Citation
The influence of language ability, age and socioeconomic status on language use

Material Information

Title:
The influence of language ability, age and socioeconomic status on language use
Creator:
Lieberman, Rita Jane, 1946-
Publication Date:
Language:
English
Physical Description:
x, 280 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Child psychology ( jstor )
Foreign language learning ( jstor )
Kittens ( jstor )
Language ( jstor )
Learning ( jstor )
Linguistics ( jstor )
Nonnative languages ( jstor )
Reasoning ( jstor )
Spoken communication ( jstor )
Students ( jstor )
Children -- Language ( lcsh )
Dissertations, Academic -- Speech -- UF ( lcsh )
Language acquisition ( lcsh )
Speech and social status ( lcsh )
Speech thesis Ph. D ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1981.
Bibliography:
Bibliography: leaves 267-278.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Rita Jane Lieberman.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
028138383 ( ALEPH )
07956698 ( OCLC )

Downloads

This item has the following downloads:


Full Text










THE INFLUENCE OF LANGUAGE ABILITY,
AGE AND SOCIOECONOMIC STATUS
ON LANGUAGE USE














by

RITA JANE LIEBERMAN












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


UNIVERSITY OF FLORIDA

1981

































In Loving Memory of My Father, Edward 1. Plotle














ACKNOWLEDGEMENTS


During my tenure as a graduate student, many individuals contributed to my personal and professional growth. To Dr. Thomas Abbott, I am especially grateful for his encouragement and guidance throughout my graduate career and his willingness to assume the "long distance" chairmanship of my supervisory committee in the final stages of my research. To the remaining members of my committee, Drs. Anthony Clark, Jayne Harder and Leonard LaPointe, I wish to express my sincere thanks for their suggestions and helpful comments during the initial stages of this project and their forebearance and understanding during the hectic final days. Dr. Edward Hutchinson, who introduced me to the important study of language use, who served as the original director of my doctoral committee and who has been my understanding department head since I joined the faculty at Appalachian State University, deserves my heartfelt thanks. He has been my colleague, my friend and my mentor. He has seen me through the dark times that accompany any project which makes one feel tiny and incapable at its inception. He has encouraged me through the gray days of writing and rewriting. With scholarly advice and with a wise blend of sympathy and threats, he has helped me inestimably to bring this project to its conclusion.

I would also like to extend my appreciation to Deanna Bowman for her excellent statistical assistance. She was always available to help and her expertise was invaluable. Valerie Buice, Geneva Henson and Susan Payne unselfishly gave of their time to type the original









manuscript and Dinah Lanning provided outstanding editorial assistance while typing the final draft. Thanks must also be extended to the faculty, staff and children of the schools which participated in this project. They made my job a pleasant and enjoyable one.

Finally, my love and endless gratitude to my husband, Chuck, who never stopped believing in me; to my parents and my husband's parents whose long and patient vigil has at last been rewarded with --nachas' from heaven; to my son, Justin, whose unconditional hugs and kisses never diminished even in the face of a crabby mony; to my dear friends, Ann Michael, whose kind and enthusiastic words were a constant source of encouragement on dreary Boone days; Vicky Breedlove, who shared my teaching responsibilities during the summer of '81 to provide me with precious time; Cathy and Wayne Huber, who hand-carried this manuscript from the Trailways bus station to the University of Florida editorial office on several occasions and whose kind hospitality gave me the peace of mind to remain calm during the week of my oral defense; and to all my other friends and family, without whose support this project would not have been possible.















TABLE OF CONTENTS


Page

ACKNOWLEDGEMENTS ..................... . . iii

ABSTRACT . .. .. . . . .. . . . . . . . . . . . . . . . . . ix

Chapter

1. INTRODUCTION . ...... .... .... . . ... . .

Significance of the Problem ......... .. . I

Statement of the Problem ............... 6

Delimitations ............. . ... 7

Limitations . . . . . . . . . . . . . . . . . . . . 8

Assumptions . . . . . . . . . . . . . . . . . . . . . 8

Hypotheses ............... . .... . 9

Hypothesis I ............ . .... . 9

Hypothesis 2 . . . . . . . . . . . . . . . . . . . . If

Hypothesis 3 .... .. .................... .... 13

Hypothesis 4 . . . . . . . . . . . . . . . . . . . . 15

Hypothesis 5 . . . . . . . . . . . . . . . . . . . . 17

Hypothesis 6 .............. . . .. 18

Hypothesis 7 .... .. .................... .... 19

2. REVIEW OF RELATED LITERATURE ........ . . .. 20

The Interdisciplinary Nature of Recent
Research into Language Use .. ........ .. 20

Expanding the Notion of Competence ...... ... 21 Origins of the Study of Language Use ..... . .. 23









Chapter Page

The Domain of Language Use .... .............. .... 23

Functions of Language ........ ................ 26

Speech Acts Theory ...... .................. .... 37

Development of Language Use in
Language-Normal Children .... .................. 53

Development of Languase Use in
Language-Impaired Children ...... ............. 66

Primary Forms ..... ................... ..... 66

Conventional Forms ........ ................. 68

Language Use in the Culturally Diverse .... ........ 72

Measurement of Language Use ... ............. .... 83

Standardized Measurement Strategies .... ........ 84

Nonstandardized Measurement Strategies ....... ... 85

The Literature in Retrospect ... ................ 89

3. METHODS AND PROCEDURES ................. ......... 92

Subjects ....... ....................... ..... 92

Language-impaired Group (LI) .. ............ .... 92

Language-Normal Groups .... ............... .... 93

Materials ........ ...................... .... 102

Classification of the Cognitive
Uses of Language ........ ................. 102

The Functional Inventory of Cognitive
Communication Strategies (FICCS) .... ......... 102

Procedures ........... ...................... I08

Collection of the Language Sample .. ........... 108

Transcription and Segmentation of
the Language Sample .... ............... . ].....I

Analysis of the Language Sample ..... .......... Ill









Chapter Page

Reliability . . . . . . . . . . . . . . . . . . . 119

Order of Presentation of All Tasks ...... .. 120

Session I ................ ... 120

Session 2 ................ ... 120

Session 3 ...... ..................... ..... 121

Session 4 ................ ... 121

Selection Measures . . . . . . . . . . . . . . . . 121

Independent Measures ........... ... 122

4. ANALYSIS OF THE RESULTS ................ 128

Group Comparisons ............. ... 128

Language Used by Language-Normal and
Language-impaired Children ........ .. 128

Language Used by Language-Noral Children of Different Ages and Socioeconomic Levels . . .. . . . . .. . . . . . . 154

Correlation and Regression Analyses ........ 175

Language Use and Standardized Measures of
Language Content and/or Form . ........ 175

Language Use and Nonstandardized
Measures of Language Form . .......... 182

Language Use and Measures of Academic
Achievement .................. 186

Reliability ................ ... 19

Internal Consistency ........... ... 194

Intra-rater Reliability ........ . ... 198

Inter-rater Reliability ......... ... 198

Summary of Findings ............ ... 199

5. SUMMARY AND CONCLUSIONS ............... 206









Chapter Page

Summary . . . . . . . . . . . . . . . . . . . . . . . . 206

Discussion . . . . . . . . . . . . . . . . . . . . . . 209

Conclusions and Suggestions for Future Research .... 217 APPENDIXES

A. FAMILY SIZE AND INCOME SCALES FOR FREE MEALS
AND REDUCED-PRICE MEALS .......... . ... 221

B. TOUGH'S FRAMEWORK FOR THE CLASS IFICATION
OF LANGUAGE USE .............. . ... 223

C. LANGUAGE SAMPLE ELICITATION PROTOCOL ALA TOUGH ..... 231 D. STIMULUS MATERIALS FOR LANGUAGE SAMPLE ELICITATION . . . 240

E. PROTOCOL FOR TRANSCRIPTION AND SEGMENTATION
OF LANGUAGE SAMPLES .................. 248

F. RULES FOR COUNTING MORPHEMES .......... ... 254

G. RULES FOR COUNTING WORDS .......... . ... 257

H. RAW DATA ON STANDARDIZED LINGUISTIC MEASURES AND THE
OVERALL SCORE ON THE FUNCTIONAL INVENTORY OF COGNITIVE COMMUNICATION STRATEGIES ACHIEVED BY
ALL SUBJECTS .................. 258

I. RAW DATA ON NONSTANDARDIZED LINGUISTIC MEASURES
OBTAINED IN RESPONSE TO THE FUNCTIONAL INVENTORY OF COGNITIVE COMMUNICATION STRATEGIES
(FittS) BY ALL SUBJECTS .......... ... 261

J. RAW DATA ON MEASURES OF ACADEMIC ACHIEVEMENT
FOR ALL SUBJECTS .. ............ . ... 264

BIBLIOGRAPHY ....................... ... 267

BIOGRAPHICAL SKETCH . .................. ... 279










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


THE INFLUENCE OF LANGUAGE ABILITY, AGE AND SOCIOECONOMIC STATUS ON LANGUAGE USE

By

Rita Jane Lieberman

August 1981

Chairman: Thomas B. Abbott
Major Department: Speech

Language is the process by which children succeed or fail in school. Because language is an important subject of instruction as well as the process by which that instruction is achieved, it is imperative that strategies of language use necessary for school success be identified and measured.

The purpose of this study was (I) to determine the influence of language ability, age and socioeconomic status on children's language use; and (2) to examine the relationship and predictive accuracy between measures of linguistic performance, academic achievement and language use..

Language use was evaluated with the Functional Inventory of

Cognitive Communication Strategies (FICCS), a structured interview which elicited four language use strategies: Reporting, Reasoning, Predicting and Projecting.

The influence of language ability on language use was examined by comparing language-impaired children to two groups of languagenormal children, one matched for age and one matched for utterance









length. Performance of the language-normal group matched for age was significantly superior to that of the language-impaired group. No significant differences were observed between the language-normal group matched for utterance length and the language-impaired group. The language-impaired group achieved significantly lower overall scores than their language-normal peers but higher scores than their younger, normal counterparts matched for utterance length, suggesting that the communicative function of the impaired children was better than their linguistic skills would imply.

The influence of age and socioeconomic status on language use was evaluated using a factorial design, with two age levels (6 years and

7 years) and two social class levels (lower and higher). The languagenormal 6-year-olds achieved significantly higher scores than the language-normal 7-year-olds on Projecting strategies.

Correlation analyses between performance on FICCS and measures of linguistic ability Indicated a strong relationship between FICCS and nonstandardized measures of language ability but not for standardized. These findings suggest that spontaneous language sampling, through its preservation of the interactive nature of communication, provides a more powerful correlate of language use than acontextual standardized tools. The relationship between FICCS and measures of academic achievement was modest, indicating that language use and other factors contribute to success in the classroom.














CHAPTER I

INTRODUCTION


Significance of the Problem


Communication is the process by which children succeed or fail in school. The modern classroom, more than ever before, provides a laboratory for the development and practice of communication skills whose application extends far beyond the confines of the school environment. Present-day education techniques have turned the classroom into a microcosm of the world where all aspects of human comunication can and do occur. The student serves as encoder, processor and decoder of the cognitive and social content of a variety of messages.

Clearly, communication underlies the major portion of classroom learning. Not only is it an important subject of instruction, but it is also the process by which that instruction is achieved. Because communication is central to both the means and ends of education, a failure in classroom communications will have the most serious consequences. The child who does not succeed in school may be one whose use of communication in the academic setting is restricted in some way and whose communication efforts may reflect an impoverishment of vocabulary or syntactic options, or, skills which are present but either inappropriate to or unavailable for a given classroom situation. Any of these problems may seriously alter a child's participation in the classroom. Traditionally, those disciplines concerned with the study of child









language and learning have focused on the adverse effects of linguistic deficits on classroom participation, while the broader, and perhaps more elusive notion of conmnunicative competence has been largely ignored. Joan Tough (1976), a British educator who has done considerable research on communication skills in early childhood, acknowledges the difference between linguistic and communicative competence. In her book, Listening to Children Talking, she distinguishes between children whose talk evidences inappropriate phonological, syntactic and semantic patterns and children whose talk is limited to a restricted set of uses.

The development of evaluative measures of communication problems In children has been primarily confined to those which measure specific linguistic aspects: phonology (Fisher & Logemann, 1971; Goldman & Fristoe, 1969; McDonald, 1968; Weiner, 1978), syntax (Carrow, 1974; Foster, Giddan & Stark, 1969; Lee, 1969), and semantics (Boehm, 1971; Dunn, 1964; MacDonald & Nickols, 1974). The emphasis on the description and evaluation of language form and content in children is reflected in current assessment strategies and much of the language research literature carried out during the 1960s and early 1970s. Influenced largely by Chomsky's (1957, 1965) theory of transformational grammar and Filimore's (1968) notions about case grammar, investigators performed linguistic analyses or designed psycholinguistically motivated research paradigms to discover new information about the development of form (Bloom, 1970; Brown, 1973; Greenfield & Smith, 1973; Menyuk, 1968) and content (Bloom, Lightblown & Hood, 1975; Clark, 1973; Leonard, 1976; Nelson, 1973, 1974).

This early influence of psycholinguistics on the study of child language focused our attention on the language which children use rather









than on children as language users. Descriptions of the linguistic competence of children were abundant, but the broader notion of communicatie competence-how children use language in socially appropriate ways--was largely overlooked. Recent research trends in the study of child language demonstrate that we are on the threshold of moving from an interest in the description of isolated lexicogrammatical forms to an interest in the description of communicative function in context. The contextualist approach emphasizes the importance of the ability to use language to convey a variety of intentions and meanings dependent upon the context and social setting of the communication.

This shift in perspective represents a major accomplishment in a movement towards relevance and functionality in communication measurement and management. Although psycholinguistic theory accounts for an important aspect of what the child comes to know during attempts to use speech for communicative purposes, it does not account for all the knowledge which underlies communication development. Specifically, it does not explain how the child not only comes to speak grammatically but simultaneously acquires the ability to apply linguistic knowledge in functionally appropriate and predictable ways in a variety of communication situations (Hopper, 1971).

Sociolinguistic theory, on the other hand, recognizes that the ability to speak appropriately is "part of the same developmental matrix" (Hymes, 1970, p. 14) as the ability to speak grammatically. Both these abilities are necessary conditions for the attainment of communicative competence. It is only through an examination of linguistic competence (the rules of grammar) as well as what might be called a pragnaticaspect of competence (the rules of social









interaction) that we can account for the development of order which underlies human communication behavior (Hopper, 1971).

Because of the recent shift in focus to a sociolinguistic theory of language learning and language behavior, applied communicologists-speech, language and hearing clinicians, special educators, early childhood specialists, language arts teachers, foreign language teachers, and hilingual/bicultural consoltants--have begun to develop informal strategies for the measurement and management of communicative competence. Those strategies can be grouped into one of three types: (a) interviewing [Fogel, 1976; Ricillo (in Larson, Backlund, Redmond & Barbour, 1978); Ritti, 1978; Tough, 1976]; (b) roleplaying [Bates, 1976a; Brenneis & Lein (in Ervin-Tripp & Mitchell-Kernan, 1977)]; and

(c) natural language sampling (Bloom, 1978; Dore, 1977; Epstein, Schwartz, Meece, Lanbie, Dukes, Crawford & Phillips, 1976; Halliday, 1975; Hanes, 1978; Schachter, Kirshner, Klips, Friedericks & Sanders, 1974; Sinon, 1979; Wollner & Geller, 1976). Using these elicitation strategies, investigators have categorized children's use of language according to a variety of functional classification schemes including those of Dore (1975, 1977, 1978), Halliday (1975), Schachter et al. (1974), Soskin & John (1963), Tough (1976, 1977), and Wells (1975).

As in the case with most research on language parameters, the majority of these investigations have gathered information on language use from middle-class, language-normal, preschool-aged children (Bates, 1976b; Bates, Camaloni & Volterra, 1975; Bruner, 1974, 1975; Dore, 1974a, 1974b, 1975, 1976, 1977, 1978; Dore, Gearhart & Newman, 1978; ErvinTripp & Mitchell-Kernan, 1977; Greenfield & Smith, 1976; Halliday, 1975). A few studies have observed the language use of linguistically








deficient children and adults. Since 1974, investigations involving functional modes of language use have been carried out with the mentally handicapped (Bedrosian � Prutting, 1978; Owens, 1979; Sabsay, 1975), the hearing-impaired (Curtiss, Prutting � Lowell, 1979; Skarakis � Prutting, 1977), the neurologically-involved (Ulatowska, Macaluso-Haynes � MendelRichardson, 1976; Wilcox & Davis, 1977), the behaviorally-disordered (Bartolucci & Albers, 1974), and the specific linguistically-handicapped (Geller & Wollner, 1976; Miller, 1978; Snyder, 1975). Even fewer studies have focused on the general language use of older school-aged children (Fogel, 1976; Tough, 1976, 1977). While additional studies of language use in older children exist in the sociolinguistic literature, their emphases have generally been narrowly confined to limited aspects of language use such as arguments (Brenneis & Lein, 1977), narratives (Kernan, 1977) or directives (MitchelI-Kernan & Kernan, 1977; Garvey, 1975, Ervin-Tripp, 1975); thus, their implications, though relevant for child literature, remain fragmentary (Rees, 1978).

From sociolinguistics also cocs an abundance of literature on the effects of social class and culture on language use. Because of differences in design, sampling protocols, language tasks and analysis, the results of these studies present conflicting evidence relative to the influence of social class on language use. Several of the studies demonstrate that children of disparate sociocultural backgrounds use language differently (Blank, Rose & Berlin, 1978; Bruck & Tucker, 1974; Tonkovich & Adler, 1978; Tough, 1977; Williams & Naremore, 1969a). Others reveal less clear-cut differences in language use (Edwards, 1977; Schachter et al., 1974; Wells, 1978). This perusal of the available literature on language use uncovers several avenues of investigation









which remain to be charted, particularly with regard to the older language-impaired child who lives in a less than advantageous social situation.


Statement of the Problem


The problem of this investigation is (a) to assess differences in language use between groups of children who differ in age, social class, and linguistic ability; and (b) to evaluate the relationship and predictive accuracy between measures of linguistic and academic competence and the Functional Inventory of Cognitive Communication Strategies (FICCS), a researcher-developed instrument, which assesses language use or communicative competence. Pore specifically, answers to the following questions will be sought;

I. Is there a difference in language use between languageimpaired children and language-normal children, matched on the basis of chronological age?

2. Is there a difference in language use between languageimpaired children and language-normal children, matched for utterance length?

3. Is there a difference in language use between languagenormal children at two different age levels, 6 years (plus or minus 3 months) and 7 years (plus or minus 3 months)?

4. Is there a difference in language use between languagenormal children of two different socioeconomic levels, lower and upper?

5. What is the relationship of language use to performance on standardized measures of language form and/or content?









6. What is the relationship of language use to performance on nonstandardized measures of language form?

7. What is the relationship of language use to performance on measures of academic achievement? Delimitations

I. The study was confined to four groups of children, one language-impaired and three language-normal.

2. Subjects were selected from the prekindergarten, kindergarten and first grade populations of the Wilkes and Avery County school systems in North Carolina according to the following criteria:

a. They were enrolled in a school program for a period of not

less than four weeks prior to inclusion in the study.

b. They demonstrated average intellectual functioning (Age

Deviation Score = 89 to 110) on the Columbia Mental Maturity

Scale (Burgemeister, Blum & Lorge, 1972).

c. They were judged to be language-lmpaired (517th percentile)

or language-normal (>70th percentile) on the basis of performance on the Bankson Language Screening Test (Bankson, 1977)

and a clinical opinion of a certified speech, language and

hearing clinician.

d. They were native speakers of English from monolingual homes

who did not exhibit any gross peripheral defects of audition
or vision.

3. The total testing time was confined to four sessions, or 2/u hours for each subject.









4. Data regarding the criterion variable of language use were confined to that gathered from a researcher-developed instrument, the Functional Inventory of Cognitive Communication Strategies (FICCS), based on Tough's (1976) functional classification scheme.

5. Data regarding the predictor variables of linguistic and academic competence were confined to three standardized measures of language form and/or content, eight nonstandardized measures of language form., and four measures of academic achievement. Limitations

1. To the extent that pretesting sensitizes the subjects to the language sampling procedure, results will not be generalizable to unpretested groups.

2. To the extent that the subjects selected are not representative of the language-impaired or language-normal population at large, results will not be generalizable beyond the sample investigated.

3. To the extent that knowledge of subject status affects the objectivity of the researcher's observations and judgments, or causes her to influence the subjects' reactions to the tasks, results may be biased in favor of one group or the other.

4. To the extent that the subjects are aware of participation in a research study, results may not be generalizable beyond the experimentally accessible population.


Ass umptions


The following assumptions were made in this study;

I. That the groups of language-impaired and language-normal

children were matched on the major critical variables affecting language









use: age, socioeconomic status, intellectual ability, and general linguistic level and that other extraneous variables which may affect language use, such as motivation and personality characteristics, were randomly distributed between the language-impaired and language-normal groups.

2. That the exposure to intervening variables in the school environment, such as supportive relationships with teachers, was equivalent for all groups.

3. That the researcher-developed inventory of cognitive connunication strategies did, in fact, measure ability to use language in response to a structured interview; and that this ability is representative of the way in which children use language in the classroom.

4. That the researcher, being a practiced speech, language and hearing clinician, was qualified to administer, score and interpret all testing procedures used in this study.


Hypotheses


To give direction to the data analysis, the following hypotheses were developed. These hypotheses, stated in the null form, were tested at the .05 level of significance. Hypothesis 1


Ho 1.1. There is no significant difference in performance on the FICCS overall score between the language-impaired (LI) and languagenormal (LN1) groups who are matched for chronological age (CA).


Ho 1.2. There is no significant difference in performance on the FICCS Reporting subcore between the LI and LNI groups.










Ho 1.3. There is no significant difference in performance on the FICCS Logical Reasoning subscore between the LI and LNI groups.


Ho 1.4. There is no significant difference in performance on the FICCS Predicting subscore between the LI and LNI groups.


Ho 1.5. There is no significant difference in performance on the FICCS Projecting subscore between the LI and LNI groups.


Ho 1.6. There is no significant difference in performance for the 24 individual communication strategies on FICCS between the LI and LNi groups.


Ho 1.7. There is no significant difference in performance for lower-level strategies overall on FICCS between the LI and LNI groups.


Ho 1.8. There is no significant difference in performance for lower-level Reporting strategies on FICCS between the LI and LNI groups.


Ho 1.9. There is no significant difference in performance for lower-level Logical Reasoning strategies on FICCS between the LI and LNI groups.


Ho .S0. There is no significant difference in performance for lower-level Predicting strategies on FICCS between the LI and LNi groups.


Ho 1.11. There is no significant difference in performance for lower-level Projecting strategies on FICCS between the LI and LNI groups.


Ho 1.12. There is no significant difference in performance for higher-level strategies overall on FICCS between the LI and Li groups.










Ho 1.13. There is no significant difference in performance for higher-level Reporting strategies on FICCS between the LI and LNI groups.


Ho 1.14. There is no significant difference in performance for higher-level Logical Reasoning strategies on FICCS between the LI and LNI groups.


Ho 1.15. There is no significant difference in performance for higher-lenel Predicting strategies on FICCS between the LI and LNI groups.


Ho 1.16. There is no significant difference in performance for higher-level Projecting strategies on FICCS between the LI and LNI groups.


Hypothesis 2


Ho 2.1. There is no significant difference in performance on the FICCS overall score between the language-impaired (LI) and languagenormal (LN2) groups who are matched for utterance length.


Ho 2.2. There is no significant difference in performance on the Reporting subscore between the LI and LN2 groups.


Ho 2.3. There is no significant difference in performance on the Logical Reasoning subscore between the LI and LN2 groups.


Ho 2.4. There is no significant difference in performance on the FICCS Predicting subscore between the LI and LN2 groups.










Ho 2.5. There is no significant difference in performance on the Projecting subscore between the LI and LN2 groups.


Ho 2.6. There is no significant difference in performance for the 24 individual communication strategies on FICCS between the LI and LN2 groups.


Ho 2.7. There is no significant difference in performance for lower-level strategies overall on FICCS between the LI and LN2 groups.


Ho 2.8. There is no significant difference in performance for lower-level Reporting strategies on FICCS between the LI and LN2 groups.


Ho 2.9. There is no significant difference in performance for lower-level Logical Reasoning strategies on FICCS between the LI and LN2 groups.


Ho 2.10. There is no significant difference in performance for lower-level Predicting strategies on FICCS between the LI and LN2 groups.


Ho 2.11. There is no significant difference in performance for lower-level Projecting strategies on FICCS between the LI and LN2 groups.


Ho 2.12. There is no significant difference in performance for higher-level strategies overall on FICCS between the LI and LN2 groups.


Ho 2.13. There is no significant difference in performance for higher-level Reporting strategies on FICCS between the LI and LN2 groups.










Ho 2.14. There is no significant difference in performance for higher-level Logical Reasoning strategies on FICCS between the LI and LN2 groups.


Ho 2.15. There is no significant difference in performance for higher-level Predicting strategies on FICCS between the LI and LN2 groups.


Ho 2.16. There is no significant difference in performance for higher-level Projecting strategies on FICCS between the LI and LN2 groups.


Hypothesis 3


Ho 3.1. There is no significant difference in performance on the FICCS overall score between the language-normal 6- and 7-year-olds.


Ho 3.2. There is no significant difference in performance on the FICCS Reporting subscore between the language-normal 6- and 7-yearolds.


Ho 3.3. There is no significant difference in performance on the FICCS Logical Reasoning subscore between language-normal 6- and 7year-olds.


Ho 3.4. There is no significant difference in performance on the FICCS Predicting subscore between language-normal 6- and 7-year-olds.


Ho 3.5. There is no significant difference in performance on the FICCS Projecting subscore between language-normal 6- and 7-year-olds.










Ho 3.6. There is no significant difference in performance for the 24 individual communication strategies on FICCS between laeguagenormal 6- and 7-year-olds.


Ho 3.7. There is no significant difference in performance on the lower-level strategies overall on FICCS between the language-normal 6- and 7-year-olds.


Ho 3.8. There is no significant difference in performance on the lower-level Reporting strategies on FICCS between language-normal 6- and 7-year-olds.


Ho 3.9. There is no significant difference in performance on the lower-level Logical Reasoning strategies on FICCS between languagenormal 6- and 7-year-olds.


Ho 3.10. There is no significant difference in performance on the lower-level Predicting strategies on FICCS between language-normal 6- and 7-year-olds.


Ho 3.11. There is no significant difference in performance on the lower-level Projecting strategies on FICCS between language-normal 6- and 7-year-olds.


Ho 3.12. There is no significant difference in performance on the higher-level strategies overall on FICCS between language-normal 6- and 7-year-olds.










Ho 3.13. There is no significant difference in performance on the higher-level Reporting strategies on FICCS between language-normal 6- and 7-year-olds.


Ho 3.14. There is no significant difference in performance on the higher-level Logical Reasoning strategies on FICCS between languagenormal 6- and 7-year-olds.


Ho 3.15. There is no significant difference in performance on the higher-level Predicting strategies on FICCS between language-normal 6- and 7-year-olds.


Ho 3.16. There is no significant difference in performance on the higher-level Projecting strategies on FICCS between language-normal 6- and 7-year-olds.


Hypothesis 4


Ho 4.1. There is no significant difference in performance on the FICCS overall score between the lower (LSES) and higher (HSES) socioeconomic status language-normal groups.


Ho 4.2. There is no significant difference in performance on the FICCS Reporting sbuscore between the LSES and HSES language-normal groups.


Ho 4.3. There is no significant difference in performance on the FICCS Logical Reasoning subscore between the LSES and HSES languagenormal groups.










Ho 4.4. There is no significant difference in performance on the FICCS Predicting subscore between the LSES and HSES language-normal groups.


Ho 4.5. There is no significant difference in performance on the FICCS Projecting subscore between the LSES and HSES language-normal groups.


Ho 4.6. There is no significant difference in performance for the 24 individual communication strategies on FICCS between the LSES and HSES language-normal groups.


Ho 4.7. There is no significant difference in performance on the lower-level strategies overall on FICCS between the LSES and HSES language-normal groups.


Ho 4.8. There is no significant difference in performance on the lower-level Reporting strategies on FICCS between the LSES and HSES language-normal groups.


Ho 4.9. There is no significant difference in performance on the lower-level Logical Reasoning strategies on FICCS between the LSES and HSES language-normal groups.


Ho 4.10. There is no significant difference in performance on the lower-level Predicting strategies on FICCS between the LSES and HSES language-normal groups.










Ho 4.11. There is no significant difference in performance on the lower-level Projecting strategies on FICCS between the LSES and HSES language-normal groups.


Ho 4.12. There is no significant difference in performance on the higher-level strategies overall on FICCS between the LSES and HSES language-normal groups.


Ho 4.13. There is no significant difference in performance on the higher-level Reporting strategies on FICCS between the LSES and HSES language-normal groups.


Ho 4.14. There is no significant difference in performance on the higher-level Logical Reasoning strategies on FICCS between the LSES and HSES language-normal groups.


Ho 4.15. There is no significant difference in performance on the higher-level Predicting strategies on FICCS between the LSES and HSES language-normal groups.


Ho 4.16. There is no significant difference in performance on the higher-level Projecting strategies on FICCS between the LSES and HSES language-normal groups.


Hypothesis 5


Ho 5.1. There is no significant relationship between performance on a measure of language use, FICCS, and the Test of Auditory Comprehension of Language (TACL).










Ho 5.2. There is no significant relationship between performance on FICCS and the Test of Language Development. (TOLD).


Ho 5.3. There is no significant relationship between performance on FICCS and the Carrow EIcited Language Inventory (CELl).


Hypothesis 6


Ho 6.1. There is no significant relationship between performance on a measure of language use, FICCS, and the total number of communication units (#CU's) used in response to a structured interview.


Ho 6.2. There is no significant relationship between performance on FICCS and the total number of words (TNW) used in response to a structured interview.


Ho 6.3. There is no significant relationship between performance on FICCS and the mean length of utterances in words (MLU-W) used in response to a structured interview.


Ho 6.4. There is no significant relationship between performance on FICCS and the total number of morphemes (TNM) used in response to a structured interview.


Ho 6.5. There is no significant relationship between performance on FICCS and the mean length of utterances in morphemes (MLU-M) used in response to a structured interview.


Ho 6.6. There is no significant relationship between performance on FICCS and the mean length of the five longest utterances (MLUS5-M) used in response to a structured interview.










Ho 6.7. There is no significant relationship between performance on FICCS and the percentage of complete communication units (%CUc) used in response to a structured interview.


Ho 6.8. There is no significant relationship between performance on FICCS and the Developmental Sentence Score (DSS) obtained in response to a structured interview. Hypothesis 7


Ho 7.1. There is no significant relationship between performance on a measure of language use, FICCS, and a measure of academic achievement, the Test of Basic Experiences language subtest (TBEL).


Ho 7.2. There is no significant relationship between performance on FICS and the Test of Basic Experiences mathematics subtest (TBEM).


Ho 7.3. There is no significant relationship between performance on FICS and the Test of Basic Experiences social studies subtest (TBESS).


Ho 7.4. There is no significant relationship between performance on FICES and the Test of Basic Experiences science subtest (TBES).














CHAPTER 2

REVIEW OF RELATED LITERATURE


The Interdisciplinary Nature of Recent Research into Language Use


There is no single corpus of research which explains the child's developing ability to use the phonological, syntactic and semantic structures of language to communicate effectively a variety of messages in diverse social situations. Several disciplines, including linguistics, sociolinguistics, developmental psychology, philosophy and speech pathology have been approaching this issue from similar but separate points of view. Evidence for these ''separate but nearly equal' vantage points comes from the many terms which have evolved to describe essentially the same set of behaviors: communication competence (Hymes, 1970; Slobin, 1971; Wood, 1976), pragmatics (Bates, 1976a, b; Dore, 1974; Hopper & Naremore, 1973; Prutting, 1979; Rees, 1978) and language use (Bloom & Lahey, 1978; Dale, 1972; Halliday, 1975). Although these terms are not precise synonyms, the cluster of behaviors referred to by such terms "has intruded into many areas of child study, highlighting the importance of rule-governed ways in which children use code items to meet the demands of conunication situations" (Allen & Brown, 1976, p. 153).

Each discipline which has undertaken the task of studying language using a contextual approach has made significant contributions concerning the language use of real speakers and listeners. From

20









philosophy has arisen a theory of speech acts which focuses on the communicative intent realized through language. Linguists have developed a specialized branch of study called pragmatics with its own terminology and taxonomy of types. The speaker's goal in using an utterance (performatives) and the rules for relating utterances to their contexts (presuppositions and conversational postulates) are the constructs treated by this discipline. The influence of social class, role and situational context on language use has been addressed by sociolinguists in their attempt to provide a fuller account of the notion of communicative competence. Developmental psychologists have put the theories of speech acts, pragmatics and cormunicative competence to work to determine how children learn their roles as communicators from early social interactions. In speech pathology, clinicians, drawing from theory as well as methodology, have evolved programs of evaluation and management of disorders of language use. For this review, the various terms associated with the contextualist approach will be subsumed under the more broadly inclusive expression, "language use."


Expanding the Notion of Competence


Chomsky (1965) introduced the notion of "competence" into linguistic theory to refer to the abstract knowledge of the rules of language which enable the speaker to understand and produce an infinite set of novel sentences. This knowledge or intuition about how language should sound and what it means accounts for the ideal speaker/hearer's ability to judge the grammaticality of sentences; however, it says







22

little about real children's everyday use of language in particular situations.

Hymes (1971, 1972b) has been critical of the linguist's tendency to emphasize the notion of competence while neglecting the issue of performance or use. He proposed an expansion of the notion of linguistic competence into "competence for use" or 'communicative competence"-how children perceive and categorize the social situations of their world and differentiate their ways of speaking accordingly. This broader account of the facts provides an explanation of what the language user knows about "who can say what, in what way, where and when, by what means and to whom" (p. 15). The important issue is that children's verbal behavior is a reflection of their knowledge of language use as well as of grammar and the manifestation of that knowledge will vary according to the communication situation. This influential role of the communication situation in determining verbal behavior is an important contribution which sociolinguistic theory makes to our understanding of language development and language behavior.

From a sociolinguistic point of view, there is more to "competence for use" than language ability. Applied communication specialists no longer use linguistic competence as the ultimate goal in communication evaluation and management programs. Davies (1977) comments,

there is more to using, learning and teaching a
language than knowledge of the linguistic rules indicates.
Knowing the grammar is just not enough as a model of how
speakers behave... (p. 2)










Origins of the Study of Language Use


In 1938, Charles Morris incorporated the notion of language use into a tripartite organization of language. According to this model, a comprehensive description of language required a statement about three principal components:

1. Syntactics--the relation of signs to one another;

2. Semantics--the relation between signs and their referents;

3. Pragnatics--the relation between signs and their users.

In the late 1950s and early 1960s, scholars in the field of the philosophy of language expanded the notion of language use to include a practical system for classification of the knowledge which underlies socially appropriate speech. The work of two linguistic philosophers in speech acts theory, J. L. Austin (1962) and J. R. Searle (1969) has served as the basis for much of the current research in the area of language use.


The Domain of Language Use


In their three-dimensional characterization of language, Bloom and Lahey (1978) echo Morris' earlier writing by describing language as consisting "of some aspect of content or meaning that is coded by linguistic form for some purpose or use in a particular context" (p. II). Through the integration of content, form and use, children become competent communicators as they acquire the ability to use alternative forms of a message for achieving the same purpose, according to differences in the situational context. Earlier depictions of child language accounted for the mastery of content and form, but it is to the









contextualist movement that we owe the additional concern with use. The contextual ists noted that "there was a structure in the use of language that went beyond the aspect of structure dealt with in grammars. (Hymes, 1972a, p. xxii).

Within the broad framework of language use, Bloom and Lahey (1978) recognize two main areas, function and context. Function refers to the goals or intent of a communication, "the reasons why people speak" (Bloom & Lahey, 1978, p. 19). Context encompasses the influence of specific situational parameters, such as the time and place of the communication, in determining the form which the message will take. 'Speakers of a language have alternative means for saying the same thing or achieving the same purpose, and which alternative is used depends on the context" (Bloom & Lahey, 1978, p. 10).

Traditionally, the functions of language have been represented in linguistic terms and associated with the syntactic structures for the declarative, interrogative, imperative or exclamative utterance. Halliday (1975) and others have described the functions of language "in more social terms involving interaction, regulation and personal control" (Bloom & Lahey, 1978, p. 20). To achieve these communicative functions, speakers choose appropriate verbal and nonverbal behaviors to bring about desired effects in a wide variety of social situations. The verbal and nonverbal choices which speakers make have been called communication strategies or communication acts by some investigators (Wood, 1976) and evolved out of the more restricted notions of "speech acts,'' "performatives" or illocutionary force.'' This category of language use involves the specification of the speaker's goal in producing







25

an utterance--it describes intention to make a statement, ask a question or issue a command.

The second aspect of language use focuses on the influential factors involved in the selection of alternative means for conveying messages in varied social contexts as well as the communication processes by which these messages are produced. The form of an utterance changes, depending upon the ability of the speaker to adapt the message to the needs of the listener and to the immediate linguistic and nonlinguistic contexts. These elements of the communication process are necessary ''to keep meaning flowing in conversation" (Rees, 1978, p. 208). Within the realm of pragmatic theory, these constructs have been labeled presuppositions and conversational postulates. Although there is considerable controversy regarding the nature of presuppositions, in general they involve -'information that is not contained in the message itself but must be known and understood if the message is to make sense' (Bates, 1976a, p. 419). Presuppositions are the determining factor by which speakers adapt their messages to the listener's needs--how they decide what to say and what not to say. According to Bates (1976a), the major conversational task for children in a communicative exchange is learning "when not to presuppose, when it is necessary to provide the listener with explicit cues about the information that is being assumed as background for a comment" (p. 445). This process has been referred to as thematic structure (Halliday, 1973) or relating new information to old (Clark & Clark, 1977; Rees, 1978).

Conversational postulates have been described by Bates (1976a) as the ''boy scout code of conversation' (p. 46). They are a special class of pragmatic presuppositions which "attempt to make explicit the







26

rules speakers and listeners must observe if meaning is to be conveyed" (Rees, 1978, p. 208). Grice (1976) has reduced these rules to four maxims which contribute to what he calls the "cooperative principle" of conversation. In summary, these maxims are: (a) be informative;

(b) be truthful; (c) be relevant; and (4) be clear.

These three constructs--performatives, presuppositions and conversational postulates--"contribute to both the form and context of messages; to how people do things with words; and to bow languages work" (Bloom & Lahey, 1978, p. 21). In 1976, Bates commented on the limited amount of pragmatic research in child language. The greater proportion of this work centered on the acquisition of performatives or comunicative intent although, as Rees (1978) noted, "the conduct of conversation is probably the most complex as well as the most important target of the pragmatic approach to the study of language" (p. 208).

The present investigation focuses on the study of a single aspect of language use--the communication strategies or goals used by children of differing linguistic ability, age, and socioeconomic level.


Functions of Language


Although philosophers, linguists and psychologists have long been interested in functional taxonomies of language use, their popularity has only recently gained momentum with applied disciplines. Various taxonomies have been described, each according to the author's own unique interests. In 1923, Malinowski developed an ethnographic account of the functions of language to describe the use of language in organizing fishing expeditions in the Trobriand Islands. Since







27

that time, other functional systems have appeared with different bases, e.g., psychological (DeLaguna, 1927; Buhler, 1934; Soskin & John, 1963), cognitive (Praget, 1923; Skinner, 1957; Bruner, 1975), literary (Jakobson, 1960), ethological (Morris, 1967), linguistic (Halliday, 1970, 1975) and educational (Britton, 1971). On the surface, the categories and terminologies embodied in each of these functional classification schemes differ, but each incorporates a basic distinction between a cognitive (ideational, representational, referential) and a social (interpersonal, expressive-conative, evocative) function of language.

Two well-known accounts of the broad functions which language serves are those offered by DeLaguna (1927) and Bruner (1975). In 1927, Delaguna wrote "men do not speak simply to relieve their feelings or to air their views, but to awaken a response in their fellows and to influence their attitudes and acts" (p. 20). Bruner (1975) maintained that "Language is acquired as an instrument for regulating joint activity and joint attention" (p. 2).

The correspondence between these two global characterizations of communicative function is not surprising, since, as McLean and McLean (1978) point out, they describe the same general underlying phenomenon. Both accounts recognize two broad functions of language: (a) the regulation or influencing of a listener's actions; and (b) the regulation or influencing of a listener's attention or attitudes. The first function type includes communication acts such as requesting or demanding which require an overt response from the listener, thereby allowing the speaker to judge the success of the communicative exchange. The second function type is not as easily explained. According to McLean







28

and McLean (1978), it is often difficult to extract from the content of the communicative exchange the speaker's underlying purpose or intent.

These simplistic global taxonomies of language provide an organizational framework for the development of more sophisticated classification schemes. As Soskin and John (1963) observed, "a simple dichotimization of something so complex as verbal behavior is nothing more than a first step; it merely helps "clear the brush" (p. 254).

The majority of these functional classification schemes were developed for the categorization of adult communication. McNeill (1970) adopted the threefold characterization of language functions described by Buhler (1934) and Jakobson (1960) to determine the purpose of young children's holophrastic utterances. In his analysis, he noted that all three aspects of the classification scheme were present in their one-word utterances including the referential, expressive and conative functions. The referential function, most closely associated with the cognitive or denotative aspects of language, is manifested in its purest form by naming or labeling. According to McNeill (1970), purely referential utterances never occurred. When children appeared to be engaged in labeling, they were actually using words predicatively as comments about the situations in which they found themselves (DeLaguna, 1927). The expressive function is the use of verbal comments and paralinguistic devices to reveal the speaker's feelings and attitudes about what is being said. The conative function influences the behavior of others and is manifested by vocative and imperative constructions.

Piaget (1923) was the first investigator to develop a functional classification scheme specifically for analysis of children's language.









In his book, The Language and Thought of the Child, he attempted to answer the following question: "What are the needs which a child tends to satisfy when he talks?" (p. 25). Piaget (1923) classified the talk of children into two main categories: egocentric and socialized. Egocentric talk is intrapersonal in nature because the child, while engaged in it, does not talk at a specific audience nor require a response from a listener who happens to overhear the utterance. The child "talks either for himself or for the pleasure of associating anyone who happens to be there with the activity of the moment" (p. 32). Egocentric speech accompanies, reinforces and supplements the speaker's actions and is manifested in monologues and the pleasurable repetition of syllables and words. Socialized talk is interpersonal in nature. "It addresses the listener, considers his point of view, tries to influence him, or exchanges ideas with him'' (Johnson, Darley & Spriestersbach, 1963, p. 180). Within the socialized function, children may use language to express adapted information, emotionally toned remarks, questions, answers, social phrases and dramatic imitations.

Recent additions to the child-oriented functional classification schemes include one developed for preschool educational settings by Schachter et al. (1974) and another designed by Halliday (1975) which can be applied from the earliest stages of infancy to adulthood. Schachter and her colleagues (1974) developed an instrument or scoring scheme called the Functions of Interpersonal Spontaneous Preschool Speech (FIS-P) to identify developmental changes in the patterns of spontaneous speech in children, ages 2 to 5, and to determine the effects of specific sociolinguistic variables such as race and socioeconomic status on these patterns. The instrument was developed









empirically based on 2,000 actual statements, from 100 preschoolers observed during freeplay activities. Only spontaneous initiations by the child, with no prodding or shaping by adults and no direct questioning by peers, were considered acceptable for scoring.

The FIS-P consists of a comprehensive set of nlne function categories which may be further subdivided into subcategory scores. According to the authors, the category scores cover motives for all spontaneous interpersonal statements including categories for personal motives (I to IV), social motives (V to VII) and other motives (VIII and IX). The subcategory scores designate the main means for implementing the functions, including the following:

I. Expressive--functions to express an emotion, e.g., "ouch'; "I hate this''; "Goodie.''

II. Desire Implementing--functions to implement a personal desire for an object, for help, for permission, for general reassurance or attention, e.g., 'Can I have some''; ''Fix it'; 'Stop it."

111. Possession Rights Implementing--functions to implement possession rights involving objects, territory, turns or roles (fantasy or real), e.g., "This is my dolly"; 'I had it first"; "I want to be mommy."

IV. Ego-Enhancing--functions to enhance the subject's ego, with context and tone showing evident pride, e.g., 'Look at my big house"; "We're sharing"; "That's junk''; ''You're cheating."

V. Self-Referring-lncluding--functions to join subject to other by self-referring the other's statements, activity or characteristics, e.g., "Me too"; "I listen to batman, too.









VI. Joining-functions to join other to subject with speaker actively initiating the union, e.g., ''Hello"; ''Are you my friend?"; "Let's play blocks''; 'Im not playing with you."

VII. Collaborative--functions to initiate or maintain a roledifferentiated social interaction, with two or more subjects participating in a project, discussion or game, e.g., "I'll shoot and you fall down."

VIII. Learning Implementing--functions to implement learning about objective world, social world, biological world or how to proceed in a task, e.g., ''What does that say?"; ''This is a dump truck.

IX. Reporting--functions to share an observation, thought or experience with another, e.g., "I went to the circus; 'Amy has new boots'; "It's raining."

In a critical review of the FIS-P, Cazden (1974) commented,

This is a unique study of developmental change in what
Schachter calls 'self-motivated' social speech, that is speech
addressed to but not elicited by, other children and adults.
The use of developmental theory, here Piaget's, as a base
for strong hypotheses about developmental changes is in
sharp contrast to research in which some category system is
applied and developmental trends found in the data then discussed. The extent to which the hypotheses were confirmed is
very encouraging, and Schachter's discussion of her results is
sound. The FIS-P, with some modifications as the author
suggests, merits further use by herself and others.(p. 81)

Halliday's (1970) functional classification of language was the first system to be developed for linguistic reasons--to shed light on the nature of linguistic structure and to attempt to answer the question, "Why is language as it is?" (p. 141).

In 1975, Halliday offered an integrative explanation of language function in children and adults. Through his intensive study of the language development of his son, Nigel, Halliday (1975) traced the









evolution of language function from earliest infancy to adulthood. Based on a sociosemiotic theory, his functional approach to language emphasized the importance of language for the education and socialization of the child. According to Halliday (1978), "a child learning language is at the same time learning other things through language building up a picture of the reality that is around him and inside him'' (p. I). Language function rather than form is at the center of this educating and socializing process since it is the structural organization of language which reflects the changing functions to which language is put and not vice versa.

Halliday's (1975) analysis of the development of language function proceeds through three phases and shows how the child's initial functional-linguistic system develops and changes into "an adult system that is both similar to and different from that of the child" (Rees, 1978, p. 252). During Phase I, which covered the period from i0/2 months to 18 months in Nigel's life, an idiosyncratic but consistent set of vocal symbols was used to express six initial functions:

1. The Instrumental or the "I want" function. In this function, the child uses language to obtain objects and services to satisfy needs.

2. The Regulatory or the "Do as I tell you" function. This function of language is used "to get someone to do something and is different from the instrumental function in focusing on the agent rather than on the object'' (Rees, 1978, p. 252).

3. The Interactional or the ''me and you" function. This is language used by the child to interact with the significant others in the environment.









4. The Personal or "Here I come" function. This is language used to express and develop the child's uniqueness as an individual and to intrude the child as a personality into the speech situation.

5. The Heuristic or ''Tell me why" function. This is the function of language to explore and learn about the environment which develops from its most basic form, the request for names of objects, into the entire range of questioning forms that the young child uses.

6. The Imaginative or "Let's Pretend" function. This is the use of language to create a make-believe environment, including soundplay, songs, rhymes, story-tellings and eventually language as a verbal art form.

These six functions represent what the child can mean during this earliest stage of linguistic development, when the child's twolevel language system represents each utterance in terms of a content (meaning) and an expression (sound). This system differs from the three-part adult language system which contains an intermediate level of organization between meaning and sound consisting of a grammar and a vocabulary. The utterances of Phase I consist of vocal postures which have neither structure nor words. Although word-like elements may be identified, these units are not necessarily imitations of the permissible sound sequences in the adult language. Each idiosyncratic vocal symbol used by Nigel during this phase was associated with one specific function. By the end of Phase I, utterances become more recognizable as words in the adult language system even though each utterance still performs only one function. It is impossible to mean more than one thing at a time. With Nigel, the first four functions of the list appeared first, followed by the appearance of the other two functions.







34

In Phase II, the transition into the adult system begins. For Nigel, this transitional period began between 160I2 and 18 months and continued until the end of his second year. During this phase, a seventh specific function emerged--the informative or "'I've got something to tell you" function. In this function, language is used as a means of communicating an experience to someone who did not share it.

The dominant characteristic of Phase II is functional generalization. Through this process, two broad functional categories or "macro-functions" emerge. The pragmatic function or "language as doing" derives from the instrumental and regulatory functions of Phase I. This function is reflected in utterances which demand a response such as "more meat," "momny come" and "fix train." The mathetic function or "language as learning" arises from the personal and heuristic functions, and includes utterances which require no response, "green car," "two book,'' and "tiny red light." The interactional function of Phase II contributes to both derivative functions.

It is interesting to note that Nigel made the distinction between these two functions more explicit by consistently producing pragmatic utterances with a rising intonation and mathetic utterances with a falling tone. This grouping of Phase I functions into pragmatic and mathetic macro-functions permitted Nigel to play two very different roles as a language user, an observer role and an intruder role. In the observer role, language served to express the experiences of his external and internal worlds. In the intruder role, language embodied his participation in the speech situation--his roles, attitudes, wishes and judgments. At the beginning of Phase II, all utterances were either pragmatic or mathetic. Gradually by the end of this phase, every







35

utterance was both pragmatic and mathetic, a significant advance toward the adult system. Nigel had learned that language may be used to both observe and interact with the environment at the same time.

Phase II was also marked by two major linguistic developments:

(a) rapid growth in vocabulary and structure, and (b) the emergence of dialogue. In Phase I, the intermediate level of linguistic form, which Halliday labels "lexico-grammer' begins to develop. For the first time, Nigel's language consisted primarily of standard lexical items. Initially, these items were restricted in meaning to one function only, e.g., "cat" meant only "hello cat!" (interactional). However, as Phase II progressed, Nigel learned to use the same word to express different functions in different grammatical contexts. The impetus for this rapid expansion in vocabulary seemed to be motivated not by the pragmatic concerns of obtaining objects and services, but rather by the mathetic concern for learning about the environment. New vocabulary is used, at first primarily in the context of observation and recall. The pragmatic function, however, seemed to contribute more to the development of language structure, the device which enables the speaker to play both observer and intruder roles at the same time. Structure frees the speaker to mean two things at once. Simultaneous with the acquisition of structure, the child learns to engage in dialogue. By dialogue, Halliday means the ability to adapt and assign communication roles in the context of a verbal interaction. For Nigel, this skill was mastered during a two-week period around 18 months when he learned to imitate dialogue and respond to a Wh-question, a command a statement, and a response.









Phase III marks the entry into the adult linguistic system, a system which is multifunctional and multistratal. For Nigel, this phase began around 24 months. It includes two major components or, according to Halliday, 'metafunctions," the ideational and the interpersonal. A third function, the textual, provides the cohesive framework within which the meaning of the other two components is organized. The ideational component derives from the mathetic, the observer function of language. It is the functional compoenet concerned with the representation of experience, "language as a means of talking about the real world" (Halliday, 1975, p. 17). The pragmatic, or Intruder function of language provides the context for the development of the interpersonal component. This component of meaning is concerned with ''the communication process as a form and as a channel of social action" (Halliday, 1975, p. 53). It is "language as a means whereby the speaker participates in the speech situation'' (Halliday, 1975, p. 17). The textual function implies genre, a mode of organizing meaning that relates language use to social context. For example,

. . . the texture of discourse depends not only on structuring the parts in an appropriate way and joining them together, but on doing so in a way that relates to the context-as narrative, as dialogue, or whatever generic mode is
selected. (Halliday, 1975, p. II1)

In Phase III, the notion of "function" is no longer synonymous with ''use." The child entering Phase III has acquired unlimited uses for language. Every utterance, whatever its use, has both an ideational and interpersonal component of meaning. Furthermore, the original Phase I functions and the informative function of Phase II have become the uses of language, "the generalized social contexts of language use" (Halliday, 1975, p. 58).







37

Global taxonomies of language function provide only the broadest outline of language use. Many of the finer discriminations of language use are not captured by their all-inclusive categorizations. A simple classification of child language using any of the functional schemes described should accurately characterize the purposes which language serves in the child's social and cognitive growth, but would not necessarily be descriptive of the functions that individual utterances May serve. It is this distinction which is often difficult to draw and maintain. Many authors reserve the term 'function" to describe the small number of higher order categories just defined and offer the terms "strategy' or "use" to describe the larger number of speech or communicative acts through which the functions of language are realized. A refined description of language use would venture beyond a mere description of language functions to a classification of the many interrelated options from which speakers choose when they communicate. These options, according to Halliday (1970), represent the meaning potential of language and involve the creative and repetitive selection of communication strategies in the context of a wide variety of social contexts.


Speech Acts Theory


The philosophers J. L. Austin (1962) and John Searle (1969)

have stated that speakers produce sentences because they are attempting to accomplish something with words. This emphasis on language use rather than language form or content probably stems from Wittgenstein's (1958) prophetic maxim that the meaning of an utterance is its use. Austin (1962), in his work How To Do Things with Words, distinguished









between the locutionary, illocutionary and perlocutionary acts which speakers perform when they use language. In saying something,

. . . we perform a locutionary act, which is roughly
equivalent to uttering a certain sentence with a certain
sense and reference . . .'meaning' in the traditional sense.
Second, . . . we also perform illocutionary acts such as
informing, ordering, warning, undertaking, etc., i.e.,
utterances which have a certain conventional force. Thirdly, we may also perform perlocutionary acts: what we bring about
or achieve by saying something, such as convincing, persuading,
deterring . . . (Austin, 1962, p. 108)

Austin's (1962) componential analysis of speech emphasized that all utterances are produced with the purpose or intent (illocution) of conveying a specific message or content (locution) to a listener upon whom they have some effect (perlocution) in modifying behaviors, thoughts or beliefs. In his original analysis of utterances, he used the term "speech acts" to refer to all three aspects of a message-locutionary, illocutionary and perlocutionary. More recently, the term has come to be reserved for only the illocutionary force of an utterance. The illocutionary speech act has been defined by Bates (1976b) as "a conventional social act, recognized as such by both speaker and hearer, that takes place when a sentence is uttered, e.g., a command is issued, a child is baptized" (p. 14).

Austin (1962) was the first to note that there were some utterances in which saying something could also be regarded as doing something or performing some action with words, as in the following examples:

Minister to couple: I now pronounce you man and wife.

Judge to prisoner: I sentence you to 30 days in the county jail.

One baseball fan to another: I bet you that the Pirates will beat the Dodgers.









In each instance, by saying something, the speaker performs an act: a couple is married; a prisoner is sentenced; a bet is placed. Each of these examples contains what Austin (1962) called a "performative" verb that makes explicit the acts the speaker intended the utterance to perform. As long as these acts are executed correctly, completely and with sincere intent by authorized persons under appropriate circumstances, then the specified acts are performed. However, should any of these conditions be violated, then the act is vold. These conditions, usually referred to as "felicity'' or ''happiness" conditions, must be respected or the utterances spoken do not perform the acts they encode.

Austin (1962) went on to develop the first taxonomy of speech acts, though by his own admission, his classification scheme was preliminary and tentative. In this first taxonomy of speech acts, every act could be classified under one of five very general categories: verdictives, exercitives, comissives, behabitves and expositives.

In 1969, J. R. Sear1e expanded upon and refined Austin's work according to the speech act, the important role which it now plays in the analysis of language use. To Searle (1969), the speech act is the basic unit of human communication upon which content and form are mapped for the purpose of conveying an infinite number of messages. As Searle (1969) states, 'speaking a language is performing speech acts . . . that � . . are in general made possible by and performed in accordance with certain rules for the use of linguistic elements" (p. 16). These rules delimit one use of language from another so that there exist a limited number of basic things which speakers do with language.







4o

Searle (1975) took issue with Austin's (1962) classification of illocutionary acts and offered an alternative taxonomy in its place. His main criticism centered around the fact that he could find no consistent set of principles upon which the taxonomy was constructed. This weakness, in turn, led to a great deal of overlap from one category to another and considerable heterogeneity within some of the categories. Throughout the system, there is a persistent confusion between verbs and acts and in some instances, the verbs listed are not even illocutionary verbs.

Searle's (1975) taxonomy of illocutionary acts presupposes at least twelve significant criteria for distinguishing one type of illocutionary act from another, including (a) differences in the purpose of the act; (b) differences in how the propositional content of an utterance relates to the world; and (c) differences in expressed psychological states toward the propositional content associated with various types of illocutionary acts. Applying these criteria, Searle (1975) delineated five categories of speech acts, including:

1. Representatives. Speakers convey propositions which they believe to be true, including suggestions, hypotheses and assertions.

2. Directives. Speakers attempt to get listeners to do something when they utter directives such as ordering, requesting, begging or pleading.

3. Commissives. The goal of commissives is to bind the speaker to some future course of action, as in promises, vows, pledges, contracts and guarantees.

4. Expressives. To express a psychological state about something, speakers utter acts such as apologies, welcomes, congratulations.









These acts specify how good or bad speakers feel about a state of affairs.

5. Declarations. When speakers utter declarations, they bring about a new state of affairs. By saying "You're fired" or "I now pronounce you man and wife, speakers change employment and civil status. Declarations are a special type of illocutionary act which Austin (1962) originally referred to as "performatives."

Each of these categories serves speakers in different ways to do things with words. Representatives tell listeners how things are. Directives attempt to get listeners to do things while commissives bind the speaker to do things. Expressives reveal the speaker's feelings and attitudes and declarations bring about changes in status and conditions.

A number of investigators have adapted Searle's (1969) classification scheme to describe how young children use their utterances. Greenfield and Smith (1976) focused on the assertions (representatives) and requests (directives) which children produce at the one-word stage. Antinucc and Parisi (1973), and Slobin (1970) performed a similar analysis on children's two-word productions. Grimm (in Cark& Clark, 1977) observed the ability of older children to use directives and commissives while Berko-Gleason (1973) and Berko-Gleason and Weintraub (1976) were concerned with the acquisition of expressives embodied in social routines.

Dore (1974, 1975, 1976, 1978) was the first investigator to develop a speech acts classification framework based on the utterances which children produce. While Halliday (1975) evolved an integrated theory of pragmatic development from the prelinguistic period through









the production of multiword grammatical utterances, Dore (1974, 1975, 1976, 1978) developed several classification systems which may be applied to the child's output at different linguistic stages.

Dore (1974, 1975) used a speech acts framework to analyze the single word utterances of two children at approximately 15 months of age. ge described early one-word messages as "primitive speech acts" (PSAs) consisting of two elements: a "rudimentary referring expression" and a "primitive force." The "rudimentary referring expression'' is the single word, the semantic component of the message, while the "primitive force" refers to the communicative intent of the utterance and is expressed by the prosodic pattern which accompanies production of single words. In this model of children's early utterances, a "rudimentary referring expression'' may be used to express several primitive forces or communicative intentions. For example, "mama" produced with falling intonation while the child pats mother on the knee serves as a label and requires no response. "Mama'' produced with rising intonation while the child unsuccessfully attempts to open a jar of peanut butter serves as a request for action and requires mother to perform the action. Finally, "mama said loudly with an abrupt rising-falling intonation while mother is across the room serves to call mother to the child. By varying the intonational contour of the utterance, it is possible for the child to use the same word to perform three different PSAs--labeling, requesting and calling. Dore (1974, 1975) used the segmental and suprasegmental phonemic features of the child's utterance as well as the child's nonlinguistic behavior, the relevant context of the utterance and the anticipated adult response to classify







43

children's single-word utterances into nine distinct PSA types. Each of these types is described with typical examples in Table 1.

Dore defines a PSA as "an utterance, consisting formally of a single word 2r a single prosodic pattern, which functions to convey the child's intention before he acquires sentences" (1974, p. 349). He emphasizes that 'a PSA is not merely an elliptical adult speech act, but a qualitatively different entity that possesses only some features similar to full speech acts" (1975, p. 32). For example, a PSA does not contain a predicating expression. It conveys a child's intention without containing propositional structure. ''. . . The components of PSAs eventually develop into the propositions and illocutionary forces of speech acts, but this occurs only after the child has acquired most of the grammatical structure of his language" (Dore, 1975, p. 32).

From 1976 to 1978, Dore revised and refined his "grammaticalillocutionary-interactional" model to classify the structure and function of nursery school conversation. According to the author, "the model is intended to show how grammatical forms are chosen to convey illocutionary intentions in the service of accomplishing social interaction (Dore, Gearhart & Newman, 1978, p. 339). Central to the model is a class of illocutionary acts which serve as the primary units of conversation and mediate between the grammatical forms that signal them and the interactional purposes for which they are used.

In this three-part model, utterances are classified as conversational acts, each consisting of a propositional content, grammatical form and an illocutionary function. Thrty-five individual conversational acts have been identified through the analysis of videotaped sessions in which seven 3-year-old children engaged in a wide variety










Table 1

Dore's Primitive Speech Act Types


Primitive
Speech Act Descriptive mf Evample Labeling Child says "eyes'' while touching doll's eyes Repeating Child says "data" after overhearing mother say doctor
in a conversation with the teacher

Answering Child says "bow wow" after mother points to a picture
of a dog and says ''What's this?"

Requesting Child says ''uh? uh? uh?'' while unsuccessfully trying (action) to push a peg through a hole and mother responds by helping

Requesting Child says "book?" while picking up a book and looking (answer) at mother

Calling Child loudly shouts "mama" with distinct intonation
while mother is across the room

Greeting Child says "hi" when teacher enters room Protesting Child says "no" while resisting mother's attempts to
put on his shoes

Practicing Child says ''daddy'' when he is not present; mother does
not respond


Modified from: J. Dore's "Holophrases, Speech Acts and Language Universals." Journal of Child Language, 2, 21-40 (1975).









of spontaneous interactions at school. Over a period of four months, 3,000 child utterances were collected and coded for conversational act. The 35 conversational acts identified accounted for 92 percent of the utterances produced by the children. To determine the conversational act which best characterized each utterance, Dore (1976, 1977) used information both internal and external to the utterance, e.g., the literal meaning of the utterance, its grammatical and prosodic characteristics, how it relates new information to old, the speaker's nonlinguistic behavior and related utterances, the verbal and nonverbal behavior of the speaker's interlocutors and the relevant situational context.

The 35 individual acts may be grouped together into six general conversational classes, including: (a) Requestives which solicit inforeation or actions; (b) Assertives which report facts, state rules, or convey attitudes; (c) Performatives which accomplish acts and establish acts by being said; (d) Responsives which supply solicited information or acknowledge remarks; (e) Regulatives which control the conversational sequence; and (f) Expresslves which nonpropositionally convey attitudes or repeat others. Three primary conversational functions are conveyed by these general classes: the transmission of content, the regulation of conversation, and the expression of attitudes. Figure I presents a network representation of Dore's (1978) threelevel coding scheme.

Two other classification schemes which have combined a speech acts approach within a functional framework for the purpose of describing language use in children are those developed by Wells (1973) and Tough (1976; 1977).





46





PRIMARY GENERAL PARTICULAR CONVERSATIONAL CONVERSATIONAL CONVERSATIONAL FUNCTION CLASS ACT

solict ..Cho.t Quest Inforeutin Prodct Quest lots



rdeuifutios
(renquost lot) i


ci tastinternal itnterno I Reports
0.) -t ernoma Eluatiocs
Attributios
sociall" Rules phesnon- "- tEplianotioos
C1 i's
roltal Jokes C noty .lea s
I(Perforetltt).Lre P "rotests


1Suppl Choice Anssoers solicited Podt Asos ...or.at.o Pcess Ansoers
j .... ,......




se xs upply rcia,ifRicat lo
respond - (responsivt)--- oddisltral1 Respooses nitortaioo Tual Iftatloos
Agreeents
f. -Is IlQe "I... Ak ldg s

-tos-requst ve4 tsoeqoet
... I'g - [............
Att- Getr
soi ide Steaker Selections oter Rhetoica I uestios regulate --_ _ rguar_ y)- _t on
(rtegIulotio Qnstls ear Bundoroysarkers coolest Pol stress rke-s

,pre,,Ecla.t lots attitude (eopressive) Ac~ooanties Repetitions

odified I r- J. tore's o .arlatio IR Preschool Childre's C.-vrsational Perforancms." In K E Nelst (ed.). ChIldrets Lanouane V. te cork: Gardner Press (197 ).

Figure I

Dore's Codes, Definitions and Examples of Conversational Acts









The Wells (1975) coding scheme classifies the interpersonal function, cognitive content and discourse structure of preschool children's talk (Dore, Gearhart & Newman, 1980, p. 36o). Most relevant to the present research, he posited 133 functions of communication which can he categorized under several sequence and subsequence conversational nudes. Wells (1973) described these functions as "acts" that a particular utterance performs. Like Austin (1962), he was concerned with the purposes that individual utterances serve. However, he went on to emphasize that his interest was not in the analyses of utterances in isolation but rather in the description of utterance function in the context in which it occurs. Wells (1973) viewed communication functions or acts as "the smallest units of verbal interaction--the building blocks from which the edifice of conversation is constructed" (p. 39).

In his system, Wells has identified six conversational sequence and subsequence modes. Sequences indicate the dominant purpose or function of a conversation; subsequences mark the subsidiary functions of smaller units of conversation within the overall purpose of the sequence. The sequences or functions, along with exemplary "acts" are as follows:

I. Control. The regulation of the present or future behavior of one or more of the participants through acts such as wanting, offering, commanding, suggesting and permitting.

2. Expressive. The expression of feelings and attitudes as an affective response to situations through acts such as exclaming, taunting, challenging, approving and disapproving.

3. Representational. The exchange of information through acts such as labeling, cormnenting, questioning, responding and justifying.









4. Social. The use of conversation to maintain social relationships through acts such as greeting, leave taking and ritualizing.

5. Tutorial. The use of conversation for didactic purposes through the acts of correcting, modeling and imitating.

6. Procedural. The use of conversation to initiate or end a sequence or to resolve a breakdown in communication through acts such as calling, requesting a repetition and reformulating.

In a conversational sequence between parent and child, the parent's dominant purpose may be control--to get the child to shut the door upon entering the house. To accomplish this purpose, it may be necessary to proceed through several conversational subsequences. "There may be an initiating (Procedural) subsequence to gain the child's attention, followed by parent's exclamation of surprise or horror (Expressive) before reverting to the original request to which the child complies (Control) (Wells, 1973, p. 81).

