an extra variable, (7), because /7/ is a separate phoneme in all varieties and it is

realized as [7] in all varieties. This information will become useful when we study the contrast

between [q] and [7] in RCA in comparison to the neutralization that takes place in HCA in

Section 3.2.1.

Table 2-2. Variants of the variables (q), (6), and (a) in the speech of non-migrant rural speakers,
native Himsi speakers, and migrant rural speakers
Variable Variants of non- Variants of native Variants of migrant
migrant rural Himsi speakers rural speakers
speakers (RCA) (HCA)



(8) [t] ~ [s] [t] ~ [s] [t] ~ [s]
(a) [d] ~ [z] [d] ~ [z] [d] ~ [z]





SCheck Appendix F for tables that include the IPA symbols of SA, RCA, and HCA. These tables will help
throughout this study in understanding the sounds used in phonetically transcribed words.










2.3.1.1 The variable (q)

One of the dependent variables of the study is (q), which is realized in the speech of the

Himsi community, including native Himsi speakers and rural migrants, as two variants: [q] and

[7] (Table 2.2). Native Himsi speakers always use [7], whereas rural migrants may vary between


[q] and [7], taking into account that [q] is their native form. There is no specific phonological


context in which [7] occurs as a replacement for [q] in the speech of rural migrants. It can occur


in many phonological contexts except in certain lexical borrowings from SA, such as [qur?aan]


'Qur'an', [1iqaa?] 'meeting', and [6aqaafe] 'cultural'. For example, the rural migrants' words


[qalem] 'pen', [raqbi] 'neck', and [wareq] 'paper' become [?alam], [ra?bi], and [wara?]


respectively in the speech of those who adopt the Himsi variant [7]. These examples show that

the change could occur word-initially, word-internally, and word-finally. One can also observe in

the given examples that the change from [q] to [7] could also be accompanied by vowel changes,


such as the change of [e]6 to [a] in [qalem] and [7alam] and [wareq] and [wara?]. This is not the


case with all speakers. The change from [q] to [7] could act independently from vowel changes.

However, this is beyond the scope of this study.


6 This is what is called an Emala. This feature exists in some Arabic dialects: it is defined as the use of [e] instead of
[a] or [ee] instead of [aa]. Emala is a feature of the rural dialects under investigation, and it usually undergoes
change in the speech of rural migrant speakers who accommodate the Himsi forms, [a] and [aa], to appear
prestigious. Emala exists in some central Syrian dialects (Versteegh 1997:153), which include the rural dialects
under investigation. It is a historical change from a > e in different linguistic environments. In Northern Syrian
dialects, such as the dialect of Aleppo, this change takes place in the neighborhood of an [i] vowel or when the [a]
vowel occurs after consonants (e.g., lisan > 1sen 'tongue' g ami? > Zeme? 'mosque', kateb > keteb 'writing', and
taleb > teleb 'striving') (Versteegh 1997). The change in the speech of migrant rural speakers takes the opposite
direction of that historical change, i.e. [e] becomes [a], in order to sound urbane. Nonetheless, this vowel change is
beyond the scope of this study and requires further investigation.









It is worth noting that lexical borrowings are excluded from this study because rural and

urban people pronounce them the same. Including them may skew the results, particularly in the

speech of professionals who use a number of jargon words from their own profession and repeat

them many times in their speech. Hence, excluding lexical borrowings from the data yields better

results and gives a better picture of the variation among the various groups of speakers. Ferguson

(1997) and Daher (1998a, 1998b) emphasize that the use of [q] in some words in Damascene

Arabic, which is characterized by the use of the glottal stop as a replacement to the /q/ sound in

Classical/Standard Arabic, is the result of lexical borrowing. Haeri (1991, 1996) also attributes

the existence of [q] in some words in Cairene Arabic to lexical borrowing. In a study (2005) of

'The role of social factors, lexical borrowing and speech accommodation in the variation of [q]

and [7] in the colloquial Arabic of rural migrant families in Hims, Syria', I investigated lexical

borrowings in Himsi colloquial speech. I found that a native Himsi speaker also uses [q] in

borrowed words from SA. I compared the words produced by that Himsi speaker with the [q]

sound with other studies and found that they are similar to the borrowed words used in Cairene

Arabic (Haeri 1991, 1996) and they are similar to the words produced with the [q] sound by the

younger generation. Given the similarity among urban dialects characterized with [7] and based

on my (2005) study of lexical borrowings in HCA, I have no reason to believe that lexical

borrowings should be included in this study.

2.3.1.2 The variable (6)

The second dependent variable is (6), which occurs in the colloquial speech of the Himsi


community as either [t] or [s] (Table 2.2) (e.g., [6al3] 'snow' in SA is pronounced as [tal31 in

HCA and RCA, whereas [6aflab] 'fox' in SA occurs as [saflab] in HCA and RCA). Some words










may occur using either variants (e.g., [6aaluu6] 'Trinity' in SA could occur as either [taaluut] or

[saaluus] in HCA and RCA). One can observe that when one [6] changes to [t], the other [6] also

changes to [t] and so is the case with respect to [s]. Since conversations were very natural, the

SA variant [6] rarely occurred and was mostly in lexical borrowings from SA. Even then, it was


mostly produced as [s], rather than [8]. For example, the form [saaluus] 'Trinity' could be


considered an attempt on the part of the speaker to imitate the SA form. The lack of [6] in the

speech of speakers results in the choice of the closest less marked form possible, in this case [s]

that has a phonemic representation, /s/, in the speakers' repertoire. This is comparable to the Thai

and Russian use of [t] and German and Japanese use of [s] instead of [6] because of the absence

of this sound from their inventory. Lombardi (2003) attributes this phenomenon to the interaction

of faithfulness and markedness constraints. Similarly, markedness could have played a maj or

role historically in producing the SA [6] and [a] as [t] and [d] respectively. The merger of [6]


and [a] with [t] and [d] respectively was completed around the 14th century (Daher, 1998a,

1999). We know that fricatives are more marked than stops cross-lingui stically, and interdental

fricatives are even more marked than other fricatives because of the extra distributed feature

[+Distributed] they have. Consequently, people in the past apparently preferred to use the less

marked forms in their colloquial dialects, deviating from the SA pronunciation. Recently, the

minor reappearance of [6] and [a] in the speech of some Damascene speakers who hold writing

jobs is attributed to increased education and lexical borrowings from SA (Daher, 1998a, 1999).

Daher' s hypothesis may also apply to the production of [6] and [a] as [s] and [z] respectively in

the speech of some speakers. In their attempt to imitate the SA variety, speakers fail to produce









the more marked sounds [6] and [a], replacing them instead with the less marked sounds [s] and


[z] respectively. This can be evidence of the absence of the two sounds /6/ and /8/ from their

inventory, given the fact that those speakers have the two phonemes /s/ and /z/ in their inventory.

In this case, education as a social factor may play a role alongside markedness to yield an output

that is close to the input, but lacking a feature that adds markedness to it (i.e., [+ Distributed]).

This situation mirrors a struggle between markedness and faithfulness constraints that might be

militating against the change of the continuant feature in [6] and [a] but not against the change of

the distributed feature, resulting in [s] and [z] respectively instead.

2.3.1.3 The variable (6)

The third dependent variable is (a), which often occurs as either [d] or [z] in HCA and


RCA (Table 2.2) (e.g., [aaqn] in SA occurs as [da?n] in HCA and [daqn] in RCA, whereas

[aauuq] 'taste, propriety' in SA occurs in HCA as [zoo?] and in RCA as [zooq]). Notice also the

vowel differences between the two words. Some words may occur using either variants (e.g.,

[ai?b] 'wolf' in SA is pronounced [diib] or [zi?b] in HCA and RCA). Notice when the [a]

changes to [d], vowels tend to change too. On the other hand, when [a] changes to [z], vowels

and other sounds existent in the words are retained. This might be an indication that the speaker

is trying to imitate the SA pronunciation but because of the high markedness of [a] or the lack of

this sound in the HCA and RCA phonemic repertoire, s/he tends to resort to the closest sound

possible (i.e., [z]), which exists in their phonemic inventory. At the same time, they maintain the

vowel system and possibly other sounds in that word. One can observe in words such as the SA

[aabiaba] 'vibration, oscillation', which are produced as [zabzabi] in HCA and RCA, that when










one [a] is produced as [z], the other [a] is also produced as [z]. Since conversations were very

natural, the SA variant [a] rarely occurred and was mostly in lexical borrowings from SA. Even


then, it was mostly produced as [z], rather than [a]. As in the case of the variable (q), the very

few lexical borrowings with [6] and [a], which occur in the data are excluded for better

understanding of this phenomenon.

2.3.2 Social Variables

Independent or extralinguistic variables included are as follows:

1. Sex (24 males and 28 females).

2. Age (two age groups: 18-35 and 52+). Twenty-four participants are in the younger age

group, and twenty-eight participants are in the older age group. The older age group

consists of 13 males and 15 females; the younger age group consists of 11 males and 13

females.

3. Social class (two social classes: lower-middle and upper-middle). The social class of

participants is determined based on the community's general classification of them as

somewhat rich or somewhat poor (Section 4.2). This social classification will be examined

in Section 4.2 against the following social indicators: family income (mainly breadwinner

income), education, occupation and residential area. Twenty-three participants are in the

upper-middle class, and twenty-nine participants are in the lower-middle class. However,

education and occupation may affect the person's social class with time. If a person is a

medical doctor or an engineer who comes from a poor family, his/her social status may

change when s/he becomes more known and starts making more money. This kind of

social mobility (Haeri 1991, 1996) is taken into account.









4. Residential areas (two residential areas in the City of Hims: Al-Hameeddieh and Akrama).

Only 13 speakers are from Akrama; the remaining 39 speakers are from Al-Hameeddieh.

2.3.2.1 Overview of Al-Hameeddieh

Al-Hameeddieh is one of the oldest residential areas in Hims connected to the central

downtown area of Hims (Figure 2-2). It is in the center of the old city of Hims, which is

surrounded by a wall that has seven gates that connect the old part of the city to the newer one.

Very few remains of the wall exist on the eastern side of the old city. The gates surround the Al-

Hameeddieh area and continue to play symbolically the role of a connecting passageway

between the old city and the new extensions of the city of Hims. Al-Hameeddieh is mainly a

Christian residential area with cultural and traditional values, which include the linguistic

behavior of the native Himsi inhabitants (i.e., the use of [7]). It obtains many of its cultural and

traditional values from the many historical and residential palaces and historical sites, which

exist in it and surround it. These sites include Umm Al-Zunnar (The Virgin Mary's Girdle)

Church; Mar Elian (St. Elias) Church; Al-Arba'in Shahid (Forty Martyrs) Church; Al-Zahrawi

Palace, a tourist site; Farkouh Palace among others that are turned into beautiful restaurants that

maintain traditional decoration traits; Hims Citadel; Al-Nour Mosque; and the Mosque of Khalid

Ibn Al-Walid, the Muslim conqueror to whom the city fell in 636 A.D. This mosque is also used

as a logo to represent the city of Hims. The residential houses, or, as referred to, palaces, stand

witness to the prominent people that have lived historically in this area. Up to this day, people

who live in Al-Hameeddieh are conceived of by other inhabitants of the city of Hims as upper

class; thus, as a residential area, it is imbued with prestige. The cultural richness of Al-

Hameeddieh contributes to this general view. Being a Christian residential area, many of the

Christian rural migrants, who constitute the participants of this study, prefer living in it to living










in the suburbs. Furthermore, kinship, family ties, and social ties with friends, relatives and

neighbors are highly valued in most of the Arab countries and particularly for rural people.

Hence, it is important for most rural migrants to the city to live in an area, such as Al-

Hameeddieh, where they can maintain connection with their own Christian tradition, practices

and rituals as well as keep their strong ties with relatives and friends who come from the same

background and live in the same area. Living in the same area enables them to see each other

more often and keep up with each other's life as well as have a solid support system.

2.3.2.2 Overview of Akrama

On the other hand, Akrama is a newly developing residential area in the suburbs of Hims

(Figure 2-2). It started developing and growing about thirty years ago. Its development is

concurrent with the establishment and development of Al-Baath University (founded 1979),

which is located in that suburban area. Akrama is mainly occupied by rural migrants, mainly

Alawites whose speech is characterized by the use of [q] Hence, it is more diverse in terms of

inhabitants than Al-Hameeddieh. Therefore, the two residential areas differ with respect to their

history. As a new residential area, Akrama has not yet acquired the prestige associated with Al-

Hameeddieh. The well-established linguistic tradition and prestige associated with Al-

Hameeddieh are expected to have a greater influence on the newcomers, especially since the

maj ority of the residents are native Himsis. This influence not only includes cultural and


SReligion could be considered a variable, as there is always the possibility that Christian rural migrants may behave
differently from other rural migrants from other religions, such as Alawites. This variable is not tested in this study,
but it could be a good source of information in future studies and comparisons among various rural migrant speakers
to the city of Hims or any other urban area in Syria. It is also worth noting here that the [q] sound used by the
Alawites is observed to be stronger and more prominent than the [q] sound produced by the Christians whose speech
is characterized with [q]. Some people ascribe the strength of the [q] sound of the Alawites to their desire to
distinguish themselves as Alawites. They feel proud to be Alawites because the leader of Syria is Alawite; they feel
that they could obtain power over others through their speech. If they sound like an Alawite, other people may fear
them or surrender to their wishes. Consequently, one would expect Alawites to behave differently from Christian
rural migrants. In other words, they may cling to their native linguistic features more than Christian rural migrants.
Thus, one may find that the Alawite rural migrants' use of [q] is higher than that of Christian rural migrants.









traditional values, but also salient linguistic features and values. This influence might be minor in

Akrama, since the maj ority of the residents are not originally Himsis. Those residents have

moved in recently and most of them maintain their native linguistic features since they come

from diverse backgrounds. This, however, does not exclude the possibility that there may be

some influence of the city linguistic features on some residents in Akrama. This could be due to

exposure to the city linguistic features through school, university, workplace and acquaintances

from different parts of the city.

2.4 Analysis

I listened to all recordings more than once for intra-rater reliability, transcribing all the

words that contain the variants under investigation. Since the study deals with quantitative

analysis of the data, there was no need to transcribe all the recordings; it was sufficient to

transcribe features that are relevant to the study.

After transcribing all the relevant words, the number of occurrences of [q] and [7] (Chapter


4); [a], [d], and [z]; and [6], [t], and [s] (Chapter 5) in the speech of each informant is calculated.

The raw numbers of observations are transformed into percentages to have balanced comparisons

among individuals. Percentages within groups and among groups are also calculated to have an

estimate of the difference in variation between males and females, the two age groups, the two

residential areas, and the two social classes. An SPSS program is used to enter the various

dependent and independent variables. The independent variables are classified using a nominal

measure that is assigned a numeric value in SPSS. For example, females and males are given the

numbers 1 and 2 respectively and so is the case with respect to the two age groups, the two social

classes, and the two residential areas. The numbers of observations of each of the dependent

variants for each informant are entered as a scale measure. For example, if the [q] sound occurs









100 times in the speech of a participant, the number 100 is entered as it is for that participant

under the variant column specified for [q] and so is the case for the other variants and for all the

participants. These raw observations are entered into a Generalized Linear Model (GZLM) to

determine the effect of all of the extralinguistic variables on the usage of the linguistic variants.

The GZLM is used when the dependent variable is quantitative and the independent variables are

quantitative and/or qualitative. These models are regression models in which dummy variables

are used for qualitative variables (Agresti and Finlay 1997). The GZLM has an advantage over

the multivariate procedure of the General Linear Model (GLM), which only assumes a normal

distribution of the data. The GZLM does not assume normal distribution, which may not be the

case in variable data. The GZLM can deal with both normally and non-normally distributed

data. Like the GLM, the GZLM enables us to enter more than one independent variable at the

same time and see the effect of each of the social factors on each of the dependent variants. It

also enables us to select the most significant social factors in fomenting the change and to

exclude the non-significant ones. Further, within the same model, we are able to measure the

main effects of each social variable as well as the interaction among those variables. Thus, the

results will show which independent variable(s) is(are) the most significant and which one(s)

is(are) the least significant. Chapter 4 gives more details about the use of these statistical models.

Other analyses were performed, such as investigating the linguistic environment for each

of the four variants, [t] and [s] and [d], and [z], to see if the linguistic environment influences the

choice of one variant over another (Section 3.3). I also manually entered the transcribed words

for each variant for each speaker into the computer. I then calculated the occurrence of each

word in the speech of each individual. I tabulated this information for each of the fifty-two

speakers, including the words, their meaning, the number of occurrence of each word, the total









number of words for each variant, and the total number of words for each set of alternating

variants (i.e., for [t] and [s] and for [d] and [z]). I also calculated the percentages of the total

number of words for each variant with respect to the total number of words for each set of

alternating variants. Then, I manually calculated the number of occurrence of each word in the

speech of the fifty-two speakers to check if frequency plays a role in determining the choice of

one variant over another. I tabulated this information separately, including the words, their

meaning, the number of occurrences of each word, and at the end the total number of words for

each variant. This information forms a small corpus of words for the four variants. I performed

this process first for the variants [t] and [s] and [d], and [z] (Chapter 5). Then, I performed the

same process for the variants [q] and [7] (Chapter 4).

The following chapter will present a theoretical analysis of the data, applying OT and the

GLA. I will introduce a number of social constraints and incorporate them with linguistic

constraints to explore the possibilities that this model can offer us in accounting for

sociolinguistic variation and change.



























AI-H ame eddi eh





~For: of the gates surrournding
Al-H-amee~ddeh. They are from
right to left: Tadmur, DDreeb,
SShaa', ~and Hood.


Akrama


Al-Baath Unliversity


Figure 2-2. City Plan of Hims. Adopted from the Homs City Council (2008). I point to the areas of concern in this study and give their
name in English, as a point of reference.









CHAPTER 3
THEORETICAL ANALYSIS

3.1 Introduction to the Theoretical Analysis of the Data

Having discussed the working mechanism of the GLA and its advantages over other

approaches to language variation, I will now present the theoretical analysis that motivated this

study. In the GLA and OT, mainly linguistic, phonetic, phonological, morphological, or syntactic

constraints have been taken into account. Thus, to answer the question about how we could

model sociolinguistic variation using these theoretical frameworks and how we could intersect

the social and the linguistic within the same theory or framework, I propose a number of social

constraints and integrate them into the GLA. These constraints will interact with linguistic

markedness and faithfulness constraints that militate against a certain sound or against the

change of a certain sound respectively.

I will start with modeling the observed variation between [q] and [7] in the speech of rural

migrant speakers to the city of Hims, proposing and implementing linguistic constraints and

social constraints that are pertinent to this variation, stating the motivation behind them. I will

show examples with and without the intersection of social constraints with phonological

constraints. This will allow the reader to see different results when social constraints are

involved, results that match or approximate real life occurrences. The inclusion of social

constraints in the computation will provide an explanation of the observed intra- and inter-

speaker variation (i.e., the speakers' different grammars within the same social group or

community or among different social groups).

Then, I will take on the constraints that are required to model the stable variation we

observe with respect to [t] and [s] and [d] and [z] in the speech of all speakers. With stable

variation, I mean that certain words are consistently pronounced with one of the four variants and









never with one of the other variants; thus, the four variants are not interchangeable. I will explain

this point more when I explore these four variants in depth in Section 3.3. We will observe later

on that the latter variation is led by different factors from the former variation, in that the

interaction between social constraints and linguistic constraints does not make a difference. What

are at work in this type of variation are mainly faithfulness constraints that militate against the

change of some sound features to maintain contrast in pronunciation among various words that

historically stemmed from SA [6] and [a].


3.2 Modeling Variation between [q] and [7]

In this section, I will first explore the linguistic constraints that are required to account for

the observed variation between [q] and [7] in the speech of rural migrant speakers to the city of

Hims. Then, I will propose a number of social constraints and intersect them with the proposed

linguistic constraints to account for the intra- and interpersonal variation observed in my data.

The proposition of social constraints and intersecting them with linguistic constraints will show

that the social and linguistic can be intertwined in one linguistic theory. It will also show that

social factors do influence our choice of a grammar at a certain time and place. Above all, social

constraints determine the degree of intra- and interpersonal speech variation, the ranking value of

other constraints, and the percentages of the occurrence of each of [q] and [7]. They provide

explanation of the speakers' different grammars and give us expectations on what the speech of a

certain speaker or group of speakers should sound like. In this study, we will see tangible

evidence of the influence of social constraints on our grammar, evidence that makes them not

only influential but also part of our grammar.










3.2.1 Linguistic Constraints Pertinent to the Variation between [q] and [7]

In SA, the two sounds /q/ and /7/ are contrastive; /q/ is pronounced [q] (Haeri 2003) in


speaking, writing and reading, while /7/ is pronounced [7]. The case is similar in the

original/native dialect of rural migrant speakers to the city of Hims, which I will refer to

hereafter as the rural colloquial Arabic (RCA). Thus, in RCA, /q/ and /7/ are also two contrastive


sounds; each is pronounced as [q] and [7] respectively. In the speech of original/native speakers

of Hims, however, these two sounds are neutralized. I will refer to the dialect spoken by Himsi

people hereafter as the Himsi colloquial Arabic (HCA). This difference between HCA, on the

one hand, and RCA and SA, on the other hand, is illustrated in Table 3-1.

Table 3-1. Variants of the variable (q) in SA, HCA, and RCA
Variable Variants in SA Variants in RCA Variants in HCA

(q) [q] [q] [?]


The absence of the [q] sound in the speech of Himsi speakers can be considered an

indication of the absence of /q/ form their phonemic inventory. In other words, only the phoneme

/7/ is present in their phonemic repertoire. For example, the SA words /7alam/ 'pain' and /qalam/


'pen' are pronounced the same in the Himsi dialect, [7alam]s. Where the two words are


phonologically distinct in SA and RCA, they are neutralized in HCA. The meaning of Pallamn is

inferred from the context (Appendix F, Himsi Colloquial Arabic Phonemic System in IPA):

(1) li'ndi allamn Pah~mar. I have a red pen.'


SOther examples of words that can have the same pronunciation in HCA are: /qitaar/ 'train' and /7itaar/ 'frame' are
pronounced as [7itaar]; /qamar/ 'moon' and /7amar/ 'ordered' as [7amar]; /qisim/ 'department, part' and /7isim/
'name': [qaal] 'he said' and [7aal] 'lineage, the family of'; /qalb/ 'heart' and /7alb/ 'Alps Mountains' as [7alb]; and
[daaq] 'taste' and [daa?] 'illness'. [daaq] 'he tasted' is in SA [6aaq]. Compare the colloquial [d] to the SA [6].









(2) li'ndi Pallam maa maf2aul. 'I have an unbelievable pain.'

The supposition that /q/ is absent from the phonemic inventory of HCA speakers is

consonant with Haeri's (1991, 1996) findings which were based on an experiment that was

performed on twenty-seven children, ages five to twelve. Although the experiment was done on

Cairene children, it shows that it is very difficult to elicit words containing [q] from children who

are exposed only to the [7] in childhood. Haeri concludes that the reappearance of [q] in Cairene

colloquial speech is the result of lexical borrowings from SA, and that children acquire [q] later

on in life through formal education. The situation regarding the use of [q] in lexical borrowings

from SA and [7] in all other lexical items seems to be an urban phenomenon that applies in maj or

urban centers, such as Damascus (Daher 1998), Amman (Abdel-Jawad 1981), and Hims (Habib

2005) just to mention a few. Given the similarities among urban Arabic dialects of which [7] is

the prestige marker, there is no reason to believe that HCA is different. In this sense, HCA

speakers lack /q/ in their phonemic system and tend to acquire it later on in life and use it only in

lexical borrowings from SA that are adopted through formal education (Habib 2005, Section

4.1). The difference between HCA and RCA regarding their phonemic systems is illustrated in

Table 3-2. All other consonants are the same in both dialects (Appendix F).

Table 3-2. Phonemic difference between HCA and RCA speakers
Dialect Phoneme Corresponding Realization
RCA /q/ [q]

HCA /?/ [?]
RCA migrant speech /q/ [sl ~ [?]










Moreover, the two words /7alam/ and /qalam/ contrast in their underlying and surface


representation in SA and RCA, but their surface representation is the same in HCA, [7alam],

indicating neutralization. Thus, an RCA speaker who has the underlying form /qalem/9

neutralizes the /q/ sound of his/her surface form [qalem] into [7] if s/he adopts the HCA


grammar. In order to understand the mechanism of acquiring the new form [7] by RCA speakers

who migrate to the city, I will present first the individual grammar of each dialect pertinent to the

use of [q] and [7]. Then, I will show what happens when an RCA speaker tries to adopt the new

form and make it part of his/her own grammar.

3.2.1.1 Grammar of RCA

As mentioned earlier, in RCA, like SA, the two sounds [q] and [7] are contrastive. In order

to preserve this contrast, we need a set of constraints to depict the RCA grammar. Since we are

dealing with the two sounds [q] and [7], I propose that there should be two context-free

markedness constraints pertaining to the two sounds in question. These two constraints are *q

(Voiceless uvular stop is marked) and *7 (Glottal stop is marked). Each of the two constraints

militates against one of the two sounds. The preservation of the contrast requires some

faithfulness constraintss. The feature Retracted Tongue Root (RTR) is characteristic of the [q]

sound, as indicated in Davis (1993, 1995), Shahin (1998), and Zawaydeh (1998). Hence, to

preserve this feature in /q/, we need the faithfulness constraint IDENT-IO (RTR) (Let a be a

segment in the input, and p a correspondent of a in the output. If a is [yRTR], then P is [yRTR]).

9 NOtice here that the underlying form is /qalem/, not /qalam/ that is the underlying form in HCA, because this is the
surface form in RCA. Speakers change the [e] to [a] when they switch from RCA to HCA. Explanation of this
change is beyond the scope of this study.









We may need one further faithfulness constraint to protect the change of /7/ into [q] in RCA,

since /7/ exists alongside /q/ as an underlying form that has the surface form [7]. This constraint

should be related to the maintenance of the glottal feature in the underlying form or to the place

of articulation in general. We can call this constraint IDENT-IO (Glottal/Place) (Let a be a

segment in the input, and p a correspondent of a in the output. If a is [yGlottal/Place], then P is

[yGlottal/Place]). Thus, the four main constraints that are of concern to us are repeated here:

(3) *q Voiceless uvular stop is marked

(4) *7 Glottal stop is marked

(5) IDENT-IO (RTR) Let a be a segment in the input, and p a correspondent of a in the
output. If a is [yRTR], then P is [yRTR]

(6) IDENT-IO (yGlottal/Place) Let a be a segment in the input, and P a correspondent of a in the
output. If a is [yGlottal/Place], then P is [yGlottal/Place]

However, the constraint IDENT-IO (Glottal/Place) may be redundant, because the constraint

IDENT-IO (RTR) can be construed in a way that the addition or deletion of the [RTR] feature is a

violation of that constraint. Al-Khatib (2008) also calls this phenomenon in urban dialects such

as the one in Amman "depharyngealization". That is, those who use the urban form [?] lose the

feature [RTR] in their speech. In this sense, we can depict the grammar of RCA by using only

the three constraints in (3, 4, & 5). To maintain the contrast between [q] and [7] in RCA, we need

the faithfulness constraint IDENT-IO (RTR) to be higher ranked than the other two constraints, *7

and *q. The ranking between *7 and *q is irrelevant because both sounds are possible in RCA

and both constraints are violable; hence, the ranking in (7) represents the grammar of RCA.

(7) IDENT-IO (RTR) >> *7, *q










The ranking in (7) guarantees a corresponding output form equal to the input. Violation of

IDENT-IO (RTR) is fatal. If the input contains /q/ as depicted in Tableaulo 3-3, using the word

/qaraabi/'kinship', the output will contain [q]. In contrast, if the input contains /7/, the output

will contain [7] as illustrated in Tableau 3-4, using the word /7alem/ 'pain'. Thus, a contrast is

maintained in RCA by employing the ranking in (7).

Tableau 3-3. Ranking of RCA: Underlying form /qaraabi/
/qaraabi/ IDENT-IO (RTR) *? q
Sa. qaraabi *
b. ?araabi *! *

Tableau 3-4. Ranking of RCA: Underlying form /7alem/
/7alem/ IDENT-IO (RTR) *? *q
Sa. ?alem
b. galem *! *


3.2.1.2 Grammar of HCA

For the HCA grammar, we will use the same set of constraints that were employed in

Section 3.2.1.1 to be able to compare how the two grammars differ with respect to the ranking of

the three pertinent constraints in (3, 4, & 5). As HCA exhibits neutralization, a different ranking

of the constraints in (3, 4, & 5) is required. Since the [q] sound is not desirable in HCA and is

absent from the speech of Himsi speakers with the exception of lexical borrowings from SA, the

constraint *q should be higher ranked than the constraint *7 that is violable in HCA.


Consequently, *7 should have a very low ranking in HCA grammar"l. The constraint IDENT-IO



'0 Tableau is the name for table in Optimality Theory. The plural is tableaux. These tableaux show which
constraints are relevant in the analysis and the ranking of these constraints with respect to each other.

'' The L1 learner of HCA has learned to demote *7, so s/he can pronounce [7]. Not being exposed to [q], the L1
learner of HCA has no reason to demote *q. *q starts high in his/her grammar and maintains that high ranking.










(RTR) could be higher ranked than *7 as well because we would like to preserve /7/ in the output

as [7]. In HCA grammar, the ranking between IDENT-IO (RTR) and *q is irrelevant because they

are both violated by the loser and neither is violated by the winner. This leads to the following

ranking of the HCA grammar:

(8) *q, IDENT-IO (RTR) >> *7

The ranking in (8) guarantees the [7] pronunciation of the input /7/ in HCA. For a Himsi


speaker, the input is constantly /7/ rather than the SA and RCA /q/, because as I mentioned

earlier /q/ is absent from his/her phonemic repertoire and /7/ is the first form that s/he is exposed

to in childhood (Prince and Smolensky 2004 [1993]). For example, if we take the input /7araabi/

'kinship' that has the underlying form /qaraabi/ in SA and RCA and apply the ranking in (8), the

winning output will be [?araabi], not [qaraabi], as illustrated in Tableau 3-5. The suboptimal

output [qaraabi] violates the higher ranked constraints, *q and IDENT-IO (RTR), allowing for the

optimal output in (b) to win although it violates the constraint *7.

Tableau 3-5. Ranking of HCA: Underlying form /7araabi/
/7araabi/ *9 IDENT-IO (RTR) *?
a. garaabi *! *
b. ?araabi *


3.2.1.3 Acquisition of the HCA form by an RCA speaker

One may ask the question why [q] is replaced by [7] in the HCA, not [k] (Abdel-Jawad


1981; Sawaie 1994), [g] (Al-Ani 1976; Abdel-Jawad 1981; Ferguson 1997), [d3112 (Palfreyman


12 I use the IPA transliteration instead of other forms used in the original sources, such as [j] in Palfreyman and al-
Khalil (2003) and [g] in Versteegh (1997).









and al-Khalil 2003; Versteegh 1997:149), or its more fronted variant [dz] (Versteegh 1997:149).

/g/ used to exist in Old Arabic, but with time, it was lost for another place of articulation,

palatoalveolar affricate /d3/ (Moscati, Spitaler, Ullendorff, and von Soden 1964). Nowadays, [g]

still exists in some Bedouin or village dialects, where it replaces the /q/ sound (Abdel-Jawad

1981; Ferguson, 1997). Because [7], not [k], [g], [d3], [dz], or [q], is the sound used in HCA, we

can assume that if a set of context-free markedness constraints exist for all the unused sounds in

HCA, these constraints will be ranked higher than *7 in HCA, as illustrated in (9).


(9) *k, *g, *d3, *dz, *q >> *7

On the other hand, in the RCA dialect [q] and [7] have equal weight because they are

contrastive. Thus, the same set of context-free markedness constraints will be ranked in RCA as

in (10) because the other four sounds do not occur in RCA.

(10) *k, *g, *d3, *dz >> *7, *q

Since neither dialect allows [k], [g], [d3] or [dz] as a replacement for [q], I do not include


the constraints *k, *g, *d3, and *dz in the ranking of constraints for each dialect. In both dialects,

these four constraints are undominated. Thus, we should not concern ourselves with them. These

constraints could be used to describe changes that occur in other Arabic dialects that make use of

the sounds [k], [g], [d3], or [dz] in place of [q]. Our main concern here is the change of grammar

that occurs as a result of the switch from RCA to HCA. Since the change is taking place from the

RCA form to the HCA form, there must be some process of reranking of the three constraints in

the RCA grammar (IDENT-IO (RTR) >> *q, *7) for an RCA speaker to sound like an HCA

speaker. Because the change is taking place from the village dialect to the city dialect, the









assumed input is the surface form of the RCA dialect, in accordance with Prince and

Smolensky's (2004 [1993]) idea of Lexicon Optimization, which is an elaboration on Stampe's

(1972) idea. Lexicon Optimization holds that the input should be the form closest to the surface

form that a child is first exposed to, unless there is evidence otherwise as from alternations, and

it is this input that undergoes change.

It is worth noting here that the acquisition of the HCA form does not necessarily involve

complete acquisition of the HCA grammar because of the difference in input form for an HCA

speaker and an RCA speaker. It also seems that the RCA ranking (IDENT-IO (RTR) >> *7, *q)


promotes faithfulness over markedness to maintain the contrast between [q] and [7], whereas the

HCA ranking (*q, IDENT-IO (RTR) >> *7) partially promotes markedness to avoid the marked


sound [q]. This implies that rural speakers who attempt to learn the new form [7] of the HCA

dialect are inclined towards the unmarked. It may appear to the observer at first sight that the

variation observed in the speech of rural migrant speakers is only a kind of change towards the

unmarked.

However, the situation is more complex than it may appear. It is true that the variation

observed in the speech of rural migrant speakers to the city of Hims is an indication of an attempt

on the part of the learners to acquire the HCA grammar. However, the degree of variation differs

from one speaker to another and from one group of speakers to another. The degree of use of the

new form may range from zero to one hundred percent based on a number of social factors, such

as age, gender, social class, and residential area. Before delving into the social influence on our

choice of a grammar, let us portray what we mean by complete acquisition of or competence in

the HCA form by an RCA speaker. In other words, let us examine what results from the

interaction of the two grammars.










For an RCA speaker to produce his/her input form /qaraabi/ as [7araabi], the output form

uttered by a Himsi speaker, s/he has to be capable of adapting his/her grammar to accommodate

the HCA form. In this case, the RCA ranking (IDENT-IO (RTR) >> *7, *q) should take the form


of (*q >> *7, IDENT-IO (RTR)). *q should be higher ranked than the other two constraints and

the faithfulness constraint IDENT-IO (RTR) is demoted at the expense of the markedness

constraint *q to avoid the output form [q]. The ranking between *7 and IDENT-IO (RTR) is

irrelevant because both of them are violable by the winner. In other words, an RCA speaker has

to rerank two of the three constraints of his/her grammar (i.e., the RCA ranking) to accommodate

the HCA form, sound like a Himsi person, and fit the image that s/he wants to display in front of

the interlocutor. Consequently, we arrive at a ranking that does not totally mirror the HCA

ranking due to the input difference between an HCA speaker and an RCA speaker. This ranking

is exemplified in (11) and Tableau 3-6.

(11) *q >> *7, IDENT-IO (RTR)

Tableau 3-6. Acqisition of HCA formby RCA spaer: Underlyingform /qaraabi/
/qaraabi/ *q "? I DENT-IO (RTR)
a. garaabi *!
,b. ?araabi *

Although the winner (b) violates two lower ranked constraints, the violation of the *q constraint

is fatal because it is ranked higher than the other two constraints, yielding [qaraabi] in (a) a loser.

What is portrayed in Tableau 3-6 is a case in which an RCA speaker has completely acquired the

HCA form. This is not the case of each migrant speaker, though. Some speakers tend not to

acquire the new form at all while others acquire it with one hundred percent accuracy. The latter

ones may switch between the RCA grammar and the new grammar successfully according to the

interlocutor: a family member or a stranger. There are also speakers who show intrapersonal









variation. In other words, those speakers oscillate in their use of the two forms in their own

speech with the same interlocutor. Thus, intrapersonal variation exists alongside interpersonal

variation. The interpersonal variation is existent not only among individuals but also among

groups of speakers. This brings us to the question of how we could deal with these types of

variation that are based on social factors within the same framework.

It has been mentioned in Chapter 1 that the GLA proposed by Boersma and Hayes (2001)

can account for optionality and free variation. If we take the case of two speakers: one who

acquired the new form one hundred percent and another one who acquired the new form only

fifty percent, we will have completely different representations of each speaker when applying

the GLA. According to the GLA, a speaker who can completely switch between the RCA and

HCA forms in accordance with her/his interlocutor is capable of setting and resetting the ranking

values of the three constraints with great success. This speaker has been able to adjust the

ranking of two of his/her constraints to adopt the new form and proj ect acquisition of that form.

Thus, s/he is capable of switching between the RCA ranking of constraints (Figure 3-1) and the

HCA new form ranking (Figure 3-2) according to the interlocutor or at a certain time or place.




IDENT-IO (RTR)






More strict Less strict

Figure 3-1. Ranking of RCA














IDENT-IO (RTR)


More strict Less strict

Figure 3-2. Ranking of HCA form

The case of a speaker who oscillates between the RCA ranking (IDENT-IO (RTR) >> *7,


*q), represented in Figure 3-1, and the HCA new form ranking (*q >> IDENT-IO (RTR), *7),

represented in Figure 3-2, is different. The two rankings have equal weight if each one of them

occurs fifty percent of the time during speech with one interlocutor. In other words, a speaker

who oscillates between the two forms does not have a strong grip on one ranking or the other.

Reranking of constraints and reselection of ranking values for those constraints is a continuous

process. In other words, oscillating speakers have not acquired the new ranking values of

constraints completely. Sometimes they approximate the right ranking value of the two shifted

constraints and produce the new form correctly. At other times, they revert to their original

ranking value associated with their native dialect grammar and produce the old form. In this

case, selection points remain fixed in the center of the range or strictness band of the three

constraints, but the ranking values of those constraints are very close to each other. The

closeness of the ranking values of the constraints leads to overlapping of constraints (Figure 3 -3)

and thus variation. Given that the ranges of those constraints are equal in width, when they are

given close ranking values whose difference from each other is less than 8-10 points, the

constraints inevitably overlap on a continuum/continuous scale. This overlapping may, however,

be optional and depend on personal choice and perhaps self-correction, which may influence the

ranking value assigned by the speaker to one constraint at a certain time or a place.













IDENT-IO (RTR)


More strict Less strict

Figure 3-3. Ranking of the three constraints IDENT-IO (RTR), *7 and *q in an oscillating speaker

This overlapping results in different outputs for the same input, depending on the speaker' s

personal conscious choice. The faithfulness constraint IDENT-IO (RTR) in Figure 3-3 is

struggling between the two context-free markedness constraints *7 and *q. This means that an

oscillating speaker fails to adjust the ranking values of constraints completely to one of the

rankings of the two forms. S/he is stuck in an intermediate stage where s/he oscillates between

two or more different ranking values. One cannot tell if s/he will ever get over this oscillating

position (i.e., leaning towards one ranking value or the other). Nonetheless, what determines the

percentages of oscillation between two grammars or rankings? In other words, why may some

speakers choose to use, for example, one ranking 30% of the time and another ranking 70% of

the time and vice versa? In addition, what are the factors that motivate people to acquire a new

form and to what degree do they influence a speaker' s choice of one grammar over another?

These questions motivated me to examine the social influence in the observed variation in

my data between [q] and [7]. A number of social factors appeared to have great influence on the

choice or preference of one variant over another. This led me to propose a number of social

constraints that I will be stating in Section 3.2.2. I will show how these social constraints play a

maj or role in answering the above questions. Their intersection with linguistic constraints alters

the results; the number and kind of social constraints involved offer various grammars, which we

will see as we delve into more details in Section 3.2.2.










3.2.2 Social Constraints Pertinent to the Variation between [q] and [7]

In this section, I propose a number of social constraints based on social factors that have

been often investigated and stressed as essential factors in language variation and change (e.g.,

Laboy 2001). Among these factors are age, gender, residential area, and social class (e.g., Labov

1966, 1972a; Milroy 1980; Eckert 1989, 1991). I choose these factors because I believe that they

play a maj or role in this study, and they are applicable to other studies and languages or dialects.

In the following list of constraints, each of the social factors included in the study is split

into four social constraints. Of course, this division of constraints may differ in other languages

or when dealing with different variables. As I am dealing here with the variation between [q] and

[7], I state the social constraints that can restrict the use of each of these two sounds. For


example, the category of gender consists of males and females who can either use [q] or [7]. This

motivates the division of the gender social factor into four separate gender constraints, each

stating that one of the sounds is socially marked when used by one of the genders. Thus, gender

is factored first into males and females. Then, males are factored into marked usage of [q] and

marked usage of [7]. Females are factored in the same way, resulting in four distinct social


constraints pertaining to gender and to the use of [q] and [7]. Each social factor is treated in the

same way, resulting in sixteen different social constraints. Each social factor consists of two

categories, and each category is split into those to whom [q] is marked socially and those to

whom [?] is marked socially. We need a balanced number and type of constraints for each social

factor, even in cases in which certain constraints are inactive or not activated by a certain

individual or a group of individuals because the theory should be as general as possible. In

addition, the social values of linguistic sounds may differ from one speaker to another or from









one group of speakers to another. Thus, the application of these constraints depends on what

variant each speaker or group of speakers values the most.

Having explained how social constraints were developed and factored for this study, it is

important to familiarize ourselves with these social constraints before starting the analysis. I,

therefore, would like to present them first in bulk and then manipulate them later individually or

in groups, intersecting them with the previously proposed linguistic constraints to see the role

they play in determining the choice between [q] and [7]. The sixteen social constraints that I

propose and that are deemed necessary for the variable use of [q] and [7] are:


Gender:


*F[q]

*F[7]

*M[q]

*M[7]

*OLD[q]

*OLD[?]

*YOUNG[q]

*YOUNG[?]

*AKR[q]

*AKR[7]

*HAM[q]

*HAM[7]

*UMc [q]

*UMc [7]


[q] is marked socially in the speech of females

[7] is marked socially in the speech of females

[q] is marked socially in the speech of males

[7] is marked socially in the speech of males

[q] is marked socially in the speech of older generation

[7] is marked socially in the speech of older generation

[q] is marked socially in the speech of younger generation

[7] is marked socially in the speech of younger generation

[q] is marked socially in the speech of Akrama residents

[7] is marked socially in the speech of Akrama residents

[q] is marked socially in the speech of Al-Hameeddieh residents

[7] is marked socially in the speech of Al-Hameeddieh residents

[q] is marked socially in the speech of upper-middle class

[7] is marked socially in the speech of upper-middle class


Age:








Area:








Social class:









*LMC[q] [q] is marked socially in the speech of lower-middle class

*LMC[7] [7] is marked socially in the speech of lower-middle class

3.2.3 Praat Analysis of the Variation between [q] and [7]

All coding and data analysis was carried out using the Praat software (Boersma and

Weenink 2007), which was fed with scripts written by me. After developing and introducing a

number of social and linguistic constraints, these constraints should be written into scripts that

are readable by the Praat software. In the script, I define my grammar by providing the relevant

constraints and suggest a ranking value for each constraint as well as the possible input forms

and all their corresponding output forms. I also indicate which constraint each input/output

violates. An example of a script is:

File type = "ooTextFile"

Obj ect class = "OTGrammar 1 "





6 constraints

Constraint [1]: "*?" 100 100 *7

Constraint [2]: "*q" 100 100 *q
Constraint [3]: "I\s(DENhT)-IO (R TR)" 100~C 100~C IET -IOrT (R TR)


Constraint [4]: "*M[7]" 100 100 *M[7]

Constraint [5]: "*O\s(LD)[7]" 100C 100C *OLD[?]3


Cnonsraint [6]: "*A\s/(KR)[7]" 100~C 100!~ *AK(R [7]









0 fixed rankings


1 tableau

Input [1]: "qiffli" 2

Candidate [1]: "qiffli" 0 10 00 0

Candidate [2]: "?iffli" 1 01 11 1

In the above script, I develop an OT grammar that includes six constraints: three linguistic

constraints and three social constraints. I give all constraints the same starting ranking value,

100. The numbers written to the right of Candidates [1] and [2] refer to the number of violations

that each candidate incurs of each constraint in the order they are listed in the script, although

their order before being listed is arbitrary because Praat cares only about ranking values and

number of violations. Once constraints are listed, the location of violations should correlate with

the order in which these constraints occur in the list. Then, Praat is asked to read this script from

a file. Since my data is variable, I follow the evaluation process of stochastic OT in which I

choose a 2.0 noise that is representative of the standard deviation according to a Gaussian

distribution (Boersma and Weenink 2007). I choose 100,000 times as the number of repetition of

evaluation or learning trails.

In order to discover the distribution of each possible output, one can follow either of the

following processes. First, one can select the defined OT grammar and choose Input to Outputs,

which results in two string obj ects that could be sent to distribution, yielding the percentage of

each output. Then one can select the OT grammar and the two strings and learn with a plasticity

of 0. 1 to see the gradual learning process of a grammar. One can further go to check the output

distribution to see if any change or gradual learning has taken place. However, one can take a









shortcut to grammar learning (Boersma and Weenink 2007) by skipping the strings step and

immediately choose to go to pair distribution from selecting the OT grammar, then select the OT

grammar and the pair distribution and leamn with a plasticity of 0. 1. This not only allows us to

observe learning but also to find out the output distribution (i.e., the percentage of the occurrence

of each variant) by choosing to go to strings from pair distribution and then from strings to

distribution. Consequently, we will get input distribution and output distribution. This process is

applied to a number of grammars defined by me to see how social constraints could affect our

choice of a grammar at a certain time and place and the degree to which they affect our

grammars and the ranking values as well as the rankings of other constraints.

Because I am interested in modeling real life variation, after applying the previous process

I chose to apply learning from partial outputs that become inputs for speakers. I select the OT

grammar and the output distribution or the output string and learn from partial outputs. The

motivation behind this type of learning is that speakers are constantly exposed to variation in the

output they receive as input. This should definitely influence their learning process and their next

set of outputs. In other words, it will influence the percentages of the occurrence of each of the

variants in their speech. The complete process of learning from partial outputs is that we first

read an OT grammar from a script file prepared and defined by the author, select it, and go to

pair distribution. Then, we select the pair distribution that appears in the obj ect window and go to

strings. We can select only the output string and go to output distribution. This output

distribution is selected with the previously defined OT grammar, and we let Praat leamn from

partial outputs with a plasticity of 0. 1 and a number of learning trails set to 100,000. In order to

explain this learning process, I will present a number of examples and show how it can model

real life occurrences of the variants under investigation. It is worth noting here that the examples









that are presented here are just a small fraction of the great number of experimental trials

performed with the GLA, not only the intersection of social constraints with linguistic constraints

but also comparing the exclusive use of linguistic constraints to the intersection of constraints.

These experimental trials also include manipulation of the output distributions or pair

distribution to show that learning from different output distributions result in different degrees of

variation, and thus different grammars. However, even when we do not manipulate the output

distribution, we still get different results when social constraints are involved. The degree of

variation depends on the type and number of social constraints involved. The following examples

will exhibit just some of what we can do and what we can model with the GLA regarding the

social phenomenon of sociolinguistic variation and change. They will also manifest the maj or

role played by social constraints in determining the percentage of each output.

In all examples, the input is the RCA surface form, the form that migrant speakers are first

exposed to in childhood, unless specified otherwise. The ranking value of all constraints at the

start of a trail is 100 unless specified otherwise. Pair distributions are attained from this impartial

grammar unless manipulation in the distribution is specified otherwise. All learning trials run

100000 times, that is 100000 steps of learning. In all trials, the evaluation noise is 0.2 and the

plasticity is 0.1. Let us look first on what variation looks like when social constraints are not

involved.

Example 1: Starting with an OT grammar that contain the three linguistic constraints *7,

*q, and IDENT-IO (RTR) and setting their ranking values to 100. Learning from partial outputs

after 100000 trials leads to the ranking values in Table 3-7 and the following output percentages:

33% of [7] and 67% of [q].









Table 3-7. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR)
Constraint Ranking value Disharmony
'F? 100.052 102.829
IDENT-IO (RTR) 100.052 99.313
*q 99.948 100.538


Disharmony in Table 3-7 and in other upcoming tables refers to the effective ranking value

during stochastic evaluation (i.e., with non-zero evaluation noise) and it is usually different from

ranking (Boersma and Weenink 2007). According to Boersma (1997a: 1),

If every constraint in an Optimality-Theoretic grammar has a ranking value along a
continuous scale, and the disharmony of a constraint at evaluation time is randomly
distributed about this ranking value, the phenomenon of optionality in determining the
winning candidate follows automatically from the finiteness of the difference between the
ranking values of the relevant constraints. The degree of optionality is a descending
function of this ranking difference.

In this sense, disharmony is the effective ranking of a constraint "determined at evaluation time

from the ranking value and a random variable" (Boersma 1997a: 3). It is more or less the ranking

value added to it the degree of spread of the ranking value of the constraint multiplied by a

Gaussian random variable with mean 0 and standard deviation 1, which is referred to as z. the

formula for calculating disharmony is:

disharmony = ranking + rankingSpreading z (Boersma 1997a: 3)

The result in Example 1 is expected because there are two constraints that are in support of

the [q] sound (i.e., *7 and IDENT-IO (RTR)) and only one in support of the [7] (i.e., *q). This


gives a fraction of two-to-one out of three, that is approximately 67% of [q] and 33% of [7].

Table 3-7 could be misleading because one may infer from it that the two constraints *7 and

IDENT-IO (RTR) are always higher ranked than *q. In fact, there is no categorical ranking of the

aforementioned three constraints. They overlap and their overlapping leads to the observed

variation in the output distribution. The close ranking values observed in Table 3-7 must









inevitably result in this overlapping and the non-categorical ranking. The percentages indicate

the number of times the ranking that allows for [q] is chosen and the number of times the ranking

that allows for [7] is chosen. But in real life this is not the case. In real life, a person who is

exposed to the overt form [q] 100% of the time, should have 100% pronunciation of [q]; his/her

underlying form, /q/, and surface form [q] should match. There should be a categorical ranking

that determines the grammar of that speaker and identical to the RCA grammar ranking, IDENT-

IO (RTR) >> *q, *7. The disparity between real life occurrences and the results of the Praat

software lies in that the software is given equal rankings to the three constraints and two output

candidates for one input (e.g., the outputs [qaraabi] and [7araabi] for the input /qaraabi/).

Consequently, the number of constraints that favor one of the two outputs over the other

determines the output distribution or percentages. This situation mirrors Anttila's SG model for

free variation in which the number of constraints that favors one sound over another determines

the frequency of its occurrence (Anttila 1997a,b, 2002; Anttila and Cho 1998). In Section

1.3.1.3, I mentioned that Anttila' s SG model runs into problems if there is a large disparity in the

output frequencies. The GLA can overcome this problem because of its stochastic process of

learning and the ability to manipulate the ranking values of constraints or the output pair

distribution from which a speaker can learn and mirror his/her environment. The ability to

manipulate the output pair distribution to match real life occurrences can give us the ranking

values of constraints in a speaker or a group of speakers (i.e., the grammar implemented by a

speaker or a group of speakers). Thus, instead of having multiple grammars from which a

speaker chooses one as in the case of SG, the GLA presents one stochastic grammar for each

speaker or group of speakers through manipulation of the pair distribution or the ranking values

of constraints.









Thus, to solve the unrealistic results of Example 1, we can either set the ranking values of

the constraints according to the speaker' s supposed grammar or feed the GLA with the real

output percentages from a pair distribution file. Applying the first method, we can model the

RCA grammar by giving the two constraints *q and *7 a lower ranking value than IDENT-IO

(RTR) by at least 10 points to guarantee that the latter constraint will be higher ranked than the

former two and that the winning candidate will remain the same (Boersma and Weenink 2007).

For example, we can set IDENT-IO (RTR) to 100 and the other two constraints to 90 or lower.

This manipulation of the ranking value of constraints secures the RCA categorical ranking of

IDENT-IO (RTR) above *q and *7 and the equal ranking of the latter two as well as an output

distribution of [q] that is equal to 100%. This same result can be achieved by applying the second

method ( i.e., manipulating the pair distribution values to 100% occurrence of [q] and [7] and 0%

occurrence of [?] and [q] when the input contains /q/and /7/ respectively). The latter

manipulation will ensure the occurrence of the RCA ranking 100% of the time. That is, it will

gaurantee a much higher ranking value of IDENT-IO (RTR) than *q and *7 as Table 3-8 shows.

There is at least a difference of thirteen points between the ranking value of IDENT-IO (RTR) and

the ranking values of each of *q and *7. Consequently, [q] occurs 100% when the input has /q/


and [7] occures 100% of the times when the input has /7/.

Table 3-8. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR), manipulating the
output distribution to 100% use of [q] and [7] and 0% use of [?] and [q] when the
input has /q/ and /7/ respectively
Constraint Ranking value Disharmony
IDENT-IO (RTR) 113.702 111.934
*q 100.594 101.045
*7? 99.594 103.045









Moreover, the percentages of Example 1 33% of [7] and 67% of [q] and their

discrepancy with real life occurrences were the result of equal rankings of the three constraints,

two output candidates for one input, and the number of constraints that favor one output/sound

over another. As I mentioned earlier, manipulating the output pair distribution of the three

aforementioned linguistic constraints can give us results that match the real life occurrences of

each speaker or group of speakers. For example, If we take the same three linguistic constraints

and manipulate the pair distribution of [q] and [7] to match the real life occurrences of older

female speakers from Akrama (i.e., 82% of [q] and 18% of [?]), we get the ranking values in

Table 3-9 and the approximate output percentages 83% of [q] and 17% of [7].

Table 3-9. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR), manipulating the
values of the pair distribution to 18% [7] and 82% [q]
Constraint Ranking value Disharmony
"? 100.695 101.943
IDENT-IO (RTR) 100.695 101.748
*q 99.305 101.139


Manipulating the pair distribution of [q] and [7] to match the real life occurrences of older

male speakers from Akrama (i.e., 96% of [q] and 4% of [7]) gives the ranking values in Table 3-

10 and the approximate output percentages 94% of [q] and 6% of [7].

Table 3-10. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR), manipulating the
values of the pair distribution to 4% [7] and 96% [q]
Constraint Ranking value Disharmony
*7? 101.634 104.884
IDENT-IO (RTR) 101.634 98.528
*q 98.366 99.081









Furthermore, manipulation of the pair distribution of [q] and [7] to match the real life

occurrences of younger female speakers from Al-Hameeddieh (i.e., 2% of [q] and 98% of [7])

leads to the ranking values in Table 3-11 and the approximate output percentages 2% of [q] and

98% of [7].

Table 3-11. Ranking values and disharmonies of *7, *q, and IDENT-IO (RTR), manipulating the
values of the pair distribution to 98% [7] and 2% [q]
Constraint Ranking value Disharmony
*q 103.289 103.766
'F? 96.711 97.238
IDENT-IO (RTR) 96.711 95.237


The last three manipulations give us three different grammars for three different groups of

speakers all of which are rural migrants to the city of Hims. Further manipulations will show us

that even individual speakers may have different grammars from the grammar of the social group

to which they belong. For example, Speaker-30 (and so is the case concerning Speakers 29 & 31)

is a young male residing in Akrama. He uses [q] 96% and [7] 4%. His grammar and the ranking

values of his constraints will be the same as those of older male speakers residing in Akrama

(Table 3-10). Although Speaker-30 belongs to the younger age group, his grammar differs from

that of the younger age group who uses [q] 3% and [7] 97% (Section 4.4. 1). The younger age

gourp grammar will be similar to that of young female speakers from Al-Hameeddieh and the

ranking values in Table 3-1 1. This disparity between the members of the same social group

raises the issue of the unrelatedness of the grammar of some members of a social group to the

grammar of that social group. Manipulating the pair distribution and creating different grammars

for each speaker or group of speakers from the same linguistic constraints does not explain why

speakers that belong to the same social group behave differently and have different grammars. It









does not also explain why the grammar of one social group differs from the grammar of another

social group (e.g., the difference between older female speakers and younger female speakers).

This is one of the reasons that motivated me to intersect social constraints with linguistic

constraints within the same framework.

Including social constraints in the computation provides some kind of explanation of what

takes place in the mind of a speaker because of social factors. In other words, including social

constraints provides explanation for such disparity among members of the same social group or

disparity among different speakers or social groups. Interacting social constraints with linguistic

constraints in the same framework is a simple comprehensive method to depict and explain the

mental process of a speaker at a certain time and/or place. One can see what happens in the mind

of a varying speaker in one simple step in the GLA. Feeding the GLA with the right output

di sribution/percentages gives the specific ranking values or grammar of each speaker or group of

speakers. In other words, the model takes advantage of the stochastic grammar embedded in the

GLA to generate grammars that match real life output percentages without trying all the possible

rankings of constraints and counting the times those rankings give a certain output, as it is the

case in SG. Some real life percentages are hard to achieve exactly, when applying only all the

possible rankings of the pertinent constraints. This is the downfall of the SG, although one

should take into consideration that knowing the number and type of social constraints involved

can help in giving approximate percentages to real life occurrences. Thus, knowing the social

constraints involved and manipulating the output distribution in the GLA to match real life

output percentages provide us with more precise grammars that match the target output

percentages in a sociolinguistic variation situation.









Another reason for including social constraints in the computation is that speakers may

rank social constraints that do not affect their speech. Ranking these constraints gives speakers

and listeners alike expectations about what the speech of others would sound like. For example,

an older speaker may rank *YOUNG[q] higher than other constraints, although this constraint may

not affect the way s/he speaks. Such ranking gives him/her expectations about what the speech of

younger speakers should sound like. This higher ranking of *YOUNG[q] tells him/her that

younger speakers do not use or are not expected to use the [q] sound in their speech. The same is

applicable to other speakers or groups of speakers who may rank constraints that are not active in

their speech but are active in others, and thus know the grammar of the other/interlocutor.

Thus, in the case of sociolinguistic variation, it is wise to include social constraints in the

mental process of a speaker because they influence the way s/he speaks. These social constraints

play a role in the linguistic discrepancy among speakers (even members of the same social

group) or groups of speakers. This means the different degrees of variation that are encountered

in the speech of rural migrant speakers to the city of Hims are influenced by a number of social

constraints, and it is the ranking values of those constraints that lead to change and determine the

degree of variation. Therefore, there is much more involved than the manipulation of the pair

distribution or the ranking values of the aforementioned three linguistic constraints. I believe that

social constraints should be taken into consideration to see their simultaneous degree of effect on

the variation between [q] and [7] in the speech of rural migrant speakers to a city in which [7] is

the prestigious form. In Example 2, we will see a difference in percentages from Example 1 as a

result of the intersection of one social constraint with the aforementioned three linguistic

constraints.









Example 2: If we add the social constraint *F[q] ([q] is marked socially in the speech of

female speakers) to the computation in Example I with the various criteria noise (2.0),

plasticity (0. 1), number of learning trials (100000), and ranking values (100) set to the same

values as in the previous experiment and in the following experiments, learning from partial

outputs results in the ranking values in Table 3-12 and the following percentages 49% [7] and

5 1% [q]. If the same process is repeated we get fluctuating results ranging between 49% and

51% for each variant. That is an average of 50% occurrence of each variant. Thus, the result was

influenced by adding the social constraint that militates against the use of [q] among female

speakers. In this example, two constraints militate against the use of [q] (i.e., *q and *F[q]) and

two militate against the use of [7] (i.e., *7 and IDENT-IO (RTR)), yielding 50-50% choice

between the two variants. This example shows that the inclusion of just one social constraint,

*F [q], can have a determining effect on the results. It leads not only to changing the ranking

values of the constraints, but also to different percentages of the occurrence of each variant. The

50-50% choice of each variant may or may not match real life occurences, but it reflects on what

the inclusion of social constraints could do. In real life, it is most likely that more than one social

constraint is involved. I will elaborate on this point as we move on with the discussion.

Table 3-12. Ranking values and disharmonies of *7, *q, *F[q], and IDENT-IO (RTR), without
manipulation of the values of the pair distribution
Constraint Ranking value Disharmony
*q 100.192 99.092
*F[q] 100.192 100.635
'F? 99.808 99.939
IDENT-IO (RTR) 99.808 96.564


Example 3: Having observed the influence that one social constraint can exert on the

ranking values and the degree of optionality in the choice between two variants, it would be of









interest to observe what results would look like if we manipulated the values of the pair

distribution. Conversely, we take the real life percentages of each of [q] and [7] in the speech of

female speakers in this study, which are 35% and 65% respectively, and put these percentages as

pair distribution values of [q] and [7] respectively in the computation. In this case, learning from


partial outputs in which [7] occurs 65% as a result of the *F [q] constraint, and the learner hears

these outputs as inputs, the ranking values of the four constraints in Table 3-12 will be after

100000 learning trials as in Table 3-13. The output distribution will be as follows: 66.5% [7] and

33.5% [q], which is very close to the real life percentages. The slight difference in results is

expected because of the learning process and optionality that is held here as a core feature of

variation. Learning from partial outputs that has a higher percentage of [7] may result in higher


learning percentage of the [7] sound. Thus, our surroundings and the people we interact with may


determine our grammar because they determine the percentage of the input of each of [7] and [q].

Further, this alone does not explain why some people may have similar attributes and similar

interactions, yet one switches between the two forms with great success according to the

interlocutor, while one may oscillate between the two forms with the same interlocutor. This is

why I believe that social constraints play a maj or role in determining our grammars. It depends

on what social constraints are activated at a certain time and place in addition to the input intake.

This will be further examined by the addition of further social constraints to see how results shift

as we add social constraints of various types.

Including the right social constraints and manipulating the pair distribution of [q] and [7] to

arrive at the observed variation in the data is important to model the various speakers' or groups

of speakers' different grammars. Following this process determines that, for example, the group









of female speakers gets an input of such percent of [q] and such percent of [7] from our data.

However, a group of female speakers could have outliers or females from different age groups

that are influenced by different social factors. Once those social constraints are determined and

included in the GLA grammar, one should be able to model individual variation as well as

smaller group variation (e.g., younger female group or older female group). In other words, the

results become closer to real life variation, when more specification or detail is offered and this

usually applies to smaller groups and individuals (i.e., intraspeaker variation). Sometimes we

could expect results that are comparable with real life production without manipulating the pair

distribution of [q] and [7] if the number of constraints that favor one sound over another

correspondes with the output percentages of real life. This can be achieved by determining the

social constraints involved and adding them to the computation. For example, having the real life

percentages 85% of [q] and 15% of [7] and five constraints favoring [q] and only one constraint

favoring [7] will lead to similar percentages without manipulating the pair distribution (Example

5). In order to explain this point, we explore a further example that deals with the female group

from the older generation. This age group according to the data should be influenced by the

constraint *OLD[?] ([7] is marked socially in the speech of older generation) because most older

female speakers use more [q] than [7] in their speech.

Table 3-13. Ranking values and disharmonies of *7, *q, *F[q], and IDENT-IO (RTR),
manipulating the values of the pair distribution to 65% [7] and 3 5% [q]
Constraint Ranking value Disharmony
*q 100.457 104.274
*F[q] 100.457 100.407
'F? 99.543 99.154
IDENT-IO (RTR) 99.543 94.586









Example 4: With the addition of the constraint *OLD[?], the results are also shifted

towards 60% use of [q] and 40% use of [7] when learning from partial outputs. In this example, it

seems that the number of constraints that support the use of [7] are three out of five constraints


(*7, *OLD[?], and IDENT-IO (RTR)), which correlates with the 60% use of [q] 20% for each

constraint in support of [q]. Only two constraints out of five constraints support the use of [7]


(*q, *F[q]), which correlate with the 40% use of [7]. These results that we observe without

manipulating the pair distribution of a grammar whose value of constraints is set to 100 are very

similar to those observed in my data. Calculating the percentages of the use of [q] and [7] in the


group of older female speakers, I got 63% use of [q] and 37% use of [7]. This kind of result

shows that the GLA can, when the right number of constraints is involved, model real-life

sociolinguistic variation without even manipulating the pair distribution. We are arriving at

results that approximate real life occurrences from just adding the respective pertinent social

constraints to the computation. If we look also at the ranking values of constraints in Table 3-14,

we will find that they are very close to those if we manipulate the pair distribution by setting the

[7] output distribution to 37% and the [q] output distribution to 63% (Table 3-15).

Table 3-14. Ranking values and disharmonies of *7, *q, *F[q], *OLD[?], and IDENT-IO (RTR),
without manipulation of the values of the pair distribution
Constraint Ranking value Disharmony
'F? 100.031 98.742
OLD[?] 100.031 101.672
IDENT-IO (RTR) 100.031 96.666
*q 99.969 100.337
*F[q] 99.969 103.072









Table 3-15. Ranking values and disharmonies of *7, *q, *F[q], *OLD[?], and IDENT-IO (RTR),
manipulating the values of the pair distribution to 37% [7] and 63% [q]
Constraint Ranking value Disharmony
*7? 100.036 101.789
*OLD[?] 100.036 100.449
IDENT-IO (RTR) 100.036 98.743
*q 99.964 99.410
*F[q] 99.964 97.268


Example 4 is one case in which it is possible to model a speaker or a group of speakers'

grammar without manipulating the pair distribution because the number of constraints that favor

one sound over another corresponds with the real life occurences of that sound. Learning from

partial outputs without the manipulation of the pair distribution yielded percentages that are close

to those with the manipulation 39% of [7] and 61 % of [q] because the number of constraints

that favor [q] are three and those that favor [7] are two. In addition, if we repeat the 100000

learning trials from partial outputs a couple of times without manipulating the values of the pair

distribution, we will get results that range between 40-43% for [7] and 57-60% for [q]. This is

further expected because of the degree of optionality allowed and the instability of input intake

in real life.

Example 4 is a good illustration of how older female speakers model their surroundings;

the percentage of the occurrence of [q] and [7] in their speech corresponds with the input

percentages of the two variants that they are exposed to in their speech environment. The

addition of the two social constraints *OLD[?] and *F [q] to the computation has provided us with

the overlapping variable grammar of older female speakers, their input intake/percentages (i.e.,

their speech environment) and output percentages (i.e., their acutal use). The comparable results

to real life occurrences of Example 4 without manipulating the pair distribution are not entirely









possible when there is a great disparity in the output percentages (Examples 6 & 7). In such

cases, manipulation of the pair distribution gives a more exact ranking values of constraints (i.e.,

a more defined grammar for specific speakers or groups of speakers).

Example 5: The addition of a sixth constraint *AKR[7] ([7] is marked socially in the

speech of Akrama residents) will allow us to observe more interaction among social and

lingusitic constraints and the importance given to the weight or ranking values of social

constraints in comparison to some linguistic constraints. Learning from partial outputs after

manipulating the pair distribution of the grammar, assigning the output of [q] 82% and the

output of [7] 18%, results in the ranking values in Table 3-16 and the percentages 82% use of [q]

and 18% use of [7]. This swerve in real life percentages is due to the social constraints *AKR[?]

and *OLD[?], not so much to the social constraint *F[q] because this constraint militate against

the use of [q].

Table 3-16. Ranking values and disharmonies of *7, *q, *F[q], *OLD[?], *AKR[?], and IDENT-IO
(RTR), manipulating the values of the pair distribution to 18% [7] and 82% [q]
Constraint Ranking value Disharmony
*AKR[?] 100.492 102.851
IDENT-IO (RTR) 100.492 101.989
*OLD[?] 100.492 100.561
*7 100.492 99.992
*F[q] 99.508 100.202
*q 99.508 99.236


Learning from partial outputs without the above manipulation gives us the ranking values

in Table 3-17 and the following percentages 32% of [7] and 68% of [q]. Repetition of this


process could give a range 30.5-35% of [7] and 69.5-65% of [q]. These percentages result

obviously from the number of constraints that support [q], that is four in comparison with two









constraints supporting [7], which gives a fraction of two-to-one out of three respectively (i.e.,


approximately 67% of [q] and 33% of [7]). In other words [q] is chosen two thirds of the times

and [?] only one third of the times.

Table 3-17. Ranking values and disharmonies of *7, *q, *F[q], *OLD[?], *AKR[?], and IDENT-IO
(RTR), without manipulation of the values of the pair distribution
Constraint Ranking value Disharmony
*AKR[?] 100.041 102.167
IDENT-IO (RTR) 100.041 102.040
*OLD[?] 100.041 98.697
"? 100.041 100.371
*F[q] 99.959 94.916
*q 99.959 100.851


This result does not appear to match real life occurrences. But if we look closely at the

characteristics or social values of the Akrama residential area, we should reconsider the

constraints that are activated by old female speakers in that area. It seems that both old age and

area of residence require that the constraint *F[7] be activated instead of *F[q]. Apparently,


*AKR[?] and *OLD[?] work together along similar constraints for a certain age group that reside

in a certain area and require that they be accompanied by a constraint that is a member of the

same family of constraints, that is the *7 family or the family of constraints that militate against

the use of [?] to which *F [7] belongs. In this case, we formulate a grammer that include the six

constraints in Table 3-18 and set the value of all of them to 100. Then, we go through the same

process of going to pair distribution, to strings, to distribution, selecting the OT grammar and the

output distribution, and finally learn from partial outputs. We will arrive at the ranking values in

Table 3-18 and at results that approximate real life occurrences: 85% use of [q] and 15% use of









[7]. These results match real life occurrences without manipulating the pair distribution because

there are five constraints that favor [q] and only one constraint that favors [7].

Table 3-18. Ranking values and disharmonies of *7, *q, *F[7], *OLD[?], *AKR[?], and IDENT-IO
(RTR), without manipulation of the values of the pair distribution
Constraint Ranking value Disharmony
*AKR[?] 100.049 102.470
IDENT-IO (RTR) 100.049 100.065
*OLD[?] 100.049 103.691
"? 100.049 100.073
*F[7] 100.049 101.517
*q 99.951 100.793


The ranking values observed in Table 3-18 and the percentages that resulted from the non-

manipulation of the pair distribution are comparable to those in Table 3-19 and the percentages if

we manipulate the pair distribution: 84% [q] and 16% [7]. This is another case in which it is

possible to model real life variation without manipulating the values of the output pair

distribution because the number of constraints that favor [q] and [7] correspond with their real

life percentages. It is enough to know the social constraints that are involved in order to achieve

results comparable with the variation we observe in real life.

Table 3-19. Ranking values and disharmonies of *7, *q, *F[7], *OLD[?], *AKR[?], and IDENT-IO
(RTR), manipulating the values of the pair distribution tol8% [7] and 82% [q]
Constraint Ranking value Disharmony
IDENT-IO (RTR) 100.008 101.412
*AKR[?] 100.008 99.437
*OLD[?] 100.008 98.799
*F[7] 100.008 97.505
"? 100.008 96.376
*q 99.992 95.583










Thus, like SG, the GLA can easily model sociolinguistic variation. It can give us

probabilities of variation that could reflect real life variation based on incorporating the right

type and number of social constraints. However, it exceeds SG in that it is capable of modeling

any speaker' s or group's variable grammar through manipulation of the output pair distribution.

The GLA dispenses with trying all possible rankings of constraints to arrive at the frequencies of

each variant. Even all possible rankings of constraints may not reflect the exact real life

percentages of variants and do not provide explanation of why such inter- and intra-speaker

sociolinguistic variation exists. I have also argued that the inclusion of social constraints in the

GLA, in addition to the manipulation of pair distribution, is essential to explain linguistic

differences among speakers from the same social group or different social groups and to model

sociolinguistic variation in formal theory. In addition, certain social constraints could give us

expectations about what other speakers' speech should sound like. In Example 5, the

constraints*F[7], *OLD[?], and *AKR[7] make us expect limited use of [7] in the speech of older

female speakers from Akrama. What is expected from these constraints is what we observe in the

speech of older female speakers from Akrama in real life. In this sense, listeners may also be

able to discover the grammar used by a certain speaker or group of speakers and the social

constraints activated by those speakers at a certain time and place, as some interviewees'

comments indicate. That is, they observe variation in the speech of their relatives based on the

interlocutor. These interviewees point out their awareness of such shift in pronunciation, when it

takes place, and with whom it takes place. This awareness may be an indication of their

knowledge of the various social constraints and various grammars that exist in their community

and that result from different activations of different social constraints. In support of this

argument, I will explore one further example.










Example 6: Similarly, if we only replace *F [7] with the constraint *M[7] ([7] is marked

socially in the speech of male speakers) and learn from partial outputs, the percentages of each of

[q] and [7] will be after 100000 learning trials: 84% and 16% respectively. These percentages are

also close to real life occurences. In our case, [q] occurs 96% in the speech of older male

speakers from Akrama. This precentage may be lower and closer to 84% if the sample is taken

from more than three speakers, which is the number of older male speakers from Akrama that are

part of my original data. It seems that the following social constraints *AKR[7], *M[7] and

*OLD[?] play a maj or role in determining the results and the precentages of each of [7] and [q].

They make us expect a low percentage of [?] and this is the case in older male speakers from

Akrama. Since the percentage of [7] in older male speakers from Akrama is lower than that of

older female speakers from Akrama and the only constraint that differs between the two social

groups is *M[7] instead of *F[7], we can conclude that *M[7] gives lower percentages of [7] in

the speech of male speakers than does *F [7] in the speech of female speakers. In other words,


*M[7] has a higher ranking value than *F[7] in all speakers. It makes speakers and listeners

expect a difference between the speech of males and females (i.e., less [7] and more [q] in the

speech of males than in the speech of females).

If we change the percentages of the pair distribution we get from the OT grammar that

include the six constraints in Table 3-20 and that resulted in 16% [7] and 84% [q] to 96% [q] and


4% [7], we will get from learning from partial outputs the ranking values in Table 3-20 and the

approximate percentages: 94% use of [q] and 6% use of [7]. Table 3-20 shows that *M[7] has a









slightly higher ranking value than that of *F [7] in Table 3-19. This last manipulation of the pair

distribution of [q] and [7] presents us not only with the grammar of these three male speakers

from Akrama but also with a model of their input environment. It provides us with information

on their social networks and surroundings as well as the type of interaction they are exposed to.

In other words, these three speakers have 96% input of [q] and thus more interaction with people

that use this sound; the same is true of [7]: 4% input of [7]. Thus, manipulation of the pair

distribution enables us to model speakers' input intake and the characteristics of the speech of

their social networks, surroundings, or area of residence or workplace.

Table 3-20. Ranking values and disharmonies of *7, *q, *M[7], *OLD[?], *AKR[?], and IDENT-
IO (RTR), manipulating of the values of the pair distribution to 4% [7] and 96% [q]
Constraint Ranking value Disharmony
*AKR[?] 100.878 103.730
IDENT-IO (RTR) 100. 878 97.323
*OLD[?] 100. 878 100.783
*M[7] 100. 878 98.379
"? 100. 878 101.061
*q 99.122 99.309


Example 7: Let us take another case in which young female speakers use 98% [7] and 2%

[q] in real life data. We will assume six constraints stated in Table 3-21, which should enable

female speakers from getting a high percentage of pronunciation of [7]. Most of the young

female speakers are occupants of Al-Hameeddieh area; hence, the constraint *HAM[q] ([q] is

marked socially in the speech of Al-Hameeddieh residents) is used. *YOUNG [q] and *F[q] are

used because we are dealing with young female speakers who rarely use the [q] sound. Using the

RCA input and learning from partial outputs before adjusting the probabilities of the occurrence

of [q] and [7] in the pair distribution, the approximate use percentages are 67% [7] and 33% [q].









This is a case in which four constraints support the use of [7] (i.e., *q, *F[q], *YOUNG[q], and

*HAM[q]) and two constraints support the use of [q] (i.e., *q and IDENT-IO (RTR)), forming a

fraction of two-to-one out of three, which is approximately 67% to 33%. These percentages are

quite disparate from the real life occurrences of young female speakers from Al-Hameeddieh.

This disparity in percentages could be indicative of two things.

First, social constraints should have higher weight or ranking values in order to match

reality. Once we change the probabilities of the pair distribution to match the occurrence of each

of the [q] and [7] in real life (i.e., 98% and 2% respectively), we get the ranking values in Table

3-21 and the precentages 98.5% use of [7] and 1.5% use of [q] after 100000 trials of learning

from partial outputs. This is further evidence that social constraints play a maj or role in

determining the winner and in influencing our grammar and the ranking values that each speaker

assigns to certain constraints. It is also possible to change the ranking values of constraints in the

defined grammar. This change could give a similar result to the observed sociolinguistic

variation among young female speakers from Al-Hameeddieh. We can, for example, set the

value of the four constraints *q, *F[q], *YOUNG[q], and *HAM[q] to 100 and the value of the

other two constraints to 95 or less. Giving the constraints these values results in approximately

99% use of [7] and 1% use of [q].

Table 3-21. RCA input: Ranking values and disharmonies of *7, *q, *F[q], *YOUNG[q],
*HAM[q], and IDENT-IO (RTR), with manipulation of the values of pair distribution
to 98% [7] and 2% [q]
Constraint Ranking value Disharmony
*HAM[q] 101.846 101.573
*q 101.846 101.121
*YOUNG[q] 101.846 97.161
*F[q] 101.846 98.518
*7? 98.138 94.280
IDENT-IO (RTR) 98.138 97.922









Second, the input of the younger generation is the HCA form, not the RCA form.

Changing the input to /7/ instead of /q/ in the above computation yields without manipulation of


the pair distribution of [7] and [q] the approximate percentages 83% use of [7] and 17% use of


[q]. These percentages are the result of five constraints in support of the use of [7] (i.e., *q,

*F[q], *YOUNG[q], *HAMv[q], and IDENT-IO (RTR)) and only one constraint in support of the use

of [q], that is *7. Changing the young generation's input from /q/ to /7/ without manipulating the

pair distribution gives percentages that are closer to those of the real life occurrences of [q] and

[7] in the speech of young female speakers, although the resulting percentages are not the exact

percentages of those speakers. In the initial stage of Example 7, we got the percentages 67% use

of [7] and 33% use of [q] when the input was the RCA form, which are very different from the

real life occurrences of [q] and [7] in the speech of young female speakers from Al-Hameeddieh.

This raises issue of the ability of the younger generation to restructure their input. I believe that

not only do social constraints exert an effect on the younger generation's grammar but also that

the younger generation has succeeded in restructuring their input. Their input becomes the HCA

input instead of the RCA input. Conversely, the input should contain /7/ instead of /q/. Being

born in the city of Hims, although to a rural family that uses the overt form [q] at home, and

being exposed from an early age to the overt form [7] at school, the younger generation may

have been able to restructure their parents' input from listening to the maj ority of their

environment' s input (i.e., friends at school). This means the younger generation, perhaps, do not

have /q/ in their phonemic inventory any more; all they have is /7/ like HCA speakers and they

tend to neutralize in the same way.









Modeling the real life occurences of young female speakers from Al-Hameeddieh, whose

input is the HCA form, requires a lower ranking value for the constraint *7 than the other five

constraints. It is expected that this constraint starts with a very low ranking in the speech of

HCA speakers and those who are able to use the HCA form with 100% accuracy. This is exactly

what happens if we lower the ranking value of the constraint *7 to 90 or less; we get 100% use of


[7] and 0% use of [q]. In other words, *7 should have a lower ranking value than the other

constraints because it is violated constantly by HCA speakers, young female speakers from Al-

Hameeddieh, and all speakers who can use the [7] with 100% accuracy. We can get the exact

ranking values of the constraints (i.e., the grammar of young female speakers from Al-

Hameedieh) by manipulating the pair distribution of [q] and [7] to 2% and 98% respectively.

Table 3-22 shows the grammar of young female speakers from al-Hameeddieh and that the

constraint *7 has a lower ranking value than the rest of the constraints. There is approximately a

three-point difference.

Table 3-22. HCA input: Ranking values and disharmonies of *7, *q, *F[q], *YOUNG[q],
*HAM[q], and IDENT-IO (RTR), with manipulation of the values of pair distribution
to 98% [7] and 2% [q]
Constraint Ranking value Disharmony
*HAM[q] 101.382 104.275
*q 101.382 104.237
*F[q] 101.382 103.628
*YOUNG [q] 101.382 100.204
IDENT-IO (RTR) 101.382 99.495
'F? 98.618 97.191


Example 7 can be taken as evidence for the occurrence of restructuring of the input in the

younger generation as a result of the influence of social factors, such as social networks, and

learnability out of exposure to the new form at an early age. It can also be taken as evidence for










the dormancy of some linguistic constraints, such as *7, in some speakers and the activation of

other constraints instead. The example further shows the importance of social constraints in

comparison to some linguistics constraints in determine a speaker's grammar. It is worth noting

here that even in cases where the percentages are a little different from real life data, these

examples show that social constraints do influence our selection of a ranking value for a certain

linguistic constraint and do play a role in our choice of a grammar at a certain time or place.

These social constraints determine the approximate number of times each output occurs (i.e., the

percentage of the occurrence of each variant).

3.2.4 Concluding Remarks on the Sociolinguistic Variation between [q] and [?]

The above examples have shown that the manipulation of the pair distribution can model

the different grammars of different speakers or groups of speakers. To explain the differences

among those speakers or groups of speakers, it was essential to include social constraints in the

computation. Manipulating the pair distribution of the linguistic constraints alone does not

explain why speakers' speech differs from each other in a sociolinguistic variation situation or

why speakers from the same social group or different social groups have different grammars.

Furthermore, the above examples have shown that certain social constraints may give us

expectations about what a speaker' s or a group of speakers' speech should sound like. For

example, the bundling of the constraints *F[7], *Old[?], and *AKR[7] in Example 5 tells us to


expect very low percentage of [7] in the speech of older female speakers from Akrama but a little


higher than the percentage of [7] in the speech of older male speakers from Akrama. As a result,


the [7] sound is probably more marked among older male speakers from Akrama, leading to the


suggestion that the constraint *M[7] is probably ranked higher than *F[7] and is assigned a










higher ranking value by male and female speakers alike. Ranking constraints that are activated

by other speakers highlight the importance of the inclusion of social constraints in the

computation in the case of a sociolinguistic variation. Having the constraint *M[7] ranked higher

than other constraints in the grammar of all speakers means that the listener should expect what

the interlocutor' s speech should sound like based on whether the interlocutor is a rural male or

female migrant speaker to the cityl3. In the same way that an RCA speaker expects an HCA

speaker to produce [?] as a result of a different grammar, an RCA speaker may expect what an

older male migrant speaker should sound like and simultaneously the social constraints that

allow him to speak in a certain way or to make choices of his grammar. A listener who listens to

a speaker in two different setting using a different form in each of the settings is aware of that

speaker' s change of grammar and is most likely aware of the social contraints involved in that

switch according to the interlocutor.

In addition, the above examples present how each speaker or group of speakers can model

their environment. Examples 4, 5 and 7 show that female speakers model their environment and

so do male speakers in Example 6. It also seems from Examples 5 and 6 that female speakers are

capable of modeling their environment, the HCA setting, better than male speaker. Thus, it may

be possible to predict that males use the [q] sound more than females because of their activation



13 This does not exclude the possibility that males may be exposed more to the [q] sound than females are. In other
words, they associate more with people (particularly males) from their own background and maintain their [q] sound
to express solidarity. Such associations with people who use their own native sound expose males to higher output
(i.e., input for the learner) percentages of the [q] sound (i.e., the pair distribution from which they learn is inclined
more towards the [q] sound than it is in females). Learning from higher percentages of [q] leads to higher production
percentages of [q] because the surrounding linguistic environment of speakers influences their degree of learnability
of a form. In other words, males do not learn the new form, [?], as much as females do because they are not as
exposed to it as females are. Their production (i.e., learnability) is affected by the pair distribution of [q] and [7] in
their environment. Even in the case of different pair distribution of [q] and [7] for males and females, *M[7] will
have in the computation a slightly higher ranking value than *F[7] to match this difference (Compare Table 3-20 to
Table 3-19).










of the *M[7] constraint that is as aforementioned more fatal or has higher ranking among gender


constraints than *F [7]. Thus, social constraints within one social factor may have a variety of

strength and ranking values based on the social significance or markedness of a sound to a

particular gender, age group, residential area, social class, etc. In my data, *M[7] may be more

fatal than *F [7], but in other social settings or linguistic environment, *F [7] may be more fatal


because [7] is considered more marked among females than among males. This does not only

apply to the constraints proposed in this study, but could apply to other social constraints that are

pertinent to other varieties or languages. All what has been mentioned so far support the

psychological reality of the model and the constraints involved in the computation, particularly

when we speak of choice, expectations, and awareness of one' s grammar as well as the other' s

grammar.

3.3 Modeling Variation between [t] and [s] and between [d] and [z]

After discussing the variable use of [q] and [7] and their modeling in the GLA by

proposing social constraints and incorporating them in the computation process, I move our

attention to the variable use of [t] and [s] and [d] and [z] in the colloquial speech of HCA and

RCA speakers. I refer to this type of variation as stable because each variant occurs in particular

words of the RCA and HCA dialects; there is no interchangeability between the variants. This

type of variation is different from the variation observed between [q] and [7]. The latter can be

considered of the gradient type because it varies based on a number of social factors. The former

is more of the categorical type because it does not vary within a speaker, among speakers, or

among groups of speakers. The same variant is used by all speakers to pronounce certain words









(Chapter 5). In this sense, it could be referred to as a phenomenon in its current state rather than

variation. This current stable phenomenon may have resulted from variation historically.

The first two variants, [t] and [s], are pronounced [6] in SA and the latter two variants, [d]


and [z] are pronounced [a]. In other words, the SA variants do not exist anymore in HCA or

RCA speech except in very rare use of lexical borrowings from SA that contain those two

variants. The SA variants [6] and [a] merged completely with [t] and [d] respectively around the

14th century (Daher, 1998a, 1999). This was the first historical change in the history of the SA

variants [6] and [a] (Schmidt 1974: 94, cited in Daher 1999, 164). The second change took place

some time after the 14th century (Birkeland 1952; Schmidt 1986: 57; Schulz 1981:33; all cited in

Daher 1999); some [6] and [a] variants in borrowed words from SA were replaced in colloquial


speech with [s] and [z] respectively. According to Daher (1999), the sounds [6] and [a] might

only appear in the speech of some Damascene speakers who are highly educated and whose jobs

involve extensive use of the written language, SA. Daher tries to argue that the appearance of the

SA variants [6] and [a] in colloquial speech as [s] and [z] respectively is due to attempting to

imitate the SA variants but pronouncing them as [s] amd [z] instead (p. 164). According to

Daher, the use of [s] and [z] is more elevated than the use of [t] and [d] respectively, despite

being also dialectal variants. Thus, some speakers may produce [s] and [z] when aiming at

producing the SA variants [6] and [a] respectively. However, he discovered that the observed

variation between [t] and [s] and between [d] and [z] is "stable" (p. 180) with no indication of

any change in progress. Daher dealt with the variants of interest from a different perspective; he

treated them as ternary and binary variables. The ternary variables are [t] and [d] and the binary

variables are [s] and [z]. Ternary refers to the possibility of using the three forms in one word









interchangeably. For example, [t], [s], or [6] can occur in the word for 'snow': [tal3], [sal3], or


[6al3]. In this case, the underlying forms for all the variants would be /t/. It would be /d/ for

words that can be pronounced with either [d], [z], or [a]. On the other hand, the underlying form


for binary variables would be /s/ and /z/ because words can be pronounced with either [s] and [6]


or [z] and [a]. Daher's findings are different from the findings in this study because his primary

purpose was comparing females' and males' use of standard and colloquial variants. These

findings are based on his data that were probably elicited in a less natural way, in that people

used the [6] or the [s] sounds instead of [t] and the [a] and [z] sounds instead of [d] in words that

are normally pronounced in colloquial speech with [t] and [d] respectively. According to Daher

(1998: 2), the sociolinguistic interview could elicit spoken SA from educated speakers outside a

formal setting, such as television and radio broadcasting, public speeches, journalism, etc. This is

not the case in my data. All speakers spoke naturally using their dialectal variants all the time.