In Wells' (1973) system, the same act such as requesting, responding and justifying may occur under different functions or sequence modes. This, in part, accounts for the large number of acts. More importantly, ''the acts are formulated to cover a wide array of phenomena, ranging from the purely linguistic (question) to the social (silence filler, evasion). As a consequence,

a generality crucial to act types may be lost
. . . and the acts are formulated at different functional
levels. Thus, the central problem with this scheme, as well
as with the other empirically based speech act codes for
children . . . is finding principled criteria for (a) motivating the level of linguistic act identified, (b) constraining
the kinds of acts to be included, and (c) defining a decision
procedure for classifying utterances into acts. (Dore, Gearhart
& Newman, 1980, p. 370)









Tough's (1976) bilevel classification system of language use evolved as a result of a longitudinal study undertaken to account for the role of language in the educational differences of young children from disparate social environments. Borrowing from the theoretical viewpoints of Bernstein (1971, 1975), Bruner (1964, 1966), Luria (1959), Piaget (1923) and Vygotsky (1962), Tough (1977) devised a taxonomy of language use to fulfill two major objectives: (a) to discover differevens in the range of meanings that a child attaches to his environment; and (b) to determine the purposes language serves for young children. Table 2 presents an outline of Tough's (1976) functional classification schema. Operational definitions and examples of the various cornmunication strategies can be found in Appendix B.

At the broadest level of classification in Tough's (1976, 1977) schema, utterances are categorized, by use, into seven major types:

1. Self-maintaining. The use of language to create an awareness of the speaker's identity and to promote the speaker's position in relation to others.

2. Directing. The use of language to control or regulate the physical actions and operations performed by oneself and others.

3. Reporting. The use of language to provide information about past and present experiences.

4. Reasoning. The use of language which employs rational thought and argument to interpret experiences.

5. Predicting. The use of language to extend communication beyond immediate present or past experiences to events that have not occurred and which may never take place.










Table 2

Tough's Framework for the Classification of Language Use



1. SELF-MAINTAINING

Strategies

a. Referring to needs
b. Protecting the self and self-interests
c. Justifying behavior and claims
d. Criticizing others e. Threatening others

2. DIRECTING

Strategies

a. Monitoring own actions
b. Directing the actions of the self
c. Directing actions of others
d. Collaborating in action with others

3. REPORTING

Strategies

a. Labeling
b. Referring to detail
c. Referring to incidents
d. Referring to the sequence of events
e. Making comparisons
f. Recognizing related aspects
g. Extracting or recognizing the central meaning
h. Reflecting on the meaning of experiences

4. TOWARDS LOGICAL REASONING

Strategies

a. Explaining a process
b. Recognizing causal and dependent relationships
c. Recognizing problems and solutions
d. Justifying judgments and actions
e. Reflecting on events and drawing conclusions
f. Recognizing principles







5I


Table 2 (continued)



5. PREDICTING

Strategies

a. Anticipating/forecasting
b. Anticipating the detail of event c. Anticipating a sequence of events
d. Anticipating problems and possible solutions
e. Anticipating and recognizing alternative courses of action
f. Predicting the consequences of actions or events

6. PROJECTING

Strategies

a. Projecting into the everpiences of others
b. Projecting into the feelings of others
c. Projecting into the reactions of others
d. Projecting into situations never experienced

7. IAGINING

Strategies

a. Developing an imaginary situation based on real life
b. Developing an imaginary situation based on fantasy
c. Developing an original story


From: J. Tough's Listening to Children Talking. London: Ward Lock Educational (1976).









6. Projecting. The use of language within an unfamiliar or external context.

7. imagining. The use of language to create an environment of make-believe.

Language use is defined as "the means by which different kinds of meaning or thinking are made evident" (Tough, 1977, p. 46). In a child's development, the purpose of language is to construct and express these meanings so that "language functions in relation to the child's developing conceptualization of the world around him" (Tough, 1977, p. 44) and reflects different modes of thinking. Five of the basic uses of language defined above-reporting, reasoning, predicting, projecting, and imagining--convey meaning reflective of an ideational appraisal of experience and could be collectively referred to as cognitive uses. The self-maintaining and directing uses convey meaning of an interpersonal nature by indicating the kind of relationship maintained between speakers and listeners and can be referred to as social uses.

At the second level of classification, utterances are categorized according to a number of strategies that serve each language use. The strategies bear some resemblance to speech acts (Austin, 1962; Searle, 1969); however, in this instance, it seems more appropriate to characterize them as "thought acts," since it is through the employment of these strategies that inferences can be made about the use of language characteristic of a number of modes of thinking. It is interesting to note that Tough (1977) claims to have arranged the strategies serving a particular use of language in an ascending order of cognitive complexity.







53

Tough's (1976, 1977) system seemed particularly well-suited to the present investigation since it was developed to resolve issues about the relationship of language use to academic success. It is the only system which claims to have identified the communication strategies which children need to master to succeed in the classroom.


Development of Language Use in
Language-Normal Chi ldren


The rules governing the use of language in context form the foundation of the study of pragmatics. According to Bates (1976a),

- * . all of language is pragmatic to begin with. We
choose our meanings to fit contexts and build our meanings
onto those contexts in such a way that the two are inseparable
in the same way that 'figure' is definable only in terms of
'ground.' (p. 420)

Studies in the acquisition of language use begin prior to speech itself, in the first year of life. "The studies indicate that semantics emerges, developnentally and logically, from pragmatics, in much the same way that syntax has been shown to emerge from semantic knowledge" (Bates, 1976a, p. 420). Miller and Yoder (1972) state, "Before the child becomes a language user, he needs to have something to say (concepts) and a reason for saying it (semantic intent) as well as a way to say it (linguistic structure)" (p. 210).

Halliday (1975), Bloom and Lahey (1978), and Prutting (1979) have identified a series of stages which characterize the development of language use from infancy to adulthood. At the first level of primary forms, the infant uses a limited repertoire of behaviors (e.g., gazing, crying, touching) to meet several basic physiologic and/or affective needs, regardless of the context. During this period, from







54

birth to approximately 9 months, infants accomplish nonlinguistically those communicative functions which they will later learn to convey linguistically (Brunner, 1974; Halliday, 1975; Mahoney, 1975). Bates, Camaionl and Volterra (1975) have labeled this the periocutionary stage of development, corresponding to one aspect of Austin's (1962) tripartite characterization of speech acts. Perlocutions have an effect on the listener but they are not recognized by both speaker and listener as conventional communications (Bates, 1976a). In early communications between parent and child, there is no evidence that the infants, themselves, are aware of the communicative value of their signals. Parents react to the infant's signals as if they were produced intentionally, thereby bringing about the desired effect, but the infants, themselves, are not aware of the conventional purpose of these signals. Therefore, 'these early sigmals function as signals only for the listener" Bates et al ., 1975, p. 212).

By the second level of conventional forms, children have learned sounds and gestures which they apply in different ways. At the beginning of this stage, just prior to the emergence of the infant's first words, meaning is transmitted intentionally through sound and movement. Infants are unable to use the conventions of language but they ''come to know that their behavior can influence the behavior of others and they behave with the intention of achieving that influence" (Bloom & Lahey, 1978, p. 204). Bates et al. (1975) have labeled this the illocutionary stage of development, since it is during this period, from approximately It to 14 months, that infants intentionally use vocalizations and gestures such as showing, giving and pointing to regulate joint attention and joint actions (Bruner, 1974). These behaviors, performed in at least two different cormmunicative contexts,









have been termed "proto-declaratives" and "proto-imperatives" by Bates et al. (1975). Through the proto-declarative, the infant directs the listener's attention to some object or event. The proto-imperative serves as a means to use the listener to obtain a desired object. These early communicative behaviors are carried out in the absence of speech and are the precursors to the development of symbolic communication.

The acquisition of conventional mans of communication begins to emerge around 14 months of age in what Bates et a]. (1975) have labeled the locutionary stage. At this point, infants acquire words which they substitute for the nonlinguistic performative schemes developed during the illocutionary stage. Their communication, in this stage, is intentional as well as conventional, even though the very first words which they produce seem to be an integral part of the gestures which accompany them, e.g., "bye-bye + hand wave" or "up + hands outstretched." The goals of these early communications continue to be linked to establishing joint attention and joint actions and are classified as assertions and requests. Assertions correspond to the proto-declaratives described earlier and requests correspond to proto-imperatives. Clark and Clark (1977) have speculated that,

. . . gestures and speech acts are parallel in function:
pointing gestures require acknowledgement from the listener
just as assertions do, and reaching gestures require an action
or answer just as requests do. This hypothesis gains added
support from the fact that children continue to use pointing
and reaching gestures together with single words and even combinations of words. (p. 313-314)

In addition to these two principal categories of early speech acts, children at the one-word stage of development have been observed to use a few primitive expressives such as greeting someone with "hi'' or expressing dismay with "uh-oh."









As the form of children's early utterances becomes more elaborate, their functional capabilities expand also. Slobin (1970),

Bowerman (1973) and Brown (1973) have observed that children at the two-word stage use assertions to acknowledge the existence or nonexistence of an object, to describe the location, possession or quality of an object and to comment about an ongoing event. Requests are of two types: requests for action and requests for information.

At the third level of conventional use, children elaborate their use of language along two dimensions. First, as they grow in linguistic ability, they add new types of speech acts to their repertoire. According to Clark and Clark (1977), they begin with representatives (assertions) and directives (requests) and then add expressives, commissioes and lastly, declarations. Second, they learn that alternative linguistic forms exist for achieving the same purpose in different situational contexts. To get a cookie, it's possible to ask a question, ''May I have a cookie?"; to make a statement, "I want a cookie"; to issue a command, "'Gmmie a cookie!'; or hint in a less direct fashion, ''That cookie sure sme lls yummyl'' The present study focuses on the former dimension--the development of a set of intentional options for communication in a wide variety of cognitive situations. This stage of development generally begins in the third year of life and continues into adulthood.

Only a few investigators have described the older child's

increasing ability to communicate a large number of intentions through language. Piaget (1923) was among the first. His system was adopted and modified by McCarthy (1930) and Davis (1937) in their studies of language development in children. Although there was some disparity







57

in the results obtained by each of these investigators, and in addition, by Day (1932) and Smith (1935), the sampling protocol in the studies was different. Piaget's (1923) subjects experienced complete freedom of movement as their communicative exchanges were recorded during free play situations with other children at school. Smith's (1935) data were collected under similar circumstances of an unstructured nature. The communicative interactions of her subjects were recorded either while children were at play with one another or in the home environment with adults. In contrast, McCarthy (1930), Day (1932) and Davis (1937) gathered their samples from children who were looking at books or playing with toys in highly structured situations with adults.

As a result of these differences in sampling strategies, the overall ratio of egocentric to socialized speech was different for the studies. Piaget (1923) and Smith (1935) observed egocentric speech occurring from 26 to 40 percent of the time; the other investigators reported this type of speech in less than 10 percent of the utterances, with children beyond 4 years of age engaging in its use I to 2 percent of the time. Conversely, socialized speech constituted between 60 to 84 percent of the conversation used by Piaget's (1923) and Smith's (1935) subjects and between 93.5 and 99.8 percent of the speech used by the subjects in the other investigations (Davis, 1937; Day, 1932; McCarthy, 1930). This wide discrepancy in the functions which language serves my be one of the first pieces of evidence to support the now widely held belief in the important influence of context or situation on what is being said, how it is being said and for what purpose. Children engaged in highly structured communication with adults have little opportunity to use egocentric language. The situation demands that








their remarks be socialized. On the other hand, when children are playing alone or in the presence of other children, it is not uncommon to hear them talking to themselves "or for the pleasure of associating anyone who happens to be there at the moment'' (Piaget, 1923, p. 32). Taken together, the four studies (Davis, 1937; Day, 1932; McCarthy, 1930; Smith, 1935) demonstrated a decrease with age in the amount of egocentric speech used. This finding lends support to the claims by Piaget (1923) and Vygotsky (1962) that in later life, inner language or thought supplants this type of talk to self.

With regard to the present study, the observation that "answers" comprised the most important category at all ages is of particular importance (Davis, 1937). The frequency of occurrence of this type of speech ranged from 25.5 percent at age 5192 to 37.1 percent at age 91/2. McCarthy (1930) also found the percentage of answers to be increasingly important in her upper two groups. Answers comprised from .3 percent of the sample at 1 1/2 years to 31 and 26 percent, respectively, at 4 and 4/ years. That answers play a vital role in the child's communicative system, both at school and at home, is further supported by estimates of the relative rates of parental questions to their young children. These range from 1O percent of the speech addressed to 12-month-old children (Friedlander, Jacobs, Davis & Wetstone, 1972) to 32 percent of the speech addressed to 13-month-old children (Nelson, 1973) and 37 percent of the speech addressed to children between the ages of 2 and 5 years (Nelson, 1973; Baldwin & Baldwin, 1973). Labov (1970) has been critical of the question-answer format in obtaining information about the language development of disadvantaged children. He has observed language which is more elaborate in both







59

form and function collected under less structured and more familiar circumstances. Although his observations would certainly indicate that, given the right set of conditions, these children do demonstrate complex linguistic skills, it does not offer a solution to the persistent problem of their limited achievement in the most common interactive mode between teacher and child and parent and child.

Schachter et al. (1974) identified developmental changes in the pattern of language use in 170 preschool children, ages 2 to 5. Using the FIS-P, a Piagetian-based taxonomy of language use, children's spontaneous initiations of speech were analyzed. The results of this analysis revealed a pivotal shift in the development of speech function patterns at age 3. Before this age, patterns of speech function consisted mainly of more immature forms such as desire implementing, reporting on self and things, "me too" self-referring and learningimplementing (naming words), with adult-addressed speech at its highest level. Schachter (1974) and her colleagues labeled this collection of early speech functions "Primary Socially Interdependent Speech,' noting that it was reflective of young children's tendency to use language egocentrically rather than sociocentrically. Functionally, these patterns served to insure "mutual gratification during the primary interdependent attachment between child and caretaker" (Schachter et al., p. i).

After age 3, the early speech patterns maintained their frequency but speech reflecting increased levels of self-other differentiation showed a marked rise. The use of ego-enhancing boasting statements increased abruptly and a number of peer-addressed collaborative and joining statements called "Secondary Social Speech"







6o

progressively rose with age. Finally, speech patterns including modulations (explanations and justifications) and collaborative disagreeing, which seemed to be adapted to the needs of the listener, emerged with high frequency at age 4 or 5. These patterns were labeled "Tertiary Socialized Speech." The authors concluded that the passage from primary to tertiary patterns of speech usage was "consistent with both Piagetian and psycholinguistic theorizing on the emergence of a significant degree of ego differentiation around 3'' (Schachter et al., 1974, p. 47).

Sore (1976) videotaped seven middle-class children over a period of seven months as they engaged in a wide variety of activities at their nursery school, in order to classify their communicative intentions. The children ranged In age from 34 to 39 months and produced almost 3,000 illocutionary acts. These acts could be classified into 32 different types grouped under six major categories. A distributional analysis of the relative proportions of use of each of the major category types revealed the following results: requests (27 percent), responses (18.5 percent), descriptions (22.3 percent), statements (13.8 percent), conversational devices (5.8 percent) and performatives (10.8 percent). Another 7.9 percent of the remarks were coded as uninterpretable and 5.8 percent were double coded, as the intent of the utterances was equivocal. It can be seen that requests accounted for the largest proportion of the children's illocutionary acts, with requests for action (10.0 percent) and requests for information (7.3 percent) outnumbering their request types. Descriptions constituted the second largest category of illocutionary acts, with major contributions to the overall proportion made by identifications (7.9 percent)







61

and descriptions of events (4.0 percent). The third largest category of acts was statements with internal reports about feelings and attitudes (7.7 percent) contributing most to the overall frequency of occurrence.

Fogel (1976) interviewed 288 children between the ages of five and fifteen to study developmental patterns of functional communication. His interview technique paired verbal prompts with pictorial stimuli to elicit speech acts reflective of language use for two different functions: controlling and feeling (Wells, 1973). Twelve stimulus items were designed to represent a wide variety of "real life" communication situations. In one item, for example, children had to pretend that they were a teacher trying to explain why a student could not have a book to read. The situations provided information about three important factors in any comnmunicative exchange: (a) the function of the communication (controlling or feeling); (b) the significant other in the situation (mother, teacher or peer); and (c) the perspective taken within the situation (commenting on behalf of oneself or someone else).

Three general findings emerged from the analysis of the speech act data. First, there was a significant age effect represented mainly by a change in the communication behavior of the fifth grade children. Second, there was a significant effect involving context which indicated that the primary basis upon which children differentiated between comnunication situations was the function of the connunication employed and not the participants involved. Third, there was a significant interaction between context and perspective which Fogel (1976) interpreted to mean that perceptions of communication differences between the self and other are dependent upon the context in which one finds oneself.









Further analysis of the results revealed that the shift in communication behavior at the first grade level was accounted for by increased diversity in speech act usage and increased use of parasituational comments--statements about the communication situations rather than statements reflective of actual dialogue. These comments set the scene for dialogue by explalelng aspects of the situation to the interviewer in order to enhance understanding. With regard to this behavior, Fogel (1976) hypothesized "that older children are more likely to present rules about situations than to present actual behaviors that may occur" (p. 209).

Three speech act types accounted for the developmental trends observed in this study including direct requests, cajoling and contractual acts. The use of direct requests for permission to do something or to go somewhere declined from the first and third grade levels (64 percent) to the ninth grade (6 percent). According to Fogel (1976), "this finding supports previous indications in the literature that younger chlldren are more likely to employ "head-on" communication strategies when attempting to persuade others (pp. 209-210). Similarly, the speech act type of cajoling, i.e., pleading to change the mind of someone about a decision, decreased in total usage from the first and third grade levels (50 percent) to the ninth grade level (3 percent), and occurred more frequently in communication situations involving the mother. Finally, contractual type acts were used very infrequently by first (12.5 percent), seventh (0.78 percent) and ninth (17.13 percent) grade children, but peaked in use by third (32.41 percent) and fifth (30.09 percent) grade children.







63

Fogel (1976) also observed children in the earlier grades using a wider variety of speech acts when communicating from another's perspective than when communicating from their own perspective. This effect was most pronounced when children took the role of teacher and least pronounced from the mother's perspective. In addition, feeling situations showed this effect more strongly than controlling situations. In other words, children saw the other as having a more diverse repertoire of responses than the self, particularly with regard to feeling situations. Fogel (1976) hypothesized that this greater discrepancy between self and other in the feeling situations might be due to the relative unfamiliarity of these situations for the young child. He concluded that younger children attributed to others a greater variety of speech act capability than they attributed to themselves, which demonstrated that they have difficulty reflecting on themselves as communicators in a variety of situations, particularly those which are unfamiliar.

Several investigators have studied older children-s use of specific types of speech acts such as directives (Ervin-Tripp, 1977; Garvey, 1975; Grim, 1975; Mitchell-Kernan, 1977), commissives (Grimm, 1975) and expressives (Berko-Gleason & Weintraub, 1976). Grin- (1975) had children of 5 and 7 years roleplay situations with Felix, a large toy cat, in order to get him to comply with four directives and a commissive. Both the 5- and 7-year-old children performed consistently well when asking Felix for something, ordering him to do something and forbidding him to do something, but the 5-year-olds had some difficulty formulating acceptable utterances for permitting Felix to do something. These results showed that 5-year-old children had not yet mastered a








complete repertoire of directives. The commissive presented problems for both the 5- and 7-year-olds, with 57 and 55 percent of their remarks achieving acceptable standards. Clark and Clark (1977) suggested "that children find it much easier to work out the conditions under which the listener is expected to do something than the conditions under which the speaker is" (p. 366). These findings support Chomsky's (1969) comprehension studies of directives and commissives which showed that children understood the directive tell, long before they understood the commissie, promise. Under 8 years of age, children interpreted the verb .promise" as if it were a directive.

Garvey (1975), Ervin-Tripp (1977), and Ilitchell-Kernan and Kernan (1977) observed the use of directives in young children. Children's ability to convey and respond to requests for action based on spontaneous dyadic interactions of younger children (3.9-4.4 years) and older children (4.7-5.7 years) were explored by Garvey (1975). Both younger and older dyads were observed to produce equal numbers of successful direct requests (e.g., "Open the doorl"). Fewer indirect requests (e.g., "Can you open the door?") appeared in the data, with the older dyads achieving, on the average, twice as many successful attempts as the younger dyads. Among the direct requests produced, no examples of requests with performative verbs were noted (e.g., I order you to open the door!"). Garvey (1975) concluded that for children in the age range studied,

the request for action rests on a set of meaning
factors which are relatively specific, which may be expressed
in variant forms, and which are available to the child in either
the discourse role of requester or of recipient of the request.
(p. 62)







65

Ervin-Tripp (1977) traced the evolution of children's directives from their earliest expression prior to 2 years of age. At this age, children produced directives with combinations of gestures, names of desired objects and some limited linguistic markers like "more" and 'want." By 3 years, they were capable of using embeddings and structural modifications to produce a number of alternative directive forms such as "Can I have my big boy shoes?'' and by 4, they employed verbal strategies requiring several steps as well as hints to negotiate interactions, e.g., "We haven't had candy in a long time." The hardest forms for children to master were those that did not explicitly identify what was wanted--question directives of an indirect type and affirmative hints, e.g., "It's hot in here." Ervin-Tripp (1977) concluded that "wide use of tactful deviousness is a late accomplishment" (p. 188).

In a study of older children from 7 to 12 years of age,

Mitchell-Kernan and Kernan (1977) examined the use of directives by a group of black children as they occurred in conversation with one another and with adults. Using Ervin-Tripp's (1977) classification scheme of directive variants, the authors found that these children had mastered all of the forms appropriate for directives and showed an awareness of at least some of the social factors which help to determine the situational appropriateness of the use of one form over another. The types of directives included: (a) need statements, e.g., "I don't want no more fighting out of the girls''; (b) imperatives, e.g., "Let my brother alone"; (c) imbedded imperatives, e.g., "John, would you please tell that lady to quit?"; (d) permission directives, e.g., "Can I speak to her?"; (e) question directives, e.g., "Hey, you







66

got a quarter, Mac?'; and (f) hints, e.g., "Last person talk to me like that is in his grave."

Although a few early expressives appear in the language of 2and 3-year-old children, their use, in general, is not mastered or fully comprehended until much later in the child's development. This aspect of language is one that parents teach by rote in the appropriate social contexts by insisting that children say "please," "thank you," "you're welcome," "I'm sorry." The use of expressives places no obligation on either speaker or listener; rather they express the feeling expected within a particular society for a particular situation. "As such, expressives are often hard to explain or justify to small children, and parents don't usually try" (Clark & Clark, 1977, p. 366).

As the literature demonstrated, knowledge about the numerous communicative intentions which older children successfully employ is fragmentary. Several investigators have used more comprehensive classification schemes to collect the broadest data base possible while others have focused on singular examples of specific speech act types. In the majority of investigations, samples have been small with considerable variety and overlap from age to age; nevertheless, some developmental trends are beginning to emerge.


Development of Language Use in
Language-impaired Children


Primary Forms

During the first nine months of life, normal infants use their voices and their bodies to attract attention, express emotion and to engage in social exchanges with familiar adults. Children at high risk







67

for later language impairment do not demonstrate the same facility with the nonverbal conunication system. Fisichelli and Karelitz (1963) observed abnormI crying patterns in brain-damaged infants during the first week of life. In interviews with mothers of autistic children, Schaeffer (1971) recorded almost a complete absence of crying in the early months of life. Mothers consistently revealed that their autistic infants never cried for attention nor when they were hungry and their responsiveness to pain and discomfort was impaired.

Ricks (1975) investigated the way in which autistic children and retarded children expressed emotional meaning in four situations:

(a) requesting; (b) frustration; (c) greeting; and (d) pleasant surprise. The parents of the autistic children were asked to listen to the recordings of their own child, two other autistic children and one retarded child and identify the meaning of the messages conveyed. Although the parents accurately identified the sounds made by the retarded child, who they said sounded "normal,'- they could not identify the sounds made by the other two autistic children. Ricks (1975) concluded that autistic children were capable of expressing the four emotions but they did so in a personal idiosyncratic way dissimilar from that of normal babies. Whereas normal babies used intoned vocalizations in response to the situations, autistic children's sounds were articulated.

Kinesic differences in the nonverbal communication systems of children at high risk for language impairment have also been noted. Schmidt and Erikson (1973) interviewed mothers of retarded children, 16 percent of whom reported delayed smiling responses in their infants. Gaze aversion was a common phenomenon observed by mothers of autistic







68

children--a characteristic which interrupts substantially normal motherchild interactions (Schaeffer, 1971). Facial expression, hand movements and body postures also were slow to evolve in these children. Ricks and Wing (1975) noted that autistic children pass through a concrete denonstration phase (e.g., pushing mother's hand to the lightswitch to get her to turn on the light) on the slow and imperfect path toward acquisition of symbolic pointing. In fact, though 50 percent of autistic children remain mute all their lives, they do not use gestures as a substitute for speech.


Conventional Forms

In the stage of conventional forms, normal children use sounds and gestures in a variety of ways to communicate a message. Snyder (1975) studied the pragmatic performance of 15 language-normal and 15 language-impaired children at the one-word stage of development. The children were matched for socioeconomic level and utterance length and were screened for cognitive development. The language-normal subjects had a mean age of 14.9 months, while the language-impaired subjects had a mean age of 24.2 months. Three experimental measures were developed to elicit presuppositional, declarative performative and imperative performative responses.

The analysis of the data revealed that the older languageimpaired children, while at the same stage of development in terms of utterance length and lexicon, were deficient in their use of language. On the presuppositional measures, the language-impaired children could encode the most informative element in a context almost as often as they encoded a more imformative one. On both the declarative and









imperative performative measures, the language-impaired children performed more poorly than the language-normal children, and the languageimpaired children generated significantly fewer verbal and nonverbal performatives in response to both types of measures, with nonverbal responses predominating over verbal ones. Snyder (1975) offered her findings as support for 'a specific representational deficit in the language-disabled child which affects the dynamic aspects of symbolization" (p. 167).

Snyder's (1975) study bridges the transition between unconventional and conventional language use. With regard to the use of unconventional forms, she found that language-impaired children used nonverbal performatives, both declarative and imperative, more often than verbal performatives. These children preferred grasping, reaching, smiling, etc., to linguistic symbols to get the examiner's attention and they substituted looking, looking and reaching, looking and fussing, etc., for words to get the examiner to perform a desired action. Even though the number of nonverbal imperatives was significantly greater for the language-impaired children than the normal children, the level of these nonverbal imperatives was significantly lower. In the use of conventional forms, the language-impaired children had difficulty generating verbal performatives. In fact, the differences between the performatives generated by the two groups approached the greatest significance when the performatives compared were linguistic.

Geller and Wollner (1976) investigated language use in a group of three older language-impaired children, ages 3.11 to 5 years, who were functioning at a mean length of utterance level from 1.1 to 1.6. The authors videotaped the children during snacktime in a typical









preschool setting and analyzed their communication skills in a variety of ways. The most significant analysis for the purposes of the present study was a classification of the types of communication intentions conveyed by the children's verbal acts. When the results of this study were compared to those of Bore (1975), collected for 3-year-olds in a similar communicative context, it appeared that the language-impaired children were deficient in language structure as well as language use.

The findings of Snyder (1975) and Geller and Wollner (1976) suggest that some language-impaired children "may be more deficient in the use of language for communication than even their limited mastery of vocabulary and syntactic structures would allow" (Rees, 1978, p. 258). Other studies do not share this observation about the language use of language-impaired children.

Owens (1978) devised a modification of Dore's (1974) primitive speech act taxonomy to evaluate the language use of twelve children, six with Downls syndrome and six nondelayed. Three of the children in each group were performing at Brown's stage I (MILL) = 1.0 to 2.0) and three at Brown's stage II (MIL = 2.5 to 3-0) with regard to linguistic competence. Audio and video recordings were made of the children in a freeplay situation with their mothers in the home in order to obtain 100 intelligible child utterances for analysis. A comparison of speech act distribution among the four groups revealed essentially the same patterns. Only two speech acts, declaration and practice, achieved significant group differences.

Two studies have investigated the young hearing impaired child's ability to communicate, either verbally or nonverbally, a variety of pragmatic intentions (Curtiss, Prutting & Lowell, 1979; Skarakis &









Prutting, 1977). Communication samples were videotaped during four different situations--freeplay, snacktime, group lesson and individual lesson--and were analyzed using a modification of Dore's (1974, 1975) taxonomy to include gestural behavior. Although Skarakis and Prutting (1977) found all communicative intentions expressed by all subjects, with requests/demands, descriptions, attention and responses occurring most frequently for all children, the realization of these intentions was carried out primarily through nonverbal means. These nonverbal behaviors are similar to those used by younger normal-hearing subjects and have been found to be precursors to later linguistic development. The authors concluded that hearing-impaired children expressed the same communicative intentions as normal-hearing children and demonstrated that they had acquired the basic foundation on which later language develops.

Curtiss et al. (1979) investigated a larger sample of hearingimpaired children who spanned a wider range in age, 22 to 60 months. They found that, overall, hearing-impaired children coded a variety of communicative intentions using both verbal and nonverbal means. All age groups exhibited all communicative intentions, but the distribution of specific category types varied with age. It was interesting to note that the number of communicative intentions expressed increased with age, but the mean length of utterance remained the same. This finding seems to indicate that pragmatic development precedes development in the other components of language and provides the foundation upon which these later linguistic skills are constructed.

Only a few investigations of the communicative intentions expressed by language-impaired children have been undertaken. The







72

findings have been scattered over a wide range of ages and developmental levels and have been insufficient in determining the true nature of language use in these children. The results of some of the studies are suggestive of a delay in language use among the language-impaired. The same features of language use have been observed in their communications but they occur less frequently and at a later age. Other studies are more indicative of disordered language use. Snyder (1975), for example, found differences in language use among her subjects who had been matched for general linguistic level.


Language Use in the Culturally Diverse


A child entering school is well on the way to mastering the essential vocal symbols and the complex grammatical system of English. The typical first grade child already knows the language of home, neighborhood and community. The child is already generating sentences effortlessly and spontaneously through unconscious use of grammatical "rules" induced from language as acquired and developed, but does not yet have skillful control of language use in a wide variety of everyday interactions (Hansen, 1974).

Children acquiring language can only acquire that language to which they have been exposed. The child learning language must acquire the rules of pronunciation, grammar and usage. The rules of usage vary fro culture to culture, but they will be learned in the immediate family during the preschool and elementary years (Brown, Ecroyd, Hopper & Narenore, 1976). Even though children from different environments may have access to the same set of language forms, they will employ these forms differently in comunication situations. In other







73

words, different groups of children might be said to show differences in language use--that is, in their knowledge of the rules for what is appropriate communication in a given situation. Language use is derived from one's communication experience, and this experience is, of course, shaped by the environment (Naremore, 1976, p. 23).

For the majority of children, family communication patterns

coincide with those of larger units of society, so what the child learns at home is reinforced by contacts outside the home, especially at school.

For other children, communication roles and norms and
various aspects of that code learned in the family settings
are different from those of the larger culture. These
latter children are likely to encounter communication difficulties in school. (Brown, Ecroyd, Hopper & Naremore, 1976,
p. 155)

Although it is apparent that the home environment in which a child is reared is relevant to emerging speech, language and comnunication skills, it has been difficult to find a valid system that allows for rigorous identification of discrete home subcultures in order to systematically examine their influence on communication development (Adler, 1979). The majority of studies in this area have merely related socioeconomic status (SES) to language and communication development.

In 1966, Cazden reviewed a series of studies that had, over the years, counted everything from vocabulary size to transformations and concluded that "on all measures, in all studies, children of upper socioeconomic status, however defined, are more advanced than the lower socioeconomic children" (p. 191). The literature on language use is not as clear-cut. Whereas some investigators still attribute the communicative advantage to children from higher socioeconomic levels, others do not. Children from lower socioeconomic levels are capable of using









language for the same purposes as children from higher socioeconomic levels, but they do so less frequently, in part because they misinterpret the communicative demands associated with the situation.

McCarthy (1930) analyzed the language use of preschool children as it occurred in samples of running conversation using a modified verslon of Piaget's (1923) functional classification scheme. She found that adapted information and questions occurred with greater frequency at all ages among children of higher socioeconomic levels. These differences in language use persisted even when mental age was held constant.

Davis (1937) also used the Piaget-McCarthy functional classification framework to analyze the language use of older children, ages 5'/2 to 91/2. Overall, the children from the higher socioeconomic levels made more spontaneous remarks than the children from the lower socoeconomic levels, but the discrepancy decreased with age. In addition, the higher socioeconomic status children asked more questions than the lower socioeconomic status children, with the difference disappearing by 9'/2 years. This finding was comparable to that observed in the younger children studied by McCarthy (1930). Two other trends emerged from the data on social class differences. First, the naming category appeared to be used with greater frequency by the children from the lower socioeconomic levels. Second, the percentage of answers was higher for the lower socioeconomic groups although the differences decreased with age. Davis (1937) attributed these results to differences in the socialization processes associated with the two strata of society.









Schachter et al. (1974) also provided data on the qualitative differences in everyday speech usage as they occurred naturally in the preschool setting for advantaged and disadvantaged black and white children. The results indicated significant sociolinguistic differences for only two scores, modulations and asserting desires to adults. Both advantaged groups (black and white) showed higher modulation scores than both disadvantaged black groups (higher and lower IQ). The authors suggested that this finding was consistent with Hess's (1969) research formulated in the framework of Bernstein's (1962, 1965) sociolinguistic theories.

Hess (1969) found that lower-class black mothers were more apt to use imperative-normative control strategies with their children, while middle-class blacks use more cognitive-rational and personalsubjective control strategies. Imperative-normative strategies involved orders, accompanied by appeals to existing norms, e.g., "Do it because I told you to do it." Cognitive-rational strategies involved appeals to reason, e.g., "Do it because it may fall," and personal-subjective strategies involved appeals to reason of an emotional nature, e.g., "Do it so you won't hurt her feelings." The latter two strategies seem to require much greater use of Modulation--explanation, justification, etc. The present data suggest that the social class differences in the use of verbal modulations are evident as early as the preschool years.

In asserting desires to adults, both advantaged groups scored consistently higher than the lower-IQ disadvantaged blacks but showed no significant difference relative to the higher-IQ disadvantaged group. These results appeared consistent with White's (1972) findings on the









competence of young children. His less competent preschool children showed less instrumental dependence on adults in their overall social behavior just as the disadvantaged children with lower IQs showed less instrumental dependence in their social speech. These data suggest that the preschool children who are least likely to turn to adults for help in fulfilling their desires are most vulnerable to school problems.

In a longitudinal study of language function, Tough (1977) compared the language used by advantaged and disadvantaged children at the ages of 3, 5/ and 7%'v. Analysis of linguistic form revealed that the 3-year-old children in the advantaged group were producing longer and more complex utterances with greater frequency on all features measured than their disadvantaged peers. These differences were demonstrated at a high level of significance and supported other research findings in showing that there are differences in the language forms used by children from lower and higher socioeconomic groups. Tough's (1977) data indicated that these differences are already apparent in the language of children by the age of 3 and are maintained in their language at 5'/ and 7 years. One of the most significant findings of the linguistic analyses was 'that the disadvantaged children had greater resources of language than their typical performances revealed" (Tough, 1977, p. 169). Although the mean scores for the disadvantaged children tended to be lower on all measures, the range of scores was not necessarily similarly restricted. All disadvantaged children produced long utterances, elaborated noun and verb phrases and a greater number of nouns than pronouns in certain communication contexts. The generally lower scores produced by the disadvantaged children could not be explained by a failure to develop and use complex language, but rather by a









difference in general orientation to language use which also was reflected in language form. The disadvantaged children lacked the appropriate expectations for certain communication situations and thus were not readily disposed "to search for or recognize information that they held as appropriate for answering the questions put to them" (Tough, 1977, p. 167).

A functional analysis of the language used by advantaged and disadvantaged children showed that by the age of 3, differences existed in the kinds of meaning that they were imposing on their experiences through language, although there was little difference in total output of language. The disadvantaged children tended to limit language use to monitoring the ongoing situation and to maintaining status in relation to others. The advantaged children used language more frequently to: (a) recall and give detail to past experiences; (b) reason about past and present experiences; (c) anticipate future events and predict outcomes; (d) survey alternative courses of action; (3) project into the feelings and experiences of others; and (f) create imagined scenes through language.

These differences, first apparent in the unstructured conversation between 3-year-olds, remained unresolved in the talk of 5/2- and 7/z-year-olds in response to structured interviews. In all situations, the disadvantaged groups used language reflective of less complex levels of thinking than that of their advantaged peers. Through their use of language, the disadvantaged groups demonstrated that they were less aware of alternative interpretations of situations and were less inclined to project beyond the immediate requirements of a task. Tough (1977) concluded that in all probability, the limited responses of the









disadvantaged children originated from attitudes which prevented them from recognizing as appropriate or relevant their extended thinking or expression of awareness. She offered as evidence for her conclusions the fact that frequently when probed for additional information, the disadvantaged group gave more complete responses which resembled the kinds of responses produced spontaneously by the advantaged groups. This finding seems to support Bernstein's (1971) observations on the relationship of early social experiences and language use. He noted how different social contexts, which provide the early experiences through which children develop language, result in different orientations toward social relationships and the use of language.

This difference in orientation toward language use has been observed in infants as young as IS months. Howe (in Bruner, 1974) has found that middle-class mothers assume an instructive role toward their infants, both by responding more to their infants' efforts to vocalize by speaking in return and by attempting more often to initiate exchanges. The working-class mothers in her study were more often laissez-faire in their approach.

In a partial replication of Tough's (1977) approach, Wells (1978) could find no clear-cut relationship between language use and either social class or educational success after one year of schooling. Wells (1978) attributed the discrepancies between the results to

(a) the distortion introduced into Tough's study by the
comparison of polarized social-class groups, and (b) the noninteractive conception of communication that underlies her
analysis of language use.(p. 9)

In Wells' (1978) study, children represented four classes of family backgrounds which spanned the full spectrum of social class. When the







79

full spectrum of family background was considered, differences in language use were far less clear-cut and there were few simple linear trends over the four social classes. In fact, in some categories of language use, the trend was in the opposite direction from that predicted by Tough. Combined scores on two tests of reading at the end of the first year of school were used to determine the extent of the relationship between language use, family background and a measure of reading achievement. Significant correlations were found between: language use and reading; language use and social class; or reading and social class. When the contribution of social class to the correlation between language use and reading was partialed out, the correlation was reduced to .19, which is not significant. On the other hand, even when the effect of language use was partialed out, the correlation between social class and reading was .47, which is still significant (p < .05). These results suggest that

- ' , whilst there is a significant relationship between
social class and both language use and educational success
(as measured by reading attainment after one year) it is not the differential use of language for the purposes identified
by Tough that is the main mediator between home background
and school success. (p. 17)

Two studies of a more experimental nature were undertaken by Edwards (1977) and Bruck and Tucker (1974) to examine the influence of social class on language use. Edwards (1977) investigated the effect of context on the linguistic form and function of ten advantaged and ten disadvantaged 11-year-old children. An interview technique was used to stimulate conversation which could be classified as narrative or explanation and the resulting language samples were analyzed using linguistic indices of form and function. On traditional measures of language form,







8o

no significant group differences emerged. There were significant differences on measures directly related to language function or the "planning principles" underlying Bernstein's (1962, 1964) restricted and elaborated codes, e.g., range of adjectival modification and explicitness of pronominal reference.

In Bernstein's (1962, 1964) early code definitions, a low proportion of exophoric references and a high ratio of adjectives to nouns were both reflective of elaborated speech and contributed to explicit unambiguous communication. In the present investigation, disadvantaged children used a narrower range of adjectives and extended modifications and had a higher proportion of exophoric references, suggesting the use of an informal, personalized mode of communication.

Task differences were significant for several of the measures in response to the narrative and explanatory modes of communication. In explanations, children used more subordination than in narratives. In addition, explanations elicited a lower proportion of simple tenses, a higher proportion of modals, and a restricted ratio of nouns to verbs, indicative of more abstract communication. A higher ratio of nouns to pronouns was also observed during explanations and could be attributed to the restriction of referents in this part of the interview.

In the present study, disadvantaged children did not exhibit the rigidity in speech suggested by some previous research. The main generalizations arising out of the data were that task differences accounted for greater and more frequent differences than social class differences and that the variation in scores did not suggest a persistent orientation to either the restricted or elaborated code. These









results offer some support for Cazden's (1970) plea to consider the important influence of contextual factors on communication.

Bruck and Tucker (1974) investigated social class differences in the acquisition of school language by twenty middle- and twenty lower-class kindergarten children. A pretest-posttest design was employed to measure changes in their grammatical and communication skills during the first year of formal schooling, including: imitation of syntactic structures, grammatical comprehension, production of whquestions, grammatical sentence completion, story-telling, description of abstract designs, vocabulary naming and vocabulary classification. These tasks were analyzed according to 26 linguistic and communication variables and the data were submitted to a factor analysis.

The results of the factor analysis suggested that "communication and grammatical abilities represent statistically independent skills in kindergarten children" (Bruck & Tucker, 1974, p. 216). Two factors emerged which contributed to grammatical abilities: comprehension of classroom English and productive knowledge of classroom grammar. Four factors emerged which contributed to communication abilities and coincided with Hymes' (1971) definitions of communication competence: speech output (knowing when to speak), communication of relevant content (knowing what to say), ambiguous and elaborated speech, and egocentric information (knowing how to say it).

Lower-class children did not perform consistently more poorly than middle-class children on all measures, but they did experience difficulty in three main areas.

First, lower-class children have the same ability as
middle-class children to comprehend grammatical structures, but have more difficulty producing them. . . . Second, the









lower-class children's speech contained fewer explicit features than the middle-class children's although it contained as many implicit features .... Third, lower-class children have particular difficulty evaluating the communicative demands of the
classroom. (Bruck & Tucker, 1974, pp. 216-217)

Bruck and Tucker (1974) concluded that,

Failure to give the expected information on demand may
contribute initially to the lower-class child's failure in school. . . . The child knows the relevant information hut
does not communicate it. This may occur because he views
the classroom situation as threatening and says nothing . . .
or because he assumes that the teacher already knows the answer and it would be redundant for him to tell her what
she already knows. (p. 217)

Two general conclusions can be drawn about changes in language abilities during the first year of school. Improvement in gramrnatical abilities was greater than improvement in communication abilities. Of particular importance, the lower-class children improved more rapidly over the year than the middle-class children. This general Improvement was not found for the communication abilities. The lower-class children "did not seem to be catching up with their middle-class peers on tests of communication abilities to the same extent that they did on tests of grammatical abilities,' (Bruck & Tucker, 1974, p. 218). Any gains in communication effectiveness appeared to he related to increased use of features which stress elaboration and explicitness and not to decreased use of ambiguous or egocentric features. In summary, Bruck and Tucker (1974) emphasized that "children need specific help in learning how to use speech more effectively in the classroom. They must learn what the communicative demands of the situation are and how to meet these demands" (Bruck & Tucker, 1974, p. 218).

Unlike developmental studies of language use in language-normal and language-impaired speakers, most investigations of the culturally







83

diverse focus on the school-aged child and the adequacy of his habitual mode of language use for success in the classroom. The results of these various investigations were not conclusive.


Measurement of Language Use


A review of existing measures of language use in children reveals only one standardized tool and a variety of nonstandardized approaches which are employed to evaluate isolated aspects of this broad area. One critical problem in the development of measurement strategies of language use has been the establishment of what Bronfenbrenner (in Cazden, Bond, Epstein, Matz & Savingnon, 1977) calls "ecological validity." Bronfenbrenner argues that much of contemporary developmental psychology is the science of the strange behavior of children in strange situations with strange adults for the briefest possible periods of time" (p. 84). Developers of oral language testing strategies, in particular, have been guilty of violations of ecological validity.

Ideally, a comprehensive assessment of language use should be designed in accordance with sociolinguistic theory to sample, in an ecologically valid way, the function and context of communication situations. To a great extent, then, the development of ecologically valid communication testing situations for children depends upon the skillful simulation of real life encounters which maintain the motivational and interactional richness of familiar academic and social communication events.










Standardized Measurement Strategims

At the present time, there is no comprehensive battery which assesses the language use of children in everyday academic and social situations. The Preschool Language Assessment Instrument (PLAI) (Blank, Rose & Berlin, 1978) is the only available test which attempts to measure functional aspects of natural language not accounted for in traditional language evaluations. Designed to evaluate young children's skills in coping with the language demands of the teaching situation, the PLAI is based on Moffet's (1968) model of classroom discourse which includes three major components: the speaker-listener dyad, the topic, and the level of discussion. The level of discussion places demands on the child that require varying levels of abstraction ordered along a continuum of perceptual-language distance. The continuum has been divided into four main levels including: (a) matching perception (e.g., "What is this called?"); (b) selective analysis of perception (e.g., "What is this part for?"); (c) reordering perception (e.g., "Look at these. How are they the same?"); (d) reasoning about perception (e.g., "If the lady wants to carry all these things into the next room at the same time, tell me what she could use."). The main problem with the PLAI is the category system on which it is based. The development of this tool seems to have been largely influenced by a desire to distinguish levels of cognitive ability rather than by strict adherence to any particular theory of language or communication development. As such, the instrument distinguishes between cognitive demands that require children to apply language to salient perceptions at a low level of abstraction, and demands that require children to restructure







85

those perceptions at a higher level of abstraction. This focus on cognition rather than communication forces the inclusion of certain items in the battery that are clearly not of an interactive nature, e.g., catching pictures, following directions and sentence completion.


Nonstandardized Measurement Strategies

The remainder of the strategies to be reviewed are all nonstandardized measurement techniques designed to answer specific experimental questions about the nature of communication development. These strategies can be grouped into one of three types: interview, roleplaying and natural language sampling.

Several investigators have used the interview type format of the PLAI with a wide variety of stimulus materials. Because looking at pictures and books is a commvon activity of pre- and school-aged children, Tough (1976) developed a picture description task for the appraisal of seven broad categories of language use: (a) self-naintaining,

(h) directing, (c) reporting, (d) logical reasoning, (6) predicting,

(f) projecting and (g) imagining.

Ritti (1978) and Fogel (1976) both employed an interview

strategy in which "real life" comunication situations were presented through pictures and verbal or written probes. In Ritti's (1978) study, children were presented story situations to which they were expected to write something which they would say in that situation and to indicate which of ten possible responses provided by the examiner they would or would not make in such a situation. Responses were developed using five social speech functions from the work of Soskin and John (1963) including information, directives, inductives, evaluations and expressires.









Fogel (1976) designed pictorial and verbal stimuli to elicit comments from children concerning their own or a significant other's communication behavior (i.e., mother, teacher or peer). He selected three relevant attributes of the communication situation around which the stimuli were designed: (a) the function of communication (social

or emotional), (b) the significant other within the situation (mother, teacher or peer), and (c) the perspective taken within the communication situation (self or other). For example, one stimulus item, a picture of a teacher who was trying to explain why a child could not have a book to read, was accompanied by the verbal probe "What would she say?"

Ricillo (in Larson, Backlund, Redmond Barbour, 1978) also used an interview format to examine children's ability to use seven functions of communication: contactive, conversative, descriptive, directive, explanatory, narrative and persuasive. In her study, a series of preestablished probes were presented to children unaccompanied by other stimulus material.

Lucas (1980) developed the Behavioral Inventory of Speech Act Performance (BISAP), a criterion-referenced tool, to evaluate the speech act production of children, ages 3 to 5. BISAP assesses the production of eight commonly used speech acts including requests for objects, requests for action, assertions, denials, statements of information, requests for information, calling or summons, and rule orders in a familiar context of the examiner's design. Lucas (1980) has used art activities, preschool academic tasks and infant-to-caregiver tasks, depending upon the background experiences of the children. Throughout the activity, the examiner presents a series of probes, comments and questions to elicit the desired speech acts. For example, to get the









children to make assertions, the examiner asks them to tell something about a picture which they have previously drawn.

A number of investigators have used role playing situations to test a variety of communication skills. Grim (in Clark & Clark, 1977) and Bates (1976a) used puppets to elicit various direct and indirect speech acts from children. Grimm (in Clark & Clark, 1977) encouraged children to convey directives--asking, ordering, forbidding and permitting--and a commissive--promising to Felix, a large toy cat for whom the experimenter acted as speaker. To get a child to ask for something, Grimm (in Clark & Clark, 1977) used the following story:

You are at a playground with Felix. He is sitting on
the swing and you are sitting on the slide. Now you'd like
Felix to let you swing too. What do you say to Felix?
(p. 365)

After each attempt to get Felix to comply with the directives, Felix refused with increasing emphasis until the child produced four versions of the relevant speech act.

Using a similar strategy, Bates (1976a) encouraged Italian children to use polite forms in asking Signora Rossi, an elderly, greyhaired puppet, for several pieces of candy. After the first request, children were rewarded with a caramel and encouraged to ask even more politely for another piece of candy.

Brenneis and Lein (1977) explored the structures and strategies of children's arguments by having pairs of children perform a variety of role plays. In two of the role plays, there was a concrete issue to be determined, i.e., who is to end up with a ball or pencil. In the other two role plays, there were more abstract issues involved-who is the strongest and who is the smartest. The authors concluded









that although certain aspects of role plays are not representative of spontaneous verbal exchanges, the techniques used in role plays also occur in everyday interactions.

Some investigators (Bloom & Lahey, 1978; Dore, 1977; Epstein et al., 1976; Hanes, 1978; Schachter et al., 1974) maintain a purist approach to language and communication sampling by collecting data through naturalistic observation. While this strategy represents the ultimate degree of ecological validity, it is not always the most effective, efficient, or practical strategy to employ. Limber (1976) raises an Important issue about using production data alone to infer the extent of linguistic knowledge in a child. Because speech is a joint function of linguistic competence, various performance factors and communicative competence, all of these together determine the final product. If a certain structure or function of communication is not observed in the sample, how can we decide whether to assume the child has not yet acquired that element or simply has not used it. Limber (1976) concludes that there are two methodological morals to be drawn from the investigation.

The first is that the spontaneous speech of children
provides at best a minimum estimate of their linguistic competence .... The second moral is that any analysis of the spontaneous speech of children should be compared
with a control sample of spontaneous speech taken from fluent
speakers in a relatively comparable situation. ... This
offers some protection against the double standard of attributing gaps in children's performance to some developmental
deficit but tacitly assumes alternative explanations when these same gaps occur in the speech of a presumed, fluent,
mature individual. (pp. 316-317)

Other, more practically oriented criticisms of natural communication sampling have been formulated by Hanes (1978). In particular,







89

Hanes (1978) remarked about the considerable time and expense required for testing, transcribing, coding and data analysis.


The Literature in Retrospect


With the recent shift in perspective from psycholinguistic to sociolinguistic emphases in the study of communication development, a number of investigators, fro a wide variety of disciplines, have begun to make significant contributions to the body of knowledge about how children use language in everyday situations. As with most literature on communications development, the initial focus of these studies has been the young child, from 0 to 3. Little effort has been directed toward an evaluation of school-aged children's skills in this area, even though it has long been suggested that a significant link may exist between skillful language use and success in school. Children whose predominant mode of language use conflicts with that of the school are in serious trouble. The likelihood that they will achieve their maximum potential in the classroom is doubtful, for language serves not only as an important subject of inst ruction, but also as the primary channel through which that instruction is carried out.

It is clear that some children bring many of their disadvantages from home .... But it is just as clear that the
school, itself, through its failures to recognize the complexity
of the language problem, creates many more disadvantaged children.(Spolsky, 1971, p. 15)

In order to circumvent some of these avoidable classroom failures, measurement strategies appropriate for assessment of the various uses to which language is put in school must be developed so that children who may be functioning at less than adequate levels can be identified and helped. Children who are language-impaired relative to the









linguistic aspects of content and form and children who are languagedifferent with regard to these same components may be among those whose use of language does not serve them adequately in the classroom.

It is apparent that a comprehensive system for the measurement of language in the classroom is not currently available, although a number of guidelines exist for the development of such a system. This system should reflect sociolinguistic trends and be capable of identifying children's mastery of a set of communication strategies appropriate for a wide variety of academic situations. It should be designed specifically to deal with the special features of communication in the classroom, above all incorporating some means of differentiation within the large class of cognitive communication strategies. Because a number of theorists have described the "speech act" or 'conmunication act" as the pivotal unit of language in communication, such acts would appear to be a logical organizing principle for the system just prescribed. Austin (1962) and Searle (1969) note that common, recurring routines such as commanding, requesting, promising and arguing are the basic units of human interaction. 'The speech act approach examines human purposes as they occur naturally when people talk to each other" (Wood, 1976).

In the present investigation, the Functional Inventory of Cognitive Communication Strategies was designed to identify the wide variety of language used and strategies reflective of the young child's growing ability to use the language system to convey meaning. The framework for the classification of the cognitive uses of language, developed by Tough (1976, 1977) served as the basis for the creation of the stimulus items on FICCS. In Tough's (1976) system, the basic




Full Text

PAGE 1

THE INFLUENCE OF LANGUAGE ABILITY, AGE AND SOCIOECONOMIC STATUS ON LANGUAGE USE by RITA JANE LIEBERMAN A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1981

PAGE 2

In Loving Memory of My Father, Edward I. Plotle

PAGE 3

ACKNOWLEDGEMENTS During my tenure as a graduate student, many individuals contributed to my personal and professional growth. To Dr. Thomas Abbott, I am especially grateful for his encouragement and guidance throughout my graduate career and his willingness to assume the "long distance" chairmanship of my supervisory committee in the final stages of my research. To the remaining members of my committee, Drs. Anthony Clark, Jayne Harder and Leonard LaPointe, I wish to express my sincere thanks for their suggestions and helpful comments during the initial stages of this project and their forebearance and understanding during the hectic final days. Dr. Edward Hutchinson, who introduced me to the important study of language use, who served as the original director of my doctoral committee and who has been my understanding department head since I joined the faculty at Appalachian State University, deserves my heartfelt thanks. He has been my colleague, my friend and my mentor. He has seen me through the dark times that accompany any project which makes one feel tiny and incapable at its inception. He has encouraged me through the gray days of writing and rewriting. With scholarly advice and with a wise blend of sympathy and threats, he has helped me inestimably to bring this project to its conclusion. 1 would also like to extend my appreciation to Deanna Bowman for her excellent statistical assistance. She was always available to help and her expertise was invaluable. Valerie Buice, Geneva Henson and Susan Payne unselfishly gave of their time to type the original i i i

PAGE 4

manuscript and Dinah Lanning provided outstanding editorial assistance while typing the final draft. Thanks must also be extended to the faculty, staff and children of the schools which participated in this project. They made my job a pleasant and enjoyable one. Finally, my love and endless gratitude to my husband, Chuck, who never stopped believing in me; to my parents and my husband's parents whose long and patient vigil has at last been rewarded with "nachas" from heaven; to my son, Justin, whose unconditional hugs and kisses never diminished even in the face of a crabby mommy; to my dear friends, Ann Michael, whose kind and enthusiastic words were a constant source of encouragement on dreary Boone days; Vicky Breedlove, who shared my teaching responsibilities during the summer of '81 to provide me with precious time; Cathy and Wayne Huber, who hand-carried this manuscript from the Trailways bus station to the University of Florida editorial office on several occasions and whose kind hospitality gave me the peace of mind to remain calm during the week of my oral defense; and to all my other friends and family, without whose support this project would not have been possible. i V

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iil ABSTRACT ix Chapter 1. INTRODUCTION 1 Significance of the Problem 1 Statement of the Problem 6 Delimitations 7 Limitations 8 Assumptions 8 Hypotheses 9 Hypothesis 1 9 Hypothesis 2 11 Hypothesis 3 13 Hypothesis k 15 Hypothesis 5 17 Hypothesis 6 18 Hypothesis 7 19 2. REVIEW OF RELATED LITERATURE 20 The Interdisciplinary Nature of Recent Research into Language Use 20 Expanding the Notion of Competence 21 Origins of the Study of Language Use 23 V

PAGE 6

Chapter Page The Domain of Language Use 23 Functions of Language 26 Speech Acts Theory 37 Development of Language Use in Language-Normal Children 53 Development of Languase Use in LanguageI mpai red Children 66 Primary Forms 66 Conventional Forms 68 Language Use in the Culturally Diverse 72 Measurement of Language Use 83 Standardized Measurement Strategies 84 Nonstandardized Measurement Strategies 85 The Literature in Retrospect 89 3. METHODS AND PROCEDURES 92 Subjects 92 LanguageIrapa i red Group (Ll) 92 Language-Normal Groups 93 Materials 102 Classification of the Cognitive Uses of Language 102 The Functional Inventory of Cognitive Communication Strategies (FICCS) 102 Procedures 108 Collection of the Language Sample 108 Transcription and Segmentation of the Language Sample 110 Analysis of the Language Sample Ill

PAGE 7

Chapter Page Rel iabil ity 119 Order of Presentation of All Tasks 120 Session 1 120 Session 2 120 Session 3 121 Session k 121 Selection Measures 121 Independent Measures 122 k. ANALYSIS OF THE RESULTS 128 Group Comparisons 128 Language Used by Language-Normal and LanguageI mpai red Children 128 Language Used by Language-Normal Children of Different Ages and Socioeconomic Levels 15^ Correlation and Reg ression Analyses 175 Language Use and Standardized Measures of Language Content and/or Form 175 Language Use and Nonstandard i zed Measures of Language Form 182 Language Use and Measures of Academic Achievement 186 Reliability 19^* Internal Consistency ]Sk Intra-rater Rel iabi I i ty 198 Inter-rater Reliability 198 Summary of Findings 199 5, SUMMARY AND CONCLUSIONS 206 vii

PAGE 8

Chapter Page Summary 206 Discussion 209 Conclusions and Suggestions for Future Research .... 217 APPEMDIXES A. FAMILY SIZE AND INCOME SCALES FOR FREE MEALS AND REDUCED-PRICE MEALS 221 B. TOUGH'S FRAMEWORK FOR THE CLASSIFICATION OF LANGUAGE USE 223 C. LANGUAGE SAMPLE ELI CITATION PROTOCOL ALA TOUGH 231 D. STIMULUS MATERIALS FOR LANGUAGE SAMPLE ELICITATION ... 2k0 E. PROTOCOL FOR TRANSCRIPTION AND SEGMENTATION OF LANGUAGE SAMPLES 248 F. RULES FOR COUNTING MORPHEMES 254 G. RULES FOR COUNTING WORDS 257 H. RAW DATA ON STANDARDIZED LINGUISTIC MEASURES AND THE OVERALL SCORE ON THE FUNCTIONAL INVENTORY OF COGNITIVE COMMUNICATION STRATEGIES ACHIEVED BY ALL SUBJECTS 258 I. RAW DATA ON NONSTANDARDI ZED LINGUISTIC MEASURES OBTAINED IN RESPONSE TO THE FUNCTIONAL INVENTORY OF COGNITIVE COMMUNICATION STRATEGIES (FICCS) BY ALL SUBJECTS 26l J. RAW DATA ON MEASURES OF ACADEMIC ACHIEVEMENT FOR ALL SUBJECTS 264 BIBLIOGRAPHY 267 BIOGRAPHICAL SKETCH 279 vi i i

PAGE 9

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE INFLUENCE OF LANGUAGE ABILITY, AGE AND SOCIOECONOMIC STATUS ON LANGUAGE USE By Rita Jane Lieberman August 1981 Chairman: Thomas B. Abbott Major Department: Speech Language is the process by which children succeed or fail in school. Because language is an important subject of instruction as well as the process by which that instruction is achieved, it is imperative that strategies of language use necessary for school success be identified and measured. The purpose of this study was (l) to determine the influence of language ability, age and socioeconomic status on children's language use; and (2) to examine the relationship and predictive accuracy between measures of linguistic performance, academic achievement and language use. Language use was evaluated with the Functional Inventory of Cognitive Communication Strategies (FICCS), a structured interview which elicited four language use strategies: Reporting, Reasoning, Predicting and Projecting. The influence of language ability on language use was examined by comparing languageimpai red children to two groups of languagenormal children, one matched for age and one matched for utterance i X

PAGE 10

length. Performance of the language-normal group matched for age was significantly superior to that of the languageimpai red group. No significant differences were observed between the language-normal group matched for utterance length and the language-impaired group. The 1 anguageimpa i red group achieved significantly lower overall scores than their language-normal peers but higher scores than their younger, normal counterparts matched for utterance length, suggesting that the communicative function of the impaired children was better than their linguistic skills would imply. The influence of age and socioeconomic status on language use was evaluated using a factorial design, with two age levels (6 years and 7 years) and two social class levels (lower and higher). The languagenormal 6-year-olds achieved significantly higher scores than the language-normal 7"year-olds on Projecting strategies. Correlation analyses between performance on FICCS and measures of linguistic ability indicated a strong relationship between FICCS and nonstandard i zed measures of language ability but not for standardized. These findings suggest that spontaneous language sampling, through its preservation of the interactive nature of communication, provides a more powerful correlate of language use than acontextual standardized tools. The relationship between FICCS and measures of academic achievement was modest, indicating that language use and other factors contribute to success in the classroom. X

PAGE 11

CHAPTER 1 INTRODUCTION Significance of the Problem Communication is the process by which children succeed or fail in school. The modern classroom, more than ever before, provides a laboratory for the development and practice of communication skills whose application extends far beyond the confines of the school environment. Present-day education techniques have turned the classroom into a microcosm of the world where all aspects of human communication can and do occur. The student serves as encoder, processor and decoder of the cognitive and social content of a variety of messages. Clearly, communication underlies the major portion of classroom learning. Not only is it an important subject of instruction, but it is also the process by which that instruction is achieved. Because communication is central to both the means and ends of education, a failure in classroom communications will have the most serious consequences. The child who does not succeed in school may be one whose use of communication in the academic setting is restricted in some way and whose communication efforts may reflect an impoverishment of vocabulary or syntactic options, or, skills which are present but either inappropriate to or unavailable for a given classroom situation. Any of these problems may seriously alter a child's participation in the classroom. Traditionally, those disciplines concerned with the study of child 1

PAGE 12

2 language and learning have focused on the adverse effects of 1 i ngui st i c deficits on classroom participation, while the broader, and perhaps more elusive notion of communicative competence has been largely ignored. Joan Tough (1976), a British educator who has done considerable research on communication skills in early childhood, acknowledges the difference between linguistic and communicative competence. In her book, L i sten i ng to Chi Idren Talking , she distinguishes between children whose talk evidences inappropriate phonological, syntactic and semantic patterns and children whose talk is limited to a restricted set of uses. The development of evaluative measures of communication problems in children has been primarily confined to those which measure specific linguistic aspects: phonology (Fisher S Logemann, 1971; Goldman & Fristoe, 1969; McDonald, 1968; Weiner, 1978), syntax (Carrow, 197^; Foster, Giddan & Stark, 1969; Lee, 1969), and semantics (Boehm, 1971; Dunn, 1964; MacDonald & Nickols, 197^). The emphasis on the description and evaluation of language form and content in children is reflected in current assessment strategies and much of the language research literature carried out during the 1960s and early 1970s. Influenced largely by Chomsky's (1957, 1965) theory of transformational grammar and Fillmore's (1968) notions about case grammar, investigators performed linguistic analyses or designed psychol inguistical ly motivated research paradigms to discover new information about the development of form (Bloom, 1970; Brown, 1973; Greenfield & Smith, 1973; Menyuk, 1968) and content (Bloom, Lightblown 6 Hood, 1975; Clark, 1973; Leonard, 1976; Nelson, 1973, 197^). This early influence of psychol i ngu i st i cs on the study of child language focused our attention on the language which children use rather

PAGE 13

than on children as language users. Descriptions of the linguistic competence of children were abundant, but the broader notion of communicative competence--how children use language in socially appropriate ways--was largely overlooked. Recent research trends in the study of child language demonstrate that we are on the threshold of moving from an interest in the description of isolated lexicogrammati cal forms to an interest in the description of communicative function in context. The contextual i St approach emphasizes the importance of the ability to use language to convey a variety of intentions and meanings dependent upon the context and social setting of the communication. This shift in perspective represents a major accomplishment in a movement towards relevance and functionality in communication measure' ment and management. Although psychol i ngui s t i c theory accounts for an important aspect of what the child comes to l
PAGE 14

k interaction) that we can account for the development of order which underlies human communication behavior (Hopper, 197'). Because of the recent shift in focus to a sociol inguistic theory of language learning and language behavior, applied communicologists-speech, language and hearing clinicians, special educators, early childhood specialists, language arts teachers, foreign language teachers, and bi 1 ingual/bicul tural consul tants--have begun to develop informal strategies for the measurement and management of communicative competence. Those strategies can be grouped into one of three types: (a) interviewing [Fogel, 1976; Ricillo (in Larson, Backlund, Redmond S Barbour, 1978); Ritti, 1978; Tough, 1976]; (b) roleplaying [Bates, 1976a; Brenneis & Lein (in Ervin-Tripp & Mi tchel 1 -Kernan , 1977)]; and (c) natural language sampling (Bloom, 1978; Dore, 1977; Epstein, Schwartz, Meece, Lambie, Dui
PAGE 15

deficient children and adults. Since 197^, investigations involving functional nx5des of language use have been carried out with the mentally handicapped (Bedrosian & Prutting, 1978; Owens, 1979; Sabsay, 1975), the hearing-impaired (Curtiss, Prutting & Lowell, 1979; Skarakis & Prutting, 1977), the neurological ly-involved (Ulatowska, Macal uso-Haynes & MendelRichardson, 1976; Wilcox S Davis, 1977), the behavioral 1 y-d i sordered (Bartolucci S Albers, 197'*), and the specific linguistically-handicapped (Geller SWollner, 1976; Miller, 1978; Snyder, 1975). Even fewer studies have focused on the general language use of older school-aged children (Fogel, 1976; Tough, 1976, 1977). While additional studies of language use in older children exist in the sociol inguistic literature, their emphases have generally been narrowly confined to limited aspects of language use such as arguments (Brenneis & Lein, 1977), narratives (Kernan, 1977) or directives (Mi tchel 1-Kernan £ Kernan, 1977; Garvey, 1975, Ervin-Tripp, 1975); thus, their implications, though relevant for child literature, remain fragmentary (Rees, 1978). From sociol inguistics also comes an abundance of literature on the effects of social class and culture on language use. Because of differences in design, sampling protocols, language tasks and analysis, the results of these studies present conflicting evidence relative to the influence of social class on language use. Several of the studies demonstrate that children of disparate sociocul tural backgrounds use language differently (Blank, Rose & Berlin, 1978; Bruck & Tucker, 197^; Tonkovich &Adler, 1978; Tough, 1977; Williams 6 Naremore, 1969a). Others reveal less clear-cut differences in language use (Edwards, 1977; Schachter et al., 197^; Wells, 1978). This perusal of the available literature on language use uncovers several avenues of investigation

PAGE 16

which remain to be charted, particularly with regard to the older 1 anguageimpa i red child who lives in a less than advantageous social s i tuat ion . Statement of the Problem The problem of this investigation is (a) to assess differences in language use between groups of children who differ in age, social class, and linguistic ability; and (b) to evaluate the relationship and predictive accuracy between measures of linguistic and academic competence and the Functional Inventory of Cognitive Communication Strategies (FICCS), a researcher-developed instrument, which assesses language use or communicative competence. More specifically, answers to the following questions will be sought: 1. Is there a difference in language use between languageimpaired children and language-normal children, matched on the basis of chronological age? 2. Is there a difference in language use between languageimpaired children and language-normal children, matched for utterance 1 ength? 3. Is there a difference in language use between languagenormal children at two different age levels, 6 years (plus or minus 3 months) and 7 years (plus or minus 3 months)? 4. Is there a difference in language use between languagenormal children of two different socioeconomic levels, lower and upper? 5. What is the relationship of language use to performance on standardized measures of language form and/or content?