They used specifically [t] in some words and specifically [s] in other words in place of the SA

variant [6]. The same applies for the use of [d] and [z]; each is used specifically in certain words

instead of the SA variant [a]. For this reason, I call this kind of variation stable variation, and I

suggest that each variant has its own underlying form that is the surface form of each one of

them, which is found in the colloquial speech of RCA and HCA speakers. In this sense, /t/ is the

underlying form for [t]; /s/ for [s]; /d/ for [d]; and /z/ for [z]. I will also argue that the two SA

phonemes /6/ and /8/ are absent from the inventory of RCA and HCA speakers (Appendix F).

The difference between HCA and RCA, on the one hand, and SA, on the other hand, is

illustrated in Table 3-23.










Table 3-23. The variants of the variables (6) and (a) in SA, HCA, and RCA
Variable Variants in SA Variants in RCA Variants in HCA
(6) [6] [t]/[s] [t]/[s]
(a) [a] [d]/[z] [d]/[z]


To support my argument, I checked the linguistic environment for each variant and I

noticed that all four variants could occur in similar environments, and there is no natural class

that can be identified before or after any one of them. This tells us that the linguistic environment

does not play a role in determining which variant should be used and when. It is obvious from

the list of environments in Table 3-24 that each of the variants is a separate phoneme; it is the

underlying form for its surface form (i.e., the underlying form for [t] is /t/ and so on). The

environments that are written in blue in Table 3-24 are identical to the environment of the

counterpart of each variant; the ones written in pink coincide in all four variants.

Examination of the linguistic environments justify my earlier suggestion that /6/ and /8/ are

not part of the RCA and HCA inventory of phonemes and that each of the variants represents the

basic allophone (i.e., the phoneme or the underlying form). It seems that there is a fixed and

stable division among words that have historically come to be pronounced with one of the four

variants: [t] and [s] and [d] and [z]. For example, the words [ktiir] 'a lot/many/much/plenty',

[masalan] 'for example', [haadaa] 'this (M)', and [7izaa] 'if' are always produced with the

sounds [t] and [s] and [d] and [z] respectively by all speakers. [t] and [s] are not interchangeable

in [ktiir] and [masalan]14, and so is the case with [d] and [z] in [haadaa] and [7izaa].




14 This is the HCA form that is pronounced with [an] word-finally. Some speakers use [masalen] instead because of
what is called in Arabic Pi'mala, i.e. the change of [a] to [e]. It is a feature of the RCA dialect, a feature that also
undergoes change in the speech of those who acquire the HCA dialect completely. However, this is beyond the
scope of this study and will be left for further research.










Table 3-24. List of environments for [t] and [s] and [d] and [z]
[t] [s] [d] [z]
[k i] [a -a] [i e] [i a]
[k a] [a i] [a a] [# a]
[i t] [? i] [a u] [a a]
[# 1] [u a] [x V] [t i]
[a i] [# a] [i a] [i #]
[k o] [# u] [a #] [a #]
[a r] [o #] [i #] [a k]
[i r] [a #] [o #] [t -a]
[a a] [i #] [a i] [n u]
[e i] [t -a] [a o] [i n]
[a #] [d i] [u e] [9 a]
[i #] [i y] [i u] [i k]
[# i] [h #] [e a] [# i]
[# a] [u #] [a ?] [a b]
[# e] [9 #] [a q] [a i]
[# m] [9 a] [a t] [9 e]
[# n] [t i] [a n] [i i]
[# o] [b b] [a r] [i r]
[o #] [a n] [# a] [yd i]
[i ] [i -n] [u -a] [i -t]
[m a] [e #] [b d] [a n]
[# ?] [i m] [# r] [i b]
[# q] [r i] [i o] [a l]
[# y] [9 i] [i b] [k o]
[9 a] [i i] [i i] [9 i]
[a u] [a b] [a b] [x i]
[r o] [u #] [x u]
[i u] [# i] [r a]
[i b] [y b] [a h]
[i a] [u a] [i e]
[ u] [3- m]
[a m] [o o]
[a r] [u a]
[m i] [3 i]
[# -i] [ u]
[a q] [a r]
[n a] [k i]
[a b] [# e]
[d -o] [n -i]
[y i] [a e]
[b a]
[# ]
[n a]
[e -#]









3.3.1 Linguistic Constraints Pertinent to the Variation between [t] and [s] and [d] and [z]

Because of the fixed and stable variation that is observed in the data, I propose a number of

linguistic constraints that are required for the maintenance of the four variants in their respective

words. Since the SA variants [6] and [a] are not observed in the data, I assume that they are not


part of the phonemic system of speakers. In other words, the SA variables /6/ and /8/ are absent

from the speakers' phonemic inventory. The context-free markedness constraints militating

against these two sounds (i.e., *6 and *8) will be undominated in the grammar of speakers'

speech. What is present in speakers' phonemic repertoire are the observed variants in the data:

/t/, /s/, /d/, and /z/. For example, the SA words /6alaaea/ 'three' and /haaaaa/ 'this (M)' are

always pronounced in RCA and HCA as [tlaati] and [haadaa] respectively. Since this applies to

all speakers, there is no reason to believe there is an underlying form that is different from the

surface form. The existence of all four variants (/t/, /s/, /d/, and /z/) in the speech of speakers

impels us to suggest four context-free markedness constraints that militate against each of the

four variants in attempt to eliminate it from or prevent it from occurring in certain words. These

four context-free markedness constraints are:

(12) *t Voiceless alveolar stop is marked

(13) *s Voiceless alveolar fricative is marked

(14) *d Voiced alveolar stop is marked

(15) *z Voiced alveolar fricative is marked

A switch from the use of [t] to the use of [s], for example, or vice versa involves a manner

of articulation change: a change from a stop to a fricative or from a fricative to a stop, and so is

the case with respect to switching between [d] and [z]. Hence, we need the following faithfulness

constraints to maintain the friction and stop features in the input:










(16) IDENT-IO (Fric) Let a be a segment in the input, and p a correspondent of a
in the output. If a is [yFric], then P is [yFric]

(17) IDENT-IO (Stop) Let a be a segment in the input, and p a correspondent of a
in the output. If a is [yStop], then P is [yStop]

Any change from /t/ into [s], /s/ into [t], /d/ into [z], and /z/ into [d] will result in violation of one

of the two constraints in (16) and (17). Since both constraints in (16) and (17) militate against the

change of some manner of articulation, we can combine them in one IDENT-IO constraint for

manner:

(18) IDENT-IO (Manner) Let a be a segment in the input, and P a correspondent of a in the
output. If a is [yManner], then P is [yManner]

The constraint in (18) is sufficient to prevent any change in the manner of articulation and

guarantees that all input forms will have equivalent corresponding output forms. In other words,

the faithfulness constraint must be higher ranked than the four markedness constraints to

preserve the input in the output. The ranking among the four markedness constraints is irrelevant

because they are all violable since all fours sounds are possible in natural speech.

3.3.2 Faithfulness and the Variants [t] and [s] and [d] and [z]

Here, I will follow the same process that I followed in modeling the variable use of [q] and

[7]. I define a number of grammars. The ranking value of constraints will be set at 100 unless

specified otherwise. Each grammar will be forwarded to pair distribution, then to strings and to

distributions. We learn from partial outputs by selecting the output string or the output

distribution with the defined OT grammar. The number of learning trials is 100,000 unless

specified otherwise with 0.2 noise and 0.1 plasticity. The underlying form is assumed to be the

surface form of the speakers, that is /ktiir/ for [ktiir] and /masalan/ for [masalan], for example. I

will start examining the variable use of [t] and [s], using the two mentioned underlying forms as










inputs and two possible outputs for each input: [ktiir] and [ksiir] and [masalan] and [matalan]

(the latter of each pair of words does not occur in speech) respectively.

Example 8: Without manipulation of the pair distribution or output distribution, we will

get varied results of 15% of [ksiir], 36% [ktiir], 33% of [masalan], and 16% of [matalan] and the

ranking values in Table 3-25.

Table 3-25. Ranking values and disharmonies of *s, *t, and IDENT-IO (Manner), without
manipulation of the values of the pair distribution
Constraint Ranking value Disharmony
IDENT-IO (Manner) 100.107 102.446
*s 100.042 99.465
*t 99.958 98.453


Although we get about 50% use of each word, we do not get results that match real life

occurrences. That is, in real life, [matalan] and [ksiir] are not used, but they still appear as

winners in the above percentages. To avoid this contradiction with real life, we need either to

change the ranking values of constraints, allowing for a definitive higher ranking of IDENT-IO

(Manner) over the other two constraints. For example, we can set the ranking value of IDENT-IO

(Manner) at 100 and give the other two constraints a ranking value of 90 or less. In this way, we

get after learning from partial outputs 50% use of [ktiir] and 50% use of [masalan], which in

actuality means 100% use of [ktiir] and 100% use of [masalan] if we had only one representative

tableau of one of the inputs rather than both tableaus in the grammar. We also get the ranking

values in Table 3-26 and Tableaus 3-27 and 3-28.

Table 3-26. Ranking values and disharmonies of *s, *t, and IDENT-IO (Manner) when ranked at
90, 90, 100 respectively
Constraint Ranking value Disharmony
IDENT-IO (Manner) 103.161 103.616
*t 91.193 89.535
*s 88.807 89.521











Tableau 3-27. Ranking of *s, *t, and IDENT-IO (anner): Underlyn form /ktiir/
/ktiir/ IDENT-IO (Mner) *t *s
Sa. [ktiir] *
b. [ksiir] *! *

Tableau 3-28. Ranking of *s, *t, and IDENT-IO (anner): Underlyi form /masalan/
/masalan/ IDENT-IO (Manner) *t *s
Sa. [masalan]
b. [matalan] *! *


The second option besides changing the ranking value of constraints is that we can

manipulate the pair distribution and we will still get similar results to those in Table 3-26 and

similar percentages. If we change the pair distribution to 100% use of [ktiir] and 100% use of

[masalan], and we learn from partial outputs, we get 50% use of each word (i.e., 100% actual use

of each) and the ranking values in Table 3-29.

Table 3-29. Ranking values and disharmonies of *s, *t, and IDENT-IO (Manner), manipulating
the pair distribution to 100% use of [ktiir] and 100% use of [masalan]
Constraint Ranking value Disharmony
IDENT-IO (Manner) 112.132 112.332
*t 100.315 101.213
*s 99.685 96.749


The large difference between the ranking value of IDENT-IO (Manner) and the other two

constraints apparent in Table 3-29 guarantees its higher ranking than the other two constraints: *t

and *s. This is indicative of the role that faithfulness plays in maintaining the input in the output.

Thus, one can conclude that in this type of variation, faithfulness constraints are the leading

factors for maintaining a stable variation. This point will be highlighted more as we examine the

effect of social constraints on this type of variation. Nonetheless, a small clarification may be

needed for why the 50% is equal to 100% in actuality. Having two tableaus in the grammar (one

with the input /masalan/; the other with the input /ktiir/) and a number of input strings of 1000









will result in equal number of input for each input: 500 inputs of /ktiir/ and 500 inputs of

/masalan/. This further results in a 1000 output strings divided equally between [masalan] and

[ktiir]: 500 outputs of [masalan] and 500 outputs of [ktiir]. This explains the 50-50% occurrence

of each word. If we had the three constraints *t, *s, and IDENT-IO (Manner) and one tableau in

the grammar in which the input and the winning output has, for example, a t only, not an s, like

ktiir. In this case, all the input strings (i.e., 1000) will be /ktiir/ and all the output strings (i.e.,

1000) will be [ktiir], yielding 100% use of [ktiir]. The only difference in this case is that IDENT-

IO (Manner) and *s will have equal ranking or non-categorical ranking with regards to each

other. The difference between their ranking value and the ranking value of *t is around 10, which

is considered a high difference that will secure higher, categorical ranking between those two

constraints and *t. Having higher rankings than *t, violation of any one of the former two

constraints is fatal, whereas violation of *t is endurable as it is violated by the winner. Moreover,

learning from partial outputs in which [t] occurs 100% in some words and the learner hears these

outputs as inputs results in 100% use of [t] and the ranking values in Table 3-30 after 100000

learning trials.

Table 3-30. Ranking values and disharmonies of *s, *t, and IDENT-IO (Manner) with
manipulation of the pair distribution to 100% use of [ktiir]
Constraint Ranking value Disharmony
*s 106.018 106.817
IDENT-IO (Manner) 106.118 104.400
*t 93.982 92.073


The same applies to [masalan]; its occurrence 50% equals 100% in actuality if we had the

three constraints *t, *s, and IDENT-IO (Manner) and one tableau in which the input and the

winning output have an s only, not a t, such as masa~lan2. Without manipulation of the pair

distribution, we should get results that match a fraction of two-to-one because the [masalan]









output will be violating only one constraint out of three constraints; the suboptimal output,

[matalan] would be violating two constraints out of three. This should result in approximately

67% use of [masalan] and 33% use of [matalan], which is far apart from real life occurrences.

Therefore, in the case of this type of variation, it is essential to manipulate the pair distribution or

the ranking value of constraints to achieve results that match real life production. All that has

been discussed so far applies also to the variable use of [d] and [z]. There is no need to present

more examples for the latter pair of variants because they behave exactly like [t] and [s]

concerning their occurrence in speech.

After showing the importance of faithfulness constraints in determining the winner with

regards to the variants [t] and [s] and [d] and [z], it is important to test the influence of social

constraints on this type of variation. Do social constraints play the same role observed in the

variable use of [q] and [7]? In the next section, I will intersect some social constraints with the

linguistic constraints proposed in this section to see if they also play a role in determining our

grammar at a certain time and a place regarding the variants [t] and [s] and [d] and [z]. However,

a slight change should be done to the constraints to accommodate the variants [t] and [s] and [d]

and [z].

3.3.3 Social Constraints Pertinent to the Variants [t] and [s] and [d] and [z]

The social constraints pertinent to the variants [t] and [s] and [d] and [z] are factored in the

same way we factored the social constraints for [q] and [7]. There are 32 social constraints; eight

constraints for each social factor because there are four variants and each one is marked with

respect to two categories within one social factor. Hence, each social factor should be factored

into eight social constraints as follows:

Gender: *F[t] [t] is marked socially in the speech of females









*F [s]

*M[t]

*M[s]

*F[d]

*F [z]

*M[d]

*M[z]

*OLD [t]

*OLD [s]

*YOUNG[t]

*YOUNG[s]

*OLD[d]

*OLD [z]

*YOUNG[d]

*YOUNG [z]

*AKR[t]

*AKR[s]

*HAM[t]

*HAM[s]

*AKR[d]

*AKR[z]

*HAM[d]

*HAM[z]


[s] is marked socially in the speech of females

[t] is marked socially in the speech of males

[s] is marked socially in the speech of males

[d] is marked socially in the speech of females

[z] is marked socially in the speech of females

[d] is marked socially in the speech of males

[z] is marked socially in the speech of males

[t] is marked socially in the speech of older generation

[s] is marked socially in the speech of older generation

[t] is marked socially in the speech of younger generation

[s] is marked socially in the speech of younger generation

[d] is marked socially in the speech of older generation

[z] is marked socially in the speech of older generation

[d] is marked socially in the speech of younger generation

[z] is marked socially in the speech of younger generation

[t] is marked socially in the speech of Akrama residents

[s] is marked socially in the speech of Akrama residents

[t] is marked socially in the speech of Al-Hameeddieh residents

[s] is marked socially in the speech of Al-Hameeddieh residents

[d] is marked socially in the speech of Akrama residents

[z] is marked socially in the speech of Akrama residents

[d] is marked socially in the speech of Al-Hameeddieh residents

[z] is marked socially in the speech of Al-Hameeddieh residents


Age:


Area:









Social class: *UMc[t] [t] is marked socially in the speech of upper-middle class

*UMc[s] [s] is marked socially in the speech of upper-middle class

*LMC[t] [t] is marked socially in the speech of lower-middle class

*LMC[s] [s] is marked socially in the speech of lower-middle class

*UMc[d] [d] is marked socially in the speech of upper-middle class

*UMc[z] [z] is marked socially in the speech of upper-middle class

*LMc[d] [d] is marked socially in the speech of lower-middle class

*LMC[z] [z] is marked socially in the speech of lower-middle class

3.3.4 Will Social Constraints Affect the Previous Results?

In this stable variation, it is not expected for social constraints to have much influence on

the variable use of [t] and [s] or [d] and [z]. All speakers use the same variant in the same word

regardless of age, gender, social class or area (e.g., /ktiir/ is pronounced [ktiir] by all speakers

and so on). To erase doubt, it may be useful to take an example in which social constraints are

involved. In this case, the social constraints involved should balance each other. That is, two

social constraints of the same category (e.g., female gender) and of two different marked usages

will be involved (e.g., *F[t] and *F[s]), since females use both [t] and [s] equally and the

variation is based on the word the variant occurs in, not on whether they are females or not.

Example 8: Following the same process in the GLA and without manipulation of the

ranking values of constraints, we get after 100000 learning trails from partial outputs the

percentages: 21% of [ksiir], 30% of [ktiir], 29% of [masalan], and 20% of [matalan] and the

ranking values in Table 3-31. Moreover, these results do not match real life occurrences because

[ksiir] and [matalan] do not occur in real life. The percentages are almost 50% for the total

outputs for each input, which seems at first sight accurate since 50% output for each input equals

100% in actuality. However, the problem lies in the three-to-two fraction of the use of [ktiir] and









[ksiir] and of [masalan] and [matalan] respectively. This is further due to the number of

constraints involved in the computation: three constraints in support of [ktiir] and [masalan] and

two constraints in support of [ksiir] and [matalan]. This is very informative; it informs us that we

should manipulate either the pair distribution or the ranking values to arrive at real life values of

variation. This manipulation will be explored in Example 9.

Table 3-31. Ranking values and disharmonies of *s, *t, IDENT-IO (Manner), *F[t], and *F[s]
without manipulation of the values of the pair distribution
Constraint Ranking value Disharmony
*F [s] 100.004 98.465
*s 100.004 95.193
*t 99.996 98.453
*F [t] 99.996 100.739
IDENT-IO (Manner) 99.656 97.660


Example 9: If we manipulate the pair distribution into 100% occurrence of each of [ktiir]

and [masalan], we get an almost 50-50% occurrence of each word after learning from partial

outputs, which is equal to 100% occurrence of each output in actuality. The ranking values of

constraints will be as in Table 3-32 after the 100000 learning trials from partial outputs.

Table 3-32. Ranking values and disharmonies of *s, *t, IDENT-IO (Manner), *F [t], and *F [s],
manipulating the values of the pair distribution to 100% use of both [ktiir] and
[masalan]
Constraint Ranking value Disharmony
IDENT-IO (Manner) 114.282 117.236
*s 101.666 103.724
*F[s] 101.666 101.099
*F [t] 98.334 100.931
*t 98.334 99.206


These results show again that the faithfulness constraint IDENT-IO (Manner) should have much

higher-ranking value than the other four constraints. Violation of this constraint is fatal. The

ranking of the other four constraints with respect to each other is irrelevant because the winners

violate at least two of them. From this example, we can conclude that social constraints are not









very important in determining our grammar with respect to the variable use of [t] and [s] and [d]

and [z]. What are important in this type of stable variation are faithfulness constraints; these

constraints play a maj or role in maintaining an equal corresponding output to the input.

3.3.2.1 Concluding remarks on the stable variation among [t] and [s] and [d] and [z]

In conclusion, Examples 8 and 9 show that the inclusion and exclusion of social constraints

regarding the use of the four sounds [t] and [s] and [d] and [z] do not make a difference. Whether

the social constraints *F [s] and *F [t] are included or not, the results are the same (i.e., the

faithfulness constraint is higher ranked than the other constraints (social or linguistic) to preserve

the input in the output). Thus, IDENT-IO (Manner) plays a maj or role in determining the winners;

it must always have a higher-ranking value than the other markedness constraints to ensure the

observed stable variation between [t] and [s] as well as between [d] and [z]. However, this kind

of lexical conditioning could still be attributed to markedness because even when someone is

reading (which is the case of one speaker who was reading a document to his cousin during the

recording), the person uses [s] instead of [6]. This presents evidence that speakers prefer less

marked sounds and use them even in casual reading. In other words, speakers may find it easier

to access sounds in their native phonemic inventory and use them instead of those learned later

on in life (i.e., access /s/ and use its surface form [s] rather than use [6] that was learned through

formal education of SA as a second language).









CHAPTER 4
QUANTITATIVE ANALYSES OF [q] VS. [7]

4.1 Introduction

In this chapter, I present quantitative analyses for the observed variation in the use of [q]

and [7]. The purpose of these quantitative analyses is to examine associations and correlations

between both the dependent variants and the independent variables. First, I present tests of

association that could inform us on what determines social class in the city of Hims. Next, a

Bivariate test will be performed on the two dependent variants [q] and [7] to see if they are

correlated. To examine the main effects of and the interaction effects among the four

independent variables age, gender, residential area, and social class on the variable use of [q]

and [7], I apply a number of statistical methods. These statistical procedures will help understand

which of these four social factors plays a maj or or a minor role in the variation. Such

investigations may help in confirming the Eindings of the theoretical analysis in Chapter 3. A

number of statistical procedures are required for each variant as the following sections will

reveal. The chapter will conclude with a discussion of the Eindings of the statistical analyses,

trying to relate these Eindings to previous studies and the Eindings in Chapter 3. It will also

present implications of these Eindings.

4.2 Tests of Association of Social Class with Education, Income, Residential Area, and
Occupation

These tests are important because social class is a very hazy concept in the Arab World;

many factors may play a role in determining someone's class. In addition, the social mobility

existent in most societies affects a person's classification within one social class or another

(Haeri 1996; Laboy 2001). This is particularly true of Arab societies because of the great influx

of migration from rural areas to urban centers. Most urban centers are associated with more










prestige, better life style, and civilization. Consequently, a change from being a farmer to being a

white-collar worker influences a person's social classification. Also, being the son of a farmer

and becoming a medical doctor affects that person's social status. Thus, implementing these

association tests may give an answer to what factors contribute to a person's social status not

only in the city of Hims, but also in other Syrian cities and urban centers in other parts of the

Arab World. In addition, if social class emerges as a significant factor in the statistical analysis

of the influence of the independent variables on the variable use of [q] and [7], the associated

factors will by inheritance have significant effect and vice versa.

In Table 2-1 in Chapter 2, I assigned social class to participants based on the community's

general evaluation of them as somewhat rich or somewhat poor. Rich in the Syrian sense does

not mean extremely rich as it is in the English sense of the word. Rich could refer to someone

who lives a comfortable lifestyle, equivalent to what would be termed upper-middle class in the

West. Poor does not refer to those who are in need or cannot afford to live. It refers to those who

can support themselves, but cannot lead the same comfortable lifestyle that the rich or upper-

middle class leads. Poor in its Syrian sense would be equivalent to what is termed lower-middle

class in the West. Given the hazy definition of social class in the Arab world, the lack of studies

that refer to class division, and the possibility of social mobility because of migration from rural

to urban areas and obtaining higher academic degrees, I decided to examine my assignment of

social class to participants based on the general evaluation of the community to them against four

possible social indicators. These social indicators are family income (mainly breadwinner

income), education, occupation and residential area. Such social indicators are assumed by many

studies to play a role in social class assignment (e.g., Labov 1972a; Eckert 1991b).










In order to see the degree of association among each of these four social indicators and

social class, I use contingency tables that include Chi-squared tests. These tests are appropriate to

measure the strength of association between a qualitative variable and other qualitative variables

(i.e., categorical variables) (Agresti and Finlay 1997). In our case the response variable is social

class that has two categories: lower middle and upper middle. The predictor variables are

income, education, occupation and residential area. Income is divided into three categories: low,

mid, and high. Education is divided into six categories: elementary school, middle school, high

school, associate degree, bachelor' s degree, and professional. The latter includes medical

doctors, dentists, and those seeking master's or doctoral degrees. Occupation is divided into four

categories: unemployed, government employee, private j ob/business, and professional. The latter

includes teachers, engineers, medical doctors, and dentists. Each of the categorical variables is

tested independently against social class because the Chi-squared test allows for testing only one

categorical variable against another categorical variable. In our case, it is testing the effect of

each of the four categorical variables on social class classification.

4.2.1 Test of Association between Social Class and Education

The Chi-squared test shows lack of association between education and social class

(Appendix A). The value of the Pearson Chi-squared test is 6.21 and the p-value is 0.29, which is

higher than 0.05, the probability value adopted throughout the study to rej ect the null hypothesis.

Rej ecting the null hypothesis is an indication of association. Since the p-value for education is

higher that 0.05, the null hypothesis holds. That is, there is no association between social class

and education. If we look at the bar chart in Figure 4-1, we will see that the six educational

groups have slight differences between the two social classes. Only the Associate degree group

shows great disparity between lower- and upper-middle classes. Although there are more











professionals and bachelor' s degrees in the upper-middle class category, the difference is slight

in that it does not reflect an association between education and social class.




10] MElementary
Middle school
O High school
Associate degree
OBachelor's degree
8- M Professional
















Lower-rniddle Upper-middle
SocClass



Figure 4-1. Bar chart for the distribution of the six educational groups between lower-middle and
upper-middle classes

4.2.2 Test of Association between Social Class and Income

The Chi-squared test shows the existence of association between income and social class


(Appendix A). The value of the Pearson Chi-squared test is 3 5.2 and the p-value is 0.000. The


very low p-value indicates strong association between social class and income. The bar chart, in


Figure 4-2, shows that low income is associated with lower-middle class, whereas high income is

associated with upper-middle class. In fact, there is not a high-income person included among


lower-middle class. The mid income is divided equally between the two social classes.












251 1 Low
II Mid
O H-igh




















Lower-rniddle Upper-middle
SocClass


Figure 4-2. Bar chart for the distribution of the three income groups between lower-middle and
upper-middle classes

4.2.3 Test of Association between Social Class and Occupation

The Chi-squared test shows the existence of association between occupation and social

class (Appendix A). The value of the Pearson Chi-squared test is 9.578 and the p-value is 0.023.

The low p-value indicates association between social class and occupation. The bar chart, in

Figure 4-3, shows that the occupational category government employee shows significant

difference in distribution between the two social classes. More government employees are in the

lower-middle class category. The other three occupational categories do not show significant

difference of distribution between the two social classes.
















12.1 Occupation
IIunemployed
SGovernmert employee
OJPrivate work/busniess
10- I Professional

















Lower middle Upper-middle
SocClass




Figure 4-3. Bar chart for the distribution of the four occupational groups between lower-middle
and upper-middle classes

4.2.4 Test of Association between Social Class and Residential Area

The Chi-squared test shows the existence of association between residential area and social

class (Appendix A). The value of the Pearson Chi-squared test is 10.3 17 and the p-value is 0.001.

The very low p-value indicates strong association between social class and residential area. The

bar chart, in Figure 4-4, shows that much more upper-middle class people live in Al-Hameedieh

than in Akrama. On the other hand, the number of lower-middle class people living in Al-

Hameeddieh and Akrama is not much different.











25_ Area
I Al-Hameeddieh
Akrama






















Lower-mriddle Upper middle
SocClass

Figure 4-4. Bar chart for the distribution of the lower-middle and upper-middle classes in the two
residential areas: Akrama and Al-Hameeddieh

4.2.5 Concluding Remarks on the Tests of Association between Social Class and Education,
Income, Occupation, and Residential Area

The Chi-squared tests showed lack of association between social class and education. On

the other hand, they showed association between social class and income, occupation, and

residential area. Nonetheless, the Chi-squared values are not enough to tell us the strength of

association between social class and the other variables. It is true that the larger the Chi-squared

value, the stronger the association. However, this value could be affected by the sample size; the

larger the sample size, the larger this value is. In order to eliminate doubt, we look at the adjusted

residual values. If the adjusted residual absolute value is higher than two, there is strong

association. Also, the higher the absolute value than two, the stronger is the association. Since

only income, occupation, and residential area showed association, we look at their adjusted









residuals (Appendix A) to see which one of them has the strongest association. Income has an

adjusted residual absolute value of 5.4 for low income; 5.2 for high income; and only 0.6 for mid

income. Occupation has an adjusted residual absolute value of 3 for government employee; 0.7

for unemployed; 1.2 and 1.3 for business/private job and professional respectively. Residential

area has an adjusted residual absolute value of 3.2 for each of the residential areas. In this sense,

income shows the strongest association followed by residential area. Occupation has the smallest

association; there is only strong association with the government employee category. This fact

was observed in the bar chart in Figure 4-3.

These results are confirmed by the Pearson correlation coefficient, which has the symbol

R. The value of R is usually between -1 and 1; the closer the value to -1 or 1, the stronger the

association. If the value is close to 0, the association is weak. The value of R for education is 0.2,

indicating lack of association. The value of R for income is 0.82, indicating strong association.

The value ofR for occupation is 0. 1, indicating lack of association. This last result contradicts

the Chi-squared test result for occupation. Hence, one should consider other statistics in addition

to the Chi-squared value and p-value to determine the strength of association. The value of R for

residential area is 0.45, indicating moderate association. The R values for the different

categorical variables investigated for strength of association with social class correlate with the

adjusted residuals results. According to both the R values and the adjusted residual values,

income has the strongest association, followed by residential area. Occupation and education

have no association with social class according to the R values; occupation only shows

significant association with respect to the government employee category according to the

adjusted residual values.









In conclusion, one can assume that social class is mainly determined by income and area of

residence in urban centers in the Arab World. It is slightly affected by occupation. Education

does not seem to be a good determining factor of social class. These findings may be helpful for

other researchers who tend to study the influence of social class on linguistic variation in the

Arab World. They could be a starting point for further investigation in urban center, not only in

Syria, but also in other Arab countries.

4.3. Speakers' Distribution of [q] and [?]

Before applying statistical procedures to the data to see the effect of the four social factors

on the variable use of [q] and [7], I present the distribution of [q] and [7] in the speech of each

speaker in Table 4-1. Table 4-1 shows the number of occurrence of each sound from the total

number of sounds for each speaker under the columns [q] and [7]. Next to each of these columns,

the percentage of the occurrence of each sound is calculated and presented under the percentage

columns. The speakers' distribution regarding the four factors tested for effect on the observed

variation is also included under separate columns for each factor. Table 4-1 shows that the total

number of tokens for [q] and [7] is 11,548. 5874 tokens of [q], which constitute 5 1% of the total


number of tokens; 5674 tokens of [7], which constitute 49% of the total number of tokens.

I entered the information included in Table 4-1 into SPSS to examine the effect of age,

gender, residential area, and social class on the variable use of [q] and [7]. I created two separate

SPSS data files to perform different tests. I entered social factors in the same way in both data

files. Each of the social factors contains two categories and is entered as a nominal variable. Age

is assigned 1 for the younger age group and 2 for the older age group. Gender is assigned 1 for

females and 2 for males. Residential area is assigned 1 for Al-Hameeddieh and 2 for Akrama.










Table 4-1. Distribution of [q] and [7] in the speech of each speaker
Speaker Gender Age Social Area No. of % of No. of % of
class [q] [q] [?] [?]


77 LM
67 LM
64 LM
60 LM
70 LM
67 LM
64 LM
53 LM
70 UM
69 UM
62 UM
62 UM
64 UM
75 LM
61 LM
61 LM
61 LM
59 LM
56 LM
52 LM
53 LM
67 LM
58 LM
58 UM
57 UM
61 UM
58 UM
57 UM
31 LM
25 LM
35 LM
30 LM
23 LM
19 LM
24 UM
23 UM
24 UM
36 UM
27 UM
35 LM
28 LM


Akrama 222/232
Akrama 264/264
Al-Hameeddieh 467/470
Al-Hameeddieh 204/204
Al-Hameeddieh 80/102
Al-Hameeddieh 70/111
Al-Hameeddieh 122/122
Akrama 183/193
Al-Hameeddieh 79/129
Al-Hameeddieh 273/273
Al-Hameeddieh 286/305
Al-Hameeddieh 308/308
Al-Hameeddieh 205/205
Akrama 170/170
Akrama 278/278
Al-Hameeddieh 44/130
Al-Hameeddieh 0/154
Al-Hameeddieh 421/421
Al-Hameeddieh 56/131
Akrama 61/129
Al-Hameeddieh 7/94
Al-Hameeddieh 115/116
Al-Hameeddieh 44/65
Al-Hameeddieh 375/375
Al-Hameeddieh 163/218
Al-Hameeddieh 0/137
Al-Hameeddieh 361/361
Al-Hameeddieh 103/133
Akrama 239/271
Akrama 290/303
Akrama 254/255
Al-Hameeddieh 32/317
Akrama 2/120
Akrama 9/220
Al-Hameeddieh 2/294
Al-Hameeddieh 32/315
Al-Hameeddieh 2/181
Al-Hameeddieh 3/59
Al-Hameeddieh 6/215
Al-Hameeddieh 5/475
Al-Hameeddieh 11/143


96
100
99
100
78
63
100
95
61
100
94
100
100
100
100
34
0
100
43
47
8
99
68
100
75
0
100
77
88
96
100
10
2
4
1
10
1
5
3
1
8


10/232
0/264
3/470
0/204
22/102
41/111
0/122
10/193
50/129
0/273
19/305
0/308
0/208
0/170
0/278
86/130
154/154
0/421
75/131
68/129
87/94
1/116
21/65
0/375
55/218
137/137
0/361
30/133
32/271
13/303
1/255
285/317
118/120
211/220
292/294
284/315
179/181
56/59
209/215
470/475
132/143


4
0
1
0
22
37
0
5
39
0
6
0
0
0
0
66
100
0
57
53
92
1
32
0
25
100
0
23
12
4
0
90
98
96
99
90
99
95
97
99
92











Speaker Gender Age Social Area No. of [q] % of No. of [7] % of
class [q] [?]
42 F 24 LM Al-Hameeddieh 0/308 0 308/308 100
43 F 18 LM Akrama 0/65 0 65/65 100
44 F 29 LM Akrama 5/421 1 416/421 99
45 F 28 UM Al-Hameeddieh 1/124 1 123/124 99
46 F 33 UM Al-Hameeddieh 3/479 0 476/479 100
47 F 32 UM Al-Hameeddieh 0/114 0 114/114 100
48 F 28 UM Al-Hameeddieh 7/385 2 378/385 98
49 F 23 UM Al-Hameeddieh 6/118 5 112/118 95
50 F 25 UM Al-Hameeddieh 2/178 1 176/178 99
51 F 21 UM Al-Hameeddieh 2/127 2 125/127 98
52 F 26 UM Al-Hameeddieh 0/230 0 23 0/23 0 100
Total 5874/ 51% 5674/ 49%
11548 11548

Social class is assigned 1 for lower-middle class and 2 for upper-middle class. In addition, in

both files I entered [q] and [7] as two separate scale variables. However, I treated the sounds [q]


and [7] differently in each of the data files. I used the percentages of [q] and [7] for each speakers

in one file. I used this file in Section 4.3.1 to explore the statistical distribution of the data with

respect to each social group. The use of percentages provides more balanced comparison among

individuals and groups. I used the raw frequency of [q] and [7] for each speaker in the second

file. I used this file in the regression tests performed in Sections 4.3.2 and 4.3.3. These tests only

accept raw frequency data because they themselves calculate probabilities and percentages.

4.3.1 Exploring the Type of Statistical Distribution among the Various Groups

The choice of an appropriate statistical procedure for the data requires understanding the

type of statistical distribution of those data with regards to each social group. In other words, it is

essential to know if the distribution is normal or not before selecting a procedure. Normality or

non-normality is an important factor to arrive at accurate statistical results. In order to discover

the type of distribution in the data, we need to examine some descriptive statistics. We can run









the Explore option from Descriptive statistics in the Analyze menu of SPSS. This procedure

allows us to examine visually and numerically the distribution of each of the dependent variants

against each group. That is, the distribution with respect to gender, social class, residential area,

and age. This method also allows us to discover any outliers or extreme cases in the data. A

separate SPSS data file was created for the Explore procedure in which the percentages rather

than the raw frequencies of [q] and [7] in Table 4-1 are used. Social factors are entered in the

same way described in Section 4.3. I select the two dependent variants and place them in the

dependent list and the four independent variables in the factor list. I choose the statistics

Descriptives, Mean (M)-Estimates, and Outliers. Boxplots (factor levels together), histograms,

stem and leaf plots, and normality plots with tests are also selected. The results of the explore

procedure (Appendix B) show that for most groups the distribution is not normal. The stem and

leaf plots and the normality plots are not included in Appendix B due to space limits and because

histograms and boxplots are sufficient to show the data distribution.

The results of the Explore procedure show that the data is not normally distributed. There

is skewness either to the right, a positive type of skewness, or to the left, a negative type of

skewnes, in most cases (Appendix B contains tables, histograms, and boxplots for all the social

groups with regards to the use of each of [q] and [7]). If we look, for example, at the distribution

of the female use of [q] and [7], we observe the following. There is a great disparity between the

mean, trimmed mean, and median in the female use of [q]. They are respectively: 83.7, 86.49,

and 100. The skewness (-1.5) statistic provides evidence of disproportionate values at the

lower/left tail of the distribution. The degree of skewness may differ from one group to another.

However, the overall results show lack of normality in the distribution. The M-Estimators Table

shows that the results of all four tests (Huber' s M-Estimator, Tukey's Biweight, Hampel's M-











Estimator, and Andrews' Wave) for the female use of [q] are closer to the median than they are

to the mean. This further confirms that the data is not normally distributed. In addition, the Tests

of Normality Table shows significance (p = 0.000), which means we rej ect the null hypothesis

that the data is normally distributed. The histogram in Figure 4-5 shows this left tail skewness for

the females' use of [q]. The boxplots show that this skewness could be due to some outliers that

are marked with small circles or some extreme cases that are marked with asterisks.


Consequently, these results influence the statistical model used for examining the effect of the


social factors on the dependent variants, [q] and [7].







for Gender- Female
500.00.- .Normal
Mean =83.70
Std. Dev. =24.38
400.00-N =972


20000





100.



0.00 20 .00 40.00 60.00 80 .00 10( .00

qalf





Figure 4-5. Histogram showing the skewness to the left in the females' use of [q]

I will not discuss the distribution of all the groups. They are comparable to the distribution

of the female use of [q] with slight differences in the values of mean, trimmed mean, and

median. There are also different degrees of skewness, the farther the statistic from 0, whether it










is positive or negative, the larger the skewness and the less the normality. For example, the

skewness (-3.69) statistic for Akrama's speakers' usage of [q] provide evidence of high skewness

to the left. This is the result of four extreme cases as the boxplots show (Appendix B).

4.3.2 Bivariate Correlations Procedure between the Two Dependent Variants [q] and [?]

This section shows that the two variants [q] and [7] are highly correlated regarding their


variable usage. The Bivariate Correlations procedure is used to measure degree of association

between two scale variables. In our case, these two count variables are the two variants [q] and


[7]. I choose the Spearman's rho correlation coefficient because it does not require normal


distribution like the Pearson correlation coefficient. It is also not affected by outliers if there are

any in the data. Running this procedure in SPSS, we get the results in Table 4-2.

Table 4-2 shows that there is a strong negative correlation between the two variants as the

p-value (p < 0.01) indicates statistical significance. This correlation is not perfect, but is close

enough to -1 (R = 0.82). This type of negative correlation is referred to as inverse relationship

(Patten 2004: 117). That is, those who have high use of [q] have low use of [7] and vice versa.


Table 4-2. Bivariate correlations between the dependent scale variants [q] and [7]
Correlations

qaf hamza

Spearman's rho qaf Correlation Coefficient 1.000 -.820"

Sig. (2-tailed) .0
4 52 52

hamza Correlation Coefficient -.820" 1.000

Sig. (2-tailed) .0
4 52 52

**. Correlation is significant at the 0.01 level (2-tailed).









4.3.3 Generalized Linear Models (GZLM) Procedure for the Effect of the Four
Independent Variables on the Two Dependent Variants [q] and [?]

The GZLM includes statistical procedures such as linear regression for normally

distributed response variables, logistic models for binary response or event count data, and

loglinear models for event count responses. One can also use a mix of statistical models through

the very general model formulation. The GZLM allow us to measure the main effects of one or

more independent variables on a dependent variable. In our case, the independent variables are

four: age, gender, residential area, and social class. The dependent variants are [q] and [7]. The

GZLM also allow us to measure the effect of interaction among the independent variables on the

dependent variable. The general linear model (GLM) multivariate procedure allows us to

measure the main effects and interaction of a number of independent variables on one or more

dependent variables through a multivariate test. The GLM, however, assumes normal distribution

and constant variance. The GZLM, thus, has an advantage over the GLM. They do not

necessarily "assume a normal distribution or constant variance" (Agresti and Finlay 1997: 550).

Having discovered in Section 4.2. 1 that the distribution of the dependent variants among the

different groups is not normal, it is more appropriate to use one of the GZLM. In addition to

allowing non-normal distribution, the GZLM expands the GLM by specifying a link function

that relates the dependent variable linearly to the independent variables.

To choose the appropriate GZLM model, we check the type of data in addition to the

distribution. In this study, the data are count. Count data refer to the number of events of a

number of trials. The number of events in my data is the number of occurrence of each of the two

dependent variants in the speech of each speaker. The number of trials is the total number of

occurrences for all speakers/subj ects for each variant. The number of trials, in my data, is the

fixed value 5874 for the [q] variant and 5674 for the [7] variant (Table 4-1). One can use binary









logistic to account for count data by specifying the Eixed value of the variable (i.e., the total

number of trials) for each variant, rather than treating the variant as binary. However, a Poisson

distribution is more commonly used for non-normally distributed count response variable. The

link function that is appropriate for count data and is usually used with a Poisson distribution is a

Log link. The Log link, unlike the identity link, allows the mean to relate non-linearly to the

independent variables. The Log link is often referred to as loglinear model. Hence, my initial

choice of procedure for my statistical analysis is the Poisson loglinear regression, which is one

type of regression tests that are used in probabilistic linguistics for count data (Manning

2003:335).

However, after performing the Poisson loglinear procedure, I realized that in the Goodness

of Fit Table (Appendix C) that the Deviance and the Pearson Chi-squared (4798.992 and

5687.238 respectively) are too high. The value/df is also too high (102. 106 for the Deviance and

121.005 for the Pearson Chi-squared), indicating overdispersion of the data. The value/df should

be close to 1. If it is higher than 1, it could still indicate a good fit of the model. However, if it is

too high, it indicates overdispersion and probably the need for a better fitting model for

overdispersion. In addition, the Log Likelihood is too low (-2527.807). The Akaike's Information

Criterion (AIC) (5065.613) and the Bayesian Information Criterion (BIC) (5075.369) are also too

high. The smaller the AIC and BIC, the better the fit. The values of the Deviance, Pearson Chi-

squared, Log Likelihood, the value/df, AIC, and BIC cast doubt on the goodness-of-fit of the

model~, althoughI lthe LikeihoodU ratio Chi1-squaredU (G2) value: (28.042) 1 in the Omib11us Tetl Table

shows significance (p = 0.000). The observed significance of the G2 meanS that we can rej ect the

null hypothesis that the independent variables have no effect on the variable use of the dependent

variant [q]. To evade this doubt, I checked the goodness-of-fit criteria of the standard Poisson










regression procedure against those of a negative binomial regression procedure with all other

settings equal. Usually, negative binomial regression is recommended for overdispersed data. By

changing the type of model to negative binomial with Log link, we actually get a better fitting

model. The Log Likelihood (Appendix C) becomes higher (-265.034), and the value/df is higher

than 1, but not extremely higher than one (1.932 for the deviance and 1.598 for the Pearson Chi-

squared). In addition, the Deviance and the Pearson Chi-squared (90.806 and 75.09 respectively)

are much smaller. The AIC (540.068) and BIC (549.825) are also much smaller than those in the

Poisson regression. The Omnibus Test Table also shows significance (p = 0.000) of the G2 ValUe

(41.3 13). In other words, the negative binomial procedure also rej ects the null hypothesis and

indicates that at least some of the social factors influence the variable use of [q]. Since both

Poisson and negative binomial regressions show significant G2 ValUeS, negative binomial

regression provides a better fit for the overdispersed data by having other supporting goodness-

of-fit criteria.