PAGE 17

6. What is the relationship of language use to performance on nonstandard i zed measures of language form? 7. What is the relationship of language use to performance on measures of academic achievement? Del i mi tat ions 1. The study was confined to four groups of children, one languageimpai red and three language-normal. 2. Subjects were selected from the preki ndergarten, kindergarten and first grade populations of the Wilkes and Avery County school systems in North Carolina according to the following criteria: a. They were enrolled in a school program for a period of not less than four weeks prior to inclusion in the study. b. They demonstrated average intellectual functioning (Age Deviation Score = 89 to 110) on the Columbia Mental Maturity Sea 1 e (Burgemei ster , Blum & Lorge, 1972). c. They were judged to be language-impaired ($17th percentile) or language-normal (^70th percentile) on the basis of performance on the Bankson Language Screening Test (Bankson, 1977) and a clinical opinion of a certified speech, language and hearing cl inician. d. They were native speakers of English from monolingual homes who did not exhibit any gross peripheral defects of audition or vision. 3. The total testing time was confined to four sessions, or 2V2 hours for each subject.

PAGE 18

k. Data regarding the criterion variable of language use were confined to that gathered from a researcher-developed instrument, the Functional Inventory of Cognitive Communication Strategies (FICCS), based on Tough's (1976) functional classification scheme. 5. Data regarding the predictor variables of linguistic and academic competence were confined to three standardized measures of language form and/or content, eight nonstandard! zed measures of language form, and four measures of academic achievement. Limi tations 1. To the extent that pretesting sensitizes the subjects to the language sampling procedure, results will not be general izable to unpretested groups. 2. To the extent that the subjects selected are not representative of the languageimpai red or language-normal population at large, results will not be general izable beyond the sample investigated. 3. To the extent that knowledge of subject status affects the objectivity of the researcher's observations and judgments, or causes her to influence the subjects' reactions to the tasks, results may be biased in favor of one group or the other. k. To the extent that the subjects are aware of participation in a research study, results may not be general izable beyond the experimentally accessible population. Assumpt ions The following assumptions were made in this study: 1. That the groups of languageimpai red and language-normal children were matched on the major critical variables affecting language

PAGE 19

9 use: age, socioeconomic status, intellectual ability, and general linguistic level; and that other extraneous variables which may affect language use, such as motivation and personality characteristics, were randomly distributed between the languageimpai red and language-normal groups . 2. That the exposure to intervening variables in the school environment, such as supportive relationships with teachers, was equivalent for al 1 groups. 3. That the researcher-developed inventory of cognitive communication strategies did, in fact, measure ability to use language in response to a structured interview; and that this ability is representative of the way in which children use language in the classroom. k. That the researcher, being a practiced speech, language and hearing clinician, was qualified to administer, score and interpret all testing procedures used in this study. Hypotheses To give direction to the data analysis, the following hypotheses were developed. These hypotheses, stated in the null form, were tested at the .05 level of significance. Hypothesis 1 Ho 1.1 . There is no significant difference in performance on the FICCS overall score between the languageimpai red (Ll) and languagenormal (LNl) groups who are matched for chronological age (CA) . Ho 1.2 . There is no significant difference in performance on the FICCS Reporting subscore between the LI and LNl groups.

PAGE 20

10 Ho 1 . 3 ' There is no significant difference in performance on the FICCS Logical Reasoning subscore between the LI and LNl groups. Ho 1 . A . There is no significant difference in performance on the FICCS Predicting subscore between the LI and LNl groups. Ho 1.5 There is no significant difference in performance on the FICCS Projecting subscore between the LI and LNl groups. Ho 1.6 . There is no significant difference in performance for the 2k individual communication strategies on FICCS between the LI and LNl groups. Ho 1,7. There is no significant difference in performance for 1 ower1 evel strategies overall on FICCS between the LI and LNl groups. Ho 1.8. There is no significant difference in performance for lower1 evel Reporting strategies on FICCS between the LI and LNl groups Ho 1.9. There is no significant difference in performance for lower-level Logical Reasoning strategies on FICCS between the LI and LNl groups. Ho 1.10 . There is no significant difference in performance for lower-level Predicting strategies on FICCS between the LI and LNl groups Ho 1.11 . There is no significant difference in performance for lower-level Projecting strategies on FICCS between the LI and LNl groups Ho 1.12 . There is no significant difference in performance for higher-level strategies overall on FICCS between the LI and LNl groups.

PAGE 21

11 Ho 1.13 . There is no significant difference in performance for higher-level Reporting strategies on FICCS between the LI and LNl groups. Ho 1.14 . There is no significant difference in performance for higher-level Logical Reasoning strategies on FICCS between the LI and LNl groups. Ho 1.15 There is no significant difference in performance for higher-level Predicting strategies on FICCS between the LI and LNl groups. Ho 1.16 . There is no significant difference in performance for higher-level Projecting strategies on FICCS between the LI and LNl groups. Hypothesis 2 Ho 2.1 . There is no significant difference in performance on the FICCS overall score between the language-impaired (Ll) and languagenormal (LN2) groups who are matched for utterance length. Ho 2.2 . There is no significant difference in performance on the Reporting subscore between the LI and LN2 groups. Ho 2.3 . There is no significant difference in performance on the Logical Reasoning subscore between the LI and LN2 groups. Ho 2.k . There is no significant difference in performance on the FICCS Predicting subscore between the LI and LN2 groups.

PAGE 22

12 Ho 2 . 5 • There is no significant difference in performance on the Projecting subscore between the LI and LN2 groups. Ho 2.6 . There is no significant difference in performance for the 2k individual coiranunication strategies on FICCS between the LI and LN2 groups. Ho 2.7 . There is no significant difference in performance for lowerlevel strategies overall on FICCS between the LI and LN2 groups. Ho 2.8. There is no significant difference in performance for lowerlevel Reporting strategies on FICCS between the LI and LN2 groups Ho 2.9. There is no significant difference in performance for lower-level Logical Reasoning strategies on FICCS between the LI and LN2 groups. Ho 2.10 . There is no significant difference in performance for lowerlevel Predicting strategies on FICCS between the LI and LN2 groups. Ho 2.11 . There is no significant difference in performance for lower-level Projecting strategies on FICCS between the LI and LN2 groups . Ho 2.12 . There is no significant difference in performance for higher-level strategies overall on FICCS between the LI and LN2 groups. Ho 2.13 . There is no significant difference in performance for higher-level Reporting strategies on FICCS between the LI and LN2 groups.

PAGE 23

13 Ho 2.1^ . There is no significant difference in performance for higher-level Logical Reasoning strategies on FICCS between the LI and LN2 groups. Ho 2 . 1 5 • There is no significant difference in performance for higher-level Predicting strategies on FICCS between the LI and LN2 groups. Ho 2.16 . There is no significant difference in performance for higher-level Projecting strategies on FICCS between the LI and LN2 groups. Hypothesis 3 Ho 3 • 1 . There is no significant difference in performance on the FICCS overall score between the language-normal 6and 7~year-olds. Ho 3 • 2 . There is no significant difference in performance on the FICCS Reporting subscore between the language-normal 6and 7-yearolds. Ho 3. 3 There is no significant difference in performance on the FICCS Logical Reasoning subscore between language-normal 6and 7year-ol ds . Ho 3. A . There is no significant difference in performance on the FICCS Predicting subscore between language-normal 6and 7-year-olds. Ho 3-5 . There is no significant difference in performance on the FICCS Projecting subscore between language-normal 6and 7-year-olds.

PAGE 24

14 Ho 3.6 . There is no significant difference in performance for the 24 individual communication strategies on FICCS between languagenormal 6and 7"year-olds. Ho 3.7 . There is no significant difference in performance on the lower-level strategies overall on FICCS between the language-normal 6and J-year-olds. Ho 3«8 . There is no significant difference in performance on the lower-level Reporting strategies on FICCS between language-normal 6and 7-year-olds. Ho 3 • 9 • There is no significant difference in performance on the lowerlevel Logical Reasoning strategies on FICCS between languagenormal 6and 7-year-olds. Ho 3. 1 0 . There is no significant difference in performance on the lower-level Predicting strategies on FICCS between language-normal 6and 7-year-olds. Ho 3' 1 1 . There is no significant difference in performance on the lower-level Projecting strategies on FICCS between language-normal 6and 7-year-olds. Ho 3.12 . There is no significant difference in performance on the higher-level strategies overall on FICCS between language-normal 6and 7-year-olds.

PAGE 25

15 Ho 3 • 1 3 • There is no significant difference in performance on the higherlevel Reporting strategies on FICCS between language-normal 6and 7"year-olds. Ho 3 • 1 ^ • There is no significant difference in performance on the higher-level Logical Reasoning strategies on FICCS between languagenormal 6and 7~year-olds. Ho 31 5 There is no significant difference in performance on the higher-level Predicting strategies on FICCS between language-normal 6and 7~year-o1ds. Ho 3-16 . There is no significant difference in performance on the higher-level Projecting strategies on FICCS between language-normal 6and 7-year-olds. Hypothesis k Ho 4.1 . There is no significant difference in performance on the FICCS overall score between the lower (LSES) and higher (HSES) socioeconomic status language-normal groups. Ho h.2 . There is no significant difference in performance on the FICCS Reporting sbuscore between the LSES and HSES language-normal groups . Ho k.3 . There is no significant difference in performance on the FICCS Logical Reasoning subscore between the LSES and HSES languagenormal groups.

PAGE 26

16 Ho k.k . There is no significant difference in performance on the FICCS Predicting subscore between the LSES and HSES language-normal groups. Ho A. 5 . There is no significant difference in performance on the FICCS Projecting subscore between the LSES and HSES language-normal groups. Ho k.6 . There Fs no significant difference in performance for the 2k individual communication strategies on FICCS between the LSES and HSES language-normal groups. Ho A . 7 . There is no significant difference in performance on the lower-level strategies overall on FICCS between the LSES and HSES language-normal groups. Ho ^.8 . There is no significant difference in performance on the lower-level Reporting strategies on FICCS between the LSES and HSES language-normal groups. Ho ^.9 . There is no significant difference in performance on the lower-level Logical Reasoning strategies on FICCS between the LSES and HSES language-normal groups. Ho A . 1 0 . There is no significant difference in performance on the lower-level Predicting strategies on FICCS between the LSES and HSES language-normal groups.

PAGE 27

17 Ho A. 1 1 . There is no significant difference in performance on the lower-level Projecting strategies on FICCS between the LSES and HSES language-normal groups. Ho 4.12 . There is no significant difference in performance on the higher-level strategies overall on FICCS between the LSES and HSES language-normal groups. Ho A. 1 3 . There is no significant difference in performance on the higher-level Reporting strategies on FICCS between the LSES and HSES language-normal groups. Ho 4 . 1 A . There is no significant difference in performance on the higher-level Logical Reasoning strategies on FICCS between the LSES and HSES language-normal groups. Ho A . 1 5 . There is no significant difference in performance on the higher-level Predicting strategies on FICCS between the LSES and HSES language-normal groups. Ho 4.16 . There is no significant difference in performance on the higher-level Projecting strategies on FICCS between the LSES and HSES language-normal groups. Hypothesis 5 Ho 5.1 . There is no significant relationship between performance on a measure of language use, FICCS, and the Test of Audi tory Comprehension of Language (TACL) .

PAGE 28

18 Ho 5.2 . There is no significant relationship between performance on FICCS and the Test of Language Development (TOLD). Ho 5 . 3 » There is no significant relationship between performance on FICCS and the Carrow Elicited Language Inventory (CELl). Hypothesis 6 Ho 6.1 . There is no significant relationship between performance on a measure of language use, FICCS, and the total number of communication units (#CU's) used in response to a structured interview. Ho 6.2 . There is no significant relationship between performance on FICCS and the total number of words (TNW) used in response to a structured interview. Ho 6.3 . There is no significant relationship between performance on FICCS and the mean length of utterances in words (MLU-W) used in response to a structured interview. Ho 6.^ . There is no significant relationship between performance on FICCS and the total number of morphemes (TNM) used in response to a structured interview. Ho 6.5 . There is no significant relationship between performance on FICCS and the mean length of utterances in morphemes (MLU-M) used in response to a structured interview. Ho 6.6 . There is no significant relationship between performance on FICCS and the mean length of the five longest utterances (MLU5-M) used in response to a structured interview.

PAGE 29

• > 1 9 Ho 6.7 ' There is no significant relationship between performance on FICCS and the percentage of complete communication units (^CUc) used in response to a structured interview. Ho 6.8 . There is no significant relationship between performance on FICCS and the Developmental Sentence Score (DSS) obtained in response to a structured interview. Hypothesis 7 Ho 7.1 . There is no significant relationship between performance on a measure of language use, FICCS, and a measure of academic achievement, the Test of Basic Experiences language subtest (TBEL) . Ho 7.2 . There is no significant relationship between performance on FICCS and the Test of Basic Experiences mathematics subtest (TBEM) . Ho 7-3 . There is no significant relationship between performance on FICCS and the Test of Basic Experiences social studies subtest (TBESS). Ho 7-^ . There is no significant relationship between performance on FICCS and the Test of Basic Experiences science subtest (TBES).

PAGE 30

CHAPTER 2 REVIEW OF RELATED LITERATURE The Interdisciplinary Nature of Recent Research into Language Use There is no single corpus of research which explains the child's developing ability to use the phonological, syntactic and semantic structures of language to communicate effectively a variety of messages in diverse social situations. Several disciplines, including linguistics, sociol inguistics, developmental psychology, philosophy and speech pathology have been approaching this issue from similar but separate points of view. Evidence for these "separate but nearly equal" vantage points comes from the many terms which have evolved to describe essentially the same set of behaviors: communication competence (Hymes, I97O; Slobin, 1971 ; Wood, 1976), pragmatics (Bates, 1976a, b; Dore, 197'*; Hopper £ Naremore, 1973; Prutting, 1979; Rees , 1978) and language use (Bloom S Lahey, 1978; Dale, 1972; Halliday, 1975). Although these terms are not precise synonyms, the cluster of behaviors referred to by such terms "has intruded into many areas of child study, highlighting the importance of rule-governed ways in which children use code items to meet the demands of communication situations" (Allen & Brown, 1976, p. 153). Each discipline which has undertaken the task of studying language using a contextual approach has made significant contributions concerning the language use of real speakers and listeners. From 20

PAGE 31

21 philosophy has arisen a theory of speech acts which focuses on the communicative intent realized through language. Linguists have developed a specialized branch of study called pragmatics with its own terminology and taxonomy of types. The speai
PAGE 32

22 little about real children's everyday use of language in particular si tuations. Hymes (1971, 1972b) has been critical of the linguist's tendency to emphasize the notion of competence while neglecting the issue of performance or use. He proposed an expansion of the notion of linguistic competence into "competence for use" or "communicative competence"-how children perceive and categorize the social situations of their world and differentiate their ways of speaking accordingly. This broader account of the facts provides an explanation of what the language user knows about "who can say what, in what way, where and when, by what means and to whom" (p. 15). The important issue is that children's verbal behavior is a reflection of their knowledge of language use as well as of grammar and the manifestation of that knowledge will vary according to the communication situation. This influential role of the communication situation in determining verbal behavior is an important contribution which sociol inguistic theory makes to our understanding of language development and language behavior. From a sociol inguistic point of view, there is more to "competence for use" than language ability. Applied communication specialists no longer use linguistic competence as the ultimate goal in communication evaluation and management programs. Davies (1977) comments, . . . there is more to using, learning and teaching a language than knowledge of the linguistic rules indicates. Knowing the grammar is just not enough as a model of how speakers behave. . . .(p. 2)

PAGE 33

23 Origins of the Study of Language Use In 1938, Charles Morris incorporated the notion of language use into a tripartite organization of language. According to this model, a comprehensive description of language required a statement about three principal components: 1. Syntactics-the relation of signs to one another; 2. $emantics-the relation between signs and their referents; 3. Pragmat i cs — the relation between signs and their users. in the late 1950s and early 1960s, scholars in the field of the philosophy of language expanded the notion of language use to include a practical system for classification of the knowledge which underlies socially appropriate speech. The work of two linguistic philosophers in speech acts theory, J. L. Austin (1962) and J, R. Searle (1969) has served as the basis for much of the current research in the area of language use. The Domain of Language Use In their three-dimensional characterization of language. Bloom and Lahey (1978) echo Morris' earlier writing by describing language as consisting "of some aspect of content or meaning that is coded by linguistic form for some purpose or use in a particular context" (p. II). Through the integration of content, form and use, children become competent communicators as they acquire the ability to use alternative forms of a message for achieving the same purpose, according to differences in the situational context. Earlier depictions of child language accounted for the mastery of content and form, but it is to the

PAGE 34

~^^^T " y . 2k contextual ist movement that we owe the additional concern with use. The contextual i sts noted that "there was a structure in the use of language that went beyond the aspect of structure dealt with in grammars. . ." (Hymes, 1972a, p. xxii). Within the broad framework of language use, Bloom and Lahey (1978) recognize two main areas, function and context. Function refers to the goals or intent of a communication, "the reasons why people speak" (Bloom & Lahey, 1978, p. 19). Context encompasses the influence of specific situational parameters, such as the time and place of the communication, in determining the form which the message will take. "Speakers of a language have alternative means for saying the same thing or achieving the same purpose, and which alternative is used depends on the context" (Bloom & Lahey, 1978, p. 10). Traditionally, the functions of language have been represented in linguistic terms and associated with the syntactic structures for the declarative, interrogative, imperative or exclamative utterance. Halliday (1975) and others have described the functions of language "in more social terms involving interaction, regulation and personal control" (Bloom & Lahey, 1978, p. 20). To achieve these communicative functions, speakers choose appropriate verbal and nonverbal behaviors to bring about desired effects in a wide variety of social situations. The verbal and nonverbal choices which speakers make have been called communication strategies or communication acts by some investigators (Wood, 1976) and evolved out of the more restricted notions of "speech acts," "performatives" or "i 1 locutionary force." This category of language use involves the specification of the speaker's goal in producing

PAGE 35

25 an utterance-i t describes intention to make a statement, ask a question or issue a command. The second aspect of language use focuses on the influential factors involved in the selection of alternative means for conveying messages in varied social contexts as well as the communication processes by which these messages are produced. The form of an utterance changes, depending upon the ability of the speaker to adapt the message to the needs of the listener and to the immediate linguistic and nonlinguistic contexts. These elements of the communication process are necessary "to keep meaning flowing in conversation" (Rees, 1978, p. 208). Within the realm of pragmatic theory, these constructs have been labeled presuppositions and conversational postulates. Although there is considerable controversy regarding the nature of presuppositions, in general they involve "information that is not contained in the message itself but must be known and understood if the message is to make sense" (Bates, 1976a, p. 419). Presuppositions are the determining factor by which speakers adapt their messages to the listener's needs--how they decide what to say and what not to say. According to Bates (1976a), the major conversational task for children in a communicative exchange is learning "when not to presuppose, when it is necessary to provide the listener with explicit cues about the information that is being assumed as background for a comment" (p. kkS) . This process has been referred to as thematic structure (Halliday, 1973) or relating new information to old (Clark & Clark, 1977; Rees, 1978). Conversational postulates have been described by Bates (1976a) as the "boy scout code of conversation" (p. ^6). They are a special class of pragmatic presuppositions which "attempt to make explicit the

PAGE 36

rules speakers and listeners must observe if meaning is to be conveyed (Rees, 1978, p. 208). Grice (1976) has reduced these rules to four maxims which contribute to what he calls the "cooperative principle" of conversation. In summary, these maxims are: (a) be informative; (b) be truthful; (c) be relevant; and (k) be clear. These three constructs--performatives, presuppositions and conversational postul ates--"contri bute to both the form and context of messages; to how people do things with words; and to how languages work" (Bloom & Lahey, 1978, p. 21). In 1976, Bates commented on the limited amount of pragmatic research in child language. The greater proportion of this work centered on the acquisition of performatives or communicative intent although, as Rees (1978) noted, "the conduct of conversation is probably the most complex as well as the most impor tant target of the pragmatic approach to the study of language" (p. 208). The present investigation focuses on the study of a single aspect of language use--the communication strategies or goals used by children of differing linguistic ability, age, and socioeconomic level Functions of Language Although philosophers, linguists and psychologists have long been interested in functional taxonomies of language use, their popularity has only recently gained momentum with applied disciplines. Various taxonomies have been described, each according to the author's own unique interests. In 1923, Malinowski developed an ethnographic account of the functions of language to describe the use of language in organizing fishing expeditions in the Trobriand Islands. Since

PAGE 37

27 that time, other functional systems have appeared with different bases, e.g., psychological (DeLaguna, 1927; Buhler, \33^; Soskin & John, 1963), cognitive (Piaget, 1923; Skinner, 1957; Bruner, 1975), literary (Jakobson, I960), ethological (Morris, 1967), linguistic (Halliday, 1970, 1975) and educational (Britton, 1971). On the surface, the categories and terminologies embodied in each of these functional classification schemes differ, but each incorporates a basic distinction between a cognitive (ideational, representational, referential) and a social (interpersonal, express i ve-conat i ve , evocative) function of language. Two well-known accounts of the broad functions which language serves are those offered by DeLaguna (1927) and Bruner (1975). in 1927, Delaguna wrote "men do not speak simply to relieve their feelings or to air their views, but to awaken a response in their fellows and to influence their attitudes and acts" (p. 20). Bruner (1975) maintained that "Language is acquired as an instrument for regulating joint activity and joint attention" (p. 2). The correspondence between these two global characterizations of communicative function is not surprising, since, as McLean and McLean (1978) point out, they describe the same general underlying phenomenon. Both accounts recognize two broad functions of language: (a) the regulation or influencing of a listener's actions; and (b) the regulation or influencing of a listener's attention or attitudes. The first function type includes communication acts such as requesting or demanding which require an overt response from the listener, thereby allowing the speaker to judge the success of the communicative exchange. The second function type is not as easily explained. According to McLean

PAGE 38

28 and McLean (1978), it is often difficult to extract from the content of the communicative exchange the speaker's underlying purpose or intent. These simplistic global taxonomies of language provide an organizational framework for the development of more sophisticated classification schemes. As Soskin and John (1963) observed, "a simple dichotimization of something so complex as verbal behavior is nothing more than a first step; it merely helps "clear the brush" (p. 25^)' The majority of these functional classification schemes were developed for the categorization of adult communication. McNeill (1970) adopted the threefold characterization of language functions described by Buhler (193^) and Jakobson (I960) to determine the purpose of young children's holophrastic utterances. In his analysis, he noted that all three aspects of the classification scheme were present in their one-word utterances including the referential, expressive and conative functions. The referential function, most closely associated with the cognitive or denotative aspects of language, is manifested in its purest form by naming or labeling. According to McNeill (1970), purely referential utterances never occurred. When children appeared to be engaged in labeling, they were actually using words predlcatively as comments about the situations in which they found themselves (DeLaguna, 1927). The expressive function is the use of verbal comments and paral inguistic devices to reveal the speaker's feelings and attitudes about what is being said. The conative function influences the behavior of others and is manifested by vocative and imperative constructions. Piaget (1923) was the first investigator to develop a functional classification scheme specifically for analysis of children's language.

PAGE 39

29 In his book, The Language and Thought of the Child , he attempted to answer the following question: "What are the needs which a child tends to satisfy when he talks?" (p. 25). Piaget (1923) classified the talk of children into two main categories: egocentric and socialized. Egocentric talk is i nt rapersona 1 in nature because the child, while engaged in it, does not talk at a specific audience nor require a response from a listener who happens to overhear the utterance. The child "talks either for himself or for the pleasure of associating anyone who happens to be there with the activity of the moment" (p. 32). Egocentric speech accompanies, reinforces and supplements the speaker's actions and is manifested in monologues and the pleasurable repetition of syllables and words. Socialized talk is interpersonal in nature. "It addresses the listener, considers his point of view, tries to influence him, or exchanges ideas with him" (Johnson, Darley & Spr iestersbach, 1963, p. 180). Within the socialized function, children may use language to express adapted information, emotionally toned remarks, questions, answers, social phrases and dramatic imitations. Recent additions to the child-oriented functional classification schemes include one developed for preschool educational settings by Schachter et al. (197^) and another designed by Halliday (1975) which can be applied from the earliest stages of infancy to adulthood. Schachter and her colleagues (197^) developed an instrument or scoring scheme called the Functions of Interpersonal Spontaneous Preschool Speech (FiS-P) to identify developmental changes in the patterns of spontaneous speech in children, ages 2 to 5, and to determine the effects of specific sociol inguistic variables such as race and socioeconomic status on these patterns. The instrument was developed

PAGE 40

30 empirically based on 2,000 actual statements, from 100 preschoolers observed during freeplay activities. Only spontaneous initiations by the child, with no prodding or shaping by adults and no direct questioning by peers, were considered acceptable for scoring. The FIS-P consists of a comprehensive set of nine function categories which may be further subdivided into subcategory scores. According to the authors, the category scores cover motives for all spontaneous interpersonal statements including categories for personal motives (I to IV), social motives (V to Vil) and other motives (VIII and IX). The subcategory scores designate the main means for implementing the functions, including the following: I. Express i vef unct ions to express an emotion, e.g., "ouch"; "I hate this"; "Goodie." II. Desire lmplementing-functions to implement a personal desire for an object, for help, for permission, for general reassurance or attention, e.g., "Can I have some"; "Fix it"; "Stop it." ill. Possession Rights lmplementing-functions to implement possession rights involving objects, territory, turns or roles (fantasy or real), e.g., "This is my dolly"; "I had it first"; "1 want to be nrammy . ' ' IV. Ego-Enhanci ng — functions to enhance the subject's ego, with context and tone showing evident pride, e.g., "Look at my big house"; "We're sharing"; "That's junk"; "You're cheating." V. Sel f-ReferringI ncl udi ngfunctions to join subject to other by se 1 freferr i ng the other's statements, activity or characteristics, e.g., "Me too"; "I listen to batman, too."

PAGE 41

31 VI. Join]jT£-functions to join otiier to subject with speaker actively initiating the union, e.g., "Hello"; "Are you my friend?"; "Let's play blocks"; "I'm not playing with you." VII. Col laborati ve-functions to initiate or maintain a roledifferentiated social interaction, with two or more subjects participating in a project, discussion or game, e.g., "I'll shoot and you fall down." VIII. Learning I mpl ement i ngf unct i ons to implement learning about objective world, social world, biological world or how to proceed in a task, e.g., "What does that say?"; "This is a dump truck." IX. Report i ng-f unct ions to share an observation, thought or experience with another, e.g., "I went to the circus"; "Amy has new boots"; "It's raining." In a critical review of the FIS-P, Cazden (197^) commented. This is a unique study of developmental change in what Schachter calls 'self-motivated' social speech, that is speech addressed to but not elicited by, other children and adults. The use of developmental theory, here Piaget's, as a base for strong hypotheses about developmental changes is in sharp contrast to research in which some category system is applied and developmental trends found in the data then discussed. The extent to which the hypotheses were confirmed is very encouraging, and Schachter's discussion of her results is sound. The FIS-P, with some modifications as the author suggests, merits further use by herself and others. (p. 8l), Halliday's (1970) functional classification of language was the first system to be developed for linguistic reasons — to shed light on the nature of linguistic structure and to attempt to answer the question, "Why is language as it is?" (p. 1^1). In 1975, Halliday offered an integrative explanation of language function in children and adults. Through his intensive study of the language development of his son, Nigel, Halliday (1975) traced the

PAGE 42

32 evolution of language function from earliest infancy to adulthood. Based on a sociosemioti c tiieory, his functional approach to language emphasized the importance of language for the education and socialization of the child. According to Halliday (1978), "a child learning language is at the same time learning other things through language building up a picture of the reality that is around him and inside him" (p. 1). Language function rather than form is at the center of this educating and socializing process since it is the structural organization of language which reflects the changing functions to which language is put and not vice versa. Halliday's (1975) analysis of the development of language function proceeds through three phases and shows how the child's initial functional-linguistic system develops and changes into "an adult system that is both similar to and different from that of the child" (Rees, 1978, p. 252). During Phase I, which covered the period from I0V2 months to I8 months in Nigel's life, an idiosyncratic but consistent set of vocal symbols was used to express six initial functions: 1. The Instrumental or the "I want" function. In this function, the child uses language to obtain objects and services to satisfy needs . 2. The Regulatory or the "Do as I tell you" function. This function of language is used "to get someone to do something and is different from the instrumental function in focusing on the agent rather than on the object" (Rees, 1978, p. 252). 3. The Interactional or the "me and you" function. This is language used by the child to interact with the significant others in the environment.

PAGE 43

33 ^. The Personal or "Here I come" function. This is language used to express and develop the child's uniqueness as an individual and to intrude the child as a personality into the speech situation. 5. The Heuristic or "Tell me why" function. This is the function of language to explore and learn about the environment which develops from its most basic form, the request for names of objects, into the entire range of questioning forms that the young child uses. 6. The Imaginative or "Let's Pretend" function. This is the use of language to create a make-believe environment, including soundplay, songs, rhymes, story-tellings and eventually language as a verbal art form. These six functions represent what the child can mean during this earliest stage of linguistic development, when the child's twolevel language system represents each utterance in terms of a content (meaning) and an expression (sound). This system differs from the three-part adult language system which contains an intermediate level of organization between meaning and sound consisting of a grammar and a vocabulary. The utterances of Phase I consist of vocal postures which have neither structure nor words. Although word-like elements may be identified, these units are not necessarily imitations of the permissible sound sequences in the adult language. Each idiosyncratic vocal symbol used by Nigel during this phase was associated with one specific function. By the end of Phase I, utterances become more recognizable as words in the adult language system even though each utterance still performs only one function. it is impossible to mean more than one thing at a time. With Nigel, the first four functions of the list appeared first, followed by the appearance of the other two functions.

PAGE 44

3k In Phase II, the transition into the adult system begins. For Nigel, this transitional period began between I6V2 and I8 months and continued until the end of his second year. During this phase, a seventh specific function emerged--the informative or "I've got something to tell you" function. In this function, language is used as a means of communicating an experience to someone who did not share it. The dominant characteristic of Phase II is functional generalization. Through this process, two broad functional categories or "macro-functions" emerge. The pragmat i c function or "language as doing" derives from the instrumental and regulatory functions of Phase I. This function is reflected in utterances which demand a response such as "more meat," "mommy come" and "fix train." The mathet i c function or "language as learning" arises from the personal and heuristic functions, and includes utterances which require no response, "green car," "two book," and "tiny red light." The interactional function of Phase II contributes to both derivative functions. It is interesting to note that Nigel made the distinction between these two functions more explicit by consistently producing pragmatic utterances with a rising intonation and mathetic utterances with a falling tone. This grouping of Phase I functions into pragmatic and mathetic macro-functions permitted Nigel to play two very different roles as a language user, an observer role and an intruder role. In the observer role, language served to express the experiences of his external and internal worlds. In the intruder role, language embodied his participation in the speech s i tuat ion--h i s roles, attitudes, wishes and judgments. At the beginning of Phase II, all utterances were either pragmatic or mathetic. Gradually by the end of this phase, every

PAGE 45

35 utterance was both pragmatic and mathetic, a significant advance toward the adult system. Nigel had learned that language may be used to both observe and interact with the environment at the same time. Phase II was also marked by two major linguistic developments: (a) rapid growth in vocabulary and structure, and (b) the emergence of dialogue. In Phase II, the intermediate level of linguistic form, which Halliday labels " 1 ex ico-g rammer" begins to develop. For the first time, Nigel's language consisted primarily of standard lexical items. Initially, these items were restricted in meaning to one function only, e.g., "cat" meant only "hello cat!" (interactional). However, as Phase II progressed, Nigel learned to use the same word to express different functions in different grammatical contexts. The impetus for this rapid expansion in vocabulary seemed to be motivated not by the pragmatic concerns of obtaining objects and services, but rather by the mathetic concern for learning about the environment. New vocabulary is used, at first primarily in the context of observation and recall. The pragmatic function, however, seemed to contribute more to the development of language structure, the device which enables the speaker to play both observer and intruder roles at the same time. Structure frees the speaker to mean two things at once. Simultaneous with the acquisition of structure, the child learns to engage in dialogue. By dialogue, Halliday means the ability to adapt and assign communication roles in the context of a verbal interaction. For Nigel, this skill was mastered during a two-week period around I8 months when he learned to imitate dialogue and respond to a WH-question, a command a statement, and a response.

PAGE 46

36 Phase III marks the entry into the adult linguistic system, a system which is multifunctional and mul t i s t ratal . For Nigel, this phase began around 2k months. It includes two major components or, according to Hal li day, "metafunctions ," the ideational and the i nterpersonal . A third function, the textua 1 , provides the cohesive framework within which the meaning of the other two components is organized. The ideational component derives from the mathetic, the observer function of language. It is the functional component concerned with the representation of experience, "language as a means of talking about the real world" (Halliday, 1975, p. 17). The pragmatic, or intruder function of language provides the context for the development of the interpersonal component. This component of meaning is concerned with "the communication process as a form and as a channel of social action" (Halliday, 1975, p. 53). It is "language as a means whereby the speaker participates in the speech situation" (Halliday, 1975, p. 17). The textual function implies genre, a mode of organizing meaning that relates language use to social context. For example, . . . the texture of discourse depends not only on structuring the parts in an appropriate way and joining them together, but on doing so in a way that relates to the context-as narrative, as dialogue, or whatever generic mode is selected. (Hal 1 iday, 1975, p. Ill) In Phase ill, the notion of "function" is no longer synonymous with "use." The child entering Phase III has acquired unlimited uses for language. Every utterance, whatever its use, has both an ideational and interpersonal component of meaning. Furthermore, the original Phase I functions and the informative function of Phase II have become the uses of language, "the generalized social contexts of language use" (Halliday, 1975, p. 58).

PAGE 47

37 Global taxonomies of language function provide only the broadest outline of language use. Many of the finer discriminations of language use are not captured by their all-inclusive categorizations. A simple classification of child language using any of the functional schemes described should accurately characterize the purposes which language serves in the child's social and cognitive growth, but would not necessarily be descriptive of the functions that individual utterances may serve. It is this distinction which is often difficult to draw and maintain. Many authors reserve the term "function" to describe the small number of higher order categories just defined and offer the terms "strategy" or "use" to describe the larger number of speech or communicative acts through which the functions of language are realized. A refined description of language use would venture beyond a mere description of language functions to a classification of the many interrelated options from which speakers choose when they communicate. These options, according to Halliday (1970), represent the meaning potential of language and involve the creative and repetitive selection of communication strategies in the context of a wide variety of social contexts . Speech Acts Theory The philosophers J. L. Austin (1962) and John Searle (I969) have stated that speakers produce sentences because they are attempting to accomplish something with words. This emphasis on language use rather than language form or content probably stems from Wittgenstein's (1958) prophetic maxim that the meaning of an utterance is its use. Austin (1962), in his work How To Do Things with Words , distinguished

PAGE 48

38 between the locutionary, i 1 locut ionary and per locut ionary acts which speakers perform when they use language. In saying something, ... we perform a locutionary act , which is roughly equivalent to uttering a certain sentence with a certain sense and reference . . .'meaning' in the traditional sense. Second, ... we also perform i 1 locut ionary acts such as informing, ordering, warning, undertaking, etc., i.e., utterances which have a certain conventional force. Thirdly, we may also perfor m perlocutionary acts : what we bring about or ach ieve ^ say i ng something, such as convincing, persuading, deterring . . . .(Austin, 1962, p. 108) Austin's (1962) componential analysis of speech emphasized that all utterances are produced with the purpose or intent (il locution) of conveying a specific message or content (locution) to a listener upon whom they have some effect (perlocution) in modifying behaviors, thoughts or beliefs. In his original analysis of utterances, he used the term "speech acts" to refer to all three aspects of a message-locutionary, i 1 locutionary and perlocutionary. More recently, the term has come to be reserved for only the i 1 locutionary force of an utterance. The i 1 locutionary speech act has been defined by Bates (1976b) as "a conventional social act, recognized as such by both speaker and hearer, that takes place when a sentence is uttered, e.g., a command is issued, a child is baptized" (p. 14). Austin (1962) was the first to note that there were some utterances in which saying something could also be regarded as doing something or performing some action with words, as in the following exampl es : Minister to couple: I now pronounce you man and wife. Judge to prisoner: I sentence you to 30 days in the county jail. One baseball fan to another: I bet you that the Pirates will beat the Dodgers.

PAGE 49

39 In each instance, by saying something, the speaker performs an act: a couple is married; a prisoner is sentenced; a bet is placed. Each of these examples contains what Austin (1962) called a "performative" verb that makes explicit the acts the speaker intended the utterance to perform. As long as these acts are executed correctly, completely and with sincere intent by authorized persons under appropriate circumstances, then the specified acts are performed. However, should any of these conditions be violated, then the act is void. These conditions, usually referred to as "felicity" or "happiness" conditions, must be respected or the utterances spoken do not perform the acts they encode. Austin (1962) went on to develop the first taxonomy of speech acts, though by his own admission, his classification scheme was preliminary and tentative. In this first taxonomy of speech acts, every act could be classified under one of five very general categories: verdictives, exercitives, commissives, behabitives and expos itives. In 1969, J. R. Searle expanded upon and refined Austin's work according to the speech act, the important role which it now plays in the analysis of language use. To Searle (I969), the speech act is the basic unit of human communication upon which content and form are mapped for the purpose of conveying an infinite number of messages. As Searle (1969) states, "speaking a language is performing speech acts . . . that . , . are in general made possible by and performed in accordance with certain rules for the use of linguistic elements" (p. 16). These rules delimit one use of language from another so that there exist a limited number of basic things which speakers do with language.

PAGE 50

ko Searle (1975) took issue with Austin's (1962) classification of i 1 locutionary acts and offered an alternative taxonomy in its place. His main criticism centered around the fact that he could find no consistent set of principles upon which the taxonomy was constructed. This weakness, in turn, led to a great deal of overlap from one category to another and considerable heterogeneity within some of the categories. Throughout the system, there is a persistent confusion between verbs and acts and in some instances, the verbs listed are not even i 1 locutionary verbs. Searle' s (1975) taxonomy of i 1 locutionary acts presupposes at least twelve significant criteria for distinguishing one type of illocutionary act from another, including (a) differences in the purpose of the act; (b) differences in how the propos i tional content of an utterance relates to the world; and (c) differences in expressed psychological states toward the proposi tional content associated with various types of i 1 locutionary acts. Applying these criteria, Searle (1975) delineated five categories of speech acts, including: 1. Representati ves . Speakers convey propositions which they believe to be true, including suggestions, hypotheses and assertions. 2. Pi recti ves . Speakers attempt to get listeners to do something when they utter directives such as ordering, requesting, begging or pleading. 3Commi ss i ves . The goal of commissives is to bind the speaker to some future course of action, as in promises, vows, pledges, contracts and guarantees. k. Express i ves . To express a psychological state about something, speakers utter acts such as apologies, welcomes, congratulations.

PAGE 51

41 These acts specify how good or bad speakers feel about a state of af fai rs. 5. Declarations . When speakers utter declarations, they bring about a new state of affairs. By saying "You're fired" or "i now pronounce you man and wife," speakers change employment and civil status. Declarations are a special type of i 1 locutionary act which Austin (I962) originally referred to as "performatives." Each of these categories serves speakers in different ways to do things with words. Representatives tell listeners how things are. Directives attempt to get listeners to do things while commissives bind the speaker to do things. Expressives reveal the speaker's feelings and attitudes and declarations bring about changes in status and conditions . A number of investigators have adapted Searle's (I969) classification scheme to describe how young children use their utterances. Greenfield and Smith (1976) focused on the assertions (representatives) and requests (directives) which children produce at the one-word stage. Antinucci and Parisi (1973), and Slobin (1970) performed a similar analysis on children's two-word productions. Grimm (in Clark S Clark, 1977) observed the ability of older children to use directives and commissives while Berko-Gleason (1973) and Berko-Gleason and Weintraub (1976) were concerned with the acquisition of expressives embodied in social rout i nes . Dore (197^, 1975, 1976, 1978) was the first investigator to develop a speech acts classification framework based on the utterances which children produce. While Halliday (1975) evolved an integrated theory of pragmatic development from the pre 1 i ngui s t i c period through

PAGE 52

42 the production of multiword grammatical utterances, Dore (197^, 1975, 1976, 1978) developed several classification systems which may be applied to the child's output at different linguistic stages. Dore (197^, 1975) used a speech acts framework to analyze the single word utterances of two children at approximately 15 months of age. He described early one-word messages as "primitive speech acts" (PSAs) consisting of two elements: a "rudimentary referring expression" and a "primitive force." The "rudimentary referring expression" is the single word, the semantic component of the message, while the "primitive force" refers to the communicative intent of the utterance and is expressed by the prosodic pattern which accompanies production of single words. In this model of children's early utterances, a "rudimentary referring expression" may be used to express several primitive forces or communicative intentions. For example, "mama" produced with falling intonation while the child pats mother on the knee serves as a label and requires no response. "Mama" produced with rising intonation while the child unsuccessfully attempts to open a jar of peanut butter serves as a request for action and requires mother to perform the action. Finally, "mama" said loudly with an abrupt rising-falling intonation while mother is across the room serves to call mother to the child. By varying the intonational contour of the utterance, it is possible for the child to use the same word to perform three different PSAs--label ing, requesting and calling. Dore (197^, 1975) used the segmental and suprasegmenta 1 phonemic features of the child's utterance as well as the child's nonl ingui stic behavior, the relevant context of the utterance and the anticipated adult response to classify

PAGE 53

k3 children's single-word utterances into nine distinct PSA types. Each of these types is described with typical examples in Table 1. Dore defines a PSA as "an utterance, consisting formally of a single word or_ a single prosodic pattern, which functions to convey the child's intention before he acquires sentences" (197^, p. 3^5)He emphasizes that "a PSA is not merely an elliptical adult speech act, but a qualitatively different entity that possesses only some features similar to full speech acts" (1975, p. 32). For example, a PSA does not contain a predicating expression. It conveys a child's intention without containing propos i t iona 1 structure. ". . . The components of PSAs eventually develop into the propositions and i 1 locutionary forces of speech acts, but this occurs only after the child has acquired most of the grammatical structure of his language" (Dore, 1975, p. 32). From 1976 to 1978, Dore revised and refined his "grammaticali 1 locutionaryi nteract ional " model to classify the structure and function of nursery school conversation. According to the author, "the model is intended to show how grammatical forms are chosen to convey i 1 locutionary intentions in the service of accomplishing social interaction (Dore, Gearhart S Newman, 1978, p. 339). Central to the model is a class of i 1 locutionary acts which serve as the primary units of conversation and mediate between the grammatical forms that signal them and the interactional purposes for which they are used. In this three-part model, utterances are classified as conversational acts, each consisting of a prepositional content , grammatical form and an i 1 locutionary function . Thirty-five individual conversational acts have been identified through the analysis of videotaped sessions in which seven 3-year-old children engaged in a wide variety

PAGE 54

Table 1 Dore's Primitive Speech Act Types P r i m i t i ve Speech Act Description of Example Label i ng Child says "eyes" while touching doll's eyes Repeat i ng Child says "data" after overhearing mother say doctor in a conversation with the teacher Answering Child says "bow wow" after mother points to a picture of a dog and says "What's this?" Requesting (act ion) Child says "uh? uh? uh?" while unsuccessfully trying to push a peg through a hole and mother responds by hel ping Request i ng (answer) Child says "book?" while picking up a book and looking at mother Cal 1 i ng Child loudly shouts "mama" with distinct intonation while mother is across the room Greet i ng Child says "hi" when teacher enters room Protesting Child says "no" while resisting mother's attempts to put on his shoes Practicing Child says "daddy" when he is not present; mother does not respond Modified from: J. Dore's "Holophrases, Speech Acts and Language Universals." Journal of Child Language, 2, 21-40 (1975).

PAGE 55

of spontaneous interactions at school. Over a period of four nxjnths, 3,000 child utterances were collected and coded for conversational act. The 35 conversational acts identified accounted for 32 percent of the utterances produced by the children. To determine the conversational act which best characterized each utterance, Dore (1976, 1977) used information both internal and external to the utterance, e.g., the literal meaning of the utterance, its grammatical and prosodic characteristics, how it relates new information to old, the speaker's nonl inguistic behavior and related utterances, the verbal and nonverbal behavior of the speaker's interlocutors and the relevant situational context. The 35 individual acts may be grouped together into six general conversational classes, including: (a) Request ives which solicit information or actions; (b) Assertives which report facts, state rules, or convey attitudes; (c) Performatives which accomplish acts and establish acts by being said; (d) Responsives which supply solicited information or acknowledge remarks; (e) Regul at i ves which control the conversational sequence; and (f) Expressives which nonpropos i t ional ly convey attitudes or repeat others. Three primary conversational functions are conveyed by these general classes: the transmission of content, the regulation of conversation, and the expression of attitudes. Figure 1 presents a network representation of Dore's (1978) threelevel coding scheme. Two other classification schemes which have combined a speech acts approach within a functional framework for the purpose of describing language use in children are those developed by Wells (1973) and Tough (1976; 1977).

PAGE 56

PRIMARY CONVERSATIONAL FUNCTION GENERAL CONVERSATIONAL CLASS PARTICULAR CONVERSATIONAL ACT i-initiateconvey content (request! ve) solicit i nformat ion sol icit_ 'action (assertive)percei vable phenomena internal phenomena "social" _ phenomena (performative) — initial reactive^respond (responsive) supply sol ici ted — information supply addi tiona) information acknowledge non r eq ue s t i ve Choice Questions Product Questions Process Questions 'Action Requests Permission Requests Sugges t ions * Identi f ications Descriptions ' Internal Reports Eval uations Attributions ' 'Rules Explanations Claims Jokes Teases ' Protests Warnings Choice Answers Product Answers Process Answers Compl iances Clari f icat ion Responses Qual ifications Agreements — [ Acknowledgements regulate conversation (regulati ve)sol ici t other mark "-content" Attention-Getters Speaker Selections Rhetorical Questions Clari fi cation Questions Boundary Markers Politeness Markers express att i tude (expressive) Exclamat ions Accompan iments Repeti tions Modified from J. Bore's "Variation in Preschool Children's Conversational Performances." In K. E. Nelson (ed.). Children's Langua ge. Vol. 1. New York: Gardner Press (1978). — Figure 1 Dore's Codes, Definitions and Examples of Conversational Acts

PAGE 57

47 The Wells (1975) coding scheme classifies the interpersonal function, cognitive content and discourse structure of preschool children's talk (Dore, Gearhart & Newman, I98O, p. 36O) . Most relevant to the present research, he posited 133 functions of communication which can be categorized under several sequence and subsequence conversational modes. Wells (1973) described these functions as "acts" that a particular utterance performs. Like Austin (I962), he was concerned with the purposes that individual utterances serve. However, he went on to emphasize that his interest was not in the analyses of utterances in isolation but rather in the description of utterance function in the context in which it occurs. Wells (1973) viewed communication functions or acts as "the smallest units of verbal interaction — the building blocks from which the edifice of conversation is constructed" (p. 39). In his system. Wells has identified six conversational sequence and subsequence modes. Sequences indicate the dominant purpose or function of a conversation; subsequences mark the subsidiary functions of smaller units of conversation within the overall purpose of the sequence. The sequences or functions, along with exemplary "acts" are as fol lows: 1. Control . The regulation of the present or future behavior of one or more of the participants through acts such as wanting, offering, commanding, suggesting and permitting. 2. Express i ve . The expression of feelings and attitudes as an affective response to situations through acts such as exclaming, taunting, challenging, approving and disapproving. 3. Representational . The exchange of information through acts such as labeling, commenting, questioning, responding and justifying.

PAGE 58

A8 A. Social. The use of conversation to maintain social relationships through acts such as greeting, leave taking and ritualizing. 5. Tutorial . The use of conversation for didactic purposes through the acts of correcting, modeling and imitating. 6. Procedural . The use of conversation to initiate or end a sequence or to resolve a breakdown in communication through acts such as calling, requesting a repetition and reformulating. In a conversational sequence between parent and child, the parent's dominant purpose may be control--to get the child to shut the door upon entering the house. To accomplish this purpose, it may be necessary to proceed through several conversational subsequences. "There may be an initiating (Procedural) subsequence to gain the child's attention, followed by parent's exclamation of surprise or horror (Expressive) before reverting to the original request to which the child complies (Control) (Wells, 1973, p. 8l). in Wells' (1973) system, the same act such as requesting, responding and justifying may occur under different functions or sequence modes. This, in part, accounts for the large number of acts. More importantly, "the acts are formulated to cover a wide array of phenomena, ranging from the purely linguistic (question) to the social (silence filler, evasion). As a consequence, ... a generality crucial to act types may be lost . . . and the acts are formulated at different functional levels. Thus, the central problem with this scheme, as well as with the other empirically based speech act codes for children ... is finding principled criteria for (a) motivating the level of linguistic act identified, (b) constraining the kinds of acts to be included, and (c) defining a decision procedure for classifying utterances into acts. (Dore, Gearhart & Newman, 1980, p. 370)

PAGE 59

^3 Tough's (1976) bi level classification system of language use evolved as a result of a longitudinal study undertaken to account for the role of language in the educational differences of young children from disparate social environments. Borrowing from the theoretical viewpoints of Bernstein (1971, 1975), Bruner OSG^, I966), Luria (1959), Piaget (1923) and Vygotsky (1962), Tough (1977) devised a taxonomy of language use to fulfill two major objectives: (a) to discover differences in the range of meanings that a child attaches to his environment; and (b) to determine the purposes language serves for young children. Table 2 presents an outline of Tough's (1976) functional classification schema. Operational definitions and examples of the various communication strategies can be found in Appendix B. At the broadest level of classification in Tough's (1976, 1977) schema, utterances are categorized, by use, into seven major types: 1. Sel f-ma i nta in i ng . The use of language to create an awareness of the speaker's identity and to promote the speaker's position in relation to others. 2. Pi recting . The use of language to control or regulate the physical actions and operations performed by oneself and others. 3. Reporting . The use of language to provide information about past and present experiences. k. Reasoning . The use of language which employs rational thought and argument to interpret experiences. 5. Predi ct i ng . The use of language to extend communication beyond immediate present or past experiences to events that have not occurred and which may never take place.

PAGE 60

50 Table 2 Tough's Framework for the Classification of Language Use 1. SELF-MAINTAINING Strategies a. Referring to needs b. Protecting the self and self-interests c. Justifying behavior and claims d. Cr i ti ci zi ng others e. Threatening others 2. DIRECTING Strateg i es a. Monitoring own actions b. Directing the actions of the self c. Directing actions of others d. Collaborating in action with others 3. REPORTING Strategies a. Label ing b. Referring to detail c. Referring to incidents d. Referring to the sequence of events e. Making comparisons f. Recognizing related aspects g. Extracting or recognizing the central meaning h. Reflecting on the meaning of experiences A. TOWARDS LOGICAL REASONING Strategies a. Explaining a process b. Recognizing causal and dependent relationships c. Recognizing problems and solutions d. Justifying judgments and actions e. Reflecting on events and drawing conclusions f. Recognizing principles

PAGE 61

51 Table 2 (continued) PREDICTING Strateg ies a. Anticipating/forecasting b. Anticipating the detail of event c. Anticipating a sequence of events d. Anticipating problems and possible solutions e. Anticipating and recognizing alternative courses of action f. Predicting the consequences of actions or events PROJECTING Strategies a. Projecting into the exerpiences of others b. Projecting into the feelings of others c. Projecting into the reactions of others d. Projecting into situations never experienced IMAGINING Strategi es a. Developing an imaginary situation based on real life b. Developing an imaginary situation based on fantasy c. Developing an original story From: J. Tough's Listening to Children Talking . London: Ward Lock Educational (1976) .

PAGE 62

52 6. Project i ng . The use of language within an unfamiliar or external context. 7. Imagining . The use of language to create an environment of makebe 1 ieve. Language use is defined as "the means by which different kinds of meaning or thinking are made evident" (Tough, 1977, p. ^6). In a child's development, the purpose of language is to construct and express these meanings so that "language functions in relation to the child's developing conceptualization of the world around him" (Tough, 1977, p. ^h) and reflects different modes of thinking. Five of the basic uses of language defined above--reporting, reasoning, predicting, projecting, and imagini ng--convey meaning reflective of an ideational appraisal of experience and could be collectively referred to as cognitive uses. The self-maintaining and directing uses convey meaning of an interpersonal nature by indicating the kind of relationship maintained between speakers and 1 isteners and can be referred to as social uses. At the second level of classification, utterances are categorized according to a number of strategies that serve each language use. The strategies bear some resemblance to speech acts (Austin, 1962; Searle, 1969); however, in this instance, it seems more appropriate to characterize them as "thought acts," since it is through the employment of these strategies that inferences can be made about the use of language characteristic of a number of modes of thinking. It is interesting to note that Tough (1977) claims to have arranged the strategies serving a particular use of language in an ascending order of cognitive complexity.

PAGE 63

53 Tough's (1976, 1977) system seemed particularly well-suited to the present investigation since it was developed to resolve issues about the relationship of language use to academic success. It is the only system which claims to have identified the communication strategies which children need to master to succeed in the classroom. Development of Language Use in Language-Normal Children The rules governing the use of language in context form the foundation of the study of pragmatics. According to Bates (1976a), . . . all of language is pragmatic to begin with. We choose our meanings to fit contexts and build our meanings onto those contexts in such a way that the two are inseparable in the same way that 'figure' is definable only in terms of 'ground.' (p. 420) Studies in the acquisition of language use begin prior to speech itself, in the first year of life, "The studies indicate that semantics emerges, developmental ly and logically, from pragmatics, in much the same way that syntax has been shown to emerge from semantic knowledge" (Bates, 1976a, p. 420). Miller and Yoder (1972) state, "Before the child becomes a language user, he needs to have something to say (concepts) and a reason for saying it (semantic intent) as well as a way to say it (linguistic structure)" (p. 210). Halliday (1975), Bloom and Lahey (1978), and Prutting (1979) have identified a series of stages which characterize the development of language use from infancy to adulthood. At the first level of primary forms , the infant uses a limited repertoire of behaviors (e.g., gazing, crying, touching) to meet several basic physiologic and/or affective needs, regardless of the context. During this period, from

PAGE 64

St* birth to approximately 9 months, infants accomplish nonl inguistical ly those communicative functions which they will later learn to convey linguistically (Brunner, 197^; Halliday, 1975; Mahoney, 1975). Bates, Camaioni and Volterra (1975) have labeled this the periocutionary stage of development, corresponding to one aspect of Austin's (1962) tripartite characterization of speech acts. Perlocutions have an effect on the listener but they are not recognized by both speaker and listener as conventional communications (Bates, 1976a). In early communications between parent and child, there is no evidence that the infants, themselves, are aware of the communicative value of their signals. Parents react to the infant's signals as if they were produced intentionally, thereby bringing about the desired effect, but the infants, themselves, are not aware of the conventional purpose of these signals. Therefore, "these early signals function as signals only for the 1 istener" Bates et al ., 1975, p. 212). By the second level of conventional forms , children have learned sounds and gestures which they apply in different ways. At the beginning of this stage, just prior to the emergence of the infant's first words, meaning is transmitted intentionally through sound and movement. Infants are unable to use the conventions of language but they "come to know that their behavior can influence the behavior of others and they behave with the intention of achieving that influence" (Bloom & Lahey, 1978, p. 20k). Bates et al . (1975) have labeled this the i 1 locutionary stage of development, since it is during this period, from approximately 10 to 14 months, that infants intentionally use vocalizations and gestures such as showing, giving and pointing to regulate joint attention and joint actions (Bruner, I97A). These behaviors, performed in at least two different communicative contexts.

PAGE 65

55 have been termed "proto-decl arat i ves" and "protoimperatives" by Bates et al. (1975). Through the proto-declarat i ve, the infant directs the listener's attention to some object or event. The protoimperative serves as a means to use the listener to obtain a desired object. These early communicative behaviors are carried out in the absence of speech and are the precursors to the development of symbolic communication. The acquisition of conventional means of communication begins to emerge around 14 months of age in what Bates et al. (1975) have labeled the locutionary stage. At this point, infants acquire words which they substitute for the nonl inguistic performative schemes developed during the i 1 locut ionary stage. Their communication, in this stage, is intentional as well as conventional, even though the very first words which they produce seem to be an integral part of the gestures which accompany them, e.g., "bye-bye + hand wave" or "up + hands outstretched." The goals of these early communications continue to be linked to establishing joint attention and joint actions and are classified as assertions and requests. Assertions correspond to the proto-decl arat i ves described earlier and requests correspond to protoimperatives. Clark and Clark (1977) have speculated that, . . . gestures and speech acts are parallel in function: pointing gestures require acknowledgement from the listener just as assertions do, and reaching gestures require an action or answer just as requests do. This hypothesis gains added support from the fact that children continue to use pointing and reaching gestures together with single words and even combinations of words, (p. 313-31^) In addition to these two principal categories of early speech acts, children at the one-word stage of development have been observed to use a few primitive expressives such as greeting someone with "hi" or expressing dismay with "uh-oh."

PAGE 66

56 As the form of children's early utterances becomes more elaborate, their functional capabilities expand also, Slobin (1970), Bowerman (1973) and Brown (1973) have observed that children at the two-word stage use assertions to acknowledge the existence or nonexistence of an object, to describe the location, possession or quality of an object and to comment about an ongoing event. Requests are of two types: requests for action and requests for information. At the third level of conventional use , children elaborate their use of language along two dimensions. First, as they grow in linguistic ability, they add new types of speech acts to their repertoire. According to Clark and Clark (1977), they begin with representatives (assertions) and directives (requests) and then add expressives, commissives and lastly, declarations. Second, they learn that alternative linguistic forms exist for achieving the same purpose in different situational contexts. To get a cookie, it's possible to ask a question, "May I have a cookie?"; to make a statement, "I want a cookie"; to issue a command, "Gimmie a cookie!"; or hint in a less direct fashion, "That cookie sure smells yummy!" The present study focuses on the former dimension — the development of a set of intentional options for communication in a wide variety of cognitive situations. This stage of development generally begins in the third year of life and continues into adulthood. Only a few investigators have described the older child's Increasing ability to communicate a large number of intentions through language. Piaget (1923) was among the first. His system was adopted and modified by McCarthy (1930) and Davis (1937) in their studies of language development in children. Although there was some disparity

PAGE 67

57 in the results obtained by each of these investigators, and in addition, by Day (1932) and Smith (1935), the sampling protocol in the studies was different. Piaget's (1923) subjects experienced complete freedom of movement as their communicative exchanges were recorded during free play situations with other children at school. Smith's (1935) data were collected under similar circumstances of an unstructured nature. The communicative interactions of her subjects were recorded either while children were at play with one another or in the home environment with adults. In contrast, McCarthy (1930), Day (1932) and Davis (1937) gathered their samples from children who were looking at books or playing with toys in highly structured situations with adults. As a result of these differences in sampling strategies, the overall ratio of egocentric to socialized speech was different for the studies. Piaget (1923) and Smith (1935) observed egocentric speech occurring from 26 to 'O percent of the time; the other investigators reported this type of speech in less than 10 percent of the utterances, with children beyond k years of age engaging in its use 1 to 2 percent of the time. Conversely, socialized speech constituted between 60 to 8^ percent of the conversation used by Piaget's (1923) and Smith's (1935) subjects and between 93.5 and 99.8 percent of the speech used by the subjects in the other investigations (Davis, 1937; Day, 1932; McCarthy, 1930). This wide discrepancy in the functions which language serves may be one of the first pieces of evidence to support the now widely held belief in the important influence of context or situation on what is being said, how it is being said and for what purpose. Children engaged in highly structured communication with adults have little opportunity to use egocentric language. The situation demands that

PAGE 68

58 their remarks be socialized. On the other hand, when children are playing alone or in the presence of other children, it is not uncommon to hear them talking to themselves "or for the pleasure of associating anyone who happens to be there at the moment" (Piaget, 1923, p. 32). Taken together, the four studies (Davis, 1937; Day, 1932; McCarthy, 1930; Smith, 1935) demonstrated a decrease with age in the amount of egocentric speech used. This finding lends support to the claims by Piaget (1923) and Vygotsky (1962) that in later life, inner language or thought supplants this type of talk to self. With regard to the present study, the observation that "answers" comprised the most important category at all ages is of particular importance (Davis, 1937). The frequency of occurrence of this type of speech ranged from 25-5 percent at age 5V2 to 37-1 percent at age 9V2. McCarthy (1930) also found the percentage of answers to be increasingly important in her upper two groups. Answers comprised from .3 percent of the sample at 1 V2 years to 31 and 26 percent, respectively, at k and k% years. That answers play a vital role in the child's communicative system, both at school and at home, is further supported by estimates of the relative rates of parental questions to their young children. These range from 10 percent of the speech addressed to 12-month-old children (Fri edlander, Jacobs, Davis & Wetstone, 1972) to 32 percent of the speech addressed to 13-month-old children (Nelson, 1973) and 37 percent of the speech addressed to children between the ages of 2 and 5 years (Nelson, 1973; Baldwin S Baldwin, 1973). Labov (1970) has been critical of the question-answer format in obtaining information about the language development of disadvantaged children. He has observed language which is more elaborate in both

PAGE 69

59 form and function collected under less structured and more familiar circumstances. Although his observations would certainly indicate that, given the right set of conditions, these children do demonstrate complex linguistic skills, it does not offer a solution to the persistent problem of their limited achievement in the most common interactive mode between teacher and child and parent and child. Schachter et al. (197^) identified developmental changes in the pattern of language use in 170 preschool children, ages 2 to 5. Using the FIS-P, a Piagetian-based taxonomy of language use, children's spontaneous initiations of speech were analyzed. The results of this analysis revealed a pivotal shift in the development of speech function patterns at age 3. Before this age, patterns of speech function consisted mainly of more immature forms such as desire implementing, reporting on self and things, "me too" sel freferr i ng and learningimplementing (naming words), with adult-addressed speech at its highest level. Schachter (197^) and her colleagues labeled this collection of early speech functions "Primary Socially Interdependent Speech," noting that it was reflective of young children's tendency to use language egocentrical ly rather than sociocentrical ly. Functionally, these patterns served to insure "mutual gratification during the primary interdependent attachment between child and caretaker" (Schachter et al . , p. i ) , After age 3, the early speech patterns maintained their frequency but speech reflecting increased levels of self-other differentiation showed a marked rise. The use of ego-enhancing boasting statements increased abruptly and a number of peer-addressed collaborative and joining statements called "Secondary Social Speech"

PAGE 70

60 progressively rose with age. Finally, speech patterns including modulations (explanations and justifications) and collaborative disagreeing, which seemed to be adapted to the needs of the listener, emerged with high frequency at age k or 5. These patterns were labeled "Tertiary Socialized Speech." The authors concluded that the passage from primary to tertiary patterns of speech usage was "consistent with both Piagetian and psychol inguistic theorizing on the emergence of a significant degree of ego differentiation around 3" (Schachter et al., 197^, p. ^7). Dore (1976) videotaped seven middle-class children over a period of seven months as they engaged in a wide variety of activities at their nursery school, in order to classify their communicative intentions. The children ranged in age from 3^ to 39 months and produced almost 3,000 i 1 locut ionary acts. These acts could be classified into 32 different types grouped under six major categories. A distributional analysis of the relative proportions of use of each of the major category types revealed the following results: requests (27 percent), responses (I8.5 percent), descriptions (22.3 percent), statements (13.8 percent), conversational devices (5.8 percent) and performatives (10.8 percent). Another 7-9 percent of the remarks were coded as uninterpretable and 5.8 percent were double coded, as the intent of the utterances was equivocal. It can be seen that requests accounted for the largest proportion of the children's i 1 locutionary acts, with requests for action (10.0 percent) and requests for information (7.3 percent) outnumbering their request types. Descriptions constituted the second largest category of i 1 locut ionary acts, with major contributions to the overall proportion made by identifications (7.9 percent)

PAGE 71

61 and descriptions of events (^.0 percent). The third largest category of acts was statements with internal reports about feelings and attitudes (7.7 percent) contributing most to the overall frequency of occurrence. Fogel (1976) interviewed 288 children between the ages of five and fifteen to study developmental patterns of functional communication. His interview technique paired verbal prompts with pictorial stimuli to elicit speech acts reflective of language use for two different functions: controlling and feeling (Wells, 1973). Twelve stimulus items were designed to represent a wide variety of "real life" communication situations. In one item, for example, children had to pretend that they were a teacher trying to explain why a student could not have a book to read. The situations provided information about three important factors in any communicative exchange: (a) the function of the communication (controlling or feeling); (b) the significant other in the situation (mother, teacher or peer); and (c) the perspective taken within the situation (commenting on behalf of oneself or someone else). Three general findings emerged from the analysis of the speech act data. First, there was a significant age effect represented mainly by a change in the communication behavior of the fifth grade children. Second, there was a significant effect involving context which indicated that the primary basis upon which children differentiated between communication situations was the function of the communication employed and not the participants involved. Third, there was a significant interaction between context and perspective which Fogel (1976) interpreted to mean that perceptions of communication differences between the self and other are dependent upon the context in which one finds oneself.

PAGE 72

62 Further analysis of the results revealed that the shift in communication behavior at the first grade level was accounted for by increased diversity in speech act usage and increased use of parasituational comments — statements about the communication situations rather than statements reflective of actual dialogue. These comments set the scene for dialogue by explaining aspects of the situation to the interviewer in order to enhance understanding. With regard to this behavior, Fogel (1976) hypothesized "that older children are more likely to present rules about situations than to present actual behaviors that may occur" (p. 209) . Three speech act types accounted for the developmental trends observed in this study including direct requests, cajoling and contractual acts. The use of direct requests for permission to do something or to go somewhere declined from the first and third grade levels {6k percent) to the ninth grade (6 percent). According to Fogel (1976), "this finding supports previous indications in the literature that younger children are more likely to employ "head-on" communication strategies when attempting to persuade others" (pp. 209-210). Similarly, the speech act type of cajoling, i.e., pleading to change the mind of someone about a decision, decreased in total usage from the first and third grade levels (50 percent) to the ninth grade level (3 percent), and occurred more frequently in communication situations involving the mother. Finally, contractual type acts were used very infrequently by first (12.5 percent), seventh (O.78 percent) and ninth (17.13 percent) grade children, but peaked in use by third (32.41 percent) and fifth (30.09 percent) grade children.

PAGE 73

63 Fogel (1976) also observed children in the earlier grades using a wider variety of speech acts when communicating from another's perspective than when communicating from their own perspective. This effect was most pronounced when children took the role of teacher and least pronounced from the mother's perspective. In addition, feeling situations showed this effect more strongly than controlling situations. In other words, children saw the other as having a more diverse repertoire of responses than the self, particularly with regard to feeling situations. Fogel (1976) hypothesized that this greater discrepancy between self and other in the feeling situations might be due to the relative unfami 1 iari ty of these situations for the young child. He concluded that younger children attributed to others a greater variety of speech act capability than they attributed to themselves, which demonstrated that they have difficulty reflecting on themselves as communicators in a variety of situations, particularly those which are unfami 1 iar. Several investigators have studied older children's use of specific types of speech acts such as directives (Ervi n-Tr ipp, 1977; Garvey, 1975; Grimm, 1975; Mi tchel 1 -Kernan , 1977), commissives (Grimm, 1975) and expressives (Berko-Gleason S Weintraub, 1976). Grimm (1975) had children of 5 and 7 years roleplay situations with Felix, a large toy cat, in order to get him to comply with four directives and a commissive. Both the 5" and 7-year-old children performed consistently well when asking Felix for something, ordering him to do something and forbidding him to do something, but the 5-year-olds had some difficulty formulating acceptable utterances for permitting Felix to do something. These results showed that 5-year-old children had not yet mastered a

PAGE 74

64 complete repertoire of directives. The commissive presented problems for both the 5" and 7-year-olds, with 57 and 55 percent of their remarks achieving acceptable standards. Clark and Clark (1977) suggested "that children find it much easier to work out the conditions under which the listener is expected to do something than the conditions under which the speaker is" (p. 366). These findings support Chomsky's (1969) comprehension studies of directives and commissives which showed that children understood the directive, tell, long before they understood the commissive, promise. Under 8 years of age, children interpreted the verb "promise" as if it were a directive. Garvey (1975), Ervin-Tripp (1977), and Mi tchell -Kernan and Kernan (1977) observed the use of directives in young children. Children's ability to convey and respond to requests for action based on spontaneous dyadic interactions of younger children (3.9-4.4 years) and older children (4.7-5-7 years) were explored by Garvey (1975). Both younger and older dyads were observed to produce equal numbers of successful direct requests (e.g., "Open the door!"). Fewer indirect requests (e.g., "Can you open the door?") appeared in the data, with the older dyads achieving, on the average, twice as many successful attempts as the younger dyads. Among the direct requests produced, no examples of requests with performative verbs were noted (e.g., "1 order you to open the door!"). Garvey (1975) concluded that for children in the age range studied, . . . the request for action rests on a set of meaning factors which are relatively specific, which may be expressed in variant forms, and which are available to the child in either the discourse role of requester or of recipient of the request (p. 62)

PAGE 75

65 Ervin-Tripp (1977) traced the evolution of children's directives from their earliest expression prior to 2 years of age. At this age, children produced directives with combinations of gestures, names of desired objects and some limited linguistic markers like "more" and "want." By 3 years, they were capable of using embeddings and structural modifications to produce a number of alternative directive forms such as "Can I have my big boy shoes?" and by k, they employed verbal strategies requiring several steps as well as hints to negotiate interactions, e.g., "We haven't had candy in a long time." The hardest forms for children to master were those that did not explicitly identify what was wanted--question directives of an indirect type and affirmative hints, e.g., "It's hot in here." Ervin-Tripp (1977) concluded that "wide use of tactful deviousness is a late accomplishment" (p. 188). In a study of older children from 7 to 12 years of age, Mi tchel 1-Kernan and Kernan (1977) examined the use of directives by a group of black children as they occurred in conversation with one another and with adults. Using Ervi n-Tr i pp' s (1977) classification scheme of directive variants, the authors found that these children had mastered all of the forms appropriate for directives and showed an awareness of at least some of the social factors which help to determine the situational appropriateness of the use of one form over another. The types of directives included: (a) need statements, e.g., "I don't want no more fighting out of the girls"; (b) imperatives, e.g., "Let my brother alone"; (c) imbedded imperatives, e.g., "John, would you please tell that lady to quit?"; (d) permission directives, e.g., "Can I speak to her?"; (e) question directives, e.g., "Hey, you

PAGE 76

66 got a quarter, Mac?"; and (f) hints, e.g., "Last person talk to me like that is in his grave." Although a few early express ives appear in the language of 2and 3~year-old children, their use, in general, is not mastered or fully comprehended until much later in the child's development. This aspect of language is one that parents teach by rote in the appropriate social contexts by insisting that children say "please," "thank you," "you're welcome," "I'm sorry." The use of expressives places no obligation on either speaker or listener; rather they express the feeling expected within a particular society for a particular situation. "As such, expressives are often hard to explain or justify to small children, and parents don't usually try" (Clark & Clark, 1977, p. 366). As the literature demonstrated, knowledge about the numerous communicative intentions which older chi Idren successful ly employ is fragmentary. Several investigators have used more comprehensive classification schemes to collect the broadest data base possible while others have focused on singular examples of specific speech act types. In the majority of investigations, samples have been small with considerable variety and overlap from age to age; nevertheless, some developmental trends are beginning to emerge. Development of Language Use in LanguageI mpa i red Children Primary Forms During the first nine months of life, normal infants use their voices and their bodies to attract attention, express emotion and to engage in social exchanges with familiar adults. Children at high risk

PAGE 77

67 for later language impairment do not demonstrate the same facility with the nonverbal communication system. Fisichelli and Karelitz (1963) observed abnormal crying patterns in brain-damaged infants during the first week of life. In interviews with mothers of autistic children, Schaeffer (1970 recorded almost a complete absence of crying in the early months of life. Mothers consistently revealed that their autistic infants never cried for attention nor when they were hungry and their responsiveness to pain and discomfort was impaired. Ricks (1975) investigated the way in which autistic children and retarded children expressed emotional meaning in four situations: (a) requesting; (b) frustration; (c) greeting; and (d) pleasant surprise. The parents of the autistic children were asked to listen to the recordings of their own child, two other autistic children and one retarded child and identify the meaning of the messages conveyed. Although the parents accurately identified the sounds made by the retarded child, who they said sounded "normal," they could not identify the sounds made by the other two autistic children. Ricks (1975) concluded that autistic children were capable of expressing the four emotions but they did so in a personal idiosyncratic way dissimilar from that of normal babies. Whereas normal babies used intoned vocalizations in response to the situations, autistic children's sounds were articulated. Kinesic differences in the nonverbal communication systems of children at high risk for language impairment have also been noted. Schmidt and Erikson (1973) interviewed mothers of retarded children, 16 percent of whom reported delayed smiling responses in their infants. Gaze aversion was a common phenomenon observed by mothers of autistic

PAGE 78

68 children--a cha racter i s t i c wh i ch interrupts substantially normal rnotherchild interactions (Schaeffer, 1971). Facial expression, hand movements and body postures also were slow to evolve in these children. Ricks and Wing (1975) noted that autistic children pass through a concrete demonstration phase (e.g., pushing mother's hand to the lightswitch to get her to turn on the light) on the slow and imperfect path toward acquisition of symbolic pointing. In fact, though 50 percent of autistic children remain mute all their lives, they do not use gestures as a substitute for speech. Conventional Forms In the stage of conventional forms, normal children use sounds and gestures in a variety of ways to communicate a message. Snyder (1975) studied the pragmatic performance of 15 language-normal and 15 languageImpai red children at the one-word stage of development. The children were matched for socioeconomic level and utterance length and were screened for cognitive development. The language-normal subjects had a mean age of lA.9 months, while the language-impaired subjects had a mean age of 2k. 2 months. Three experimental measures were developed to elicit presupposi tional , declarative performative and imperative performative responses. The analysis of the data revealed that the older languageimpaired children, while at the same stage of development in terms of utterance length and lexicon, were deficient in their use of language. On the presupposi tional measures, the languageimpai red children could encode the most informative element in a context almost as often as they encoded a more imformative one. On both the declarative and

PAGE 79

69 imperative performative measures, the 1 anguageimpai red children performed more poorly than the language-normal children, and the languageimpaired children generated significantly fewer verbal and nonverbal performatives in response to both types of measures, with nonverbal responses predominating over verbal ones. Snyder (1975) offered her findings as support for "a specific representational deficit in the language-disabled child which affects the dynamic aspects of symbol ization" (p. 167). Snyder's (1975) study bridges the transition between unconventional and conventional language use. With regard to the use of unconventional forms, she found that 1 anguagei mpa i red children used nonverbal performatives, both declarative and imperative, more often than verbal performatives. These children preferred grasping, reaching, smiling, etc., to linguistic symbols to get the examiner's attention and they substituted looking, looking and reaching, looking and fussing, etc., for words to get the examiner to perform a desired action. Even though the number of nonverbal imperatives was significantly greater for the languageimpai red children than the normal children, the level of these nonverbal imperatives was significantly lower. In the use of conventional forms, the 1 anguagei mpa i red children had difficulty generating verbal performatives. In fact, the differences between the performatives generated by the two groups approached the greatest significance when the performatives compared were linguistic. Geller and Wollner (1976) investigated language use in a group of three older 1 anguagei mpa i red children, ages 3.11 to 5 years, who were functioning at a mean length of utterance level from 1.1 to 1.6. The authors videotaped the children during snacktime in a typical

PAGE 80

70 preschool setting and analyzed their communication skills in a variety of ways. The most significant analysis for the purposes of the present study was a classification of the types of communication intentions conveyed by the children's verbal acts. When the results of this study were compared to those of Dore (1975), collected for 3-year-olds in a similar communicative context, it appeared that the 1 anguagei mpai red children were deficient in language structure as well as language use. The findings of Snyder (1975) and Geller and Wollner (1976) suggest that some languageimpai red children "may be more deficient in the use of language for communication than even their limited mastery of vocabulary and syntactic structures would allow" (Rees, 1978, p. 258). Other studies do not share this observation about the language use of languageimpai red children. Owens (1978) devised a modification of Dore's (197^) primitive speech act taxonomy to evaluate the language use of twelve children, six with Down's syndrome and six nondelayed. Three of the children in each group were performing at Brown's stage I (MLU = 1.0 to 2.0) and three at Brown's stage II (MLU = 2.5 to 3-0) with regard to linguistic competence. Audio and video recordings were made of the children in a freeplay situation with their mothers in the home in order to obtain 100 intelligible child utterances for analysis. A comparison of speech act distribution anx5ng the four groups revealed essentially the same patterns. Only two speech acts, declaration and practice, achieved significant group differences. Two studies have investigated the young hearing impaired child's ability to communicate, either verbally or nonverbally, a variety of pragmatic intentions (Curtiss, Prutting S Lowell, 1979; Skarakis 6

PAGE 81

71 Prutting, 1977). Communication samples were videotaped during four different s i tuat ions--f reepl ay , snacktime, group lesson and individual lesson--and were analyzed using a modification of Dore's (197A, 1975) taxonomy to include gestural behavior. Although Skarakis and Prutting (1977) found all communicative intentions expressed by all subjects, with requests/demands, descriptions, attention and responses occurring most frequently for all children, the realization of these intentions was carried out primarily through nonverbal means. These nonverbal behaviors are similar to those used by younger normal -hea ri ng subjects and have been found to be precursors to later linguistic development. The authors concluded that hear i ngimpa i red children expressed the same communicative intentions as normal-hearing children and demonstrated that they had acquired the basic foundation on which later language develops. Curtiss et al. (1979) investigated a larger sample of hearingimpaired children who spanned a wider range in age, 22 to 60 months. They found that, overall, hear i ngimpa i red children coded a variety of communicative intentions using both verbal and nonverbal means. All age groups exhibited all communicative intentions, but the distribution of specific category types varied with age. It was interesting to note that the number of communicative intentions expressed increased with age, but the mean length of utterance remained the same. This finding seems to indicate that pragmatic development precedes development in the other components of language and provides the foundation upon which these later linguistic skills are constructed. Only a few investigations of the communicative intentions expressed by languageimpa i red children have been undertaken. The

PAGE 82

findings have been scattered over a wide range of ages and developmental levels and have been insufficient in determining the true nature of language use in these children. The results of some of the studies are suggestive of a delay in language use among the languageimpai red. The same features of language use have been observed in their communications but they occur less frequently and at a later age. Other studies are more indicative of disordered language use. Snyder (1975), for example, found differences in language use among her subjects who had been matched for general linguistic level. Language Use in the Culturally Diverse A child entering school is well on the way to mastering the essential vocal symbols and the complex grammatical system of English. The typical first grade child already knows the language of home, neighborhood and community. The child is already generating sentences effortlessly and spontaneously through unconscious use of grammatical "rules" induced from language as acquired and developed, but does not yet have skillful control of language use in a wide variety of everyday interactions (Hansen, \37k) . Children acquiring language can only acquire that language to which they have been exposed. The child learning language must acquire the rules of pronunciation, grammar and usage. The rules of usage vary from culture to culture, but they will be learned in the immediate family during the preschool and elementary years (Brown, Ecroyd, Hopper S Naremore, 1976). Even though children from different environments may have access to the same set of language forms, they will employ these forms differently in communication situations. In other

PAGE 83

73 words, different groups of children might be said to show differences in language use — that is, in their knowledge of the rules for what is appropriate communication in a given situation. Language use is derived from one's communication experience, and this experience is, of course, shaped by the environment (Naremore, 1976, p. 23). For the majority of children, family communication patterns coincide with those of larger units of society, so what the child learns at home is reinforced by contacts outside the home, especially at school , For other children, communication roles and norms and various aspects of that code learned in the family settings are different from those of the larger culture. These latter children are likely to encounter communication difficulties in school . (Brown , Ecroyd, Hopper £ Naremore, 197^, p. 155) Although it is apparent that the home environment in which a child is reared is relevant to emerging speech, language and communication skills, it has been difficult to find a valid system that allows for rigorous identification of discrete home subcultures in order to systematically examine their influence on communication development (Adler, 1979). The majority of studies in this area have merely related socioeconomic status (SES) to language and communication development. In 1966, Cazden reviewed a series of studies that had, over the years, counted everything from vocabulary size to transformations and concluded that "on all measures, in all studies, children of upper socioeconomic status, however defined, are more advanced than the lower socioeconomic children" (p. I9I). The literature on language use is not as clear-cut. Whereas some investigators still attribute the communicative advantage to children from higher socioeconomic levels, others do not. Children from lower socioeconomic levels are capable of using

PAGE 84

language for the same purposes as children from higher socioeconomic levels, but they do so less frequently, in part because they misinterpret the communicative demands associated with the situation. McCarthy (1930) analyzed the language use of preschool children as it occurred in samples of running conversation using a modified version of Piaget's (1923) functional classification scheme. She found that adapted information and questions occurred with greater frequency at all ages among children of higher socioeconomic levels. These differences in language use persisted even when mental age was held constant . Davis (1937) also used the Piaget-McCarthy functional classification framework to analyze the language use of older children, ages 5V2 to 3^/2. Overall, the children from the higher socioeconomic levels made more spontaneous remarks than the children from the lower socioeconomic levels, but the discrepancy decreased with age. In addition, the higher socioeconomic status children asked more questions than the lower socioeconomic status children, with the difference disappearing by 9V2 years. This finding was comparable to that observed in the younger children studied by McCarthy (1930). Two other trends emerged from the data on social class differences. First, the naming category appeared to be used with greater frequency by the children from the lower socioeconomic levels. Second, the percentage of answers was higher for the lower socioeconomic groups although the differences decreased with age. Davis (1937) attributed these results to differences in the socialization processes associated with the two strata of society.

PAGE 85

75 Schachter et al . (197^) also provided data on the qualitative differences in everyday speech usage as they occurred naturally in the preschool setting for advantaged and disadvantaged black and white children. The results indicated significant sociol inguistic differences for only two scores, modulations and asserting desires to adults. Both advantaged groups (black and white) showed higher modulation scores than both disadvantaged black groups (higher and lower IQ). The authors suggested that this finding was consistent with Hess's (1969) research formulated in the framework of Bernstein's (1962, 1965) sociol inguistic theories . Hess (1969) found that lower-class black mothers were more apt to use imperative-normative control strategies with their children, while middle-class blacks use more cogni t i verat ional and persona 1 subjective control strategies. Imperative-normative strategies involved orders, accompanied by appeals to existing norms, e.g., "Do it because I told you to do it." Cogn i t i verat ional strategies involved appeals to reason, e.g., "Do it because it may fall," and personal -subject! ve strategies involved appeals to reason of an emotional nature, e.g., "Do it so you won't hurt her feelings." The latter two strategies seem to require much greater use of Modulation--explanation, justification, etc. The present data suggest that the social class differences in the use of verbal modulations are evident as early as the preschool years . In asserting desires to adults, both advantaged groups scored consistently higher than the lower-IQ disadvantaged blacks but showed no significant difference relative to the l\igher-IQ disadvantaged group. These results appeared consistent with White's (1972) findings on the

PAGE 86

76 competence of young children. His less competent preschool children showed less instrumental dependence on adults in their overall social behavior just as the disadvantaged children with lower IQs showed less instrumental dependence in their social speech. These data suggest that the preschool children who are least likely to turn to adults for help in fulfilling their desires are most vulnerable to school problems. In a longitudinal study of language function, Tough (1977) compared the language used by advantaged and disadvantaged children at the ages of 3, 5V2 and 1^/^. Analysis of linguistic form revealed that the 3-year-old children in the advantaged group were producing longer and more complex utterances with greater frequency on all features measured than their disadvantaged peers. These differences were demonstrated at a high level of significance and supported other research findings in showing that there are differences in the language forms used by children from lower and higher socioeconomic groups. Tough's (1977) data indicated that these differences are already apparent in the language of children by the age of 3 and are maintained in their language at 5V2 and 7 years. One of the most significant findings of the linguistic analyses was "that the disadvantaged children had greater resources of language than their typical performances revealed" (Tough, 1977, p. 169). Although the mean scores for the disadvantaged children tended to be lower on all measures, the range of scores was not necessarily similarly restricted. All disadvantaged children produced long utterances, elaborated noun and verb phrases and a greater number of nouns than pronouns in certain communication contexts. The generally lower scores produced by the disadvantaged children could not be explained by a failure to develop and use complex language, but rather by a

PAGE 87

77 difference in general orientation to language use which also was reflected in language form. The disadvantaged children lacked the appropriate expectations for certain communication situations and thus were not readily disposed "to search for or recognize information that they held as appropriate for answering the questions put to them" (Tough, 1977, p. 167). A functional analysis of the language used by advantaged and disadvantaged children showed that by the age of 3, differences existed in the kinds of meaning that they were imposing on their experiences through language, although there was little difference in total output of language. The disadvantaged children tended to limit language use to monitoring the ongoing situation and to maintaining status in relation to others. The advantaged children used language more frequently to: (a) recall and give detail to past experiences; (b) reason about past and present experiences; (c) anticipate future events and predict outcomes; (d) survey alternative courses of action; (3) project into the feelings and experiences of others; and (f) create imagined scenes through language. These differences, first apparent in the unstructured conversation between 3-year-olds, remained unresolved in the talk of sVaand 7V2-year-olds in response to structured interviews. In all situations, the disadvantaged groups used language reflective of less complex levels of thinking than that of their advantaged peers. Through their use of language, the disadvantaged groups demonstrated that they were less aware of alternative interpretations of situations and were less inclined to project beyond the immediate requirements of a task. Tough (1977) concluded that in all probability, the limited responses of the

PAGE 88

disadvantaged children originated from attitudes which prevented them from recognizing as appropriate or relevant their extended thinking or expression of awareness. She offered as evidence for her conclusions the fact that frequently when probed for additional information, the disadvantaged group gave more complete responses which resembled the kinds of responses produced spontaneously by the advantaged groups. This finding seems to support Bernstein's (1971) observations on the relationship of early social experiences and language use. He noted how different social contexts, which provide the early experiences through which children develop language, result in different orientations toward social relationships and the use of language. This difference in orientation toward language use has been observed in infants as young as 18 months. Howe (in Bruner, 1974) has found that middle-class mothers assume an instructive role toward their infants, both by responding more to their infants' efforts to vocalize by speaking in return and by attempting nnore often to initiate exchanges. The working-class mothers in her study were more often laissez-faire in their approach. In a partial replication of Tough's (1977) approach, Wells (1978) could find no clear-cut relationship between language use and either social class or educational success after one year of schooling. Wells (1978) attributed the discrepancies between the results to (a) the distortion introduced into Tough's study by the comparison of polarized social-class groups, and (b) the noninteractive conception of communication that underlies her analysis of language use. (p. 9) In Wells' (1978) study, children represented four classes of family backgrounds which spanned the full spectrum of social class. When the

PAGE 89

79 full spectrum of family background was considered, differences in language use were far less clear-cut and there were few simple linear trends over the four social classes. In fact, in some categories of language use, the trend was in the opposite direction from that predicted by Tough. Combined scores on two tests of reading at the end of the first year of school were used to determine the extent of the relationship between language use, family background and a measure of reading achievement. Significant correlations were found between: language use and reading; language use and social class; or reading and social class. When the contribution of social class to the correlation between language use and reading was partialed out, the correlation was reduced to .19, which is not significant. On the other hand, even when the effect of language use was partialed out, the correlation between social class and reading was .47, which is still significant (£ < .05). These results suggest that . . . whilst there is a significant relationship between social class and both language use and educational success (as measured by reading attainment after one year) it is not the differential use of language for the purposes identified by Tough that is the main mediator between home background and school success, (p. 17) Two studies of a more experimental nature were undertaken by Edwards (1977) and Bruck and Tucker (197^) to examine the influence of social class on language use. Edwards (1977) investigated the effect of context on the linguistic form and function of ten advantaged and ten disadvantaged ll-year-old children. An interview technique was used to stimulate conversation which could be classified as narrative or explanation and the resulting language samples were analyzed using linguistic indices of form and function. On traditional measures of language form.

PAGE 90

no significant group differences emerged. There were significant differences on measures directly related to language function or the "planning principles" underlying Bernstein's (1962, 196^) restricted and elaborated codes, e.g., range of adjectival modification and explicitness of pronominal reference. In Bernstein's (1962, 196A) early code definitions, a low proportion of exophoric references and a high ratio of adjectives to nouns were both reflective of elaborated speech and contributed to explicit unambiguous communication. In the present investigation, disadvantaged children used a narrower range of adjectives and extended modifications and had a higher proportion of exophoric references, suggesting the use of an informal, personalized mode of communication. Task differences were significant for several of the measures in response to the narrative and explanatory modes of communication. In explanations, children used more subordination than in narratives. In addition, explanations elicited a lower proportion of simple tenses, a higher proportion of modals, and a restricted ratio of nouns to verbs, indicative of nnore abstract communication. A higher ratio of nouns to pronouns was also observed during explanations and could be attributed to the restriction of referents in this part of the i nterview. In the present study, disadvantaged children did not exhibit the rigidity in speech suggested by some previous research. The main generalizations arising out of the data were that task differences accounted for greater and more frequent differences than social class differences and that the variation in scores did not suggest a persistent orientation to either the restricted or elaborated code. These

PAGE 91

81 results offer some support for Cazden's (1970) plea to consider the important influence of contextual factors on communication. Bruck and Tucker (IS?'*) investigated social class differences in the acquisition of school language by twenty middleand twenty lower-class kindergarten children. A pretest-posttest design was employed to measure changes in their grammatical and communication skills during the first year of formal schooling, including: imitation of syntactic structures, grammatical comprehension, production of whquestions, grammatical sentence completion, story-telling, description of abstract designs, vocabulary naming and vocabulary classification. These tasks were analyzed according to 26 linguistic and communication variables and the data were submitted to a factor analysis. The results of the factor analysis suggested that "communication and grammatical abilities represent statistically independent skills in kindergarten children" (Bruck & Tucker, 197^, p. 216). Two factors emerged which contributed to grammatical abilities: comprehension of classroom English and productive knowledge of classroom grammar. Four factors emerged which contributed to communication abilities and coincided with Hymes' (1971) definitions of communication competence: speech output (knowing when to speak), communication of relevant content (knowing what to say), ambiguous and elaborated speech, and egocentric information (knowing how to say it). Lower-class children did not perform consistently more poorly than middle-class children on all measures, but they did experience difficulty in three main areas. First, lower-class children have the same ability as middle-class children to comprehend grammatical structures, but have more difficulty producing them. . . . Second, the

PAGE 92

82 lower-class children's speech contained fewer explicit features than the middle-class children's although it contained as many implicit features. . . , Third, lower-class children have particular difficulty evaluating the communicative demands of the classroom. (Bruck S Tucker, 197^*, pp. 216-217) Bruck and Tucker (197^) concluded that, Failure to give the expected information on demand may contribute initially to the lower-class child's failure in school. . . . The child knows the relevant information but does not communicate it. This may occur because he views the classroom situation as threatening and says nothing . . . or because he assumes that the teacher already knows the answer and it would be redundant for him to tell her what she already knows, (p. 217) Two general conclusions can be drawn about changes in language abilities during the first year of school. improvement in grammatical abilities was greater than improvement in communication abilities. Of particular importance, the lower-class children improved more rapidly over the year than the middle-class children. This general improvement was not found for the communication abilities. The lower-class children "did not seem to be catching up with their middle-class peers on tests of communication abilities to the same extent that they did on tests of grammatical abilities" (Bruck S Tucker, 197^, p. 218). Any gains in communication effectiveness appeared to be related to increased use of features which stress elaboration and explicitness and not to decreased use of ambiguous or egocentric features. In summary, Bruck and Tucker (197^) emphasized that "children need specific help in learning how to use speech more effectively in the classroom. They must learn what the communicative demands of the situation are and how to meet these demands" (Bruck 6 Tucker, 1974, p. 218). Unlike developmental studies of language use in language-normal and 1 anguageimpa i red speakers, most investigations of the culturally

PAGE 93

83 diverse focus on the school-aged child and the adequacy of his habitual mode of language use for success in the classroom. The results of these various investigations were not conclusive. Measurement of Language Use A review of existing measures of language use in children reveals only one standardized tool and a variety of nonstandardized approaches which are employed to evaluate isolated aspects of this broad area. One critical problem in the development of measurement strategies of language use has been the establishment of what Bronf enbrenner (in Cazden, Bond, Epstein, Matz 6 Savingnon, 1977) calls "ecological validity." Bronfenbrenner argues that "much of contemp>orary developmental psychology is the science of the strange behavior of children in strange situations with strange adults for the briefest possible periods of time" (p. 84). Developers of oral language testing strategies, in particular, have been guilty of violations of ecological va 1 i d i ty . Ideally, a comprehensive assessment of language use should be designed in accordance with sociol inguistic theory to sample, in an ecologically valid way, the function and context of communication situations. To a great extent, then, the development of ecologically valid communication testing situations for children depends upon the skillful simulation of real life encounters which maintain the motivational and interactional richness of familiar academic and social communication events.

PAGE 94

84 Standardized Measurement Strategies At the present time, there is no comprehensive battery which assesses the language use of children in everyday academic and social situations. The Preschool Language Assessment Instrument (PLAl) (Blank, Rose & Berlin, 1978) is the only available test which attempts to measure functional aspects of natural language not accounted for in traditional language evaluations. Designed to evaluate young children's skills in coping with the language demands of the teaching situation, the PLAl is based on Moffet's (1968) model of classroom discourse which includes three major components: the speaker1 istener dyad, the topic, and the level of discussion. The level of discussion places demands on the child that require varying levels of abstraction ordered along a continuum of perceptua 1 1 anguage distance. The continuum has been divided into four main levels including: (a) matching perception (e.g., "What is this called?"); (b) selective analysis of perception (e.g., "What is this part for?"); (c) reordering perception (e.g., "Look at these. How are they the same?"); (d) reasoning about perception (e.g., "If the lady wants to carry all these things into the next room at the same time, tell me what she could use."). The main problem with the PLAl is the category system on which it is based. The development of this tool seems to have been largely influenced by a desire to distinguish levels of cognitive ability rather than by strict adherence to any particular theory of language or communication development. As such, the instrument distinguishes between cognitive demands that require children to apply language to salient perceptions at a low level of abstraction, and demands that require children to restructure

PAGE 95

85 those perceptions at a higher level of abstraction. This focus on cognition rather than communication forces the inclusion of certain items in the battery that are clearly not of an interactive nature, e.g., matching pictures, following directions and sentence completion. Nonstandardized Measurement Strategies The remainder of the strategies to be reviewed are all nonstandardized measurement techniques designed to answer specific experimental questions about the nature of communication development. These strategies can be grouped into one of three types: interview, roleplaying and natural language sampling. Several investigators have used the interview type format of the PLAI with a wide variety of stimulus materials. Because looking at pictures and books is a common activity of preand school-aged children. Tough (1976) developed a picture description task for the appraisal of seven broad categories of language use: (a) self-maintaining, (b) directing, (c) reporting, (d) logical reasoning, (ei) predicting, (f) projecting and (g) imagining. Ritti (1978) and Fogel (1976) both employed an interview strategy in which "real life" communication situations were presented through pictures and verbal or written probes. In Ritti 's (1978) study, children were presented story situations to which they were expected to write something which they would say in that situation and to indicate which of ten possible responses provided by the examiner they would or would not make in such a situation. Responses were developed using five social speech functions from the work of Soskin and John (I963) including information, directives, inductives, evaluations and express i ves .

PAGE 96

86 Fogel (1976) designed pictorial and verbal stimuli to elicit comments from children concerning their own or a significant other's communication behavior (i.e., mother, teacher or peer). He selected three relevant attributes of the communication situation around which the stimuli were designed: (a) the function of communication (social or emotional), (b) the significant other within the situation (mother, teacher or peer), and (c) the perspective taken within the communication situation (self or other). For example, one stimulus item, a picture of a teacher who was trying to explain why a child could not have a book to read, was accompanied by the verbal probe "What would she say?" Ricillo (in Larson, Backlund, Redmond & Barbour, 1978) also used an interview format to examine children's ability to use seven functions of communication: contactive, conversative, descriptive, directive, explanatory, narrative and persuasive. In her study, a series of preestabl i shed probes were presented to children unaccompanied by other stimulus material. Lucas (1980) developed the Behavioral Inventory of Speech Act Performance (BISAP), a cri ter ionreferenced tool, to evaluate the speech act production of children, ages 3 to 5. BISAP assesses the production of eight commonly used speech acts including requests for objects, requests for action, assertions, denials, statements of information, requests for information, calling or summons, and rule orders in a familiar context of the examiner's design. Lucas (1980) has used art activities, preschool academic tasks and i nfant-to-careg i ver tasks, depending upon the background experiences of the children. Throughout the activity, the examiner presents a series of probes, comments and questions to elicit the desired speech acts. For example, to get the

PAGE 97

children to make assertions, the examiner asi
PAGE 98

88 that although certain aspects of role plays are not representative of spontaneous verbal exchanges, the techniques used in role plays also occur in everyday interactions. Some investigators (Bloom £ Lahey, 1978; Dore, 1977; Epstein et al., 1976; Hanes, 1978; Schachter et al., 197^) maintain a purist approach to language and communication sampling by collecting data through naturalistic observation. While this strategy represents the ultimate degree of ecological validity, it is not always the most effective, efficient, or practical strategy to employ. Limber (1976) raises an important issue about using production data alone to infer the extent of linguistic i
PAGE 99

89 Hanes (1978) remarked about the considerable time and expense required for testing, transcribing, coding and data analysis. The Literature in Retrospect With the recent shift in perspective from psychol ingu i st i c to sociol ingui st i c emphases in the study of communication development, a number of investigators, from a wide variety of disciplines, have begun to make significant contributions to the body of knowledge about how children use language in everyday situations. As with most literature on communication development, the initial focus of these studies has been the young child, from 0 to 3. Little effort has been directed toward dn evaluation of school-aged children's skills in this area, even though it has long been suggested that a significant link may exist between skillful language use and success in school. Children whose predominant mode of language use conflicts with that of the school are in serious trouble. The likelihood that they will achi eve their maximum potential in the classroom is doubtful, for language serves not only as an important subject of instruction, but also as the primary channel through which that instruction is carried out. It is clear that some children bring many of their disadvantages from home. . . . But it is just as clear that the school, itself, through its failures to recognize the complexity of the language problem, creates many more disadvantaged children. (Spolsky, 1971 , p. 15) In order to circumvent some of these avoidable classroom failures, measurement strategies appropriate for assessment of the various uses to which language is put in school must be developed so that children who may be functioning at less than adequate levels can be identified and helped. Children who are language-impaired relative to the

PAGE 100

linguistic aspects of content and form and children who are languagedifferent with regard to these same components may be among those whose use of language does not serve them adequately in the classroom. It is apparent that a comprehensive system for the measurement of language in the classroom is not currently available, although a number of guidelines exist for the development of such a system. This system should reflect sociol ingui st i c trends and be capable of identifying children's mastery of a set of communication strategies appropriate for a wide variety of academic situations. It should be designed specifically to deal with the special features of communication in the classroom, above all incorporating some means of differentiation within the large class of cognitive communication strategies. Because a number of theorists have described the "speech act" or "communication act" as the pivotal unit of language in communication, such acts would appear to be a logical organizing principle for the system just prescribed. Austin (1962) and Searle (I969) note that common, recurring routines such as commanding, requesting, promising and arguing are the basic units of human interaction. "The speech act approach examines human purposes as they occur naturally when people talk to each other" (Wood, 1976). In the present investigation, the Functional Inventory of Cognitive Communication Strategies was designed to identify the wide variety of language used and strategies reflective of the young child's growing ability to use the language system to convey meaning. The framework for the classification of the cognitive uses of language, developed by Tough (1976, 1977) served as the basis for the creation of the stimulus items on FICCS. In Tough's (1976) system, the basic

PAGE 101

91 units of communication called "speech acts" or in this instance "thought acts" are subsumed under four major functions of communication: Reporting, Logical Reasoning, Predicting and Projecting. Using a structured interview, children's habitual strategies of language use are elicited in response to a series of pictures which illustrate the storyline of The Black Kitten Gets Lost . It is believed that this technique most closely simulates the interaction that occurs between teacher and child in the classroom setting. Because questions and answers account for approximately percent of the exchanges that take place in the classroom, this format would appear to result in the most ecologically valid indication of how children will fare when confronted with the communication demands of the classroom.

PAGE 102

CHAPTER 3 METHODS AND PROCEDURES Subjects The four groups of subjects in this study included one group of languageimpai red children and three groups of language-normal children. The groups were selected as follows. LanguageImpa i red Group (Li) Ten 1 anguageimpai red children, five at each of two age levels, 6 years (plus or minus 3 months) and 7 years (plus or minus 3 months) were selected after screening 37 language-impaired children relative to several selection criteria. At the time of testing, all subjects had been enrolled in an elementary school program in either Wilkes or Avery counties in North Carolina for a period of not less than four weeks and were receiving speech and language therapy. Subjects were identified as languageimpai red on the basis of depressed performance (^17th percentile) on the Bankson Language Screening Test (Bankson, 1977) and the clinical judgment of a certified speech, language and hearing clinician. A child was regarded as 1 anguageimpa i red if he demonstrated a disruption of uncertain origin within or among the language components of content, form and use, resulting in reduced comprehension and production of linguistic messages. This primary linguistic deficit occurred in children who did not otherwise evidence deficits of cognitive (mental 92

PAGE 103

93 retardation), sensory (hearing impairment), psychosocial (autism) and/or neurologic (aphasia) function or differences with respect to bilingual or bicultural language systems. To avoid the possible influence of certain variables which might affect linguistic and communicative performance, all subjects were to have been identified as Caucasian and middle-class. Although it was possible to control for racial differences, it was not possible to control for the influence of social class on linguistic and communicative performance. All language-impai red children used in this study came from families of lower-class socioeconomic status (SES) as determined by Hoi 1 ingshead ' s (1957) two-factor index of social position, a measure which operationally defines social position in terms of occupation and education of the head of the household. The social class index of children in the languageimpa i red group ranged from kk to 69, placing them either in Class V of VI or Holl i ngshead ' s six-class scale. Tables 3 and A present a descriptive summary of pertinent subject characteristics for the languageimpai red group. Language-Normal Groups Two language-normal control groups were employed. Leonard, Bolders and Miller (1976), in a study of semantic relations used by language-normal and 1 anguageimpa i red children, demonstrated the efficacy of making comparisons under both matched age and matched utterance length conditions. Comparisons by matched age showed how the language sl
PAGE 104

94 Table 3 Subject Characteristics for LanguageImpaired (LI) 6-Year-Olds SUBJECT AGE IN MONTHS SEX SOCIOECONOMIC INDEX CLASS ADS CMMS STANINE BLST RAW Z 1 LE MLU-M 1 . 75 M 51 IV 89 4 88 2 4.583 2. 75 F 65 V 89 4 88 2 6.524 3. 73 F 69 V 89 4 72 4 5.584 f 75 M 51 IV 99 5 69 2 5.026 5. 74 M 58 IV 99 5 92 6 5.235 RANGE 73-75 51-69 89-99 69-92 4.5836.524 MEAN 74.4 58.8 93 81.8 5.390 CMMS = Col umbia Mental Matur i ty Scale ADS = Age Devi a ion Score BLST = Bankson Language Screening Test MLU-M = Mean Length of Utterance

PAGE 105

95 Table k Subject Characteristics for LanguageImpaired (LI) 7-Year-Olds SUBJECT AGE IN MONTHS SEX SOCIOECONOMIC INDEX CLASS CMMS ADS STAN 1 NE BLST RAW %\LE MLU-M 6. 87 M 69 V 108 6 96 5 5.183 7. 83 M k8 IV 89 93 3 5.192 8. 82 F 51 IV 90 k 96 17 6.085 9. 81 F 58 IV 89 88 6 4.838 10. 81 M 58 IV 96 5 87 8 3.882 RANGE 81-87 A^-69 89-108 8A-96 3.8826.085 MEAN 82.8 56 Sk.k 91.2 4.956 CMMS = Co 1 umb i a Mental Maturity Scale ADS = Age Deviation Score BLST = Banl
PAGE 106

96 the same general stage of linguistic development. Differences in language skill observed between the groups matched for age but not between those matched for utterance length suggest that the nature of the difference could best be described as a general developmental delay in language acquisition. Conversely, differences in language skill observed between the groups matched for age as well as those matched for utterance length suggest that the obtained differences are better characterized as deviant. The investigators (Leonard et al., 1976) concluded that the employment of both matching strategies in comparative studies of children's syntax should be useful in resolving the question of delayed versus deviant structural development. It follows that a similar research design should clarify the nature of an impairment in the selection and use of specific communicative strategies. Language-Normal Group 1 (LNl) . The first language-normal group included ten lower SES, Caucasian children, five at each of two age levels, 6 years (plus or minus 3 months) and 7 years (plus or minus 3 months). All subjects were selected after screening A6 lower SES elementary school children enrolled in kindergarten and first grade in Avery and Wilkes counties in North Carolina according to several selection criteria. At the time of testing, all subjects had been enrolled, for a period of not less than four weeks, in a public school setting and none were currently receiving speech and language therapy. All subjects were judged language-normal on the basis of performance on the Bankson Language Screening Test (>70th percentile) (Bankson, 1977) and clinical observation. Other criteria for selection included:

PAGE 107

97 1. Normal intelligence as measured by the performance on the Columbia Mental Maturity Scale (CMMS) (Burgemei ster , Blum & Lorge, 1972). 2. Native speakers of English from monolingual homes as determined from school records. 3Absence of gross peripheral defects of audition or vision as determined by teacher interview. No previous history of speech, language and/or hearing problems as determined by school records and teacher interview. Tables 5 and 6 present a descriptive summary of pertinent subject characteristics for Language-Normal Group I. Language-Normal Group 2 (LN2). The second language-normal group included ten lower SES, Caucasian children matched on the basis of general language level with the 1 anguageimpai red group. General language level was determined by calculating the mean number of morphemes (adapted from Brown, 1973) per communication unit (Loban, I963, 1976). Subjects were selected after screening 37 lower SES children relative to several selection criteria. These children had been enrolled, for a period of not less than four weeks, in a preschool setting in Avery and Wilkes counties in North Carolina. At the time of testing, none were enrolled in speech and language therapy. Subjects were judged languagenormal on the basis of performance on the Bankson Language Screening Test (>70th percentile) (Bankson, 1977) and clinical observation. Other criteria for inclusion were the same as those for LNl. Table 7 presents a descriptive summary of pertinent subject characteristics for Language-Normal Group 2.

PAGE 108

98 Table 5 I Subject Characteristics for Lower Socioeconomic Language-Normal (LNl) 6-Year-Olds SUBJECT AGE IN MONTHS SEX SOCIOECONOMIC INDEX CLASS CMMS STAADS NINE BLST RAW %\LE MLU-M 1 1 . 75 M 58 IV 93 k 131 7k 5.550 12. 71 F 51 IV 107 6 119 7^4 5.70'» 13. 72 F kk IV 97 5 13^ 93 7.826 ]k. 7k M 58 IV 108 6 130 70 7. 161 15. 69 M 58 IV 92 k 119 7k 5.^490 RANGE 69-75 kk-S8 92-108 119-13^ A.5907.826 MEAN 72.2 53.8 33. k 126.6 6.3't6 CMMS = Co 1 umb i a Men ta 1 Maturi ty Scale ADS = Age Deviation Score BLST = Bankson Language Screening Test MLU-M = Mean Length of Utterance

PAGE 109

99 Table 6 Subject Characteristics for Lower Socioeconomic Language-Normal (LNI) 7-Year-Olds SUBJECT AGE IN MONTHS SEX SOCIOECONOMIC INDEX CLASS CMMS STAADS NINE BLST RAW ^ILE MLU-M 16. 87 M kk IV 89 i» 135 72 7.201 17. 81 F 51 IV 101 5 135 81 8.302 18. 87 M 58 IV 98 5 136 79 9. 176 19. 81 F 51 IV 102 5 136 86 6.061 20. 8k M IV 108 6 133 75 8.625 RANGE 81-87 kk-5S 89-108 133-136 6.0619.176 MEAN 8^4 99.6 135 7.873 CMMS = Col umbia Mental Maturity Scale ADS = Age Deviation Score BLST = Bankson Language Screening Test MLU-M = Mean Length of Utterance

PAGE 110

100 Table 7 Subject Characteristics for Language-Normals (LN2) Matched for Utterance Length SUBJECT AGE IN MONTHS SEX SOCIOECONOMIC INDEX CLASS CMMS STAADS NINE BLST RAW ^ILE MLU-M 21. 46 F 51 IV 105 6 89 85 4.552 22. 55 F 55 1 v 98 5 99 85 6.483 23. 52 F 47 IV 105 6 76 76 5.500 24. 48 F 62 V 110 6 76 76 5.315 25. 62 M 58 IV 104 6 87 87 5.357 26. 62 F 73 V 101 5 70 70 5.350 27. 50 F 62 V 103 5 76 76 5.589 28. 63 F 44 IV 102 5 75 75 6.024 29. 51 M 58 IV 97 5 70 70 4.232 30. 51 F 58 IV 89 4 70 70 3.831 RANGE 46-63 44-73 89-1 10 82-121 3.8316.483 MEAN 54 56.8 101 .4 99.8 5.223 CMMS = Columbia Mental Maturity Scale ADS = Age Deviation Score BLST = Bankson Language Screening Test MLU-M = Mean Length of Utterance

PAGE 111

10] Language-Normal Group 3 (LN3) Because of the generally depressed economic level in the semi-rural area in which the study was conducted, all subjects previously identified came from families of lower socioeconomic status. These subjects represented the average economic level of the majority of inhabitants in this geographic area, as substantiated by the fact that approximately 70 percent of all children enrolled in the elementary schools included in this study were participants in the subsidized school lunch program (Burleston, personal communication). Eligibility for this program is determined by family size and income scales. Children from families whose income is at or below the levels indicated may receive free milk and meals or meals at a reduced price. Appendix A contains a reproduction of family size and income scales for Avery and Wilkes County Schools. Since many investigators (Bruck & Tucker, 197A; Cazden, 1970; Schachter, Kirschner, Klips, Friedricks & Sanders, 197^; Tough, 1976) have noted the interrelatedness of language and culture, a third Language-Normal Group (LN3) was included in the study to examine the Infl uence of social class on language use. Ten Caucasian, upper-middle SES, language-normal children, five at each of two age levels, 6 years (plus or minus 3 months) and 7 years (plus or minus 3 months), were selected after screening 20 upper-middle SES children relative to several selection criteria used for the establishment of LNl and LN2. The designation of upper-middle SES was made on the basis of Hoi lingshead's (1957) two-factor index of social position. Assignment to this social class category required a Social Index which ranged from 11 to 27, placing the children in Class I or II of Hoi 1 ingshead's six-class scale.

PAGE 112

102 Tables 8 and 9 present a descriptive summary of pertinent subject characteristics for Language-Normal Group 3. Mater i al s Classification of the Cognitive Uses of Language The framework for the classification of the cognitive uses of language developed by Tough (1976, 1977) served as the basis for the creation of the stimulus items on the Functional Inventory of Cognitive Communication Strategies (FICCS). Table 2 presents an outline of Tough's (1976) functional classification schema. Operational definitions and examples of the various communication strategies can be found In Appendix B. The Functional Inventory of Cognitive Communication Strategies (FICCS) In developing an inventory of the cognitive uses of language, it is necessary to identify the kinds of meanings that the child is able to express through the use of language. Tough's (1977) functional classification of language use permits the identification of a wide variety of language uses and strategies reflective of the child's growing ability to use the language system to convey meaning. As Tough (1976) notes: Language provides the medium through which thinking can be expressed and, . . . the very use of language and the continuous experience of being amongst users of language, influences not only the way in which the child will use language but more important, the way in which he will think, and the kind of interpretation he will make of his experiences. (p. 76) Tough (1976) has developed two sets of six pictures each. Black Kitten Gets Lost and Dad Forgets His Lunch , for the purpose of helping

PAGE 113

103 Table 8 Subject Characteristics for Higher Socioeconomic Language-Normal (LN3) 6Year-Olds SUBJECT AGE IN nUN 1 no C C V SOCIOECONOMIC INDEX CLASS CMMS ADS STAN 1 NE BLST RAW «ILE U 1 1 1 u MLU-M 31. 75 M 22 1 1 92 it 130 70 32. 75 F 22 1 1 108 6 131 7^ 7.685 33. 69 M 22 1 1 9^ li 128 88 6.198 3A. 69 F 22 1 1 100 5 126 80 5.673 35. 71 F 22 1 1 106 6 126 80 6.815 RANGE 69-75 92-108 126-131 5.6737.685 MEAN 71 .8 22 100 128.2 6.569 CMMS = Columbia Mental Maturity Scale ADS = Age Deviation Score BLST = Bankson Language Screening Test MLU-M = Mean Length of Utterance

PAGE 114

104 Table 9 Subject Characteristics for Higher Socioeconomic LanguageNormal (LN3) 7-Year-Olds SUBJECT AGE IN MONTHS SEX SOCIOECONOMIC INDEX CLASS CMMS ADS STANINE BLST RAW ^ILE MLU-M 36. 84 M 15 1 104 6 1 04 1 00 6.991 37. 81 F 22 1 1 106 6 140 92 8.048 38. 82 M 22 1 1 no 6 147 97 9.083 49. 82 M J 1 11 c 1 ill D . y 40. 91 M 22 1 1 c J ft iinc O . MUp RANGE 81-84 15-22 96-1 10 136-149 6.9249.083 MEAN 82 20.6 103.4 142.6 7.890 CMMS = = Co 1 umb i a Mental Maturity Scale ADS = Age Deviation Score BLST = = Banl
PAGE 115

105 teachers to make appraisals of children's language use. Her pictures were employed for the elicitation of language uses in this study so that the data obtained would be of the same denomination, and thus, easily usable, by other investigators embracing Tough's (1976) system (see Figure 2). The Black Kitten Gets Lost pictures were chosen for this study because of the sufficiently detailed nature of the subject matter and the familiarity of its content for young children. The pictures which illustrate the storyline of the Black Kitten Gets Lost are 7Vi»" by sVi*" black-and-white line drawings. Each drawing contains sufficient detail involving actions and incidents to provide children with the wide basis they need to demonstrate the full range of their thinking. The strong storyline depicted in each offers children evidence for making interpretations that carry forward from one picture to the next. Although the problems associated with collecting a representative sample of language use from static pictorial representations is recognized, it was nonetheless the task of choice in this study for a number of reasons. First, in the elicitation of oral language samples, pictures are the stimuli most frequently employed (BarrieBlackley, Musselwhite 6 Rogister, 1978). Second, looking at pictures and books is a common preand elementaryschool activity which children seem to enjoy (Tough, 1976). Third, the use of a standard set of stimulus pictures insures the comparability of responses within a single subject and across subject populations. And, finally, picture description is a task easily adaptable to a variety of sampling settings. A series of eight question pairs, designed to elicit a wide variety of cognitive communication strategies, was developed to accompany each picture. A complete list of these quest ions is included in

PAGE 116

106

PAGE 117

107 Appendix C. A similar set of questions also was developed to accompany the first picture in the story sequence Dad Forgets His Lunch for the purpose of modeling and practice in the elicitation protocol. A list of these questions is also included in Appendix C. Many of the questions used with each picture were selected from those suggested by Tough (1976). Tough (1976) subscribes to Piaget's (1923) "clinical method" of Interviewing children in which the interviewer poses a question, then induces the child to talk more spontaneously by following rather than leading the discussion. Since Piaget (1953) and Tough (1976) are most interested in the child's underlying thought and logic, oral language serves as a window into the child's cognitive processes (Barr ie-Blackley et al., 1978). Although the interest of this study is similar, the clinical method could not be adopted since it requires the interviewer to frame questions in response to the child's "lead" and thus, no two children necessarily receive the same questions or the same experimental treatment. Although the clinical value of such a strategy is without question, it does not readily lend itself to analysis by the statistical procedures necessary for testing the research hypotheses in this study. Therefore, a standard set of questions to be presented in a structured interview to each child was developed. Labov (1970) has criticized all adult efforts to elicit representative speech from children using a quest ion-and-answer format. The alternative which he provides, though ingenious, was not acceptable for this study. Labov (1970) asked a group of children to take care of his pet rabbit for a few minutes while he left the room, and to talk to it so that it would remain calm and unafraid. While he was gone, he left the

PAGE 118

108 tape recorder running to pick up the children's conversation. Labov's (1970) observations emphasize the important influence of certain situational parameters on children's conversation and underscore the necessity for collecting language samples in a variety of contexts if truly representative samples are to be obtained. The question-answer format was selected as the sampling strategy in this study for two reasons. Question-answer exchanges between teacher and child account for kO percent of all communication in the classroom, while child initiations account for less than 10 percent (Dunkin 6 Biddle, ]37^) . Although other sampling conditions might produce language which is structurally and conceptually more complex, they would not represent the type of classroom situation to which the child must adjust to achieve academic success. Secondly, the imposition of structure upon the sampling strategy permits determination of qualitative differences between any single child's responses and those of his peers or between normal and impaired language users. Procedures Collection of the Language Sample A language sample was obtained from each subject following the protocol in Appendix C. Collection of the samples occurred at the subject's school in a quiet room normally used for speech and language evaluations. During the sampling procedures, only the clinician and child were present in the room. The same clinician collected all samples. The language samples were recorded using a Panasonic, model RQ-320 S, portable tape recorder with a Lafayette, model ML-1,

PAGE 119

109 omnidirectional lavalier microphone. Total recording time ranged between 17 and 36 minutes. After a two-minute period of informal conversation, the clinician presented the following standard instructions and modeled appropriate responses to a set of practice stimuli: I'd like you to tell me some stories. I'm going to show you some pictures and I'd like you to make up a story for each picture. But first, let me show you how to do it. (Present Dad Forgets His Lunch , Picture #1.) Set 1: Suppose I asked you to tell me all about this picture. You might say, it looks like Mom has made some sandwiches for lunch and Dad has gone off and forgotten them. Set 2: Then suppose I asked you, what is the man doing here? You might say, he's going to work. And suppose I said, why is he doing that? You might say, so he can earn some money for his family. Set 3: And then suppose I said, how do you think Mora feels? You might say, she feels very sad cause Dad has forgotten his lunch. Following these activities, the clinician presented a set of three practice items to screen all subjects for the ability to perform the task as instructed and to allow for adaptation to the language sampling setting. To be included in the investigation, subjects were required to respond appropriately to two out of the three practice stimuli either as a result of the initial stimulus or the prompt. Prompts were supplied when a subject refused to respond after a tensecond pause following the examiner's stimulus, produced a response such as "I don't know," or emitted a response which was more than 50 percent unintelligible. Encouragement was offered throughout the child's response by using certain paral i ngui s t i c devices such as head nodding, smiling, and vocalizing "uh-huh."

PAGE 120

no Practice I terns Set 4: a . What do you think the boy is shouting? b. If you were the boy, what would you shout? Set 5: a . Can the boy make Dad hear him? Why do/don't you think so? b. Does Dad hear the boy? Why do/don't you think he hears him? Set 6: a. What do you think the children will do now? b. Will the children try to catch their Dad? How will they do it? To insure consistency in the method of obtaining the language sample, a series of eight "questions" were presented to each subject for each of six test pictures. Pictures are included in Appendix D. Prompts were offered for the same instances of unacceptable response detailed above and encouragement was freely given to enhance verbal productivity. At the conclusion of each set of eight "questions," a statement which summarized the central meaning displayed in each picture was presented by the clinician so that responses to subsequent items would not be influenced by unusual interpretations to previous questions. The "questions" are included in Appendix C. Transcription and Segmentation of the Language Sample same clinician, within a week of the recording date. Loban ' s (1963, 1976) segmentation protocol was used to identify all analyzable language units. An outline of Loban 's system may be found in Appendix E. All language samples were orthographical ly transcribed by the

PAGE 121

in Analysis of the Language Sample Based on the developmental level of the subjects, linguistic and functional measures, which would best reflect the linguistic and communicative performance of children at the stage of syntactic expansion and refinement (Hannah, 197^; McLean & McLean, 1978), were selected to analyze the language samples. Some of the critical aspects of children's advanced linguistic and communicative skills included verbal productivity, syntactic completeness, syntactic length, syntactic complexity and language use. To quantify these five areas, 37 measures were used to analyze each language sample. For a list of these measures, see Table 10. Verbal productivity . These measures evaluated quantitative variations in verbal output by examining the number of units produced in response to a predetermined number of stimuli in the sampling protocol . 1. Total Number of Communication Units (#CU's) . A communication unit may be defined semantical ly and structurally. According to Loban (1976), Watts (19^8) first described the communication unit in semantic terms as "... a group of words which cannot be further divided without the loss of their essential meaning" (p. 9). Because this definition proved to be insufficiently objective for reliable application, Loban (1963, 1976) redefined the communication unit in structural terms as an ". . . independent clause with all of its modifiers" (p. 9), Most communication units fit this structural description; however, because of the reciprocal nature of spoken language, Loban (1963, 1976) expanded the definition to include two

PAGE 122

1 12 Table 10 Language Sample Analysis Measures VERBAL PRODUCTIVITY 1. Total Number of Communication Units (#CU's) 2. Total Number of Words (TNW) 3. Total Number of Morphemes (TNM) SYNTACTIC COMPLETENESS k. Percentage of Complete Communication Units (^CUc) 5. Mean Length of Communication Units in Words (MLU-W) 6. Mean Length of Communication Units in Morphemes (MLU-M) 7. Mean Length of the Five Longest Communication Units (MLU5-M) SYNTACTIC COMPLEXITY 8. Developmental Sentence Score (DSS) LANGUAGE USE 9. Overall Functional Language Score 10. Reporting Subscore 11. Logical Reasoning Subscore 12. Predicting Subscore 13. Projecting Subscore ]k. Label ing (Str-1) 15. Referring to detail (Str-2) 16. Referring to incidents (Str-3) 17. Referring to the sequence of events (Str-4) 18. Making comparisons (Str-5) 19. Recognizing related aspects (Str-6) LANGUAGE USE (continued) 20. Extracting the central meaning (Str-7) 21. Reflecting on the meaning of experiences (Str-8) 22. Explaining a process {Str-9) 23. Recognizing causal and dependent relationships (Str-lO) 2^. Recognizing problems and sol utions (Str1 1 ) 25. Justifying judgments and actions (Str-12) 26. Reflecting on events and drawing conclusions (Str-13) 27. Recognizing principles (Str-14) 28. Anticipating/forecasting (Str-15) 29. Anticipating the detail of events (Str-l6) 30. Anticipating a sequence of events (Str-17) 31. Anticipating problems and possible solutions (Str-18) 32. Anticipating and recognizing alternative courses of action (Str-19) 33. Predicting the consequences of actions or events (Str-20) 3^*. Projecting into the experiences of others (Str-21) 35. Projecting into the feel i ngs of others (Str-22) 36. Projecting into the reactions of others (Str-23) 37. Projecting into situations never experienced (Str-24)

PAGE 123

113 other structural categories: (a) answers to questions which lack only the repetition of the question elements to satisfy the criterion of independent predication; and (b) answers of "yes" and "no" given in response to yes/no questions. For the purpose of the present investigation, a communication unit was defined either as an answer to a question or as a grammatically independent predication, together with any related subordinate clauses, which could not be further divided without the loss of essential meaning. These units were segmented and tallied following the protocol presented in Appendix E. 2. Total Number of Words (TNW) . The standards adapted from McCarthy (1930) and refined by Davis (1937) were used to determine how words should be counted. The total number of words produced in response to the six stimulus pictures was calculated and recorded. For a detailed account of these standards, see Appendix F. 3. Total Number of Morphemes (TNM) . Rules for counting morphemes have been established by Brown (1973). A description of these rules is presented in Appendix G. Syntactic completeness . This measure evaluated the structural integrity of the child's remarl
PAGE 124

developmental characteristic for younger children" (p. l^*). Others (Leonard, 1972; Shotick & Blue, 1971) have noted that language-normal speakers produce significantly more complete communication units than languageimpai red speakers. For the present investigation, utterance completeness was defined as the coordination of a noun phrase element with a verb phrase element. The formula employed to determine the percentage of complete communication units was / COMPLETE CU's ^ ^ TOTAL CU's ' Syntactic length . These measures evaluated the average and maximum length of the subject's spontaneous utterances and provided an indirect indicator of early syntactic maturity. 5. Mean Length of Communication Units in Words (MLU-W) . McCarthy (1930) originally described average sentence length as "the simplest and most objective measure of the degree to which children combine words at the various ages" (p. 50). More recently, she has stated that no measure "seems to have superseded the mean length of sentence for a reliable, easily determined, objective, quantitative, and easily understood measure of linguistic maturity" (McCarthy, 195^, pp. 550-551). Sharf (1972) found this measure correlated highly with other measures of language development. The formula for calculating mean length of communication units in words was / NUMBER OF WORDS PER CU x ^ TOTAL NUMBER OF CU's ' 6Mean Length of Communication Units in Morphemes (MLU-M) . Several authors (Barr ie-Blackley, Didow & Faurest, 1974; Brown, 1973;

PAGE 125

115 Tyack & Gottsleben, ]S7k) have used mean length of utterance in morphemes as a measure of utterance length. Barrie-Blackley et al. (1978) indicated that MLU-M "emphasizes linguistic complexity more than does MLU in words, presumably making it a more sensitive measure" (p. 10). In the present study, general language level was operationally defined as the mean number of morphemes (Brown, 1973) per communication unit (Loban, 1963, 1976). This measure was selected as the general indicator of linguistic maturity because it has been shown to reflect developmental changes at least up to 7 years of age, and in some tasks, apparently beyond age 7 (Barrie-Blackley et al., 1978). Although Brown (1973) claimed that developmental level is not reflected by increasing average utterance length past an MLU-M of about h.O, other investigators (Loban, 1976; Musselwhite, 1975; O'Donnell, 1967) demonstrated developmental increments in MLU-M in young school-aged children. Musselwhite (1975) found that MLU-M discriminated between children in kindergarten and those in second grade. Loban (1976) found a relatively steady increase in mean number of morphemes per communication unit for the same subjects from kindergarten through twelfth grade. He concluded that "from the standpoint of obtaining a simple, straightforward method to measure the degree of fluency with language, the average number of words per communication unit appears to be an exceptionally good device" (Loban, 1976, p. 26). In a similar analysis, O'Donnell, Griffin and Norris (1967) used the T-unit, an equivalent measure to the communication unit, as the basis for calculating mean length measures. They found increments in T-unit length of varying magnitude from grade to grade, and concluded that mean length of utterance "is a sensitive measure of development

PAGE 126

116 toward maturity in children's language production" (O'Donnell et al . , 1967, p. The formula for the mean number of morphemes per communication unit is , NUMBER OF MORPHEMES > WOTAL NUMBER OF CU's^ Rules for counting morphemes appear in Appendix C. 7. Mean Length of the Five Longest Utterances (MLU5~M ) . Davis (1937) first suggested that the mean length of the five longest utterances might serve as good Indicator of maximum linguistic skill--the best performance a subject can produce in a given situation. This measure was calculated by totaling the number of words in the five longest communication units and dividing by five. Syntactic complexity . This measure evaluates the level of grammatical sophistication which the child's utterances represent and categorizes them according to structural type. 8. Developmental Sentence Score (PSS) . The DSS provides a quantitative index of the expressive use of selected syntactic structures in spontaneous speech, including indefinite pronouns and noun modifiers, personal pronouns, main verbs, secondary verbs, negatives, conjunctions, interrogative reversals and wh-questions . The last 50 complete, different, consecutive, intelligible, nonecholalic sentences in each sample were used for this analysis. Utterances in this part of the sample were segmented using Lee's (197A) rules for separating sentences, which differ from Loban's (1976) with regard to coordinated sentences. A sentence was judged complete and included in this analysis if it consisted of a noun and a verb in

PAGE 127

117 subject-predicate relationship. Of the ^0 samples collected, seven did not contain 50 complete, analyzable sentences. The number of analyzable sentences in these samples ranged from hO to 49, with a mean of hS. Following Lee's (IS?'^) scoring procedure, each sentence containing any of the eight specified grammatical structures received a weighted score from 1 to 8 for each structure produced appropriately, plus an additional sentence point if the sentence were correct in all respects. Individual sentence scores for the sample were totaled and divided by the number of analyzable sentences in the sample to arrive at the mean score per sentence or the Developmental Sentence Score (DSS). Language use . To examine the functional use conveyed by the child's language, communication units were categorized into one of 2k cognitive communication strategies defined within Tough's (1976) framework for the functional analysis of language use. (See Appendix B for operational definitions and examples of each of the strategies.) Categories were mutually exclusive and exhaustive. Measures 9-37 correspond to the 2k individual cognitive communication strategies, the four superordiinate uses of language — Reporting, Logical Reasoning, Predicting, and Project i ng--and the overall functional language score.