4.3.3.1 Negative binomial regression showing the main effects of the four independent
variables, age, gender, residential area, and social class, on the variable use of the
dependent variant [q]

In this section, mainly age and gender emerge as significant. Although residential area

emerges as significant in one test, it shows otherwise in another test. On the other hand, social

class emerges as insignificant. The negative binomial regression procedure is run though the

SPSS software. From the analyze menu, I choose Generalized Linear Models. Then, I choose

negative binomial regression as the type of model. Like the Poisson regression, negative

binomial assumes non-normal distribution of the count data and the link function Log. For the

response variable, I choose [q]. For the predictors, I choose the four independent variables as

factors: social class, age, gender, and residential area. From options, I choose descending. For the

model, I choose first the main effects of the four factors to see the effect of each independent









variable on the response variable independently. For the estimation, I use the Pearson Chi-

squared scale parameter method. This method is better for overdispersed data. For statistics, I

check the case processing summary statistics, descriptive statistics, model information, goodness

of fit, model summary statistics, and parameter estimates including exponential parameter

estimates. Under the estimated mean (EM)-estimates, I select all four factors to measure simple

contrast between each of the two categories included in the factor (e.g., contrast between the

female and male categories in the gender factor). I also identify the category of reference, usually

the highest number (2). For example, I chose the [Gender = 2], that is males who have been

assigned to the category 2, as the reference category. Finally, I save the predicted value of linear

predictor and the standardized deviance residual. These values are saved to the active data base

and later used in a scatter plot to perform an informal model check. From graphs, I select simple

scatter plot and drag it to the open window. The predicted value of linear predictor is dragged to

the horizontal/X-axis and the standardized deviance residual is dragged to the vertical/Y-axis.

The resultant plot should show us if there are any outliers.

The results of the negative binomial regression (Appendix C) show in the Tests of Model

Effects Table that age (p = 0.000), gender (p = 0.009), and residential area (p = 0.008) play

significant roles in the observed variation in the use of [q]. In contrast, social class (p = 0.862) do

not seem to play a significant role in the variable use of [q]. The role played by age seems to be

more important than the role played by gender and residential area. In the Parameter Estimate

Table, we can read the coefficients under the column entitled B. These coefficients enable us to

interpret the degree of use of each variant by the various groups of speakers. We can also read

the exponential of the coefficient B under the column entitled Exp(B), which tells us the odds

ratio of the occurrence of a certain sound in the speech of a group of speakers. The B coefficient









for age is 2.866; it is positive. This means that older speakers [Age = 2] use more [q] than

younger speaker [Age = 1]. The Exp(B) for age is 17.564. This number informs us that the odds

that older speakers would use the [q] are 18 times the odds that the younger speakers would use

the [q]. The coefficient for gender is 1.117; it is also positive. This means that males [Gender =

2] use more [q] than females [Gender = 1]. The Exp(B) for gender is 3.055. This number informs

us that the odds that males would use the [q] are 3 times the odds that females would use the [q].

The coefficient for residential area is 1.481; it is also positive. This means that Akrama speakers

[Area = 2] use more [q] than Al-Hameeddieh speakers [Area = 1]. The Exp(B) for residential

area is 4.398. This number informs us that the odds that Akrama speakers would use the [q] are 4

times the odds that Al-Hameeddieh speakers would use the [q]. Although social class did not

show significant effects, reading the coefficient may reveal some variation between the two

social classes in their use of [q]. Speakers from the upper-middle class use [q] slightly less than

speakers from the lower-middle class (coefficient B = 0.072). The Exp(B) for social class is

0.93. This number informs us that the odds that upper-middle class would use the [q] are 93% of

the odds/amount that lower-middle class would use the [q]. It is also useful to look at the

estimated marginal means that result from the simple contrast performed on the two categories

within each factor. The estimated marginal means indicate the difference between the two

categories within a factor.

Age. With regards to age, the estimated marginal mean for older speakers is 333.42 and

18.98 for younger speakers. If we multiply 18.98 by the odds ratio, 17.564, we get 333.36. This

result confirms that older speakers use [q] 18 times the amount used by younger speakers. In this

sense, older speakers are expected to use [q] much more than younger speaker do. The Individual

Result Table shows that the difference between older and younger speakers is highly significant










(p = 0.003). The Overall Test Table also reports significant contrast between older and younger

speakers (p = 0.003).

Gender. As for Gender, the estimated marginal mean for males is 139.06, whereas it is

45.51 for females. Multiplying 45.51 by the odds ratio, 3.055, we get 139.03, indicating males

are expected to use 3 times what females use. The Individual Result Table displays the

significance (p = 0.023) of this difference between males' and females' use of [q]. The Overall

Test Table also reports significant contrast between males and females (p = 0.023).

Residential area. Although the estimated marginal means for residential area show great

difference (166.85 for Akrama speakers and 37.93 for Al-Hameeddieh speakers), this difference

is not significant according to the Individual Result Table and the Overall Test Table. Both tables

report a p-value that is larger than 0.05 (p = 0.092).

Social class. The same applies to social class. The estimated marginal mean for upper-

middle class is 76.74 and 82.48 for lower-middle class). This difference is not significant

according to the Individual Result Table and the Overall Test Table. Both tables report a p-value

that is much larger than 0.05 (p = 0.86).

The significance shown in the contrast tests for age and gender indicate that the variation is

not due to chance; rather, it is a variation that is affected by those two social factors. The

difference is still significant even after the p-values are adjusted by the sequential Sidak method.

However, the insignificance of the contrast test for residential area shows that the variation

between Akrama and Al-Hameeddieh speakers is a chance variation, not necessarily a variation

affected by residential area. Thus, the simple contrast results for all social factors do not

completely accord with the results of the Parameter Estimates Table. Age and gender are

significant predictors of the variable use of [q], but residential area is not as significant.









Looking also at the scatter plot resulting from saving the standardized deviance residuals

and the estimated linear predictors, we note that there is strong variability of residuals across all

values of the linear predictor. However, variability increases towards the central value of the

mean and decreases as the mean increases. There may be a possible cause for concern regarding

outliers. There are 3 points that are higher than 2. Usually, residuals that are higher than 2 are

considered outliers. Some sources suggest that for a point to be consider an outlier, it should be

higher than 3 (Agresti and Finlay 1997). In this case, the scatter plot does not show any outliers.

4.3.3.2 Negative binomial regression showing the effect of the interaction of the
independent variables, age, gender, and residential area, on the variable use of the
dependent variant [q]

In this section, we will see interaction between age, gender, and residential area. However,

this interaction does not hold for all social categories. I apply a three-way interaction model to

see if there is interaction among the social factors that have shown significant main effects in

Section 4.2.3.1. I only exclude social class because it showed the least effect among the four

independent variables. I do not exclude residential area, although it did not show significant

difference between Akrama and Al-Hameeddieh speakers in the contrast test. The three-way

interaction model of the negative binomial regression seems to be more fitting than the main

effects model in Section 4.2.3.1. The Log Likelihood value is (-254.43), which is higher than the

Log Likelihood value of the main effects model (-265.034). The Deviance, Pearson Chi-squared,

AIC, and BIC show smaller values (69.59, 36.965, 524.86, and 540.47 respectively) than the

main effects model. Also, the value/df are smaller than the main effects model (1.582 for

Deviance and 0.84 for Pearson Chi-squared). The Omnibus Test Table also shows significance at

p < 0.05 and the Likelihood ratio Chi-squared is 103.808. The Test of Model Effect Table shows

that age (p = 0.000) and gender (p = 0.000) have significant effects independently from each

other. The p-value for residential area (p = 0.052) approaches significance. Thus, residential area









is included in the interaction model. The interaction of age, gender, and area is also statistically

significant (p = 0.000). The Parameter Estimates Table shows that age (p = 0.000), gender (p =

0.007) and the interaction between the [Age=1], [Gender=2], and [Area=2] (younger males from

Akrama) (0.005) are statistically significant. Residential area and other interactions of the

categories within the three factors do not show significance. Looking at the coefficient B values

may, however, be useful in determining the difference between the two categories within a factor

or the interaction between the various categories. The coefficient B for age is (3.734), which

indicates much more use of [q] by the older age group [Age = 2] than the younger age group.

The odds ratio under column Exp(B) shows that the odds that older speakers would use the [q]

sound is 41.845 times the odds that the younger speakers will use the [q] sound. The coefficient

B for gender is (1.339), which indicates much more use of [q] by males [Gender = 2] than

females group. The odds ratio under column Exp(B) shows that the odds that males would use

the [q] sound is 3.815 times the odds that females would use the [q] sound. The interaction

among [Age=1], [Gender=2], and [Area=2] seems to have the highest coefficient B (2.812)

among all other interactions. This means that younger male speakers who reside in Akrama use

more [q] than younger female speakers who reside in Akrama. This is supported by the fact that

in Table 4-1, there are three younger male speakers (29, 30, & 31) who reside in Akrama and use

almost 100% [q], in contrast to younger female speakers from the same area. In addition, the

coefficient (B = 484) of the interaction among [Age=2], [Gender=1], and [Area=2] shows that

older female speakers from Akrama use slightly more [q] than older female speakers from Al-

Hameeddieh and younger female speakers from both Akrama and Al-Hameeddieh.









4.3.3.3 Negative binomial regression showing the effect of the interaction of the two
independent variables, age and gender, on the variable use of the dependent variant [q]

This section will show strong interaction between age and gender. A two-way interaction

model is employed to investigate this interaction. This model shows a worse fit than the three-

way interaction model discussed in Section 4.2.3.2, but a better fit than the main effects model.

The Log Likelihood value for the two-way interaction is (-261.794). The Omnibus Test Table

shows significance (p = 0.000) and a Likelihood ratio Chi-squared of (71.153). The Test of

Model Effect Table shows that both age (p =0.000) and gender (p = 0.000) have significant

effects independently from each other. The interaction between age and gender is also significant

(p = 0.000). The Parameter Estimates Table shows similar results to the Test of Model Effect

Table. Age (p =0.000) and gender (p = 0.000) are statistically significant as well as the

interaction between gender and age (p = 0.000). The coefficient B (3.815) shows that older age

group uses [q] more than younger age group. The odds ratio shows that the odds that older

speakers would use [q] are 45.356 times the use of [q] by younger speakers. The coefficient B

(3.119) shows that males use [q] more than females. The odds ratio shows that the odds that

males would use [q] are 24.509 times the use of [q] by females. The coefficient B (-2.827) for the

interaction between age and gender shows negative association. That is, older male speakers use

[q] less than older female speakers and younger female and male speakers. The odds ratio shows

that the odds that older male speakers would use [q] are 5.9% of the use of other speakers. This

result does not sound right because my observations of speakers and all the previous statistical

tests show that older speakers and male speakers use more [q] than younger speakers and female

speakers .









4.3.3.4 Negative binomial regression showing the main effects of the four independent
variables, age, gender, residential area, and social class, on the variable use of the
dependent variant [?]

As it is the case regarding the use of [q], in this section age, gender, and residential area

emerge as significant concerning the use of [7]. I follow the same procedure of the negative

binomial regression that I followed for [q]. The only difference is that the response variable is [?]

instead of [q]. In the Goodness-of-Fit Table (Appendix D), Deviance and Pearson Chi-squared

are 104.98 and 50.341 respectively. The value/df are higher than 1 (2.234 for Deviance and

1.071 for Pearson Chi-squared), indicating the fitness of the model for the observed variation.

The Log Likelihood is -268.473. The Omnibus Test Table shows significance at p < 0.05 and a

Likelihood ratio Chi-squared of (51.853), indicating the existence of main effects of at least

some of the independent variables.

The results of the negative binomial regression procedure (Appendix D, Tests of Model

Effects Table) show that age (p = 0.000), gender (p = 0.005), and area (p = 0.035) play

significant roles in the observed variation in the use of [7]. In contrast, social class (p = 0.321) do

not seem to play a significant role in the variable use of [?].The role played by age seems to be

more important than the role played by gender and area. These results are similar to those when

examining the main effects of age, gender, residential area, and social class on the variable use of

[q]. In the Parameter Estimate Table, we can read the coefficients under the column entitled B.

Those coefficients enable us to interpret the degree of use of the variant [7] by the various groups

of speakers. The coefficient for age is (-2. 186); it is negative. This means that older speakers

[Age = 2] use much less [7] than younger speakers [Age =1]. The odds ratio for age (0. 112)

shows that the odds that older speakers would use [?] is very small (i.e., 1.12% of the amount of










use of [7] by the younger speakers). The coefficient for gender is (-0.858), which means that


male speakers [Gender = 2] use slightly less [7] than female speakers [Gender = 1]. The odds


ratio for gender (0.424) shows that the odds that male speakers would use [?] is very small (i.e.,


4.24% of the amount of use of [7] by female speakers). The coefficient for area is (-0.887); it is


also negative. This means that speakers from Akrama [Area = 2] use less [7] than speakers from

Al-Hameeddieh [Area =1]. The odds ratio for area (0.412) shows that the odds that speakers

from Akrama would use [?] is very small (i.e., 4. 12% of the amount of use of [7] by the speakers

from Al-Hameeddieh). Although social class does not play a significant role in the variation,

reading the coefficient column shows that there is a slight difference between the two social

classes. Speakers from the upper-middle class use [7] slightly less than speakers from the lower-

middle class (coefficient B = -0.353). The odds ratio for social class (0.703) shows that the odds

that the upper-middle class would use [?] is very small (i.e., 7.03% of the amount of use of [7] by

the lower-middle class). It is also useful to look at the estimated marginal means that result from

the simple contrast performed on the two categories within each factor. The estimated marginal

means indicate the difference between the two categories.

Age. With regards to age, the estimated marginal mean for younger speakers is 157.38,

whereas it is 17.68 for older speakers. Thus, younger speakers are expected to use [7] much more

than older speakers. Both the Individual Result Table and the Overall Test Table show that the

difference between the two age group in their use of [7] is significant (p = 0.000).

Gender. The estimated marginal means for gender also show great difference (81.01 for

female speakers and 34.35 for male speakers). This difference is significant according to the









Individual Result Table and the Overall Test Table. Both tables report a p-value smaller than

0.05 (p = 0.017).

Residential area. As for residential area, the estimated marginal mean for Al-Hameeddieh

speakers is 82.2 and 33.85 for Akrama speakers. In this sense, Al-Hameeddieh speakers' use of

[7] is more than double Akrama speakers' use of it. Both the Individual Result Table and the

Overall Test Table show that the difference Al-Hameeddieh and Akrama speakers is highly

significant (p = 0.016).

Social class. In contrast, the estimated marginal mean do not show a great difference

between the two classes (44.22 for the upper-middle class and 62.93 for the lower-middle class).

This difference is not significant according to the Individual Result Table and the Overall Test

Table. Both tables report a p-value that is larger than 0.05 (p = 0.304).

The significance of age, gender, and area shown in the contrast tests indicate that the

variation is not due to chance; rather, it is a variation that is affected by the three social factors:

age, gender and residential area. The difference is still significant even after the p-values are

adjusted by the sequential Sidak method. Thus, the simple contrast results for all social factors

accord with the results of the Parameter Estimates Table. That is, age, gender, and residential

area are significant predictors of the variable use of [?].

Looking also at the scatter plot resulting from saving the standardized deviance residuals

and the estimated linear predictors, we note that there is strong variability of residuals across all

values of the linear predictor. There may be a possible cause for concern regarding outliers.

There are 9 points that are higher than 2. Usually, residuals that are higher than 2 are considered

outliers. Some sources suggest that for a point to be consider an outlier, it should be higher than

3 (Agresti and Finlay 1997). In this case, the scatter plot does not show any outliers.









4.3.3.5 Negative binomial regression showing the effect of the interaction of the
independent variables, age, gender, and residential area, on the variable use of the
dependent variant [?]

In this section, we note that unlike the case of [q], there is no interaction between age,

gender, and residential area regarding the use of [7]. Like in the case of [q], I apply a three-way

interaction model to see if there is interaction among the social factors that have shown

significant main effects in Section 4.2.3.4. I only exclude social class because it showed the least

effect among the four independent variables. The three-way interaction model of the negative

binomial regression seem to be more fitting than the main effects model in Section 4.2.3.4. In the

Goodness-of-Fit Table (Appendix D), Deviance and Pearson Chi-squared are 101.808 and

45.578 respectively, which are smaller than the ones in the main effects model and closer to the

degree of freedom (df) value. The value/df are higher than 1 (2.3 14 for Deviance and 1.036 for

Pearson Chi-squared), which shows smaller dispersion than the main effects model and a better

fit. The Log Likelihood value is (-266.887), which is higher than the Log Likelihood value of the

main effects model. The Omnibus Test Table shows significance at p < 0.05 and a Likelihood

ratio Chi-squared 56.678. The Test of Model Effect Table shows that only age (p = 0.000) and

gender (p = 0.006) are statistically significant. Residential area (p = 0.077) does not have

significant effects. The interaction among the three social factors is also insignificant (p = 0.395).

In the Parameter Estimates Table shows only age (p = 0.000) is statistically significant. Gender

(p = 0.847) and residential area (p = 0.999), in contrast, do not show significance as well as the

interaction among the various categories. The coefficient B for age (-1.496) from the Parameter

Estimates Table shows that older speakers use [7] less than younger speakers. The odds ratio for


age shows that the odds that older speakers would use [?] is smaller than the odds that the


younger speakers would use [?] (i.e., 2.24% of the amount used by younger speakers). Although









the other two factors and the interaction among the three factors are not significant, I will discuss

them in terms of what the coefficients show us. It is worth noting that the coefficients and the

odds ratio show that males use less [7] than females (B = 0. 1; Exp(B) = 0.905 = 9.05% of the


females use of [7]). The coefficients and the odds ratio show that younger males from Akrama


use less [7] than younger males from Al-Hameedieh (B = 1.065; Exp(B) = 0.345 = 3.45% of the


use of [7] by younger males from Al-Hameeddieh). In addition, older female speakers form


Akrama use less [?] than older females from Al-Hameedieh and younger female speakers from

both Akrama and Al-Hameeddieh (B = 0.866; Exp(B) = 0.421 = 4.21% of the other females'

use of [?]).

4.3.3.6 Negative binomial regression showing the effect of the interaction of the two
independent variables, age and gender, on the variable use of the dependent variant [?]

This section will also show that unlike the case of [q], there is no interaction between age

and gender regarding the use of [7]. A two-way interaction model has a worse fit than both the

main effects model and the three-way interaction. In the Goodness-of-Fit Table (Appendix D),

Deviance and Pearson Chi-squared are 107.027 and 48.573 respectively, which are higher than

the other two models. The value/df are higher than 1 (2.23 for Deviance and 1.012 for Pearson

Chi-squared), an indication of a good fit. The Log Likelihood value is (-269.497), which is lower

than the Log Likelihood value of the other two models. The Omnibus Test Table shows

significance at p < 0.05 and a Likelihood ratio Chi-squared 52.86. The significance of Likelihood

ratio indicates rejection of the null hypothesis that the independent variables are insignificant.

The Test of Model Effect Table shows that only age (p = 0.000) and gender (p = 0.002) are

statistically significant independently from each other. The interaction among the two social









factors is insignificant (p = 0.103). In the Parameter Estimates Table only age (p = 0.000) is

statistically significant. Gender (p = 0.272) as well as the interaction between the two

independent variables (p = 0. 102) are not significant. The coefficients and the odds ratio show

that older speakers use less [7] than younger speakers (B = 1.619; Exp(B) = 0.198 = 1.98% of


the younger speakers' use of [7]). The coefficients and the odds ratio show that males use less [7]

than females (B = 0.454; Exp(B) = 0.63 5 = 6.3 5% of the female speakers' use of [7]). In


addition, older males use less [7] than younger males and older and younger females (B = -


0.931; Exp(B) = 0.394 = 3.94% of other speakers' use of [7]). It seems that a three-way

interaction (Section 4.2.3.5) and a two-way interaction (this section) do not show statistically

significant interaction among the social factors regarding the observed variation in the use of [7].

The three social factors, age, gender, and area, seem to work mostly independently from each

other.

4.4 Discussion of the Statistical Results

After investigating the degree of influence of each of the four social factors age, gender,

residential are, and social class on the variable use of [q] and [7], I will present a summary of

those findings, their implications, and their relation to other variationist studies and to the

theoretical proposal of this study. First, I will explore each variable separately. Then, I will

conclude with a general summary.

4.4.1 Age

Age has been investigated by many sociolinguistic studies and found to play a maj or role

in linguistic variation (e.g., Walters 1991, 1992; Miller 2005; Sankoff and Blondeau 2007). In

most studies, younger speakers are more likely to adopt the prestigious form. From the statistical










analysis above, age emerges as a maj or factor influencing the variable use of [q] and [7]. Table

4-3 presents manually calculated differences in percentage usage of [q] between younger and

older speakers. The table shows that there is a great difference between the two age groups.

Older speakers use [q] 85% of the time; younger speakers use it 15% of the time. This usage is

almost reversed with respect to [7]. Younger speakers use [7] 84%; older speakers use it 16%.

The difference between the usage of [q] and [7] between the two age groups is 69%. That is, the

older speakers use [q] 69% more than younger speakers; younger speakers use [7] 69% more


than older speakers.

Table 4-3. Distribution of [q] and [7] according to age group
Variant No. of Tokens % Younger No. of Tokens
for Younger Age Group for Older Age
Age Group Group

[q] 913/5718 16% 4961/ 5830
r31 4805/5718 84% 869/5 830


% Older
Age
Group
85%
15%
70%


Difference
in
Percentage
69%
69%


Diff. in percent
between use of
[q] and [7]


68%


It is worth noting that if we exclude the three young male speakers (29, 30, & 3 1) that

appear to be exceptions to the other young speakers, regarding their use of [q], we will have a

greater difference. Those three speakers live in Akrama and use [q] almost 100%. It seems that

they show more solidarity with their surroundings than younger male speakers from Al-

Hameeddieh who seem to be more influenced by the prestige of that area. Those three speakers'

use of [q] constitutes 783 tokens out of the 913 tokens used by all 24 young speakers. The

remaining 130 tokens are used by the other 21 speakers. If we divide 130 on the total number of

tokens (4889) for the 21 young speakers, we get less than 3% use of [q] by the younger










generation. Calculating also the three speakers' use of [7] leads to a total of 46 tokens out of the

4805 tokens of the 24 speakers. Subtracting these 46 tokens from 4805 and dividing the resulting

number (4759) on the total number of tokens for the 21 young speakers (4889) gives a 97% use

of [7] by the younger speakers.

These findings have a number of implications. First, the great difference between the two

age groups indicates that the shift towards the urban, prestigious form is quick. The younger

generation is more inclined towards the new form than older speakers. It took only one

generation to adopt the new form. Although the younger generation, in this study, are the sons

and daughters of the older generation and have been exposed to their parents' linguistic forms

from birth, they show complete shift of linguistic interest. This situation is comparable to

Miller' s (2005:924) study of migrant speakers to Cairo. She suggests that the contact situation in

Cairo leads "to a long-term accommodation for the first migrant generation and to total

accommodation or dialect shift for the second generation, for example, those born in Cairo who

speak almost dominantly CA." It seems that the younger speakers' exposure to schools, which

involves mixing with native Himsi children, at an early age, has greatly influenced their choice

of a variant. In this sense, they started understanding the social stigma associated with [q] at an

earlier stage and were able to evade it by completely adopting the Himsi form, [7]. The city

social values along with their influence on the selection of a linguistic variant became their

values, leading to their selection of [?] over [q]. This is the core of Mufwene' s (2005) theory of

language evolution and selection. Selection is very much influenced by the species' surroundings

and their relationship to their environment. Adopting the social values of their surroundings leads

to restructuring of their parents' initial input to adapt to their environment, an environment that

could be demanding linguistically and socially.









Second, some parents may be showing a struggle with this stigma, particularly older

females who show greater variation than older males. They are trying to adopt the new form.

Intra- and inter-speaker variation results from such attempts. Not all older speakers adopt the

prestigious form to the same degree. Such variation may be very indicative of a number of

issues. Older speakers are aware of the social stigma associated with [q], but they are unable to

employ their social knowledge to its fullest. Some of them show almost complete acquisition of

the new form (e.g., Speakers 17 & 21). Others do not show any kind of adaptation to the new

form. There are also those who oscillate between the two forms within the same conversation

with the same interlocutor. In the latter case, speakers probably have an internal struggle between

their original social values and the present social values that impose pressure on their linguistic

behavior. This internal struggle among social values leads to linguistic struggle because the

social and linguistic are highly associated. Those who do not show any adaptation to the new

environment could be the result of a fully developed social system that is very difficult to

interfere with after full development. This is similar to species that may easily change genetically

at their early stage of development for the sake of environmental adaptation, but once developed,

they hardly change.

This brings us to the third point. These findings are consonant with my theoretical proposal

in Chapter 3, in that, social factors should be treated on a continuous scale. They may be

employed completely in the adoption process, resulting in 100% usage of the new form. In other

words, their ranking values have been acquired and applied accurately in speech. In varying

speakers, their new ranking values are not totally acquired; hence, the observed oscillation

between the two forms. Speakers who do not show any kind of adoption of the new form seem to

cling to their original social values, and thus, the ranking values of their native form. Such









clinging could be an indication of their wish to maintain their original identity and show

solidarity with their native community. Those points are only applicable to socially conditioned

variations, not to other types of variation that may be conditioned by other factors, such as

linguistic factors including markedness and faithfulness factors.

4.4.2 Gender

Gender has played a maj or role in linguistic variation as well (e.g., Fischer 1958; Trudgill

1974; Macaulay 1977, 1978; Romaine 1978; Gal 1979; Milroy 1980; Abdel-Jawad 1981, 1986;

Milroy and Milroy 1985, 1992; Newbrook 1982; Eisikovits 1987, 1988; Walters 1991, 1992;

Eckert 1991; Sawaie 1994; Coates 1996; Haeri 1996; Daher 1998a, 1998b; Al-Wer 1999, 2002).

Most of these studies have shown that women are more inclined towards the prestigious forms.

Abdel-Jawad (1981) and Al-Wer (1999, 2002), for example, showed that women in Jordan use

the urban prestigious forms more than men do. Abdel-Jawad (1981) generalizes this to include

the Arab World. His generalization seems to be borne out. The statistical analysis above has

shown that gender is statistically significant in the variable use of [q] and [7]. Men tend to use

[q], the rural form, more than women do. On the other hand, women are more inclined towards

the urban form, [7]. Table 4-4 shows that males use [q] 66% of the time, whereas females use [q]

37%. There is a difference of 29%. Furthermore, women use [7] 63%, whereas men use it 34%

of the time with the same difference in usage.

Table 4-4. Distribution of [q] and [7] according to gender
Variant No. of Tokens % No. of Tokens % Difference in
for Males Males for Females Females Percentage
[q] 3634/5469 66% 2240/6079 37% 29%

[?] 1835/5469 34% 3839/6079 63% 29%
Diff. in percent 32% 26%
between use of [q]
and [7]









This difference between males and females is an indication that "women are more aware of

the social significance" (Habib 2005:26) of [7] than males are. In other words, women are more


sensitive to the new form (Labov 1972a:303). The higher use of [7] by rural females seems to

indicate that rural females like to climb up the social scale and sound city-like. It seems more

important for women to appear prestigious in a society that implements more limitations and

restrictions on women. Women in restricted communities usually learn to adapt more to their

environment than men do. Consequently, this leads to higher levels of linguistic adaptation.

Language becomes an escape gate to the world around them. It becomes a tool to declare their

difference (Ayres-Bennett 2004) from and superiority over men in one aspect of life, as Habib

(2005:27) asserts

Women probably compensate for their general social inferiority in Syrian society by
presenting themselves as more linguistically capable and prestigious ..... They may be
more inclined towards the prestigious forms because of the social pressure that is imposed
on them: sounding pleasant and aspiring to appear more educated and urban, so that they
can attract a good husband from a good social status and prosperous economic position.

Older women have shown linguistic insecurity (Laboy 2001) in their interviews. Most of

the varying women speakers from the older age group employed correction towards the

prestigious forms. With regards to some words, they may show a reverse type of correction or

rather slips of the tongue, which probably they could not control. Further, this could be due to

their incomplete acquisition of the ranking values of the social constraints and consequently the

ranking values of the relevant linguistic constraints. For example, Speaker-20 shows great

variation in the same conversation. Table 4-5 presents her use of a number of words in the same

conversation, sometimes with the sound [q] and at other times with the sound [7]. The table

shows the number of occurrence of each word and the corresponding meaning. One can notice

that the only difference between the two words in one row is the use of [q] or [?].This is just one









example of an older female speaker who shows great variation in her speech, and thus, some

linguistic insecurity.

Table 4-5. Variability in the speech of Speaker-20
Words with [q] No. of word with Words with [7] No. of word with Glossary

qabl 1 7abl 1 Before
qilt 3 ?ilt 17 I said
hallaq 3 halla? 12 Now
qiddainaa 1 ?iddainaa 1 We spent time
qallee 4 alleee 10 He told me
waqt 2 wa?t 2 Time
qal 2 ?al 2 Discourse marker


Statistics have also shown that the interaction of age and gender is statistically significant.

Table 4-6 shows that the difference between older female and male speakers is very small with

respect to the use of [q]. Although the number of older males (i.e., 13) is smaller than the number

of older females (i.e., 15), they use [q] 12% more than older females. This difference is greater

regarding the use of [?]. Older females use [7] 64% more than older males. As for younger male

and female speakers, the difference with respect to the use of [q] is greater between them than

between older speakers. Although the number of younger males (i.e., 1 1) is smaller than the

number of younger females (i.e., 13), they use [q] 90% more than younger females. Younger

females use [7] 30% more than younger males.

However, if we exclude the three young male speakers (Speakers 29, 30, and 3 1) that seem

to behave differently from all other younger speakers, we get completely different percentages.

These three male speakers are brothers and have lived all their lives in Akrama, a linguistically

less influential area because of its abundance in rural migrants whose speech is characterized

with [q]. Living in Akrama seems to have hindered them from adopting the prestigious form, [?].









Table 4-7 shows that excluding those three speakers brings the difference between younger male

and female speakers down to 36% more use of [q] by younger males. In addition, the difference

regarding the use of [?] goes up to 32% more use of [7] by younger female speakers. Regardless

of which table we adopt, there is difference when gender is grouped according to age. In other

words, gender and age work together in influencing this apparent variation.

Table 4-6. Distribution of [q] and [7] according to age and gender
No. of No. of [q] % No. of [7] %
speakers Tokens [q] Tokens [ ]
Older males 13 2763/4961 56% 155/869 18%
Older females 15 2198/4961 44% 714/869 82%
Difference in 12% 64%
percentage
Younger males 11 871/913 95% 1680/4805 35%
Younger females 13 42/913 5% 3125/4805 65%
Difference in 90% 30%
percentage

Table 4-7. Distribution of [q] and [7] according to age and gender without speakers 29, 30, & 31
No. of No. of [q] % No. of [7] %
speakers Tokens [q] Tokens [?]
Older males 13 2763/4961 56% 155/869 18%
Older females 15 2198/4961 44% 714/869 82%
Difference in 12% 64%
percentage
Younger males 8 88/130 68% 1634/4759 34%
Younger females 13 42/130 32% 3125/4759 66%
Difference in 36% 32%
percentage


4.4.3 Residential area

Residential area showed significance in the above statistics. Table 4-8 also shows

difference in the use of [q] between Akrama and Al-Hameeddieh speakers. Akrama speakers use

[q] 23% more than Al-Hameeddieh speakers, while Al-Hameeddieh speakers use [7] 23% more.

The results of this study support Miller' s (2005) findings in Cairo. Miller found that the speech









of migrant speakers was affected by the area of residence. Those who lived in the suburban area,

Giza, showed less accommodation to the Cairene forms. The reason is that the Giza area is

occupied with more rural people than the Cairo area. Consequently, people have less contact

with the new forms than those who live in the Cairo area and show more accommodation

towards the Cairene forms.

Table 4-8. Distribution of [q] and [7] according to residential area
Variant No. of Tokens % No. of Tokens % Al- Difference
for Akrama Akrama for Al- Hameeddieh in
Hameeddieh Percentage
[q] 1977/2921 68% 3897/8627 45% 23%
[?] 944/2921 32% 4730/8627 55% 23%
Diff. in percent 36% 10%
between use of [q]
and [7]


Furthermore, examining the interaction between age and area is useful because age has

shown throughout the study to be the most important factor in influencing the observed variation.

Table 4-9 shows that older people from Akrama use [q] 20% more than younger speakers from

Akrama. Younger speakers from Akrama use [?] 82% more than older speakers form Akrama.

Older speakers from Al-Hameeddieh use [q] 94% more than younger speakers from Al-

Hameeddieh. Although the number of younger speakers from Al-Hameeddieh (i.e., 17) is smaller

than the number of older speakers form Al-Hameeddieh (i.e., 22), they use [7] 66% more than

older speakers form Al-Hameeddieh. These results indicate that younger speakers from Akrama

are less susceptible to [7] in their environment. Thus, their degree of use of [q] is not much

different from that of older speakers. On the other hand, younger speakers who reside in Al-

Hameeddieh seem to have less exposure to [q]. Consequently, they use [q] much less than older

speakers. Furthermore, the smaller difference between older and younger speakers from Al-










Hameeddieh with regards to their use of [7] indicates greater exposure of the older speaker in Al-


Hameeddieh to [7]; hence, their greater use of [7] than those who live in Akrama.


Table 4-9. Distribution of [q] and [7] according to age and residential area
No of No. of [q] % [q] No. of [7] % [7]
speakers Tokens Tokens
Older speakers from Akrama 6 1178/1977 60% 88/944 9%
Younger speakers from Akrama 7 799/1977 40% 856/944 91%
Difference in percentage 20% 82%
Older speakers from Al-Hameeddieh 22 3783/3897 97% 781/4730 17%
Younger speakers from Al- 17 1 14/3 897 3% 3949/4730 83%
Hameeddieh
Difference in percentage 94% 66%

4.4.4 Social class

Social class did not show any significance in the statistical analyses with respect to the

variable use of both [q] and [7]. Table 4-10 shows that there is only a 16% difference between

the two social classes. The upper-middle class uses [q] 16% less than the lower-middle class and

[7] 16% more than the lower-middle class. These differences emerged as insignificant in

Sections 4.3.3.1 and 4.3.3.4.

The fact that the upper-middle class is more inclined towards the prestigious form is not

surprising. Many studies have shown that the lower-middle class aspires to appear prestigious.

Members of the lower-middle class try to imitate those from the upper classes. For example,

Labov (1972a) found that the use of (r) increases by social class and formality of style. However,

he found that women from the lower-middle class use the upper class form, r-1, more than other

speakers in word-lists and minimal pairs. He describes this phenomenon as hypercorrection

towards the more prestigious form among lower-middle class females. Speakers from the lower-

middle class realize the importance of the r-pronunciation. Consequently, they outperform the

upper-middle class in the r-pronunciation when they are able to monitor themselves in formal









styles, such as reading word-lists or minimal pairs. This phenomenon is also termed as the

"crossover pattern" (Labov 1972a) or "apparent deviation" (Labov 1966:227). Such crossover

patterns are taken as a sign for change in progress. Like Labov (1966, 1972a), Trudgill (1974)

found a crossover pattern in the use of the variable (ing) in Norwich. Female speakers from the

lower-middle class showed a great shift from the use of -in in casual style to the use of -ing in

formal style. This shift is much greater in the lower-middle class than it is in any other social

group. However, in other studies like Milroy's (1980), males from working classes showed less

use of the prestigious form in Belfast because of their strong connection to their working class

community. It is a way of expressing solidarity with their fellow workers.

Table 4-10. Distribution of [q] and [7] according to social class
Variant No. of Tokens % No. of Tokens % Difference
for LM LM for UM UM in
Percentage
[q] 3655/6284 58% 2219/5264 42% 16
[?] 2629/6284 42% 3045/5264 58% 16
Diff. in percent between use 16% 16%
of [q] and [7]


Since social class emerged as statistically insignificant, the associated factors such as

income and occupation are assumed, in Section 4.2, not to play a role on their own, by

inheritance, in determining variation in the use of [q] and [7]. A GZLM procedure was

performed to examine the influence of income, occupation, and education on each of the variants

[q] and [7]. The results are included in Appendix E. The results show that income (p = 0.318),

occupation (p = 0.804), and education (p = 0.353) do not play a significant role in the variable

use of [q]. However, the results are slightly different with respect to the variable use of [?].

Income (p = 0. 114) and occupation (p = 0. 148), which are associated with social class, do not









play a significant role. In contrast, Education, which is not associated with social class, plays a

significant role (p = 0.000). Speakers with elementary education [Education = 1] (p = 0.531) and

middle-school education [Education = 2] do not show significance according to the Parameter

Estimates Table. The other four educational categories show significance. The least significant

category is speakers with high-school education [Education = 3] (p = 0.043). The other three

categories associate degree [Education = 4], bachelor's degree [Education = 5], and

professional [Education = 6] are highly significant, all of which have the p-value (p = 0.000).

Examining the coefficients in the Parameter Estimates Table, we notice that the higher the

education, the higher the use of [7]. The coefficient for professional [Education = 6] is 4.029,

indicating that professionals use [?] much more than other categories. The Exp(B) is 56.201,


which informs us that the odds that professionals will use the [7] sound is 56 times the odds that

other educational categories will use the [?].The coefficient for bachelor' s degree [Education =


5] is 3.588, indicating that speakers with a bachelor' s degree use [7] much more than those with

less education. The Exp(B) is 36. 169, which informs us that the odds that speakers with a

bachelor' s degree will use the [7] sound is 36 times the odds that lower educational categories

will use the [7]. The coefficient for associate degree [Education = 4] is 3.473, indicating that


speakers with an associate degree use [7] much more than those with less education. The Exp(B)

is 32.249, which informs us that the odds that speakers with an associate degree will use the [7]


sound is 32 times the odds that lower educational categories will use the [7]. The coefficient for


high school [Education = 3] is 1.836, indicating that speakers with high school education use [?]

more than those with less education. The Exp(B) is 6.274, which informs that the odds that










speakers with high school education will use the [7] sound is 6 times the odds that speakers with

middle school or elementary school education will use the [7].

These findings implicate that schooling greatly affects a speaker' s adoption of the

prestigious form, [7]. Although in formal education the [q] sound is the basis for learning reading

and writing (Daher 1998a; Haeri 2003), it seems that the social influence of the people we

interact with in the school setting is greater, resulting in increased use of [?] and decreased use of

[q]. This finding is very interesting because it emphasizes the need to separate, in the Arab

World, between what is considered prestigious in speech, [7] in our case, and what is prestigious

or standard in formal education, [q]. This separation was first suggested by Ibrahim (1986: 125)

"Standard Arabic has a certain degree of prestige and its religious, ideological, and educational

values are undeniable, but its social evaluative connotations are much weaker than those of

locally prestigious varieties." Walters (1996: 177), likewise, differentiates between two norms, a

norm that is chosen for "free speech" and is "local", the other is chosen for reading aloud and is

"external". Hachimi (2001:30) further refers to the possibility of having "competing prestige

varieties". Thus, even local varieties may carry different types of prestige and may be evaluated

differently by different speakers based on the social class they belong to or their educational and

occupational background.

4.4.5 Relationship of the Statistical Findings to the Theoretical Analysis

In Chapter 3, I showed that the three social constraints *F [7], *OLD[?], and *AKR[?] are


involved in giving us results that match real life occurrences of [q] (82%) and [7] (18%) in the

speech of older female speakers from Akrama. The quantitative analysis of this chapter has

shown that there is significant interaction between gender, age, and residential area with respect










to the use of [q], not with respect to the use of [7]. This finding slightly contradicts with the

Endings of the theoretical analysis. Although, it reveals that there is interaction among the three

factors, it eliminates this interaction in the case of the variable use of [?]. Consequently, the

theoretical analysis has an advantage over the quantitative analysis, in that it shows that there is

partial interaction among the three factors with regards to the variable use of each of the

independent variants. The theoretical analysis splits each social factor into four different

constrains and specifies exactly which of the four constraints is involved. In this sense, only one

of the constraints may be involved at one time or used by a certain speaker or group of speakers,

whose effect may not show clearly in the quantitative analysis. Furthermore, not all speakers

may use the same social factor in the same way. Speakers may activate different constraints

within the same social factor. For example, I showed that younger female speakers from Al-

Hameeddieh activate *F[q], *YOUNG[q], and *HAM[q], as opposed to older female speakers

from Akrama, to arrive at their real life occurrences of [q] (2%) and [7] (98%). These results tell


us that residential area interacts with age and gender with regards to the use of [7], in contrast to

the results of the quantitative analysis. This finding highlights one further advantage of the

theoretical analysis. Although a whole factor may not be involved in the interaction with regards

to the use of one variant, part of it may be implemented by some speakers. Thus, the theoretical

proposal of this study allows for more options and accounts more closely for the differences

among different groups of speakers as well as individual speakers, which is very hard to do in

one step in the statistical analyses. The theoretical analysis shows that it is not sufficient to say

that this factor does not interact with other factors. It informs us specifically what constraints)

is(are) involved. It also gives us the mental representation of how these factors interact with the

linguistic constraints and how they influence the ranking values of those constraints.









Furthermore, the theoretical proposal allows us to see that certain social constraints may have

higher ranking values in some speakers than others. For examples, we have seen that *M[7] may

have a higher ranking value than *F [7] in the speech of older male speakers from Akrama


because of their higher use of [q] (96%) and lower use of [7] (4%) than older female speakers

from Akrama. This is a further advantage of the theoretical analysis. It can indicate that one

social factor, such as gender, may be more important to one group of speakers than another

group of speakers. In the case of the variable use of [q] and [7], it may explain the masculine

connotation of [q] and the feminine connotation of [7] in the literature (Sawaie 1994; Daher

1998a). Moreover, it may explain the masculine meaning that is associated with non-prestigious

forms versus the feminine meaning associated with the prestigious forms (Abdel-Jawad 1986).

Nonetheless, the quantitative analysis cannot be disregarded completely because it has also

its advantages, such as giving us a head start on what social factors may be involved or not,

particularly for someone who may not be very familiar with the community. For someone who is

from a community under investigation, like me, s/he will be able in the theoretical analysis to

know the social constraints that are involved in the learning process for each group of speakers.

Knowing the gender, area, and age of the speaker and the linguistic behavior of that speaker or

group of speakers would enable a researcher to easily discover the social constraints involved in

the speech of that speaker. Furthermore, I have mentioned in Chapter 3 that the GLA gives

predications on what the speech of speakers would sound like if certain social constraints are

involved. A listener is also capable of discovering the social constraints employed by a certain

speaker at a certain time and place because they are familiar with their environment and the

social constraints that are usually activated by different speakers. The GLA also gives









representations of the amount of input of each variant that a speaker or a group of speakers is

exposed to. Thus, it reflects on the social networks of the speakers. It can tell us that those older

female speakers from Akrama receive approximately 82% input of [q] and 18% input of [7],

implicating that the Akrama environment is more dense with rural people and it is not a good

environment for acquiring the new form quickly. This is apparent in the speech of the three

young male speakers from Akrama (29, 30, & 31). They differ from their counterparts who

reside in Al-Hameeddieh and who only took one generation to acquire the new form completely.

The complete acquisition of [7] by young female speakers from Al-Hameeddieh is a further

indication of the density of Himsi people in that area and that Al-Hameeddieh is a facilitative

learning environment of the new form.

This discussion leaves us with one further question. How does the learning process

proceeds? In other words, do speakers learn some words quicker than others? If that is the case,

does the frequency of certain words influence such acquisition? We have seen that the variable

use of [q] and [7] is influenced by social factors. However, it is worth examining if this variation

proceeds first in the direction of the most common words than it does in the direction of the less

frequent words.