PAGE 128

118 STRATEGIES 9. Label ing 10. Referring to detail 11. Referring to incidents 12. Referring to the sequence of events 13. Making comparisons I'*. Recognizing related aspects 15. Extracting central meaning 16. Reflecting on the meaning of experiences 17Explaining a process 18. Recognizing causal and dependent relationships 19Recognizing problems and solutions 20. Justifying judgments and actions 21. Reflecting on events and drawing conclusions 22. Recognizing principles 23. Ant i c ipat i ng/forecas t i ng 2k. Anticipating the detail of events 25. Anticipating a sequence of events 2§. Anticipating problems and possible solutions 27. Anticipating and recognizing alternative courses of action 28. Predicting the consequences of actions or events 29. Projecting into the experiences of others 30. Projecting into the feelings of others 31. Projecting into the reactions of others 32. Projecting into situations never experienced USES 33. Reporting 3^*. Logical Reasoning 35. Predicting 36. Projecting 37. OVERALL In this Investigation, only the cognitive uses of language, with the exception of Imagining, were studied. Whenever a question arose about the classification of an utterance within a particular language use, the utterance was always assigned to the highest level strategy appropriate. To modulate the effects of distortion of both the talkative and the quiet child, an "interval score" was adapted from the work of Schachter et al. (197^) as the quantitative unit of analysis. According to Schachter et al. (197^), raw frequency scores used to describe developmental language data "distort in the direction of giving excess weight to the scores of the talkative child . . ." (p. 11). Converting raw frequency scores to express the percentage of occurrence of individual

PAGE 129

language categories does not resolve the problem. "Percent conversion distorts in the direction of giving excess weight to the scores of the quiet child" (Schachter et al., 197^, p. 11). In this study, the "interval score" was defined as the number of observation intervals in which no more than two examples of a communication strategy occurred out of a total of six observation intervals per child. Each of six pictures in the Black Kitten Gets Lost series was considered to be a separate observation interval, and within any given observation interval, a communication strategy contributed to the interval score no more than two times. Therefore, the total possible interval score for a specific communication strategy across all six intervals was 12. Interval scores were computed for each of the 24 individual communication strategies defined by Tough's (1976) system, as well as for the superord i nate categories of language use--Report ing , Logical Reasoning, Predicting and Projecting--and the overall functional language score. Schachter et al. (197^) commented that the interval score seemed to offer "a promising compromise between raw frequency data and percent conversion data, identifying qualitative differences in speech within a greatly attenuated range (maximum = 12) of quantitative differences and thus modulating the effects of distortions of both the talkative and the quiet child" (p. 11). Rel iabi 1 i ty Both intrarater and interrater reliability were calculated for each of the measures of language use. For intrarater reliability, the author reanalyzed four language samples selected at random, one from each of the subject groups. For interrater reliability, two raters

PAGE 130

120 trained in the method of analysis, independently analyzed the same four language samples. For each measure of language use, the percentage of rater agreement was computed by a point-by-point percentage of agreement formula: NUMBER OF AGREEMENTS NUMBER OF AGREEMENTS + DISAGREEMENTS ^ These percentages of agreement represented the ratio of agreed observations to total observations, using each analyzed utterance as an observation. In the present investigation, a percentage of agreement of .70 was considered to be acceptable. Order of Presentation of All Tasks Four sessions were scheduled for all subjects during which the following measures were administered in the sequence indicated. All independent measures demonstrated adequate reliability and validity and were standardized on representative samples. Session 1 Approximate time: kS minutes 1 . Columbia Mental Maturity Scale 2. Bankson Language Screening Test Session 2 Approximate time: kO minutes ^. Functional Inventory of Cognitive Communication Strategies 5 . Carrow Elicited Language Inventory

PAGE 131

121 Session 3 Approximate time: kO minutes 6. Test of Auditory Comprehension of Language 7. Test of Language Development Session k Approximate time: 25 minutes 8. Test of Basic Experiences Selection Measures Columbia Mental Maturity Scale (CMMS) . The CMMS (Burgemeister, Blum & Lorge, 1972) is an individually administered test of general reasoning ability, most commonly used when there is a need for a relatively short, easily administered and scored measure of general matur i ty . The CMMS consists of 92 pictorial and figural classification items arranged in a series of eight overlapping scales or levels of increasing complexity. Each item consists of a series of from three to five drawings to which the subject responds by pointing to the drawing which is unrelated to the others. This use of the classification item type format, which requires no verbal response and a minimal motor response, makes the CMMS especially appropriate for administration with a wide variety of developmental ly delayed children. The CMMS was administered according to the standard testing protocol. Raw scores (total number of correct responses) were converted to age deviation scores (ADS) and stanines for purposes of individual and group comparisons. In order to be considered for participation in

PAGE 132

122 the present study, children needed to achieve a raw score equivalent to the fourth, fifth or sixth stanine (ADS = 89 to 110). Banl
PAGE 133

of a wide variety of grammatical forms and syntactic constructions. It was developed to circumvent problems inherent in clinical language sampling by allowing an analysis of forms and structures which the child may be capable of producing but which may not occur in a particular spontaneous sample. The CELI is a diagnostic tool which may be used to (a) identify children with language problems; (b) determine which specific linguistic structures contribute to inadequate linguistic performance; and (c) quantify language status by assignment of a numerical error score. The inventory consists of 52 stimuli, Including 51 sentences and one phrase which range in length from two to ten words. The stimuli were developed to contain basic sentence types, simple transformations and specific grammatical morphemes. The stimulus items were read by the examiner and repeated by the subject. The subject's imitations were recorded on audio-tape using a Panasonic, model RQ-320 S, portable tape recorder with a Lafayette, model ML-1, omnidirectional lavalier microphone. A total error score was computed for purposes of individual and group compari son . Test of Auditory Comprehension of Language (TACL) . The TACL (Carrow, 1973) was designed to evaluate auditory comprehension of a variety of language forms and structures. The objectives of the instrument are: (a) to assign a child to a developmental level of comprehension based on test performance; and (b) to identify specific areas of linguistic difficulty in need of intervention. The TACL consists of 101 items grouped according to grammatical category:

PAGE 134

1. Vocabulary . This section evaluates comprehension of contentive vocabulary items including nouns, adjectives, verbs and adverbs. 2. Morphology . Comprehension of function words and derivational and inflectional suffixes is measured in this section. 3. Syntax . A limited number of syntactic structures is evaluated in this section. These include imperatives, noun-verb agreement, complementation, modification and coordination. The test items were presented verbally by the examiner after which the subject responded by selecting a line drawing which represented the referent for the linguistic form being tested from a field of three. Raw scores (total number of correct responses) were tallied for individual and group comparisons. Test of Language Development (TOLD) . The TOLD (Newcomer 6 Hammi 1 1 , 1977) is a measure of language ability designed: (a) to identify children who have significant problems in comprehension and use of spoken language; (b) to isolate specific areas of language deficit, e.g., oral vocabulary, grammatic understanding; and (c) to serve as the basis for planning a program of criterion-testing and diagnostic teaching. The test is based on a two-dimensional model of language structure. The primary dimension of the model encompasses the major constructs of 1 ingui st i cs--semant i cs , syntax and phonology. The secondary dimension refers to the receptive and expressive modalities of language. These dimensions provide the basis for the development of the five principal and two supplemental subtests of the TOLD:

PAGE 135

125 Principle Subtests 1. Picture Vocabulary . This receptive subtest measures the ability to understand the meaning of spoken words. 2. Oral Vocabulary . On this expressive subtest, the ability to define comnxsn words is evaluated. 3. Grammatic Understanding . This recepti ve subtest measures the ability to comprehend morphological markers and syntactic structures . A. Sentence Imitation . The ability to repeat acceptable English sentences is evaluated on this expressive subtest. 5. Grammatic Completion . This subtest examines both receptive and expressive skills by requiring the subject to supply the morphological form necessary to complete spoken sentences. Supplemental Tests 6. Word Discrimination . This receptive subtest measures the ability to differentiate between orally presented word pairs that are either the same or minimally different. 7. Word Articulation . On this expressive subtest, the ability to produce relatively difficult sounds in isolation and blends is evaluated. For the purposes of the present study, only the five principal subtests were administered to each subject. The standardized procedures for administration and scoring of each of the subtests were followed. Raw scores (total number of correct responses) were converted to scaled scores for making individual and group comparisons.

PAGE 136

126 Test of Basic Experiences (TOBE). The TOBE (Moss, 1972) is a set of standardized group tests designed to assess the richness of conceptual background of children in preschool, kindergarten or first grade. The tests provide an indication of "how well a child's experiences have prepared him for his introduction to many of the scholastic activities that he will encounter" (Moss, 1972, p. 6). The TOBE battery consists of four separate tests at each of two 1 evel s--Mathemat i cs , Language, Science and Social Studies — and one composite test of General Concepts which includes items from the other four. Level K is designed for children in preschool or kindergarten and Level L is designed for children in kindergarten or first grade. Each subtest consists of 28 items which seem to have been selected using a combination of normreferenced and criterion-referenced criteria. The TOBE Mathematics Test measures the mastery of fundamental mathematical concepts and terms prerequisite to much of the primary mathematics curriculum. The Language Test assesses basic language comprehension skills in vocabulary, sentence structure, verb tenses, sound-symbol relationships and letter recognition. It also includes items pertaining to listening skills and perception of symbols as the carriers of meaning. The TOBE Science Test is designed to measure the extent of early experiences with animals, humans, plants, machinery, weather and other phenomena. The TOBE Social Studies Test evaluates how wel 1 a child recognizes and understands concepts pertaining to social groups, social roles, social customs, rules of safety and human emotions. The General Concepts Test is composed of items from each of the other four areas and may be used as a gross measure of a child's

PAGE 137

127 experiences and familiarity with various concepts. The General Concepts Test was not administered as part of the present investigation. In this study, the presentation of the TOBE was modified for individual administration. Subjects were instructed as follows: I am going to show you some pictures [open test booklet to p. 1]. Look at all the pictures [point to each of the pictures on the page from left to right] and then point to the picture which I describe. After presentation of the general directions, the examiner administered four of the TOBE subtests: Mathematics, Language, Science, and Social Studies. On all four subtests, the 28 test items were preceded by four practice items. If a subject had difficulty following the directions or understanding the task, some or all of the practice items were repeated. Subjects were required to "point to," "touch" or "put a finger on" the item described by the examiner. Directions for specific items were read by the examiner from the examiner's manual. For each item, the words "point to," "touch" or "put a finger on" were substituted for the word "mark" in the directions. A single score sheet was used for all subtests on a particular level. The examiner marked the subject's response for each item by recording an A, B, C or D corresponding to the first, second, third or fourth pictures on the page. Raw scores (total number of correct responses) were tallied and converted to scaled scores for individual and group comparisons.

PAGE 138

CHAPTER k ANALYSIS OF THE RESULTS Group Comparisons Language Used by Language-Normal and LanguageI mpa i red Children Since Hypotheses 1 and 2 relate to the issue of whether language-normal and language-impai red children perform similarly on an inventory of language use, they will be treated together in this sect ion. To test subhypotheses 1.1 to 1.6 and 2.1 to 2.6, the performance on FICCS for the languageimpai red group (Li) and each of the two language-normal control groups, LNl matched on the basis of chronological age and LN2 matched for utterance length, was compared. The scores used for comparison were "interval" scores defined as the number of observation intervals in which no more than two examples of a communication strategy occurred out of a possible six observation intervals per child. Interval scores were computed for each of the 2k individual communication strategies defined by Tough's (1976) system, as well as for the superordinate categories of language use — Reporting, Logical Reasoning, Predicting, and Projecting — and the overall functional language score. Tables 11, 12, and 13 show raw performance data for the four major categories of language use and the overall functional language score on FICCS achieved by the LI, LNl, and LN2 groups. The data were submitted to a one-way analysis of variance (ANOVA) to determine 128

PAGE 139

129 Table 11 Interval Scores on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the LanguageI mpa i red Group FICCS LOGICAL SUBJECT REPORTING REASONING PREDICTING PROJECTING OVERALL 1 35 n 14 20 80 2 33 9 21 22 85 3 30 6 15 13 64 k 25 11 12 11 59 5 15 9 20 10 54 6 22 6 6 17 51 7 h2 12 18 16 88 8 32 11 16 17 76 9 27 8 13 18 66 10 20 k 16 14 54 RANGE : 15-42 4-12 6-21 10-22 51-88 MEAN 28.1 8.9 15.1 15.8 67.9 MEDIAN 28.5 9.0 15.5 16.5 65.0

PAGE 140

130 Table 12 Interval Scores on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the Language-Normal (LNl) Group Matched for Chronological Age FICCS LOGICAL SUBJECT REPORTING REASONING PREDICTING PROJECTING OVERALL 11 31 7 19 71 12 31 11 11 20 73 13 k] 15 16 18 90 14 16 14 14 20 64 15 22 12 16 18 68 16 27 15 15 16 73 17 36 13 21 19 89 18 30 ]k 16 14 74 19 2k 16 16 18 74 20 32 22 28 22 104 RANGE 16-41 7-22 11-28 14-22 64-104 MEAN 28.9 13.9 16.7 18.5 78.0 MEDIAN 30.5 14.0 16.0 18.5 73.5

PAGE 141

131 Table 13 Interval Scores on the Functional Inventory of Cognitive Communication Strategies (FlCCS) Achieved by the Language-Normal (LN2) Group Matched for Utterance Length FlCCS LOGICAL CI IR IFfT O U D J t ^ 1 vX. rU r\ 1 1 Nu pp AcnM 1 wr KLHoUIN 1 INu pern 1 TT 1 wr DDH ICTTIMP r KU J tL 1 1 INb U V t (\M L L 21 13 6 12 10 41 1 n 1 u 1 1 I 1 1 U oO 23 26 7 9 14 56 24 21 7 9 n 48 25 23 16 13 15 67 26 30 8 17 17 72 27 33 12 13 16 74 28 25 11 11 16 63 29 33 6 10 9 58 30 22 2 12 15 51 RANGE 13-33 2-16 9-17 9-17 41-74 MEAN 25.5 8.6 11.7 13.3 59.1 MEDIAN 25.5 7.5 11.5 14.5 59.0

PAGE 142

132 whether differences in language use existed among the groups. According to the data in Table ]k, the results of this test were statistically significant ( = 6.108, df = 2/27, £= .007); therefore, planned orthogonal comparisons were performed to examine the overall language use of the language-impaired (Li) group in relation to each of the two language-normal control groups, LNl, matched for chronological age, and LN2, matched for utterance length. The comparison between LI and LNl yielded statistically significant results using a one-tailed test {t_ = 1.867, £< .05), whereas the comparison between LI and LN2 was not significant U = 1.627, £> .05, two-tailed test). A one-tailed test was used for comparing LI to LNl, since previous and current research indicate that 1 anguagei mpa i red children perform more poorly than their language-normal colleagues on all measures of linguistic performance. In contrast, a two-tailed test was used to examine the differences between LI and LN2, since research trends do not unequivocally determine the nature of the prediction to be made between these two groups. Although the LI and LNl groups performed as expected, with LNl achieving superior overall scores on FICCS, LI and LN2 did not. The LI group achieved higher overall scores on FICCS than did their younger, normal counterparts matched for utterance length. The findings of this analysis, then, resulted in the rejection of Ho 1.1 and acceptance of Ho 2.1. A comparison of the ability to use language for the purposes of Reporting, Logical Reasoning, Predicting and Projecting was also performed for LI, LNl, and LN2 using a one-way ANOVA. The summary data in Tables 15 through 18 show statistically significant differences among the groups for Logical Reasoning ( F_ = 7.427, df = 2/27, £= .003),

PAGE 143

133 Table ]k Summary of One-Way ANOVA for the Overall Functional Language Score on the Functional Inventory of Cognitive Communication Strategies (FiCCS) Achieved by the LI , LNI, and LN2 Groups SOURCE df SS MS P Between Groups 2 1788.868 894. 43A 6. 108 .007 Within Groups 27 3953.798 U6.if37 Total 29 57^2.664 LI = LanguageI mpa i red LNI = Language-Normal s matched for chronological age LN2 = Language-Normals matched for utterance length

PAGE 144

Table 15 Summary of One-Way ANOVA for the Logical Reasoning Subscore on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the LI, LNl, and LN2 Groups SOURCE df ss MS £ Between Groups 2 177.267 88.633 l.hll .003 Within Groups 27 322.200 11.933 Total 29 L! = Languageimpa i red LNl = Language-Normals matched for chronological age LN2 = Language-Normals matched for utterance length

PAGE 145

135 Table I6 Summary of One-Way ANOVA for the Predicting Subscore on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the LI, LNl, and LN2 Groups SOURCE df SS MS I £ Between Groups 2 130.399 65.200 h.2k\ .025 Within Groups 27 ^15.010 15.3^7 Total 29 5^5.^99 LI = LanguageI mpai red LNl = Language-Normals matched for chronological age LN2 = Language-Normals matched for utterance length

PAGE 146

136 Table 17 Summary of One-Way ANOVA for the Projecting Subscore on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the LI, LNl, and LN2 Groups SOURCE df SS £ Between Groups 2 135 .267 67.633 7.357 .003 Within Groups 27 248 .200 9.193 Total 29 383 .467 LI = LanguageImpai red LNl = Language-Normals matched for chronological age LN2 = Language-Normals matched for utterance length

PAGE 147

137 Table I8 Summary of One-Way ANOVA for the Reporting Subscore on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the LI, LNl, and LN2 Groups SOURCE df ii MS £ Between Groups 2 63.195 31.598 .62^4 .5'»3 Within Groups 27 1366.298 50.60i» Total 29 \k2S.k3k LI = LanguageI mpai red LNl = Language-Normals matched for chronological age LN2 = Language-Normals matched for utterance length

PAGE 148

138 Predicting ( £ = 4.2^41, df = 2/27, £= .025) and Projecting {F_ = 7.357, df = 2/27, 2_ = .003), but no significant differences for Reporting {F_ = 0.62A, d£ = 2/27, £= -5^3). To evaluate the significant findings, planned orthogonal comparisons were performed between LI and LNl, and Li and LN2 for the Logical Reasoning, Predicting and Projecting subscores. Results of these comparisons revealed significant differences between LI and LNl in the use of Logical Reasoning (_t = 3.236, £ < .005, one-tailed test) and Projecting ( t^ = -1.991, £< .05, one-tailed test) strategies and no significant differences between LI and LN2 for any use category. During the language sampling tasl<, both LI and LN2 used comparable numbers of Logical Reasoning strategies; and LI used significantly fewer of these strategies than did LNl. In view of these findings, subhypotheses 1.3 and 1.5 were rejected, whereas subhypotheses 1.2, l.^t and 2.2 through 2.5 were retained. Further analysis was performed to determine if there were any significant differences in the use of the 2k individual communication strategies on FiCCS between LI and LNl and LI and LN2. Data were subjected to a two-tailed Mann-Whitney U-test, the results of which, including mean score ranks and U-tests corrected for ties, are found in Tables 19 and 20. As shown, only four communication strategies, one from each major language use category, achieved a significant statistical difference between the LI and LNl groups. These included strategy 7, Extracting or Recognizing Central Meaning (jj = 10.5, £ = .002); strategy 13, Reflecting on Events and Drawing Conclusions (L[ = 15.0, £= .00k); strategy 18, Anticipating Problems and Possible Solutions {U = 25.0, £= .013); and strategy 22, Projecting into the Feelings of Others (U^ = 19.5, £= .018). In each case, the LNl group

PAGE 149

139 Table 19 A Comparison of Communication Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Used by LanguageImpaired (LI) and Language-Normal (LNl) Groups Matched for Chronological Age COMMUNICATION STRATEGIES LANGUAGEIMPAIRED (N = 10) Mean Rank LANGUAGENORMAL (N = 10) Mean Rank MANNWHITNEY U-test £ REPORTING 1 . Label ing 12.10 8.90 3^.0 .224 2. Referring to detail 11.10 9 90 44 0 644 3. Referring to incidents ]].^5 9.55 40.5 .460 k. Referring to the sequence of events 9.75 11.25 42.5 .550 5. Making comparisons 10.25 10. 75 47 5 815 6. Recognizing related aspects 9.45 11.55 39.5 .393 ?• Extracting or recognizing the central meaning A ... ~j 6.55 10.5 . 002** 8. Reflecting on the meaning or experiences 1 n fin 1 U . oU 1/ . 0 . 0 1 0 LOGICAL REASONING 9. Explaining a process 8.95 12.05 34.5 .202 10. Recognizing causal and dependent relationships 9.95 1 1 .05 44.5 . 503 11. Recognizing problems and sol ut ions 9.50 n .50 40.0 .146 12. Justifying judgments and act ions 8.95 12.05 34.5 232 13Reflecting on events and drawing conclusions 7.00 U.oo 15.0 .004** lA. Recognizing principles 9.15 1 1 .85 36.5 .211 PREDICTING 15. Anticipating/forecasting 10.70 10.30 48.0 .878 16. Anticipating the detail of events 10.50 10.50 50.0 1 .000 17. Anticipating a sequence of events 9.60 11. 'o 41.0 .483 18. Anticipating problems and possible solutions 8.00 13.00 25.0 .013** 19. Anticipating and recognizing alternative courses of action 9.^5 11.55 39.5 .382 20. Predicting the consequences of actions or events 9.85 11.15 43.5 .606

PAGE 150

]kO Table 19 (continued) COMMUN ICATION STRATEGIES LANGUAGEIMPAIRED (N = 10) Mean Rank LANGUAGENORMAL (N = 10) Mean Rank MANNWHITNEY P PRO IPTT 1 MP r r\U J QUI I INu 21. Projecting into the experiences of others 9. 10 11 .90 35.0 .28if 22. Projecting into the feelings of others 7.^5 13.55 19.5 .018* 23. Projecting into the reactions of others 10.^5 10.55 '»9.5 .969 2^. Projecting into situations never experienced 9.75 n .25 k2.5 .552 *Significant at or beyond the .05 level. **Signif icant at or beyond the .01 level.

PAGE 151

Table 20 A Comparison of Communication Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Used by LanguageImpaired (LI) and Language-Normal (LN2) Groups Matched for Utterance Length LANGUAGELANGUAGECOMMUNICATION IMPAIRED NORMAL uhitmcv STRATEGIES (N = 10) (N = 10) WHHNtY Mean Rank Mean Rank ^"^^^t REPORTING 1 . Label ing 10 .65 10 .35 48 5 .908 2. Referring to detail 11 .30 9 .70 42 0 11 0 — . 548 3. Referring to incidents 10 .95 10 .05 ^5 .5 .728 h. Referring to the sequence of events 1 I .30 9 .70 42 0 .527 5. Making comparisons 1 1 .55 9.A5 39 5 .255 6. Recognizing related aspects 8.65 JIT J J 31 5 .143 7. Extracting or recognizing the central meaning 10 90 10 .10 46 0 .750 8. Reflecting on the meaning of experiences 13 00 8 00 25 0 .050* LOGICAL REASONING 9. Explaining a process 12 05 8 95 34 5 . 121 10. Recognizing causal and dependent relationships 10 10 10 90 46. 0 . 626 1 1 . Recognizing problems and solut ions 10 00 1 1 00 45. 0 .317 12. Justifying judgments and act ions 1 1 15 9 85 43. 5 .616 13. Reflecting on events and drawing conclusions 9. 30 1 1 . 70 38.0 .261 ]k. Recoqnizinq principles 10. 75 10. 25 47. 5 .831 PREDICTING 15. Ant ici pat in g/fo recast ing 12. 70 8. 30 28. 0 .091 l6. Anticipating the detail of events 1 1 . ho 9. 60 41. 0 .473 17. Anticipating a sequence of events 1 1 . 05 9. 95 44. 5 .665 18. Anticipating problems and possible solutions 10. 00 1 1 . 00 45. 0 .317 19. Anticipating and recognizing alternative courses of action U. 05 6. 95 14. 5 . 002** 20. Predicting the consequences of actions or events 12. 70 8. 30 28. 0 .056

PAGE 152

1^2 Table 20 (continued) COMMUNICATION STRATEGIES . LANGUAGEIMPAIRED (N = 10) Mean Rank LANGUAGENORMAL (N = 10) Mean Rank MANNWHITNEY £ rKUJcL 1 1 N\i 21. Projecting into the experiences of others 9.25 11.75 37.5 .339 22. Projecting into the feel ings of others 10.^5 10.55 A9.5 .968 23. Projecting into reactions of others 14.30 6.70 12.0 . 004** m. Projecting into situations never experienced 13.25 7.75 22.5 ,022** *Significant at or beyond the .05 level. "Significant at or beyond the .01 level.

PAGE 153

used significantly more of these strategies during tlie language sample task than did the LI group. A similar analysis was performed on the data generated by the LI and LN2 groups for the 2h individual communication strategies on FICCS. The mean score ranks and U-tests, corrected for ties, are shown in Table 20. Results of the analysis revealed a statistical significance between the two groups for strategy 8, Reflecting on the Meaning of Experiences (U^ = 25.0, £= .05); strategy 19, Anticipating and Recognizing Alternative Courses of Action = 14.5, £= .0002; strategy 20, Predicting the Consequences of Actions or Events (]J = 28.0, p = .056); strategy 23, Projecting into the Reactions of Others (lU = 12.0, £ = .004); and strategy Ik, Projecting into Situations Never Experienced (£ = 22.5, £= .022). On each of the two Predicting strategies (19, 20), the two Projecting strategies (23, Ik), and the Reporting strategy (8), the LI group used significantly more examples than did the LN2 group who were matched on the basis of utterance length. To test subhypotheses 1.7 to 1.16 and 2.7 to 2.16, the data were regrouped into lower-level and higher-level cognitive communication strategies. For the Reporting and Projecting uses, the number of strategies contained in each was divided by four and those which appeared in the first quartile were designated lower-level strategies, and those which appeared in the fourth quartile were designated higher-level strategies. For the Logical Reasoning and Predicting uses, the number of strategies in each set was divided by three and those strategies below the 33rd centi le were labeled lower-level strategies; those above the 67th centi le were labeled higher-level strategies. An overall score for lower-level strategies (LoFICCS) was computed by

PAGE 154

summing the interval scores achieved on all strategies designated lowerlevel, and an overall score for higher-level strategies (HiFICCS) was computed by summing the interval scores achieved on all strategies designated higher-level. Tables 21-26 show the interval scores for lowerand higher-level strategies on FiCCS achieved by the LI, LNl and LN2 groups. These data were also subjected to a Mann-Whitney U-test to make a between-subjects comparison. Results of this analysis are presented in Tables 27 and 28, and reveal no statistically significant differences between groups for lowerlevel strategies. Both LI and LNl, and LI and LN2 used comparable numbers of lower-level strategies during elicitation of the language sample. Therefore, Ho's 1.7 to 1,11, and 2.7 to 2.11 were not rejected. In contrast, differences of statistical significance were obtained between LI and LNl, and LT and LN2 for the use of higher-level strategies. As shown in Table 23, LI and LNl performed in a significantly different manner on HiFICCS overall (U = 2A.5, £= .053), Hi Reporting (U = 2k. S, £ = .052), and Hi Logical Reasoning (_U = 12.0, £= .003). In each category, LNl used significantly more higher-level strategies than did their language-impaired peers. Similarly, Table 2k shows a significant difference in performance between LI and LN2 on HiFICCS overall ( LI = 21.0, £= .027), Hi Predicting = 9.5, £= .001), and Hi Projecting ( £ = 22.5, £= .022). Interestingly, the languageimpaired group used more higher-level strategies overall, and in particular. Predicting and Projecting strategies. These findings resulted in the rejection of subhypotheses 1.12-1.1^*, 2.12, 2.15 and 2.16, and the acceptance of 1.15, 1.16, 2.13 and 2.]k. In a post hoc analysis, a series of ANOVAs for repeated measures was performed to determine if significant differences existed between

PAGE 155

1^5 Table 21 Interval Scores for Lower-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the LanguageI mpa i red (LI) Group SUBJECT REPORTING LOWER-LEVEL LOGICAL REASONING STRATEGIES PREDICTING PROJECTING OVERALL 1 19 0 12 6 37 z ID 1 13 6 36 3 13 0 1 1 0 2A k 9 2 9 1 21 5 3 1 17 0 21 6 8 2 k 3 17 7 18 1 10 5 3A 8 15 0 12 3 30 9 13 0 12 k 29 10 7 0 0 21 RANGE 3-19 0-2 i»-17 0-6 17-37 MEAN 12. 1 0.7 11. i» 2.8 27.0 MEDIAN 13.0 0.5 12.0 3.0 26.5

PAGE 156

li»6 Table 22 Interval Scores for Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the LanguageImpa i red (Li) Group SUBJECT REPORTING HIGHER-LEVEL LOGICAL REASONING STRATEGIES PREDICTING PROJECTING OVERALL 1 1 1 0 0 1 2 c a 3 u 1, 3 1 0 i / 2 2 9 6 2 2 2 12 5 k 1 2 1 8 6 2 0 1 0 3 7 7 1 2 1 n 8 2 1 2 9 9 I 0 1 1 3 10 5 1 2 6 lA RANGE 1-7 0-2 o-/» 0-6 2-1 A MEAN 3. A 0.8 2.0 1.9 8.1 MEDIAN 3.0 1.0 2.0 1.5 9.0

PAGE 157

U7 Table 23 Interval Scores for Lower-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FlCCS) Achieved by the Language-Normal (LNl) Group Matched for Chronological Age LOWER-LEVEL STRATEGfES SUBJECT REPORTING LOGICAL REASONING PREDI CTI NG PROJECTING nVFRAI 1 V/ V L. V\r\ L, L. ] 1 15 1 10 7 33 12 16 0 10 h 30 13 1 1 5 7 1 2k 14 5 1 12 5 15 6 0 13 k 23 16 9 2 10 3 Ik 17 13 1 12 k 30 18 11 I 9 1 22 19 11 0 13 k 28 20 9 16 5 3^* RANGE 5-16 0-5 7-16 1-7 22-3A MEAN 10.6 1.5 11.2 3.8 27.1 MEDIAN 11.0 1.0 1 1 .0 4.0 26.0

PAGE 158

U8 Table 2k Interval Scores for Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the Language-Normal (LNl) Group Matched for Chronological Age HIGHER-LEVEL STRATEGI ES LOGICAL SUBJECT REPORTING REASON 1 NG PRFH 1 TT 1 MR r r\iL L/ 1 1 1 iHu PRH IFPT 1 Mr r\\/c DA 1 1 1 1 k 0 1 1 6 12 5 ] n u 1 1 7 13 8 k k 3 19 14 3 3 1 2 9 15 k i| 3 1 q 16 6 2 2 14 17 8 3 6 1 18 18 8 3 3 1 15 19 2 k 1 2 9 20 6 7 6 3 22 RANGE 3-8 0-7 0-6 1-3 6-22 MEAN 3.3 2.7 2.0 13.4 MEDIAN 5.5 3.5 2.5 2.0 14.5

PAGE 159

1^9 Table 25 Interval Scores for Lower-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the Language-Normal (LN2) Group Matched for Utterance Length SUBJECT REPORTING LOWER-LEVEL LOGICAL REASON 1 NG STRATEGIES PREDICTING PROJECTI NG OVERALL 21 5 0 10 2 17 22 7 1 8 0 16 23 1 1 0 10 5 26 2k 8 0 9 ii 21 25 7 2 10 7 26 26 16 0 12 9 37 27 16 0 12 5 33 28 10 0 7 0 17 29 22 0 10 2 3^ 30 15 0 12 7 3A RANGE 5-22 0-2 7-12 0-9 16-37 MEAN 11.7 0.3 10.0 i<.l 26.1 MEDIAN 10.5 0 10.0 26.0

PAGE 160

150 Table 26 Interval Scores for Higher-Level Strstegies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the Language-Normal (LN2) Group Matched for Utterance Length SUBJECT REPORTING HIGHER-LEVEL LOGICAL REASONING STRATEGIES PREDICTING PROJECTING OVERALL 2] 1 0 1 0 2 22 2 0 0 6 23 3 1 0 5 24 1 1 0 25 3 3 0 7 26 2 1 1 5 27 2 I 1 5 28 2 2 0 5 29 0 0 0 0 0 30 0 1 0 1 2 RANGE 0-k 0-3 0-1 0-1 0-7 MEAN 1.8 1.2 0.3 0.7 4.0 MEDIAN 2.0 1.0 0.0 1 .0 5.0

PAGE 161

151 Table 27 A Comparison of Lowerand Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Used by LanguageI mpai red (Ll) and Language-Normal (LNl) Groups Matched for Chronological Age HI/LO STRATEGIES LHIN uUMut IMPAIRED (N = 10) Mean Rank LMlNliUMut NORMAL (N = 10) Mean Rank MANNWHITNEY U-test p LOFICCS OVERALL 10.85 10.15 i»6.5 .790 LO REPORTING 9-50 11.50 'O.O .m LO LOGICAL REASONING 11.75 9.25 37.5 .317 LO PREDICTING 10.00 1 1.00 .702 LO PROJECTING 1 1 .90 9.10 36.0 .2BA HIFICCS OVERALL 13.05 7.95 2^4.5 .053* HI REPORTING 13.05 7.95 2'f.5 .052* HI LOGICAL REASONING ^.30 6.70 12.0 .003** HI PREDICTING 11.20 9.80 43.0 .588 HI PROJECTING 11.25 9.75 k2.5 .552 *Signif leant at or beyond .05. **Signif icant at or beyond .01.

PAGE 162

152 Table 28 A Comparison of Lowerand Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Used by LanguageI mpa i red (Li) and Language-Normal (LN2) broups Matched for Utterance Length HI/LO STRATEGIES LANGUAGEIMPAIRED (N = 10) Mean Rank LANGUAGENORMAL (N = 10) Mean Rank MANNWHITNEY U-test £ in Pirrc n\/PPAii LU rILLo UVtKMLL 11 in 11.10 9.90 .648 LO REPORTING 10.90 10. 10 46.0 .761 LO LOGICAL R FACnw 1 Kill 11 Q C 9.05 35.5 .195 LO PREDICTING 12.50 8.kO 29.0 .106 LO PROJECTING 9.25 11.75 37.5 .339 HI FICCS OVERALL 13.^0 7.60 21 .0 .028* HI REPORTING 12.75 8.25 27.5 .084 HI LOGICAL REASONING 9.30 1 1 .70 38.0 .332 HI PREDICTING 1A.55 6.45 9.5 .001** HI PROJECTING 13.25 7.75 22.5 .022* *Significant at or beyond ,05. **Signi f icant at or beyond .01.

PAGE 163

153 the use of lowerand higher-level strategies by subject groups LI, LNl, and LN2. Overall differences as well as differences within the major use categories of Reporting, Logical Reasoning, Predicting, and Projecting were evaluated. All three groups used significantly more overall lower level strategies than higher level strategies with results as follows: LI ( £ = 46.151, df_ = 1/9, £= .001); LNl (£ = 38.037, df = 1/9, p = .0002); and LN2 (£ = U.778, df = 1/9, p = .004). When the overall functional language score was divided into subcategory scores of language use, the three groups maintained their differential use of lowerversus higher-level strategies for the Reporting and Predicting uses, yielding the following results for the LI group: Reporting (F = 2. 1405, df = 1/9, £ = .001) and Predicting (£ = 80.817, df = 1/9, £= .00001); the LNl group: Reporting {F_ = 20.373, df_ = 1/9, £= .001) and Predicting (F = 9^.927, df_ = 1/9, £= .00000); and the LN2 group: Reporting (£ = 26.028, d£ = 1/9, £= .001) and Predicting {F_= 406.588, df_= 1/0, £= .00000), Two of the groups used significantly greater numbers of lower-level Projecting strategies than higher. The results of this analysis yielded the following: LNl ( F = 7.944, df = 1/9, £= .02) and LN2 ( F = 14.778, df = 1/9, £= .004). Finally, both language-normal groups used significantly more higherthan lower-level Logical Reasoning strategies: LNl = 10.565, df^ = 1/9, £= .01) and LN2 (f = 25.138, df = 1/9, £= .001). A summary of these findings appear in Tables 56 through 58 at the conclusion of this chapter.

PAGE 164

15^ Language Used by Language-Normal Children of Different Ages and Socioeconomic Levels To test subhypotheses 3-1 to 3.16, and 4.1 to A. 16, performance on FICCS of language-normal children at two different ages, 6 years (plus or minus 3 months) and 7 years (plus or minus 3 months), and two socioeconomic levels, lower and higher was compared. Interval scores, computed for the overall functional language score and the subcategory scores of Reporting, Logical Reasoning, Predicting, and Projecting, are shown in Tables 29-32 for each subject group. These data were analyzed using a two-way analysis of variance (ANOVA) to determine the significance of the two main effects, age and socioeconomic level, as well as their two-way interaction. Summaries of the ANOVAs for language use parameters are found in Table 33. The two main effects of age and socioeconomic level were not found to be significant for the FICCS overall functional language score and the subscores of Reporting, Logical Reasoning and Predicting, nor was the two-way interaction of age x socioeconomic level. A significant main effect for age (f. = 5.663, df_ = 1/16, £= .03) was obtained for the Projecting subscore but the main effect for socioeconomic level and the interaction effect were not statistically significant. The language-normal 6-year-olds used significantly more Projecting strategies in response to the Black Kitten Gets Lost than the 7-year-olds. These findings resulted in the acceptance of subhypotheses 3.1 to 3.4, and A.l to k.5, and the rejection of subhypothesis 3.5. To examine differences in the use of the 24 individual communication strategies on FICCS between the language-normal 6and

PAGE 165

155 Table 29 Interval Scores on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by Language Normal 6-Year-Olds FICCS LOGICAL SUBJECT REPORTING REASONING PREDICTING PROJECTING OVERALL ]] 31 7 19 71 12 31 11 11 20 73 13 k\ 15 16 18 90 1*1 16 14 ]k 20 6i» 15 22 12 16 18 68 31 37 20 25 2k 106 32 33 16 16 22 87 33 15 8 13 15 51 3A 21 12 13 19 65 35 21* 18 15 17 7k RANGE 15-Al 7-20 11-25 15-2'» 51-106 MEAN 27.1 13.3 15.3 19.2 7k. 3 MEDIAN 25.5 13 1A.5 19 72

PAGE 166

156 Table 30 Interval Scores on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by Language-Normal 7" Year-Olds ^1 IR IFPT r\ t r U r\ 1 1 N li LOGICAL D C A C A M 1 M ^ KtAbON 1 Nb FICCS PRED 1 CT 1 NG PROJECTING OVERALL 16 27 15 15 16 73 17 36 13 21 19 89 )8 30 14 16 14 74 1 0 1 0 1 o 1 o 74 20 32 22 28 22 104 36 '9 12 17 12 60 37 20 17 14 1 U 38 30 19 29 19 97 39 26 13 12 17 68 40 23 11 9 13 56 RANGE 19-36 11-22 9-29 12-22 56-104 MEAN 26.7 15.2 17.7 16.4 76 MEDIAN 26.5 H.5 16 16.5 73.5

PAGE 167

157 Table 31 Interval Scores on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by LSES Language-Normal (LNI) 6and 7-Year-Olds FICCS LOGICAL SUBJECT REPORTING REASONING PREDICTING PROJECTING OVERALL n 31 7 li» 19 71 12 31 11 11 20 73 13 41 15 16 18 90 \k 16 14 14 20 64 15 22 12 16 18 68 16 27 15 15 16 73 17 36 13 21 19 89 18 30 14 16 ]h 74 19 2k 16 16 18 74 20 32 22 28 22 104 RANGE 16-Al 7-22 11-28 14-22 64-104 MEAN 26 13.9 16.7 18.4 78 MEDIAN 30.5 14 16 18.5 73.5 LSES = Lower Socioeconomic Status

PAGE 168

158 Table 32 Interval Scores on the Functional Inventory of Cognitive Communication Strategies (FiCCS) Achieved by HSES Language-Normal (LN3) 6and 7-Year-Olds FICCS LOGICAL SUBJECT REPORTING REASONING PREDICTING PROJECTING OVERALL 31 37 20 25 2k 106 32 33 16 16 22 87 33 15 8 13 15 51 3* 21 12 13 19 65 35 24 18 15 17 74 36 19 12 1 2 60 37 20 17 14 ]k 65 38 30 19 29 19 97 39 26 13 12 17 68 i»0 23 11 9 13 56 RANGE 15-37 8-20 9-29 12-24 51-106 MEAN 2k. B 1A.6 16.3 17.2 72.9 MEDIAN 23.5 1A.5 1^.5 17 66.5 HSES = Higher Socioeconomic Status

PAGE 169

159 Table 33 Summaries of Two-Way ANOVAS for the Overall and Subcategory Scores on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by Lower (LSES) and Higher (HSES) Socioeconomic Status Language-Normal 6and 7-Year-Olds OVERALL FICCS df 11 MS F_ P Main Effects 2 115.600 57.800 .241 .789 Status 1 105.800 105.800 .441 .516 Age I 9.800 9.800 .041 .842 Interaction of Status x Age 1 30A.200 304.200 1 .268 .277 Explained 3 419.800 139.933 .583 .635 Res i dual 16 3838.396 239.900 Total 19 4258. 195 224.116 REPORTING d_f SS MS IF £ Main Effects 2 85.300 42.650 .789 .471 Status 1 84.050 84.050 1.554 .230 Age 1 1 .250 1 .250 .023 .881 Interaction of Status x Age 1 18.050 18.050 .334 .571 Expl ai ned 3 103.350 34.450 .637 .602 Res i dua 1 16 865. 198 54.075 Total 19 968.548 50.976 LOGICAL REASONING df li MS f; £ Main Effects 2 32.500 16.250 1 .213 .323 Status 1 6.050 6.050 .451 .511 Age I 26.450 26.450 1.974 .179 Interaction of Status x Age 1 18.050 18.050 1.347 .263 Explained 3 50.550 1 6.850 1.257 .322 Residual 16 214.400 13.400 Total 19 264.950 13.945

PAGE 170

160 Table 33 (coot 1 nued; PREDICTING clf SS MS F £ Main Effects 2 31 .700 15.850 .551 .587 Status 1 .450 .450 .016 .902 Age 1 31 .250 31.250 1 .087 .313 Interaction of Status X Age 1 31 .250 31.250 1 .087 .313 Explained 3 62.950 20.983 .730 .549 Res idual 16 459.999 28.750 Total 19 522.959 27.524 PROJECTING df SS MS £ Main Effects 2 50. 500 25.250 3 .401 .059* Status 1 8. 450 8.450 1 .138 .302 Age 1 42. 050 42.050 5 .663 .030* Interaction of Status x Age 1 1 1 . 250 1 1 .250 1 .515 .236 Expl a i ned 3 61. 750 20.583 2 .772 .075 Res idual 16 118. 800 7.425 Total 19 180. 550 9.503

PAGE 171

7-year-ol ds , a Mann-Whitney U-test was applied to tlie data generated by the groups and the results are reported in Table 3k. Three strategies, strategy 3, Referring to Incidents (L[ = 19-5, £= .016); strategy 7, Extracting and Recognizing Central Meaning {U_ = 25.5, £= .057); and strategy 23, Projecting into Reactions of Others (£ = 21.0, £= .026), differentiated between the groups to a statistically significant degree. The 6-year-old group produced more examples of strategies 3 and 23, whereas the 7-year-old group produced greater numbers of strategy 7. A similar analysis was performed on the Interval scores achieved by the lower and higher socioeconomic groups. Results of the MannWhitney U-tests, shown in Table 35, yielded a significant difference between the two groups in the use of only one strategy, strategy 9, Explaining a Process ( = 25.5, £= .046). The higher socioeconomic group used a significantly greater number of this strategy type than did the lower socioeconomic group. To test subhypotheses 3-7 to 3-16, and k.7 to 4.16, the data were regrouped into lower-level and higher-level cognitive communication strategies. Tables 36-39 show the interval scores for lowerand higher-level strategies on FICCS for the 6and 7-year-old languagenormal groups. According to the data in Table kO, the 6and 7-yearold groups used comparable numbers of lowerand higher-level strategies on FICCS. Since there were no significant differences between the groups in the use of these strategies, subhypotheses 3.7 to 3.16 were not rejected. Tables 41-44 show the interval scores for lowerand higher-level strategies on FICCS achieved by the lower and higher socioeconomic groups. With one exception, there were no significant differences evidenced between the groups. The mean ranks and

PAGE 172

162 Table 3^ A Comparison of Communication Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Used by Language-Normal 6and 7-Year-Olds LANGUAGELANGUAGECOMMUNICATION , ^^^^^"^ NORMAL MANNSTRATEGIES 6-YEAR-OLDS 7-YEAR-OLDS WHITNEY (N = 10) (N = 10) U-test Mean Rank Mean Rank REPORTING 1 . Label ing 12 10 "iL n 99 1 . ZZ 1 2. Referr i ng to deta i 1 9.55 11.45 40.5 .462 3. Referring to incidents 13.55 7.45 19.5 .016 4. Referring to the sequence of events 10.30 10.70 48.0 .877 5. Making comparisons in \^ 1 n Ri; 'to . 5 6. Recognizing related aspects O . jU \ i.. IK) Zo . U 7. Extracting or recognizing 8. the central meaning ft nt; 1 z . 9 C C . Ui)/ Reflecting on the meaning of experiences 1 u . up in Qc 1 u . 79 "7 . /z/ LOGICAL REASONING 9. Explaining a process 9.^0 11.60 39.0 .369 10. Recognizing causal and dependent relationships 11.55 9.45 39.5 .256 1 1 . Recognizing problems and solutions 9.20 1 1 .80 37.0 .194 12. Justifying judgments and act ions 9.10 1 1 .90 36.0 .282 13. Reflecting on events and \k. drawing conclusions 8.55 12.45 30.5 .132 Recognizing principles 8.60 12.40 31.0 .136 PREDICTING 15. An 1 1 c 1 pa t i n g/ f o reca s t i n g 1 1 .90 9.10 36.0 .282 16. Anticipating the detail 17. of events 8.25 12.75 27.5 .079 Anticipating a sequence 18. of events 8.85 12.15 33.5 .199 Anticipating problems and possible solutions 9.15 1 1 .85 36.5 .229

PAGE 173

Table 3^* (continued) COMMUNICATION STRATEGIES LANGUAGENORMAL 6-YEAR-OLDS (N = 10) Mean Rank LANGUAGENORMAL 7-YEAR-OLDS (N = 10) Mean Rank MANNWHITNEY U-test 19. Anticipating and recognizing alternative courses of action 20. Predicting the consequences of actions or events PROJECTING 21. Projecting into the experiences of others 22. Projecting into the feelings of others 23. Projecting into the reactions of others 2k. Projecting into situations never experienced 8.20 12.80 27.0 .069 10. 15 10.85 46.5 .783 12.45 8.55 30.5 .133 9.50 1 1 .50 40.0 .433 13.40 7.60 21.0 .026* 12.15 8.85 33.5 .194 *Signif leant at or beyond the .05 level.

PAGE 174

164 Table 35 A Comparison of Communication Strategies on the Functional inventory of Cognitive Communication Strategies (FICCS) Used by the Lower (LSES) and Higher (HSES) Socioeconomic Language-Normal Groups LSES HSES COMMUNICATION LANGUAGELANGUAGEMANNSTRATEGIES NORMAL NORMAL WHITNEY (N = 10) (N = 10) U-test Mean Rank Mean Rank REPORTING 1 . Label ing 8.40 12.60 29.0 .109 2. Referring to detai 1 9.05 11.95 35.5 .261 3. Referring to incidents 11.15 9.85 43.5 .609 k. Referring to the sequence of events in At; 48.5 .908 . 5. Making comparisons 1 u . /U 1 U . 3U Ho. 0 • 859 6. Recognizing related aspects 1 n ftc 1 u . op 46.5 .780 7. Extracting or recognizing 8. the central meaning Q An 1 1 lin 1 1 . 4U li 1 n .485 Reflecting on the meaning of experience 9 30 1 1 70 J) u . u LOG 1 GAL REASONING 9. Explaining a process 12.95 8.05 25.5 .046* 10. Recognizing causal and dependent relationships 10.40 10.60 49.0 .914 11. Recognizing problems and 12. solutions 11.10 9.90 44.0 .549 Justifying judgments and act ions 10.10 10.90 46.0 .759 13. Reflecting on events and 14. drawing conclusions 10. 15 10.85 46.5 .787 Recognizing principles 11.05 9.95 44.5 .666 PREDICTING 15. Anticipating/forecasting 9.45 11.55 39.5 .419 16. Anticipating the detail 17. of events 10.60 10.40 49.0 .938 Anticipating a sequence 18. of events 11.35 9.65 41.5 .508 Anticipating problems and possible solutions 9.00 12.00 35.0 .181

PAGE 175

165 Table 35 (continued) LSES HSES COMMUNICATION STRATEGIES LANGUAGENORMAL (N = 10) Mean Rank LANGUAGEIII n kj A 1 NORMAL (N = 10) Mean Rank MANNWH ITNEY U-test £ ij. MnLicipaLiny ana recog nizing alternative courses or ac L 1 on Q on y . zo I 1 . oO 37.0 ^u. rreaiCLing tne consequences of actions or events 10.75 10.25 A5.5 PROJECTING 21. Projecting into the experiences of others 9.30 1 1 .70 38.0 .355 22. Projecting into the feelings of others 8. AO 12.60 29.0 .100 23. Projecting into the reaction of others 10.70 10.30 A8.0 .878 Ik. Projecting into situations never experienced 10.80 10.20 A7.0 .813 "Significant at or beyond the .05 level.

PAGE 176

166 Table 36 Interval Scores for Lower-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the Language-Normal 6-Year-Olds LOWER-LEVEL STRATEGIES LOGICAL SUBJECT REPORTING REASONING PREDICTING PROJECTING OVERALL n 15 1 10 7 33 12 16 0 10 4 30 \3 11 5 7 1 24 14 5 1 12 5 23 15 6 9 13 4 23 31 14 4 17 7 42 32 10 4 8 4 26 33 4 1 12 4 21 34 5 2 12 3 22 35 10 3 11 3 27 RANGE ^-16 0-5 7-17 1-7 21-42 MEAN 9.6 2.1 11.2 4.2 27.1 MEDIAN 10.0 1.5 11.5 4.0 25.0

PAGE 177

167 Table 37 Interval Scores for Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the Language-Normal 6Year-Olds HIGHER-LEVEL STRATEGIES LOGICAL SUBJECT REPORTING REASONING PREDICTING PROJECTING OVERALL 11 k 0 1 1 6 12 5 1 0 I 7 13 8 k 4 3 19 l^i 3 3 1 2 9 15 k k 3 15 31 5 6 4 3 18 32 7 1 3 2 13 33 1 0 0 4 5 34 5 2 0 3 10 35 2 k k 1 11 RANGE 1-8 0-6 0-4 1-A 5-19 MEAN 2.5 2.0 2.i» 11.3 MEDIAN 2.5 2.0 2.5 10.5

PAGE 178

168 Table 38 Interval Scores for Lower-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the Language-Normal 7"Year-01ds LOWER-LEVEL STRATEGIES SUBJECT REPORTING LOGICAL REASONING PREDICTING PROJECTING OVERALL 1 0 9 2 10 3 17 13 1 1 12 30 1 it 1 0 1 1 1 1 9 1 1 22 19 M 0 13 4 28 20 9 4 16 5 34 36 5 1 10 2 18 37 2 5 3 14 38 9 I 14 0 24 39 8 2 9 5 24 40 7 2 9 3 21 RANGE 4-13 0-4 5-16 0-5 14-34 MEAN 8.6 1.6 10.7 3.0 23.9 MEDIAN 9.0 1.5 10.0 3.0 24.0

PAGE 179

169 Table 39 Interval Scores for Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by the Language-Normal 7-Year-Olds SUBJECT REPORTING HIGHER-LEVEL LOGICAL REASONING STRATEGIES PREDICTING PROJECTING OVERALL 1 o 6 2 2 14 0 3 6 1 18 1 o 1(5 8 3 3 I 15 19 2 4 1 2 9 20 6 7 6 3 22 36 4 3 4 3 14 37 7 6 3 1 17 38 5 7 12 3 27 39 5 2 1 0 8 i»0 3 1 0 1 5 RANGE 2-8 1-7 0-12 0-3 5-27 MEAN 4.0 3.8 1.7 14.9 MEDIAN 5.5 3.5 3.0 1.5 14.5

PAGE 180

170 Table kO A Comparison of Lowerand Higher-Level Strategies on tiie Functional Inventory of Cognitive Communication Strategies (FICCS) Used by Language-Normal 6and 7-Year-Olds HI/LO STRATEGIES LANGUAGENORMAL 6-YEAR-OLDS (N = 10) Mean Rank LANGUAGENORMAL 7-YEAR-OLDS (N = 10) Mean Rank MANNWHITNEY U-test LO FICCS OVERALL 1 1 .80 9.20 37.0 .321 LO REPORTING 9.60 k].0 .kSk LO LOGICAL REASON ING 11 .00 10.00 ks.o .695 LO PREDICTING 11 .00 10.00 ks.o .703 LO PROJECTING 8.55 30.5 .133 HI FICCS OVERALL 8.90 12. 10 3^.0 .226 Hi REPORTING 9.05 11.95 35.5 .267 HI LOGICAL REASONING 8.25 12.75 27.5 .085 HI PREDICTING 9.00 12.00 35.0 .250 HI PROJECTING 12.15 8.85 33.5 . \3k

PAGE 181

171 Table ^41 Interval Scores for Lower-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by Lower Socioeconomic Status (LSES) Language-Normals LOWER-LEVEL STRATEGIES SUBJECT REPORTING LOGICAL REASONING PREDICTING PROJECTING OVERALL 1 1 15 1 10 7 33 12 16 0 10 k 30 13 1 1 5 7 1 Ik 14 5 1 12 5 23 15 6 0 13 k 23 16 9 2 10 3 Ik 17 13 1 12 k 30 18 11 1 9 1 22 19 11 0 13 h 28 20 9 k 16 5 34 RANGE 5-16 0-5 7-16 1-7 22-34 MEAN 10.6 1.5 1 1 .2 3.8 27.1 MEDIAN 11 .0 I.O 11 .0 4.0 26.0 LSES = Lower Socioeconomic Status

PAGE 182

172 Table k2 Interval Scores for Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by Lower Socioeconomic Status (LSES) Language-N'ormals HIGHER-LEVEL STRATEGIES SUBJECT REPORTING LOGICAL REASONING PREDICTING PROJECTING OVERALL 11 k 0 1 1 6 12 5 1 0 1 7 13 8 k 4 3 19 \k 3 3 I 2 9 15 It k 3 k 15 16 6 k 2 2 lA 17 8 3 6 1 18 18 8 3 3 1 15 19 2 4 1 2 9 20 6 7 6 3 22 RANGE 2-8 0-7 0-6 1-A 6-22 MEAN 3.3 2.7 2.0 13.4 MEDIAN 5.5 3.5 2.5 2.0 14.5 LSES = Lower Socioeconomic Status

PAGE 183

173 Table ^3 Interval Scores for Lower-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by Higher Socioeconomic Status (USES) Language-Normals LOWER-LEVEL STRATEGIES SUBJECT REPORTING LOGICAL REASONING PREDICTING PROJECTING OVERALL 31 ]k k 17 7 k2 32 10 k 8 4 26 33 k 1 12 4 21 3A 5 2 12 3 22 35 10 3 11 3 27 36 5 1 10 2 18 37 k 2 5 3 ]k 38 9 I 14 0 24 39 8 2 9 5 24 40 7 2 9 3 21 RANGE k-]k 1-4 5-17 0-7 14-42 MEAN 7.6 2.2 10.7 23.9 MEDIAN 7.5 2.0 10.5 3.0 23.0 HSES = Higher Socioeconomic Status

PAGE 184

Table Interval Scores for Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Achieved by Higher Socioeconomic (HSES) Language-Normals HIGHER-LEVEL STRATEGIES SUBJECT REPORTING LOGI CAL REASONING PREDICTING PROJECTING OVERALL 31 5 6 k 3 18 32 7 1 3 2 13 33 1 0 0 5 3A 5 2 0 3 10 35 2 Jil k 1 n 36 k 3 k 3 37 7 6 3 1 17 38 5 7 12 3 27 39 5 2 1 0 8 40 3 1 0 I 5 RANGE 1-7 0-7 0-12 0-k 5-27 MEAN 3.2 3.1 2.1 12.8 MEDIAN 5.0 2.5 3.0 2.5 12.0 HSES = Higher Socioeconomic Status

PAGE 185

Mann-Whitney U-tests reported in Table kS between the lower and higher socioeconomic language-normal groups revealed a significant difference for Lo Reporting strategies {U_ = 25.0, £= .058), with the lower socioeconomic group using greater numbers of this type of strategy. On the basis of these data, subhypothes i s ^.8 was rejected; subhypotheses k.7, and k.S to A. 16 were accepted. Tables 56-58, at the conclusion of this chapter, present a summary of these findings. Correlation and Regression Analyses To test Hypotheses 5, 6 and 7, performance on FICCS was correlated with performance on standardized measures of language content and/or form, nonstandard i zed measures of language form, and measures of academic achievement. The predictive value of these three types of measures in estimating performance on a measure of language use, FICCS, was also explored. Language Use and Standardized Measures of Language Content and/or Form To test Hypothesis 5, the relationship between performance on the Test of Language Development (TOLD) (Newcomer S Hammi 1 1 , 1977), the Test of Auditory Comprehension of Language (TACL) (Carrow, 1973), the Carrow Elicited Language Inventory (CELI) (Carrow, 197'*) and FICCS was analyzed for all subjects using a Pearson product-moment correlation (0 . The raw performance data for each subject on the criterion measure and each of the predictor measures is included in Appendix H and the intercorrelations among the four measures is presented in Table A6. As is shown, there is a marked degree of mul t icol 1 inear i ty among the

PAGE 186

176 Table k5 A Comparison of Lowerand Higher-Level Strategies on the Functional Inventory of Cognitive Communication Strategies (FICCS) Used by the Lower (LSES) and Higher (HSES) Socioeconomic Language-Normal Groups HI/LO STRATEGIES LSES LANGUAGENORMAL (N = 10) Mean Rank HSES LANGUAGENORMAL (N = 10) Mean Rank MANNWH 1 TNEY U-test n E. LO FICCS OVERALL 12.70 8.30 28.0 .093 LO REPORTING lA.O 8.00 25.0 .058* LO LOGICAL REASONING 8.50 12.50 30.0 .118 LO PREDICTING 11.15 9.85 A3. 5 .620 LO PROJECTING 11.70 9.30 38.0 .355 HI FICCS OVERALL 10.95 10.05 45.5 .733 HI REPORTING 1 1 .80 9.20 37.0 .320 HI LOGICAL REASONING 10.40 10.60 49.0 .939 HI PREDICTING 10.65 10.35 48.5 .908 HI PROJECTING 10.20 10.80 47.0 .813 '"Significant at or beyond the .05 level. LSES = Lower Socioeconomic Status HSES = Higher Socioeconomic Status

PAGE 187

177 Table Correlations anrong Three Standardized Linguistic Measures and the Functional inventory of Cognitive Communication Strategies (FICCS) for All Subjects FICCS TACL TOLD CELI FICCS 1 .00** .30 .16 TACL 1 .00** .60** .35* TOLD 1.00** .76** CELI 1 .00** *Significant at or beyond the .05 level. **Signif icant at or beyond the .01 level. TACL = Test of Auditory Comprehension of Language TOLD = Test of Language Development CELI = Carrow Elicited Language Inventory

PAGE 188

178 variables, _r = -.35 to .76. Table kj summarizes the relationship between FICCS and the three standardized measures of language content and/or form relative to the obtained correlation coefficients (r) , coefficients of determination {r_^) , and standard errors (SE) . These results indicated that only performance on the TACL [£ (38) = .kk, £< .005] was significantly correlated with performance on FICCS when a two-tailed test of significance was applied. Although the correlation coefficient and coefficient of determination suggest a moderate degree of positive relationship between TACL and FICCS, only 19 percent of the variance on FICCS can be predicted from knowledge of the variance on TACL. The measure of standard error between TACL and FICCS was .13* This measure indicates the probable extent to which a correlation is apt to vary on future samplings; that is, it provides an indication of reliability, "the degree of stability which any found measure is likely to have when it is derived from a sample drawn from a larger population about which we wish to generalize" (Franzblau, 1958, pp. 76-77). Since, by chance alone, this may be expected to vary on future sampling, and since, in 99-8 percent of the cases, the variations may be expected to fall within ±3 Standard Error ( SE ) of the found measure, these limits define the range of the probabilities for the total population. Because the observed correlations between TACL and FICCS are smaller than three times the SE, they are not statistically reliable relative to size or sign, although they are sufficiently significant to warrant further study. Franzblau (1958) encourages follow-up study, using a larger sample, on any correlation which is equal to or exceeds .30, since chance factors are much more likely to be present in small samples. On future sampling, the range of variations due to chance alone would

PAGE 189

179 Table Relationship of Standardized Measures of Language to the Functional Inventory of Cognitive Communication Strategies (FICCS) for Al 1 Subjects MEASURES r SE ±3SE TOLD .30 .14 -. 12 72 .09 ns TACL Ak .13 .05. 83 .19 .01 CELI -.16 .15 -.61. 29 .03 ns TOLD = Test of Language Development TACL = Test of Auditory Comprehension of Language CELI = Carrow Elicited Language Inventory

PAGE 190

1 80 produce the following: r_ = .05 to .83. This ranges from almost no correlation to a marked or high correlation. Finally, to determine the predictive value of the independent variables, TACL, TOLD, and CELI, in relation to the criterion variable. Ft CCS, the raw performance data were submitted to a stepwise linear multiple regression analysis. A summary of the results of the procedure is included in Table '48. According to these data, the regression of FICCS on three standardized measures of language content and/or form is statistically significant (£ = 2.953, d_f = 3/36, £< .05. Of the total variance of the FICCS scores of ^0 children studied, 20 percent is accounted for by a linear combination of the TACL, TOLD, and CELI (R^ = .kk, = .20). Further analysis of the results indicated, however, that only the TACL (£ = 9. 136, d_f = I/38, £ < .01) significantly contributed to the prediction of performance on FICCS. The other two variables in the equation, TOLD and CELI, did not make a statistically significant contribution to the prediction; that is, there was no improvement in the accuracy of estimating performance on FICCS with the additional variables. Franzblau (1958) believes that "the predictive value of the correlation coefficient is almost never put to any practical use to estimate one measure from another" (p. 88). Coefficients below .40 do not yield accurate predictions even 10 percent better than chance. To yield a prediction which is 25 percent better than chance, the correlation must be at least .66; to achieve a prediction which is 50 percent better than chance, the correlation must be at least .86; while to be 75 percent better than chance, the coefficient must approximate .97. These figures explain why "the requirements for making sound predictions are rea 1 1 y too h i gh for ordinary circumstances"

PAGE 191

181 Table kS ANOVA and R^s of the Regressions of the Three Standardized Linguistic Measures on the Functional Inventory Cognitive Communication Strategies (FICCS) SOURCE df 11 MS f; £ R^ Xi X3 3/36 1729.287 576.i*29 2.953 .05 .197 Residual 7028.313 195.231 Xi X2 2/37 1715.962 857.981 4.508 .025 .196 Res i dual 70AI .638 190.315 Xi 1/38 1697.369 1697.369 9.136 .005 Residual 7060. 231 185.796 Xi = Test of Auditory Comprehension of Language (TACL) X2 = Test of Language Development (TOLD) X3 = Car row El ici ted Language Inventory (CELl)

PAGE 192

182 (Franzblau, 1958, p. 88). The correlation between performance on TACL and FICCS, and TOLD and FICCS were .kk and .30, respectively. In view of these findings, null subhypotheses 5.1 and 5.2 were rejected, and 5.3 was not. Language Use and Nonstandard! zed Measures of Language Form The relationship between the overall interval score achieved on FICCS and each of eight nonstandard i zed linguistic measures of form was analyzed for all subjects using the Pearson product-moment correlation (r) . The raw performance data for all subjects on #CU's, TNW, MLU-W, TNM, MLU-M, MLU5-M, %CUc and DSS appear in Appendix I, and the intercorrelations among the measures are presented in Table kS. Again, the results indicated a high degree of mul ticol 1 i neari ty among the variables, high intercorrelations of the independent variables, .31 to .99. A summary of the correlational analysis is shown in Table 50, including the correlation coefficients (jr) and coefficients of determination {r_^) obtained between FICCS and each of the eight measures of linguistic form, as well as the standard errors (SE) . These data revealed statistically significant, positive correlations between FICCS and all of the nonstandard linguistic measures when a two-tailed test of significance was applied. The highest correlation was obtained between FICCS and #CU's (38) = .86, p < .001]. The TNM and TNW both correlated II (38) = .8A, £< .001]; MLU5-M [r_ (38) = .70, p < .001]; MLU-M and MLU-W [£ (38) = .5A, p < .001]; DSS [r_ (38) = .i,l, p < .01]; and %CUc [jl (38) = .35, £< .05]. Although each of the obtained correlation coefficients was significantly different from zero, only the coefficients for #CU's, TNM, and TNW showed a high degree of correlation with

PAGE 193

183 Table kS Correlations among Eight Nonstandard i zed Linguistic Measures and the Functional Inventory of Cognitive Communication Strategies (FICCS) for All Subjects FICCS #CU's TNW MLU-W TNM MLU-M MLU5-M ^CUc DSS FICCS 1 . 00** . 86** . 8^** .5^** .84** .5/,** .70** .35+ .i»l* #CU's I. 00** .91** .46* .91** .^7* .70** .36+ .31 TNW 1 .00** .77** .99** .77** .89** .50* .63** MLU-W I .00** .76** .99** .60** .85** TNM 1 .00** .77** .89** .50* .62** MLU-M 1 .00** .88** . 6 1 ** . 85** MLU5-M 1 .00** .51** .77** %C[}c 1 .00** .60** DSS 1.00 +Significant at or beyond the .05 level. *Significant at or beyond the .01 level. **Signi f leant at or beyond the .001 level.

PAGE 194

184 Table 50 Relationship of Measures of Academic Achievement to the Functional Inventory of Cognitive Communication Strategies (FICCS) for Al 1 Subjects MEASURES £ SE^ ±3SIE r^ £ #CU's .86 .04 .Ih .98 .73 <.001 TNW .Zk .OA .72 .96 .70 <.001 MLU-W .54 .11 .21 .87 .29 <.001 TNM .84 .05 .69 .99 .70 <.00I MLU-M .54 .11 .21 .87 .29 <.001 MLU5-M .70 .08 .49 .94 .kS <.00I °^CUc .35 .14 -.07 .77 .20 <.05 OSS .41 .13 .02 .80 .17 <.01 #CU's = Total Number of Communication Units TNW = Total Number of Words MLU-W = Mean Length of Communication Units in Words TNM = Total Number of Morphemes MLU-M = Mean Length of Communication Units in Morphemes MLU5-M = Mean Length of the Five Longest Communication Units in Morphemes ICUc = Percentage of Complete Communication Units DSS = Developmental Sentence Score

PAGE 195

185 FICCS, with shared variances of 73 percent, 70 percent and 70 percent, respectively. The coefficient for MLU5~M showed a marked degree of correlation with FICCS and accounted for ^9 percent of the variance on FICCS. The variables, MLU-M, MLU-W, and DSS correlated moderately with FICCS, accounting for 29 percent, 29 percent, and 17 percent of the shared variance, respectively. Finally ^CUc evidenced the lowest correlation with FICCS, a low positive relationship in which 12 percent of the variance on FICCS could be predicted from knowledge of the variance on ^CUc. On the basis of the observed findings, null subhypotheses 6.1 to 6.8 were rejected. To determine the reliability of the obtained coefficients, the measures of standard error for each variable were consulted. Four of the observed correlations appeared statistically reliable relative to size and sign, including #CU's, TNW, TNM, and MLU5-M. The respective SE's of .Ok, .Ok, .05, and .08 indicated that on future samplings, 99.8 percent of the samples would yield a correlation with FICCS, of the same direction and within one level of the original correlation. Changes in the correlations of MLU-W and MLU-M with FICCS on subsequent samplings would range between .21 and .87, indicating stability relative to sign, but not relative to size. The correlations between %CUc and DSS, and FICCS proved unreliable in terms of both size and sign. On future sampling, the range of variations due to chance alone would produce the following: _r = -.07 to .77and£= ,02 to .80, respectively. This ranges from a negligible correlation to a marked or high correlation, and in the former example, the sign is reversed at the lower end. The predictive value of the eight independent variables, #CU's, TNW, MLU-W, TNM, MLU-M, MLU5-M, ^CUc and DSS, in relation to the

PAGE 196

186 criterion variable, FICCS, was evaluated for all subjects using a stepwise linear multiple regression analysis. The results of this analysis, as shown in Table 51, indicated that the regression of FICCS on the eight nonstandardi zed measures of language form was statistically significant ( F_ = 16.^*82, df = 8/31, £< .001), yielding a multiple correlation coefficient of .90 and a multiple coefficient of determination of .81. In other words, of the total variance of the FICCS scores of 'O children studied, 81 percent can be accounted for by a linear combination of the eight predictor variables. Further analysis of the data indicated, however, that only #CU's ( F = 71-958, df = 2/37, £< .01) and MLU-W (£ = 3.919, d£ = 2/37, £< .01) significantly contributed to the prediction of performance on FICCS. After these first two predictor variables have been entered into the equation, little is added to by successively entering additional variables. These two variables, #CU's and MLU-W, acting in concert, achieved a multiple correlation of = .87 and multiple coefficient of determination of R^^ = .076, i.e., 76 percent of the variance on FICCS can be predicted from knowledge of the variance on the two predictor variables. Moreover, because the multiple correlation approximated .86, predictions which are 50 percent better than chance may be achieved. Therefore, it appears that the combination of two variables may predict the criterion variable almost as well as the combination of eight variables. Language Use and Measures of Academic Achievement The relationship between four measures of academic achievement and FICCS were determined in order to test Hypothesis 7. The Pearson product-moment correlation (r) was used to examine the relationship for

PAGE 197

187 Table 51 . ANOVA and ' s of the Reg ressions of the Eight Nonstandardized Linguistic Measures on the Functional Inventory of Cognitive Communication Strategies (FICCS) SOURCE df SS MS r r n r2 Xl Xs 8 7090.575 886.322 16.482 < .001 .806 Res i dual 31 1667.025 53.775 Xl Xy 7 7083.760 1011 .966 19.346 < 001 .809 Res idual 32 1673. S'to 52.308 Xl Xe 6 6727.330 1 121 .222 18.224 < 001 .768 Res i dual 33 2030.270 61 .523 Xl Xs 5 6701.3^5 1340.269 22. 161 < 001 .765 Res idual 3^ 2056.225 60.478 if 9 ft n77 UU 1 . /bZ Res i dual 35 2080.796 59.451 Xl X3 3 6659.241 2219.747 38.083 <• 001 .760 Res idual 36 2098.358 58.288 Xl X2 2 6636. 117 3318.059 57.869 <. 001 .758 Res i dual 37 2121 .483 57.337 1 641 1 .431 641 1 .431 103.844 <. 001 .732 Res idual 38 2346.169 61 .741 Xl = Total Number of Commun i cat ion Units (#CU s) X2 = Mean Length of Communication Units in Words (MLU-W) X3 = % of Complete Communication Units {%CUc) Xit = Developmental Sentence Score (DSS) X5 = Total Number of Morphemes (TNM) Xe = Total Number of Words (TNW)

PAGE 198

Table 51 (continued) X7 = Mean Length of Communication Units in Morphemes (MLU-M) Xs = Mean Length of Five Longest Communication Units in Morphemes (MLU5-M)

PAGE 199

all subjects between performance on FICCS and four subtests of the Test of Basic Experiences (Moss, 1972): language, mathematics, social studies and science. The raw performance data for all subjects on these measures appears in Appendix J, and the intercorrelations among the five measures are included in Table 52. Again, the data demonstrate a high degree of mul t i col 1 i near i ty among the variables, which is indicative of high intercorrelations among the independent variables ranging from .78 to .88. Table 53 presents a summary of the correlational analysis, including the correlation coefficients (r) , coefficients of determination {r^) and measures of standard error ( SE ) obtained between FICCS and each of the four subtests of academic achievement. When a two-tailed test of significance was applied to the obtained correlations, the results indicated that all correlations were statistically different from zero and worthy of continued investigation: TBES [r_ (38) = .50, £ < .01]; TBEM W (38) = .k7, p < .01]; TBEL [r_ (38) = .39, £ < .02]; and TBESS [r_ (38) = .36, £< .05]. All coefficients were positive, with TBES and TBEM correlating to a moderate degree and TBEL and TBESS to a low degree. The amount of shared variance between FICCS and the four measures of academic achievement ranged from a low of 13 percent to a high of 25 percent, suggesting that very little variation in the independent variables is linked to variation in the criterion variable. On the bases of these results, null subhypotheses 7.1 to J.h were rejected. The reliability of the obtained coefficients was evaluated by examining the measures of standard error for each variable. These ranged from .12 between TBEM and FICCS and .l^t between TBESS and FICCS. The respective SE_' s indicated unreliable correlations in terms of size

PAGE 200

190 Table 52 Correlations among Four Measures of Academic Achievement and the Functional Inventory of Cognitive Communication Strategies (FICCS) for All Subjects FICCS TBES TBEM TBESS TBEL FICCS 1.00** .50* .47* .36+ .39+ TBES 1 .00** . 79** .79** .78** TBEM 1 .00 .83** .88** TBESS 1.00** .88** TBEL I .00** +S i gn i f i cant at or beyond the .05 1 evel . *Significant at or beyond the . 01 1 evel . **Signif icant at or beyond the . 001 level . TBES = Test of Basic Experi ences (Science) TBEM = Test of Bas i c Exper iences (Mathemat i cs) TBESS = Test of Bas i c Experi ences (Social Sciences) TBEL = Test of Bas i c Experiences (Language)

PAGE 201

Table 53 Relationship of Measures of Academic Achievement to the Functional Inventory of Cognitive Communication Strategies (FICCS) for Al 1 Subjects MEASURES r SIE ± :3SE £ TBEL .39 .13 .00 .78 .15 <• .02 TBEM .^7 .12 .11 .82 .22 <, .01 TBESS .36 .14 -.07 .77 .13 <, .05 TBES .50 .12 .lA .86 .25 <, .01 TBEL = Test of TBEM = Test of TBESS = Test of TBES = Test of Basic Experiences Basic Experiences Basic Experiences asic Experiences (Language) (Mathemat i cs) (Social Studies) Sci ence)

PAGE 202

192 and sign between all four measures of academic achievement and FICCS. On future samplings, the range of variations due to chance alone would produce the following: TBEL and FICCS (£ = .00 .78); TBEM and FICCS (£ = .11 .82); TBES and FICCS = .14 .86); and TBESS and FICCS {r_ = -.07 " .77). This ranges from no or negligible correlation to a marked or high correlation, and in the last example, the sign is reversed at the lower end. A stepwise linear multiple regression analysis was used to examine the predictive value of the four independent variables, TBEL, TBEM, TBES and TBESS in relation to the criterion variable FICCS. A summary of this analysis is presented in Table 5^. These data showed that the regression of FICCS on the four measures of academic achievement was statistically significant {•_ = 3.379, d_f = A/35, £< .01) and yielded a multiple correlation coefficient of .53 and a multiple coefficient of determination of .28. This means that 28 percent of the total variance on FICCS can be accounted for by a linear combination of the four predictor variables. When the F_ ratios of the regression weights were consulted, however, it was noted that only TBES (F = 12. 76^*, df = 1/38, £< .005) significantly contributed to the prediction of performance on FICCS. Little is added to R^^ by successively entering the other three variables into the equation. The multiple correlation coefficient and multiple coefficient of determination between TBES and FICCS were .50 and .25, respectively. These data indicate that accurate predictions of FICCS scores from knowledge of scores on TBES can only be achieved at approximately 20 percent better than chance.