4.5 Frequency Effects on the Acquisition of [7]

It is expected that highly frequent words will be acquired faster by speakers than less

frequent words (Pierrehumbert 2001; Medoza-Denton, Hay, & Jannedy 2003). First, I entered the

words for each speaker in a separate table, grouping similar words together. Then, I created one

huge table in which I grouped similar words from all speakers. I added the number of similar

words to examine the frequency of their occurrence. I extracted from that huge table the most

frequent words produced with [7], words that occur 20 times or more (Table 4-11). I also









extracted the most frequent words produced with [q], words that occur 20 times or more (Table

4-11). I performed this extraction because it may help in making comparison between the two

lists the [7] list and the [q] list. Having similar words in the two frequent lists may be a further

indication that the frequency of these words leads to faster acquisition of them. The frequency of

words produced with [q] does not necessary mean that it will negatively affect the acquisition.

Rather, it informs us that those words are very frequent in speech in society in general and that

people are exposed to them more often than to other words.

The process of learnability of the new form [7] will be noticed significantly if we examine

the words produced with [7] by older speakers whose speech is characterized with variation. For


this reason, I decided to investigate the words used with the [7] sound by those speakers. The

speakers who show greater variability are speakers 5, 6, 9, 16, 19, 20, 23, 25, and 28. There are

older speakers who show minor variability, such as speakers 1, 3, 8, 11, and 22. I will examine

first those with lower variability and then those with higher variability to see which words with

the [7] sound are used by them. Then, I will present in Table 4-12, the number of occurrence of

some of the frequent words extracted in Table 4-11 in their speech in comparison to less frequent

words. I will also calculate the percentage of the occurrence of these frequent words in relation

to other words uttered with [7] in their speech to arrive at conclusions regarding whether

frequency has influenced their acquisition of certain words before other words.

Table 4-12 strongly suggests that the more frequent the word, the more likely to occur in

the speech of varying speakers. This implies that frequently occurring words are acquired faster

than non-frequent words. In most of the varying speakers the word halla2'now' shows the


highest percentages because it is the most frequent word in Table 4-11 (73 5 tokens with [7] and









457 tokens with [q]). Even in speakers whose variation is minor, this word seems to penetrate

into their system because of its high frequency, as is the case with regards to speakers 1, 3, 8, and

11 (Table 4-12). The second most frequent word in Table 4-11 is wa~t 'time' (295 occurrences

with [7]). Table 4-12 shows that this word occurs in high percentages in the speech of most


speakers. For example, it is the only occurring word besides halla2'now' in the speech of


Speaker-8 and has a higher percentage than halla2'now' (Table 4-12). It has the highest


percentage after halla2'now' in the speech of Speaker-16 (Table 4-12). The next four words in


frequency are 2aal 'said', MZ't 'I/you said', ba~aa 'so/such/yet', and Palllee 'he told me' (264,


234, 124,, and 115 occurrences with [7] respectively) (Table 4-11). Paal 'said', for instance, has

the second highest percentage in Speaker-25, the third highest percentage in Speakers 16 & 19 ,

and the fourth highest percentage in Speaker-20 (Table 4-12). It occurs in the speech of most

other speakers. MZ't 'I/you said' has the second highest percentage in Speakers 19 & 28. It also


occurs in the speech of most other speakers. ba~aa 'so/such/yet' has the second highest

percentage in Speaker-23, the third highest percentage in Speaker-28, the fourth highest

percentage in Speakers 6 & 16, and the fifth highest percentage in Speakers 9 & 25. This word

appears in the speech of most other speakers. Palllee 'he told me' has the third highest percentage

in Speaker-20 and it occurs in the speech of most other speakers. It is worth noting here that a

word like 2alllee 'he told me' is related semantically to other frequently occurring words like


2allaa~~~~~~~~111111111 'he told her', Palhtuu 'he told him', Palul 'say/I say', etc. All of these words carry the

semantic meaning of 'saying'. They are calculated independently from each other. However, it is

possible to group them together because they are derived from the same root in Arabic. Arabic










morphology depends mainly on root and patterns; patterns are added to roots to derive other

words which are semantically related. Grouping them together will give higher frequencies of

them. Although the various forms of this verb and other words have been calculated separately, I

do not eliminate the possibility that speakers may operate on the basis of the frequency of the

root, not the word. Feldman, Frost, and Pnini (1995) have argued that morphemes and rules are

stored in the brain, comparing English, a concatenative language, with Hebrew, a non-

concatenative language like Arabic. Their study showed morphological relatedness effects in a

repetition priming study among words that share the same root, even at long lags. This indicates

that roots are stored separately in the lexicon and are accessed quickly by speakers who have

encountered them earlier on in speech. Once accessed, speakers apply stored rules to build or

decipher semantically related words. Nonetheless, Beret, Vaknin, and Marcus (2006) argued that

in Hebrew, stems, not roots, are stored in the brain, although their findings do not completely

eliminate the possibility of roots being stored. On the other hand, Davis and Zawaydeh (2001)

and Arad (2003) propose that both stems and roots are stored in the brain. This point is beyond

the scope of this study and requires further research.

In conclusion, from comparisons between Table 4-11 and Table 4-12, one can confirm that

frequency plays a maj or role in the acquisition process. Mostly, high frequency words are the

ones that occur in the speech of varying speakers. This is apparent in the high percentages of

these words in their speech. We have seen that the more frequent the word, the higher its

percentage in the speech of varying speakers. This leads to the conclusion that the more frequent

the word, the faster the acquisition of that word and the more likely for it to be acquired by

learners. Highly frequent words are even acquired by those whose speech can be characterized

with almost 100% use of [q]. Knowing the frequency of a word allows us to make predications









on whether this word is used by varying speakers or not. In other words, the frequency of words

gives expectations on what a varying speaker' s speech is likely to contain. Moreover, frequency

should not be understood as the main reason for acquisition of the new form. Social factors

impose pressure on speakers to adopt the new form. The role of frequency is leading this

adoption process towards the most common words first.

4.6 Conclusion

This chapter has shown the influence of social factors on the variable use of [q] and [7].

The results have shown that age, gender, and residential area play maj or roles in affecting this

variation. Social class emerged as an insignificant factor in this variation. The results have also

shown some interaction between the three significant factors with respect to the use of [q], but

not with regards to the use of [7]. The interaction between gender and age proved significant for

both variants. The chapter has also tested the association between social class and education,

income, occupation, and residential area. Education did not show significant association with

social class. Income and residential area showed the most significance. Occupation proved to be

significant with respect to the government employees' category. The factors that showed

association with social class, occupation and income, did not have an effect on the variable use

of [q] and [7]. However, education showed some significance with respect to the use of [7]

independently from social class, to which it is not associated. The chapter has also shown some

advantages of the theoretical analysis over the quantitative analysis. The theoretical analysis

allows more specificity with regards to social factors; it gives clearer picture of what revolves in

the mind of a speaker at a certain time and place. While some social constraints within one social

factor may be activated by a speaker or a group of speaker, they may not be activated by others.

While the quantitative analysis did not show interaction among age, gender, and residential area










with regards to the use of [7], the theoretical analysis was able to show this interaction because of

its specificity. Furthermore, the theoretical analysis has shown that one social constraint within

one social factor may have a higher ranking value within the same speech community than its

counterpart. This specificity is not as achievable in the quantitative analysis as it is in the

theoretical analysis. Finally, this chapter has shown that word frequency influence the acquisition

process. Highly frequent words are acquired before less frequent words. The following chapter

will present a quantitative analysis of the four variants [t] and [s] and [d] and [z]. Those four

variants will be dealt with differently from the two variants [q] and [7] because as we have seen

they are not socially conditioned.

























riflaat/rif?eet
?uulee
ha??
tlaa?ee
rfli?
?illik
bii?uuloo
?addeef
tarii?
ta?riiban/ta?riiban
?alluu
far?
?al

?afad
?issa
?imt/7umit

?add
?allaa
?alb
?uul
?aa~di
foo?
?ilnaa
?iddaam/7iddeem


7]


Table 4-11. Most frequent words produced with [q] and [
Word produced No. of Matching word
with [7] tokens produced with [q]
I halla?/halle? 735 hallaq/halleq
2 wa?t 295 waqt
3 ?aal 264 qaal
4 ?ilt 234 qilt
5 ba?aa 124 baqaa
6 alleee 115 qallee
7 7abl 96 qabl
8 ?aam 77 qaam


lossary


No. of
tokens
457
82
424
397
98
172
87
41

53
61
49
22
24
110
47
19
48
59
148
13
139

7
21
3

34
42
39
20
10
55
88
60


Now
Time
Said
I/you said
So/such/yet
He told me
Before
Discourse marker (Lit. 'got
up/did')
Friends
You can say/say
Right/price
You can find
Friend
I tell you
You (pl)/they say
How much
Road/way
Almost/approximately
He told him
Difference
Discourse marker (Lit. 'it has
been said')
He sat/lived
Story
Discourse marker (Lit. 'I got
up/did')
As much
He told her
Heart
Say/I say
Sitting/living (F)
Up/above/upstairs
We said
Before/in front of


rifqaat
quulee
haqq
tlaaqee
rfliq
qillik
quuloo
qaddeef
(ariiq
taqriiban/taqriiben
qalluu
farq
qal

qafad
qissa
qimt

qadd
qallaa
qalb
quul
qaa~di/dee~di
fooq
qilnaa
qiddaam/qiddeem











Glossary


No. of
tokens
43
36
32
7
19
7


28
25
8
10
10
6
12
20
10
9
25
22
20
20
20
27
28
29
33
24


Close/relative
She/you tell(s) me
We moved
Method/way
Papers
I find
He sits/lives
Possible
We sit/live
Living/sitting
He told them
Less
Social class
Mind
Coffee
He tells me
He tells me
Relation
He is sitting down
Floor/story


Little
Old/archaic
Paper/leaf/Syrian pound
Town name
Market/drive
Area


Table 4-11. Continued.
Word produced
with [7]
33 ?aliil
34 7adiim
35 war?a/wara?a
36 ?affiini
37 suu?
38 manti?a/
manta?a/mant?a
39 7ariib
40 t?illee
41 nta?alnaa
42 tarii?a
43 wraa?
44 laa?ee
45 y?Yod
46 maf?uul
47 ni?Yod
48 ?aa~diin
49 ?allon
50 ?a?all
51 taba?a
52 Ta?l
53 ?ahwi
54 y?illee
55 bii?illee
56 Talaa?a
57 ?ee~id/7aafid
58 taabi?


No. of
tokens
27
12
26
29
28
28


24
24
24
23
23
22
21
20
20
20
5
7
12
15
8
18
13
13
15
13


Matching word
produced with [q]
qaliil
qadiim
warqa/waraqa
qaffiini
suuq
mantgaq/mantqa

qariib
tqillee
ntaqalnaa
tariiqa
wraaq
laaqee
yq~od
ma~quul
niq~od
qaa~diin/qee~diin
gallon
?aqall
Jabaqa
Taql
qahwi
yqillee
biiqillee
Talaaqa
qee~id/qaafid
taabiq









Table 4-12. Percentage of the occurrence of frequently occurring words in the speech of varying
spaers
Speaker Word with [7] No. of Total No. of % of the word' s occurrence to
tokens tokens with [7] the total of words with [7]
1 halla? 8 10 80%
3 halla? 2 3 67%
8 halla? 3 10 30%
wa?t 7 70%
11 halla? 9 19 57%
22 ?imt 1 1 100%
5 ?afad 8 22 36%
?abl 2 9%
rif?aat 1 4.5%
wa?t 1 4.5%
halla? 1 4.5%
?aam 1 4.5%
?aal 1 4.5%
?illik 1 4.5%
?umit 1 4.5%
6 halla? 9 41 22%
taba?a 5 12%
foo? 3 7%
ba?aa 2 5%
manti?a 2 5%
?al 1 2%
tarii?a 1 2%
tarii? 1 2%
far? 1 2%









Table 4-12. Co atinued.
Speaker Word with [7] No. of Total No. of % of the word' s occurrence to
tokens tokens with [7] the total of words with [7]
9 ?aany 11 50 2294
halla? 9 1896
vva?t 5 1094
?uul 4 894
ba?aa 4 894
?uulee 3 6%
?alluu 3 6%
?addeef 2 495
7aal 2 495
?imt 1 2%
suu? 1 295
?abl 1 294
gallon 1 2%


halla? 12
vva?t 6
7aal 5
ba?aa 4
?ilt 3
?add 3
?al 2
tarii? 2
7aliil 2
?allaa 2
taba?a 2
?illik 2
?illaa 2
?alluu 2
rif?aat 1
?ahwri 1
tlaa?ee 1
?uulee 1
?illon 1
?adiing I


295
2%
1%









Table 4-12. Continued.
Speaker Word with [7] No. of Total No. of % of the word' s occurrence to
tokens tokens with [7] the total of words with [7]
19 halla? 12 75 16%
?ilt 9 12%
?aal 9 12%
?alluu 7 9%
rfli? 3 4%
rif?aat 2 3%
?add 2 3%
?aa~diin 1 1%
?aam 1 1%
?ilnaa 1 1%
?ahwi 1 1%
?afad 1 1%
alleee 1 1%
foo? 1 1%
20 ?ilt 17 61 28%
halla? 12 20%
alleee 10 16%
?aal 4 7%
wa?t 2 3%
?al 2 3%
?ilnaa 2 3%
?imt 1 2%
Ya?l 1 2%
wara?a 1 2%
?abl 1 2%
?add 1 2%
23 halla? 14 21 67%
ba?aa 1 5%
t?illee 1 5%
?ilt 1 5%
?ilnaa 1 5%
?aal 1 5%









Table 4-12. Continued.
Speaker Word with [7] No. of Total No. of % of the word' s occurrence to
tokens tokens with [7] the total of words with [7]
25 halla? 16 55 2995
?aal 13 2494
?al 7 1394
ni?Yod 2 4%
ba?aa 2 494
vva?t 2 494
?aan, 1 294
alleee 1 295
?allaa 1 295
?abl 1 294
rnaf?uul 1 295
28 halle? 10 29 3594
?ilt 3 1095
ba?aa 1 395
Ta?l 1 394
Talaa?a 1 394
?ilt 1 395
?aal 1 394









CHAPTER 5
QUANTITATIVE ANALYSIS OF THE VARIANTS [t] AND [s] AND [d] AND [z]

5.1 Introduction

Because all speakers behave in the same way regarding the use of the variants [t] and [s]

and [d] and [z], social factors do not seem to play a role in this type of variation. If social factors

were to play a role, we would notice variation among speakers and groups of speakers. On the

contrary, what we notice is that certain words are always used by all speakers with only one of

the variants, [t] or [s] or [d] or [z]. In Chapter 3, I have shown that the linguistic environment

does not influence the use of one variant over another. This led me to examine the frequency of

these words to see if the number of occurrences of words has diachronically led to the present

stable phenomenon.

To investigate the effect of frequency on the use of the four variants, I entered the words

pronounced with each of the four variants in a separate table for each speaker, grouping similar

words together. Then, I grouped similar words from all speakers in one large table to arrive at the

frequency of each word in the speech of all 52 speakers. Then, I summarized the most frequent

words in Tables 5-1 and 5-6. The results of this process will be explored in the following

sections. I will first start with the frequency analysis of the two sounds [t] and [s] followed by a

discussion of the results. Then, I will present the frequency analysis of the two sounds [d] and [z]

followed by a discussion of the results. The general discussion and conclusion Sections 5.6 and

5.7 will touch on usage-based theory proposed by Bybee (2001) and on how this theory is able to

explain the present situation as a result of two different diachronic changes.

5.2 Analysis of the Use of [t] and [s]

I extracted from the large table, in which I grouped all similar words produced with [t] and

[s] and [d] and [z] together, the most frequent words produced with [t] and [s], words whose










frequency is 10 or higher (Table 5-1). The total number of words produced with [t] is 2767.

Table 5-1 shows that the most frequent word produced with [t] is ktiir 'much/a lot'; it occurs

1 159 times, which constitutes 42% of the total number of words produced with [t]. The second

most frequent word is mitl 'like/as' (288 tokens = 10%). The third word in frequency is

2alktar PaktorlS 'more' (245 tokens = 9%). tlat 'three' (153 tokens), taanee ~~tt~~tteeneettt~~tt 'second (M)'


(149 tokens), and tlaati~t~lt~ltlt~lt~lt~ 'three' (123 tokens) are respectively next highest in frequency. They

constitute 6%, 5%, and 4% of the total number of words produced with [t] respectively. Though

tlat and tlaati~t~lt~ltlt~lt~lt~ have exactly the same meaning, their syntactic position is usually different, and

thus their usage. tlat is used as the first term of a construct state (two consecutive nouns) (e.g.,

tlat kutub 'three books'). It cannot stand as a number on its own or be connected to an enclitic

pronoun. tlaatilt~lt~ltlt~lt~lt~ on the other hand, is used as a number on its own and can be suffixed to an

enclitic pronoun (e.g., tlaartiton 'the three of them'). Other words such as tlaa~lt~ltlt~lt~ltiintleetiin 'thirty'

(93 tokens = 3%), tneen 'two' (71 tokens = 3%) and taaniivii ~~tt~~tteeniivii~t~t 'second (F)' (64 tokens =

2%) also show high frequency.

As for the variant [s], the total number of words produced with [s] is 870. The most

frequent word produced with [s] is ma;salin ma;salan 'for example/instance' (530 tokens = 61%).

This seems to be the most frequent word among all words produced with [s]. Other words such

as saa2ir 'a proper noun' (19 tokens), iEcdiis 'modern/talk' (18 tokens), and ihadis 'accident'


(16 tokens) constitute about 2% each.



15 Paktar occurs in free variation with Paktor. Paktar is the HCA: Paktor is the RCA from. This example shows one
of the vowel differences between the two dialects. As Paktor and Paktar occur in free variation, other subsequent
words in the text such as taanee teenee, tlaatiin dleetiin, taaniyiviiteeniivii, masalan masalin occur in free variation.
In each of these pairs of words, the first term is the HCA form: the second term is the RCA form. These are
examples that show vowel differences between the two dialects. RCA speakers who reside in Hims may vary their
speech regarding the use of these vowels, which is beyond the scope of this study.
























musaqqaf
bahs
?assar
t?assir
saqaafli
siqaa
Total


30




27
26
23

22
22
16

15
13
12
11
2671/2767


ord produced
ith [s]
asalin/masalan

a?ir

Idiis

ladi s
anawiivii


Glossary No. of
tokens


I


Table 5-1. Most frequent words produced with [t] and [s]
Word produced Glossary No. of W
with [t] tokens wj
1 ktiir Much/a lot 1159 m;


sa

ha

ha
sa


530

19

18

16
13

12
11
11
10
10
10
660/870


For
example
A proper
noun
Modern/
talk
Accident
High
school
Educated
Research
Affected
Effect (V)
Education
Trust


2 mitl


Like/as


288

245

153
149

123
93
71
64
41
37
31


3 ?aktar/7aktor

4 tlat
5 taanee/teenee

6 tlaati
7 tlaatiin/tleetiin
8 tneen
9 taaniivii/teeniivii
10 tmaaniin/tmeeniin
11 tminn
12 tmaanaauu




13 tlaataauu




14 tmaanii/tmeenii
15 taalit/teelit
16 kattir

17 taaltii/teeltii
18 tnteen
19 tlaatmiivii

20 tnaff
21 t?iil/tqiil
22 tyaab/tyeeb
23 tmintaff
Total


More

Three
Second

Three
Thirty
Two (M)
Second
Eighty
Eight
The monomial
'eight' digit of
a two digit
number, not
decimal
The monomial
'three' digit of
a two digit
number, not
decimal
Eight
Third (M)
Make more/
mecrease
Third (F)
Two (F)
Three
hundreds
Twelve
Heavy
Clothes
Eighteen









If we compare the number of words whose frequency is less than 10 times for each of the

variants [t] and [s], we realize a larger number of words produced with [s] than with [t].

However, most of those words occur once or twice. 23 words out of 50 words produced with [t]

have very high frequency. That is, about half of the words have very high frequency. They add

up to 2671 words out of 2767 words (Table 5-1), which constitute 97% of the total number of

words produced with [t] (Table 5-2). On the other hand, only 11 words out of the 109 words

produced with [s] have high frequency. They total 660 out of 870 (Table 5-1), which constitutes

76% of all the words produced with [s] (Table 5-2). In addition, their frequency is not as high as

those produced with [t]. They have very low frequency compared to the words produced with [t],

which is obvious from each word count and from the number of words that have much higher

frequency. Furthermore, a proper noun such as saa2ir mainly occurs in the speech of two

speakers, his parents. His name is repeated many times during relating personal facts or stories

about their son. Even the most frequent word produced with [s], masai~lin ma;salan, adds to less

than half the frequency of the word ktiir.

This comparison will be made clearer by constructing a contingency table that show the

frequency of words pronounced with [t] and [s] in comparison to each other. Table 5-2 shows a

two-category division of frequency: words with frequency less than 10 and words with frequency

higher than 10. The table also shows the percentage of those frequencies to the total number of

words. The two sets of percentages for the [t]-words and [s]-words are called the "conditional

distribution" (Agresti and Finlay 1997:25 1) on the frequency. If the conditional distribution on

the frequency is identical, the two categorical variables are statistically independent (i.e.,

frequency and type of word are independent form each other). In other words, frequency does

not have an effect on the type of sound used in a word. In contrast, if the conditional distribution









is not identical, the two categorical variables are statistically dependent (i.e., frequency and type

of word are dependent on each other). In other words, frequency influences the type of sound

used in a word. Table 5-2 shows that there is discrepancy in the conditional distribution; hence,

frequency and the type of sound used in a word are dependent on each other. Frequency plays a

role in this type of variation. Frequent words are mostly pronounced with the [t] sound in

contrast to infrequent words that are mostly pronounced with the [s] sound.

Table 5-2. Contingency table showing the frequency use of [t] and [s] words and their
conditional distribution in all speakers
Word Type Freq < 10 & % Freq > 10 & % Total & %

[t]-words 96 (3%) 2671 (97%) 2767 (100%)

[s]-words 210 (24%) 660 (76%) 870 (100%)

Total & % 306 (8%) 3331 (92%) 3637



To have a visual presentation of this comparison, I entered all the words produced with [t]

and [s] into SPSS as two separate variables. Their frequency category was entered as a separate

categorical nominal variable with 1 assigned for low frequency (<10) and 2 for high frequency

(> 10). The result of the examination of frequencies from Analyze-Descriptive Statistics-

Frequencies is summarized in Figures 5-1 and 5-2. Figure 5-1 shows clearly that the difference

between the number of high and low frequency words produced with [t] is not very big. Figure 5-

2 shows a great disparity between the numbers of high and low frequency words produced with

[s]. In addition, comparing the two figures reveals a great difference in frequency between the

high frequency words produced with both [t] and [s]. From the two figures, one can conclude

that the higher the frequency of the word, the higher the possibility that it is pronounced with [t].


















































Frequency of [t] words




Figure 5-1. Distribution of [t]-words between high and low frequency


30-







S20-






l-


21
u60-
r
o


u
40r





~LO-


low high

Frequency orf [s] wvords




Figure 5-2. Distribution of [s]-words between high and low frequency



205










Furthermore, to test whether the difference in frequency between [t]-words and [s]-words

is significant, I conducted two two-tailed one-sample t-tests to compare the raw frequency of the

[t]-words to the raw frequency of [s]-words. In this case, I entered into SPSS all words

produced with [t] and [s] as two separate variables and their raw frequencies as two separate

variables. The comparison yielded a highly significant result ([t]-words frequency: t = 2.294, df=

60, p = .025; [s]-words frequency: t = 2.334, df = 1 09, p = .02 1). In general, the mean for [t]-

words is 45.36 which is much higher than the mean of [s]-words, 7.93 (Tables 5-3). The high

difference in mean is a good indicator of the influence of frequency on the current use of [t]-

words vs. [s]-words.

Table 5-3. Two tailed One-Sample T-Test comparing the significance of the difference in the
frequency of [t]-words and [s] words
One-Sample T-Test
Test Value = 0
9% Confidence Interval of the Dfeec

t df ISig. (2-tailed)Ma Difference Lower Upper

Frequency [t]-words 29460 .025 45.361 5.81 84.91
Feuny [s]-wod .3 109 .021 7.927 1.19 14.6


5.3 Discussion of the Frequency Analysis of the Stable Variants [t] and [s]

The above observations are an indication that frequency may have played a role

diachronically in determining this synchronic type of stable phenomenon (i.e., which words are

produced with [t] and which ones are produced with [s]). Nonetheless, if we take, at face value,

the complete merger theory of the SA variants [6] and [a] with [t] and [d] respectively around the


14th century (Daher, 1998a, 1999; Schmidt 1974), how do we explain the presence of words with

[s] pronunciation instead of [t]? Daher (1999: 164) suggests that the use of [s] and [z] is more

elevated than the use of [t] and [d] respectively in colloquial speech. If his suggestion is correct,









we would expect variation among speakers and groups of speakers. This is not the case in my

data. All speakers behave similarly. According to some studies, a second change occurred some

time after the 14th century (Birkeland 1952; Schmidt 1986:57; Schulz 1981:33; all cited in Daher

1999). Borrowed words form SA with the [6] and [a] variants were replaced in colloquial speech

with [s] and [z] respectively. This seems plausible given the following observations.

The non-frequent words with the [t] sound are related semantically to the frequent words.

They are derived from the same root and most of them are related to the numbers three and eight

as well as the word for much. On the other hand, the low frequency words with the sound [s] are

not necessarily semantically related to those of high frequency, but they may be semantically

related to each other. A close examination of these words reveals that many of them have the

vowel system of the SA corresponding words, indicating that they are mainly borrowings from

SA. However, due to the high markedness of the sound [6], the sound [s] is used instead. Table

5-4 presents just a few examples that show the similarity in vowel system between the observed

forms and their SA corresponding forms. Other words may have shown some assimilation to the

vowel system of the dialects, but this change is less likely to happen word-internally. It mostly

occurs word-finally as a result of suffixing a colloquial suffix in place of the SA suffix, for

example, the colloquial feminine marker [-i] in place of the SA feminine marker [-a] (Examples

1-3 in Table 5-5); the relational adjective marker [-ee] in place of [-iiy] (Examples 4 & 5 in

Table 5-5); or a colloquial pronominal in place of a SA pronominal (Examples 6 & 7 in Table 5-

5). There is a number of word-internal vowel assimilation to the vowel system of the dialects, as

in Examples 8 & 9 in Table 5-5. The integration of the SA root with the colloquial dialects

patterns could be an indication that roots and patterns are stored separately in the brain because

speakers could interlock the root of the SA with a pattern from the colloquial dialects. Thus, the









root is borrowed from SA, not the pattern, in some cases. Furthermore, the previously observed

semantic relatedness among words that are produced with one of the two variants provides

evidence for the possibility that roots and patterns are stored separately in the brain, because

roots are "semantic abstractions" from which semantically related words are derived (Holes

2004:99). Speakers seem to use the same root with its dialectal phonemic representation (i.e., the

same sound [t] or [s]) in different words that are semantically related.

Table 5-4. Similarity between the vowel system of observed words with [s] with their
counterparts in SA
Observed word with [s] SA corresponding word with [6] Glossary
1 ?asar ?aear Trace
2 ?asnaa? ?aenaa? During
3 ?isbaat ?iebaat Proof
4 7ahdaas ?ahdaa6 Incidents
5 saabit Baabit Firm/fixed
6 tamsiil tameiil Acting

Table 5-5. Assimilated words with [s] to the vowel system of the dialects
Observed word with [s] SA corresponding word with [6] Glossary
1 wiraas-i wiraa6-a Heredity/ inheritance
2 ba~si baf6-a Delegation
3 lissi li66-a Gum
4 sawaanee Bawaan-iiy Seconds
5 saanawee Baanaw-iiy High school
6 saqaaft-uu 6aqaafatu-huu His education
7 ythadds-oo yatabadda6-uuna They talk
8 tuusaq-ee taeiq-iina You (F) trust
9 tuuras tarie Inherit


In Chapter 3, I have established that /t/ and /s/ are the underlying form for [t] and [s]

respectively in colloquial speech. The phonological representations of these two sounds are

accessed when using similar words. Because phonemes constitute morphemes and roots are

separate morphemes in Arabic, one can conclude that speakers access these root morphemes with









their phonemic representations. The status of the root and the pattern in the brain is beyond the

scope of this study and requires further research. They are mentioned here to show that there is

the possibility that some words (more specifically roots/morphemes) are borrowed from SA and

are intersected with colloquial morphology and phonology. The borrowings show assimilation to

the vowel system of the dialects and adoption of the phonemic system of those dialects, which

consists of the phonemes /t/ and /s/, not /6/.

5.4 Analysis of the Use of [d] and [z]

I also extracted from the large table that contains all words produced with [t] and [s] and

[d] and [z] the most frequent words produced with [d] and [z], words whose frequency is 10 or

higher (Table 5-6). The total number of tokens produced with [d] is 1740. Table 5-6 shows that

the most frequent word produced with [d] is haadaa 'this (M)'; it occurs 607 times, which

constitutes 35% of the total number of words produced with [d]. The second most frequent word

is hadool 'these' (135 tokens = 8%). The third word in frequency is haadii 'this (F)' (116 tokens

= 7%). aya~d 'took' (114 tokens = 7%), hadiik 'that (F)' (67 tokens = 4%), and Pag~d 'take' (59

tokens = 3%) are respectively next highest in frequency. Other words such as riay;J'lI take' (48


tokens = 3%), yaagod 'he takes' (48 tokens = 3%) and riay/l'take' (47 tokens = 3%) also show

high frequency. It is also worth noting here that there are a number of variants of haadaa 'this

(M)', hadool 'these', and haadii 'this (F)', all pronounced with [d], but they show various vowel

patterns. These words are haad 'this (M)' (2 tokens), hadaa 'this (M)' (1 token), hedaaddddd~~~~~~~dddddd 'this

(M)' (1 token), hidaakdddd~~~~~~~dddddd 'that (M)' (1 token), hauud 'these' (1 token), hauudaalee~~~~dddd~~~~ddd 'these' (2

tokens), haduuk 'those' (4 tokens), and haidii 'this (F)' (8 tokens). If we add these words to their

respective variants (i.e., those with which they show free variation), we will find even higher





So/like this
Remember
(V)
For this
reason
Teacher/
professor
Lying
Proper noun
Suffered
Intelligence
So/like this

Permission
Same


frequency of those words. It may not make much difference in this small corpus of words, but it


may make a difference when looking at a larger corpus of words.


Table 5-6. Most frequent words produced with [d]
Word produced Glossary No. of
with [d] tokens
1 haadaa This (M) 607
2 hadool These 135
3 haadii This (F) 116


and [z]
Word produced
with [z]
?izaa
kazaa
tzakkar


Glossary


No. of
tokens
567
231
35

29

27

26
22
15
14
14

12
10
1002/1257


Took

That (F)

Take
I take
He takes
Take
I take/ Ihave
taken
I took
That (M)
You take/she
takes
Has taken/
taking/ I take
She takes/ you
take
I take
Has taken/
taking/ I take
We take
Name of a city

We took
He takes
These
Has taken
She takes
Take (imp)
This
Corn
Virgin Mary


lizaalik

?staaz

kizb
j'azaa
t~zzab
zakaa
kizza

?izn
zaat
Total


4 ?aXad

5 hadiik

6 ?ayd
7 ?aaXod
8 yaaXod
9 yaayd
10 ?aayd

11 Xadt
12 hadaak
13 taayd

14 ?aaXid

15 taaXod

16 baaXod
17 ?eexid

18 naaXod
19 Ilaad?iivii/
Ilaadqiivii
20 Xadnaa
21 yaaXid
22 hauudee
23 7eeXd
24 taaXid
25 Xood
26 haidaa
27 daraa
28 Tadraa
Total


29

27

24
24

23
22

20
17
15
15
14
13
13
11
10
1640/1740









As for the variant [z], the total number of tokens produced with [z] is 1257 (Table 5-6).

The most frequent word produced with [z] is izaa 'if (567 tokens = 45%). The second most

frequent word is kazaa 'so/like this' (23 1 tokens = 18%). tzakkarzt~t~ztzt~t~ztzt~ 'remember' (3 5 tokens)

constitutes 3%. Other words such as lizaalik 'for this reason' (29 tokens), and 2staazttt~~~~~ttttt~~~~

'teacher/Mr./professor' (27 tokens), and kizb 'lying' (26 tokens) constitute about 2% each.

If we compare the number of words whose frequency is less than 10 times for each of the

variants [d] and [z], we realize a larger number of words produced with [z] than with [d].

However, most of those words occur once or twice. 28 words out of 64 words produced with [d]

have very high frequency. That is, about half of the words have very high frequency. They add

up to 1640 words out of 1740 words (Table 5-6), which constitute 94% of the total number of

words produced with [d] (Table 5-7). On the other hand, only 12 words out of the 138 words

produced with [z] have high frequency. They total 1002 out of 1257 (Table 5-6), which

constitutes 80% of all the words produced with [z] (Table 5-7). In addition, the frequency of

most of the words produced with [z] is not as high as those produced with [d]. Most of them have

very low frequency compared to the words produced with [d], which is obvious from each word

count and from the number of words that have higher frequency. However, two words produced

with [z] could be considered exceptions regarding their high frequency: izaa 'if that occurs 567

times and kazaa 'so/like this' that occurs 23 1 times. I attribute the high frequency of these two

words to their being a conditional expression and a discourse marker respectively. Consequently,

these two words occur frequently in speech.

As in the case of words with [t] and [s], this comparison will be made clearer by

constructing a contingency table that show the frequency of words pronounced with [d] and [z]

in comparison to each other. Table 5-7 shows the two-category division of frequency and the









conditional distribution on the frequency for each type of word. Table 5-7 shows that there is

some discrepancy in the conditional distribution, although it is smaller than that in the case of [t]

and [s] words. The difference in the conditional distribution indicates that frequency and the type

of sound used in a word are somewhat dependent on each other. Frequency seems to play a role

in this type of phenomenon. Frequent words are mostly pronounced with the [d] sound in

contrast to infrequent words that are mostly pronounced with the [s] sound.

Table 5-7. Contingency table showing the frequency use of [d] and [z] words in all speakers
Word Type Freq < 10 & % Freq > 10 & % Total & %

[d]-words 100 (6%) 1640 (94%) 1740 (100%)

[z]-words 255 (20%) 1002 (80%) 1257 (100%)

Total & % 355 (12%) 2642 (88%) 2997


To have a visual presentation of this comparison, I also entered all the words produced

with [d] and [z] into SPSS as two separate variables. Their frequency category was entered as a

separate categorical nominal variable with 1 assigned for low frequency (<10) and 2 for high

frequency (> 10). The result of the examination of frequencies from Analyze-Descriptive

Stati stics-Frequencies is summarized in Figures 5-3 and 5-4. Figure 5-3 shows that the difference

in number between high and low frequency words produced with [d] is insignificant. Figure 5-4

shows a great disparity between the numbers of high and low frequency words produced with

[z]. In addition, comparing the two figures reveals a great difference in frequency between the

high frequency words produced with both [d] and [z]. From the two figures, one can conclude

that the higher the frequency of the word, the higher the possibility that it is pronounced with [d].

The case of [d] and [z] words is not much different from that of [t] and [s] words, although the

low frequency words with [d] may be a little more than those with [t].







































Frequency of [d] words

Figure 5-3. The distribution of [d]-words between high and low frequency


~2-


10-


126-



100-







50-



25-


Frequency of [z] words

The distribution of [z]-words between high and low frequency


Figure 5-4.


I I










To test whether the difference in frequency between [d]-words and [z]-words is significant,

I conducted two two-tailed one-sample t-tests to compare the raw frequency of the [d]-words to

the raw frequency of [z]-words. In this case, I entered into SPSS all words produced with [d] and

[z] as two separate variables and their raw frequencies as two separate variables. The comparison

yielded a highly significant result ([d]-words frequency: t = 2.762, df= 64, p = .007; [z]-words

frequency: t = 2.063, df = 137, p = .041). In general, the mean for [d] words is 26.77, which is

much higher than the mean of [z]-words, 9.12 (Table 5-8). The high difference in mean is a good

indicator of the influence of frequency on the current use of [d]-words vs. [z]-words as a

consequence of two separate diachronic changes.

Table 5-8. Two tailed One-Sample T-Test comparing the significance of the difference in the
frequency of [d]-words and [z]-words
One-Sample Test
Test Value = 0
95% Confidence Interval of the Dfeec

t df ISig. (2-tailed) ea Difference Lower Upper

Frequency [d]-word 2.762 6.07 26.769 7.40 46.13
Feuny [z]-words 2063 137 .041 9.116 .38 17.85



5.5 Discussion of the Frequency Analysis of the Stable Variants [d] and [z]

Like in the case of [t]- and [s]-words, frequency seems to have played a role diachronically

in determining the present synchronic stable phenomenon (i.e., which words are produced with

[d] and which ones are produced with [z]). Like in the case of [t]- and [s]-words, Birkeland's

(1952), Schmidt's (1986), and Schulz's (1981) suggestions that a second change has occurred

some time after the 14th century seems to apply in the case of the use of [d] and [z]. Borrowing

words form SA with the [6] and [a] variants was followed by a replacement of those two marked


sounds with the less marked sounds, [s] and [z] respectively.









The non-frequent words with the [d] sound are related semantically to the frequent words.

They are derived from the same root and most of them are related to the word takettttt~~~~~~~tttttt and variants

of the word for this. On the other hand, the low frequency words with the sound [z] are not

necessarily semantically related to those of high frequency, but they may be semantically related

to each other. Similar to the case of [t] and [s], a close examination of these words reveals that

they mostly have the vowel system of the SA corresponding words, indicating that they are

mainly borrowings from SA. However, it seems that the high markedness of the sound [a] has

been disliked and replaced by the less marked sound [z]. In other words, speakers find it easier to

access a sound that is present in their phonemic system, /z/, and resembles the SA variant but

lacks the [+distributed] feature. Phonological reduction is not foreign to frequent words (e.g.,

Hooper 1976; Patrick 1992; Bybee 2000, 2001, 2002; Pierrehumbert 2001; Coetzee 2008). Table

5-9 presents just a few examples that show the similarity in vowel system between the observed

forms and their SA corresponding forms. Other words may have shown some assimilation to the

vowel system of the dialects. The assimilation is more obvious in the case of [z] than it is in the

case of [s]. It occurs word-internally as well as word-finally as a result of suffixing a colloquial

suffix in place of the SA suffix, for example, the colloquial feminine marker [-i] in place of the

SA feminine marker [-a] (Examples 1-3 in Table 5-10); the relational adjective marker [-ee] in

place of [-iiy] (Examples 4 & 5 in Table 5-10); or a colloquial pronominal in place of a SA

pronominal (Examples 6 & 7 in Table 5-10). The word-internal vowel assimilation to the vowel

system of the dialects are numerous. Some examples are stated in 8 & 9 in Table 5-10.

Interweaving the SA root with the colloquial dialects patterns and the existent semantic

relationship among words that are produced with one of the two variants, [d] or [z], provide

evidence that roots and patterns are stored separately in the brain because speakers could









interlock the root of the SA with a pattern from the colloquial dialects. In many cases, the root is

borrowed from SA, not the pattern.

Table 5-9. Similarity between the vowel system of observed words with [z] with their
counterparts in SA
Observed word with [s] SA corresponding word with [6] Glossary
1 mazhab maahab School of thought
2 ?itigaaz ?itigaai making (decision)
3 mafzuuriin mafauuriin They are forgiven
4 ?inzaar ?iniaar Warning
5 tanfliz tanfliia Doing a
project/performance
6 haakazaa haakaaaa Such/this way

Table 5-10. Assimilated words with [s] to the vowel system of the dialects
Observed word with [s] SA corresponding word with [6] Glossary
1 Icizaa?iiy-i Icizaaaiiy-i nutritious
2 lizz-i lii6-a Pleasure
3 zakiiy-i aakiiy-a Clever (F)
4 zakar-ee aakar-iiy Masculine
5 zak-ee aak-iiy Clever
6 btizzil-uu tayiulu-huu It let him down
7 zakkr-oo aakkar-uu They reminded
8 zooq Sawq Taste
9 mitzakkr-a mutaaakkir-a I remember (N)


5.6 General Discussion

The above frequency analysis not only informs us about the effect of frequency on the use

of certain words, but also enables us to make kind of predications about what may have

happened diachronically. It allows us to answer some, not all, of the questions about the current

stable phenomenon. It is puzzling to see that the words that are produced with the fricative

sounds, [s] and [z], are higher in number than those produced with the stops, [t] and [d]. If the

former words are borrowed from SA, we would expect a smaller number of words to be

borrowed. The data contradict that. The data give the impression that there has been some kind









of influx of SA borrowings into the dialects, but this influx stopped at a certain point in time for

some reason. The data give a sense of creeping foreign elements that were stopped by an

opposing force. This leads to the question: what is this opposing force?

We have observed one list of words for each of the sounds, [t] and [s] and [d] and [z]. The

most frequent words maintain the stops; the least frequent ones maintain the fricatives, although

few words produced with the fricatives show very high frequency. If we only take markedness or

"the emergence of the unmarked" (McCarthy and Prince 1994) into consideration, we would

expect to have more words pronounced with [t] and [d] than with [s] and [z] because fricatives

are more marked than stops (*[+cont] >> *[-cont]) cross-lingui stically (Clement 1990; Roca and

Johnson 1999; Lombardi 2003). In addition, [t] is the least marked consonant because it is

"unmarked for all features" (Frisch, Broe, and Pierrehumbert 1998:23). In other words, it is

underspecified for dorsal, labial, [-anterior], [+distributed], lateral, round, continuantnt], and

continuantnt, -strident]. All of these features add some kind of markedness to a sound. [t] is

simply characterized as a [coronal] and [-voice], which are the least marked among all the

mentioned features. For example, in English, words such as [6in] 'thin' and [ais] 'this' are

replaced by some speakers with [tin] and [dis] respectively (Labov 1972a), not [*sin] and [*sis].

One would expect more words produced with [t] and [d] not only based on markedness but also

due to the complete merger that happened in the 14th century between the SA interdentals [6] and


[a] and the stops [t] and [d] respectively (Daher, 1998a, 1999; Schmidt 1974). The situation is

the contrary; we find more words produced with the fricatives [s] and [z] as a result of the second

historical change that occurred sometime after the 14th century (Birkeland 1952; Schmidt

1986:57; Schulz 1981:33; all cited in Daher 1999). However, this change did not seem to affect

certain words that are produced with [t] and [d] (i.e., the most frequent words of all). It stopped









at a certain point, affecting mainly the least frequent words, which are supposedly borrowed

from SA and replaced by the colloquial Arabic phonology, in our case /s/ and /z/.

To explain the current split between lexical items that are totally produced with the stops

and others that are completely produced with the fricative, we may have to follow one of two

possibilities. The first possibility will be to refute the theory of a complete merger in the 14th

century between stops and interdentals (Daher, 1998a, 1999; Schmidt 1974), that is traces of

interdentals continue to surface, although not as interdentals, rather as fricatives. Refuting the

first theory leads to refuting the second theory of a second change after the 14th century, which

introduced the fricatives in place of the interdentals (Birkeland 1952; Schmidt 1986:57; Schulz

1981:33; all cited in Daher 1999). Refuting this theory will be accompanied with then suggesting

that the historical change only affected highly frequent words on the basis of usage-based models

(Bybee 2001) or exemplar-based models (Pierrehumbert 2001). These models propose that

words of higher frequency are more prone to reduction. Non-frequent words are less affected by

change and usually retain their original form or features. Thus, one can assume on these grounds

that highly frequent words, such as haadaa and ktiir, historically underwent reduction of the SA

features continuantnt] and [+distributed], but the least frequent words did not. However, there

are no traces of interdentals in natural speech. In addition, refuting this theory does not explain

the two lists of words with fricatives, such as masa~lan2 and Pizaa, instead of the SA interdentals.

These two lists also show reduction of the feature [+distributed]. The few high frequency words

with the fricatives do not behave like other highly frequent words that have also reduced the

continuantnt] feature. They seem to have resisted this type of reduction. Why would some

frequent words resist one reduction and not the other? In other words, why would some frequent

words show reduction of two features and others show reduction of only one feature?









Furthermore, if there was not a complete merger, we would expect some kind of variation and

probably a balanced number of words with the stops and the fricatives. We actually have no

variation and continue to have a large number of words produced with fricatives and less words

produced with stops. It seems this possibility cannot completely explain the present split among

stops and fricatives. Consequently, we have to take the second theory into consideration. That is,

a second change happened some time after the 14th century and led to this stable phenomenon.