PAGE 203

193 label S'i ANOVA and R^'s of the Regressions of the Four Measures of Academic Achievement on the Functional Inventory of Cognitive Communication Strategies (FICCS) SOURCE II ii !li £ p r2 Xi Xi, V35 2k33.(>kk 609.91 1 3.379 <. 025 .528 Res i dual 6317.956 180.513 Xi X3 3/36 2^*31 .204 810.401 4.612 <. 01 .527 Res i dua 1 6326.396 175.733 Xi X2 2/37 2282.625 1141 .313 6.522 < . 005 .510 Res i dual 6474.975 174.999 Xi 1/38 2201 .969 2201 .969 12.764 <. 005 .501 Res i dual 6555.631 172.517 Xi = Test of Bas i c Experiences-Science (TBES) X2 = Test of Bas ic Exper iences-Mathematics (TBEM) X3 = Test of Bas i c Exper iences-Social Studies (TBESS) Xij = Test of Bas ic Experiences-Language (TBEL)

PAGE 204

Re] iabi 1 i ty Three measures of reliability were examined in this investigation to determine the consistency or precision of measurement associated with FICCS. These included a measure of internal consistency and measures of intraand inter-rater reliability. Internal Consistency Internal consistency refers to "estimates of reliability based on the average correlation among items within a test" (Nunnaly, I967, p. 210). It is a measure of the homogeneity of test items and in this investigation is represented by coefficient alpha (r ). According to Nunnaly (1967), "coefficient alpha sets an upper limit to reliability" (p. 210) and should be applied to all new measurement strategies prior to obtaining other estimates of reliability since reliability estimates obtained using coefficient alpha are highly similar to those estimated from other techniques and suffer fewer of the disadvantages. In most situations, coefficient alpha provides a good estimate of reliability since the major source of measurement error associated with the statistic is due primarily to the sampling of content and secondarily to specific situational factors (Nunnaly, I967). Table 55 presents a summary of the reliability analysis for all subjects on the Functional Inventory of Cognitive Communication Strategies . As is shown, the inter-item correlations between the 2k individual communication strategies ranged from -.^7 to .65, with a mean r_ of .13, indicating that in general, there is a negligible relationship among the items on FICCS. Each item appears to measure a unique aspect

PAGE 205

195 Table 55 Summary of the Reliability Analysis on the Functional Inventory of Cognitive Communication Strategies (FICCS) SCALE NUMBER OF ITEMS CRONBACH'S ALPHA INTERITEM CORRELATION RANGE SQUARED MULTIPLE CORRELATION RANGE REPORTING 8 .46 -.4 .43 . 15 .39 LOGICAL REASONING 6 .52 -.17 .56 .17 .44 PREDICTING 6 .45 -.38 .37 .15 .30 PROJECTING k -.04 -.18 .20 .05 .10 LOWER-LEVEL STRATEGIES 7 .41 -.47 .40 .35 .58 HIGHER-LEVEL STRATEGIES 7 .70 -.10 .61 .49 .67 FICCS OVERALL 24 .72 -.47 .65 .51 .87

PAGE 206

196 of language use and has little in common with the other items on the scale. The squared multiple correlation range indicated that the contribution of individual items to the internal consistency of the scale ranged from 51 to 87 percent. Overall, the alpha coefficient for the 2k communication strategies on FICCS was .72, which is indicative of a marked degree of internal consistency overall. When FICCS was further subdivided into its four language use subscales--Reporting, Logical Reasoning, Predicting and Projecting — reliability estimates of internal consistency were markedly diminished. In part due to the reduced number of items in each subscale. Table 50 summarizes the reliability analysis for all subjects on the four major language use subscales. On the Reporting subscale, inter-item correlations ranged from -.^0 to .^3, with a mean £_ of .12, indicative of a low degree of relationship among items in the scale. Inter-item correlations between the items on the other subscales ranged from -.17 to .56, with a mean of .17On Predicting, £ values ranged from -.18 to .20, with a mean £ of -.01. The squared multiple correlation ranges for the subscales were as follows: Reporting (_R^ = .15 to .39), Logical Reasoning (R^^ = . 1 7 to .kk) , Predicting (R^^ = .15 to .30), and Projecting (R^^ = .05 to .10). The items on the Reporting and Logical Reasoning subscales appeared to make the greatest contribution to the internal consistency of the subscale, while the items on the Projecting subscale made the least. Because of the reduced number of items on each of the subscales and because the individual items appeared to measure very different aspects of language use, the coefficient alpha for each of the subscales was notably diminished: Reporting (r = .46); Logical Reasoning (r = .52); Predicting (r = .45); and Projecting (r

PAGE 207

.ok). In the case of the Reporting, Logical Reasoning and Predicting subscales, a moderate degree of internal consistency was demonstrated; however, no pattern of internal consistency was observed relative to the Projecting items. It should be noted, however, that the Projecting slibscale contained the fewest number of items of any scale. In an attempt to further isolate some of the factors accounting for the reduced reliability on FICCS, a split half reliability coefficient was computed for the lower versus higher strategies on the inventory. The analysis resulted in an r^^ of .41 for the lower-level strategies and an r^^ of .70 for the higher-level strategies, with a correlation between the two types of strategies of .03. These findings suggest that there is more item variance relative to overall performance on the lower-level strategies than on the higher-level ones and that on future samplings, the likelihood of obtaining a similar score on the lower-level strategies is approximately 16 percent, while there is a kS percent likelihood of achieving a similar score on the higherlevel strategies. This pattern of performance appears to be maintained for each of the four language use subscales with the exception of Projecting. The coefficients alpha for the lower-level strategies were as follows: Reporting, = .^5; Logical Reasoning, .31; Predicting, = .12; and Projecting, -.39 r ; while the coefficients for the XX higher-level strategies were: Reporting, r^^ = .46; Logical Reasoning, .kk; Predicting, r = .kO; and Projecting, r = .25. Finally, the respective correlations (r) between the lowerand higher-level strategies on each of the subscales were .11, .35, .35, and -.01.

PAGE 208

198 Intra-rater Reliability Intra-rater reliability refers to the extent to which a rater will score the same performance in the same way on two different occasions. To establish an intra-rater reliability index, four language samples selected at random, one from each of the subject groups, was rescored by the investigator. A total of 22k communication strategies was reanalyzed using a point-by-point percentage of agreement formula: NUMBER OF AGREEMENTS NUMBER OF AGREEMENTS + DISAGREMENTS ^ The obtained percentages represent the ratio of agreed observations to total observations, using each analyzed strategy as an observation. The results of this analysis yielded the following percentages of agreement for samples 1 to k: 86, 88, 90, and 79. The mean overall intra-rater reliability index was 85-5, which is indicative of a high degree of scoring precision by a single rater. inter-rater Reliability Inter-rater reliability refers to the extent of agreement between different raters classifying the same phenomena. An inter-rater reliability index for FICCS was established between two raters following the same procedure described for determining the intra-rater reliability index. Using the point-by-point percentage of agreement formula, the following percentages of agreement for samples 1 to ^4 were generated: 65, 72, 71 and 76. The mean overall inter-rater reliability index was 71.0, suggesting a high degree of scoring consistency between two raters.

PAGE 209

199 Summary of Findings The results of the group comparisons on FICCS and the correlations between predictor and criterion variables revealed the findings that follow. As shown in Table 56, the LNl group achieved significantly superior performance on the FICCS overall functional language score and the major use scores of Logical Reasoning and Projecting, as compared to their 1 anguageimpa i red peers (Ll), matched for chronological age. Significant differences were not observed between the groups in the use of language for Reporting or Predicting purposes. Out of the 2A individual communication strategies evaluated by FICCS, only four were used in a significantly different manner by LI and LNl, including strategies 7, 13, 18 and 22 (see Table 57). In each instance, the LNl group used these strategies with greater frequency than did the LI group. Finally, when the strategies were further subdivided into lowerand higher-level strategies, the LNl and LI groups used comparable numbers of lower-level strategies overall and relative to each major use of language, but differed significantly in their use of some higherlevel strategies. In particular, statistically significant differences were observed in the use of higher-level strategies overall. Reporting and Logical Reasoning. A summary of this analysis is presented in Table 58. A comparison of the languageimpa i red (Ll) and language-normal (LN2) groups, matched for utterance length, showed no significant differences between the groups on FICCS in the overall functional language score and the major use subscores (see Table 56). Statistically

PAGE 210

200 Table 56 Summary of Group Comparisons for the Overall Functional Language Scores and Subscale Scores on FICCS GROUPS OVERALL REPORTING LOGI CAL REASONING PREDICTING PROJECTING LI LNl * Li LN2 LN6 LN7 * LSES HSES *Signiflcant differences between groups LI = LanguageI mpa i red Group LNl = Language-Normal Group Matched for Chronological Age LN2 = Language-Normal Group Matched for Utterance Length LN6 = Language-Normal 6-Year-Olds LN7 = Language-Normal 7-Year-Olds LSES = Lower Socioeconomic LanguageNormal Group HSES = Higher Socioeconomic Language-Normal Group

PAGE 211

20) « 3 D "> o c o o u. 4-1 c o o >*V in o> c 0) o 4-1 l/> U 3 C3 < -atM CM CM CM O CM 00 SO oo to 0. 3 — O — Z OC _J -J \0 Ul US 00 00 Q. 3 o iO) c c (TJ O c 0) < — c 11I a> O 0 Z Z I 1 t) I/) Oi (TJ T3 nj 3 3 cn CD C C (D (D -J TJ (71 Dl fU (TJ (U (TJ 3 3 3 3 C71 CT> o> cn c c c c (TJ fQ nj (TJ u — E o c o
PAGE 212

202 CO k. to ILI C3 < C3 O ex: C3 O UJ en a. o —t z < — — o t3 CO O < o o. < LU o a. to LU < > UJ _i I LlI o (_> o LU t3 —1 z < — CJ z — o C3 to o < o a. LU OC < > o to a. =) o L9 — Z CM to UJ to to 3: in CL 3 o u c (U (U 2 4-> (U -Q tfl 0) u 0) u ai j: < w — C (D 4> U —I Ol «> O O >— c O (D C >O 0) t> JC -M o :z) o o T3 -D 0) 0) ^ j: in O U TJ M +J — (D UJ O Q3: s: I 3 10 O. Q. 03 13 3 0) C3 O O >iiI "O O C3 vO •— fO fU 03 03 E E E Q. U I1_ E O O O — z z z 1 I I I (U 0) (U 0) O) D) C7) 03 03 03 03 3 3 3 3 C7) CD Dl CJ) C C C C 03 03 03 03 a O. 3 3 O o i— 03 03 E e 1>o o z T (u 1/1 C?) 03 O 03 3 3 a> C3^ C C 03 03 -I U *i o o c — CO OJ o o E O 03 O — II II II II II II II — — CM vo to to — I Z Z Z Z LU LU -J —J —I —I to to -I X

PAGE 213

203 significant differences were noted in the use of five individual communication strategies, 8, 19, 20, 23 and 2k, with the LI group using more examples of each than the LN2 group. These results are summarized in Table 57. Significant differences were also obtained between the groups for use of higher-level strategies, whereas use of lower-level strategies was similar. As summarized in Table 58, the LI group used significantly more higher-level strategies overall, and in particular, Predicting and Projecting strategies, than did the LN2 group. A comparison of the language-normal 6(LN6) and 7" (LN7) yearolds revealed that with one exception, there was no significant difference in the overall functional language score or the major subscores of language use on FICCS. As shown in Table 56, the exception occurred relative to the use of Projecting strategies, with the LN6 group producing significantly more of these strategies than the LN7 group. Table 57 shows a significant difference in the use of three individual communication strategies, 3, 7 and 23, between the two groups, with LN6 using more examples of strategies 3 and 23, and LN7 using greater numbers of strategy 7No significant differences in the use of lowerand higher-level strategies were observed for the groups. Differences in language use between the lower (LSES) and higher (HSES) socioeconomic groups were minimal. Table 56 shows no significant differences between the groups on the overall functional language score and the four major use subscores of Reporting, Logical Reasoning, Predicting and Projecting. Only one of the individual communication strategies, strategy 9, differentiated between the groups to a statistically significant degree, with HSES using greater numbers of this strategy type than LSES. On the lowerand higher-level strategies (see

PAGE 214

204 Table 58) the groups differed significantly in the use of lower-level Logical Reasoning strategies, with LSES using these strategies more f requent ly . The results of the correlation and regression analysis revealed that of the three standardized measures of language content and/or form, only the TACL was significantly correlated with performance on FICCS {r_ = .44) and significantly contributed to the prediction of performance on FICCS. The correlation between FICCS and the other two variables, TOLD and CELI, was not significantly different from zero, and neither of these variables made a statistically significant contribution to the prediction of performance on FICCS. The correlation coefficients between FICCS and the eight nonstandardized measures of language form were all statistically significant, positive correlations ranging from r_ = .35 to .86. When the data were subjected to a stepwise linear multiple regression, it was observed that the combination of two variables, #CU's and MLU-W, could predict performance on FICCS almost as well as the combination of eight variables. Finally, the correlations among the four measures of academic achievement and FICCS were all statistically different from zero and positive: r_ = .36 to .50. The regression of FICCS on these measures revealed that only the TBES significantly contributed to the prediction of performance on FICCS; the addition of TBEL, TBEM, and TBESS to the equation added little to its predictive accuracy. Results of the reliability analysis revealed a marked degree of internal consistency for overall performance on FICCS (r = .72).

PAGE 215

Intraand inter-rater reliability also achieved an acceptable level of precision for this investigation at 85.5 and 71.0, respectively.

PAGE 216

Chapter 5 SUMMARY AND CONCLUSIONS Summary The purpose of this study was (1) to determine the influence of language ability, age and socioeconomic status on children's language use; and (2) to determine the relationship and predictive accuracy between measures of linguistic performance, academic achievement and language use. Subjects were ten 1 anguagei mpa i red children and thirty languagenormal children, all monolingual speakers of average intelligence with no gross peripheral defects of audition or vision. The children's use of language was evaluated with the Functional Inventory of Cognitive Communication Strategies (FICCS), a structured interview designed to elicit typical strategies of language use in response to a series of pictures. The framework for the classification of the cognitive uses of language, developed by Tough (1976, 1977), served as the basis for the creation of the verbal probes on FICCS. At the broadest level of classification in this schema, utterances are categorized, by use, into four major cognitive types: Reporting, Logical Reasoning, Predicting and Projecting. These four basic uses are exhibited in children's speech through 2h individual communication strategies. To examine the effects of language ability on language use, the languageimpai red children were matched to two groups of language-normal 206

PAGE 217

207 children, one on the basis of chronological age and the other on the basis of utterance length. When matched for chronological age, the language-normal children achieved significantly superior performance on the FICCS overall functional language score and the major use scores of Logical Reasoning and Projecting as compared to their languageimpaired peers. No significant differences for these measures was observed between the 1 anguageimpa i red group and the language-normal group matched for utterance length. These findings seem to support a general developmental delay in language use; however, trends in the data are indicative of a disorder rather than a delay. The 1 anguagei mpa i red group achieved higher overall scores on FICCS than did their younger, normal counterparts matched for utterance length, which suggests that the communicative function of the impaired children is better than their linguistic skills would imply. The influence of age and socioeconomic status on language use was evaluated using a 2 x 2 factorial design, with two age levels (6 years plus or minus 3 months, and 7 years plus or minus 3 months) and two social class levels (lower and higher). Ten children comprised each group. No significant differences were observed for the overall functional language score on FICCS and the subscores of Reporting, Logical Reasoning and Predicting, nor was the two-way interaction of age by socioeconomic level significant. On Projecting strategies, the language-normal 6-year-olds achieved significantly higher scores than the language-normal 7-year-olds. This general absence of significant findings may be related to the small sample size, the modest difference in age between the two groups or the fact that both the lower and

PAGE 218

208 higher socioeconomic groups attended the same schools and lived in the same rural communities. The relationship between performance on FICCS and three standardized measures of language content and/or form, including the TACL (Carrow, 1973), the TOLD (Newcomer 6 Hammi 1 , 1977) and the CELI (Carrow, 197^), was investigated for all ^0 subjects. Although the standardized measures were significantly i ntercorrelated, only performance on the TACL was significantly correlated with performance on FICCS to a moderately positive degree. These findings suggest that FICCS measures a communication skill which is not measured by the standardized instruments investigated in this study and underscores the importance of adding a measure of language use to a diagnostic battery since children's linguistic performance on standardized measures is not necessarily reflective of their language use--what they can do with the linguistic abilities that they have. The relationship between performance on FICCS and eight nonstandardized measures of linguistic form, including #CU's, TNW, MLU-W, TNM, MLU-M, MLU5-M, %CUc and DSS, was also explored for all kO subjects. These data revealed statistically significant positive correlations between FICCS and each of the nonstandardized linguistic measures. One explanation for the higher correlations between these measures and FICCS may be that they were all derived from the same language sample. Another explanation may be that spontaneous language sampling preserves the interactive nature of communication and therefore provides a more powerful correlate of language use than the acontextual standardized tools of linguistic content and/or form.

PAGE 219

209 Performance on FICCS and four measures of academic achievement, including TBEM, TBEL, TBES and TBESS, was correlated and indicated a TOderate positive relationship between the measures which was statistically significant. The shared variance between the measures ranged from a low of 13 percent to a high of 25 percent, suggesting that factors other than language use contribute to a significant extent to success in the classroom. Reliability estimates on FICCS were indicative of a marked degree of internal consistency overall (r^^ = .72) and a high degree of intrarater (.86) and interrater (.71) reliability. Pi scuss ion The influence of language ability on language use was not clearly resolved by the results of this investigation. Although the data demonstrated that languageimpa I red children performed more poorly than their language-normal age-mates on the overall functional language score and two out of four subscores of language use, the differences in performance on these measures between the languageimpa i red and language-normal children matched for utterance length were not as clear-cut. On every measure, the languageimpa i red children achieved higher scores than their younger, normal counterparts and on some measures, these results approached significance. It was also interesting to note that the three groups used comparable numbers of lowerlevel strategies on FICCS, but LNl used significantly more higher-level strategies overall than LI, and LI used significantly more higher-level strategies overall than LN2. These findings appear to indicate that with increased sample size, the nature of the impairment in language

PAGE 220

210 use should be characterized as a "disorder" rather than a "delay." These 1 anguageimpa i red children were capable of using language in much more sophisticated ways than their knowledge of linguistic content and form suggested, and yet they were not as accomplished as their colleagues at the same age; thus there appeared to be a fundamental mismatch between what they could say grammatically and what they could say pragmatically. This finding was not surprising when Bates' (1976a) observation that children arrive at conventional forms of communication by mapping semantics and syntax onto the already acquired pragmatic system was considered. The ability to function or to intend is learned prior to speech itself and forms the foundation on which the lexicogrammatical system is built. In the few studies which have compared the language use of linguistically deficient children to that of language-normal children, results have been mixed. When utterance length was controlled, Snyder (1975) found language-impaired children at the one-word stage of development to be more deficient in language use than their mastery of vocabulary and syntactic structures would indicate. Geller and Wol Iner (1976) observed similar deficiencies in the language use of older language-impaired children (3.11-5 years) when results of their study were compared to those of Dore (1975), collected for 3-year-olds in a similar communicative context. While Snyder's (1975) findings are suggestive of a "disorder" in the use of language, Geller and Wollner's (1976) results indicate a "delay" since no attempt was made to match the linguistic level of their subjects to those of Dore's (1975). In contrast, the results of three other investigations suggested similar patterns of language use in the languageimpai red and

PAGE 221

211 language-normal populations studied. Owens (1978) compared the language use of mentally-handicapped children and normal children at Brown's (1973) stage I and II and found essentially the same speech act distribution among the four groups. Skarakis and Prutting (1977) and Curtiss et al . (1979) investigated the young hearingimpai red child's ability to communicate, either verbally or nonverbal ly, a variety of pragmatic intentions. Overall, they found that hearingimpai red children coded a variety of communicative intentions using both verbal and nonverbal means and similar to the present investigation, they noted that the number of communicative intentions expressed increased with age, even though mean length of utterance remained the same. This finding seems to indicate that competence in language use or the pragmatic aspect of language precedes development in the other components of language and provides the foundation on which these later linguistic skills are constructed. In the present study, the older 1 anguageimpa i red children knew how to function but they did so with impoverished systems of form and content. Due to limitations in the sampling protocol, the influence of age and socioeconomic status on language use was not clearly resolved. The effect of age on language use was measured by comparing the performance on FICCS of children at two different levels: 6 years (plus or minus 3 months) and 7 years (plus or minus 3 months). When differences between the two groups on the overall functional language score and the major subscores of language use were analyzed, only use of Projecting strategies achieved significant differences, with the 6-year-olds using more of this type of strategy than the 7-year-olds. Two factors may have contributed to the lack of significant findings observed between

PAGE 222

212 the age groups. First, the difference in age between the two groups was modest, ranging from 6 to l8 months. This limited distinction between the groups may have been insufficient to achieve significant differences in performance. Most studies of a developmental nature either span a wider age range or allow for a greater discrepancy in age between groups, usually ranging from 18 to 30 months. Second, children may reach a high level of performance on the communication strategies studied and plateau just prior to or immediately beyond the age levels under investigation. Tough (1976) noted that although 3-year-old children varied greatly in their verbal output, their use of language already reflected complex thinking. Schachter et al. {]37^) observed a pivotal shift in language use at the age of 3, while Fogel (1976) found a change in language patterns reflected in increased diversity of speech act usage at approximately 10 years of age. Schachter et al . (197'») also noted that after age 3, early or lower-level communication strategies maintained their frequency but strategies representing increased levels of complexity showed a marked rise. While the results of the present study are in agreement with Schachter' s et al . (197't) findings on lower-level strategies, increased use of more complex strategies was not observed, possibly because this change in language use patterns would not be expected to occur until a much later developmental stage relative to the classification schema involved. In the present study, social class had no influence on children's language use. This finding is in direct disagreement with Tough (1977) and others who have found a different orientation toward language use in children from lower and higher socioeconomic levels. Bernstein

PAGE 223

213 (1971) has suggested that children who are brought up in different environments, exposed to different attitudes and values and to different outlooks on the world in general are not only developing different ways of viewing the world but are also building up different orientations toward the use of language, which reflect differences in the organization of meaning of their experiences. The children included in the present study all attended the same schools and resided in small rural communities in the mountainous northwestern region of North Carolina. The isolated nature of the communities themselves may have obscured any findings relative to social class. In these isolated mountain areas, it appears that families are more similar than different. They pursue a way of life regardless of occupation and material wealth and this is the heritage which they pass down to their children. For this reason, the continuum on which social class differences is based is necessarily restricted. The children in the present study did not represent the extremes along the continuum of social class; therefore differences in performance between groups would not be as likely. The study of the relationship between social class and language use has been subject to sampling problems. There are some who believe that the critical variables on which to stratify subjects for the purpose of studying emerging language use are the type, quantity and quality of verbal stimulation in the home and not socioeconomic status. Within a given socioeconomic level, the range of environmental stimulation may be so great as to make any generalizations about SES and development extremely tenuous. Adler (1979) reports much of the available information on home environment of the culturally different has been obtained through retrospect ivetype questionnaires and the

PAGE 224

reliability of such a data gathering technique is suspect. At this time, the relative benefits or disadvantages of the type, quality and quantity of verbal stimulation in culturally different homes is unknown. The study of the relationships of early verbal stimulation to the emergence of language use has implications for intervention. The early experience model postulates that early stimulation serves to promote or retard development of certain processes or functions which are crucial for later development. inherent in this view is the notion that early identification and management of inadequate stimulation strategies will facilitate the emergence of competence for language use. The unit of analysis on which comparisons were made represents another possible cause of the outcome of this study. In Tough's (1977) original work, the language use of children from disparate social environments was compared using frequency of occurrence data; that is, the number of times a child used each of the individual communication strategies was counted and those strategies on which one group scored two or three times as often as the other were noted. In the present investigation, interval scores were used to neutralize the distorting effects of either the talkative or quiet child. By eliminating the extremes of verbal productivity, this study showed essentially the same use of language by children of lower and higher socioeconomic status on the task presented. It appears that the interval score as adopted in the present investigation may more truly measure aspects of competence for language use while frequency of occurrence data and percentage conversions reflect language use performance. If children are capable of using

PAGE 225

215 specific communication strategies, even on a limited basis, they are demonstrating competence or knowledge about the use of those strategies. Future studies should compare all three measures of language use — the interval score, frequency of occurrence and percentage conversion on the same group of children to examine the relationship between the measures and to determine which measure provides the rrost accurate estimate of children's language use skills. Tough's (1977) results and the results of the present investigation seem to indicate that children from lower and higher socioeconomic levels are capable of using language for the same cognitive purposes; however, for some reason, lower socioeconomic status children choose not to use the language of which they are capable as frequently as their higher socioeconomic status counterparts. In fact, Tough (1977) observed that through a series of prompts, directive questions and comments, children of lower socioeconomic status could be led to interpretations which other children impose spontaneously on situations. These children have the resources with which to make complex interpretations, but either they are not aware of their own knowledge or they do not see the necessity for expressing it. Performance on FICCS was correlated with standardized and nonstandardized measures of language ability and measures of academic achievement to explore the relationship between these measures and a measure of language use. When performance on FICCS was compared with standardized measures of language content and/or form, including the TACL, TOLD and CELI, only the TACL was significantly correlated with FICCS to a moderately positive degree. This suggests that FICCS measures a skill not evaluated by the currently available tests of

PAGE 226

216 language content and/or form. Further, it implies that children may be competent users of language even though the linguistic system which they employ is deficient semantically and syntactically. These results would strongly support the addition of an assessment of language use to the clinician's diagnostic battery, as current standardized measures of language ability provide limited predictive accuracy relative to language use. The relationship between performance on FICCS and the nonstandardized measures of linguistic form was of greater magnitude. All eight measures, including #CU's, TNW, MLU-W, TNM, MLU-M, MLU5-M, ^CUc and DSS, achieved statistically significant positive correlations with FICCS. The fact that the measures of language use and the measures of linguistic form were derived from the same language sample may have accounted for the elevated correlations among the measures. Also, the interactive nature of conversation is better preserved through spontaneous sampling and provides a more powerful correlate of language use than the standardized tools which are void of natural linguistic and non 1 ingui stic contexts. It was interesting to note that FICCS was not as highly correlated with measures of linguistic complexity as it was with measures of verbal productivity. This finding lends support to the hypothesis that children can be linguistically deficient and still be competent in language use. However, unless they are verbally productive, competence in language use will be difficult to measure. The relationship between performance on FICCS and the four measures of academic achievement were all positive, ranging from a low to moderage degree. This finding seems to indicate that while language

PAGE 227

217 use makes a significant contribution to academic success, it is only one of many potentially influential factors. Conclusions and Suggestions for Future Researcii Competence in the use of language is essential for social and educational growth. Children whose use of language is inadequate for the conduct of everyday interpersonal and academic communicative exchanges will not be at home in the educational world nor in society at large. In a very general and deep sense, educational failure is often language failure. Children who do not succeed in school may not be using language in the ways required by the school. According to Halliday (1973), this does not mean that these children cannot read or write or express themselves acceptably, but rather that there exists a more profound problem — a general mismatch between their linguistic capabilities and the demands that are made upon them. Bloom and Lahey (1978) have described two characteristics of competent language use: (I) the progressive mastery of a number of basic functions of language; and (2) the ability to adapt communication to the demands of the communicative situation. Inherent in this notion of competence for use is the obligatory component of functionality — to be competent, language must "wor k" for ch i I dren . As Halliday (1978) has expla i ned. Being 'appropriate to the situation' is not some optional extra in language; it is an essential element in the ability to mean . . . Our functional picture of the adult linguistic system is of a culturally specific and s ituational ly sensitive range of meaning potential. Language is the ability to 'mean' in the situation types, or social contexts that are generated by the culture. When we talk about 'uses of language,' we are

PAGE 228

218 concerned with the meaning potential that is associated with particular situation types. (p. 3^) FICCS was designed to identify the wide variety of language used and strategies reflective of children's growing ability to use the language system to convey meaning. Based on Tough's (1976, 1977) bi level cl ass i f icai ton of language use, it evaluates in a standardized, efficient, and reliable way the cognitive communication strategies used by language-normal and 1 anguagei mpa i red children; however, it is not without its limitations. Chief among these is the fact that FICCS assesses only the use of cognitive communication strategies while ignoring the use of social strategies. Clearly, a comprehensive assessment system of language use would need to recognize and evaluate the dual functions of communication--cogni tive and social. Secondly, FICCS employs a single sampling strategy, the structured interview, to elicit all information on language use. Since specific parameters of the communication situation influence to a great extent the representativeness of the communication samples obtained, it seems advisable to gather information on language use in response to a variety of sampling strategies reflective of the diverse nature of the communicative exchange. Finally, FICCS measures a child's habitual mode of language use during a structured storytelling task. No attempt was made to insure the use of each of the cognitive communication strategies in Tough's (1976, 1977) system, although the questions employed provided approximately equal opportunities to respond in the Reporting, Logical Reasoning, Predicting and Projecting modes. Thus, if one or another of the individual strategies was not used by a child, it was impossible to judge whether the child was incapable of using it or whether the

PAGE 229

219 sampling conditions did not oblige its use. In a positive sense, FICCS provided an adequate description of children's typical use of language. It did not, however, characterize their true communication potential. FICCS described what children did with language, but not what they could do . As a result of the present investigation of FICCS, the following suggestions are made for future research: 1. An assessment tool of language use should be developed which measures the social functions of language as well as the cognitive. These functions should be evaluated within a number of changing contexts that control for influential factors of language use: the participants, the setting, the topic, and the goal of the communication. 2. Items on the tool should be constructed so that each communication strategy in the functional taxonomy upon which the instrument is based will have an equal number of opportunities of being measured. 3. The development of language use, with particular reference to the acquisition of a wide variety of communication strategies, should be traced over time at two-year intervals with a larger sample. k. The relationship of social class to language use should be investigated by stratifying the sample on the amount, type, quantity, and quality of verbal stimulation within a given socioeconomic class and not by a simple dichotimization of children into lower and higher social class groups. 5. The efficacy of a stimulation program for the development of competence in language use should be investigated. The goal of the stimulation program would be to enhance the ability of children from different cultural backg rounds to analyze communication situations, to

PAGE 230

220 predict outcomes of potential communication strategies and to encode effectively. This would not entail "making all children behave in similar ways in sets of situations, but rather would enable all children to perceive what is likely to be appropriate in given situations within given cultures" (Brown, Ecroyd, Hopper & Naremore, 1976, p. 155). The essence of these suggestions for future research is the development of a comprehensive system with strategies for the measurement and management of language use that are logically consistent and evolve out of the same theoretical framework. Such a system is not currently available, although a number of guidelines exist for its development.

PAGE 231

APPENDIX A FAMILY SIZE AND INCOME SCALES FOR FREE MEALS AND REDUCED-PRICE MEALS

PAGE 232

APPENDIX A FAMILY SIZE AND INCOME SCALES FOR FREE MEALS AND REDUCED-PRICE MEALS The following income scales* were used by the Avery County Schools and the Wilkes County Schools, respectively, to determine el bility for Free and Reduced-Price Meals in the 1979-1980 school year Avery County Schools Family Size Free Meals Reduced Meals I 0 k,5S0 ^,591 7,190 2 0 6,0^10 6,040 9,420 3 0 7,'i30 7,^*91 11.680 k 0 8, 9^*0 8,hS] 13,940 5 0 10,390 10,391 16,200 D f\ 11 Oil A 11 Q /i 1 II , oH 1 1 O , M/0 7 0 13,290 13,291 20,730 Q O 0 14,7^0 li»,7Hl 22,990 Each additional family member 1 , ^50 2,260 Wilkes County Schools Fami ly Si ze Free Meals Reduced Meals 1 0 k,]SO ^•,191 6,530 2 0 5,500 5,501 8,530 3 0 6,810 6,811 10,630 k 0 8,110 8,111 12,660 5 0 9,310 9,311 14,530 6 0 10,510 10,511 16,400 7 0 1 1 ,600 11,601 18, 100 8 0 12,690 12,691 19,790 9 0 13,680 13,681 21 ,330 10 0 U,660 14,661 22,870 n 0 15,6A0 15,641 24,390 12 0 16,610 16,611 25,910 Each additional family member 970 1,520 "Income scales in dol lars . 222

PAGE 233

APPENDIX B TOUGH'S FRAMEWORK FOR THE CLASSIFICATION OF LANGUAGE USE

PAGE 234

APPENDIX B TOUGH'S FRAMEWORK FOR THE CLASSIFICATION OF LANGUAGE USE Operational Definitions and Examples REPORT I NG-the use of language to provide information about past and present experiences. A. Label i ngincl udes utterances which serve the simple purpose of identifying observed phenomena. 1. gates 2. houses 3. playing k. That's the baby. 5. I see some trees. B. Referr i ng to Detai 1 — includes utterances which serve to describe the criterial attributes of objects, actions and/or events . 1 . It was a big cat. 2. There's a little path. 3. The car is going real fast. k. The dog is in the road. 5. The mother cat's got one eye closed and the other eye looking at them (kittens). C. Referring to Incidents — includes utterances which describe the occurrence of an action or event. 1. That little kitten's going out the door. 2. A dog's chasing the black kitten down the lane. 3. That cat's running across the road. 4. He brought the little kitten home to that family. 5. They gave the cat something to eat. D. Referring to the Sequence of Eventsincludes utterances which accurately reflect the serial nature of several related actions or incidents. 1. A man stopped and picked up the little black kitten. 2. When that baby kitten went out the door, the dog saw him and started chasing him. 224

PAGE 235

225 3. The cat and dog are going to get something to eat. Then, they're going back home. A. I went inside the store and I couldn't find her. Then I f i na 1 ly found her. 5. Dad stopped the car and got out and he tied the wire on the muffler and hooked it on the bottom of the car. E. Making Comparisons — includes utterances which link objects, actions or experiences through examination of similarities and di f ferences. 1. The dog's bigger than the kitten. 2. The kitten is running faster 'n the dog. 3. Snow is on the trees but not on the bushes. k. The little black kitten doesn't want to get in trouble like he did over at that house with all the kittens over there. 5. My kitty's not real; it's a play one. F. Recognizing Related Aspectsincludes utterances which posit an association between two or more actions or events. 1. I got my cat so ticklish that it scratched me. 2. When my nose is running, I wipe it off. 3. I jumped on my bed and it fell down. h. We usually take something for me to lay down on so I can go to sleep as we're riding. 5. i used to have a kitty cat named Noel ' cept it ran away. G. Extracting or Recognizing Central Meaning — includes utterances which impose a primary structure or coherence upon a situation or event and serve to unify the contributing parts into a composite whole. 1. I think she's mad cause all the cats came in and drank up the milk and started messin' up. 2. The dog is chasing the cat and if it chases it out in the road, they might both get run over. 3. The man's stopped to pick up the little black kitten so he won't get run over. k. Everybody's excited now because Dad brought the little black kitten home. 5. The family's having dinner and here comes along the lady hunting for the little kitten. H. Reflecting on the Meaning of Exper i encesi ncl udes utterances which express the speaker's attitudes or feelings about a s i tuat ion . 1. When I fly up in an airplane, it scares me to death. 2. I wouldn't keep it cause I don't like animals. 3. We were sad cause we left the hamster there overnight. h. That kitten's a sweet one. 5. it isn't nice for the dog to chase the kitten.

PAGE 236

226 II. LOGICAL REASONING — the use of language which employs rational thought and argument to interpret experiences. A. Explaining a Processincl udes utterances which described a particular method of doing something, generally involving several steps or operations. 1. I'd feed the kitten and put it in a box and put a little blanket in so it could lay down. I'd play with it too. 2. You should stop and look to each side and see if a car is coming and if a car isn't coming, you might go across. 3. The muffler started draggin' and Dad had some wire in the car and he stopped the car and got out and he tied the wire on the muffler and hooked it on the bottom of the car. That held the muffler up. k. Wait 'til the cars go by. Look in both directions. Then if there's nothing coming, you can walk across the road. 5. When the ocean comes, we get us some water and take it up there and get the seashells and drop them in. We wash them off. B. Recognizing Causal and Dependent Rel at ionsh i psincl udes utterances which acknowledge a logical and relevant connection between two situations and which express this most commonly in terms of "how" and "why." 1. I'm sleepy from waking up too early in the morning. 2. They're gonna get in the house cause she left the door open . 3. Both kittens got lost cause they went outside. k. The dog' 11 probably catch up with the kitten cause it stopped. 5. He doesn't want the kids to shout cause it hurts his ears. C. Recognizing Problems and Their Sol ut ionsi ncl udes utterances which acknowledge obstacles to a course of action and suggest ways to surmount them. 1. He is picking up the cat so it won't get run over by the traf f i c. 2. The man's holding the little black kitten so the children won't grab it and hurt it. 3. We can't have kittens inside cause my mama's allergic to ki ttens. k. It'll be real hard to find its owner so I'll keep it. 5. I'd make it a little house and lock it in so no one'd take it. D. Justifying Judgments and Actions — includes utterances which offer a reason or explanation for decisions and behavior which apply only to a particular situation. 1. She might' ve been to church cause she has on her good clothes.

PAGE 237

227 2. She might 've washed the windows cause they're real clean. 3. Some of the cats are going back out because they want to go back to their own house. 4. He looks like a carpenter cause he's got on overalls. 5. Every time a cat gets in the kitchen, mama scats it out cause I'm allergic to cats. E. Reflecting on Events and Drawing Concl us ionsi ncl udes utterances which evaluate the implications of an action or event and result in judgments. 1. The kittens have made a mess on the table and it might ruin the tablecloth. 2. The cats are the nextdoor neighbor's. They left the doors open and the cats came in and they're eating up all the food . 3. It's good that the man has stopped so the dog won't catch the little black cat. k. He's taking it inside so maybe he can give it to his chi 1 dren, 5. It's good to have a cat so what the children don't eat, it can eat. F. Recognizing Pr i nci pi esi ncl udes utterances which provide an elemental rule or rules to explain observed phenomena. 1. The cats shouldn't climb up on the table. 2. You shouldn't run across the road without looking. 3. The cats are not s'posed to be in the kitchen tearing up. k. When you get off the school bus, you shouldn't walk behind it; always walk in front of it. 5. Most cats know to climb trees when something's after them. PREDI CTI NG — the use of language to extend communication beyond immediate present or past experiences to events that have not yet occurred and which may never take place, A. Anticipating/Forecasting — includes utterances which contemplate future happenings or remote concerns. 1. The lady's gonna clean up the kitchen. 2. The dog might catch up with the cat. 3. That man's gonna find out where the cat lives. k. The lady might come in the house. 5. The cat' 11 never run off again. B. Anticipating Detail of Actions and Events — includes utterances which delineate or describe future happenings or remote concerns. 1. The kitten might turn around and run back out through here and back up through the bushes back up to its house. 2. The cat's gonna get in the milk truck.

PAGE 238

228 3. The kitten wants to go off on another walk by himself. k. The kitten' 11 tear up the house and tear the curtains down and eat up all the food and make the baby's chair fall down . 5. My dad's gonna get me a brand new bicycle. C. Anticipating a Sequence of Events — includes utterances which propose an ordered series of related actions or events. 1. I'd try to clean up all the mess they made, then rest. 2. Then she will hit the kittens and make them go out. 3. This dog will chase the cat and then he'll hide somewhere. k. She's gonna let those cars go, then she's gonna run across . 5. The bus is gonna stop and he's gonna get off and get his cat and get back on. D. Anticipating Problems and Possible Solutions — includes utterances which acknowledge possible obstacles to a planned course of action and suggest ways to surmount them. 1. I think she'll put the kittens in a box so they can't get out . 2. If she couldn't get rid of the cats, she would call the pol i ce. 3. It looks like she was having somebody over for dinner so she would call them back and tell them that everything's messed up. A. A whole bunch of people might get the kitten and handle it a lot and if it's not very old, the mama might not claim it. 5. They might build a fire for the cat to get warm in case it's cold. E . Anticipating and Recognizing Alternative Courses of Action — includes utterances which offer several different interpretations or explanations of a situation. 1. She might call the police or some kind of animal catchers. 2. She might put the kittens up for a real long time or she might put them out and not let them come in for a real long time. 3. I think the kitten's gonna go out the path or go out in the yard or go hide. k. He might take the cat home or put it in his truck or sell it to somebody. 5. Mother might want to give the kitten to one of her friends or one of her cousins or to somebody for their birthday. ^Predicting Consequences of Actions or Events — includes utterances which suggest a possible outcome of some immediate or future action or event.

PAGE 239

229 1. The cat that's up on the table might tear the tablecloth if it sticks its claws down the tablecloth and tries to run. 2. If that car backs up on that bad dog, it will get run over. 3. if the kitten doesn't run, the dog' 11 get it. k. If the kitten tries to cross the street, it'll get run over because of all the traffic. 5. If she knocks on the door and gets the cat, the kids'd cry. PROJECT ! NG-the use of language within an unfamiliar or external context. A. Projecting into the Experiences of Others — Includes utterances which contemplate everyday occurrences from another's perspective. 1. The lady's thinking about cleaning that up. 2. The little black kitten doesn't want to get caught by that dog. 3. The little black kitten's thinking that he's lost. k. The truckdr i ver ' s thinking that he's gonna hit the black ki tten. 5. I think the kitten will come up to the door and want to see the lady. B. Projecting into Feelings of Othersincl udes utterances which reflect what it feels like to be another individual. Emotions and attitudes which are representative of another's point of view are expressed. 1. The mother cat is really mad at those kittens. 2. She hates to clean up messes. 3. The kitten's afraid that the dog might catch her. k. The little black kitten's thinking that he's lost. 5. When I get my pants real muddy, mother gets mad at me. C. Projecting into Reactions of Others — includes utterances which consider how another individual would respond to a particular situation or experience. 1. The lady's gonna hurt her kittens. 2. That old lady said "Get outa here." 3. She looks like she said "Oh my goodness," but she doesn't look like she's gonna stop them. k. The mother might be saying "We can't pay for all that cat food." 5. The man's saying "I'm gonna take you home and give you some milk and some cat food."

PAGE 240

230 D. Projecting into a Situation Never Exper i encedi ncl udes utterances in which the speaker conjectures about his own or another's feelings and reactions to unfamiliar activities or events. 1. The dog's thinking about eating the cat. 2. The dog's thinking about fixing a nice little meal outa that kitty. 3. The dog's thinking "I want to kill that cat." 4. She's gonna kill them kitties.

PAGE 241

APPENDIX C LANGUAGE SAMPLE EL I CITATION PROTOCOL ALA TOUGH

PAGE 242

APPENDIX C LANGUAGE SAMPLE ELI CITATION PROTOCOL ALA TOUGH I. Preparation A. El i cl tor--language clinician B. Materials 1. Practice st imul i --"Dad Forgets His Lunch," #1 2. Sample stimul i --"Black Kitten Gets Lost," #1, 2, 3, ^, 5. 6 C. Practice and Sample Statements--see below II. Pre-el i ci tation activities A. Rapport-building (Engage the child in informal conversation for 2 minutes.) B. Instructions and modeling SAY: "I'D LIKE YOU TO TELL ME SOME STORIES. I'M GOING TO SHOW YOU SOME PICTURES AND I'D LIKE YOU TO MAKE UP A STORY FOR EACH PICTURE. BUT FIRST LET ME SHOW YOU HOW TO DO IT. (Present "Dad Forgets His Lunch," Picture #1.) Set 1: SAY: SUPPOSE I ASKED YOU TO TELL ME ALL ABOUT THIS PICTURE. YOU MIGHT SAY, IT LOOKS LIKE MOM HAS MADE SOME SANDWICHES FOR LUNCH AND DAD HAS GONE OFF AND FORGOTTEN THEM. Set 2: THEN SUPPOSE I ASKED YOU, WHAT IS THE MAN DOING HERE? (Point to man in picture.) YOU MIGHT SAY, HE'S GOING TO WORK. AND SUPPOSE I SAID, WHY IS HE DOING THAT? YOU MIGHT SAY, SO HE CAN EARN SOME MONEY FOR HIS FAM 1 LY . Set 3: AND THEN SUPPOSE I SAID, HOW DO YOU THINK MOM (Point to woman in picture) FEELS. YOU MIGHT SAY, SHE FEELS VERY SAD CAUSE DAD HAS FORGOTTEN HIS LUNCH. C. Practice 232

PAGE 243

1. Say: NOW YOU TRY SOME. 2, Follow elicitation procedure for practice and sample i terns . Ml. El ici tation A. Present practice and stimulus pictures, one at a time. 1. Ask first question of each set (see below). 2. Ask alternate question of each set if: a. Child emits a response which Is more than 50 percent unintel 1 igible. b. Child emits a response such as "I don't know." c. A ten-second pause is observed. 3. Continue until child responds to all three questions for the practice picture and all eight questions for each sample picture. B. Practice Statements Set k: (a) WHAT DO YOU THINK THE BOY IS SHOUTING? (b) IF YOU WERE THE BOY, WHAT WOULD YOU SHOUT? Set 5: (a) CAN THE BOY MAKE DAD HEAR HIM? WHY DO/DON'T YOU THINK SO? (b) DOES DAD HEAR THE BOY? WHY DO/DON'T YOU THINK HE HEARS HIM? Set 6: (a) WHAT DO YOU THINK THE CHILDREN WILL DO NOW? (b) WILL THE CHILDREN TRY TO CATCH THEIR DAD? HOW WILL THEY DO IT? SUMMARY: SO IT LOOKS LIKE MOM IS UPSET BECAUSE DAD HAS GONE OFF TO WORK AND FORGOTTEN HIS SANDWICHES. C. Sample Statements NOW LET'S LOOK AT SOME MORE PICTURES.

PAGE 244

23^ THE BLACK KITTEN GETS LOST Picture I Set 1 a. Tell me all about this picture. b. What do you think is happening here? Set 2 a. What do you think is wrong here? Why do you think that? b. What are the kittens doing? I wonder why they're doing that? Set 3 a. What else do you think is wrong here? Why do you think that? b. What else are the kittens doing? I wonder why they're doing that? Set k a. How do you think the lady feels? Why do you think that? b. Do you think the lady is mad? Why is she mad? Set 5 a. What will the lady do now? b. If this were you, what would you do? Set 6 a. What do you think the mother cat is thinking? b. If you were the mother cat, what would you be thinking? Set 7 a. Tell me about something that you did that made your mother mad. b. Have you ever made your mother mad? What did you do? Set 8 a. Let's look at the picture again. What do you think is going to happen next? Why do you think that? b. What do you think the black kitten is going to do now? Why do you think that? SUMMARY: So, it looks like the lady is upset because the kittens have made such a mess. They have spilled the milk, eaten the hot dogs and ruined her knitting. And during all the confusion, the little black kitten has wandered, unnoticed, out the back door.

PAGE 245

235 Picture II Set 1 a. Tell me all about this picture. b. What do you think is happening here? Set 2 a. What do you think is wrong here? Why do you think that? b. What is happening to the little black kitten now? Set 3 a. Tell me all about the road in this picture. b. What kind of road are the dog and the little black kitten coming to? Why do you think that? Set k a. What kinds of things might happen when there's a lot of traffic on a road? Anything else? b. Have you ever seen a busy road like the one in the picture? Tel 1 me about i t. Set 5 a. How do you think the little black kitten feels? Why do you think so? b. Do you think the little black kitten is frightened? Why? Set 6 a. What do you think the dog is thinking? b. If you were the dog, what would you be thinking? Set 7 a. What will the little black kitten do now? Why? b. If you were the little black kitten, what would you do? Why? Set 8 a. What do you think is going to happen next? Why do you think that? b. What will happen to the little black kitten now? Why do you think that? SUMMARY: So, a big dog is chasing the little black kitten towards a busy road. If the kitten runs onto the road, he might get run over.

PAGE 246

236 Picture I I I Set 1 a. Tell me all about this picture. b. What do you think black kitten is doing now? Set 2 a. What do you think is wrong here? Why do you think that? b. What will happen if the black kitten keeps running? Why do you think that? Set 3 a. What do you think the boy in the bus is thinking? b. If you were the boy in the bus, what would you be thinking? Set k a. What do you think the man in the truck is thinking? b. If you were the man in the truck, what would you be th inking? Set 5 a. What do you think the little black kitten is thinking? b. If you were the little black kitten, what would you be thinking? Set 6 a. What should you do when you come to a busy road? Why? b. What's the best way to cross a busy road? Set 7 a. Tell me about a trip that you took in a car or a bus. b. Have you ever traveled anywhere by car or by bus? Tell me about your trip. Set 8 a. What do you think is going to happen next? Why do you think that? b. What do you think will happen to black kitten now? Why? SUMMARY: Black kitten has reached the main road safely but now it looks like he might be in danger from the traffic.

PAGE 247

237 Picture IV Set I a. Tell me all about this picture. b. Now, what do you think is happening? Set 2 a. What is the man going to do? Why is he going to do that? b. Is the man going to pick black kitten up? Why is he going to do that? Set 3 a. Is it a good thing that the man has stopped? Why do you think so? b. What might happen to black kitten if no one picked him up? Anything else? Set k a. What do you think the man is saying to black kitten? b. If you were the man, what would you say to black kitten? Set 5 a. What do you think black kitten is thinking? b. If you were black kitten, what would you be thinking? Set 6 a. Tell me about a time you got lost. b. Have you ever been lost? What happened? Set 7 a. Let's look at the picture again. What kind of work do you think the man does? How do you know? b. Do you think the man might be a window cleaner? Why do you think so? Set 8 a. What do you think the man will do now? Why do you think that? b. Where will the man take black kitten now? Why do you think that? SUMMARY: The man has stopped his van to take care of the little black kitten because the kitten seems to be lost.

PAGE 248

238 Picture V Set 1 a. Tell me all about this picture. b. What do you think is happening here? Set 2 a. What is the man doing here? Why is he doing that? b. Where has the man brought the kitten? Why do you think he did that? Set 3 a. How do you think the children feel? Why do you think so? b. Do you think the children are happy? Why do you think so? Set k a. How do you think the mother feels? Why do you think that? b. Do you think the mother is pleased? Why or why not? Set 5 a. What do you think the mother will say? Why do you think that? b. What would your mother say if you brought a lost kitten home? Why would she say that? Set 5 a. What do you think the children will do now? b. How will the children take care of black kitten now? Set 7 a. What would you do if someone brought a lost animal to your house? Anything else? b. Did you ever find a lost animal and bring it home to your family? What happened? Set 8 a. What do you think will happen next? Why do you think that? b. Do you think the black kitten will stay with the family? Why do you think that? SUMMARY: The man has brought the kitten home to his family and all the children are happy to see him.

PAGE 249

239 Picture Vl Set 1 a. Tell me all about this picture. b. What do you think is happening here? Set 2 a. What is everybody doing here? b. What is/are the little black kitten (children, mother) doing now? Set 3 a. What do you think the little girl is thinking now? b. If you were the little girl, what would you be thinking? Set k a. Tell me about the lady coming up the path. b. What is the lady coming up the path doing? Set 5 a. What will the lady say to the family? b. If you were the lady, what would you say to the family? Set 6 a. What do you think the black kitten will want to do? Why do you think that? b. If you were the black kitten, what would you want to do? Set 7 a. How would you take care of a kitten if you had one? b. What kinds of things do you need to do to take care of a ki tten? Set 8 a. How do you think this story ends? Why do you think that? b. Do you think black kitten stays with the family? Why/ why not? SUMMARY: The family and the little black kitten are having dinner. The lady who lost the kitten is looking for him. (Continue summary with ending that child has provided.)

PAGE 250

APPENDIX D STIMULUS MATERIALS FOR LANGUAGE SAMPLE ELICITATION %

PAGE 251

APPENDIX D STIMULUS MATERIALS FOR LANGUAGE SAMPLE ELICITATION 2k]

PAGE 252

1 242

PAGE 254

2kk 1

PAGE 256

246

PAGE 257

2k7

PAGE 258

APPENDIX E PROTOCOL FOR TRANSCRIPTION AND SEGMENTATION OF LANGUAGE SAMPLES

PAGE 259

APPENDIX E PROTOCOL FOR TRANSCRIPTION AND SEGMENTATION OF LANGUAGE SAMPLES^ I. Transcription A. Examiner's stimuli (Record all verbal behaviors, whether or not they are part of the stimulus protocol.) 1. Use transcription recording form. 2. Record any additional remarks in capital letters. B. Subject's language (Record all verbal behaviors.) 1. Transcribe the child's language in the space provided on the transcription recording form. 2. Double-space between language units, use no caps, and no punctuat ion . 3. Note certain contextual elements in parentheses. 4. For each stimulus item, number each language unit consecutively and segment following the segmentation system (adapted from Loban, 1963, pp. ^-7)Loban's system of segmentation combines several approaches. First, the subject's speech is segmented by oral intonation patterns and then, within such intonation segments, syntacticsemantic units are identified (Loban, p. 5). I I . Segmentat ion A. PHONOLOGICAL UNIT (PU) . A PU is "judged by the contours of intonation, stress and pause in the subjects' voices" (Loban, 1963, p. 5). Silent pauses and terminal juncture can be recorded as a double cross (#) . "A PU is usually marked by a definite pause preceded by a diminishing of force and a drop in the pitch of the voice (or a rise for queries)" (Loban, 1963, p. 6). Double bar (//) and single bar (/) are used to represent silent pauses of less finality. According to this definition, the PU occurs between # junctures. In addition ^Based on Loban (I963, 1976) in Barr ie-Blackley et al . (1978). 2k3

PAGE 260

250 to these conventions, eHipses (. . .) can be used to indicate the length of a pause, by transcribing approximately one ellipse (.) for each second of pause time (Barrie-Blackley et al., 197^). B. COMMUNICATION UNIT (CU) . A CU is "always a subdivision of the larger phonological unit , . . because it must be identified by the semantic meaning which is being communicated (Loban, 1963, p. 5). According to Loban, it was defined by Watts (19'»8) as "a group of words which cannot be further divided without loss of their essential meaning." Thus, a CU would never cross a PU boundary, although there could be more than one CU within a PU. Several specific rules for segmenting are as fol lows : 1. A CU need not be complete in structure: WHAT WILL THE LADY DO NOW^ a, #clean up# This example is both a PU (because it is marked by pauses and intonational boundaries), and a CU (because it is an implicitly complete response). 2. A "sentence" that contains utterances that can "stand alone" (i.e., a compound sentence conjoined with "and," "and then," or "then") is considered two or more CUs: ^a. #the black kitten will walk out the door b. and he'll run down the road# This division is made sometimes unnaturally, within PUs, in order to isolate independent predications. When such a division is made on the basis of language structure alone, and no pause or intonation break separates the units (that is, they fall within one PU) , a line should be drawn to connect the two CUs. 3. A conjoined "sentence" in which subject or verb deletion has occurred will be counted as one CU, provided it falls within one PU: a. #the black kitten will walk out the door and run down the road# In the above example, the subject "he" has been deleted from the second clause. It is considered one CU, since it falls within one PU and because the phrase "and run down the road" cannot stand alone. Utterances in caps indicate those spoken by the elicitor.

PAGE 261

251 4. The child's use of terminal juncture may produce grammatically independent predications that are not complete in structure: a. #the black kitten walked out the door# b. #and ran down the road# In this example, each CD falls within a different PU, thus requiring segmentation. 5. "Adjoined" utterances are counted as one CD, provided all clauses fall with one PU: a. #the black kitten walked out the door because he wanted to run down the road# Though the clauses are structurally complete, they must remain as one CU to preserve the meaning. If the clauses are separated by terminal juncture, however, they will comprise two (or more) CDs: a. #the kitten likes to play with yarn# b. #because it's so much fun# 6. Some additional segmenting conventions may be necessary. a. Segmenting "adjoined" CUs. (l) If two or more clauses are linked to an adjoining clause, the entire utterance is counted as one CU: (a) #if the boy already knew what time to catch the bus (b) and he's gotta get his books# b. Segmenting "asides" in CUs. (1) When there is a CU within a CU, each is counted separately: #the kitten has a pretty collar [it shines] around her neck# would be segmented as follows: (a) #the kitten has a pretty collar [] around her neck# (b) [it shines] (2) An explanation or expansion within or at the end of a CU is counted as part of the CU to which it refers : (a) #i got a kitty ... a play one#

PAGE 262

252 c. A direct quotation is counted as part of the CU In which it occurs: (1) #the man said "I'll take you home with me" and then put the kitten in his van# Excluded verbal segments (mark through transcriptions with a line and do not number as CUs) 1. Omit, by crossing out, YES-NO RESPONSES elicited outside the protocol by the examiner (i.e., IS THAT RIGHT? "yfe%."). 2. Omit language MAZES. Loban (I963) defines a maze as a "series of words or initial parts of words which do not add up, either to meaningful communication or to structural units of communication. . . . Sometimes the subjects persevere with the ideas they are trying to formulate and at the end of the maze, do achieve a unit of communication. Other times the subjects abandon the ideas they are trying to express" (p. 8) . In this protocol, maze will be applied to segments characterized by general lack of normal English order, and will not be applied to simple repetitions or rewordings, without the confusion of order. Mazes will be set off by a box: ...he had a lot of dogs" Omit FALSE STARTS. Loban (1963) also discusses false starts as simple repetitions or slightly altered rephrasings of a form, without confusion of word order or sequence. For this protocol, false starts can occur at the beginning of a PU, or within a CU and will be set off by an encompassing circle. C'^1»&^^^ato'yt/yv&»m'l4^ the dog's chasing the cat ('ti»yv/dC)^yvoai») — the dog can't cross the street the mother cat has f T/ho^wvwtwwlnyrawaii)) four kittens. ...eat (^Wt/ao^l/f'jww^wiy&va^) some food. Omit REPETITIONS. Repetitious segments that do not function as false starts are defined separately in this protocol. A REPETITION will be defined as an exact repetition of an immediately preceding word, phrase or CU that occurs in a ritualistic manner but not for emphasis or sequencing. Repetitions will be excluded from the initial analysis by

PAGE 263

253 marking the segments through with a wavy line and indicating each repetition with a triangle (A). I have one like th i s--//'VVkfe'vtM/5. — // A I have a dog--/ A A 5. Omit STARTERS. Loban also includes starters in his analysis. For this protocol, a starter occurs when the speaker is attempting to "hold the floor" or fill silence while he organizes his thoughts. The words carry little meaning and are often stereotypic and individualistic. Starters will be marked through with a wavy line and indicated, as a un i t , by a rightsided triangle (s>) . P>wVtA1<»toW I have one ^Vh^vl^'VWvyfefe. . . I get it down >9»fvavuvwwvawd. . .he left 6. Omit UNINTELLIGIBLE UNITS. Utterances, segments of which cannot be understood for whatever reason, are excluded from the initial analysis. This is done by transcribing a raised dash for each syllable and enclosing the "segment" in brackets ([ — ]). If desired, phonetic transcription can be used in place of dashes, with the established notation for syllabification. If there are two or more unintelligible syllables in any CU, the CU is not numbered and is omitted from the initial analysis . In addition, segments that must be "guessed at" are placed in brackets ([daddy go]) but can be counted since their language is known approximatel y . 7. Omit NON-LINGUISTIC SOUND-MAKING, defined as instances when the child "plays" with sounds or melodies. Mark such vocalizations through with a straight line. [Note that "sounds" that serve a syntactic function in a CU are not included in this category (e.g., "the car went V roooom' ' ) • ] and that's al 1 . . . dadedododo 8. Omit direct repetitions of CUs.

PAGE 264

APPENDIX F RULES FOR COUNTING MORPHEMES

PAGE 265

APPENDIX F RULES FOR COUNTING MORPHEMES (adapted from Brown, 1973) 1. When a word is repeated for emphasis, count each occurrence up to a total of three (e.g., "No, No, No"). 2. Count all compound words (two or more free morphemes, proper names and ritualized reduplications) as single words unless there is evidence of independent use of these elements elsewhere in the sample (e.g., sailboat, pocketbook, birthday, choo-choo, quackquack, night-night). 3. Count all diminutives as single words unless there is evidence of creative application of this ending to other nouns (e.g., mommie, doggie) . k. Count all regular inflectional endings as separate morphemes (e.g., plural /-s/, cups; past /-D/, jumped; possessive hzl, boy's; third person singular /-z/, runs; progressive /-'\nq/ , going; comparative /-er/, bigger; superlative /-est/, biggest). 5. Count all words containing irregular inflectional endings as single words (e.g., plural, men; past, went; comparative, better; superlative, best). 6. Count overgeneral ized inflectional endings as separate morphemes (e.g., plural, firemans; past, goed; superlative, bestest). 7. Count all auxiliaries, contracted or uncontracted, as separate morphemes (is, have, will, can, won't). 8. Count all catenatives as single morphemes (hafta, wanna, gonna). 9. Count noises as single morphemes when they function as an integral part of the sentence ("The car goes vroom."). 10. Omit fillers (e.g., "um," "er," "well"). 11. Omit excogitations unless they appear to function as a question (e.g., "see," "you know," "O.K."). 255

PAGE 266

APPENDIX G RULES FOR COUNTING WORDS

PAGE 267

APPENDIX G RULES FOR COUNTING WORDS 1. Count repetitions for emphasis as directed under "morphemes" above. 2. Count hyphenated words, compound words, and proper names that function as single units as one word (e.g., "Merry-go-round," "Mother Goose," "Betty Lou") . 3. Count each word of a verbal combination as a separate word (e.g., "have been playing" = 3 words). k. Count contractions of the subject and predicate, such as "it's" and "you're" as single words. 5. Count contractions of the verb and negative (e.g., can't and don't) as single words. 6. Count as one word, words that have inflectional endings or that take a phonological ly dissimilar contracted form. 7. Count ritualized expressions as one word (e.g., "Oh boy," "my gosh," "darn it," "doggone it," "all right," "maybe," "giddy-up," "ain't," "lookit") if it occurs alone. 8. Count the following as two words: "oh yes," "oh no," "oh gee," "let's see," "on to," "Christmas tree," "kinda," "oughta," "hafta," "lookit" if followed by an object. 9. Count noises for words as directed under morphemes. 10. Count colloquialisms and neologisms as one word (e.g., "yike," "ya," "yippe," "teensy-weensy ," "naw," "yeah," "whoops," "wham"). 11. Omit fillers such as "well," "er," "um." 257

PAGE 268

APPENDIX H RAW DATA ON STANDARDIZED LINGUISTIC MEASURES AND THE OVERALL SCORE ON THE FUNCTIONAL INVENTORY OF COGNITIVE COMMUNICATION STRATEGIES ACHIEVED BY ALL SUBJECTS

PAGE 269

APPENDIX H RAW DATA ON STANDARDIZED LINGUISTIC MEASURES AND THE OVERALL SCORE ON THE FUNCTIONAL INVENTORY OF COGNITIVE COMMUNICATION STRATEGIES ACHIEVED BY ALL SUBJECTS STANDARDIZED LINGUISTIC MEASURES SUBJECT TOLD TACL CELI FICCS I 75 Bk 90 ^7 2 78 88 29 59 3 72 75 72 k 77 79 no 5 78 79 86 39 6 83 89 k8 37 7 83 9^ 105 60 8 79 67 39 51 9 80 79 75 ^7 10 79 88 19 37 11 101 90 7 ^7 12 96 90 22 48 13 108 89 1 63 \k 101 90 9 iik 15 98 81 9 68 16 108 93 ii 5^ 17 101 93 6 59 18 1 1 1 91 1 55 19 101 89 3 49 20 ]]k 93 3 75 21 92 71 27 27 22 95 73 15 42 23 102 75 9 41 24 96 79 7 32 25 100 88 16 47 26 88 81 22 48 27 95 79 21 51 28 98 88 16 44 29 88 67 ko 34 30 81 72 3^* 34 31 98 88 5 74 32 101 92 5 87 33 98 88 19 32 3h 98 85 9 46 35 107 88 3 51 259

PAGE 270

260 APPENDIX H (continued) STANDARDIZED LINGUISTIC MEASURES SUBJECT TOLD TACL CELI FICCS 36 121 95 0 49 37 122 97 0 50 38 120 95 2 64 39 117 97 1 51 40 95 88 7 57 Range 72-122 67-97 0-110 Mean 96.650 85.300 24.900 49.7 SD 13.765 8.378 30.490 12.517 TOLD = Test of Language Development TACL = Test of Auditory Comprehension of Language CELI = Carrow Elicited Language Inventory

PAGE 271

APPENDIX I RAW DATA ON NONSTANDARD! ZED LINGUISTIC MEASURES OBTAINED IN RESPONSE TO THE FUNCTIONAL INVENTORY OF COGNITIVE COMMUNICATION STRATEGIES (FICCS) BY ALL SUBJECTS

PAGE 272

APPENDIX I RAW DATA ON NONSTANDARD I ZED LINGUISTIC MEASURES OBTAINED IN RESPONSE TO THE FUNCTIONAL INVENTORY OF COGNITIVE COMMUNICATION STRATEGIES (FICCS) BY ALL SUBJECTS NONSTANDARDIZED LINGUISTIC MEASURES SUBj^^^ #CU TNW MLU-W TNM MLU-M MLU5-M %CUc DSS 1 131 570 4.34 605 2 124 733 5.91 809 3 113 580 5.13 631 4 113 550 4.87 568 5 68 256 5.24 381 6 109 540 4.95 565 7 186 911 4.87 971 8 128 677 5.28 779 9 105 417 3.97 466 10 68 252 3.71 264 11 178 852 4.79 988 12 ]k2 703 4.95 810 13 219 1529 6.98 1714 li» 93 557 5.99 666 15 110 535 4.86 604 16 129 853 6.61 929 17 172 1273 7.40 1428 18 113 939 8.31 1037 19 131 706 5.39 794 20 176 1413 8.03 1519 21 67 296 4.42 305 22 \k3 867 5.82 966 23 88 415 4.72 484 2k 73 356 4.88 388 25 126 675 5.36 743 26 132 650 4.92 707 27 112 570 5.09 626 28 123 683 5.55 741 29 95 310 3.26 402 30 8k 281 3.39 318 31 2k7 1441 5.83 1599 32 156 1 105 7.08 1199 33 106 580 5.47 657 34 101 530 5.25 573 35 120 745 6.21 825 4.62 9 .6 54.20 4.82 6.52 15 .6 59.68 8.50 5.58 11 .2 76.99 6.70 5.03 16 .0 53.10 7.26 5.60 1 1 .2 67.65 5.24 5. 18 12 .4 63.30 5.04 5.19 14 .8 67.20 4.10 6.09 15 .4. 67.19 8.02 4.44 9 .6 46.67 4.86 3.88 7 .6 66.18 5.93 5.55 14 .2 58.99 7.36 5.70 1 1 .8 73.24 6.14 7.83 21 .6 78.08 10.76 7.16 15 .0 80.65 8.90 5.49 12 .6 63.64 7.70 7.20 17 .6 83.17 9.22 8.30 23 .6 72.09 10.24 9.18 20 .4 84.96 17.14 6.06 14 .0 84.73 10.24 8.63 22 6 87.50 15.52 4.55 10 6 68.66 8. 10 6.48 13 6 83.22 7.54 5.50 12 4 61.36 7.62 5.32 1 1 4 64.38 5.94 5.90 14 2 73.81 10.06 5.36 9 6 61 .36 6.60 5.59 12 2 76.79 6.64 6.02 14 0 76.42 8.88 4.23 10 2 48.42 5.82 3.83 7 4 51.81 7.27 6.47 23 6 74.49 8.30 7.69 19 2 76.28 10.78 6.20 13 2 66.04 7.86 5.67 1 1 2 74.26 8.28 6.88 15 2 71 .67 10.26 262

PAGE 273

263 APPENDIX I (continued) NONSTANDARD I ZED LINGUISTIC MEASURES SUBJECT #CU TNW MLU-W TNM MLU-M MLU5-M ^CUc DSS 36 118 mk 6.31 825 6.99 19. ,2 69.49 1 1 .38 37 103 767 7.45 829 8.05 20. ,8 82.52 14.90 38 227 1862 8.20 2062 9.08 26. .0 82.38 17.00 39 106 661 6.24 734 6.92 16, ,2 76.42 12.46 ko 111 847 7.63 933 8.41 17. ,4 91.89 10.48 Range 67-247 2523. 272643.837.446.674. 101862 8. 31 2062 9.18 26.0 91 .89 17. 14 Mean 126.30 733.28 5. 61 81 1 . 10 6.21 14.86 70.50 8.75 SD 41 .894 359.22 1. 27 397.14 1.39 4.61 1 1 .02 3.19 #CU's = Total Number of Communication Units TNW = Total Number or Words MLU-W = Mean Length of Communication Units in Words TNM = Total Number of Morphemes MLU-M = Mean Length of Communication Units in Morphemes MLU5"M = Mean Length of the Five Longest Communication Units in Morphemes ^CUc = Percentage of Complete Communication Units DSS = Developmental Sentence Score

PAGE 274

APPENDIX J RAW DATA ON MEASURES OF ACADEMIC ACHIEVEMENT FOR ALL SUBJECTS

PAGE 275

APPENDIX J RAW DATA ON MEASURES OF ACADEMIC ACHIEVEMENT FOR ALL SUBJECTS TEST OF BASIC EXPERIENCES MATHFMAT 1 PS SOCIAL STUDIFS SCI ENCE 1 1 il7 *»/ Hy HO HI CO pp 4 3 H 1 HO yi Hp il hi HO ii7 liQ HP c 3 £;ii po Hp CO PP 7 po c? pp tt w to Hp Q 4^ hi P 1 in ^ 1 Hp 1 1 1 1 / ^ Dp oo CC PP 1 9 op Al oy 1 1 68 AA 00 1 Jl 1 H ou po /p 1 C C 1 PO P* P/ 1 O oo 7 1 / 1 Ail OH 75 17 66 71 56 59 18 69 74 82 79 19 66 68 68 57 20 61 68 64 72 21 ^7 35 46 36 22 kk 43 46 45 23 48 47 47 50 Ik 46 45 46 49 25 49 51 50 53 26 46 53 48 49 27 45 46 47 49 28 52 51 52 53 29 44 41 46 46 30 46 43 46 44 31 66 54 54 55 32 72 74 58 69 33 69 58 54 53 34 63 62 68 57 35 63 71 61 64 265

PAGE 276

266 APPENDIX J (continued) TEST OF BASIC EXPERIENCES SUBJECT LANGUAGE MATHEMATICS SOCIAL STUDIES SCIENCE 36 66 68 72 66 37 80 Jh 68 69 38 80 Ih 72 90 39 69 1^ 6i» 6A kO 61 68 56 66 Range 41-80 l^-Jk 39-82 36-90 Mean 57-07 57-65 h^.kO 57-22 SD \\M 11-61 10.81 11.55

PAGE 277

BIBLIOGRAPHY Alder, S., Poverty children and their language . New York: Grune and Stratton, Inc. (1979). Allen, R. R. , & Brown, K. L. , Developing communication competence in chi Idren . Skokie, 111.: National Textbook Company (1976). Antinucci, F. , & Parisi, D. , Early language acquisition: A model and some data. In C. Ferguson & D. Slobin (Eds.), Studies in chi Id language development . New York: Holt, Rinehart & Winston (1973). Austin, J. L., How to do things with words . Cambridge, Mass.: Harvard University Press (1962). Baldwin, A. L., & Baldwin, C. P., The study of mother-child interaction. American Scientist , 61, 7l'»-721 (1973). Bankson, N. W., Bankson language screening test . Baltimore: University Park Press (1977) . Barrie-Blackley, S., Didow, S. M. , & Faurest, A., Practical screening for language disorders through small sample analysis: Preliminary report. Communique , 4, 8-11 (197'»). Barrie-Blackley, S., Musselwhite, C. R. , Rogister, S., Clinical oral language sampl ing . Danville, 111.: The Interstate Printers and Publishers, Inc. (1978). Bartolucci, G. & Albers, R. J., Deictic categories in the language of autistic children. Journal of Autism and Childhood Schizophrenia . 4, 131-1^1 (1974). Bates, E. , Pragmatics and sociol inguistics in child language. In D. Morehead £ A, Morehead (Eds.), Normal and deficient child language . Baltimore: University Park Press ( 1 976a) . Bates, E., Language and context . New York: Academic Press, Inc. (1976b). Bates, E. , Camaioni, L. , 6 Volterra, V., Acquisition of performatives prior to speech. Merri 1-Palmer Quarterly , 21, 205-226 (1975). Berko-Gleason, J., Code switching in children's language. In T. E. Moore (Ed.), Cognitive development and the acquisition of language . New York: Academic Press (1973). Bedroslan, J. L. & Pruitting, C. A., Communicative performance of mentally retarded adults in four conversational settings. Journal of Speech and Hearing Research , 21, 79-95 (1978). 267

PAGE 278

268 Berko-Gleason, J., & Weintraub, S., The acquisition of routines in child language. Language in Society , 5, 129-136 (1976). Bernstein, B., Social class, linguistic codes, and grammatical elements. Language and Speech , 5, 221-240 (1962). Bernstein, B., Elaborated and restricted codes: Their social origins and some conclusions. American Anthropologist , 66, 55-69 (1964). Bernstein, B., A soc iol ingu i st i c approach to social learning. In J. Gould (Ed.), Penguin survey of the social sciences . Baltimore: Penguin (19657^^ Bernstein, B. (Ed.), Class, codes and control I: Theoretical studies towards a sociology of language . London: Routledge & Kegan Paul (1971). Bernstein, B., (Ed.), Class, codes and control II: Towards a theory of educational transmissions . London: Routledge & Kegan Paul (1975). Blank, M., Rose, S. A., & Berlin, S. , The preschool language assessment instrument . Baltimore: University Park Press (1978). Bloom, L., Language development: Form and function in emerging grammars . Cambridge, Mass.: MIT Press (1970)Bloom, L. , S Lahey, M. , Language development and language disorders . New York: John Wi ley and Sons , I nc. ( 1 978) . Bloom, L., Lightblown, P., & Hood, L., Structure and variation in child language, Monographs of the Society for Research in Child Development , 40, Uo. 160 (1975). Boehm, A., Boehm test of basic concepts . New York: Psychological Corporation (1971). Bowerman, M. F., Early syntactic development: A cross-linguistic study with special reference to Finnish . Cambridge: Cambridge University Press (1973). Brenneis, D., & Lein, L., You fruithead: A sociol inguistic approach to children's dispute settlement. in S. Ervin-Tripp and C. Mi tchel 1 -Kernan (Eds.), Child Discourse . New York: Academic Press, Inc. (1977). Britton, J., What's the use? Educational Review , 23, 205-219 (1971). Brown, K. L., Ecroyd, D. , Hopper, R. , and Naremore, R., A summary of communication competencies derived from literature in review. In R. R. Allen and K. L. Brown (Eds.), Developing communication competence in chi Idren. Skokie, 111.: National Textbook Co. ( 1 976) .

PAGE 279

2 69 Brown, R. , A first language: The early stages . Cambridge, Mass.: Harvard University Press (1973). Bruck, M., £ Tucker, R. G. , Social class differences in the acquisition of school language. Herr i 1 1 -Pal mer Quarterly , 20, 205-220 (197^). Bruner, J. S., The course of cognitive growth. American Psychologist , 19, 1-15 (196^). Bruner, J. S. , Studies in cognitive growth . New York: John Wiley and Sons (1966). Bruner, J. S., From communication to language: A psychological perspective. Cognition , 3, 225-287 (197^). Bruner, J. S., The ontogenesis of speech acts. Chi Id Language , 2, 1-19 (1975). Buhler, K. , Sprachtheorie: Die darstel 1 ungsf unkt ion der sprache . Jena: Fischer UsWT. Burgemei ster, B. B., Blum, L. H., 5 Lorge, I., Columbia mental maturity scale, 3rd edition. New York: Harcourt, Brace, Jovanovich, Inc. 11972). Car row, E., Test for auditory comprehension of language . Austin, Tex.: Lea r n i n g Concepts (1973). Carrow, E., Carrow elicited language inventory . Austin, Tex.: Learning Concepts (197^). Cazden, C. B., Subcultural differences in child languages: An interdisciplinary review. Merrill-Palmer Quarterly , 12, 185-219 (1966). Cazden, C. B., The neglected situation of child language research and education. Journal of Social Issues , 26, 35-60 (1970). Cazden, C. B., Commentary. In Schachter, F. F. , Ki rschner, K., Klips, B., Friedericks, M. , & Sanders, K. , Everyday preschool interpersonal speech usage: Methodological, developmental, and sociol inguistic studies. Monographs of the Society for Research in Child Development , 39, No! 156 (1974). Cazden, D. B., Bond, J. T., Epstein, A. S., Matz, R. D. , & Savignon, S. J., Language assessment: Where, what, and how. Anthropology and Education . 2, 83-91 (1977). Chomsky, N. , Syntactic structures . The Hague: Mouton (1957). Chomsky, N., Aspects of the theory of syntax . Cambridge, Mass.: MIT Press (ig^ST Chomsky, C, The acquisition of syntax in children from 5 to 10 . Cambridge, Mass. : MIT Press (1969) .

PAGE 280

270 Clark, E., What's in a word? On the child's acquisition of semantics in his first language. In T. Moore, (Ed.), Cognitive development and the acquisition of language . New York: Academic Press ( 1 973) . Clark, H., 6 Clark, E., Psychology of language . New York: Harcourt Brace Jovanovich Publ ishers (1977) • Curtiss, S., Prutting, C. A., S Lovel 1 , E. L., Pragmatic and semantic development in young children with impaired hearing. Journal of Speech and Hearing Research , 22, 53^-552 (1979). Dale, P., Language development; Structure and function . Hinsdale, 111.: The Dryden Press, Inc. (1972). Davies, A., (Ed.), Language and learning in early childhood . London: Heinemann Educationa 1 Books Ltd. (1977) . Davis, E. A., The development of linguistic skill in twins, singletons with siblings, and only children from age five to ten years. Chi 1 d Welfare Monographs , No. 1^. Minneapolis, Minn.: University of Minnesota Press (1937) • Day, E. J., The development of language in twins. I: A comparison of twins and single children. Chi Id Development , 3, 179-199 (1932). DeLaguna, G. A., Speech: Its function and development . Bloomington, Indiana: University Press (first edition 1927) (1963). Dore, J., A pragmatic description of early development. Journal of Psychol inquistic Research , 3, 3^3-350 (197^). Dore, J., Holophrases, speech acts and language universals. Journal of Child Language , 2, 21-40 (1975). Dore, J., Children's i 1 locut ionary acts. In R. Freedle (Ed.), Pi scourse: Comprehension and production . New Jersey: Lawrence Erlbaum Associates (1976) . Dore, J., Oh them sheriff: A pragmatic analysis of children's responses to questions. In S. Ervin-Tripp and C. Mi tche 1 1 -Kernan (Eds.), Child Discourse . New York: Academic Press, Inc. (1977). Dore, J., Variation in preschool children's conversational performances. In K. E. Nelson (Ed.), Children's language . Vol. 1. New York: Gardner Press (1978) . Dore, J., Gearhart, M. , & Newman, D., The structure of nursery school conversation. In K. E. Nelson (Ed.), Children's language , Vol. 1. New York: Garner Press (1978). Dunkin, M. J., & Biddle, B. J., The study of teaching . Holt, Rinehart and Winston, Inc. (197^). Dunn, L., Peabody picture vocabulary test . Circle Pines, Minn.: American Guidance Service (1964).

PAGE 281

271 Edwards, A. D. , Speech codes and speech variants: Social class and task differences in children's speech. Journal of Child Language , 3, 247-265 (1977). Epstein, A. S., Schwartz, P., Meece, J., Lambie, D. , Dukes, P., Crawford, F., & Phillips, P., The mutual problem-solving task . Ypsilanti, Mich.: High/Scope Educational Research Founcation (1976). Ervin-Tripp, S., Wait for me roller skate. In S. Ervin-Tripp and C. Mitchel 1-Kernan (Eds.), Chi Id Discourse . New York: Academic Press (1977). Ervin-Tripp, S., 5 Mi tchel 1-Kernan, C. (Eds.), Chi Id discourse . New York: Academic Press, Inc. (1977). Fillmore, C. The case for case. In E. Bach & R. T. Harms (Eds.), Universals in linguistic theory . New York: Holt, Rinehart & Winston (1968). Fisher, A., 6 Logeman, J., Fisher Logemann test of articulation competence . Geneva, 111.: Houghton-Mifflin (1971). Fisichelli, V. R., S Karelitz, S., The cry latencies of normal infants and those with brain damage. Journal of Pediatrics , 62, 72^-73'* (1963). Fogel , D., The study of functional communication competence in children. In R. R. Allen and K. L. Brown (Eds.), Developing communication competence in children . Skokie, 111.: National Textbook Company (1976). Foster, R. , Giddan, J., & Stark, J., Assessment of children's language comprehens ion . Austin, Tex.: Learning Concepts ( 1 969) . Franzblau, A. N., A primer of statistics for non-statisticians . New York: Harcourt, Brace 6 World, Inc. (1958). Friedlander, B. Z., Jacobs, A. C, Davis, B. B., & Wetstone, H. S., Time-sampling analysis of infants' natural language environments in the home. Chi Id Development , k3, 730-7^0 (1972). Garvey, C, Requests and responses in children's speech. Journal of Child Language , 2, i*l-60 (1975). Geller, E. F., S Wollner, S. G., A preliminary investigation of the communicative competence of three linguistically impaired children. Paper presented at the New York State Speech and Hearing Association, Liberty, New York (1976). Goldman, R. , S Fristoe, M., The Gol dmanFr i stoe test of articulation . Circle Pines, Minn.: American Guidance Service, Inc. ( 1 969) .

PAGE 282

27? Greenfield, P., S Smith, J., The structure of communication in early language development . New York: Academic Press (1976)Grice, H. P., Logic and conversation. In P. Cole and J. L. Morgan (Eds.), Syntax and semantics, Vol. 3: Speech acts . New York: Academic Press (1975) . Grimm, H., Analysis of short-term dialogues in 5-7 year-olds: Encoding of intentions and modifications of speech acts as a function of negative feedback loops. Paper presented to the Third International Child Language Symposium, London (1975). Halliday, M. A. K. , Language structure and language function . in J. Lyons (Ed.), New horizons in linguistics . Baltimore: Penguin Books (1970). Halliday, M. A. K. , Explorations in the functions of language . London: Edward Arnold (1973) . Halliday, M. A. K. , Learning how to mean . London: Edward Arnold (1975). Halliday, M. A. K. , Language as social semiotic . Baltimore: University Park Press (19781"; Hanes, M., Validation of the High/Scope open framework curriculum model for bilingual bicultural preschool programs . Ypsilanti, Mich.: High/Scope Educational Research Foundation (1978). Hannah, E. P., Applied linguistic analysis . Northridge, Calif.: Joyce Motion Picture Co. (197^) . Hansen, H. P., Language acquisition and development in the child: A teacher-child verbal interaction. Elementary English , 51, 276-285 (1974). Hess, R. D., Parental behavior and children's school achievement; implications for Head Start. In E. Grotberg (Ed.), Critical Issues in research related to d i sadvantaged ch i 1 dren . Princeton, N.J.: Educational Testing Service (1969). Hoi 1 ingshead. A., A two-factor index of social position . New Haven: Yale Uni vers i ty Press (1957) • Hopper, R. , Expanding the notion of competence. The Speech Teacher , 20, 29-35 (1971). Hopper, R. , S Naremore, R., Children's speech: A practical introduction to communication development . New York: Harper & Row (1973). Hunt, J. McV. Intelligence and experience . New York: Ronald (1961). Hymes, D., On communicative competence . Philadelphia: University of Pennsylvania Press (1970).

PAGE 283

273 Hymes, D. , Introduction. In C. B. Cazden, V. P. John & D. Hymes (Eds.), Functions of language in the classroom . New York: Teachers College Press, Columbia University (1972a). Hymes, D. , Models of the interaction of language and social life. In J. J. Gumperz and D. Hymes (Eds.), Directions in sociol i ngu i st ics . New York: Holt, Rinehart and Winston, Inc. (1972b). Jakobson, R. , Closing statement: Linguistics and poetics. In T. A. Sebeok (Ed.), Style in language . Cambridge, Mass.: MIT Press (I960). Johnson, W. , Darley, F. L., £ Spr iestersbach, D. C, Diagnostic methods in speech pathology . New York: Harper & Row (1 9631"! Kernan, K. T. , Semantic and expressive elaboration in children's narratives. In S. Ervin-Tripp 6 C. Mi tchel 1 -Kernan (Eds.), Chi Id discourse . New York: Academic Press (1977). Labov, W. , The logic of nonstandard English. In F. Williams (Ed.), Language and poverty . Chicago: Markham (1970). Larson, C, Backlund, P., Redmond, M. , & Barbour, A., Assessing functional communication . Falls Church, Va.: Speech Communication Association (1978). Lee, L., Northwestern syntax screening test . Evanston, 111.: Northwest ern University Press (1969)Lee, L. , Developmental sentence analysis . Evanston, 111.: Northwestern University Press ( 1 37^) . Leonard, L. B., What is deviant language? Journal of Speech and Hearing Disorders , 37, ^27-^^6 (1972). Leonard, L. B., Bolders, J. G., & Miller, J. A., An examination of the semantic relations reflected In the usage of normal and languagedisordered children. Journal of Speech and Hearing Research , 19, 371-392 (1976). Limber, J., Unravelling competence, performance and pragmatics in the speech of young children. Journal of Child Language , 3, 309-318 (1976). Loban, W. , The language of elementary school children . Urbana, 111.: Nat i onal Council of Teachers ( 1 963) . Loban, W. , Language development: Kindergarten through grade twelve . Urbana , 111.: National Council of Teachers (1976). Lucas, E. v.. Semantic and pragmatic language disorders . Rockville, Md. : Aspen Systems Corporation (1980;.

PAGE 284

1 27A Luria, A. R, , The directive function of speech in development and dissolution. Word, 15, 3'»l-352 (1959). MacDonald, J. D. , 6 Nickols, M. , Environmental language inventory manual. Columbus, Ohio: Ohio State University (197M. Mahoney, G. J., An ethological approach to delayed language acquisition. American Journal of Mental Deficiency . 80, 139-1^8 (1975). Maliknowski, B., The problem of meaning in primitive languages. Supplement I. In C. K. Ogden and I. A. Richards, The meaning of meaning . London: Routledge and Kegan Paul (1923). McCarthy, D. A., The language development of the preschool child. Chi Id Welfare Monographs , No. k, Minneapolis, Minn.: University of Minnesota Press (1930). McCarthy, D. A., Language development in children . In L. Carmichael (Ed.), Manual of child psychology . New York: John Wiley & Sons (195A). McDonald, E. , McDonald deep test of articulation . Pittsburgh: Stanwix House {]SeW. ' McLean, J. E., & McLean, L. K. , A transactional approach to early language training . Columbus, Ohio: Charles E. Merrill Publishing Company (1978) . McNeill, D., The acquisition of language: The study of developmental psycho! inguistics . New York: Harper & Row (1970). Menyuk, P., Sentences children use . Cambridge, Mass.: The MIT Press (1969). Miller, J., & Yoder, D. , A syntax teaching program. In J. McLean, D. Yoder, 6 R. Sch iefel busch (Eds.), Language intervention with the retarded . Baltimore, Md.: University Park Press (1972). Miller, L., Pragmatics and early childhood language disorders: Communicative interactions in a half-hour sample. Journal of Speech and Hearing Disorders , A3, ^19-^36 (1978). Mi tchel 1-Kernan, C, 6 Kernan, K. T., Pragmatics of directive choice among children. In S. Ervin-Tripp & C. Mi tchel 1 -Kernan (Eds.), Chi 1 d di scourse . New York: Academic Press (1977). Moffet, J. A., Teaching the universe of discourse . Boston: Houghton Mifflin Morris, C, Foundations of the theory of signs . Chicago: University of Chicago Press (1938) . Morris, D., The naked ape . New York: McGraw-Hill (1967). i .1

PAGE 285

275 Moss, M. H., Test of basic experiences . Monterey: CTE/McGraw-Hi 1 1 (1972). Musselwhite, C. R., Three variations of the imperative format of Ian guage sample el icitation. Unpubl i shed Master ' s thesis, University of North Carolina at Chapel Hill (1975). Naremore, R. , The learning of communication roles and norms. In R. R. Allen and K. L. Brown (Eds.), Deve loping commun i ca t i on compe tence in chi Idren . Skokie, ill.: National Textbook Co. (1976). Nelson, K. , Structure and strategy in learning to talk. Monographs of the Society for Research in Child Development , 38, No. 1^9 (1973). Nelson, K. , Concept, word and sentence: Interrelations in acquisition and development. Psychological Review , 8l , 267-285 (197'*). Newcomber, P., & Hammi 1 1 , D., Test of language development . Austin, Tex.: Empiric Press (1977). Nunnally, J., Psychometric theory . New York: McGraw-Hill (1967). O'Donnell, R. , Griffin, W. J., & Norris, R. C, Syntax of kindergarten and elementary school: A transformational analysis . National Council of Teachers of English Research Report Number 8, Champaign, 111.: National Council of Teachers of English (1967). Owens, R. E., Jr., Speech acts in the early language of non-delayed and retarded children: A taxonomy and distributional study (Doctoral dissertation, Ohio State University, 1978). Dissertation Abstracts International , 1978, 39, 2593A-3183A. (University Microfi Imsj. Piaget, J., The language and thought of the child . New York: Harcourt, Brace and World (1923). Prutting, C. A., Process: The action of moving forward progressively from one point to another on the way to completion. Journal of Speech and Hearing Disorders , kk, 3-30 (1979). Rees, N., Pragmatics of language. In R. L. Schiefel busch (Ed.), Bases of language intervention . Baltimore: University Park Press, 191" 268 (1978). Ricks, D. M., Vocal communication in pre-verbal normal and autistic children. In N. O'Connor (Ed.), Language, cognitive deficits and retardation. London: Butler-Worth (1975). Ricks, D. M., & Wing, L. , Language, communication and the use of symbols in normal and autistic children. Journal of Autism and Childhood Schizophrenia , 5, 191-221 (1975). Ritti, A. R., Social functions of children's speech. Journal of Communication, 28, 35-^^ (1978). «

PAGE 286

276 Ryan, J., Early language development: Towards a commun i cat iona 1 analysis. In M. P. M. Richards (Ed.), The integration of a child into a social world . London: Cambridge University Press, 185-213 (197M. Sabsay, S., Communicative competence among the severely retarded: Some evidence from the conversational interaction of Down's syndrome (mongoloid) adults. Paper presented at the Linguistic Society of America, San Francisco (1975). Schachter, F. F. , Kirshner, K. , Klips, B., Friedricks, M., & Sanders, K. , Everyday preschool interpersonal speech usage: Methodological, developmental, and sociol ingui stic studies. Monographs of the Society for Research in Child Development , 39, No. 156 {^37^)Schaffer, A. R., The growth of sociability . Baltimore, Md.: Penguin Press (1971). Schmidt, R., 6 Erikson, M. T, , Early predictors of mental retardation. Mental Retardation , 11, 27-29 (1973). Searle, J. R. , Speech acts: An essay in the ph i losophy of language . London : Cambridge University Press (1969). Searle, J. R. , A taxonomy of speech acts. In K. Gunderson (Ed.), Language, mind, S knowledge . Minnesota studies in the Philosophy of Science, Vol. Vll. Minneapolis, Minn.: University of Minnesota Press (1975). Sharf, D., Some relationships between measures of early language development. Journal of Speech and Hearing Disorders , 37, 64-7A (1972). Shotick, A., 6 Blue, M., Influence of CA and IQ levels on structure and amount of spontaneous verbalization. Psychological Reports , 29, 275-281 (1971). Simon, C. S., Communicative competence: A functional-pragmatic approach to language therapy . Tucson, Arizona: Communication Skill Builders (1979)1 Skarakis, E. A., & Prutting, C. A., Early communication: Semantic functions and communicative intentions in the communication of the preschool child with impaired hearing. American Annals of the Deaf , 122, 382-391 (1977). Skinner, B., Verbal Behavior . New York: Appleton-Century-Crofts (1957). Slobin, D. I., Universals of grammatical development in children. In G. B, Flores d'Arcais & W. J. M. Leveit (Eds.), Advances in psycho1 inguistics . Amsterdam: North-Holland Publishing ( 1 970) . Slobin, D. I., Psychol i ngu i st i cs . Glenview, 111.: Scott, Foresman & Co. (1971).

PAGE 287

277 Smith, M. E., A study of some factors influencing the development of the sentence in preschool children. Pedagogical Seminary and Journal of Genetic Psychology , k(>, 182-211 (1935). Snyder, L. S., Pragmatics in language-disabled children: Their prelinguistic and early verbal performatives and presuppositions. Unpublished doctoral dissertation, University of Colorado (1975)Soskin, W. F, , 6 John, V. P., The study of spontaneous talk. In R. Barker (Ed.), Stream of Behavior . New York: Appl eton-Century-Crof ts (1963). Spolsky, B., The language barrier to education. In G. E. Perren (Ed.), Interdisciplinary approaches to language . London: Center for Information on Language Teaching Reports and Papers, 6 (1971). Tonkovich, J. D., 6 Adler, S., Content and sequence analysis of communicative interaction: A research tool. Paper presented at the American Speech and Hearing Association Convention. San Francisco, California (1978). Tough, J., Listening to children talking . London: Ward Lock Educational (1976) . Tough, J., The development of meaning . New York: John Wiley and Sons (1977) . Trantham, C. R. , & Pedersen, J. K. , Normal language development: The key to diagnosis and therapy for language disordered children . Bal t irrore : Wi 1 1 i ams and Wilkins Co. ( 1 976) . Tyack, D., & Gottsleben, R. , Language sampling, analysis and training . Palo Alto, Calif.: Consulting Psychologists Press (197^). Ulatowska, H. K. , Macal uso-Haynes, S. , s Mendel-Richardson, S., The assessment of communicative competence in aphasia. In R. H. Brookshire (Ed.), Clinical aphasiology conference proceedings . Minneapolis, Minn.: B. R. K. Publishers (1976). Vygotsky, L. S. , Thought and language . Cambridge, Mass.: MIT Press (1962). Weiner, F., Phonological process analysis . Baltimore, Md.: University Park Press (1978). Wells, G., Coding manual for the description of child speech . Bristol School of Education (1973). Wells, G., Language use and educational success: An empirical response to Joan Tough's 'The development of meaning.' Research in Education , 18, 9-3^ (1978).

PAGE 288

278 White, B. L., Fundamental early environmental influences on the development of competence. In M. E. Meyer (Ed.), Third Western Symposium on Learning: Cognitive Learning . Bellingham, Wash.: Western Washington State University Press (1972) . Wilcox, M. J., & Davis, G. A., Speech act analysis of aphasic communication in individual and group settings. In R. H. Brookshire (Ed.), Clinical aphasiology conference proceedings . Minneapolis, Minn.: B. R. K. Publishers (1977). Williams, F. , & Naremore, R. C, On the functional analysis of social class differences in modes of speech. Speech Monographs , 36, 77" 102 (1969). Williams, F. , & Naremore, R. C. , Social class differences in children's syntactic performance: A quantitative analysis of field study data. Journal of Speech and Hearing Research , 12, 777*793 (1969). Wittgenstein, L., Philosophical investigations . Oxford: Basil, Blackwell & Mott (195Hir Wollner, S. G., S Geller, E., How to evaluate communication in a somewhat pragmatic way. Paper presented at the American Speech and Hearing Association Convention. Houston, Texas (1976). Wood, B. S., Children and communication . Engl ewood-C 1 i f f s , N.J.: P rent i ce-Hall, Inc. (1976).

PAGE 289

BIOGRAPHICAL SKETCH On August 10, ]Sk(>, Rita Jane Lieberman was born in Louisville, Kentucky, where she resided until the age of six. For the next 16 years, she lived in the state of Ohio and graduated from Warren G. Harding High School in 1964. In March 1968, she graduated summa cum laude with a Bachelor of Science degree in speech pathology from Ohio State University and in December 1970, she received a Master of Science degree in speech pathology from the University of Michigan. In September 1971 i she began her doctoral studies at the University of Florida after returning from two years as a Peace Corps Volunteer in Brasilia, Brazil. As a volunteer, she worked in several rehabilitation settings to train paraprofessionals in the field of speech pathology. From 1971 to 1977, she was supported at various times by an Office of Education Fellowship, a Veteran's Administration Traineeship and a University of Florida Graduate Teaching Assistantship. In 1976, she was awarded the Outstanding Teaching Assistant of the year by the graduate school. Her major areas of study were language impairment in children and in adults. Her work experience has included clinical positions in the public schools and at the Miami Rehabilitation Center for Crippled Children and Adults. From November 1978 to the present, she has been employed as an instructor in the Department of Speech Pathology and Audiology at Appalachian State University in Boone, North Carolina. 279

PAGE 290

280 She is married to Charles B. Lieberman and they have one son, Justin Edward.

PAGE 291

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Thomas B. Abbott, Chairman Professor of Speech I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Anthony J. Clark Associate Professor of Speech I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Leonard L. LaPointe Associate Professor of Speech Coordinator of Instruction Audiology and Speech Pathology Service Veterans Administration Hospital I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. 'JayiT^ C. Harder Proressor of Linguistics

PAGE 292

This dissertation was submitted to the Graduate Faculty of the Department of Speech in the College of Liberal Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August 1981 Dean for Graduate Studies and Research


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 E22NJ9AW5_RLPRXL INGEST_TIME 2015-05-14T20:41:39Z PACKAGE AA00031504_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

TOWARD A THEORY OF LANGUAGE LEARNING: SENSORY INTEGRATION IMPLICATIONS OF THE DARTMOUTH INTENSIVE LANGUAGE MODEL BY FLORENCE L. WALTERS A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1982

PAGE 2

Copyright 1982 by Florence L. Walters

PAGE 3

ACKNOWLEDGMENTS Many people gave their time and support to the development of this study. Dr. Margaret Morgan, chairman of my doctoral committee, deserves my sincere appreciation for guiding my graduate education. I am grateful for the exceptional editorial skills and critical appraisals she applied to this dissertation. I offer heartfelt thanks to Professor Emeritus Dora Hicks for sharing hard-won time with me and for invaluable help as a committee member. My appreciation goes to Dr. James Hensel for serving on my committee and for answering my questions, both in class and out. Dr. Helga Kraft, director of the intensive language program at the University of Florida, served as a model of courage and creativity, and provided an open door to the hard-working, enthusiastic, and helpful students, tutors, graduate teaching assistants, faculty, and secretary in the Department of Germanic and Slavic Languages. Throughout the past twenty years, Kay Sieg, acting chairman of the Department of Occupational Therapy, provided personal and professional assistance toward the development of sensory integration theories expressed in this dissertation. I wish to convey special acknowledgment to Dr. Lei a Llorens, chairman of the Department of Occupational Therapy from 1976 to 1982. She provided a luminous path for research. Jim Flavin, television producer-director. Office of Instructional Resources, University of Florida, helped to perfect my videotaping i i i

PAGE 4

skills. Professor Raymond Johnson, director of the electronics communications laboratory, and Annette Lang, recipient of a fellowship to the University of Tubingen, Germany, are due special thanks for their assistance as "interobservers." I am also indebted to Carla Chesser, teacher of German, for the many hours she spent evaluating videotapes of student performances on the oral interview component of the "Zertifikat." I thank Dr. Stephen Olejnik, assistant professor. Foundations of Education, for making statistics not only understandable but enjoyable. I would also like to recognize the voluntary contributions of Dr. Norman Markel, professor of speech, psychology, and anthropology. His spirited encouragement and acknowledgment of the possibilities of communicative body movements sustained my work. My warmest appreciation goes to my friends, the Delonys, the Johnsons, the Wessels, Mary Heisler, and Earl Blekking, for their gentle support and encouragement. Finally, I wish my children, Carolyn Kolar, Nancy Pollock, William Walters, and Cynthia Padgett, to know that their daily gifts of love were a source of support throughout this endeavor. iv

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGMENTS iii LIST OF TABLES ^^^^ ABSTRACT ^ CHAPTER I. INTRODUCTION 1 Need for the Study 1 Rationale for the Study 3 Dartmouth Intensive Language Model 3 Neurophysiological Implications of the Model .... 5 Observation and Analysis of the Model 7 II. BACKGROUND LITERATURE AND THEORETICAL FRAMEWORK 10 Review of Literature in Foreign Language Education ... 10 Empirical Studies of Language Learning 12 Second Language Learner-Teacher Interaction Models . . 13 Neurophysiological ly Oriented Studies of Language Learning 17 Review of Studies Related to Underlying Concepts and Assumptions 18 Supportive Studies from Anthropology 20 Supportive Studies from Linguistics 21 Supportive Studies from the Bio-Behavioral Sciences . 23 Supportive Studies from the Neurosciences 25 Supportive Studies from Psychology 27 Supportive Studies from Education 28 Supportive Studies from Occupational Therapy 32 Summary of Supportive Studies 34 Theoretical Framework of the Study 34 Assumptions of the Study 35 Hypotheses 37 Definition of Terms 39 Teacher-Directed Sensory-Motor Stimulation 39 Student Foreign Language Communication Effectiveness . 41 Dynamic Teacher/Student Interaction 41 Model 42 Intensive Language Program 42 Traditional Language Program 43 V

PAGE 6

Page Communication 43 Kinesics 44 Language 44 Foreign Language 44 Limitations 44 Plan of the Dissertation 46 III. PROCEDURE 48 The Sample 49 Phase One: Subjects for Collection of Descriptive Data 49 Phase Two: Subjects for Collection of Inferential Statistical Data 49 Instrumentation and Collection of Data 50 Phase One: Instruments for Collection of Descriptive Data 50 Phase One: Collection of Descriptive Data 51 Phase One: Computations of Descriptive Data .... 52 Phase Two: Instruments for Collection of Inferential Statistical Data 52 Phase Two: Collection of Inferential Statistical Data 56 Analyses 60 Phase One: Descriptive Data Analyses 60 Phase Two: Inferential Statistical Analyses .... 61 IV. FINDINGS 64 Results of Descriptive Data Analyses 64 Null Hypothesis la 65 Null Hypothesis lb 67 Null Hypothesis Ic 67 Null Hypothesis Id 67 Null Hypothesis 1 69 Results of Inferential Statistical Analysis 72 Analysis of Covariance (ANCOVA) Results 72 Independent Chi-Square Test Results 79 Discussion of Descriptive Findings 87 Rejection of Null Hypotheses la, lb, Ic, and Id . . 87 Rejection of Null Hypothesis 1 89 Discussion of Inferential Statistical Findings .... 91 Rejection of Null Hypotheses 2, 2a, 2b, and 2c .. . 92 Failure to Reject Null Hypotheses 3a, 3b, and 3c . . 93 Rejection of Null Hypothesis 3d 94 Rejection of Null Hypothesis 4 95 V. CONCLUSIONS AND RECOMMENDATIONS 98 Conclusions 98 Recommendations 100 vi

PAGE 7

Page REFERENCES 106 APPENDIX A. SAMPLE PAGE FROM VIDEO BEHAVIORAL OBSERVATION CHART FOR ONE TEACHER 120 B. INTEROBSERVER RELIABILITY CHARTS .... 123 C. SAMPLE PAGE FROM VIDEO BEHAVIORAL OBSERVATION CHART FOR ONE TUTOR 126 D. STUDENT FOREIGN LANGUAGE DATA SHEET AND RATING FORM .... 129 BIOGRAPHICAL SKETCH 132 Vli

PAGE 8

LIST OF TABLES Table Page 1. Summary of Interobserver Reliability Data 65 2. Mean Seconds per Contact Hour of Kinesic and Vocal Language Behaviors by Intensive and Traditional Language Teachers . . 66 3. Total Mean Seconds per Contact Hour of Teacher-Directed Sensory-Motor Stimulation by Intensive and Traditional Language Teachers 68 4. Mean Seconds per Contact Hour of Kinesic and Vocal Language Behaviors by Tutors 70 5. Total Mean Seconds of Teacher-Directed Sensory-Motor Stimulation by Tutors 70 6. Level of Teacher-Directed Sensory-Motor Stimulation Provided for Students in Intensive and Traditional Language Programs 71 7. Summary of Covariance Analysis for Effect of Level of Teacher-Directed Sensory-Motor Stimulation on the Mean Foreign Language Communication Total Test Score 73 8. Adjusted Least Squares Means for Foreign Language Communication Total Test Scores 74 9. Summary of Covariance Analysis for Effect of Level of Teacher-Directed Sensory-Motor Stimulation on the Mean Grammar Structure-Vocabulary Test Score 75 10. Adjusted Least Squares Means for Grammar StructureVocabulary Test Scores 75 11. Summary of Covariance Analysis for Effect of Level of Teacher-Directed Sensory-Motor Stimulation on Mean Listening Comprehension Test Score 76 12. Adjusted Least Squares Means for Listening Comprehension Test Scores by Groups 77 13. Summary of Covariance Analysis for Effect of Level of Teacher-Directed Sensory-Motor Stimulation on the Mean Speaking Proficiency Test Score 78 viii

PAGE 9

Table Page 14. AdjustEd Least Squares Means for Speaking Proficiency Test Scores 15. Self-Assessed Reading Comprehension Ratings by Student Groups • 81 16. Self-Assessed Listening Comprehension Ratings by Student Groups 82 17. Self-Assessed Writing Proficiency Ratings by Student Groups 84 18. Self-Assessed Speaking Proficiency Ratings by Student Groups 85 19. Foreign Language Course Satisfaction Ratings by Student Groups 86 ix

PAGE 10

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TOWARD A THEORY OF LANGUAGE LEARNING: SENSORY INTEGRATION IMPLICATIONS OF THE DARTMOUTH INTENSIVE LANGUAGE MODEL By Florence L. Walters December 1982 Chairman: Margaret K. Morgan Major Department: Curriculum and Instruction Concern for the current need for research in foreign language education to reverse American monol inguistic tendencies prompted this study. It was designed to determine if neurophysiological implications of the Dartmouth Intensive Language Model could provide a basis for more effective language teaching. The apparent effectiveness of teaching behaviors prescribed in the model was observed and analyzed in university beginning German classrooms. A review of related studies supported assumptions of relationships between Ayres's sensory integration theories and Rassias's emphasis on the teacher as transmitter of effective foreign language communication. The theory that teacher-directed sensory-motor stimulation facilitates student foreign language communication effectiveness was tested in two procedural phases. In the preliminary stage, descriptive data collected from 12 videotapes documented that whereas instructors who used the intensive X

PAGE 11

language model provided students with 4447.67 total mean seconds per contact hour of stimulation via gross body movement, hand gestures, and vocalized German, traditional language instructors provided 1138.00 seconds of such stimulation. In the second phase, differences in achieving levels of two groups of students on grammar, listening comprehension, and speaking tests were analyzed with inferential statistics. Students provided higher levels of teacher-directed sensory-motor stimulation achieved significantly (alpha < .0001) higher test scores. Significantly (alpha < .01) greater percentages of students who were provided higher levels of teacherdirected sensory-motor stimulation rated their German-speaking skills "excellent" and their satisfaction with beginning German courses "high" than did students provided lower levels. Findings suggested that neurophysiological implications of teaching behaviors prescribed by Rassias could provide a basis for more effective language teaching. Further research was recommended to reduce internal validity and general izability limitations caused by nonrandomization. Studies are needed, not only in foreign language education, but in other aspects of education, to test the coherence of the theory that learning is a by-product of students' adaptive responses to sensory-motor stimulation provided by the kinesic and vocal language behaviors of teachers in an environment of dynamic teacher/student interaction. x1

PAGE 12

CHAPTER I INTRODUCTION The geographical location of the United States has long shielded Americans from the necessity of learning foreign languages. However, in today's shrinking world, where people of different nationalities are in close and constant contact, those who lack facility with foreign languages are culturally disadvantaged. Traditional instruction in foreign languages has been of limited value in reversing American monol inguistic tendencies. But as global changes have compelled a reversal of these tendencies, educators have had to learn how language learning can be facilitated and how to develop more effective teaching behaviors. This study was designed to determine if the neurophysiological implications of teaching behaviors prescribed in the Dartmouth Intensive Language Model could provide instructors with a basis for effective teaching. The study was based on the belief that an understanding of how brain processes interact with the processes of language learning could help to explain how languages can be taught more effectively. Need for the Study Since 1965, foreign language teachers have struggled with the problems caused by sharp declines in the national regard for academic study of foreign languages (Brod, 1980). Critical self-examination revealed that most instructors operated with no specifically defined approach to teaching. The confusing array of eclectic approaches was found to be due, at least in part, to a lack of a coherent theory of 1

PAGE 13

2 language learning and a lack of knowledge as to how to teach (Benseler & Schulz, 1980). In other words, language instructors could find no scientific foundation on which to base their teaching. Lack of research in foreign language education forced instructors to depend on theories of language acquisition from psychology and linguistics, two fields dominated by stimulus-response theories of learning. Foreign language instructors accepted cognitive theories of learning from both psychology and linguistics, but acceptance of these theories did nothing to dispel "the continuing questioning of our [instructors'] efforts and the persistent dissatisfaction among learners" (Birkmaier, 1973, p. 1295). This dissatisfaction among learners was predictable from the results of a large-scale survey conducted in U.S. colleges and universities that showed the average college senior majoring in a second language to have a "limited working proficiency" in speaking and comprehending the foreign language (Carroll, 1967, p. 200). Along with experiencing frustration over their students' lack of fluency with languages, instructors trying to teach students to enjoy languages had to contend with a lack of suitable training in teaching methodology offered by institutions of higher education (Rivers, 1975). Instructors who succeeded in teaching students to use and enjoy the cultural benefits of languages operated from intuitive bases rather than from a theoretical base or from formal training (Jorstad, 1980; Strasheim, 1971). Furthermore, informal learning methods used by children, refugees, immigrants, Europeans with close foreign-speaking neighbors, and Peace Corps volunteers often proved more effective than teaching and learning in academic settings (D'Anglejan, 1978). Yet

PAGE 14

3 the creative and natural forms of language learning were seldom studied for their potential to provide a conceptual framework for language teaching. The President's Commission on Foreign Language and International Studies focused national attention on the need to reverse American monolinguistic tendencies (Office of Education, 1979). If instructors were to improve language teaching, they would need more research on teaching behaviors that might be effective. Rationale for the Study One approach to teaching languages that appears to have been effective is the Dartmouth Intensive Language Model. In the present study, the investigator focused on neurophysiological implications for language learning as exemplified in that model. Dartmouth Intensive Language Model Current concerns for foreign language teaching were presaged by John A. Rassias when he introduced an experimental model in 1967 to remedy low and declining student enrollments and uninspired teaching at Dartmouth College, Hanover, New Hampshire. Rassias (1970) developed the Dartmouth Intensive Language Model from his belief that teachers must strive to create environments that instill in the student a desire to communicate in a foreign language. The model was based on his contention that instructors must "learn how to communicate and learn how to teach others to communicate" ( Effective Language Teaching , 1972). Instructors who studied this approach to language teaching were exposed to the precepts of Rassias in workshops where they participated actively in exercises and drills devised to reduce inhibitions

PAGE 15

4 to communication. They learned to use communication behaviors by mastering a choreographed rhythm and pace of verbal and nonverbal cues. These cues made use of chorals, finger-snaps, pointing gestures, acknowledging expressions, and large body movements. Teachers were instructed to speak to their students only in the target language and to move energetically, rather than stand still. Emphasis was placed on the communicative value of dramatization and animation for giving students positive feedback and sincere attention. Instructors who applied these communicative teaching behaviors in their classrooms expressed exhilaration at having effected student learning. Their students expressed astonishment at how easily they began to communicate in foreign languages. Both teachers and students valued opportunities to interact and communicate without pretense in a learning environment that was far removed from more passive, traditional approaches to language teaching. The effectiveness of teaching behaviors prescribed in the model was apparent from objective as well as subjective accounts. Rassias (1970) reported that after the model was implemented at Dartmouth, enrollments increased and students learned more easily to communicate effectively in languages other than their own. National recognition of the success of the model (U.S. Congress, 1974) prompted other colleges and universities with flagging language programs to adopt Rassias's precepts for teaching. These educational institutions reported an increase in enrollments and a rise in students' confidence in their oral communication skills and their enthusiasm for further language study (Byrd, 1980; Johnston, 1980; Stansfield and Hornor, 1981).

PAGE 16

5 Rassias places emphasis on communication as a key element in the art of language teaching. This emphasis stems from humanistic theories of the need people have for sharing the experiences of life. Within the context of the classroom, Rassias (1970, 1972) believes that the student learns to communicate effectively through the us£ of language, not through the analysis of grammatical structure. The teacher strives to bring out the uninhibited child in each student, so as to recreate that state of sensitivity that permits native language learning. The teacher is expected to direct energy toward moving and speaking with native-like fluency while interacting dynamically with the student. The effectiveness of the teacher is explained by the "reality" of the teacher who through energy expended in communicative body movements and vocal language behaviors, transmits commitment to communicating in foreign languages and effects student learning through all of the senses. Neurophysiological Implications of the Model In the present study, the investigator attempted to relate the humanistically based theories of Rassias to biological and neurophysiological theories of the efficacy of learning through all of the senses. Learning through all of the senses has proven to be the easy way that behavior evolves fromoneform of life to another through the development of the species and from one stage of life to another through the development of the individual. Coghill (1929) and Herrick (1956) discussed this phylogenetic and ontogenetic validity of learning through the senses and concluded that all forms of life require sensory stimulation to adapt and

PAGE 17

6 learn. Ayres (1972b) went on to develop a theory of sensory integration that is based on the assumption that sensory information from the environment effects sensory integration processes throughout the nervous system that enable human beings to interact effectively with the environment. Sensory integration processes involve lower, less complex, subcortical structures of the brain as well as higher, more complex, cortical structures. Ayres developed these theories to show the effectiveness of organism-environment interaction and the resulting sensory-motor stimulation for subcortical as well as cortical learning. Although these neurophysiologically based theories were developed in special education and occupational therapy, in the present study these theories were assumed to contain concepts that relate to the humanistic theories of dynamic teacher/student interaction developed by Rassias. Biological and neurophysiological implications of teaching behaviors are seldom related by educational theorists to effective learning behaviors. Yet, throughout history, scholars (e.g. , Bacon, Comenius, Dewey, James, Montessori, Ogden, Pestalozzi, Whitehead) have related the art of teaching to the development of human potential through the use of natural integrative learning processes of living, doing, and experiencing. More recently, biological and neurophysiological implications for developing human potential through multi sensory stimulation were expressed by Coppola (1970), Edwards (1979), Montague (1978), and Nelson (1977). Several writers, including Chappie (1981), Gorman (1974), Holt (1967, 1981), Hyman (1974), and Suzuki (1969,

PAGE 18

7 1973) , have expressed a need for dynamic teacher/student interaction for the development of human potential. Gage (1964) described the general educational need for fundamentally based theories of teaching as the desire to know "the ways in which a person influences an organism to learn" (p. 268). By reminding teachers of the role played by the central nervous system in learning, Pribram (1964) set a precedent in education that was later enunciated by Chall and Mirskey (1978) and by Wittrock (1978, 1980). This precedent in education, along with the precedent set by Ayres in occupational therapy, supported the theoretical framework for the present investigation. Observation and Analysis of the Model In the mid-1970s, the Department of Germanic and Slavic Languages at the University of Florida faced declining enrollments in beginning German courses. Findings from a large-scale departmental study (Von Gal, 1974) indicated student dissatisfaction with the instructors' emphasis on grammatical structure; 70% of students queried were dissapointed by their inability to communicate in German after three quarters of study. These weaknesses in the established program prompted the experimental implementation of the Dartmouth Intensive Language Model in 1978 in some sections of beginning German (Johnston, 1980). Since then, students who register for beginning German courses may elect to enter either the experimental program or the program using traditional instruction. The investigator observed and analyzed aspects of the intensive beginning German experiment during 1980-1982. Teachers in intensive beginning German classrooms expended high levels of energy to create

PAGE 19

8 optimum learning environments. Teachers moved about the classrooms constantly, usually using a choreographed pattern of communicative body movements and vocal language behaviors. Analysis of these behaviors differentiated communicative body movements into (a) gross body movements of the head, trunk, and appendages on both vertical and horizontal planes, and (b) gestures in the form of finger-snaps, pointing movements, acknowledging hand movements, and other nonverbal communicative hand movements. Vocal language behaviors were distinguished by (a) native or near-native German fluency, and (b) omission of English from speech— teachers spoke only in German and encouraged students to do likewise. Despite feelings of exhaustion after conducting class, instructors claimed that they were exhilarated by student learning effected by their efforts. Observation and analysis of student behaviors in intensive beginning German classrooms showed students to be highly stimulated by the teaching behaviors prescribed by Rassias. After the first few days of classes, most students lost fears of making mistakes and shed inhibitions to speaking in German. Students freely expressed enthusiasm for the teachers and methods that enabled them to communicate. Students made such positive comments as, "These classes are alive!" "You have to talk out loud in German— at first it's scary, but after a day or two, you realize you can do it!" "You can't help but learn." Negative reactions were rare and usually suggested students' mistrust of nontraditional , nonliteral learning strategies. Typical comments were, "How do they expect me to know what to say if I can't see the words ?" "They go too fast— I can't think that fast." "It takes too

PAGE 20

much of my time— I've got a heavy load of reading for my other classes this term." The alertness, attentiveness, and responsiveness noted in students in intensive beginning German class seemed to be related to the communicative body movement and vocal language behaviors of the instructors. Students appeared to be learning languages as children do in what Piaget (1926/1959) called the sensory-motor developmental stage. The importance of a developmental sequence for learning was noted by Ayres (1972b) as a need to develop the function of subcortical sensory integration processes in the brain before the higher cortical processes carv be expected to function effectively. Stated simply, students appeared to be learning to communicate in a foreign language the easy way— subconsciously , through their senses.

PAGE 21

CHAPTER II BACKGROUND LITERATURE AND THEORETICAL FRAMEWORK The present study was prompted by a need for research in foreign language education to provide instructors with knowledge of how to teach, based on a coherent theory of language learning. The rationale for this study was based on the assumption that the neurophysiological implications for language learning in communicative body movements and vocal language behaviors prescribed in the Dartmouth Intensive Language Model can explain how languages can be taught more effectively through teacher-directed sensory-motor stimulation. This chapter addressed (a) a review of literature related to research in foreign language educatibn and to other concepts underlying the present study, (b) a theoretical framework, (c) assumptions on which the study is based, (d) hypotheses, (e) definitions of terms, (f) limitations of the study, and (g) the plan for the dissertation. Review of Literature in Foreign Language Education Volumes have been written on foreign language education. Yet Hosenfeld (1979) found the results of her literature review disappointing in that "language teachers still do not know which of the many available teaching approaches to select . . . moreover, language learners do not appear to have achieved increased proficiency from the use of recommended techniques" (p. 51). Responding to a request from the President's Commission on Foreign Language and International Studies to 10

PAGE 22

11 assess the results of methodologies for foreign language education on the college and university level, Benseler and Schulz (1980) agreed with Hosenfeld. Their findings confirmed disturbing developments in the field which led the chairman of the President's Commission to write. But the hard and brutal fact is that our programs and institutions for education and training for foreign language and international understanding are both currently inadequate and actually falling further behind. (Perkins, 1979, p. 1) The gravity of the educational problem was illustrated by data showing that "only 8 percent of American colleges and universities now require a foreign language for admission, compared with 34 percent in 1966" (Office of Education, 1979, p. 7). Student enrollments in college and university foreign language programs were reported to have fallen 37.4% in German, 36.6% in French, and 31.7% in Russian since 1968 (Primeau, 1979). Figures showing an increase in enrollments at Dartmouth during these same years (Luxenberg, 1978), when coupled with evidence of inadequate teacher education (Warriner, 1979), influenced the President's Commission to state. There is an urgent need for bettertrained teachers and for extensive retraining of those already serving in the nation's classrooms, particularly in view of widespread agreement that the decline in enrollment is in large measure a response to poor instruction. (Office of Education, 1979, p. 8) Instructional problems instigated much of the foreign language education research emanating from psychology and linguistics (Oiler and Richards, 1973). Yet, the value of basing the teaching of language skills on Skinner's behavioristic theories or on Chomsky's transformational grammar theories was questioned by Chomsky himself. Addressing foreign language teachers, Chomsky (1973) warned "that suggestions from the 'fundamental disciplines' must be viewed with caution and skeptici,sm"

PAGE 23

12 (p. 29). Carroll (1975) restated this theme in wondering whether cognitive psychology could offer teachers a research base for methodology. Empirical Studies of Language Learning Since the mid-1970s, researchers have begun to explore alternative theories of language learning on which to base teaching. Examples of empirical studies that indicate the trend away from unquestioned confidence in mechanistic stimulus-response experiments from psychology and the preoccupation with syntax in linguistics include: (a) an investigation of teaching behaviors in treating errors in oral work by Fanslow (1977), (b) a study by Galloway (1980) on the perceptions of native and non-native raters of student communication effectiveness, (c) an investigation of the effects of extended imitative speech practice on fluency (Hieke, 1981), and (d) dissertations such a:S Jackson' s study (1979) of the effects of visual -aural versus aural methods of presenting interviews on listening comprehension and Dye's study (1977) of the use of the student's body movement to facilitate second language learning in secondary schools. Chun (1980), surveying the literature, found growth in the number of empirical studies addressing first and second language relationships. Even though these researchers (e.g., Bailey, Madden, & Krashen, 1974; Olson & Samuels, 1973; Snow & Hoefnagel-H6"hle, 1977) reported investigations of different affective, cognitive, and developmental aspects of language learning, their findings were similar in rejecting the popular assumption of a critical period of language learning which prevents older learners from achieving verbal fluency or native-like pronunciation.

PAGE 24

13 Eckman and Hastings (1977) introduced a collection of such studies by stating that the similarities of first and second language learning are at least as important as any dissimilarities, "and our progress in understanding either may depend, in part, on our ability to identify, elucidate, and explain these similarities" (p. ix). Sheldon (1977), a contributor to this theme, argued for additional research designed to observe "the interaction of child and parent, or second language learner and teacher" (p. 11 ). Second Language Learner-Teacher Interaction Models A review of the foreign language literature reveals methodologies based on the interactive nature of first language learning and the need for teacher/student interaction. These methodologies are claimed to be based on psychologically oriented theories of behavior. Total Physical Response . Asher was apparently the first psycholinguist to apply a theory of the interactive nature of first language learning in the classroom. Through a series of quasi-experimental studies, Asher (1969, 1977) demonstrated that it is possible to accelerate the rate at which adults and children acquire listening comprehension in a second language. The Total Physical Response approach was developed on the premise that in order to learn a foreign language "gracefully with a minimum of stress" a strategy is needed "that is in harmony with the biological program" (Asher, 1972, p. 134). Later, the concept underlying his strategy was described as the acceleration of the assimilation of information and skills through the use of the kinesthetic sensory system (Asher, 1977). The success of the approach in overcoming the often frustrating and traumatizing effects of

PAGE 25

14 traditional teaching methods is explained by Asher (1981) as the result of its "sensorimotor method" and of "right-brain learning" (p. 54). Counsel ing-Learning . Curran (1972), a priest and counselor who recognized that students have emotional investments and somatic reactions, developed the Counseling-Learning approach. The relationship of the affective domain to the cognitive domain as well as the human need to be understood, provides the theoretical foundation for methods prescribed to instill receptiveness, trust, and commitment between student and teacher. Grammar and vocabulary are considered to be less important than the interaction of persons. Suggestopaedia . Suggestopaedia is a teaching method developed by Georgi Lozanov (Racle, 1979). An original feature of this method is that students are required to engage in relaxation exercises as the teacher reads new material with varying levels of loudness while baroque music is played in the background to aid in the unconscious absorption of the material. Racle (1979) reports results of experiments in Canada to confirm that the approach has "psychological advantages" for learning to use language as a natural communication process (p. 47). Monitor Model . Although not developed as a specifically applied method of teaching, Krashen's Monitor Model (1979) has relevance for the present investigation of assumptions of subcortical learning. Krashen assembled empirical evidence suggesting that adults acquire language in the same subconscious manner as do children. He compares his rationale for distinguishing between acquisition strategies and classroom learning strategies with the rationale of Gallwey (1974) who teaches tennis by emphasizing the importance of subcortical sensory-integrative brain processes for learning a skill.

PAGE 26

15 Formal and informal learning . D'Anglejan (1978) reviewed the literature relating to the nature of formal and informal language learning from a variety of psychol inguistic and sociolinguistic perspectives for teaching children in the schools of Quebec. Even though applications of D'Anglejan' s findings may be beyond the limited scope of university teaching, their correspondence to the findings of researchers investigating adult foreign language learning indicates the significance of her con elusions to the conceptual framework of the present study. The logic used in arriving at the conclusion that complex language skills can be acquired better in the absence of formal instruction is stated as follows: Since language or communication is in fact behavior, and not a body of knowledge, it seems normal that it should essentially be acquired as the result of communicative interaction between the learner and a well-disposed native speaker. This is clearly the case in native language acquisition, which is universally successful. . . . There is reason to believe that simply rubbing shoulders with native speakers is not sufficient— the learner must receive an input of the second language directed to him by a concerned speaker. This input must be embedded in a context of social interaction. (D'Anglejan, 1978, pp. 233-234) D'Anglejan called for a better understanding of learning and increased respect for the natural capacity for language acquisition which learners bring with themselves into the classroom. Dartmouth Intensive Language Model . Respect for the natural capacity of the student to communicate in a foreign language is a foremost pee cept of the Dartmouth Language Model. The model was developed by Rassias, a professor of languages with a strong interest in drama who came to Dartmouth in 1964 after studying in France on a Full bright Scholarship and earning a doctoral degree at the University of Dijon (Rassias, 1976). Humanistic concerns for the art of teaching that grew

PAGE 27

16 out of close contact with trainees and volunteers in the Peace Corps, both in the United States and overseas, influenced Rassias to eschew scientific quantitative research in favor of the practical challenge of restoring vitality and humanity to foreign language education (Rassias, 1972). Preoccupation with the practical and qualitative aspects of teaching and a primary concern for the student's natural humanistic capacity for language learning may be among the reasons for the "strange" lack of scientific research to identify and explain the effectiveness of the Dartmouth Intensive Language Model noted by Stansfield and Hornor (1981, p. 25). Support for the Dartmouth model comes from a number of sources. Christensen (1979) in reviewing teaching behaviors that require "creativity, imagination, trust, humor, and acceptance of the student as a creative person" mentions Rassias as the developer of a "highly audiolingual approach which has gained wide recognition" (pp. 96, 127). While searching for instructional approaches that might be used to generate student interest and improve teaching and learning, Schulz (1978) included .the Dartmouth Intensive Language Model in her extensive study of foreign language programs. Schulz reported that intensive courses "rate among factors for students with high positive attitudes toward language learning . . . and contribute to initial enrollment and retention of students" (p. 43). In reporting the advantages of intensive programs, Keilstrup (1981) included the "well-known Dartmouth model" as a successful example (p. 377), In an article describing the history of the implemented DartmouthRassias model, Stansfield and Hornor (1981) reported (a) a fourfold

PAGE 28

17 increase in the number of language majors at Dartmouth, (b) a rise of 86% in foreign language enrollments at Washington University, and (c) the spread of the model to at least 55 colleges and universities. The account of Stansfield and Hornor of the success of the model at the University of Denver coincides with accounts from Emory and Henry College (Byrd, 1980) and the University of Florida (Johnston, 1980). Even though the Dartmouth Intensive Language Model is recognized in education and even though social scientists see evidence of the effectiveness of the methods prescribed by Asher and Lozanov, little in the foreign language literature explains language learning from a neurophysiological perspective. Neurophysiologically Oriented Studies of Language Learning Walsh and Diller (1978) made an attempt to link applied linguistic theory to the structure and function of language centers in the cerebral cortex of the human brain. Despite their recognition of "the possibility that language centers do far less than at first thought" (p. 3), Walsh and Diller neglected the contribution of subcortical sensory integration processes known to mediate brain functions. They limited their investigation to matching neurol inguistic pathways between Wernicke's Area, Broca's Area, and other cortical regions with cognitive strategies used in three different methods of language teaching. Lamendella (1977) made a more comprehensive analysis of neurophysiological implications for language acquisition. The significance of Lamendella's work— an effort to clarify terminologies applied to various types of first and second language learning— 1 ies in his recognition that "it is the neural systems that are responsible for

PAGE 29

18 accomplishing learning of any sort" (p. 156). Lamendella also recognized the role of communication in language learning by distinguishing between "learning to communicate in the target language" at the natural level and "learning the target language" at the cognitive-translation level (p. 182). The general principles of neurofunctional organization presented in this treatise were derived from information-processing theories in psychology. The perspectives of Lamendella and of Walsch and Diller are similar to those of Heilman (1978) who stated that his purpose in writing about language and the brain was "to provide the educator with some fundamental understanding of the neurophysiological processes underlying languages" (p. 143). The belief that the understanding of sensory integrative brain processes may provide educators with a conceptual framework for language teaching led to analysis of theoretical and empirical studies related to this belief. Review of Studies Related to Underlying Concepts and Assumptions The theory that learning to communicate in a foreign language is facilitated by teacher-directed sensory-motor stimulation is supported by underlying concepts of human communication, human development, and language learning. These concepts emerge from interdisciplinary studies in the broad fields of anthropology, linguistics, the bio-behavioral sciences, the neurosciences, psychology, education, and occupational therapy. Corroboration for assumptions that relate communicative body movement to communication and sensory integration theories derives from the work of the anthropol igist, Birdwhistell (1952, 1970, 1974), who created

PAGE 30

19 the term "kinesics" to define body movement as related to the nonverbal aspects of interpersonal communication. .. Birdwhistell stated. Human communication is much broader than the exchange of words in discrete messages with silences between them. My premise is that communication is a continuous multi sensory process. . . . Present research indicates that fascinating and complex relationships are operative in the patterned activities of the various sensory modalities utilized in the communication process. (1974, p. 203) The basis for these "fascinating and complex relationships" is found in the work of Ayres (1972b, 1981) to define the sensory integration process that organizes sensory information for use by human beings as they learn language as an adaptive response. The pioneering study of Spitz and Wolf (1946), describing the smiling response in babies, provided a focal point for the perspectives of researchers seeking a relationship among human developmental theories of communication, kinesics, and language learning. Spitz and Wolf, psychologists with a clinical interest in infant emotions and the psychosomatic effects of sensory deprivation, found that "emotional stimulation provokes a body response, and conversely somatic stimulation brings forth an emotional response" (p. 60). Additionally, they hypothesized reasons for the development of language throughout the history of the species, as well as for the development of language throughout the life of each individual. Spitz and Wolf believed that language develops phylogenetically and ontogenetically because the upright posture of humans permits face-to-face interaction and because the liberated hand in humans frees the mouth and facial muscles for purposes of speech and expressions of emotion. A similar view of the interaction of motive and emotive aspects of behavior was later taken by Tomkins and Izard (1965) who suggested

PAGE 31

20 that emotions manifested in "facial responses that communicate and motivate at once both publicly outward" and privately "backward and inward" (p. viii). The discovery that human beings are capable of interpreting an extraordinary amount of sensory information from momentary, slight facial responses prompted Tomkins and McCarter (1964) to suggest the existence of a "language of the face" (p. 125). Supportive Studies from Anthropology Birdwhistell (1970) considered facial expressions as one element of nonverbal communicative behavior that, along with speech, constistutes language learned culturally. Recent compilations of theoretical and empirical studies conducted in the field of nonverbal communication by Kendon, Harris, and Key (1975), Schiefelbusch & Hollis (1980), and Siegman and Feldstein (1978) emphasize the growing realization that body movement is not merely an accompaniment to speech. Researchers of body movement and its effects (e.g., Davis, 1975; Hall, 1959, 1966, 1968; Key, 1975, 1977; Mehrabian, 1967a, 1967b, 1972, 1976; Sarles, 1974, 1977; Scheflen, 1968) have found that verbal language does only part of the work it was previously thought to do as a function of meaning, learn ing, and communication. For example, Mehrabian reported "that the combined effect of simultaneous verbal, vocal, and facial attitude communications is a weighted sum of their independent effects— with the coefficients of .07, .38, and .55, respectively" (Mehrabian & Ferris, 1967a, p. 252). In her comprehensive study of the nonverbal communication literature. Key (1977) unequivocally stated that "human communication is body movement" (p. 5).