Taking the second theory into account implies accepting the first theory too. Furthermore, a

support of the merger theory comes from Churchyard' s (1993:33 5) historical study. He

mentions, in a footnote, thatht *0, *., and *8 actually remained distinct in Old Aramaic is seen

from the fact that in later Aramaic these sounds merged with the dental/alveolar stops *t, *., and

*d, respectively"16. These historical sound variation and changes, according to Churchyard, may

have developed in some dialects before others and then they gained prestige to the point that they

started to gain importance and prestige in other dialects later on, leading to the spread of the

change to other dialects and eventually a merger. Since Arabic is Proto-Semitic language and

most of its sound system (twenty-one out of the twenty-eight letters in Arabic were borrowed

from the twenty-two Aramaic Alphabet) was adopted from Aramaic, a similar change and a

merger may have happened in Arabic. This seems to be a feature of Proto-Semitic languages in

which similar shifts and mergers have occurred historically and have influenced each other, as is









16 The asterisk before each of the sounds in the quotation is used to indicate a variable, not a variant. The symbol *.8
is used to refer to a dental/alveolar emphatic fricative: *. is the symbol used to refer to dental/alveolar emphatic stop.
These two symbols are represented in IPA as dand t respectively.










the case of the influence of Aramaic dental/alveolar *s sound shift ([P] -, [s]) on the Hebrew *s


sound shift ( [k] [P] and [P]- [s])"7

This brings us to the second possibility and back to the role of frequency in this stable

phenomenon. According to Bybee (2001:11), "token frequency"" has two main effects. In one

effect of frequency "phonetic change often progresses more quickly in items with high token

frequency". Pierrehumbert (2001:3) argues on the basis of Bybee' s (2000, 2001) theory of usage-

based phonology that "leniting historical changes are more advanced in frequent words than in

rarer ones." Thus, high frequency words are more likely to undergo change, mainly reduction.

For example, the deletion of [t] and [d] word-finally post-consonantally is more common in

English in highly frequent words, such as went, just, and and, than in infrequent words (Guy

1991; Santa Ana 1992; Bybee 2000; Phillips 2006; Coetzee 2008; Coetzee and Pater 2008). This

is an example of a whole segment reduction, but reduction could include feature reduction for the

purpose of minimizing articulatory effort (Saussure 1959, Martinet 1964) (i.e., easier and simpler

articulation). This first frequency effect proposed by Bybee (2001) can explain the first historical

change towards the use of [t] and [d] in place of [6] and [a] respectively. The SA interdentals

were reduced to full stops; the features [+distributed] and [+cont] were eradicated. This merger

apparently happened gradually, affecting the most frequent words first and then the least frequent

ones.

The second effect of token frequency contradicts the first effect, because "it makes items

more resistant to change" (Bybee 2001:12). In other words, high frequency words become


17 *s refers to voiceless lateral fricative. [1] stands for (alveo)palatal fricative-lateral that existed in early first
millennium B.C.E. Hebrew. [9] stands for dental/alveolar fricative-lateral that existed in early first millennium
B.C.E. Aramaic.

1s "Token Frequency" to the frequency of the occurrence of a word.









resistant to a new swiping change because they have already developed an automatized

production (p. 12). Because "tokens of use map onto existing representations, high-frequency

items grow strong and therefore are easier to access" (p. 28). Bybee (2001:12) refers to this type

of resistance as "conserving" effect, rendering some words more "conservative" in the face of

change. For her, the frequency of words

gives them a high level of lexical strength. That is, they are so engrained as individual
patterns that they are less likely to change even if general changes are occurring in the
language. To account for this entrenchment effect, I have proposed (Bybee 1985) that
representations are strengthened whenever they are accessed. This strengthening makes
them subsequently easier to access and also more resistant to some forms of change.

This second effect of frequency, as proposed by Bybee (2001), can account for the

existence of [s]- and [z]-words alongside the [t]- and [d]-words that resulted from the 14th

century merger. It seems that the second historical change that occurred sometime after the 14th

century (Birkeland 1952; Schmidt 1986:57; Schulz 1981:33; all cited in Daher 1999) did not

affect the most frequent words in the dialects (i.e., words that are produced with the stops, [t] and

[d]). We have observed in the data that the most frequent words are those produced with the

stops with few exceptions that are produced with the fricatives, [s] and [z]. The findings of this

study support Bybee' s (2001) usage-based theory, which provides explanation to the current

stable situation of the use of [t] and [s] and [d] and [z]. For Bybee (2003), "repetition" and "high

frequency" develop 'automatization" and "autonomy" that lead to "entrenchment". In this sense,

the highly frequent words produced with the stops were not affected by the second historical

change because they became entrenched in their own phonemic representation. I avoid here

using Bybee' s (2001) suggestion that words are stored as whole units in the lexicon because of

their autonomy. In a language like Arabic that has a more complex morphological system, this

may not be case. Bybee's suggestion was actually critiqued by Pierrehumbert (2002) in her

review of Bybee's (2001) book. Pierrehumbert suggests that









the case for excluding morphemes in addition to words is less firm. Even in English, and
even for infants, words are normally extracted from running speech rather than learned in
isolation; positing a mental lexicon of any and all effectively extractable chunks might
provide a better basis for the analysis of more complex morphological systems. (2002:462)

The case of how words and morphemes are stored in Arabic requires further

investigation. Suggesting that words become entrenched in their phonemic representation does

not violate the view that words could be "storage units" (Bybee 2001; Beret, Vaknin, and Marcus

2006) or morphemes could be stored separately (Feldman, Frost, and Pnini 1995; Pierrehumbert

2002).

5.7 Conclusion

In conclusion, Bybee's (2001) usage-based theory with its two frequency effects provides

an explanation to the current stable phenomenon regarding the use of the four variants [t] and [s]

and [d] and [z] in the naturally occurring speech of rural migrant speakers to the city of Hims.

The two frequency effects of the usage-based theory support the theories of two different

historical changes. It is because of those two historical changes and frequency effects that we

have two different categories of words for each of the SA interdentals. The first change led to the

loss of the two features [+distributed] and continuantnt] from the SA interdentals. The second

change began with borrowing a large number of words from SA. Instead of using the marked

sounds [6] and [a], speakers apparently sought in their native phonology similar sounds that are

less marked, [s] and [z]. This is not surprising for a people that grew accustomed to producing

much less marked sound, [t] and [d]. Borrowing the SA forms was, thus, accompanied in this

case with one feature reduction instead of two, [+distributed]. This case of seeking native

phonology is observed in an example given by Morris (2000:14) in an endnote. The example

presents the case of the Peninsular dialect of Chinato, native to the province of Caceres

(Extremadura), in which there is no underlying /s/. Instead, the dialect has only /6/ (spelled with










's', 'c', or 'z'), which is realized as [h] in syllable coda (2000:14). Because [6] is highly marked

in that dialect, it is replaced with the native dialect phonology (i.e., /h/). Although I have been

able, to a great extent, to explain the reason behind this stable phenomenon, I still think that

further research is required in this area to confirm these findings.

Although the speech sample is taken from only RCA speakers, it is my belief that the same

pattern applies to HCA speakers. It is also believed to be a characteristic of big urban centers in

Arab countries, such as Amman (Abdel-Jawad 1986). This pattern may differ in Bedouin dialects

or in other Arab dialects. In Bahrain, for example, Holes (1986) actually notes a kind of

convergence towards the SA variant [6] away from the non-standard form [f. For him, this

variation is led by social factors such as religious denominations and social identity as well as the

influence of SA through education. The [6] is used by the 'Arab, the "Sunni descendents of the

Bedouin tribes" (p. 33) whose dialect is perceived as prestigious. This example among others

calls for investigating similar phonological changes, in which the marked interdentals are used. It

seems their use and variation may depend on different factors in different settings and require

explanation from different angles.









CHAPTER 6
CONCLUSION

6.1 Introduction

In the first chapter, a number of research questions were set forth for investigation. This

chapter aims not only to give a brief summary of the answers to each of those five research

questions but also to summarize other findings that emerged throughout the study. I will first

provide an answer to each question individually, based on the findings in the various analysis

chapters. I will then provide some advantages of and caveats concerning the new model

proposed in this study. Then, I will touch on the role of frequency effects in this study. I will also

present some implications, directions and recommendations for future research. I will conclude

this chapter by emphasizing the importance of combining both the social and linguistic in one

linguistic theory.

6.2 Use of OT and the GLA to Account for Switching between Two Different Dialects and
Inter- and Intra-speaker Variation

Research question 1: How can Optimality Theory account for speakers' switch between

two different dialects and their intrapersonal and interpersonal variation in the use of particular

sounds (e.g., the use of [7] in place of [q]) based on their interlocutor' s background?

To answer this question, I developed a new model for analyzing sociolinguistic variation,

employing OT and the GLA. In this model, I proposed a number of social constraints in addition

to more conventional linguistic constraints of OT and the GLA, although the particular linguistic

constraints themselves may not have been proposed previously. I treated social constraints in the

same way that linguistic constraints are treated in the GLA. They are given a continuous scale of

ranking values and are subj ect to the continuous ranking of stochastic grammar. In other words,

their ranking values may differ from one speaker to another and from one group of speakers to









another. I have shown that the interaction of those social constraints with linguistic constraints

yields results that match real life occurrences. I have also shown that the choice of the right type

and number of social constraints determines the grammar of an individual or a group of

individuals at a certain time and place. In order for a speaker to acquire the new form, [?],

completely, s/he has to change the ranking values of two constraints of his/her mother dialect,

RCA. Speakers who can switch between the two forms [q] and [7] with one hundred percent

accuracy have acquired the new ranking values with one hundred percent accuracy, in that they

do not have overlapping among certain constraints. Thus, they sound city-like because of

acquiring the new grammar, and thus, the new form completely. In those speakers, certain social

constraints rank very high, affecting their choice of a ranking value at a certain time and place.

Speakers who show intra- and inter-speaker variation in the use of [q] and [7] with the

same interlocutor have not acquired the new ranking values of the constraints completely. The

ranking values of constraints become very close to each other, and to the constraints of their

mother dialect, leading to overlapping of constraints. This overlapping leads to variable outputs.

The degree of variation relies on the type and number of social constraints involved. For

example, a speaker or a group of speakers that uses [q] 85% and [7] 15% may have more social

constraints or different social constraints involved than a speaker or a group of speakers that uses

[q] 65% and [7] 35%. I would like to add that some social constraints might rank higher than

other social constraints in some speakers, leading also to inter-speaker variation.

OT and the GLA not only account for sociolinguistics variations, such as the variable use

of [q] and [7], but also for other types of phenomena that may occur within the same speech.

They have the advantage of accounting for two different variations independently from each









other. For example, the use of the sounds [t] and [s] and [d] and [z] is a stable situation in the

naturally occurring speech of rural migrant speakers to Hims. OT and the GLA have been able to

deal with this phenomenon separately from the sociolinguistic variation between [q] and [7]. One

variation does not necessarily affect the other. In the case of the variants [t] and [s] and [d] and

[z], OT and the GLA have shown that faithfulness to the input is the maj or factor that determines

the output, in contrast to the variable use of [q] and [7], in which faithfulness constraints are

easily violated.

6.3 Incorporating Social Factors into the GLA

Research question 2: How can social factors be incorporated into the GLA?

I implemented social constraints as part of OT and the GLA like linguistic constraints.

They interact with linguistic constraints to represent what goes on in the brain of speakers during

speech production because of their social surroundings. I showed that their interaction with

linguistic constraints is essential to give percentages that match real life occurrences. I treated

these social constraints like stochastic constraints on a continuous scale. The ranking value of

these social constraints may differ from one speaker or group of speakers to another. The

difference in their ranking values among speakers reflects the difference in real life percentages.

For example, the social constraint *M[7] may be higher ranked than *F[7] because women are


more socially conscious of the prestige associated with the form [7] than males are, particularly


older females. The higher ranking value of *M[7] leads to higher percentages of use of [q] by

males than by females. In addition, some social constraints form groups together and work as

one family. For example, older female speakers from Akrama should have all of the following

social constraints higher ranked than other constraints to maintain their mother dialect form [q]:










*F [7], *OLD[?], *AKR[7]. Conflict may arise when speakers employ constraints that militate

against two opposing sounds (i.e., more variation results from such implementation).

Furthermore, incorporating social constraints in the computation explains the variable use of [q]

and [7] among members of the same social groups or different groups. In addition, certain social

constraints give expectations on what the speech of a certain speaker or group of speakers should

sound like.

6.4 Role Played by Social Factors

Research question 3: To what degree do social factors play a role in building our

grammars, or rather developing new grammars?

We have noticed that the intersection of social constraints with linguistic constraints

provides a clear picture of what goes on in the mind of a learner of a new dialect (usually

prestigious urban dialects). The social constraints that are involved determine the percentages of

the occurrence of each variant in the naturally occurring speech of rural migrant speakers. The

degree of the social awareness, which is related in this study to the ranking value of the social

constraints, affects the choice of a certain variant over another. Thus, if the social constraints that

impose the use of the urban prestigious variant have high-ranking values (i.e., are highly valued

by speakers), such as *F [q] and *OLD[q], the prestigious form is more likely to be produced. If

they are ranked low, the mother dialect form is more likely to be produced. The latter situation

reflects the importance of the mother dialect social values in the speech of those speakers. In

speakers who do not show acquisition of the new form, the ranking values of their mother dialect

social constraints are fixed and do not change, particularly in the speech of older speakers. In

such situation, social constraints such as *F[7] and *OLD[?] would be highly ranked, whereas

social constraints such as *F[q] and *OLD[q] would be low ranked. I have compared this to










species that, when fully developed in their environment, are less likely to change and develop

later on in life. They maintain their features and move on with life. One should, in any case,

consider that whether speakers adopt the new ranking values of social constraints or not, these

social constraints affect the way they speak and the type of grammar chosen at a particular time

and place. If they are affected by their new environment and the new social constraints, they tend

to vary their speech and some of them my reach complete levels of acquisition of the new form.

If they are not influenced by the urban society's values, they retain their rural community social

values and continue to produce their native form almost invariably.

6.5 Consistency of Variation

Research question 4: How consistent is the pattern of variation and sound change within

and among rural migrant speakers to the city of Hims?

Concerning the two variants [q] and [7], intra- and inter-speaker variation has been

observed. Thus, some speakers showed not only variation in their speech within the same

conversation but also correction towards the new form. In addition, varying speakers differed

from each other in the degree of usage of each variant. Other speakers used either the [q] sound

or the [7] sound invariably. Thus, the variable use of [q] and [7] is not consistent. There is a clear

inter- and intra-speaker variation. We have seen that social factors play a maj or role in this

variation, particularly age. Younger speakers are more inclined towards the urban form, [?],

whereas the maj ority of older speakers are more inclined towards the [q] sound. Females are

more inclined towards the prestigious form [7] than men are. Actually, men, mainly elderly men,

are more inclined towards the [q] sound. Furthermore, residents of Al-Hameeddieh show higher

usage of [?] than residents in Akrama because of their environment, which is more linguistically









demanding in Al-Hameeddieh than it is in Akrama. This inconsistency is reflected in both the

statistical analyses and the ranking values of social and linguistic constraints.

As for the variants [t] and [s] and [d] and [z], speakers have shown consistency in their use.

All speakers young and old, males and females, upper- and middle-class, and residents of both

residential areas, Akrama and Al-Hameeddieh use a certain variant with certain words and the

other variant in other specific words. I have called this a stable variation because of the observed

consistency among speakers. This phenomenon developed historically because of two different

changes: one that led to a merger between the SA forms [6] and [6] and the stops [t] and [d]

respectively; the second led to the introduction of [s] and [z] in place of the SA forms [6] and [6]

as a result of lexical borrowings from SA.

6.6 Markedness Constraints versus Social Constraints

Research question 5: Which factors are stronger in leading a sound change: markedness

constraints or social constraints?

We have seen that in the variable use of [q] and [7], social constraints play a maj or role in

determining the grammar chosen at a particular time and place. In this type of variation, social

constraints are the ones leading the present change towards the prestigious form more than

linguistic markedness constraints do, although one cannot completely eliminate the effect of

markedness on thi s type of variation. First, the glottal stop i s less marked cross-lingui stically

than the uvular voiceless stop. It lacks the [RTR] feature that is characteristic of the [q] sound.

According to radical underspecification, the positive value of a feature is the marked one and the

negative value is the default, unmarked one (Archangeli 1984, 1988). In an endoscopic study of

tokens of /7 h/ spoken by a Jordanian Arabic speaker, Zawaydeh (1999) found that no retraction

of the tongue root accompanied these two segments. Thus, they are considered underspecified









for place, which makes them less marked than other consonants. This finding is supported by a

number of phonological analyses of laryngeals (e.g., Clements 1985; Steriade 1987; Kenstowicz

1994; Rose 1996), which have indicated that /7 h/ are produced with only a glottal gesture,

rendering them placeless. This makes Glottal the least marked place of articulation in many

languages, allowing for the following hierarchy [LAB] >> [DOR] >> [COR] >> [Glottal].

The unmarkedness of the glottal stop /7/, for example, can be observed in its use as an

epenthetic consonant in Arabic (Holes 2004) and many other languages to avoid onsetless

syllables, such as Malay (from the Austronesian language Bahasa Melayu/Indonesia as spoken in

Malaysia) (e.g., El-Imam and Don 2005). Shahin (2002:60) argues that all word-initial glottal

stops in Arabic are epenthetic and the evidence "is that they are not observed when another

consonant can serve as onset of the word-initial syllable." For example, in Palestinian Arabic,

[7ik.tib] 'write' is possible, but not *[ik.tib] or *[bi.-?ik.tib]19

The unmarkedness of [7] may have led historically to the merger of [q] with [7] (Abdel-

Jawad 1981; Haeri 1996: 122) in major urban areas, such as Hims (Habib 2005), Damascus

(Daher 1998a), Cairo (Haeri 1991), Amman (Abdel-Jawad 1981), etc. This merger, according to

Garbell 1978 [1958]) occurred sometime between the 11Ith and the 15th centuries. However, the

role of markedness in the variable use of [q] and [7] is not limited to the historical linguistic

markedness of [q] in comparison to the unmarkedness of [7]. If we look closely at the social

constraints that have been employed in this study, we find that they are mainly markedness

constraints that militate against the use of a certain variant by a speaker or a group of speakers.



19 Dots indicate syllable boundary; hyphens indicate morpheme boundary; and asterisks indicate ill-
formedne ss/ungrammaticality .









Thus, those social constraints that are pertinent to what is prohibited, allowed or expected in

certain communities are one type of markedness constraints. In the current synchronic variation,

they play a more pronounced role than linguistic markedness constraints in determining the

grammars of the various speakers. Those affected by the urban markedness constraints tend to

vary their speech. On the other hand, those who maintain their mother dialect forms show their

maintenance or clinging to the social values of their native communities, and thus are affected by

the markedness social constraints that are pertinent to their communities. This could also be

viewed as the effect of faithfulness to their social values.

As for the four variants [t] and [s] and [d] and [z], social constraints did not seem to play

any role in this phenomenon. The study has indicated that markedness may have played a role

historically in pronouncing the SA variants [6] and [6] as [t] and [d] respectively. The merger

with these two sounds was completed around the 14th century (Daher, 1998a, 1999; Schmidt

1974). In addition, the use of [s] and [z] in place of [6] and [6] of the borrowed words from SA is

another indication of the role of markedness in the second diachronic change which took place

some time after the 14th century (Birkeland 1952; Schmidt 1986:57; Schulz 1981:33; all cited in

Daher 1999). There is ample evidence from Arabic and other languages that show that the

sounds [6] and [6] are marked. Morris (2000:3), for example, indicates that the realization of

Spanish aspirated /s/ and /6/ as [h] word-finally and sometimes word-internally are cases of the


emergence of the unmarked, e.g., diez /die6/ may surface as die~h]. Today, markedness is no

longer playing a role in the use of the four variants [t] and [s] and [d] and [z]. What maintains the

pronunciation of each of the four variants as it is are faithfulness constraints. Each of the four

variants has its identical underlying form, which represents the input in OT. This input is

protected by faithfulness constraints to appear as it is in the output.









6.7 Advantages of and Caveats about the New Model

6.7.1 Advantages of the New Model

The five questions that started this work led to the development of a new model that has

the following advantages:

1. It gives mental representation of the interaction between social and linguistic

constraints and the effect of each type of constraints simultaneously. The results have

shown how the GLA can model sociolinguistic variation by incorporating the right

type and number of social constraints. The continuous approach to constraints allows

for great degrees of optionality and variation. The results show that social constraints

should constitute an important part of linguistic theory, as they play a maj or role in

determining our grammar at a certain time and place. It is social constraints that

influence our degree of acquisition of the ranking value of constraints and the degree of

usage of each of the sociolinguistic variants, [q] and [7]. Giving mental representations

of the various individual and group grammars in a sociolinguistic variation situation

endows the model and the proposed constraints with a psychological reality (Section

1.1). The second, third, and fourth advantages of the model also draw our attention to

the psychological reality of the proposed constraints and the accompanying

sociolinguistic process.

2. This new model not only presents what goes on in the human's brain when a

sociolinguistic variation is in process but also gives predications on what variation

would look like if certain constraints are involved. A listener can predict from listening

to the speech of an interlocutor what type of grammar or constraints are involved. If

that listener is a rural migrant and the interlocutor is another rural migrant who displays









some kind of variation between the rural and the urban forms, s/he will most likely

notice the interlocutor' s variable use and the social constraints involved in this

variation. For instance, a male listener to an older female speaker from Al-Hameeddieh

knows that she activates the two constraints *F[q] and *OLD[?]. These two constraints

have also opposing effects: one of them militates against the use of [q] and the other

one militates against the use of [7]. This could be a further reason for the occurrence of

variation in a certain speaker or group of speakers. Opposing social constraints may

create conflict between the social meaning associated with each variant in a speaker' s

brain, leading to overlapping of constraints and consequently variation.

3. The model further reflects on each individual's or group of individuals' linguistic

environment because the degree of input of a form affects the degree of acquisition of

that form. A speaker who has ample contact with urban interlocutors is more likely to

have higher degree of acquisition of the new form than a speaker who is surrounded

with rural speakers most of the time, as was the case when comparing Al-Hameeddieh

speakers to Akrama speakers.

4. It has also been observed during applying this new model that because of the young

speakers' early contact with the urban form, they tend to acquire it and use it all the

time. The model, in this case, predicts that the younger speakers have reconstructed

their input despite their initial exposure to their parents' rural form, [q]. In other words,

their input becomes the urban form, /7/. They restructured their input not only because


they are in contact with a new form but also because they notice the prestigious social

value of the urban form. As many may know, children are very sensitive to comments

and criticism. They do not like to be stigmatized or ridiculed by their peers or friends









because they do not sound like them. As soon as these children notice the new form

and the stigma associated with their parents' form, they decide to adopt the new form.

In their early stage of acquisition, it is easy to restructure their input and sound more

urbane than their parents are. We have seen exceptions among three young male

speakers who have lived all their lives in Akrama. Because of their less demanding

environment to learn the new form, they retain their parents' form. However, all the

other young speakers, who are likewise the sons and daughters of rural migrants, show

complete acquisition of the new form in just the span of one generation, as is the case

of the sons and daughters of rural migrant speakers to Cairo (Miller 2005).

5. The model not only accounts for sociolinguistic variation but also other accompanying

phenomena that may be led by different factors. This is an important aspect of the

GLA, its ability to account for categorical and optional ranking values. The GLA has

enabled me to account for the use of the variants [t] and [s] and [d] and [z] separately

from the sociolinguistic variation between [q] and [7]. It has shown that social

constraints do not play a role in the case of the former four variants, although they play

a role in the variable use of [q] and [7]. It has further shown that the consistent

occurrence of each variant in certain words is the result of the categorical higher

ranking of faithfulness constraints with respect to other constraints.

6. Where statistical analyses fail to indicate a specific influence of a social factor on the

variation observed, the new model provides evidence of the effect of that factor

because of the model's more refined specificity. Each social factor is divided into a

balanced number of constraints in relation to its categories and the sociolinguistic

variants. A social factor such as residential area could be divided into four social









constraints if there are two residential areas and two variants. Thus, the constraints are

calculated in a factorial manner, multiplying the number of the social factor categories

with the number of the sociolinguistic variants. This specificity allows, for example,

for only one of the four constraints to be involved in the computation. In Section 4.4.4,

I have shown that the statistical analysis fails to show significant interaction between

residential area and gender and age, regarding the variable use of [7]. However, the

results of Example 7 in Chapter 3 have shown that young female speakers from Al-

Hameeddieh must rank all three constraints *F[q], *YOUNG[q], and *HAM[q] -

higher than *7 to achieve 98% pronunciation of [7] and 2% of [q]. Thus, the social

constraint, *HAM[q], which constitutes one of the four social constraints of residential

area, is implemented in the speech of young female speakers from Al-Hameeddieh to

produce high percentages of [?].

6.7.2 Caveats about the New Model and Other Limitations of the Study

A researcher should know the speech community well to be able to discover the social

constraints that are involved in the variation and the learning process. It was feasible for me to

investigate the speech of rural migrant speakers to the city of Hims because I am a close member

of that community. I know the participants well. I know their life styles and the community's

social evaluation for each one of them. I also know the culture and the social values and behavior

associated with each linguistic form. This is not only knowledge on my part but also on the part

of many speakers who have expressed throughout their conversations the stigma associated with

[q] and the rural dialects. They also mentioned who they think tend to change their speech the

most (i.e., older females). In addition, some of them pointed out some of the reasons that rural

speakers tend to change their speech (Habib 2005, Sections 5.2 and 5.3).









The model cannot converge on a simple categorical grammar for multiple varying

speakers, which is normal because the ranking values of constraints are very close to each other,

leading to overlapping of constraints, and hence variation. Because the model cannot converge

on a strictly dominated grammar, one can only predict continuous variation and the present

percentage usage of each variant in speakers. One cannot predict if a varying speaker can acquire

the new form completely. The only way to make such prediction is to manipulate the pair

distribution in the GLA and make speakers learn from partial outputs that are considered

potential input. Changing the amount of input that a speaker receives of each variant can change

the result in the direction of convergence towards one grammar or the other. However, in the real

world the exact percentages of input are hard to determine; there are many factors that can play a

role in determining the input intake, which is very hard to determine without observing a speaker

over a very long period of time or for the rest of his/her life. Although it is beyond the scope of

this study, it is worth mentioning that I have noticed during the examination of the speech of one

older female speaker whom I recorded in 2004 and 2006 that she has progressed more, in two

years, towards the prestigious form. More observations of this type throughout upcoming years

are essential to understand and arrive at conclusions on how language change progresses in a

society.

Although this study did not cover all the possible social factors that could be involved in

the variation under investigation, it introduced a new model of sociolinguistic analysis that could

be improved and expanded on. It may be a good idea to include social factors, such as religion,

social status, and social distance in future studies. It may also be a good idea to have more than

one interviewer conduct the interviews with the same interviewees (Walters 1991). The

background of the interviewer may have influenced the way speakers speak. I would expect










speakers to use the [7] sound more frequently with a speaker who is originally from the city of

Hims, for example. Because interviewees in this study are very close relatives and friends to me,

they may have felt more comfortable using their native form, [q], with me, particularly in the

case of Speakers 29, 30, and 31. According to their sister (i.e., Speaker-44), those three young

male speakers who reside in Akrama change their speech towards the prestigious form, [7], when

conversing with urban interlocutors or strangers, particularly from the opposite sex. This was not

the case when I interviewed them. My explanation of this situation is that they felt very

comfortable talking to me in their native form, [q], because of their strong familiarity with me.

They know that I am familiar with their rural form and the way people from Oyoun speak. They

also know that I will not be critical of that rural form because my parents use it with us all the

time.

6.8 Frequency effects

Although frequency was not a part of the research questions that led this work, I felt that it

was essential to examine the influence of frequency of words on the acquisition process of the

urban form [7]. The examination of word frequency was even more important in the case of the

four variants [t] and [s] and [d] and [z], because this stable phenomenon cannot be explained by

social factors, as all speakers are consistent in using them. The results show that frequency has

varying roles with respect to the two types of variations that are dealt with in this work.

Frequency has a facilitative learning effect concerning the sociolinguistic variation between [q]

and [7]. In other words, it is not the factor that leads the change from [q] to [7]; rather, it is the

leading factor for acquiring highly frequent words faster than less frequent words which contain










the prestigious form [7]. Thus, frequency, in this case, plays a maj or role in determining which

words a varying speaker acquires initially.

As for the use of the variants [t] and [s] and [d] and [z], frequency has two opposing

effects. The first effect led to the historical merger between the SA [6] and [6] and the stops [t]

and [d] respectively. The second effect safeguarded the most frequent words that contain the [t]

and [d] from another historical sweeping change (i.e., from the use of [s] and [z] in place of the

SA [6] and [6] in borrowed words). Very frequent words develop with time some kind of

autonomy or automatization because of continuous repetition (Bybee 2001); they develop a

strong mental representation that makes them easier to access and more resistant to change.

6.9 Implications, Directions and Recommendations for Future Research

For a long time phonological theory and sociolinguistic theory acted independently from

each other. In this work, I was able to bring these two subfields together. Linguistic theory

should be general enough to account for not only the phonological aspect of grammar but also

the social aspect of that grammar. Researchers who are interested in factors other than social

factors may find the new model helpful in accounting for psychological, pragmatic, or other

factors within the same framework. I would not want to make the claim that the model is perfect

in its present shape. I believe more work and modifications may need to be done to improve the

model. It needs to be tested on other languages and other social constraints. The development of

the model does not stop at the application of the social constraints proposed in this study. Rather,

it should extend to the proposition of other social constraints that may be pertinent to different

sound changes in different speech communities.

One important implication of this study is the possibility of generalizing the proposed

social constraints to other languages. Many sociolinguistic studies have shown similar results










regarding the increased use of prestigious forms by women, younger generations, or residents of

prestigious areas (e.g., Labov 1972a; Gal 1979; Milroy 1980; Eckert 1991b; Miller 2005). These

similar results prompt me to propose some general social constraints that can be applied to any

language and any linguistic form. At the same time, there are studies that have shown

contradictory results. Thus, the ranking values of those social constraints are the determining

factor of what form is chosen at a certain time and place by a certain speaker or group of

speakers. Those ranking values, in this sense, may differ from one speech community to another

or from one language to another. Because not all speech communities evaluate sounds in the

same way, one should always take into consideration different grammars and different social

constraints. Consequently, generalizing the social constraints does not have negative effects

because those general constraints could be higher or lower ranked in one speech community or

another. The general social constraints that I propose are as follows:

Gender:

*F [SF] Stigmatized forms are marked in the speech of females

*M[SF] Stigmatized forms are marked in the speech of males

*F[PF] Prestigious forms are marked in the speech of females

*M[PF] Prestigious forms are marked in the speech of males

Age:

*OLD[SF] Stigmatized forms are marked in the speech of older speakers

*YOUNG[SF] Stigmatized forms are marked in the speech of younger speakers

*OLD[PF] Prestigious forms are marked in the speech of older speakers

*YOUNG[PF] Prestigious forms are marked in the speech of younger speakers









Social class:

*UM[SF] Stigmatized forms are marked in the speech of upper-middle class

*LM[SF] Stigmatized forms are marked in the speech of lower-middle class

*UM[PF] Prestigious forms are marked in the speech of upper-middle class

*LM[PF] Prestigious forms are marked in the speech of lower-middle class

Residential area:

*PR[SF] Stigmatized forms are marked in the speech of prestigious areas

*NR[SF] Stigmatized forms are marked in the speech of non-prestigious areas

*PR[PF] Prestigious forms are marked in the speech of prestigious areas

*NR[PF] Prestigious forms are marked in the speech of non-prestigious areas

Of course, these constraints do not form an exhaustive list of general social constraints;

rather, they are just a subset of the possible general social constraints that can be proposed. One

can even modify these general ones to include more constraints if one is dealing with more than

two variants of the same variable. If there are three variants, then the social factor could be

divided into six constraints. Furthermore, stigmatized forms and prestigious forms in the

constraints' description could be replaced with the respective sounds under investigation. One

can also take into consideration the length of stay in the city and other factors such as religion,

education, etc.

Proposing general social constraints raises the issue of the universality of the

sociolinguistic behavior of speakers across languages. This is a wide topic to explore in future

research. Questions such as the following may occur. Are human beings endowed with a

universal set of social constraints? Do they eliminate or reactivate some of these constraints

because of their observations of the societal behavior around them? In other words, do they










acquire the ranking of these constraints in the same way they acquire the ranking of their L1

linguistic constraints? And do they acquire these social constraints in the same way an L2 learner

acquires the ranking of the linguistic constraints of an L2?

I propose that humans may have universal social constraints in the same way that they

have a universal set of linguistic constraints. Similar to how they acquire their mother tongue by

eliminating or reranking markedness and faithfulness linguistic constraints that are pertinent to

their L1, speakers acquire their social skills from their surroundings and tend to eliminate some

social constraints and rerank social constraints that are pertinent to their environment. Thus, like

linguistic constraints that are universal in nature but their ranking with respect to each other is

language specific, social constraints could be universal in nature and their ranking is language,

dialect, or community specific. In other words, the relativity of the social constraints is

maintained in the specificity of individual grammars and constraint rankings of particular

speakers or communities of speakers. When speakers learn a second language or dialect, they

attempt to change the ranking of those social constraints or reactivate some dormant social

constraints to fit into their new environment. This is true pragmatically. For example, an L2

speaker has to learn the pragmatics of the L2 in order to sound native-like and to integrate into

the new culture. Thus, the idea of universal social constraints is worth exploring in future

research.

One further suggestion for future research is applying the new model to examine the

percentage of usage of [q] and [7] in other Arab urban communities (within or outside Syria)

against those found in this study. If percentages match, then there is a similar trend in most Arab

urban communities towards not only the activation of the same social constraints but also the

same ranking values. If percentages do not match, then there might be difference in the ranking









values of the same social constraints from one community to another or from one country to

another. In addition, the difference among Arab countries with respect to the use of various

variants as a replacement for [q] calls for further studies to examine those usages against the

usage of other variants in other Arab countries to look for similar or different trends.

The variants [t] and [s] and [d] and [z] are not socially conditioned in the speech of most

Syrian speakers. However, this may not be the case in other Arab countries. Thus, further studies

are needed to investigate if the naturally occurring speech of other Arab dialects has socially

conditioned variations in the use of these four variants. I presented the example of Bahrain

(Holes 1986), in which the SA variant [6] is considered socially and colloquially prestigious and

there is a kind of convergence towards this form away from the non-standard form [f. In this

case, social constraints could be included in the GLA to account for the sociolinguistic variable

use of [6] and [f].

In addition to all the above suggestions, longitudinal observational studies are a good

source of data to make future predictions in the GLA about changes in progress. Such

longitudinal observations could complement this study, in that they could provide the usage

percentages of each variant in varying speakers throughout a number of years. The results may

be informative about the directionality of variation and the change in the input intake of each

variant if there is any. If changes in usage percentages differ throughout a period of time, then

we can conclude that some learnability is taking place despite the speakers' surrounding varying

linguistic environment. Such results may also be informative about the speakers' increased or

decreased social contact with speakers of the prestigious form.

Based on my personal observations and the comments of some speakers, I believe it is

important to trace the sound change of [q] into [7] in some villages of Wadi Al-Nasara because









of contact with urban speakers through marriage or increased commuting between these villages

and urban areas. I believe some villages may show greater use of [7] than [q] within the village


itself. Furthermore, some villages may show greater use of [7] or [q] than other villages. The


spread of the use of [?] to the villages and the difference in spread among these villages is worth

investigating. Such research should focus on children from all ages that live in the villages and

receive their education in the schools of those villages to observe the outside influence on them. I

mean by outside influence any of the following: going to visit relatives in urban areas or [7]-

speaking areas; mothers from outside the villages that have been married to men in the villages

and brought with them the prestigious form [7]; and the children's contact with relatives visiting


from [?]-speaking area. In such studies, not only children who undergo such outside influences

should be investigated but also other village children who are not directly affected by an outside

influence. They may have minor contact with [7] through exposure to television stations that use


the [7] sound instead of [q] or through contact with other village children who are exposed to

those outside influences. Starting with observing village children may lead to predicting the

future pronunciation of the villages. In other words, we may be able to foresee a change in

progress in the villages themselves. I believe there is a change in progress in the villages towards

the urban prestigious form but this change has different degrees. In one or two villages of Wadi

Al-Nasara, such as Marmarita, there is obvious difference between the older and younger

generation regarding the use of [q] and [7]. This issue will be reserved for future research that I

believe can give us fascinating results.









6.10 Conclusion

This work has presented a new model for analyzing sociolinguistic variation. It is my hope

that this model will be put to use by other researchers to investigate what goes on in the brain of

a speaker who varies his/her speech for social reasons. In the same way that this work has

brought two linguistic subfields closer together, phonologists should start accepting social factors

as part of linguistic theory, and sociolinguists should not distance themselves from theoretical

models to explore variation throughout the world. Interacting social constraints with linguistic

constraints within one theoretical framework allows us to observe the simultaneous workings of

those constraints. In some variations, social constraints are involved and required. In other

variations, no social constraints are involved or required. Both situations can be accounted for in

OT and the GLA. The GLA allows us to deal with different phenomena independently from each

other. For example, vowel variation within the same speaker may act independently from

consonant variation. Habib (in progress) has shown that the variable use of [q] and [7] may or


may not be accompanied by certain vowel changes. Some speakers may choose to vary between

[q] and [7] but use the rural dialect vowel system. Other speakers may vary in both their use of


[q] and [7] and their use of the rural and urban vowel systems. Furthermore, some speakers may

show complete acquisition of both systems. Finally, the social constraints proposed in this study

do not necessarily play a role in all types of variations. One should examine all possible causes

of a certain phenomenon, such as frequency effects and linguistic, internal and psychological

forces.











APPENDIX A
TESTS OF ASSOCIATION FOR SOCIAL CLASS WITH EDUCATION, INCOME,
OCCUPATION, AND RESIDENTIAL AREA



Cross Tabulations (Contingency Tables)



Case Processing Summary

Cases

Valid 1Vhssing Total

N Percent N Percent N Percent

SoilClass Education 52 100.0% 01 .0%1 521 100.0

SoilClass Income 52 100.0%1 01 .0%1 52 100.0

SoilClass Occupation 52 100.0% 01 .0%1 521 100.0

SoilClass Area 52 100.0% 01 .0%1 521 100.0


Crosstab

Social Class

Lower-middle Upper-middle Total

Count Adjusted Residual Count Adjusted Residual Count

Euain Elementary 3 .9 11 -.94

Middle school 5 .5 31 -.58

High school 41 -.6 5 .69

Associate degree 8 1.8 2 -1.8 10

Bachelor's degree 61 -1.3 9 1.3 15

Professional 21 -1.1 4 1.16

Total 281 24 52


Social Class Education












Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 6.207 5 .8

Likelihood Ratio 6.494 5 .261

Linear-by-Linear Association 2.006 1 .157

Nof Valid Cases 52

a. 9 cells (75.0%) have expected count less than 5. The minimum expected count is 1.85.


Symmetric Measures

Value Asymp. Std. Error" Approx. Tb Approx. Sig.

Inevlby Interval Pearson's R .198 .133 1.431 .1590

Ordinal by Ordinal Spearman Correlation .207 .135 1.499 .1400

Nof Valid Cases 52

a. Not assuming the null hypothesis.

b. Using the asymptotic standard error assuming the null hypothesis.


Crosstab

Income

Low IVid High Total

SoilClass Lower-middle Count 22 6 0 2

Adjusted Residual 5.41 -.6 -5.2

Upper-middle Count I 1 71 16 24

Adjusted Residual -5.4 .6 5.2

Total Count 23 13 16 52


c. Based on normal approximation.




Social Class Income












Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 35.1510 2 .0

Likelihood Ratio 45.608 2 .0

Linear-by-Linear Association 34.464 1 .0

Nof Valid Cases 52

a. O cells (.0%) have expected count less than 5. The minimum expected count is 6.00.



Symmetric Measures

Value Asm.Std. Erroral Approx. Tb Approx. Sig.

Inevlby Interval Pearson's R .822 .053 10.209.00

Ordinal by Ordinal Spearman Correlation .822 .055 10.193 .0000

Nof Valid Cases 52

a. Not assuming the null hypothesis.

b. Using the asymptotic standard error assuming the null hypothesis.

c. Based on normal approximation.


Crosstab

Social Class

Lower-middle Upper-middle Total

Count Adjusted Residual Count Adjusted Residual Count

Occupation unemployed 71 -.7 81 .71 15

Government employee 11 3.0 11 -3.0 12

Private work/business I 1 -1.2 3 1.24

Professional 9 -1.3 12 1.3 21

Total 28 24 52


Social Class Occupation












Chi-Square Tests

Value df Asymp. Sig. (2-sided)

Pearson Chi-Square 9.578 3 .023

Likelihood Ratio 10.98 3 .012

Linear-by-Linear Association 1.028 1 .311

Nof Valid Cases 52

a. 2 cells (25.0%) have expected count less than 5. The minimum expected count is 1.85.


Symmetric Measures

Value Asymp. Std. Error" Approx. Tb Approx. Sig.

Inevlby Interval Pearson's R .142 .140 1.014 .315e

Ordinal by Ordinal Spearman Correlation .114 .144 .811 .4210

Nof Valid Cases 52

a. Not assuming the null hypothesis.

b. Using the asymptotic standard error assuming the null hypothesis.

c. Based on normal approximation.


Crosstab

Area

Al-Hameeddieh Akrama Total

SoilClass Lower-middle Cotmt 16 12 28

Adjusted Residual -3.2 3.2

Upper-middle Cotmt 23 1 24

Adjusted Residual 3.21 -3.2

Total Coumt 39 13 52


Social Class Residential Area












Chi-Square Tests

Asymp. Sig. (2- Exact Sig. (2- Exact Sig. (1-
Value df sided) sided) sided)

Pearson Chi-Square 10.317 1 .001

Continuity Correctionb 8.357 1 .004

Likelihood Ratio 11.926 1 .001

Fisher's Exact Test .001 .001

Linear-by-Linear Association 10.119 1 .001

Nof Valid Cases 52

a. O cells (.0%) have expected count less than 5. The minimum expected count is 6.00.

b. Computed only for a 2x2 table



Symmetric Measures

Value Asymp. Std. Error" Approx. Tb Approx. Sig.

Inevlby Interval Pearson's R -.445 .099 -3.518 .001c

Ordinal by Ordinal Spearman Correlation -445 .099 -3.518 .001c

Nof Valid Cases 52

a. Not assuming the null hypothesis.

b. Using the asymptotic standard error assuming the null hypothesis.

c. Based on normal approximation.












APPENDIX B
DESCRIPTIVE STATISTICS EXAMINING TYPE OF DISTRIBUTION INT THE DATA


Gender


Case Processing Summary

Cases

Valid 1Vhssing Total

Gender N Percent N Percent N Percent

qfFemale 972 100.0%1 01 .0%1 972 100.0

Male 15061 100.0%1 01 .0% 15061 100.0

hmaFemale 972 100.0%1 01 .0%1 972 100.0

Male 15061 100.0%1 01 .0%1 15061 100.0




Descriptives

Gender Statistic Std. Error


laf Female


83.7016

82.1671

85.2362

86.4959

100.0000

594.381

24.37992

1.00

100.00

99.00

25.00

-1.498

1.504

91.5458

90.6798

92.4118

94.2044

99.0000


.7819$


Mean

95% Confidence Interval for Mean Lower Bound

Upper Bound

5% Trimmed Mean

Median

Variance

Std. Deviation

Minimum

Maximum

Range

Interquartile Range

Skewness

Kurtosis

Mean

95% Confidence Interval for Mean Lower Bound

Upper Bound


.4414$


Male


5% Trimmed Mean

Median













Variance 293.534

Std. Deviation 17.13285

Minimum 1.00

Maximum 100.00

Range 99.00

Interquartile Range 6.00

Skewness -3.266 .063

Kurtosis 12.083 .2

namza Female Mean 16.30 .782

95% Confidence Interval for Mean Lower Bound 14.76

Upper Bound 17.83

5% Trimmed Mean 13.50

Median .00

Variance 594.381

Std. Deviation 24.380

1Vhnimum 0

Maximum 99

Range 99

Interquartile Range 25

Skewness 1.498 .078

Kurtosis 1.5041 .155

Male Mean 8.45 .441

95% Confidence Interval for Mean Lower Bound 7.59

Upper Bound 9.32

5% Trimmed Mean 5.80

Median 1.00

Variance 293.534

Std. Deviation 17.133

Minimum 0

Maximum 99

Range 99

Interquartile Range 6

Skewness 3 .266 .063

Kurtosis 12.083 .2













M-Estimatorse

Huber's M-

Gender Estimatora Tukey's Biweightb Hampel's M-Estimatore Andrews' Waved

qf Female

Male 98.7437 99.6968 99.1258 99.74

hma Female

Male 1.31 .23 .84.2

a. The weighting constant is 1.339.

b. The weighting constant is 4.685.

c. The weighting constants are 1.700, 3.400, and 8.500

d. The weighting constant is 1.340*pi.

e. Some M-Estimators cannot be computed because of the highly centralized distribution around the median.

Extreme e Values

Gender Case Number Value

qf Female Highest 1 14 100.00

2 15 100.00

3 18 100.00

4 24 100.00

5 271 100.00

Lowest 1 50 1.00

2 45 1.00

3 44 1.00

4 40 1.00

5 51 2.00a

Male Highest 1 2 100.00

2 4 100.00

3 71 100.00

4 10 100.00

5 12 100.00b

Lowest 1 371 1.00

2 35 1.00

3 33 2.00

4 39 3.00





hma Female Highest 1 40 9

2 44 9

3 45 9

4 50 9

5 4898

Lowest 1 27

2 240

3 180

4 150

5 140

Male Highest 1 35 9

2 379

3 33 9

4 39 9

5 34 9

Lowest 1 310

2 130

3 120

4 100



a. Only a partial list of cases with the value 2.00 are shown in the table of lower extremes.

b. Only a partial list of cases with the value 100.00 are shown in the table of upper extremes.

c. Only a partial list of cases with the value 98 are shown in the table of upper extremes.

d. Only a partial list of cases with the value 0 are shown in the table of lower extremes.


Tests offorniality

Kolmogorov-Smirnov" Shapiro-Wilk<

Gender Statistic df Sig. Statistic df Sig.

qf Fenude .351 972 .000 .707 972 .0

Male .340 1506 .000 .535 1506 .0

hma Female .351 972 .000 .707 972 .0

Male .340 1506 .000 .535 1506 .0

a. Lilliefors Significance Correction




















for Gender= Female


0.00 20 .00 40.00 60.00 80 .00 100.00

qaf


U


'


20 .00 40.00 60.00 80 .00 100.00

qalf


qaf: Gender Histograms


-- Normal
Mean =83.70
Std. Dev. =24.38
N =972


500.0.


400.0_


S300.0
a

S200.0






000


-Normal
Mean =91.55
Std. Dev. =17.133
N =1,506


600.0-




QI 400.0




20000




0 0
0.00


aRF~RFR~f


for Gender= Male


















5


65






16








50 40 33
Female Male


Gender


hamza: Gender Histograms







for Gender= Female


~i~--RtR--R~


qaf: Gender Boxplots


1t00.00-



8. 0

80.00-



-000

60.00'



40.00-


-Normal
Mean =16.3
Std. Dev. =24.38
N =972


000

600.00'



S400.00-
-
e

2 00.00-


100.00-



0.00-


20 40 60

hamza


80 100

















for Gender= Male


00 I I nMYR
0 20 40 60 80 100


4550
51 4440 37 3435
0 "36
490 41 38@
21 3



16

















Female Male


----- Normal
Mean =8.45
Std. Dev. =17.133
N =1,506


000

800.00'


"



u*r 400.00-



200.00-


hamza: Gender Boxplots


hamza


Gender
















Case Processing Summary

Cases

Valid Missing Total

Age N Percent N Percent N Percent

qf Young 341 100.0% 0 .0% 341 100.0%

Old 21371 100.0% 01 .0%1 21371 100.0

hma Young 341 100.0% 0 .0% 341 100.0%

Old 21371 100.0% 0 .0% 21371 100.0%




Descriptives

Age Statistic Std. Error


1 I


Age


laf Young Mean

95% Confidence Interval for Mean Lower Bound

Upper Bound

5% Trimmed Mean

Median

Variance

Std. Deviation

Minimum

Maximum

Range

Interquartile Range

Skewness

Kurtosis


80.1085

76.5439

83.6731

83.3392

96.0000

1119.903

33.46495

1.00

100.00

99.00

12.00

-1.727

1.123


1.81223























.132

.263


Old Mean

95% Confidence Interval for Mean Lower Bound

Upper Bound

5% Trimmed Mean

Median

Variance

Std. Deviation


89.8030

89.0648

90.5412

92.0959

100.0000

302.835

17.40216


.37644













Minimum 8.00

Maximum 100.00

Range 92.00

Interquartile Range 22.00

Skewness -1.8611 .053

Kurtosis 2.947 .106

namza Young Mean 19.89 1.812

95% Confidence Interval for Mean Lower Bound 16.33

Upper Bound 23.46

5% Trimmed Mean 16.66

Median 4.00

Variance 1119.903

Std. Deviation 33.465

Minimum 0

Maximum 99

Range 99

Interquartile Range 12

Skewness 1.727 .132

Kurtosis 1.123 .263

Old Mean 10.20 .376

95% Confidence Interval for Mean Lower Bound 9.46

Upper Bound 10.94

5% Trimmed Mean 7.90

Median .00

Variance 302.835

Std. Deviation 17.402

Minimum 0

Maximum 92

Range 92

Interquartile Range 22

Skewness 1.8611 .053

Kurtosis 2.9471 .106












M-Estimatorse

Huber's M-

Age Estimatora Tukey's Biweightb IHampel's M-Estimatore Andrews' Waved

qf Young 94.1673 95.5674 95.0705 95.5673

Old

hma Young 6.46 4.60 5.00 4.6

Old

a. The weighting constant is 1.339.

b. The weighting constant is 4.685.

c. The weighting constants are 1.700, 3.400, and 8.500

d. The weighting constant is 1.340*pi.

e. Some M-Estimators cannot be computed because of the highly centralized distribution around the median.


Extreme Values

Age Case Number Value

qf Young Highest 1 31 100.0

2 30 96.0

3 29 88.0

4 32 10.00

5 36 10.00

Lowest 1 50 1.0

2 45 1.0

3 44 1.0

4 40 1.00

5 371 1.00a

Old Highest 1 2 100.0

2 4 100.0

3 71 100.0

4 10 100.00

5 12 100.00b

Lowest 1 21 8.0

2 16 34.0

3 19 43.0












4 20 47.00

5 9 61.00

hma Young Highest 1 35 99

2 371 99

3 40 99

4 44 99

5 45 99C

Lowest 1 31 0

2 30 4

3 29 12

4 36 90

5 32 90

Old Highest 1 21 92

2 16 66

3 19 5

4 20 53

5 9 39

Lowest 1 271 0

2 24 0

3 18 0

4 15 0

5 14 0d

a. Only a partial list of cases vith the value 1.00 are shove in the table of lover extrennes.

b. Only a partial list of cases with the value 100.00 are shown in the table of upper extremes.

c.COnly a partial list of cases vith the value 99 are shove in the tabke of upper extrennes.

d. Only a partial list of cases with the value 0 are shown in the table of lower extremes.












Tests of Normality

Kolmogorov-Smirnov" Shapiro-Wilk<

Age Statistic df Sig. Statistic df Sig.

qf Young .426 3411 .000 .569 341 .0

O*1d .336 21371 .0001 .650 2137.0

hma Young .426 3411 .000 .569 341 .0

O*1d .336 21371 .000 .650 2137.0

a. Lilliefors Significance Correction


qaf: Age Histograms


Normal
Mean =80.11
Std. Dev. =33.465
N =341


qazf





































I m- gal le li ac" '' '


41 3621


45

Young Old

Age


Normal
Mean =B9.BO
Std. Dev. =17.402
N = 2,1 37


1200.00;


1 000.00~


800 .00-


600 .00-


400 .00


200 .00-


20.00 40.00 60.[00

gaf


80.00


1 o;o


qaf: Age Boxplots


100.00-



600-

80.00-








20.00-



O. ~


for Age= Old


m~F~Ffi~TR




















for Age= Young


for Age= Old


hamza: Age Histograms


-- Normal
Mean =19.89
Std. Dev. =33.465
N =341


200.00'




150.00-


-
0* 100.00-
a

-

50.00-


U.UU


n


20 40 60

hamza


I I
80 100


SNormal
Mean =10.2
Std. Dev. =17.402
N =2,137


1250.00~


1000.00~
r

"I 750.00-


500 .00>


250 .00-


U.uu


I r
60 80


m


h -in-T~-m


20 40

hamzax















45
,04044 2





16
O
19






******


Young Old

Age




Area


Case Processing Summary

Cases

Valid 1Vhssing Total

Area N Percent N Percent N Percent

qfAl-Hameeddieh 1649 100.0% 01 .0% 1649 100.0%

Akrama 829 100.0% 0 .0% 829 100.0%

hma Al-Hameeddieh 1649 100.0% 01 .0% 1649 100.0%

Akrama 829 100.0% 01 .0% 829 100.0%


hamza: Age Boxplots


80-




60~


-

40-




20-













Descriptives

Area Statistic Std. Error


5% Trimmed Mean

Median

Variance


laf Al-Hameeddieh


Mean

95% Confidence Interval for Mean Lower Bound

Upper Bound


85.9885

84.8948

87.0822

89.1575

99.0000

512.742

22.64381

1.00

100.00

99.00

23.00

-1.918

3.420

93.4029

92.3979

94.4079

96.0271

96.0000

217.313

14.74155

1.00

100.00

99.00

5.00

-3.691

14.681

14.01

12.92

15.11

10.84

1.00

512.742


.55762
























.060

.120

.51200
























.085

.170


5% Trimmed Mean

Median

Variance

Std. Deviation

Minimum

Maximum

Range

Interquartile Range

Skewness

Kurtosis


Akrama


Mean

95% Confidence Interval for Mean Lower Bound

Upper Bound


5% Trimmed Mean

Median

Variance

Std. Deviation

Minimum

Maximum

Range

Interquartile Range

Skewness

Kurtosis


namza Al-Hameeddieh


Mean

95% Confidence Interval for Mean Lower Bound

Upper Bound













Std. Deviation 22.644

Minimum 0

Maximum 99

Range 99

Interquartile Range 23

Skewness 1.918 .060

Kurtosis 3.4201 .120

Akrama Mean 6.60 .512

95% Confidence Interval for Mean Lower Bound 5.59

Upper Bound 7.60

5% Trimmed Mean 3.9

Median 4.00

Variance 217.313

Std. Deviation 14.742

Minimum 0

Maximum 99

Range 99

Interquartile Range 5

Skewness 3.691 .085

Kurtosis 14.681 .170




M-Estimators

Huber's M- Hampel's M-

Area Estimatora Tukey's Biweightb Estimatore Andrews' Waved

qf Al-Hameeddieh 98.9315 99.8099 99.7124 99.810

Akrama 97.0764 97.4494 97.3084 97.448

hma Al-Hameeddieh 1.07 .19 .29 .19

Akrama 2.92 2.52 2.68 2.52

a. The weighting constant is 1.339.

b. The weighting constant is 4.685.

c. The weighting constants are 1.700, 3.400, and 8.500

d. The weighting constant is 1.340*pi.












Extreme Values

Area Case Number Value


Lowest 1 441 1.0(

2 331 2.0(

3 341 4.0(

4 201 47.0(

5 291 88.0(

Highest 1 351 95

2 371 95

3 401 95

4 451 95

5 501 95

Lowest 1 27 (

2 24

3 18

4 13

5 121 0

Highest 1 441 95


laf Al-Hameeddieh


Highest 1

2

3

4

5

Lowest 1

2



4



Highest 1

2

3

4

5


100.00

100.00

100.00

100.00

100.00"


100.00

100.00

100.00

100.00

96.00"


Akrama


namza Al-Hameeddieh


Akrama












2 339





5 29 12

Lowest 1 310

2 15

3 14





a. Only a partial list of cases with the value 100.00 are shown in the table of upper extremes.

b. Only a partial list of cases with the value 96.00 are shown in the table of upper extremes.

c. Only a partial list of cases with the value 0 are shown in the table of lower extremes.

d. Only a partial list of cases with the value 4 are shown in the table of lower extremes.



Tests offfornality

Kolmogorov-Smirnov" Shapiro-Wilk<

Area Statistic df Sig. Statistic df Sig.

qf Ad-Hmneeddieh .322 1649 .000 .672 1649 .0

Adxama .372 829 .000 .455 829 .0

hma Ad-Hmneeddieh .322 1649 .000 .672 1649 .0

Adxama .372 829 .000 .455 829 .0

a. Lilliefors Significance Correction




















for Area= Al-Hameeddieh


Do I F- I I I '
0.00 20.00 40.00 60.00 80 .00 100.00

gafb







for Arlea= Akrama


qaf: Area Histograms


-- Normal
Mean =85.99
Std. Dev. =22.644
N =1,649




























-Normal
Mean =93.40
Std. Dev. =14.742
N =829


80. 0

800.00-


-



r* 400.00-
a

-

200.00-


400.0_




30000




U* 200.0




10000




0.0*


I I I a
0.00 20.00 40.00 60.00


I
80.00


100.00


~FA~R~

















miliinia









20


16




36
32 2
"~"841 J434

37 44


for Area= Al-Hameeddieh

1000.0_



800.00



e 60000
au


u.. 4000



200.0


qaf: Area Boxplots


100.00




80.00~



400-

60.00-




40.00


Al-Halmeeddieh


Akrrame


Area


hamza: Area Histograms


-- Normal
Mean =14.01
Std. Dev. =22.644
N =1,649


U UV


Ilioa'
20 40 60


80 1~0









































00 j" I[~ "
0 20 40 60 80 100





for Area= Akrama


Normal
Mean =6.6
Std. Dev. =14.742
N =829


40. 0


400.00-


-



U* 200.00-

L.


100.00-


hamza: Area Boxplots


37 40 50 44
5139 93845 4
418 21
3236



16


20


Al-Hameeddieh


Akrame


Area


hiamzax
















Case Processing Summary

Cases

Valid 1Vhssing Total

Social Class N Percent N Percent N Percent

qf Lower-middle 1540 100.0% 0 .0% 1540 100.0

Upper-middle 938 100.0% 01 .0%1 938 100.0

hma Lower-middle 1540 100.0% 0 .0% 1540 100.0

Upper-middle 938 100.0% 0 .0% 938 100.0



Descriptives

Social Class Statistic Std. Error


1 I


Social Class


laf Lower-middle Mean


87.6260

86.5691

88.6829

90.5130

99.0000

447.096

21.14464

1.00

100.00

99.00

12.00

-2.052

3.698


.53882























.062

.125


95% Confidence Interval for Lower Bound

Mean Upper Bound

5% Trimmed Mean

Median

Variance

Std. Deviation

Minimum

Maximum

Range

Interquartile Range

Skewness

Kurtosis


Upper-middle Mean

95% Confidence Interval for Lower Bound

Mean Upper Bound

5% Trimmed Mean

Median

Variance

Std. Deviation


89.8529

88.5894

91.1163

92.9295

100.0000

388.783

19.71758


.64380





Minimum

Maximum

Range

Interquartile Range

Skewness

Kurtosis


1.00

100.00

99.00

23.00

-2.780

8.628

12.37

11.32

13.43

9.49

1.00

447.096

21.145

0

99

99

12

2.052

3.698

10.15

8.88

11.41

7.07

.00

388.783

19.718

0

99

99

23

2.780

8.628


namza Lower-middle


Mean

95% Confidence Interval for Lower Bound

Mean Upper Bound

5% Trimmed Mean

Median

Variance

Std. Deviation

Minimum

Maximum

Range

Interquartile Range

Skewness

Kurtosis

Mean

95% Confidence Interval for Lower Bound

Mean Upper Bound

5% Trimmed Mean

Median

Variance

Std. Deviation

Minimum

Maximum

Range

Interquartile Range

Skewness

Kurtosis


Upper-middle












M-Estimatorse

Huber's M- Hampel's M-

Social Class Estimatora Tukey's Biweightb Estimatore Andrews' Waved

qf Lower-middle 98.5813 99.3888 99.1391 99.3970

Upper-middle

hma Lower-middle 1.93 .39 .83 .39

Upper-middle

a. The weighting constant is 1.339.

b. The weighting constant is 4.685.

c. The weighting constants are 1.700, 3.400, and 8.500

d. The weighting constant is 1.340*pi.

e. Some M-Estimators cannot be computed because of the highly centralized distribution around the median.



Extreme TV lues

Social Class Case Number Value

qf Lower-middle Highest 1 2 100.0

2 4 100.0

3 71 100.0

4 15 100.0

5 18 100.00

Lowest 1 44 1.0

2 40 1.0

3 33 2.0

4 34 4.0

5 41 8.00

Upper-middle Highest 1 10 100.0

2 12 100.0

3 13 100.0

4 14 100.0

5 24 100.00

Lowest 1 50 1.0

2 45 1.0

3 371 1.0












4 35 1.00

5 51 2.00"

hma Lower-middle Highest 1 40 99

2 44 99

3 33 98

4 34 96

5 21 92d

Lowest 1 31 0

2 18 0

3 15 0

4 71 0

5 4 O"

Upper-middle Highest 1 35 99

2 371 99

3 45 99

4 50 99

5 48 98'

Lowest 1 271 0

2 24 0

3 14 0

4 13 0

5 12 Oe

a. Only a partial list of cases with the value 100.00 are shown in the table of upper extremes.

b. Only a partial list of cases with the value 8.00 are shown in the table of lower extremes.

c. Only a partial list of cases with the value 2.00 are shown in the table of lower extremes.

d. Only a partial list of cases with the value 92 are shown in the table of upper extremes.

e. Only a partial list of cases with the value 0 are shown in the table of lower extremes.

f. Only a partial list of cases with the value 98 are shown in the table of upper extremes.












Tests of Normality

Kolmogorov-Smirnov" Shapiro-Wilk

Social Class Statistic df Sig. Statistic df Sig.

qf Lower-middle .341 15401 .000 .6441 154 .0

Upper-middle .336 9381 .0001 .5661 938.0

hma Lower-middle .341 15401 .000 .6441 154 .0

Upper-middle .336 9381 .000 .5661 938.0

a. Lilliefors Significance Correction


qaf: Social Class Histograms






for Social C~lass- Lower-rniddle


-- Normal
Mean =87.63
Std. Dev. =21 .145
N =1,540


600.0_


500.0_





2 0000


1r 0000


0.00 20.00


40.00 60.00

gaf


80.00


100.00


~L~S~rrr







































"" lgr w- 'gl re arge'''


23




20
19*:

16




2132 36
4 1
t3430

4440 33 45 35


-- Normal
Mean =B9.B5
Std. Dev. =19.718
N =938


000

600.00~


5 00.00-




rr 300.00-


200.00-


1OD-


0.00 20100


40 .00 60.00 80.00 100d.00

q~af


qaf: Social Class Boxplots


-100.00-



&000

80.00-





40.00;



20.00



ao-


Lower-mriddle


Upper-m~iddle


Social Class


for Social Class- Upper-middle


~P~fft~R




















for Social Class- Lower-middle


UU~~lT -m
0 20 40 60 80 100

hamza







for Social Class- Upper-middle


hamza: Social Class Histograms


--Normal
Mean =12.37
Std. Dev. =21 .145
N =1,540




























Normal
Mean =10.15
Std. Dev. =19.718
N =938


000

800.00-


-



r* 400.00-
a

-

200.00-


U.


000

600.00~


5 00.00-




400.00-








10 .0


S-Ri-I
20 40 60

hiamzax


I I
80 100











hamza: Social Class Boxplots


1004 37453
44*21 50 51
$41 3g3
32 36


16

60- 19
N *2





20-







Lower-middle Upper-middle

Social Class











APPENDIX C
GENERALIZED LINEAR MODELS RESULTS FOR THE DEPENDENT VARIABLE [q]


Poisson Loglinear Regression Showing the Main Effects of the Independent Variables, Age,
Gender, Residential Area, and Social Class, on the Use of the Dependent Variant [q]



Model Information

Dependent Variableqa

Prbability Distribution IPoisson

Link FunctionLo


Case Processing Summary

N Percent

Inluded 521 100.0

Excluded 0.0

Total 521 100.0


Categorical Variable Infor nation

N Percent

Factor Age Old 28 53.8%

Young 24 46.2%

Total 52 100.0%

Gender Male 24 46.2%

Female 28 53.8%

Total 52 100.0%

Area Akrama 13 25.0%

Al-Hameeddieh 39 75.0%

Total 52 100.0%

Social Class Upper-middle 24 46.2%

Lower-middle 28 53.8%

Total 52 100.0%












Continuous Variable Information


I I N Minimum Maximum Mean Std. Deviation
Dependent Variable qaf 52 () 467 112.96 132.696


Goodness of Fitd

Value df Value/df

Deviance4798.992 471 1()2.1()

Scaled Deviance 39.659 47

Pearson Chi-Square 5687.238 471 121.()()

Scaled Pearson Chi-Square 47.()()( 47

LgLikelihood" -2527.8)

AdutdLog Likelihoode -2().89()

Akie's Information Criterion (AIC) 5()65.613

Finite Sample Corrected AIC (AICC) 5()66.918

Bayesian Information Criterion (BIC) 5()75.369

Consistent AIC (CAIC) 5()8().369

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Social Class

a. The full log likelihood function is displayed and used in computing information criteria.
I I I I I
c. The adjusted log likelihood is based on an estimated scale parameter and is used in the
model fitting omnibus test.

d. Information criteria are in small-is-better foma.


Omnibus Testa

Likelihood Ratio

Chi-Square Df Sig.

28.()42 4.)))

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Social Class

a. Compares the fitted model against the intercept-only model.












Tests of Model Effects

Type III

Likelihood Ratio

Source Chi-Square df Sig. F dfl df2 Sig.

(Intercept) 117.806 1 .000 117.806 1 47 .0

Ae 22.268 1 .000 22.268 1 47 .0

Gender 4.238 1 .040 4.238 1 47 .045

Area 2.843 1 .092 2.843 1 47 .9

SoilClass .379 1 .538 .379 1 47 .541

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Social Class



Param ster Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypothesis Test Interval for Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper

(Intercept) 2.995 .4833 2.04 3.942 38.393 1 .000 19.982 7.749 51.53

[ge=2] 1.623 .3996 .80 2.406 16.503 1 .000 5.069 2.317 11.094

[Age=1] On .. 1

[Gender-2] .62 .2970 .020 1.184 4.104 1 .043 1.825 1.020 3.6

[Gender-1] On .. 1

[Area=2] .583 .3386 -00 1.247 2.967 1 .085 1.792 .923 3.48

[Aea=1] On .. 1

[SocClass=2] .201 .3252 -437 .838 .381 1 .537 1.222 .646 2.312

[SocClass=1] On .. 1

(Scale) 1.210E2

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Social
Class

a. Set to zero because this parameter is redundant.

b. Computed based on the Pearson chi-

square.





Model Information


DpnetVariableQa

Prbability Distribution INegative binomial (1)

Link FunctionLo


Case Processing Summary

N Percent

Inluded 52 100.0

Excluded 0.0

Total 52 100.0%


S2 000-
e

a
*
5 1 000-


-

.000-


m
-100


80o







o 8


a o
o


a 0

oB


2.000


3.000 4.000 5.000
Predicted Value of Linear Predictor


Negative Binomial Regression to Show the Main Effects of Gender, Age, Residential Area,
And Social Class on the Use of [q]












Categorical Variable Infor ration

N Percent

Factor Age Old 281 53.8%

Young 241 46.2%

Total 52 100.0%

Gender Male 241 46.2%

Female 281 53.8%

Total 52 100.0%

Area Akrama 13 25.0%

Al-Hameeddieh 391 75.0%

Total 521 100.0%

Social Class Upper-middle 241 46.2%

Lower-middle 281 53.8%

Total 521 100.0%


Continuous Variable Information


I I N Mmnimum Maximum Mean Std. Deviation
Dependent Variable qaf 521 0 4671 112.96 132.696



Goodness of Fitd

Value df Value/df

Deviance 90.806 47 1.932

Scaled Deviance 56.8371 47

Pearson Chi-Square 75.090 47 1.598

Scaled Pearson Chi-Square 47.000 47

LgLikelihood" -265.034

Adjusted Log Likelihoode -165.890

Akie's Information Criterion (AIC) 540.068

Finite Sample Corrected AIC (AICC) 541.373

Bavesian Information Criterion (BIC) 549.825

Consistent AIC (CAIC) 554.825












Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Social Class

a. The full log likelihood function is displayed and used in computing information criteria.
I I I I I
c. The adjusted log likelihood is based on an estimated scale parameter and is used in the model
fitting omnibus test.

d. Information criteria are in small-is-better form.


Omnibus Testa

Likelihood Ratio

Chi-Square df Sig.

41.313 4 .0

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Social Class

a. Compares the fitted model against the intercept-only model.


Tests of Model Effects

Type III

Likelihood Ratio

SouceChi-Square df Sig. F dfl df2 Sig.

(Intercept) 2016.8921 11 .000 2016.892 1 47.0

Age 37.4781 11 .000 37.478 1 47.0

Gender 7.515 1 .006 7.515 1 47.0

Area 7.7941 11 .0051 7.7941 1 47.0

SoilClass .0301 1 .862 .030 1 47.6

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Social Class












Parameter Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypothesis Te t Interval for Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper

(Intercept) 1.681 .48101 .738 2.623 12.2071 11 .0001 5.3681 2.091 13.78

[ge=2] 2.8661 .42271 2.037 3.694 45.963 11 .0001 17.5641 7.670 40.221

[Age=1] OaI .I .I I I 1

[Gender-2] 1.117 .41941 .295 1.939 7.093 1 .0081 3.055 1.343 6.951

[Gender-1] OaI .I .I I I 1

[Aea=2] 1.481 .54901 .405 2.557 7.2791 11 .0071 4.3981 1.500 12.901

[Area=1] OaI .I .I I I 1

[SocClass=2] -.0721 .41461 -.8851 .7401 .0301 11 .8621 .9301 .413 2.09

[SocClass=1] OaI .I .I I I 1

(Scale) 1.598b

negativee



Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Social
Class

a. Set to zero because this parameter is
redundant.

b. Computed based on the Pearson chi-square.



Estimated Marginal Means 1: Age


Estimates

95% Wald Confidence Interval

Age Mean Std. Error Lower Upper

Old 333.42 107.7271 122.28 544.56

Young 18.981 5.502 8.20 29.7












Individual Test Results

ge Simple Contrast Contrast Estimate Std. Error Wald Chi-Square df Sig.

LvlYoung vs. Level Old -314.441 107.581 8.543 1 .003



Overall Test Results



IWald Chi-Square df Sig.
8.543 1 .003


Estimated Marginal Means 2: Gender


Estimates

95% Wald Confidence Interval

Gender Mean Std. Error Lower Upper

Mae 139.06 39.340 61.96 216.17

Female 45.51 14.88 16.34 74.69



Individual Test Results



|Gender Simple Contrast Contrast Estimate Std. Error Wald Chi-Square df Sig.
LvlFemale vs. Level Male -93.55 41.2171 5.152 1 .023


Overall Test Results



IWald Chi-Square df Sig.
5.152 1 .023


Estimated Marginal Means 3: Area


Estimates

95% Wald Confidence Interval

Area Mean Std. Error Lower Upper

Akrama166.85 73.833 22.141 311.56

Al-Hameeddieh 37.93 8.811 20.661 55.20












Individual Test Results

AraSimple Contrast Contrast Estimate Std. Error Wald Chi-Square df Sig.

LvlAl-Hameeddieh vs. Level
-128.92 76.523 2.838 1 .9




Overall Test Results



IWald Chi-Square df Sig.
2.838 1 .9


Estimated Marginal Means 4: Social Class


Estimates

95% Wald Confidence Interval

SoilClass Mean Std. Error Lower Upper

Upper-middle 76.74 26.856 24.10 129.38

Lower-middle 82.48 20.625 42.05 122.9



Individual Test Results

SoilClass Simple Contrast Contrast Estimate Std. Error Wald Chi-Square df Sig.

Level Lower-middle vs. Level
5.74 32.613 .031 1 .6
Upper-middle


Overall Test Results



IWald Chi-Square df Sig.
.031 1 .6








































3.00 'I II
2.000 4.000 6 000 8.000
Predicted Value of Linear Predictor


Model Information


Dependent Variableqa

Prbability Distribution INegative binomial (1)

Link FunctionLo


Case Processing Summary

N Percent

Inluded 52 100.0%,

Excluded 0.0

Total 52 100.0%


2.000-


o of

d 8

o 4 o
o o


o o a o

o 8

o
B0


~1.000-


-





- r
-1 .000-


E

-2.000-


Negative Binomial Regression Showing the Effect of the Interaction of Three Independent
Variables, Age, Gender, and Residential Area, on the Use of the Dependent Variant [q]












Categorical Variable Infor ration

N Percent

Factor Age Old 28 53.8%

Young 24 46.2%

Total 52 100.0%

Gender Male 24 46.2%

Female 28 53.8%

Total 52 100.0%

Area Akrama 13 25.0%

Al-Hameeddieh 39 75.0%

Total 52 100.0%

Social Class Upper-middle 24 46.2%

Lower-middle 28 53.8%

Total 52 100.0%


Continuous Variable Information


I I N Minimum Maximum Mean Std. Deviation
Dependent Variable qaf 52 0 467 112.96 132.696



Goodness of Fitd

Value df Value/df

Deviance 69.598 44 1.58

Scaled Deviance 82.842 44

Pearson Chi-Square 36.965 44 .4

Scaled Pearson Chi-Square 44.000 44

LgLikelihood" -254.430

Adjusted Log Likelihoode -302.849

Akie's Information Criterion (AIC) 524.860

Finite Sample Corrected AIC (AICC) 528.209

Bavesian Information Criterion (BIC) 540.470

Consistent AIC (CAIC) 548.470












Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Age Gender Area

a. The full log likelihood function is displayed and used in computing information criteria.
I I I I I
c. The adjusted log likelihood is based on an estimated scale parameter and is used in the model fitting
omnibus test.

d. Information criteria are in small-is-better form.


Omnibus Testa

Likelihood Ratio

Chi-Square df Sig.

103.808 7.0

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Age Gender Area

a. Compares the fitted model against the intercept-only model.


Tests of Model Effects

Type III

Likelihood Ratio

SourceChi-Square Df Sig. F dfl df2 Sig.

(Intercept) 487.2431 11 .000l 487.243 1 44 .0

Age 41.012 11 .0001 41.012 1 44 .0

Gender 17.383 11 .0001 17.383 1 44 .0

Area 3.778 11 .0521 3.778 1 441 .058

Ae *Gender *Area 25.302 41 .0001 6.326 4 44 .0

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Age Gender Area

a. The validity of the likelihood ratio chi-square is uncertain because log-likelihood convergence was not achieved for the
constrained model. Results shown are based on the last iteration.












Parameter Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypothesis Test Interval for Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper


1.830

4.541



2.319



1.328





1.808





.305





2.482










4.779


3.364

41.845

1

3.815

1

.743

1



.559





.390





1.622





1





16.649





1





1





1


:Intercept)

[Age=2]

[Age=1]

[Gender-2]

[Gender-1]

[Area=2]

[Area=1]

[Age=2] *

[Gender-2] *

[Area=2]

[Age=2] *

[Gender-2] *

[Area= 1]

[Age=2] *

[Gender-1] *

[Area=2]

[Age=2] *

[Gender-1] *

[Area= 1]

[Age=1] *

[Gender-2] *

[Area=2]

[Age=1] *

[Gender-2] *

[Area=1]

[Age=1] *

[Gender-1] *

[Area=2]

[Age=1] *

[Gender-1] *

[Area= 1]


.3148

.4118



.5000



.8290





1.2195





.6363





1.0195










1.0036


.596

2.927



.359



-1.921





-2.972





-2.189





-1.515










.845


14.851

82.214



7.172



.128





.228





2.190





.225










7.853


1.815

18.668



1.432



.146





.051





.112





.220










2.329


6.234

93.794



10.166



3.773





6.099





1.357





11.962










119.023











Scale) .840b

negative binomial)1

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Area, Age *
Gender Area

a. Set to zero because this parameter is redundant.

b. Computed based on the Pearson chi-square.


I


Case Processing Summary

N Percent

Inluded 52 100.0

Excluded 0.0

Total 52 100.0%


Regression to Show the Effect of the Interaction of Gender and Age on
the Use of [q]



Model Information



Negative binomial (1)


Dependent Variable

Probability Distribution

Linkr Function


Negative Binomial












Categorical Variable Infor ration

N Percent

Factor Age Old 28 53.8%

Young 24 46.2%

Total 52 100.0%

Gender Male 24 46.2%

Female 28 53.8%

Total 52 100.0%

Area Akrama 13 25.0%

Al-Hameeddieh 39 75.0%

Total 52 100.0%

Social Class Upper-middle 24 46.2%

Lower-middle 28 53.8%

Total 52 100.0%


Continuous Variable Information


I I N Minimum Maximum Mean Std. Deviation
Dependent Variable qaf 52 0 467 112.96 132.696


Goodness of Fitd

Value df Value/df

Deviance 84.326 481 1.757

Scaled Deviance 82.778 4

Pearson Chi-Square 48.898 481 1.019

Scaled Pearson Chi-Square 48.000 4

LgLikelihood" -261.794

Adjusted Log Likelihoode -256.988

Akie's Information Criterion (AIC) 531.588

Finite Sample Corrected AIC (AICC) 532.440

Bavesian Information Criterion (BIC) 539.393

Consistent AIC (CAIC) 543.393












Dependent Variable: qaf
Model: (Intercept), Age, Gender, Age Gender

a. The full log likelihood function is displayed and used in computing information criteria.
I I I I I
c. The adjusted log likelihood is based on an estimated scale parameter and is used in the model
fitting omnibus test.

d. Information criteria are in small-is-better form.


Omnibus Testa


Tests of Model Effects

Type III

Likelihood Ratio

SourceChi-Square Df Sig. F dfl df2 Sig.

(Intercept) 1702.422"1 1 .000 1702.422 1 48 .0

Age 54.524 1 .000 54.524 1 48 .0

Gender 35.057 1 .000 35.057 1 48 .0

Age *Gender 22.061 1 .000 22.061 1 48 .0

Dependent Variable: qaf
Model: (Intercept), Age, Gender, Age Gender

a. The validity of the likelihood ratio chi-square is uncertain because log-likelihood convergence was not achieved for the
constrained model. Results shown are based on the last iteration.


Dependent Variable: qaf
Model: (Intercept), Age, Gender, Age Gender

a. Compares the fitted model against the intercept-only model.












Parame er Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypcthesis Test Interval for Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper

(Intercept) 1.173 .3203 .545 1.801 13.402 1 .000 3.231 1.724 6.053

[ge=2] 3.815 .4135 3.004 4.625 85.095 1 .000 45.356 20.1671 102.003

[Age=1] OaI .I .I I I 1

[Gender-2] 3.199 .4432 2.330 4.068 52.108 1 .000 24.509 10.282 58.417

[Gender-1] OaI .I .I I I 1

[ge=2] *
-2.8271 .58611 -3.9761 -1.678 23.269 1 .000 .0591 .019 .18
[Gender-2]

[ge=2] *
[Gender-1]

[ge=1] *
[Gender-2]

[ge=1] *
OaI I. I. I .I 1
[Gender-1]

(Scale) 1.019b

(Negative



Dependent Variable: qaf
Model: (Intercept), Age, Gender, Age Gender

a. Set to zero because this parameter is redundant.

b. Computed based on the Pearson chi-square.






















Model Information

Dependent Variablehaz

Prbability Distribution INegative binomial (1)

Link FunctionLo


Case Processing Summary

N Percent

Inluded 52 100.0%,

Excluded 0.0

Total 52 100.0%


Categorical Variable Infor ration

N Percent

Factor Age Old 281 53.8%

Young 241 46.2%

Total 52 100.0%

Gender Male 241 46.2%

Female 281 53.8%

Total 52 100.0%

Area Akrama 13 25.0%

Al-Hameeddieh 391 75.0%

Total 521 100.0%

Social Class Upper-middle 241 46.2%

Lower-middle 281 53.8%

Total 521 100.0%


APPENDIX D

GENERALIZED LINEAR MODELS RESULTS FOR THE DEPENDENT VARIABLE [7]


Negative Binomial Regression to Show the Effect of Gender, Age, Residential Area, and
Social Class on the Use of [?]











Continuous Variable Information


I I N Mminiunn Maximum Mean Std. Deviation
Dependent Variable hamza 52 () 476 1()9.12 13().152


Goodness of Fitd

Value df Value/df

Deviance1()4.98() 471 2.23

Scaled Deviance 98.()14 47

Pearson Chi-Square 5().341 471 1.()71

Scaled Pearson Chi-Square 47.()()( 47

LgLikelihood" -268.473

AdutdLog Likelihoode -25().658

Akie's Information Criterion (AIC) 546.947

Finite Sample Corrected AIC (AICC) 548.251

Bayesian Information Criterion (BIC) 556.7()3

Consistent AIC (CAIC) 561.7()3

Dependent Variable: hamza
Model: (Intercept), Age, Gender, Area, Social Class

a. The full log likelihood function is displayed and used in computing information criteria.
I I I I I
c. The adjusted log likelihood is based on an estimated scale parameter and is used in the model
fitting omnibus test.

d. Information criteria are in small-is-better foma.


Omnibus Testa


Dependent Variable: hamza
Model: (Intercept), Age, Gender, Area, Social Class

a. Compares the fitted model against the intercept-only model.












Tests of Model Effects

Type III

Likelihood Ratio

SouceChi-Square df Sig. F dfl df2 Sig.

(Intercept) 1576.332 1 .000 1576.332 1 47.0

Age 45.865 1 .000 45.865 1 47.0

Gender 7.763 1 .005 7.763 1 47.0

Area 4.422 1 .035 4.422 1 47 .041

SoilClass .98 1 .321 .984 1 47 .32

Dependent Variable: hamza
Model: (Intercept), Age, Gender, Area, Social Class

Parameter Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypothesis Te t Interval f >r Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper

(Intercept) 6.107 .379 5.363 6.852 258.781 1 .000 449.211 213.443 945.40

[Age=2] -2.186 .3098 -2.793 -1.579 49.781 1 .000 .112 .061 .0

[ge=1] Oa .. 1

[Gender-2] -858 .3023 -1.450 -.265 8.051 1 .005 .424 .234 .767

[Gender-1] Oa .. 1

[Area=2] -.8 4130 -1.69 -078 4.615 1 .032 .412 .183 .925

[Aea=1] Oa .. 1

[SocClass=2] -.353 .3584 -1.055 .350 .968 1 .325 .703 .348 1.419

[SocClass=1] Oa .. 1

(Scale) 1.071b

(Negative
binomial)

Dependent Variable: hamza
Model: (Intercept), Age, Gender, Area, Social
Class

a. Set to zero because this parameter is
redundant.

b. Computed based on the Pearson chi-square.















Estimates

95% Wald Confidence Interval

Age Mean Std. Error Lower Upper

Old 17.681 4.445 8.97 26.4

Young 157.381 35.545 87.71 227.05


Individual Test Results

Sequential Sidak

Ae Simple Contrast Contrast Estimate Std. Error Wald Chi-Square df Sig.

LvlYoung vs. Level Old 139.701 35.1071 15.8341 1 .0



Overall Test Results



IWald Chi-Square df Sig.
15.834 1 .0


Estimated Marginal Means 2: Gender


Estimates

95% Wald Confidence Interval

Gender Mean Std. Error Lower Upper

Mae 34.35 8.046 18.58 50.12

Female 81.01 19.349 43.08 118.9



Individual Test Results

Sequential Sidak

Gender Simple Contrast Contrast Estimate Std. Error Wald Chi-Square df Sig.

LvlFemale vs. Level Male 46.65 19.546 5.6971 1 .017


Overall Test Results



IWald Chi-Square df Sig.
5.697 1 .017


Estimated Marginal Means 1: Age















Estimates

95% Wald Confidence Interval

Area Mean Std. Error Lower Upper

Akrama33.85 11.748 10.83 56.8

AlHameeddieh 82.20 14.487 53.81 110.6


Individual Test Results

Sequential Sidak
AraSimple Contrast Contrast Estimate Std. Error Wald Chi-Square df Sig.

LvlAl-Hameeddieh vs. Level
48.35 20.023 5.831 1 .016




Overall Test Results



IWald Chi-Square df Sig.
5.831 1 .016


Estimated Marginal Means 4: Social Class


Estimates

95% Wald Confidence Interval

SoilClass Mean Std. Error Lower Upper

Upper-middle 44.22 13.252 18.25 70.2

Lwer-middle 62.93 12.687 38.06 87.79



Individual Test Results

Sequential Sidak
SoilClass Simple Contrast Contrast Estimate Std. Error Wald Chi-Square df Sig.

LvlLower-middle vs. Level
18.70 18.195 1.056 1 .0
Upper-middle


Estimated Marginal Means 3: Area














Overall Test Results



IWald Chi-Square df Sig.
1.056 1 .0


Model Information


Dependent Variablehaz

Prbability Distribution INegative binomial (1)

Link FunctionLo


Case Processing Summary

N Percent

Inluded 52 100.0%,

Excluded 0.0

Total 52 100.0%


2 000-




a lao
1 00


v






-2 000~


O



O


oo
o
So o o
ooo
o



oo o
o p
oo
0o


2.000


4 000 5.000
Predicted Value of Linear Predictor


Negative Binomial Regression Showing the Effect of the Interaction of the Three

Independent Variables, Age, Gender, and Residential Area on the Use of the Dependent

Variant [?]












Categorical Variable Infor ration

N Percent

Factor Age Old 28 53.8%

Young 24 46.2%

Total 52 100.0%

Gender Male 24 46.2%

Female 28 53.8%

Total 52 100.0%

Area Akrama 13 25.0%

Al-Hameeddieh 39 75.0%

Total 52 100.0%

Social Class Upper-middle 24 46.2%

Lower-middle 28 53.8%

Total 52 100.0%


Continuous Variable Information


I I N Minimum Maximum Mean Std. Deviation
Dependent Variable hamza 52 0 476 109.121 130.152


Goodness of Fitd

Value df Value/df

Deviance 101.8081 441 2.314

Scaled Deviance 98.283 4

Pearson Chi-Square 45.5781 441 1.03

Scaled Pearson Chi-Square 44.000 4

LgLikelihood" -266.887

Adjusted Log Likelihoode -257.647

Akie's Information Criterion (AIC) 549.775

Finite Sample Corrected AIC (AICC) 553.124

Bavesian Information Criterion (BIC) 565.385

Consistent AIC (CAIC) 573.385












Dependent Variable: hamza
Model: (Intercept), Age, Gender, Area, Age Gender Area

a. The full log likelihood function is displayed and used in computing information criteria.
I I I I I
c. The adjusted log likelihood is based on an estimated scale parameter and is used in the model
fitting omnibus test.

d. Information criteria are in small-is-better form.


Omnibus Testa

Likelihood Ratio

Chi-Square df Sig.

56.678 7 .0

Dependent Variable: hamza
Model: (Intercept), Age, Gender, Area, Age Gender Area

a. Compares the fitted model against the intercept-only model.


Tests of Model Effects

Type III

Likelihood Ratio

SourceChi-Square df Sig. F dfl df2 Sig.

(Intercept) 1332.1571 11 .000l 1332.1571 144.0

Age 32.4671 11 .000 32.4671 144.0

Gender 7.665 1 .0061 7.665 144.0

Area 3.1301 11 .0771 3.130 144.8

Ae *Gender *Area 4.0801 41 .395 1.020 444.0

Dependent Variable: hamza
Model: (Intercept), Age, Gender, Area, Age Gender Area

a. The validity of the likelihood ratio chi-square is uncertain because log-likelihood convergence was not achieved for the
constrained model. Results shown are based on the last iteration.













Parameter Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypothesis Test Interval for Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper


317.826

12.268


240.364

.224

1

.905

1

1.001

1



.137





.277





.421





1





.345





1





1





1


439.157

.517


:Intercept)

[Age=2]

[Age=1]

[Gender-2]

[Gender-1]

[Area=2]

[Area=1]

[Age=2] *

[Gender-2] *

[Area=2]

[Age=2] *

[Gender-2] *

[Area= 1]

[Age=2] *

[Gender-1] *

[Area=2]

[Age=2] *

[Gender-1] *

[Area= 1]

[Age=1] *

[Gender-2] *

[Area=2]

[Age=1] *

[Gender-2] *

[Area= 1]

[Age=1] *

[Gender-1] *

[Area=2]

[Age=1] *

[Gender-1] *

[Area= 1]


5.482

-1.496

0

-.100

0

.001

0



-1.989





-1.283





-.866




0





-1.065


.3075

.4272



.5177



.7840





1.2477





.6835





1.0311











.9990


4.879

-2.334



-1.115



-1.536





-4.435





-2.623





-2.886











-3.023


6.085

-.659



.915



1.537





.456





.056





1.155











.893


131.558

.097



.328



.215





.012





.073





.056











.049


2.496



4.651





1.578





1.058





3.175











2.442


2.542





3.525





.705











1.137












Scale) 1.036b

negative binomial)1

Dependent Variable: hamza
Model: (Intercept), Age, Gender, Area, Age *
Gender Area

a. Set to zero because this parameter is redundant.

b. Computed based on the Pearson chi-square.



Negative Binomial Regression Showing the Effect of the Interaction of the Two
Independent Variables, Age and Gender, on the Use of the Dependent Variant [?]


Model Information

Dependent Variablehaz

Prbability Distribution INegative binomial (1)

Link Function ILog


Case Processing Summary

N Percent

Inluded 52 100.0%

Excluded 01 .0%

Total 52 100.0%












Categorical Variable Infor ration

N Percent

Factor Age Old 28 53.8

Young 24 46.2

Total 52 100.0

Gender Male 24 46.2

Female 28 53.8

Total 52 100.0

Area Akrama 13 25.0

Al-Hameeddieh 39 75.0%,

Total 52 100.0%

Social Class Upper-middle 24 46.2

Lower-middle 28 53.8

Total 52 100.0%


Continuous Variable Information


I I N Minimum Maximum Mean Std. Deviation
Dependent Variable hamza 52 0 476 109.12 130.152



Goodness of Fitd

Value df Value/df

Deviance 107.0271 48 2.230

Scaled Deviance 105.764 48

Pearson Chi-Square 48.573 48 1.01

Scaled Pearson Chi-Square 48.000 48

LgLikelihood" -269.497

Adjusted Log Likelihoode -266.317

Akie's Information Criterion (AIC) 546.994

Finite Sample Corrected AIC (AICC) 547.845

Bavesian Information Criterion (BIC) 554.799

Consistent AIC (CAIC) 558.799












Dependent Variable: hamza
Model: (Intercept), Age, Gender, Age Gender

a. The full log likelihood function is displayed and used in computing information criteria.
I I I I I
c. The adjusted log likelihood is based on an estimated scale parameter and is used in the model fitting
omnibus test.

d. Information criteria are in small-is-better form.


Omnibus Testa

Likelihood Ratio

Chi-Square df Sig.

52.8601 3 .0

Dependent Variable: hamza
Model: (Intercept), Age, Gender, Age Gender

a. Compares the fitted model against the intercept-only model.


Tests of Model Effects

Type III

Likelihood Ratio

SourceChi-Square df Sig. F dfl df2 Sig.

(Intercept) 2618.410 1 .000 2618.410 1 48 .0

Age 48.470 1 .000 48.470 1 48 .0

Gender 9.883 1 .002 9.883 1 481 .003

Age *Gender 2.658 11 .103 2.658 1 481 .110

Dependent Variable: hamza
Model: (Intercept), Age, Gender, Age Gender












Parame er Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypcthesis Te t Interval f >r Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper

(Intercept) 5.482 .2796 4.934 6.0301 384.505 1 .000 240.385 138.972 415.801

[ge=2] -1.619 .38351 -2.3711 -.8681 17.835 1 .000 .1981 .093 .420

[Age=1] OaI .I .I I I 1

[Gender-2] -.454 .41321 -1.264 .3561 1.205 1 .272 .635 .283 1.42

[Gender-1] OaI .I .I I I 1

[ge=2] *
-.931 .56921 -2.046 .185 2.674 1 .102 .394 .129 1.203
[Gender-2]

[ge=2] *
[Gender-1]

[ge=1] *
[Gender-2]

[ge=1] *
OaI I. I. I .I 1
[Gender-1]

(Scale) 1.012b

(Negative



Dependent Variable: hamza
Model: (Intercept), Age, Gender, Age Gender

a. Set to zero because this parameter is
redundant.

b. Computed based on the Pearson chi-square.























Model Information

DpnetVariableqf

Prbability Distribution INegative binomial (1)

Link Function ILog


Case Processing Summary

N Percent

Inluded 52 1()(.()%

Excluded ()0 .()%

Total 52 1()(.()%


Categorical Variable Information

N Percent


APPENDIX E

GENERALIZED LINEAR MODELS THE DEPENDENT VARIABLES [q] AND [7]


Negative Binomial Regression Showing the Main Effects of Education, Income, and
Occupation on the Variable Use of [q]


11.5%

28.8%

19.2%

17.3%

15.4%

7.7%

1()(.()%

3().8%

25.()%

44.2%

1()(.()%

4().4%

7.7%


Factor Education















Income


Professional

Bachelor's degree

Associate degree

High school

Middle school

Elementary

Total

High

Mid


Low

Total

Occupation Professional

Private work/business





Continuous Variable Information


I I N Mminiunn Maximum Mean Std. Deviation
Dependent Variable qaf 521 () 467 112.961 132.696



Goodness of Fitd

Value df Value/df

Deviance 132.38() 41 3.229

Scaled Deviance 57.536 41

Pearson Chi-Square 94.334 41 2.3()1

Scaled Pearson Chi-Square 41.()()( 41

LgLikelihood -285.821

AdutdLog Likelihoode -124.226

Akie's Information Criterion (AIC) 593.642

Finite Sample Corrected AIC (AICC) 6()(.242

Bayesian Information Criterion (BIC) 615.1()6

Consistent AIC (CAIC) 626.1()6

Dependent Variable: qaf
Model: (Intercept), Education, Income, Occupation

a. The full log likelihood function is displayed and used in computing information criteria.
I I I I I
c. The adjusted log likelihood is based on an estimated scale parameter and is used in the model fitting
omnibus test.

d. Information criteria are in small-is-better fom1.

Omnibus Testa


ILikelihood Ratio Chi-Square df Sig.
1().6181 1() .38

Dependent Variable: qaf
Model: (Intercept), Education, Income, Occupation

a. Compares the fitted model against the intercept-only model.


Government employee

unemployed

Total


23.1%

28.8%

1()(.()%












Tests of Model Effects

Type III

Likelihood Ratio

Source Chi-Square df Sig. F dfl df2 Sig.

(Intercept) 568.410 1 .00 568.410 1 41 .0

Education 5.542 5 .353 1.108 5 41 .371

Icme 2.291 2 .318 1.145 2 41 .2

Occupation .973 .84.329 3 41 .0

Dependent Variable: qaf
Model: (Intercept), Education, Income, Occupation



Parameter Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypcthesis Te st Interval for Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper

(Intercept) 5.062 .76471 3.563 6.560 43.816 1 .000 157.844 35.265 706.500

[Education=6] -1.950 1.21971 -4.341 .440 2.556 11 .1101 .1421 .013 1.553

[Education=5] -.728 1.00791 -2.703 1.248 .521 1 .470 .483 .0671 3.48

[Education=4] -.912 1.0938 -3.056 1.231 .696 11 .4041 .4021 .0471 3.426

[Education=3] .139 1.11341 -2.043 2.321 .016 11 .901 1.149 .130 10.189

[Education=2] .730 1.0760 -1.379 2.839 .460 11 .497 2.075 .252 17.09

[Education=1] OaI .I .I I I 1

[Income=3] -1.073 .69261 -2.431 .284 2.401 1 .121 .342 .088 1.329

[ncome=2] -.5631 .8062 -2.143 1.0171 .4871 11 .4851 .5701 .1171 2.76

[Income=1] OaI .I .I I I 1

[Occupation=4]1 .7541 .76501 -.746 2.253 .970 1 .325 2.1251 .474 9.515

[Occupation=3]1 .425 1.05121 -1.635 2.486 .164 1 .686 1.530 .195 12.011

[Occupation=2]1 .156 .70401 -1.224 1.536 .049 1 .825 1.169 .294 4.64

[Occupation=1]l OaI .I .I I I 1

(Scale) 2.301b

(Negative
binomial)






























Model Information

DpnetVariablehaz

Prbability Distribution INegative binomial (1)

Link Function ILog


Case Processing Summary

N Percent

Inluded 52 100.0%

Excluded 0 .0%

Total 52 100.0%


Dependent Variable: qaf
Model: (Intercept), Education, Income,

Occupation

a. Set to zero because this parameter is
redundant.

b. Computed based on the Pearson chi-square.


Negative Binomial Regression Showing the Main Effects of Education, Income, and
Occupation on the Variable Use of [?]


Categorical Variable Informatit n

N Percent

Factor Education Professional 61 11.5%

Bachelor's degree 151 28.8%

Associate degree 10lo 19.2%

High school 91 17.3%

Middle school 81 15.4%

Elementary 41 7.7%

Total 521 100.0%

Income High 161 30.8%

Mid 131 25.0%

Low 231 44.2%





Continuous Variable Information



I I N Minimum Maximum Mean Std. Deviation
Dependent Variable hamza 52 0 4761 109.12 130.152



Goodness of Fitd

Value df Value/df

Deviance 106.316 41 2.593

Scaled Deviance 76.209 41

Pearson Chi-Square 57.1971 41 1.395

ScldPearson Chi-Square 41.000 41

Log Likelihood" -269.141

Adjusted Log Likelihoode -192.925

Akie's Information Criterion (AIC) 560.283

Finite Sample Corrected AIC (AICC) 566.883

Baesian Information Criterion (BIC) 581.747

Consistent AIC (CAIC) 592.747

Dependent Variable: hamza
Model: (Intercept), Education, Income, Occupation

a. The full log likelihood function is displayed and used in computing information criteria.


c. The adjusted log likelihood is based on an estimated scale parameter and is used in the model fitting omnibus
test.

d. Information criteria are in small-is-better form.


Total

Occupation Professional

Private work/business

Government employee

unemployed

Total


100.0%

40.4%

7.7%

23.1%

28.8%

100.0%































Tests of Model Effects

Type III

Likelihood Ratio

Source Chi-Square df Sig. F dfl df2 Sig.

(Intercept) 510.843 1 .00 510.843 1 41 .0

Euain 31.474 5 .00 6.295 5 41 .0

Income4.352 2 .114 2.176 2 41 .126

Occupation 5.355 3 .14 1.785 3 41 .165

Dependent Variable: hamza
Model: (Intercept), Education, Income, Occupation



Parameter Estimates

95% Wald Confidence 95% Wald Confidence

Interval Hypothesis Test Interval for Exp(B)

Std. Wald Chi-

Parameter B Error Lower Upper Square df Sig. Exp(B) Lower Upper


Omnibus Testa


Dependent Variable: hamza
Model: (Intercept), Education, Income, Occupation

a. Compares the fitted model against the intercept-only model.


10.340

56.201

36.169

32.249

6.274

1.739

1

1.118

.421

1


56.732

425.144

213.394

149.509

37.032

9.830


:Intercept)

[Education=6]

[Education=5]

[Education=4]

[Education=3]

[Education=2]

[Education=1]

[Income=3]

[Income=2]

[Income=1]


2.336

4.029

3.588

3.473

1.836

.553

On

.111

-.865

On


.8685

1.0324

.9056

.7826

.9058

.8837



.5497

.5662


.634

2.005

1.813

1.940

.061

-1.179



-966

-1.975


4.038

6.052

5.363

5.007

3.612

2.285



1.189

.244


7.234

15.229

15.699

19.699

4.111

.392



.041

2.336


1.885

7.429

6.130

6.956

1.063

.308



.381

.139












[Occupation=4]1 -1.093 .5929 -2.2551 .070 3.3961 11 .0651 .3351 .105 1.072

[Occupation=3]1 .389 .74141 -1.064 1.8421 .2761 11 .6001 1.4761 .345 6.311

[Occupation=2]1 -.6671 .6469 -1.9351 .601 1.0641 11 .3021 .5131 .144 1.824

[Occupation=1]l OaI .I .I I I 1

(Scale) 1.395b

negativee



Dependent Variable: hamza
Model: (Intercept), Education, Income,

Occupation

a. Set to zero because this parameter is
redundant.

b. Computed based on the Pearson chi-square.









APPENDIX F
INTERNATIONAL PHONETIC ALPHABET (IPA) SYMBOLS FOR STANDARD ARABIC,
HIMSI COLLOQUIAL ARABIC, AND RURAL COLLOQUIAL ARABIC

Standard Arabic Phonemic System in IPA20

Bilabial Labio- (Inter) Alveolar Alveo- Palatal Velar Uvular Pharyngeal glottal
dental dentalpatl
Stop b t d k q ?
t d
Nasal mn
Affricate
d3
Fricative f 6 6 s z J X gb B h


Liquid 1

Glidewj


Himsi Colloquial Arabic Phonemic System in IPA

Bilabial Labio- Alveolar Alveo- Palatal Velar Uvular Pharyngeal glottal
dental paatal
Stop b t d k ?
t d
Nasal m n
Fricative f s z I 3 X ga B h


Liquid 1

Glide w


20 The sounds on the left within the same column are voiceless; the ones on the right are voiced. This applies to the
other two tables.









Rural Colloquial Arabic Phonemic System in IPA

Bilabial Labio- Alveolar Alveo- Palatal Velar Uvular Pharyngeal glottal
dental p alatal
Stop b t d k q ?
t d
Nasal m n
Fricative f s z I 3 X x B h


Liquid 1

Glide w










LIST OF REFERENCES


Abdel-Jawad, Hassan R. 198 1. Phonological and' social variation in Arabic in Amman.
Philadelphia, PA: University of Pennsylvania dissertation.

Abd-el-Jawad, Hassan Rashid, 1986. The emergence of an urban dialect in the Jordanian urban
centers. International Journal of the Sociology ofLanguage 6 1.5 3-63.

Abu-Haidar, Farida. 1992. Shifting boundaries: The effect of MSA on dialect convergence in
Baghdad. Perspectives on Arabic linguistics IV, ed. by Ellen Broselow, Mushira Eid, and
John McCarthy, 91-106. (Current Issues in Linguistic Theory 85). Amsterdam: John
Benj amins.

Adam, Galit. 2002. From Variable to Optimal Grammar: Evidence from language acquisition
and' language change. Tel-Aviv, Israel: Tel-Aviv University dissertation. Rutgers
Optimality Archive21 ROA-567.

Agresti, Alan, and Barbara Finlay. 1997. Statistical M~ethods for the Social Sciences, Third
Edition. Upper Saddle River, NJ: Prentice Hall International, Inc.

Al-Ani, Salman H. 1976. The development and distribution of the Arabic 'qaf' in Iraq. Essaysss~~~~ssss on
Islamic civilization, ed. by Donald P. Little, 48-55. Leiden: E. J. Brill. [Reprinted in
Readings in Arabic Linguistics (1978), ed. by Salman H. Al-Ani. Bloomington, Indiana:
Indiana University Linguistics Club, 1978].

Al-Khatib, Samer. 2008. On the directionality of emphasis spread. Paper presented at the annual
meeting of the Linguistics Society of America, Chicago.

Al-Wer, Enam. 1999. Why do different variables behave differently? Data from Arabic.
Language and' Society in the M~iddle East and North Afr~ica, ed. by Yasir Suleiman, 38-57.
London: Curzon.

Al-Wer, Enam. 2002. Education as a speaker variable. Language Contact and Language Conflict
in Arabic, ed. by Aleya Rouchdy, 41-53. London: Curzon.

Anttila, Arto. 1997a. Deriving variation from grammar. Variation, change and phonological
theory, ed. by Frans Hinskens, Roeland van Hout, and W. Leo Wetzels, 35-68.
Amsterdam: John Benjamins.

Anttila, Arto. 1997b. Variation in Finnish phonology and' morphology. Stanford, CA: Stanford
University dissertation.





21 Rutgers Optimality Archive can be accessed at this website http://roa.rutgers.edu/index.php3 .










Anttila, Arto. 2002. Variation and phonological theory. The H~andbook ofLanguage Variation
and Change, ed. by Jack K. Chambers, Peter Trudgill, and Natalie Schilling-Estes, 206-
243. Oxford: Blackwell.

Anttila, Arto, and Young-mee Yu Cho. 1998. Variation and change in optimality theory. Lingua
104.31-56.

Arad, Maya. 2003. Locality constraints on the interpretation of roots: the case of Hebrew
denominal verbs. Natural Language and Linguistic Theory 21.737-778.

Archangeli, Diana. 1984. Underspecification in Yawelmani Phonology and Molphology,
Cambridge, MA: MIT doctoral dissertation.

Archangeli, Diana. 1988. Aspects ofunderspecification theory. Phonology 5.183-208.

Aske, Jon. 1990. Disembodied rules vs. patterns in the lexicon: Testing the psychological reality
of Spanish stress rules. BLS 7.30-45.

Ayres-Bennett, Wendy. 2004. Sociolinguistic Variation in Seventeenth-Centwry France:
Methodology and Case Studies. Cambridge and New York: Cambridge University Press.

Bayley, Robert. 1994. Consonant cluster reduction in Tejano English. Language Variation and
Change 6.303-326.

Bell, Alan. 1984. Language style as audience design. Language in Society 13 (2).145-204.

Bell, Alan. 1991. The Language of News M~edia. Oxford: Blackwell.

Beret, Iris; Vered Vaknin; and Gary F. Marcus. 2007. Roots, stems, and the universality of
lexical representations: Evidence form Hebrew. Cognition 104 (2).254-286.

Bermutdez-Otero, Ricardo, and Richard M. Hogg. 2003. The actuation problem in Optimality
Theory: phonologization, rule inversion, and rule loss. In Holt, 91-119.

Birkeland, Harris. 1952. G; 1,n thr and structure ofEgyptian Arabic dialect. Oslo: Avhandlinger
utg. Av det Norske Videnskape-Acad.

Bod, Rens; Jennifer Hay; and Stefanie Jannedy (eds.). 2003. Probabilistic Linguistics, 289-341.
Cambridge, MA: The MIT Press.

Boersma, Paul, and David Weenink. 2007. Praat: a system for doing phonetics by computer
(Version 4.6.03) [Computer program]. Retrieved May 24, 2007, from
http://www. praat. org/

Boersma, Paul. 1997a. How we learn variation, optionality, and probability. The Institute of
Phonetic Sciences of the thriversity ofAmsterd amdddd~~~~~~ddddd 21.43-58.










Boersma, Paul. 1997b. Sound change in functional phonology. Rutgers Optimality Archive
ROA-237, ms.

Boersma, Paul. 1998. Functional Phonology: Formalizing the Interaction between Articulatory
and Perceptual Drives. Amsterdam: University of Amsterdam dissertation.

Boersma, Paul. 2000. Learning a grammar in Functional Phonology. In Dekkers, van
der Leeuw, and van de Weij er, 465-523.

Boersma, Paul. 2001. Praat: a system for doing phonetics by computer. Glot international
5:9/10.341-345.

Boersma, Paul, and Bruce Hayes. 2001. Empirical Tests of the Gradual Learning Algorithm.
Linguistic Inquity 32 (1).45-86.

Bybee, Joan L. 1985. Morphology as lexical organization: evidence from lexical diffusion.
Theoretical 2olphology, ed. by Michael Hammond, and Michael Noonan, 119-41. San
Diego, CA: Academic Press.

Bybee, Joan L. 2000. The phonology of the lexicon: evidence from lexical diffusion.
Usage-based models of language, ed. by Michael Barlow and Suzanne Kemmer, 65-85.
Stanford, CA: CSLI.

Bybee, Joan L. 2001. Phonology and Language Use. (Cambridge Studies in Linguistics 94).
Cambridge: Cambridge University Press.

Bybee, Joan L. 2002. Word frequency and context of use in the lexical diffusion of phonetically
conditioned sound change. Language Variation and Change 14.261-290.

Bybee, Joan L. 2003. Mechanisms of change in grammaticization: the role of frequency.
The H~andbook of Historical Linguistics, ed. by Brian D. Joseph and Richard D. Janda,
602-623. Oxford: Blackwell.

Bybee, Joan L., and James L. McClelland. 2005. Alternatives to the combinatorial paradigm of
linguistic theory based on domain general principles of human cognition. The Linguistic
Review 22.381-410.

Cameron, Deborah, 1998. Performing gender identity: Young men's talk and the construction of
heterosexual masculinity. Language and Gender. A Reader, ed. by Jennifer Coates, 270-
284. Oxford: Blackwell.

Chambers, Jack K.; Peter Trudgill; and Natalie Schilling-Estes (eds.). 2002. The H~andbook of
Language Variation and Change. Maiden, MA: Blackwell.

Chomsky, Noam. 1980. Rules and Representations. Behavioral andBrain Sciences 3. 1-61.










Churchyard, Henry. 1993. Early Arabic siin and siin in light of the proto-Semitic fricative-lateral
hypothesis. Perspectives on Arabic linguistics V: Papers f~on the Fifth Annual
Synaposiunt on Arabic Linguistics, ed. by Mushira Eid and Clive Holes, 313-342.
Amsterdam: John Benjamins.

Clements, George N. 1985. The geometry of phonological features. In Phonology Yearbook 2,
225-252. Cambridge: Cambridge University Press.

Clements, George N. 1990. The role of sonority cycle in core syllabification. Papers in
Laboratory Phonology I: Between the Grananar and Physics of Speech, ed. by John
Kingston and Mary F. Beckman, 283-333. Cambridge: Cambridge University Press.

Coates, Jennifer. 1996. Women, M~en andLanguage. London: Longman Group Limited.

Coetzee, Andries W. 2008. Phonological Variation and Lexical Frequency. Short' version to
appear in the Proceedings of the 38t Annual Meeting of the North East Linguistics
Society (NELS 38). Rutgers Optimality Archive ROA-952, ms.

Coupland, Nikolas. 1980. Style-shifting in a Cardiff work-setting. Language in Society 9. 1-12.

Coupland, Nikolas. 1985. Hark, hark the lark: Social motivations for phonological style-
shifting. Language and Conanunication 5 (3).195-72.

Coupland, Nikolas. 2001. Dialect stylization in radio talk. Language in society 30 (3).345-75.

Cutillas-Espinosa, Juan A. 2004. Meaningful variability: A sociolingui stically-grounded
approach to variation in Optimality theory. hIternational Journal of English Studies 4
(2).165-184.

Cutillas-Espinosa, Juan Antonio, and Juan Manuel Hernandez-Campoy. 2004. Audience and
speaker design in radio broadcasting: Attitudes to language as a starting script.
Universidad de Murcia, ms.

Daher, Jamil. 1998a. Linguistic variation in Damnascus Arabic: A quantitative analysis of nen's
and wonzen 's speech. New York, NY: New York University dissertation.

Daher, Jamil. 1998b. Gender in linguistic variation: The variable (q) in Damascus Arabic.
Perspectives on Arabic linguistics XI: Papers f~on the Eleventh Annual Synaposiunt on
Arabic Linguistics, ed. by Elabbas Benmamoun, Mushira Eid, and John McCarthy, 183-
206. Amsterdam: John Benjamins.

Daher, Jamil. 1999. ((theta)) and ((eth)) as Ternary and Binary Variables in Damascene Arabic.
Perspectives on Arabic linguistics XII: Papers f~on the TMI Iephl Annual Synaposiunt on
Arabic Linguistics, ed. by Elabbas Benmamoun, 163-182. Amsterdam: John Benjamins.










Davis, Stuart. 1993. Arabic pharyngealization and phonological features. Perspectives on Arabic
linguistics V, ed. by Mushira Eid and Clive Holes, 49-162. Amsterdam: John Benjamins.

Davis, Stuart. 1995. Emphasis spread in Arabic and grounded phonology. Linguistic Inquiry 26
(3).465-498.

Davis, Stuart, and Zawaydeh, Bushra A. 2001. Arabic hypocoristics and the status of the
consonantal root. Linguistic Inquiry 32.512-520.

Dekkers, Joost; Frank van der Leeuw; and Jeroen van de Weij er (eds.) 2000. Optimality Theory:
Phonology, syntax, and acquisition. Oxford: Oxford University Press.

Devitt, Michael. 2006a. ignorance ofLanguage. Oxford: Oxford University Press.

Devitt, Michael. 2006b. Intuitions in Linguistics. British Journal for the Philosophy of Science
57. 481-513.

Demuth, Katherine. 1995. Markedness and the development of prosodic structure. Proceedings
of the Twenty-Fifth Annual M~eeting of the North East Linguistics Society (NELS 25), ed.
by Jill N. Beckman, 13-25. GLSA: University of Massachusetts, Amherst.

Eckert, Penelope. 1989. Jocks and Burnouts: Social Categories and Identity n the High School.
New York: Teachers College Press.

Eckert, Penelope (ed.) 1991a. New Ways of Analyzing Sound Change. San Diego: Academic
Press.

Eckert, Penelope. 1991b. Social polarization and the choice of linguistic variants. New Ways of
Analyzing Sound Change, ed. by Penelope Eckert, 213-232. San Diego: Academic Press.

Eisikovits, Edna. 1987. Sex differences in inter-group and intra-group interaction
among adolescents. Women and Language in Australian and New Zealand Society, ed. by
Anne Paulwels 45-58. Sydney: Australian Professional Publications.

Eisikovits, Edna. 1988. Girl-talk/boy-talk: Sex differences in adolescent speech. Australian
English, ed. by Peter Collins and David Blair, 35-54. New York: University of
Queensland Press.

El-Imam, Yousif A., and Zuraidah Mohd Don. 2005. Rules and Algorithms for Phonetic
Transcription of Standard Malay. IEICE TRANS. INF. and SYST. E88-D (10). 2354-2372.

Encyclopedia Britannica. 2008. Hims. Retrieved January 27, 2008, from Encyclopedia
Britannica Online: http://www.britannica.com/eb/article-90401










Feldman, Laurie Beth; Ram Frost; and Tamar Pnini. 1995. Decomposing words into their
constituent morphemes: Evidence from English and Hebrew. Journal ofExperintental
psychology: Lea-l~rning. Mentory, and Cognition 21 (4).947-960.

Ferguson, Charles. A. 1997. Grammatical agreement in Classical Arabic and the modern
dialects: A response to Versteegh' s pidginization hypothesis. Structuralist studies in
Arabic linguistics: Charles A. Ferguson 's papers, 1954-1994, compiled by R. Kirk
Belnap and Niloofar Haeri, 81-91. Leiden and New York: E. J. Brill.

Finegan, Edward, and Douglas Biber. 1994. Register and Social Dialect Variation: An Integrated
Approach. Sociolinguistic Perspectives on Register, ed. by Douglas Biber and Edward
Finegan, 315-347. New York: Oxford University Press.

Fischer, John L. 1958. Social influences on the choice of a linguistic variant. Word 14.47-56.

Frisch, Stefan; Michael Broe; and Janet Pierrehumbert. 1998. Similarity and phonotactics
in Arabic. ROA-223.

Gal, Susan, 1978. Peasant men can't get wives: Language change and sex roles in bilingual
community. Language in Society 77 (1).1-16.

Gal, Susan. 1979. Language Shift: Social Determinants of Linguistic Change in Bilingual
Austria. New York: Academic Press.

Garbell, Irene. 1958. Remarks on the historical phonology of an Eastern Mediterranean
Arabic dialect. Word 14. 303-337. [Reprinted in Readings in Arabic Linguistics, ed. by
Salman Al-Ani, 204-241. Bloomington, Indiana: Indiana University Linguistics Club,
1978].

Giles, Howard; Nikolas Coupland; and Justine Coupland. 1991. Accommodation Theory:
Communication, context and consequence. Contexts ofAcconanodation: Developments in
Applied Sociolinguistics, ed. by Howard Giles, Justine Coupland, and Nikolas Coupland,
1-68. Cambridge: Cambridge University Press.

Giles, Howard, and Peter F. Powesland. 1975. Speech Style andSocialEducation. London:
Academic Press in association with the European association of Experimental Social
Psychology.

Gilford, Henry. 1978. Syria, A First Book. New York; London: Franklin Watts.

Gnanadesikan, Amalia E. 1995. Markedness and Faithfulness Constraints in Child Phonology.
Ms., Rutgers Optimality Archive ROA-67.

Gundel, Jeanette, and Gillian Sankoff. 2008. Pragmatics and sociolinguistics: Still in the
courtship phase. Paper presented at annual meeting of the Linguistics Society of America.











Guy, Gregory R. 1991. Explanation in variable phonology: an exponential model of
morphological constraints. Language Variation and Change 3.1-22.

Habib, Rania. 2005. The role of social factors, lexical borrowing and speech acconanodation in
the variation of [q] and [~ ]in the colloquial Arabic of rural migrant families in Hints,
Syria. Gainesville, FL: The University of Florida MA thesis.

Hachimi, Atiqa. 2001. Shifting sands: Language and gender in Moroccan Arabic. Gender across
Languages, Volume I, IM~PACT: Studies in Language and Society 9, ed. by Marlis
Hellinger and Hadumod Bussmann, 27-5 1. Amsterdam: John Benj amins.

Hachimi, Atiqa. 2005. Dialect Leveling, Maintenance and Urban Identity in Morocco. Manawa,
Hawaii: University of Hawaii dissertation.

Hachimi, Atiqa. 2007. Becoming Casablancan: Fessis in Casablanca as a Case Study. Arabic
in the City: Issues in Dialect Contact and Variation, ed. by Catherine Miller, Enam Al-
Wer, Dominique Caubet, and Janet C.E. Watson, 97-122. London: Routledge-Taylor and
Francis.

Haeri, Niloofar. 1991. Sociolinguistic Variation in Cairene Arabic: Palatalization and the Qafin
the Speech of2~en and Women. Philadelphia, PA: University of Pennsylvania Ph.D.
dissertation.

Haeri, Niloofar. 1996. The sociolinguistic market of Cairo: gender, class, and education.
London/New York: Kegan Paul International; Distributed by Columbia University Press.

Haeri, Niloofar. 2003. Second language, ordinary people: Dilenanas of culture and politics in
Egypt. New York: Palgrave Macmillan.

Hayes, Bruce. 2000. Gradient well-formedness in Optimality Theory. In Dekkers, van der
Leeuw, and van de Weij er, 88-92.

Hayes, Bruce, and May Abad. 1989. Reduplication and syllabification in Ilokano. Lingua
77.331-374.

Hayes, Bruce, and Margaret MacEachern. 1998. Quatrain form in English folk verse. Language
74.473-507.

Hernandez-Campoy, Juan M., and Peter Trudgill. 2002. Functional compensation and
southern peninsular Spanish /s/ loss. Folia Linguistica Historica XXIII. 141-167.

Holes, Clive. 1986.The social motivation for phonological convergence in three Arabic
dialects. hIternational Journal of the Sociology ofLanguage 6 1.33-5 1.










Holes, Clive. 2004. Modern Arabic: Structures, Functions, and Varieties. Washington, D.C.:
Georgetown University Press.

Holt, David Eric, 2003. Remarks on Optimality Theory and language change. In Holt, 1-30.

Holt, David Eric (ed.). 2003. Optimality Theory and Language Change. Dordrecht:
Kluwer Academic Publishers.

Homs City Council. 2008. lama f an madinat himp (A glance at the city of Hims). Retrieved
February 4, 2008, from http ://www.homscitycouncil .org.sy/modules/0/3 .php

Hooper, Joan B. 1976. Word frequency in lexical diffusion and the source of morphological
change. Current Progress in Historical Linguistics, ed. by William M. Christie, 95-105.
Amsterdam: North-Holland Publishing Co.

Hume, Elizabeth and Naomi Nagy. 2008. Phonology. Paper presented at the annual meeting of
the Linguistics Society of America.

Hymes, Dell. 1972. The Ethnography of Speaking. Anthropology and Human Behavior, ed. by
Thomas Galdwin and William C. Sturdevant, 15-53. Washington, D.C.: Anthropological
Society of Washington.

Ibrahim, H. Muhammad. 1986. Standard and prestige language: A problem in Arabic linguistics.
Anthropological Linguistics 28.115-126.

Johnson, Keith, and Nancy Niedzielski. 2008. Phonetics, sociophonetics, and the phonetics-
phonology interface. Paper presented at the annual meeting of the Linguistics Society of
America.

Kager, Rene. 1999. Optimality Theory. New York, NY: Cambridge University Press.

Kenstowicz, Michael.1994. Phonology in Generative Grammar. Oxford: Blackwell Publications

Kisseberth, Charles. 1970. On the functional unity of phonological rules. Linguistic Inquiry
1.291-306.

Kochetov, Alexei. 1998. Labial Palatalization: A Gestural Account of Phonetic Implementation.
The Canadian Linguistic Association Annual Proceeding. 38-50.

Kroch, Anthony. 1989. Reflexes of grammar in patterns of language change. Language
Variation and Change 1 (3).199-244.

Kroch, Anthony. 1994. Morphosyntactic variation. The ParasessionPP~~PP~~~PP~~PP on Variation in Linguistic
Theory, ed. by Katherine Beals, Jeanette Denton, Robert Knippen, Lynette Melnar,
Hisami Suzuki, and Erica Zeinfeld 180-201. CLS 30 (2), Chicago.










Labov, William. 1963. The social motivation of a sound change. Word 19.273-309.

Labov, William. 1966. The Social Stratification of English in New York City. Washington:
Center for Applied Linguistics.

Labov, William. 1972a. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press.

Labov, William, 1972b. Language in the Inner City: Studies in the Black English Vernacular.
Philadelphia: University of Pennsylvania Press.

Labov, William. 1994. Principles of Linguistic Change. Volume I: Internal Factors. Oxford:
Blackwell.

Labov, William. 2001. Principles of Linguistic Change. Volume II: Social Factors. Oxford:
Blackwell.

Lambert, Wallace E.; R. C. Hodgson; Robert D. Gardner; and Samuel Fillenbaum. 1960.
Evaluational Reactions to Spoken Languages. Journal ofAbnormal and Social
Psychology 60.44-51.

Liberman, Mark. 1994. Optionality and optimality. Fragment of a draft, Department of
Linguistics, University of Pennsylvania.

Lombardi, Linda. 2003. Second language data and constraints on Manner: Explaining
substitution for English interdentals. Second Language Research 19(3).225-250.

Macaulay, Ronald K. S. 1977. Language, Social Class and Education. Edinburgh University
Press.

Macaulay, Ronald K. S. 1978. Variation and consistency in Glaswegian English. In Trudgill,
132-143.

Mahayni, Riad G. 1990. A Decentralized Approach to Urbanization: The Case of Syria. Journal
ofPlan2ning Education and Research 9 (2). 117-125.

Manning, Christopher D. 2003. Probabilistic syntax. Probabilistic Linguistics, ed. by Rens Bod,
Jennifer Hay, and Stefanie Jannedy, 289-341. Cambridge: The MIT Press.

Martinet, Andre. 1964. Elements of General Linguistics. Chicago: University of Chicago Press.

McCarthy, John J., and Alan Prince. 1994. The emergence of the unmarked: Optimality in
prosodic morphology. Proceedings of the North East Linguistic Society 24, ed. by
Gonzalez Merce, 333-79. Amherst, MA: GLSA Publications. [Reprinted in Optimality
Theory in Phonology: A Reader, ed. by McCarthy, John J. Maiden, MA: Blackwell,
2004].










McMahon, April M. S. 2000. Lexical Phonology and the History ofEnglish. Cambridge:
Cambridge University Press.

Medoza-Denton, Norma, Jennifer Hay, and Stefanie Jannedy. 2003. Probabilistic
Sociolinguistics: Beyond Variable Rules. In Bod, Hay, and Jannedy, 97-138.

Miller, Catherine. 2005. Between accommodation and resistance: Upper Egyptian migrants in
Cairo. Linguistics 43 (5).903-956.

Milroy, Lesley. 1980. Language andsocial networks. Oxford: Basil Blackwell.

Milroy, James. 2006. Language change and the speaker: On the discourse of historical linguistics.
Variation and reconstruction, ed. by Thomas D. Cravens. Amsterdam: John Benj amins.

Milroy, James, and Lesley Milroy. 1985. Linguistic change and social network and speaker
innovation. Journal of Linguistics 21.3 39-3 84.

Milroy, James, and Lesley Milroy. 1992. Social network and social class: toward an integrated
sociolinguistic model. LiS 21.1-26.

Moumine, M. E. 1990. Sociolinguistic variation in Casablanca Moroccan Arabic. Casablanca,
Morocco: University Mohammed V dissertation.

Morris, Richard E. 1998. Stylistic variation in Spanish phonology. Columbus, OH: The Ohio
State University dissertation. Rutgers Optimality Archive ROA-292.

Morris, Richard E. 2000. Constraint interaction in Spanish /s/-aspiration: three peninsular
varieties. Hispanic Linguistics at the Turn of the M~illennium: Papers f~on the 3rd
Hispanic Linguistics Synaposiunt, ed. by Hector Campos, Elena Herburger, Alfonso
Morales-Front, and Thomas J. Walsh. Somerville, MA: Cascadilla. Also Rutgers
Optimality Archive ROA-391.

Moscati, Sabatino; Anton Spitaler; Edward Ullendorff; and Wolfram von Soden. 1964. An
introduction to the comparative grananar of the Semitic languages. Wiesbaden: Otto
Harrassowitz.

Mufwene, Salikoko S. 2005. Language evolution: The population genetics way.
Gene, Sprachen, und ihre Evohition, ed. by Guenther Hauska, 30-52. Universitaetsverlag
Regensburg.

Nagy, Naomi, and William T. Reynolds. 1996. Accounting for variable word-final deletion
within Optimality Theory. Sociolinguistic variation: Theory, Data and Analysis (Selected
Papers from NWAV 23), ed. by Jennifer Arnold, Renee Blake, Brad Davidson, Scott
Schwenter, Julie Solomon. Stanford, CA: CSLIPublications.










Nagy, Naomi, and William T. Reynolds. 1997. Optimality Theory and variable word-final
deletion in Faeter. Language Variation and Change 9.37-55.

Newbrook, Mark. 1982. Sociolinguistic Reflexes of Dialect Interference in West Wirral.
Reading, Berkshire, United Kingdom: Reading University Dissertation.

Ohala, John J. 1993. The phonetics of sound change. Historical linguistics: problems and
perspectives, ed. by Charles Jones, 237-278. London: Longman.

Palfreyman, David, and Muhamad al-Khalil. 2003. "A Funky Language for Teenzz to Use":
Representing Gulf Arabic in Instant Messaging. Journal of Consputer-M~ediated
Conanunication 9 (1). Retrieved November 18, 2007, from
http://jcmc.indiana.edu/vol9/issuel/palfreyman.htm

Patrick, Peter L. 1992. Creoles at the intersection of variable processes: -t, d deletion and past-
marking in the Jamaican mesolect. Language Variation and Change 3.171-189.

Patten, Mildred L. 2004. Thiderstanding research methods: An overview of the essentials, Fourth
Edition. Glendale, CA: Pyrczak Publishing.

Phillips, Betty S. 2006. Word Frequency and~exical Diffusion. New York: Palgrave Macmillan.

Pierrehumbert, Janet. 2001. Exemplar dynamics: Word frequency, lenition and contrast.
Frequency and the emergence of linguistic structure, ed. by Joan Bybee and Paul Hopper,
137-157. Amsterdam: John Benjamins.

Pierrehumbert, Janet. 2002. Review of Phonology and language use, by Joan Bybee. 200 1.
(Cambridge Studies in Linguistics 94.) Cambridge: Cambridge University Press.
Phonology 19.459-463.

Prince, Alan, and Paul Smolensky. 2004. Optimality Theory: Constraint interaction in
Generative Grananar. Maiden, MA: Blackwell. Previously available as a Ms., 1993,
Piscataway, NJ: Center for Cognitive Science, Rutgers University.

Rickford, John R., and Penelope Eckert (eds.). 2001. Style and sociolinguistic variation.
Cambridge, UK and New York: Cambridge University Press.

Reynolds, William T. 1994. Variation and phonological theory. Philadelphia, PA: University of
Pennsylvania dissertation.

Roca, Iggy, and Wyn Johnson. 1999. A Course in Phonology. Maiden, MA: Blackwell
Publishing.

Romaine, Suzanne. 1978. Postvocalic /r/ in Scottish English: Sound change in progress? In
Trudgill, 144-157.










Rose, Sharon. 1996. Variable laryngeals and vowel lowering. Phonology 13.73-117.

Rumelhart, David E., and James L. McClelland. 1987. Learning the past tenses of English verb:
Implicit rules or parallel distributed processing? Mechanisms of language acquisition, ed.
by Brain MacWhinney. Hillsdale, NJ: Lawrence Erlbaum Associates.

Sankoff, Gillian, and Hdli~ne Blondeau. 2007. Language change across the lifespan: /r/ in
Montreal French. Language 83 (3).560-588.

Santa Ana A., Otto. 1992. Chicano English evidence for the exponential hypothesis: a variable
rule pervades lexical phonology. Language Variation and Change 4.275-288.

Saussure, Ferdinand de. 1959. Course in General Linguistics. New York: McGraw-Hill. First
published 1919.

Sawaie, Mohammed. 1994. Linguistic Variation and Speakers Attitudes: A Sociolinguistic Study
of Sone Arabic Dialects. Damascus: Al-Jaffan and Al-Jabi Publishers in co-operation
with the Institut Frangais d'Etudes Arabes de Damas.

Schilling-Estes, Natalie, 1999. Situated ethnicities: Constructing and reconstructing identity in
the sociolinguistic interview. thriversity ofPennsylvania Working Papers in Linguistics 6
(2).137-51.

Schilling-Estes, Natalie, 2002. Investigating Stylistic Variation. In Chambers, Trudgill, and
Schilling-Estes, 375-401.

Schmidt, Richard W. 1974. Sociostylistic variation in spoken Egyptian Arabic: A reexantination
of the concept of diglossia. Washington, D.C.: Georgetown University dissertation.

Schmidt, Richard W. 1986. Applied Sociolinguistics: The case of Arabic as a second language.
Anthropological Linguistics 28 (1).55-71.

Schultz, David E. 1981. Diglossia and variation in formal spoken Arabic in Egypt. Madison, WI:
University of Wisconsin, Madison dissertation.

Shahin, Kimary N. 1998. Optimized postvelar harmony in Palestinian Arabic. In Benmamoun,
Eid, and Haeri, 143-166.

Slezak, Peter. 2007. Linguistic explanation and 'psychological reality'. Retrieved June 7, 2008
from: http://web.gc.cuny.edu/Philosophy/people/eitSzaDvt_20.d

Smolensky, Paul. 1996. On the comprehension/production dilemma in child language. Linguistic
Inquiry 27.720-731.

Stampe, David. 1972. How I spent my sunaner vacation (A dissertation on natural phonology).
Chicago, IL: University of Chicago dissertation.











Steriade, Donca. 1987. Locality conditions and feature geometry. Proceedings of the 1 7th annual
Meeting of the North East Linguistics Society (NELS 1 7), ed. by Joyce McDonough and
Bemnadette Plunkett, 595-618. Graduate Student Linguistic Association, University of
Massachusetts, Amherst.

Syrian Arab Republic Central Bureau of Statistics. 1960. Statistical Abstract. Damascus, Syria:
Office of the Prime Minister.

Syrian Arab Republic Central Bureau of Statistics. 1986. Statistical Abstract. Damascus, Syria:
Office of the Pr