PAGE 32

21 Supportive Studies from Linguistics • In linguistics, the emphasis on the cognitive functions of language tends to obscure the neurophysiological implications of body movements as they function in the use of language as a tool of communication. Nevertheless, abundant theoretical and scientific evidence in the literature supports the thesis that the expressive and receptive modes of language are dependent on the movement of the human anatomy and physiology as "man communicates through all his senses" (Ong, 1974, p. 166). In his classic treatise on human speech, Paget (1930) described the interaction among movements of sound, movements of speech, and movements of hearing by stating that the repeated pressures which we normally "hear" are produced by rhythmical variations of pressure of the air which is in contact with our ear drums. These rhythmical variations of air pressure are due to sound waves, which are themselves due to some rhythmical disturbance of the air at the source of sound. . . . The significant elements in human speech are the postures and gestures which produced them. We lip-read by ear. . . . It will be no disgrace to human speech if, after all, it should turn out to be (as I believe it is) a branch of human gesture, (pp. 2, 174, 196) Since this clear exposition of the interrelatedness of audition and articulation, researchers using spectrographic measurements and voice synthesizers have reached similar conclusions. In reviewing the literature on the perception of speech, StuddertKennedy (1974) cited numerous sources to defend his argument that "only through their articulatory origins can the temporally scattered and contextual variable acoustic, (and auditory) patterns of speech be understood" (p. 2371). Lieberman (1974) reviewed quantitative acoustical, anatomical, and physiological data on what linguists caTT the "suprasegmental

PAGE 33

22 prosodic" features of language. These features include the tones, modulation, pitch, stress, and other "musical" qualities of vocal sounds that occur along with, or aside from, speech. The results of Lieberman's review of data on these features of language showed that the primary acoustic cue that signals the intonation of an utterance is the contour, or the spatial quality, of the vocal frequency measured by the human listening apparatus. In other words, the acoustical cues that enable our nervous system to use verbal information from the environment are not derived from the structure of the words of a language, but from their spatial and musical tonal qualities. Of additional neurophysiological interest was the finding that the larynx assumes a different position in singing than in speech. In singing the larynx is insensitive to air pressure whereas in speaking the larynx is sensitive to "aerodynamic interactions" (Lieberman, 1974, pp. 2429-2433). Elsewhere Lieberman (1975) suggests that language skills are "automatized" skills that are similar to other aspects of motor activity, such as dancing, playing the violin, or driving a car (p. 92). Jaffe (1978) investigated neurophysiological implications of verbal and nonverbal communication, arguing that "cerebral hemisphere specialization in man has evolved under the selective pressure of face-to-face interaction" (p. 61). His theory that the left hemisphere processes verbal time-sequenced language while the right hemisphere processes paral inguistic, kinesic, and spatially oriented aspects of language has implications for developmental theories of language learning.

PAGE 34

23 Supportive Studies from the Bio-Behavioral Sciences In the study of human development, Piaget (1926/1959) considered the acquisition of language to be a by-product of the sensory-motor construction process whereby language and cognition develop organically through interaction with the environment. Bruner {1975a, 1975b, 1977) argued that the ontogenetic development of verbal language skills is rooted in earlier joint attention and joint activity behaviors observed when mother and child interact. Confirmation of these premises may be found in the observations of researchers conducting ethological or behavioral microanalysis of human interaction. The discovery by Condon and Ogston (1966) that the body of a speaker moves in a precise synchrony with the speech of the speaker led to a study by Kendon (1970) showing that the flow of movement in the listener may also be rhythmically coordinated with the speech and movement of the speaker. When this "interactional synchrony" was observed in neonates, Condon and Sander (1974) speculated that by the time these newly born infants begin to speak, they may have already laid down within themselves the form and function of a language system that encompasses "a multiplicity of interlocking aspects: rhythmic and synthetic 'hierarchies,' suprasegmental features, and paralanguage nuances, not to mention body motion styles and rhythms" (p. 101). The premise that the first stages in language acquisition are already advanced stages in the child's neurophysiological development of vocal and kinesic behaviors is supported by the work of Trevathen (1974, 1977) who filmed the interactive behaviors of mothers and their babies. Using microanalytic techniques, Trevathen identified gestures and other

PAGE 35

24 movement and vocal language behaviors performed reciprocally by mothers and babies to facilitate the exchange of motive and emotive information. By studying the speech available to children while they are learning language. Snow (1972, 1979) identified verbal patterns in the conversational mode of mothers' speech which showed that mothers believe that their babies are capable of reciprocal communication and interaction. These observations signified a shift away from viewing language as an object of knowledge to be learned as syntax. Recent collections of studies representing similar observations of the emergence of language in children (e.g.. Lock, 1978; Waterson & Snow, 1978) indicate an increasing consensus for the theory that language is a tool for the development of communicative competence learned as a by-product of natural biological processes. Biological implications of language learning were first discussed by Lenneberg (1967) who found in the neurological mechanism of language "a natural extension of very general principles of organization of behavior which are biologically adapted to a highly specific ethological function" (p. 324). In addition to suggesting that language evolves from innate capacities for learning behavior, Lenneberg suggested that aphasic symptoms were a disorder of the brain's temporally patterned activities in interchange between cortical and subcortical structures. His conclusions lend support to sensory integration theories of subcortical language learning. In part, he concluded that the neurological processes underlying language are not confined to cortical areas. ... In order to reconcile experimental findings regarding the relationship between the cerebral cortex and behavior in animals, and to relate these findings to our clinical experiences with man, it appears reasonable to assume that complex, species-specific behavior patterns, such as

PAGE 36

25 language, result partly through subcortical, highly centralized integrating mechanisms, and partly through interaction of activities on the most rostral levels. (Lenneberg, 1967, p. 222) Further evidence of the biological foundation of language may be found in ethology, the comparative study of the character of behavior patterns across species. Eibl-Eibesfeldt (1975) is one of the European ethologists whose recognition of phylogenetic adaptations has influenced sciences dealing with the biological bases of human behavior. Eibl-Eibesfeldt described observations suggesting that expressive movements between cultures lie not only in such basic expressions as smiling, laughing, crying, and facial expressions of anger, but in whole syndromes of physiological actions. For example, he observed that expressions of anger are characterized cross-culturally, as well as in those born deaf and blind, "by opening the corners of the mouth in a particular way and by frowning, and also by clenching the fists, stamping on the ground, and even by hitting at objects" (Eibl-Eibesfeldt, 1972, p. 299). Such evidence, augmented by similar evidence from the anthropologists, Ekman and Friesen (1971 ), who describe face and emotion constants across cultures, indicates that common motor patterns and their neural substrates may be phylogenetic adaptions that are useful for communication and efficient for interaction with the environment. Supportive Studies from the Neurosciences The need for interaction with the environment in order to adapt and learn is a basic tenet in the neurosciences. Even the simplest forms of life behave and respond to changes in the environment for survival. To be alive is to be sensitive to the physical world (e.g., to

PAGE 37

26 receive and measure vibrations, to calculate pressures, to detect levels of light and movement). The role of the sensory systems to detect and interpret spatial and temporal information is documented in the work of Sarnat and Netsky (1974) to trace the evolution of the nervous system and in the work of Jacobson (1978) to trace the development of the human nervous system throughout the life of the individual. Knowledge of the human nervous system has been increasing since the classic studies of Sherrington (1906/1973) to map the functional tracts in the spinal column and brain. In the process of his studies, Sherrington discovered sensory organs in skeletal muscles, isolated the point of response in the synapse, and speculated on the integrative function of neuronal networks. More recent contributions to this knowledge of neurophysiology have come from neuromolecular studies such as those reported by Eccles (1966, 1967, 1977), Edelman and Mountcastle (1978), Hubel (1979), and Szentagothai and Arbib (1976). The concept that "the brain is a complex of widely and reciprocally interconnected systems and that the dynamic interplay of neuronal activity within and between these systems is the very essence of brain function" was confirmed by the study of Mountcastle (1978, p. 7) to find organization patterns in the estimated 50 billion cortical nerve cells and their seemingly infinitely possible connections. Mountcastle's finding that neural organization occurs vertically as well as horizontally in the human cortex substantiates sensory integration theories; he concluded that "there is nothing intrinsically motor about the motor cortex, nor sensory about the sensory cortex" (p. 9).

PAGE 38

27 Supportive Studies from Psychology Gibson (1966), a psychologist, reached a similar conclusion about interdependency in sensory and motor organization in his classic study of the senses as perceptual systems. He described the senses as active, moving mechanisms for looking, listening, sniffing, tasting, and touching, rather than as previously described passive receptors for sensations. In realizing that sensory-motor responses are both affected and effected by changing patterns and transformations of energy that generate the information organized by sensoryintegrative processes, he advanced the theory "that the senses can obtain information about objects in the world without the intervention of an intellectual process" (p. 2). Gibson's theories have implications for reforming learning theories that are limited to concepts of static cortical cognitive structures and memory storage areas. Experimental support for the implied effectiveness of subcortical sensory-integrative brain processes for learning was provided by Lashley (1929, 1950/1974). His neurophysiological experiments on animal behaviors demonstrated the fallacy of the hypothesized engram, or memory trace in the brain, and the limitations of passive, mechanistic aspects of stimulus-response learning. Three other psychologically oriented scientists who have given support to neurophysiological theory of languages are (a) Hebb (1949, 1980) whose "cell assembly" theory suggests the unity of the brain's neural networks as a biological learning device requiring interaction with the environment for development; (b) Young (1971, 1978) who elucidates the biological foundations of the "programs of the brain"; and (c) Pribram (1971, 1978) whose extensive neuropsychological studies

PAGE 39

28 indicate the interrelatedness of all areas of the central nervous system that communicate in "the languages of the brain," Supportive Studies from Education In education, the significance of neurophysiological approaches to the explanation of learning was first recognized by those who applied perceptual -motor theories to promote motor skill learning and to remediate learning disorders in children (e.g, Barsch, 1967, 1968; Cratty, 1971, 1974, 1975; de Quiros & Schrager, 1978; Kephart, 1960; Le Winn, 1969). Until recently, however, educational psychologists have resisted defining learning according to neurophysiological principles (Goodman, 1979; Hilgard & Bower, 1966). Between the appearance of Pribram's "Neurophysiological Notes on the Art of Education" (1964) and the appearance of the volume Education and the Brain (Chall & flirskey, 1978), the educational literature contained few references to the implications of neurophysiological research for language learning and teaching. However, the concepts of cerebral asymmetry derived from the works of Kimura (1973), Sperry (1964, 1974), and Geschwind (1965, 1970) were referred to by writers in education. For example, implications of splitbrain research for learning strategies and educational practices were addressed by Lutz (1978), McCallum and Glynn (1979), and Rennel s( 1976) . Recent publications of Teyler (1978) and Wittrock (1980) address the relationship between theoretical studies of human behaviors in educational psychology and theoretical studies of the structure and function of the brain in the neurosciences . The accumulation of information by researchers (e.g., Bechtereva, 1978; Luria, 1973; McGuiness & Pribram, 1980; and Trevathen, 1980) who are interested in the

PAGE 40

29 neurophysiological correlates of mental activity is beginning to affect psychological views of cognition, memory, information processing, and learning styles. Educators are realizing that "the brain does not usually learn in the sense of accepting or recording information from teachers . . . [that] it actively constructs its own interpretation of information and draws inferences from it" (Wittrock, 1978, p. 101). After comparing recent findings about the brain with recent findings about the cognitive processes of learning, Wittrock (1980) concluded that in the future, it will be productive "to study how learners transform the environmental events of teaching and instruction into functional information" (p. 398). A review of literature related to "how learners transform the environmental events of teaching" neurophysiological ly in liberal arts and sciences programs at the college or university level revealed no directly related empirical research. The literature contains studies of kinesic and nonverbal behaviors of both college and K-12 teachers. For example, Esp (1978) designed a study to guage the effects of college teacher nonverbal behavior during a period of instruction. Analysis of positive and negative cues exhibited by male and female teachers while presenting a lecture on the metric system indicated that only the nonverbal behavior of female teachers influenced student learning; however, the nonverbal behavior of both sexes had an effect on student attitude. Although not conducted with college instructors, nor for the same purpose as the present study, Credell's descriptive study (1977) of nonverbal behaviors in teaching exemplifies the growing concern for

PAGE 41

30 communication in educational programs. The purpose of the study was to determine if a group of teachers who had been subjectively labeled as likeable for displaying appropriate communication behaviors in their role as teachers differed in their evaluations of such behaviors from a group of teachers who had been labeled as being disliked. Credell videotaped 21 teachers to gather data for 10 diffusely defined nonverbal communication behaviors: physical appearance, facial expressions, use of time, gestures, voice tone, proxemics, body expression, expression of eyes, actions, and expressions of mouth. Credell reported indications of differences between the positive responses of the two groups as they reviewed videotapes: the likeable teachers noticed more positive nonverbal behaviors than were observed by the less liked teachers. One study that approximates more closely the design of the present study was conducted by Wyckoff (1973). Teacher mobility, gesturing, and pausing while presenting a lecture were examined for their effect on student recall of factual information contained in the lecture. Data collected from videotapes of 12 teachers and from test scores of 48 students indicated that increases in the frequency of stimulus variation on the part of secondary school teachers resulted in improved student performances. Wyckoff noted the need for "research directed toward the observation and measurement of teacher-student behavior related to input or stimulus variables" and concluded that variables such as movement, gesture, voice variation, eye contact, and enthusiasm do affect student attitude toward instruction (p. 85). Teaching effectiveness in explaining was explored in an early large-scale study conducted by Gage, Belgard, Dell, Hiller, Rosenshine,

PAGE 42

and Unruh (1968). Rosenshine (1968) contributed to the study by investigating specific behaviors, the frequencies of which were expected to be related to teacher effectiveness in explaining lecture material in social studies classes. According to Rosenshine, "the variables were the stimuli received by the pupils, that is, the verbal and nonverbal behavior of the teachers while they lectured" (p. 36). One variable that discriminated most significantly between the high-scoring and low-scoring groups for the two different lecture topics was the frequency and amount of time the teacher gestured and moved; the teacher of high-scoring groups moved and gestured more. Rosenshine hypothesized that these gestures and movements may have the effect of arousing or focusing attention. However, verbal variables taken singly and in combinations which might have been classified as attenionarousing variables did not discriminate between high and low groups. . . . These verbal variables included phrases stating the importance of material or recalling material, cognitive reversal, and references to problems and conflicts. (Rosenshine, 1968, p. 39-40) In a later and more comprehensive review of correlational and experimental studies of teacher enthusiasm, Rosenshine (1970) found evidence that ratings given teachers on such high-inference behaviors as "stimulating," "energetic," "mobile," "enthusiastic," and "animated related to measures of pupil achievement. The results of lowinference studies suggested that the frequencies of such variables as movement, gesture, variation of voice, and eye contact are related to pupil achievement, (p. 510) Empirical research from education that is relevant to the investigation of neurophysiological implications of teaching behaviors may be found in special education and physical education where perceptual motor theories have been advanced and tested for learning and teaching effectiveness. Such research is collected in journals such as

PAGE 43

32 Perceptual and Motor Skills and the Journal of Motor Behavior , as well as in volumes of studies such as those edited by Stelmach (1976, 1978, 1980). Although the scope of the present study does not include a review of such research, studies supporting the neurological concepts of perceptual -motor behavior developed by Williams (1969) and the closed loop theory of motor learning developed by Adams (1976, 1978) are indicative of the findings of Smith (1978) who concluded that "the potential for sensorimotor integration during motor programming is tremendous. ... We must look at the capacities of the sensorimotor system and seek to define the optimal conditions for motor control" (p. 180). The unifying concept in educational studies designed to determine how students learn from what Smith calls the "sensorimotor system" and what Wittrock calls the "environmental events of teaching" may be the concept of the need organisms have for interaction in order to adapt and learn. Supportive Studies from Occupational Therapy The assumption of organism-environment interaction is central to theories of sensory integration developed by Ayres in special education and occupational therapy. Ayres (1972b) stated. Central to the concept of brain development and function is the action of the environment upon the organism and the reaction of that organism upon the environment. This interaction constitutes the essence of a sensory integrative and sensorimotor response. The primary function of the brain is to translate sensory impulses into meaningful information and to organize an appropriate motor response. The process is most direct and obvious in the young child. The sensory and integrative aspect of the process, being more subtle and covert than the motor, is often overlooked, (p. 22) The published studies of Ayres (e.g., 1964, 1977, 1981) provided occupational therapists with concepts from the neurosciences to complement

PAGE 44

33 those derived from the biological and behavioral sciences. These concepts underly the premise that therapist-directed sensory-motor stimulation has the potential to mitigate learning disabilities and to facilitate language acquisition (Clark & Steingold, 1982). Empirical studies found in the occupational therapy literature may be relevant to the present investigation of neurophysiological implications of teaching behaviors used to facilitate language acquisition. By studying the effects of sensory integration procedures on children with auditory-language disorders, Ayres (1972a) and Ayres and Mailloux (1981) found evidence to generate hypotheses that relate increases in sensoryintegrative stimulation to increases in language comprehension and expression. The results of a study by Magrun and associates (1981) showed an increase in spontaneous verbal language in developmental ly delayed children immediately after stimulation of the sensory system that provides human beings with a sense of balance and movement in space. This sensory system, the vestibular system, provides human beings with important information as to whether they are the ones moving or whether the external environment is moving. The investigators suggested that vestibular stimulation may be an effective nonverbal intervention method for facilitating spontaneous language. Stilwell, Crowe, and McCallum (1978) explored the relationship between functioning of the vestibular system and specific communication disorders. Their findings supported the conclusion: It seems logical to assume that the cortex cannot be considered as the sole processing center for speech and language. Rather, it appears that the development of the language centers is in some way dependent on previous, as well as ongoing, subcortical sensory integration. This can be viewed as a pattern of continuous reciprocal interaction between the language centers and

PAGE 45

34 the more primary subcortical structures, thereby making these language centers neither fully dependent nor fully independent of subcortical neural processing, (pp. 226-227) Particularly related to the present study of college students is the finding by Angelo (1980) that the reading skills of low-achieving college students responded to sensory-integrative treatment. In discussion, Angelo stated that the results of the study undertaken to assess the use of a sensory-integrative approach to improve the academic performance of low-achieving college students "indicated that reading was significantly improved after students participated in a program of sensory-integrative activities" (p. 674). Summary of Supportive Studies Assumptions of neurophysiological need for interaction found in the review of related studies served to unify concepts of human communication, human development, and language learning. Theoretical and empirical evidence found in related fields supported the assumption that languages may be learned easily through the natural capacity human beings have to learn through their senses. Ample theoretical evidence showed that learning to use a foreign language for effective communication may be dependent on sensory-integrative brain processes that are, in turn, dependent on organism-environment interaction. Such evidence supports the theoretical framework of the present study. Theoretical Framework of the Study According to Rassias (1972), "the teacher is the 'reality' of learning: his own commitment to communication must be experienced by his students" (p. 12). In the present study, the "reality" of the

PAGE 46

35 teacher was conceptualized as the energy the teacher directs toward students in the form of communicative body movement and vocal language behaviors that neurophysiologically correspond to the foreign language communication skills of the teacher. Such teaching behaviors are theorized to stimulate students to respond adaptively through their own communicative body movement and vocal language behaviors that neurophysiologically correspond to foreign language communication skills in students. These concepts are fundamental to the theory that learning to communicate in a foreign language can be facilitated by teacher-directed sensory-motor stimulation. The investigator attempted to test this theory through a quasiexperimental ly designed study to determine the effects of teacherdirected sensory-motor stimulation on student foreign language communication effectiveness. The independent or treatment variable was sensorymotor stimulation as defined by the energy that teachers direct toward students through kinesic and vocal language behaviors. The dependent variable was student foreign language communication effectiveness as defined by test scores. The loci for the study were (a) the intensive and (b) the traditional beginning German classrooms at the University of Florida, Gainesville. Assumptions of the Study The assumptions of this study were based on the theory of teacherdirected sensory-motor stimulation developed from the review of related studies and from direct observation of teaching behaviors in beginning German classrooms. These assumptions are as follows:

PAGE 47

36 1. Instructors who use the intensive language model provide more kinesic and vocal language behaviors than do those who use traditional methods of instruction. These kinesic and vocal language teaching behaviors are measures of teacher-directed sensory-motor stimulation. 2. Students who are provided teacher-directed sensory-motor stimulation respond adaptively with behaviors that correspond to communication effectiveness. Therefore, students who are provided a higher level of teacher-directed sensory-motor stimulation achieve a higher score on a test of foreign language communication effectiveness than do students provided a lower level. a. Based on the second assumption, students whose instructors use traditional teaching methods and textbooks emphasizing English translation and grammatical analysis of the structure of a language achieve a higher score on a grammatical structure-vocabulary test than do students whose instructors use less English translation and grammatical analysis in the intensive language program. b. Based on the second assumption, students whose instructors use more foreign language communicative body movements and vocalization in the intensive language program actii eve a higher score on a listening comprehension test than do students whose instructors use fewer such movements and vocalizations in the traditional language program. c. Based on the second assumption, students who are provided a higher level of teacher-directed sensory-motor stimulation in the intensive language program achieve a higher speaking proficiency score on an oral interview test than do students provided a lower level in the traditional language program.

PAGE 48

37 3. Students who are provided a higher level of teacher-directed sensory-motor stimulation that is appropriate for the learning of the four basic language skills feel more confident of their ability to use these skills in foreign language communication than do students provided a lower level of such stimulation. 4. Students who are provided a higher level of teacher-directed sensory-motor stimulation that is appropriate for the learning of effective foreign language communication skills express more satisfaction with the efficacy of their foreign language courses to increase these skills than do students provided a lower level of such stimulation. Hypotheses The assumptions to be tested in this study were formulated as statistical hypotheses stated in the null form as follows: 1. The total mean seconds per contact hour of teacher-directed sensory-motor stimulation provided by teachers who use the intensive language model does not differ from the total mean provided by traditional language teachers. a. The mean seconds per contact hour of gross body movement directed toward students by teachers who use the intensive language model does not differ from the mean seconds directed toward students by traditional language teachers. b. The mean seconds per contact hour of hand gestures directed toward students by teachers who use the intensive language model does not differ from the mean seconds directed toward students by traditional language teachers.

PAGE 49

38 c. The mean seconds per contact hour of spoken German directed toward students by teachers who use the intensive language model does not differ from the mean seconds directed toward students by traditional language teachers. d. The mean seconds per contact hour of spoken English directed toward students by teachers who use the intensive language model does not differ from the mean seconds directed toward students by traditional language teachers. 2. The mean foreign language communication total test score of students who are provided a higher level of teacher-directed sensory stimulation does not differ significantly from that of students provided a lower level. a. The mean grammatical structurevocabulary test score of students who are provided a higher level of teacher-directed sensorymotor stimulation does not differ significantly from that of students provided a lower level. b. The mean listening comprehension test score of students who are provided a higher level of teacher-directed sensory-motor stimulation does not differ significantly from that of students provided a lower level. c. The mean speaking proficiency test score of students who are provided a higher level of teacher-directed sensory-motor stimulation does not differ significantly from that of students provided a lower level . 3. The self-assessed language skill performance ratings of students who are provided a higher level of teacher-directed sensory-motor

PAGE 50

39 stimulation do not differ significantly from those of students provided a lower level . a. The self-assessed reading comprehension ratings of students who are provided a higher level of teacher-directed sensorymotor stimulation do not differ significantly from those of students provided a lower level. b. The self-assessed listening comprehension ratings of students who are provided a higher level of teacher-directed sensorymotor stimulation do not differ significantly from those of students provided a lower level . c. The self-assessed writing proficiency ratings of students who are provided a higher level of teacher-directed sensory-motor stimulation do not differ significantly from those of students provided a lower level . d. The self-assessed speaking proficiency ratings of students who are provided a higher level of teacher-directed sensory-motor stimulation do not differ significantly from those of students provided a lower level. 4. The course satisfaction ratings of students who are provided a higher level of teacher-directed sensory-motor stimulation do not differ significantly from those of students who are provided a lower level. Definition of Terms Teacher-Directed Sensory-Motor Stimulation Conceptually, this term refers to the teachers' kinesic and vocal language behaviors that provide visual, auditory, and tactile sensory input to stimulate adaptive responses in students. In an environment of

PAGE 51

40 dynamic teacher/student interaction, students' adaptive responses further stimulate sensory-integrative feedback processes that are neurophysiologically equivalent to language teaching for teachers, and to language learning for students. These concepts derive from theories of sensory integration by Ayres (1972b). Operationally, the term refers to the teachers' kinesic and vocal language behaviors differentiated into (a) gross body movements of the head, trunk, and appendages directed toward students while eye contact is given to at least one student; (b) hand gestures in the form of finger snaps, pointing movements, acknowledging hand movements or signs, and other gestures made with the hands while eye contact is given to at least one student; (c) use of spoken German; and (d) omission of spoken English Spoken English is considered to be contraindicated when teaching a foreign language to native English-speaking students. This assumption is supported in the literature by Warriner (1979) who listed the first prerequisite for effective language teaching as "the target language is used almost exclusively by both teacher and student" (p. 51). According to the theory of sensory integration, any information gained by the senses from the English spoken by the teacher constitutes sensory input that stimulates sensory-motor responses for the acquisition of English; and constitutes sensory input that inhibits sensory-motor responses for the learning of German. Thus, the term teacher-directed sensory-motor stimulation is operationally defined as the combination of three kinesic and vocal language behaviors appropriate for foreign language communication effectiveness: (a) gross body movement directed toward students, (b) hand gestures directed toward students, and (c) spoken German.

PAGE 52

41 Student Foreign Language Communication Effectiveness Conceptually, this term refers to the ability to use a foreign language skillfully in the receptive (reading and listening) and expressive (writing and speaking) modes of verbal communication. Early in the twentieth century, Jesperson (1904) argued that to teach a language is to teach a student to communicate effectively in real-life situations. More recently, this premise has been confirmed in the literature (e.g., Bacheller, 1980; Lapan, 1980; Rassias, 1970, 1972). There is, however, still no consensus on how to test for foreign language communication effectiveness (Oiler & Perkins, 1980; Valette, 1977; Woodford, 1980). Jones (1977) argued that "the most neglected skill in language testing is still that of speaking" (p. 243). Operationally, the term student foreign language communication effectiveness refers to students' achievements on three component test forms of the Zertifikat Deutsch als Fremdsprache: (a) the grammatical structure-vocabulary component, (b) the listening comprehension component, and (c) the oral interview or speaking proficiency component. Dynamic Teacher/Student Interaction Conceptually, this term is used as a substitute for the many concepts of humanistic teaching described in the literature and applied in the field of foreign language education under such rubrics as transactional analysis (Stevick, 1974b) and caring and sharing (Moskowitz, 1978; Stevick, 1974a). As used in this study, the term is antonymous to terms used to describe passive teaching methods, such as the "expository mode" described by Bruner (1962, p. 83).

PAGE 53

42 Operationally, Rassias expects that teachers who "adopt" his humanistically based communicative techniques will "adapt" them according to the dynamics of their ever-changing environment ( Effective Language Teaching , 1972). Thus, the operational definition of the term dynamic teacher/student interaction implies an optimum learning environment of active participation, such as created by teachers who adapt behaviors prescribed in the Dartmouth Intensive Language Model. Model The term model includes references to the Dartmouth Intensive Language Model and to teaching behaviors prescribed by Rassias. Intensive Language Program Conceptually, this term includes all aspects of the experimental beginning German courses that implement the model in the Department of Germanic and Slavic Languages at the University of Florida. Operationally, the term refers to an experimental beginning German program structured as two courses, each of which continues for one 15week semester. Each course requires students to attend four classes and four tutor groups each week for five semester hours of credit. Thus, the term intensive language program implies 120 teacher contact hours and 120 tutor contact hours to complete 10 semester credits and meet foreign language requirements of the College of Liberal Arts and Sciences. Intensive language instructors are faculty members and graduate teaching assistants who use communicative techniques prescribed in the model to teach the four basic language skills. Tutors are students who complete one semester or more of German and meet selection criteria of good

PAGE 54

43 pronunciation and skill in using the basic methodology prescribed by Rassias. Tutors assist teachers and conduct homework sessions for tutor groups composed of 6-12 students. The teaching behaviors of tutors are the same as the teaching behaviors of instructors and are included in the operational definition of the term, teacher-directed sensory-motor stimulation . Traditional Language Program Conceptually, this term includes all aspects of the established beginning German courses as traditionally taught in the Department of Germanic and Slavic Languages at the University of Florida. Operationally, the term refers to an established beginning German program structured as three 15-week semesters. In the first semester, students attend classes 4 hours each week for 60 teacher contact hours to complete 4 semester hours of credit. In the second and third semesters, students attend classes 3 hours each week for 45 contact hours and 3 hours of credit each semester. Thus, the term traditional language program implies 150 teacher contact hours to complete 10 semester hours of credit to meet foreign language requirements of the College of Liberal Arts and Sciences. Traditional language teachers are faculty members and graduate teaching assistants who use idiosyncratic methods and standard textbooks to teach the four basic language skills. Communication The term communication denotes the system used by all forms of life to share experiences. The essence of communication is said to "lie in the evolution of synergistic interplay between participants

PAGE 55

44 committed to maximizing the efficiency of interchange" (Petrovich & Hess, 1972, p. 19). Within this definition there are neurophysiological implications of human communication's dependence on body movement. According to Key (1977), human communication is body movement. Movement of the vocal apparatus results in speech, the verbal act, or paralanguage, the nonverbal act. Movement of the muscular and skeletal apparatus results in kinesic communication, another kind of nonverbal act. (p. 5) Kinesics Birdwhistell (1952, p. 3) first used the term kinesics to define body movement as related to the nonverbal aspects of interpersonal communication. In the present study, the terms, kinesic, body language, nonverbal communication, and communicative body movement, are used interchangeably. Language Conceptually, language is considered to be a behavioral "tool" that is used by human beings for the purpose of communication (Bruner, 1978, p. viii). Operationally, the term language includes foreign languages. Foreign Language The term foreign language refers to a language other than a native language. A recent trend in education is to use the term second language to include foreign language and bilingual learning and teaching issues (Richards, 1978). Limitations This study was limited to a quasi-experimental design because of the necessity of working within existing classrooms when attempting to

PAGE 56

45 determine the effects of teacher-directed sensory-motor stimulation on foreign language communication effectiveness. Student selection and assignment to treatment groups was nonrandom; therefore, full experimental control of initial group differences was not feasible. In order to reduce this limitation to internal validity, covariance analysis was used to control statistically for initial group nonequivalences (Campbell & Stanley, 1963). Although the use of naturally formed groups might have inferred that the findings in this study could be generalized to similar classroom teaching situations, the self-selection of students into these groups presented a threat to such external validity. The self-selection of students into (a) intensive and (b) traditional language programs presented a threat to both the internal and external validity of results hypothesized to show no differences between the groups. No effort was made to remove these threats by seeking to identify preexisting differences in the two groups based on personality configurations, learning styles, and other related characteristics. An effort was made to reduce the severity of this limitation by seeking to identify differences between the groups according to previous educational and practical experiences with German. These differences were used as potential covariates to control statistically for initial group inequival encies that might have been caused by the use of a nonrandomized sample of student subjects. The inability to measure all possible student behaviors that may indicate student foreign language communication effectiveness presented a further limitation. Listening comprehension, speaking proficiency, and grammatical structure-vocabulary were directly measured

PAGE 57

46 by performance on tests. Specific reading, writing, nonverbal, and paral inguistic skills were not measured because of time and endurance constraints. Another limitation was imposed by the inability to measure all possible nonverbal and verbal language teaching behaviors that might provide sensory-motor stimulation for students in an environment of dynamic teacher/student interaction. Facial expressions, voice tones, and a variety of other nonverbal paral inguistic acts were not measured because of constraints of time and measurement instruments. Prior to the present study, the apparent effectiveness of teaching behaviors used in the experimental intensive German program had begun to influence the behaviors of teachers in the traditional program (Johnston, 1980). Although this influence might have combined with a Hawthorne effect created by the presence of video equipment to record observations of teacher behaviors, such a limitation was not expected to affect the outcome of the study. Plan of the Dissertation The statement of the problem and the rationale for the study were presented in the Introduction. Chapter II contains a review of studies related to the state of research in foreign language education and to the concepts and assumptions underlying the theoretical framework of this study. The assumptions, hypotheses, definition of terms, limitations, and plan of the dissertation are also presented in Chapter II. Chapter III contains a description of the procedures used to determine the effects of teacher-directed sensory-motor stimulation on student foreign language communication effectiveness. Chapter IV contains

PAGE 58

47 results of data analyses and findings from the study. Conclusions and recommendations for further research are presented in Chapter V.

PAGE 59

CHAPTER III PROCEDURE The present investigation was designed to determine the effects of teacher-directed sensory-motor stimulation on student foreign language communication effectiveness. The subjects were 6 instructors, 3 tutors, and 92 students enrolled in the final semester of beginning German courses at the University of Florida. The assumptions to be tested in this study were formulated as statistical hypotheses in Chapter II. The study was conducted in two phases. The preliminary phase involved procedures to collect descriptive data to document the actual duration of sensory-motor stimulation directed by two groups of language teachers toward their students. The second phase involved procedures to collect inferential statistical data to (a) document the level of teacher-directed sensory-motor stimulation provided for students in beginning German programs, (b) quantify students' previous educational and practical experiences with German for use as covariate data, (c) quantify students' affective responses to their own ability to communicate in German and to their satisfaction with the efficacy of their courses to promote such abilities, and (d) measure student foreign language coimiunication effectiveness from test scores. This chapter contains descriptions of the subjects, instruments, methods, and analyses used in the two phases of investigation. 48

PAGE 60

49 The Sample Phase One: Subjects for Collection of Descriptive Data The subjects for this phase of the study were three instructors in the intensive language program and three in the traditional language program. Two of the intensive language teachers and two of those in the traditional language program were used because they were the only instructors who taught the final semester of beginning German during 19811982. The one additional intensive language teacher and the one additional traditional teacher were randomly selected from a pool of four teachers, two of whom taught first semester intensive beginning German courses and two who taught first semester traditional beginning German courses during 1981-1982. Phase Two: Subjects for Collection of Inferential Statistical Data In addition to the 6 teachers who served as subjects in the first phase of the study, 3 tutors and 92 students were subjects for the second phase. Two of the tutors were randomly selected from a pool of 7 who conducted tutor groups for students enrolled in the final semester of beginning German in the intensive language program. The other tutor was randomly selected from a pool of 6 who conducted first semester tutor groups. The total number of student subjects originally included 55 intensive language students and 37 traditional language students who were enrolled in 4 sections of the final semester of beginning German, and who were the population who attended during the 14th week of the spring 1981-1982 semester. Bonus points were awarded at the discretion of the

PAGE 61

50 instructors to all students who participated in the study and completed all instrumental procedures to measure communication effectiveness and affective responses. Although no student refused to participate, 4 students in the intensive language program and 5 students in the traditional language program were absent and did not complete all instrumental procedures. Incomplete data obtained from these students were not used in the statistical analyses, nor were extreme data obtained from one student in the intensive language program. This student reported extensive previous experience with the German language that was reflected in perfect test scores on all three components of the test used to measure student foreign language communication effectiveness. The data used in the statistical analyses, therefore, were obtained from 50 intensive language students and 32 traditional language students. Instrumentation and Collection of Data Phase One: Instruments for Collection of Descriptive Data The instruments used in the first phase of the study included the following standard videotape equipment: a Sony 4200 camera, a Sony 3600 videotape recorder, a 9-inch Sony receiver monitor, 12 one-hour Scotch reel-to-reel videotapes, and an Odetics video timer. Video Behavioral Observation Charts were designed by the investigator to record the durations in seconds that teachers were observed directing four behaviors toward students. A sample page from a Video Behavioral Observation Chart for one teacher is included as Appendix A.

PAGE 62

51 Phase One: Collection of Descriptive Data In order to collect data to document the actual duration of sensory-motor stimulation directed by teachers toward students, teachers were videotaped. Each of 6 teachers selected as subjects was videotaped for a full 50-minute class period twice during one semester. These videotapes were assumed to record a representative sample of kinesic and vocal language teaching behaviors per contact hour. A video timer signaling lOth-of-second, second, and minute durations was transposed onto each tape as a visual digital clock. From each videotape, four specific teaching behaviors were charted on Video Behavioral Observation Charts. The four behaviors, (1) gross body movement, (2) hand gestures, (3) spoken German, and (3) spoken English were charted for 2,700 seconds (45 minutes) starting with the first words spoken by the instructor to the students. The investigator observed the entire videotape for each teacher and charted 1 -second durations for each behavior. Two independent observers charted one identical 5-minute (300 second) randomly selected segment of four of the videotapes. The seconds counted by the two independent observers were correlated with each other and with the investigator's same 5-minute segment. Interobserver reliability was computed as the rate of agreement (the number of seconds in agreement divided by the total of agreements plus the total of disagreements). At least four interobserver reliability checks with at least 85% agreement for each behavior were made to ensure consistency of observation. According to Huck, Cormier, and Bounds (1974), such interobserver procedures avoid instrumental threats to applied

PAGE 63

52 behavioral analysis designs (pp. 334-335). The interobserver Reliability Charts containing data for two randomly selected intensive language teachers and for two randomly selected traditional language teachers are included as Appendix B. Phase One: Computations of Descriptive Data The number of seconds counted for each teaching behavior during a 45-minute period was recorded and totaled on Video Behavioral Observation Charts. The totals on the six charts for the three intensive language teachers were averaged to compute the mean seconds per contact hour for each behavior. The same procedure was followed to compute the mean seconds per contact hour for each behavior by traditional language teachers. In order to compute the total mean seconds per contact hour of teacher-directed sensory-motor stimulation, the behaviors were made consistent with the operational definition of the term. Thus, this variable was calculated as the total mean seconds of three behaviors: gross body movement directed toward students, hand gestures directed toward students, and spoken German. Phase Two: Instruments for Collection of Inferential Statistical Data The instruments used in the second phase of the study included (a) standard videotape equipment and Video Behavioral Charts, (b) Student Foreign Language Data Sheet and Rating Forms, and (c) forms of three components of the Zertifikat Deutsch als Fremdsprache battery of tests.

PAGE 64

53 Videotape equipment and Video Behavioral Observation Charts . These instruments were the same as those described in the first phase, except that these were used to record teaching behaviors of the three tutor subjects. A sample page from a Video Behavioral Observation Chart for one tutor is included as Appendix C. Student Foreign Language Data Sheet and Rating Form . This instrument was designed by the investigator as a 5-item questionnaire to collect data for use as (1) covariates, (2) students' self -assessments of their ability to communicate in German, and (3) students' selfassessments of their satisfaction with courses to provide them with such communication effectiveness. The first three items were constructed to collect covariate data from students' previous educational and practical experiences with German. The construction of these three items was accomplished through two field tests, using variations of these items with intensive and traditional language students in final semester beginning German classrooms during the fall semester 1981-1982. The responses of these 19 intensive language students and 28 traditional language students were used to improve the reliability of items employed to gather the information needed as covariate data in the present study. The fourth and fifth items on the Student Foreign Language Data Sheet and Rating Form were designed as rating scales. The Performance Rating Scale and the Satisfaction Rating Scale were constructed according to guidelines for the construction and appropriate use of "numerical, graphic" scales as set forth by Dohner (1974, p. 21). The Performance Rating Scale was constructed with 5 points arranged along a

PAGE 65

54 continuum from excellent (4) to poor (0). The Satisfaction Rating Scale was constructed with 5 points arranged along a continuum from high (4) to low (0). The scales were pilot tested with 19 intensive language students and 28 traditional language students in final semester beginning German courses during fall 1981-1982. Because the perceptions of the student may change over time, the Performance and Satisfaction Rating Scales designed for this study were not considered to be any more or less reliable than other rating scales for these data. A sample copy of the Student Foreign Language Data Sheet and Rating Form is included as Appendix D. Zertifikat Deutsch als Fremdsprache (DaF) . Three components of this battery of tests were used to collect data from test scores as measures of student foreign language communication effectiveness. The problem of finding a reliable and valid measure of foreign language communication effectiveness is recognized in the literature (e.g., Brod, 1980; Oiler & Perkins, 1980; Valette, 1977; Woodford, 1980). From a review of such literature, the consensus appeared to be that, except for the oral proficiency interview used by the Foreign Service institute, Americans have no commonly recognized yardstick to measure student ability to communicate in a foreign language. Valette's reference (1980) to the availability and appropriateness of the "Zertifikat" was the basis for selecting the DaF for this study. This decision was reached after consultation with intensive and traditional instructors in the Department of Germanic and Slavic Languages and after securing permission from the authorities at the Goethe Institute, Atlanta, Georgia. The rationale for including the grammar structure-vocabulary test form of the DaF was considered to be consistent with the inclusion of

PAGE 66

55 grammatical structure-vocabulary exercises in standard textbooks used by teachers in traditional classroom instruction. Students in the traditional group in this study were expected to feel more familiar with a paper and pencil grammar test than with an oral interview test of grammar usage. The inclusion of this component was, therefore, an attempt to allay content validity bias toward the more real -life listening and speaking components. Listening and speaking test forms of the DaF were included in the total test score because they were believed to be valid performance tests of communication effectiveness in real-life situations. The DaF also includes reading and writing tests that are designed to test these skills in real -life situations. These two test components were not, however, included in the tests administered in the present study because of constraints of time, cost, scoring, and attendant endurance. Moreover, the investigator assumed that elements of reading comprehension were inherent in the listening comprehension and grammar structurevocabulary test forms. These forms required the student to read directions and to choose answers printed on the forms in German. The reliability and face validity of the DaF as an achievement test for adults are discussed in the manual (Deutscher Volkschochschul-Verband & GoetheInsti tut , (1977). In addition, the reliability of the component tests was assured through correspondence with the Goethe Institutes in Atlanta and in Munich, Germany. According to these sources, the Zertifikat was the result of a cooperative effort between the Goethe Institute, the Association of Educational Institutions for Adults in the Federal Republic of Germany, and the

PAGE 67

56 Association of Educational Institutions for Adults of Switzerland to certify communication proficiency in basic German. The test batteries were prepared in Germany where item analysis was used to determine the adequacy of items to discriminate between the better and poorer students who took draft versions of the components in at least seven different schools. Phase Two: Collection of Inferential Statistical Data Computation of teacher-directed sensory-motor stimulation levels . In order to convert the total mean seconds per contact hour of teacherdirected sensory-motor stimulation to the levels that comprised the statistics for the independent variable, the total mean seconds were multiplied by contact hours per (a) the intensive and (b) the traditional language program. Contact hours per program were calculated by multiplying the number of class periods per week by the number of weeks per semester, then adding these contact hours per semester according to the number of semesters required to complete each program. In the case of the traditional language program, the calculations for the first semester were made by multiplying 4 hours per week by 15 weeks to obtain 60 contact hours for the first semester. Contact hours for the second and third semesters were each calculated as 3 hours per week multiplied by 15 weeks to obtain a product of 45 contact hours each. The addition of the calculated contact hours for the three semesters provided the sum of 150 contact hours for the traditional language program.

PAGE 68

57 In the case of the intensive language program, calculations for the two required semesters each involved 4 hours per week of classes and 4 hours per week of tutor groups for a total of 240 contact hours. The mean seconds per contact hour of kinesic and vocal language behaviors by tutors were computed in a similar manner to that used to compile data on teacher behaviors during the first phase. Three tutor subjects were each videotaped while conducting a tutor group once during the semester so as to collect a sample of behaviors by tutors. The mean seconds per contact hour of teacher-directed sensory-motor stimulation were computed and tabulated for use in computing the level of teacher-directed sensory-motor stimulation provided for students in the intensive language program. Collection and computation of covariate data . In order to collect quantifiable data on students' previous experiences with German, the investigator distributed the Student Foreign Language Data Sheet and Rating Form to the student subjects. The first three items on this questionnaire pertained to covariate data. Students reported on the first item previous beginning German courses taken at the University of Florida. This information was combined with data indicating current course status, and values of 1.5 and 1.0 weighted to intensive and traditional courses, respectively. The rationale for these weightings was based on the requirement of two intensive language courses and the requirement of three traditional language courses each for 10 semester hours. The resulting calculations for each student were designated as covariate 1 (COVl).

PAGE 69

58 Respondents reported on the second item previous German courses taken at schools other than the University of Florida. These data were quantified as school years. A school year was defined as two quarters or semesters at the postsecondary level or one SeptemberJune year at the K-12 level. Each student received a calculated score in terms of full, partial, or nil school years, designated ascovariate 2 (C0V2). Respondents reported on the third item previous hearing and speaking experiences with the German language outside of formal schooling. These data were quantified in terms of calendar months, calculated as quarter-years, and weighted differently according to where the student had the experience. One quarter-year of experience with hearing German within Germany was weighted by a factor of 2, whereas one quarter-year of experience with hearing German outside of Germany was weighted by a factor of 1. One quarter-year of experience with speaking German within Germany was weighted by a factor of 3, whereas one quarter-year of experience with speaking German outside of Germany was weighted by a factor of 2. These weightings were consistent with findings in the foreign language literature (e.g., Balke, 1980; Brod, 1980; Keilstrup, 1981; Thiel, 1980) that suggested the value of experiencing a language in its native context. Each student received a calculated quarteryear score that was designated as covariate 3 (C0V3). Collection of test score data . In order to collect data on student foreign language communication effectiveness, three components of the battery of tests comprising the Zertifikat Deutsch als Fremdsprache (DaF) were administered to all student subjects during the 14th week of classes at the end of the 1982 spring semester.

PAGE 70

59 The investigator administered the grammatical structure-vocabulary component during regularly scheduled class periods of the two intensive and the two traditional beginning German classes that comprised the experimental and control groups in this study. The listening comprehension component was administered in the language lab by the investigator and by the language lab director together during four class periods in order to accommodate all student subjects. The speaking component was administered during individually scheduled 5-minute periods throughout the week. Native-German speaking interviewers administered the speaking component at the same time as the investigator videotaped the interviews for scoring at a later date by an independent scorer. Four statistically reliable variations of this "Communication in Everyday Situations" part of the DaF were alternated during the administration. Scoring procedures were accomplished according to instructions in the DaF manual (Deutscher Vol kschochschul -Verband & Goethe-Institut, 1977). Scoring of the grammatical structure-vocabulary and the listening comprehension components was accomplished by using a National Computer System scanner. The speaking component was scored by the nativeGerman speaking teacher who taught German at a local high school. This teacher independently viewed the videotapes and rated each student subject according to instructions in the DaF manual. The 82 students who completed all three test components received (a) a grammar structure-vocabulary score designated as GRAM, (b) a listening comprehension score designated as HEAR, (c) a speaking proficiency score designated as SPEAK, and (d) a total score comprising the three

PAGE 71

60 component scores and designated as TOT. The component scores were each given equal weight (30 points) in computing a maximum total test score of 90 points. Collection of data from rating scales . The fourth and fifth items of the questionnaire distributed to student subjects were designed as rating scales. The fourth item was a Performance Rating Scale designed to collect information about the student's self-assessed ability to use the German language. Each student assigned values ranging from 4 (excellent) to 0 (poor) to each of the four language skills. The fifth item of the questionnaire was designed as a Satisfaction Rating Scale to collect information about students' satisfaction with intensive and traditional language courses at the University of Florida. Each student assigned a value ranging from 4 (high) to 0 (low) to current and previous courses. Analyses The analyses of data were completed in two phases. Data collected in the first phase were tabulated and described. Data collected in the second phase were subjected to inferential statistical analyses. Phase One: Descriptive Data Analyses The preliminary phase of this investigation involved descriptive data. The previously defined kinesic and vocal language behaviors were systematically observed, recorded, and counted in an attempt to document their character and duration in seconds per contact hour. Collected data were tabulated and used to test the first null hypothesis. These data were also used to establish the levels of teacher-directed

PAGE 72

61 sensory-motor stimulation that were the independent variable for the second phase of the investigation. Phase Two: Inferential Statistical Analyses The design for testing the second null hypothesis was a nonequivalent control group design, described by Campbell and Stanley (1963) as a quasi-experimental design, and iS) represented as Three potential pretest measures were identified as COVl , C0V2, and C0V3 to represent (a) previous beginning German courses taken at the University of Florida, (b) previous educational experience with German in schools other than the University of Florida, and (c) previous practical experience with German. The dependent variable was defined as the posttest measures of student foreign language communication effectiveness which were identified as GRAM, HEAR, SPEAK, and TOT representing test scores on three components of the DaF and the total test score for each student subject. The treatment or independent variable was defined as the level of teacher-directed sensory-motor stimulation measured as the total mean seconds of gross body movement, hand gestures, and spoken German directed toward students in 240 contact hours in the intensive language program and 150 contact hours in the traditional language program. The experimental group (Group INT) was comprised of 50 students who were provided a higher level of teacher-directed sensory-motor stimulation. The control group (Group TRAD) was comprised of 32 students who were provided a lower level of teacher-directed sensory-motor stimulation.

PAGE 73

62 Data collected from the two groups were subjected to an analysis of covariance (ANCOVA) to test for differences in foreign language communication effectiveness between the two groups. ANCOVA was used because of its ability to reduce error in the dependent variable by partial ing out the effects of differences between the groups according to the pretest covariate measures. Prior to using ANCOVA to test the second null hypothesis, the Pearson product-moment correlation procedure was used to test the significance of the correlations between the covariates and the dependent variable. The third and fourth hypotheses were designed to test the differences in affective responses between (a) students in the experimental and (b) students in the control group. In both these hypotheses, the variable, level of teacher-directed sensory-motor stimulation, was measured as the total mean seconds of gross body movement, hand gestures, and spoken German directed toward students in 240 contact hours in the intensive language program and 150 contact hours in the traditional language program. Self-assessed ratings on four language skills comprised the variables measured by values on the Performance Rating Scale. Course satisfaction ratings were the variables measured by the combined values of current and previous intensive traditional language courses taken at the University of Florida as marked on the Satisfaction Rating Scale. The data for these variables were analyzed as frequencies in contingency tables. Independent chi-square tests were used to test the significance of differences in ratings between the (a) experimental and (b) control groups.

PAGE 74

63 The Statistical Analysis System (SAS Institute Incorporated, 1979) was used in conjunction with computer facilities at the Northeast Regional Data Center to analyze data collected in this second phase of the study. The level of significance selected to test the hypotheses was alpha < .05. Results of the descriptive and inferential statistical analyses performed in the two phases of this study are presented in Chapter IV.

PAGE 75

CHAPTER IV FINDINGS The specific purpose of this investigation of neurophysiological implications for language learning identified in teaching behaviors prescribed in the Dartmouth Intensive Language Model was to determine the effects of teacher-directed sensory-motor stimulation on foreign language communication effectiveness. Data resulting from two procedural phases of the study are presented, and findings are discussed in the present chapter. Results of Descriptive Data Analyses The purpose of the preliminary phase of the study was to document the actual number of seconds per contact hour that teachers directed sensory-motor stimulation to students in the intensive and the traditional beginning German programs at the University of Florida. Subjects for this phase of the study were three intensive and three traditional language teachers. Twelve videotapes of teacher behaviors were observed to record and count seconds per contact hour of four previously defined behaviors. Confidence in the reliability of these counts was strengthened through interobserver reliability procedures. Data from interobserver Reliability Charts are summarized in Table 1. The data in Table 1 indicate that the rate of agreement per behavior was high enough (above 85%) to justify the reliability of the investigator's counts for the six language teacher subjects. 64

PAGE 76

65 Table 1 Summary of Interobserver Reliability Data Rate of agreement per behavior Language teacher uruoc) uuujf movement (300 sec.) Gestures (300 sec.) Spoken German (300 sec.) Spoken English (300 sec.) inLensive ueacner n 92% 86% 100% Intensive teacher B 90 96 98 100 Traditional teacher A 96 98 97 97 Traditional teacher B 98 95 97 94 Data presented in Table 2 are the results of the investigator's observations and counts of seconds per contact hour of kinesic and vocal language behaviors of intensive and traditional language teachers. These data were used to test the subsections of the first null hypothesis. Null Hypothesis la The mean seconds per contact hour of gross body movement directed toward students by instructors who use the intensive language model does not differ from the mean seconds directed toward students by traditional language teachers. Inspection of Table 2 indicates that the magnitude of difference between the mean seconds per contact hour of gross body movement by intensive language teachers (1473.50) and the traditional language teachers (297.83; revised at 260.60) appears to be enough to reject Null Hypothesis la. Teachers who used the intensive language model directed more gross body movement toward students than did traditional language teachers.

PAGE 77

66 Table 2 Mean Seconds per Contact Hour of Kinesic and Vocal Language Behaviors by Intensive and Traditional Language Teachers Behavior Gross body Hand Spoken Spoken movement gestures German English Language teachers Day (seconds) (seconds) (seconds) (seconds) Intensive Teacher A 1 2 1026 1120 1753 1189 1493 1472 8 0 Teacher B 1 2 1831 1366 1729 1335 1648 2042 0 3 Teacher C 1 2 1 CO o 1 boo 1815 1 <:09 1159 1 cm 1 501 1315 3 0 Total 8841 8374 9471 14 Mean 1473. 50 1395.67 1578.50 2.33 iditional Teacher A 1 2 72 121 317 202 235 550 1779 2014 Teacher B 1 2a 413 484b 872, 1029° 314. 1832° 1556 158b Teacher C 1 2 260 437 514 447 422 514 923 514 Total 1787 3381 3867 6944 Mean 297.83 563.50 644.50 1157.33 Revised total 1303 2352 2035 6786 Revised mean 260.60 470.40 407.00 1357.20 Day 2 for Teacher B was atypical. Teacher B admitted to varying from usual teaching methods on day 2 by singing and lecturing dramatically in German for at least 40 minutes. bThese data were omitted in computations of the revised total and the revised mean.

PAGE 78

67 Null Hypothesis lb The mean seconds per contact hour of hand gestures directed toward students by teachers who use the intensive language model does not differ from the mean seconds directed toward students by traditional language teachers. Inspection of Table 2 indicates that the magnitude of difference between the mean seconds per contact hour of hand gestures by intensive language teachers (1395.67) and traditional language teachers (563.50; revised at 470.40) appears to be great enough to reject Null Hypothesis lb. Teachers who used the intensive language model directed more hand gestures toward students than did traditional language teachers. Null Hypothesis Ic The mean seconds per contact hour of spoken German directed toward students by teachers who use the intensive language model does not differ from the mean seconds directed toward students by teachers using traditional language instruction. Table 2 indicates that the magnitude of difference between the mean seconds per contact hour of spoken German by intensive language teachers (1578.50) and traditional language teachers (644.50; revised at 407.00) appears to be great enough to reject Null Hypothesis Ic. Teachers who used the intensive language model directed more spoken German toward students than did traditional language teachers. Null Hypothesis Id The mean seconds per contact hour of spoken English directed toward students by teachers who use the intensive language model does not differ from the mean seconds directed toward students by traditional language teachers.

PAGE 79

68 According to Table 2, the magnitude of difference between the mean seconds per contact hour of spoken English by intensive language teachers (2.33) and traditional language teachers (1157.33; revised at 1357.20) appears to be great enough to reject Null Hypothesis Id. Teachers who used the intensive language model directed less spoken English toward students than did traditional language teachers. Data presented in Table 3 represent the total mean seconds per contact hour of teacher-directed sensory-motor stimulation by intensive and traditional language teachers. These data were used to test the main section of the first null hypothesis. Table 3 Total Mean Seconds per Contact Hour of Teacher-Directed Sensory-Motor Stimulation by Intensive and Traditional Language Teachers Teacher-directed Language teachers sensory-motor rstimulation Intensive Traditional Gross body movement (mean seconds) 1473.50 260.60 Hand gestures (mean seconds) 1395.67 470.40 Spoken German (mean seconds) 1578.50 407.00 Total mean 4447.67 1138.00 Revised mean seconds per contact hour were used to calculate the total mean seconds per contact hour by traditional language teachers. The unrevised total mean calculated at 1505.83 was considered erroneous because of inclusion of extreme raw data reported in Table 2.

PAGE 80

69 Null Hypothesis 1 The total mean seconds per contact hour of teacher-directed sensory-motor stimulation provided by teachers who use the intensive language model does not differ from the total mean provided by traditional language teachers. Inspection of Table 3 indicates that the magnitude of difference between the total mean seconds per contact hour of teacher-directed sensory-motor stimulation provided by intensive language teachers (4447.67) and traditional language teachers (1138,00) is great enough to reject Null Hypothesis 1. Teachers who used the intensive language model provided more teacher-directed sensory-motor stimulation than did traditional language teachers. In addition to data collected for the six teachers in the intensive and traditional beginning German programs, data compiled from the Video Behavioral Observation Charts of three tutors were included in the computations to establish the level of teacher-directed sensory motor stimulation used as the independent variable in the second procedural phase of the study. Table 4 contains data on tutor behaviors. The total mean seconds per contact hour of teacher-directed sensory-motor stimulation by tutors is reported in Table 5. Table 6 shows the levels of teacher-directed sensory-motor stimulation provided for students in the intensive and traditional language programs. Inspection of Table 6 indicates that a higher level of teacherdirected sensory-motor stimulation is provided for students in the intensive language program than in the traditional language program. These higher and lower levels were used as the treatment or independent

PAGE 81

70 variable in the quasi-experimental ly designed test of the second null hypothesis. Table 4 Mean Seconds per Contact Hour of Kinesic and Vocal Language Behaviors by Tutors Behavior Gross body movement Tutors (seconds) Hand gestures (seconds) Spoken German (seconds) Spoken Engl ish (seconds) Tutor A 1505 1006 1113 0 Tutor B 862 1230 1234 4 Tutor C 1323 897 973 2 Total 3690.00 3133.00 3320.00 6 Mean 1230.00 1044.33 1106.67 2 Table 5 Total Mean Seconds of Teacher-Directed Sensory-Motor Stimulation by Tutors Teacher-directed sensory-motor stimulation Tutors Gross body movement (mean seconds) 1230.00 Hand gestures (mean seconds) 1044.33 Spoken German (mean seconds) 1106.67 Total 3381.00

PAGE 82

r— Hv» •M C a> -a 3 *> to so Ma ra •rS> CD o o ssQ. a. SZ (U o 01 •In3 I— c 3 ea E -I •r4J I— oo (O c so 0 •+-> •(-> O -r2: XI 1 ta o (/) -o £1 C > -a O) -r»-> (/) o c O) 01 S_ -M •rC Q 1-1 I sc: u ra c +-> o 3 o o sO I/) to s+J 3 C O O ^ O 1^ _ o o +J s00c O SO o I S(U ^3 O E 1/1 •!O C 4-> (0 0)01 O) 1/1 o o o o r>s o r>. o IT) O O 00 CO o "53O CO CO in o OJ to o fO 0) o o o o o o o o CSJ uo o o o o «^ 01 CO o o C30 CO CO i. o +-> 3 > cu

PAGE 83

72 Results of Inferential Statistical Analyses The second phase of the study involved the use of covariance analysis of test score data and chi-square analysis of rating scale data. Subjects for this phase were 50 students in the intensive language program and 32 students in the traditional language program. Analysis of Covariance (ANCOVA) Results Prior to testing the second null hypothesis with ANCOVA, the potential covariates (COVl, C0V2, C0V3) were correlated (alpha < .05) with the dependent variable or posttest measure, TOT. The results of Pearson product moment procedures showed C0V2 to be the most highly correlated, with a correlation coefficient of .28 significance at the alpha < .01 level. The other covariates were dropped because of lack of significant correlations at the alpha < .05 level. The second covariate (C0V2) was kept as a pretest measure of previous educational experience with German at schools other than the University of Florida to covary with the posttest measures. Tests of the homogeneity of slopes (alpha < .05) were also performed. No significant interaction was found between the covariate, C0V2, and total test scores, TOT, or between C0V2 and any of the component test scores (GRAM, HEAR, SPEAK). With the assumption of homogeneity of slopes met, ANCOVA was used to test the second main null hypothesis and the three subsections of the second null hypothesis. The 50 subjects in the intensive language program were the experimental group provided a higher level of teacher-directed sensory-motor

PAGE 84

73 stimulation (Group INT). The 32 subjects in the traditional language program were the control group provided a lower level of teacherdirected sensory-motor stimulation (Group TRAD). The results of the analyses (alpha < .05) of the test score data, adjusted for unequal Ns and adjusted according to the covariate to control for initial differences between the groups, are reported in Tables 7 through 14. Tables 7 and 8 reflect the covariance analysis results for the effect of level of teacher-directed sensory-motor stimulation on the mean foreign language communication total test score. Table 7 Summary of Covariance Analysis for Effect of Level of Teacher-Directed Sensory-Motor Stimulation on the Mean Foreign Language Communication Total Test Score Source df SS MS F Previous educational experience (C0V2) 1 2365.31 2365.31 18.07* Level of stimulation (Group) 1 6556.88 6556.88 50.09* Error 79 10341.73 130.91 *Significant at alpha < .0001. Table 8 presents the adjusted least squares means for foreign language communication total scores for students who are provided a higher level of teacher-directed sensory-motor stimulation and those provided a lower level of teacher-directed sensory-motor stimulation.

PAGE 85

74 Table 8 Adjusted Least Squares Means for Foreign Language Communication Total Test Scores Group TOT INT 60.80 TRAD 42.31 Null Hypothesis 2 . The mean foreign language communication test score of students who are provided a higher level of teacher-directed sensory-motor stimulation does not differ significantly from that of students provided a lower level. Inspection of Table 7 indicates that the F ratio of 50.09 was statistically significant (alpha < .05) for difference between the groups. Therefore, the second main null hypothesis is rejected. As indicated by the means listed in Table 8, students who were provided a higher level of teacher-directed sensory-motor stimulation achieved the higher mean foreign language communication total test score. Tables 9 and 10 contain the covariance analysis results for the effect of level of teacher-directed sensory-motor stimulation on the mean grammar structure-vocabulary test scores.

PAGE 86

75 Table 9 Summary of Covariance Analysis for Effect of Level of Teacher-Directed Sensory-Motor Stimulation on the Mean Grammar Structure-Vocabulary Test Score Source df SS MS F Previous educational experience (C0V2) 1 98.94 98.94 7.63* Level of stimulation (Group) 1 239.40 239.40 18.46** Error 79 1024.30 12.95 *Significant at alpha < .01. **Significant at alpha < .0001. Table 10 presents the adjusted least squares means for grammar structure-vocabulary test scores for students who are provided a higher level of teacher-directed sensory-motor stimulation and those provided a lower level of teacher-directed sensory-motor stimulation. Table 10 Adjusted Least Squares Means for Grammar Structure-Vocabulary Test Scores Group GRAM INT 14.76 TRAD 11.22

PAGE 87

76 Null Hypothesis 2a . The mean grammatical structure-vocabulary test score of students who are provided a higher level of teacherdirected sensory-motor stimulation does not differ significantly from that of students provided a lower level. Inspection of Table 9 indicates that the F ratio of 18.46 was statistically significant (alpha < .05) for differences between the groups. Therefore, the first subsection of the second hypothesis is rejected. According to means listed in Table 10, students who were provided a higher level of teacher-directed sensory-motor stimulation achieved the higher mean grammatical structure.-vocabulary test score. Tables 11 and 12 contain the covariance analysis results for the effect of level of teacher-directed sensory-motor stimulation on the mean student listening comprehension test score. Table 11 Summary of Covariance Analysis for Effect of Level of Teacher-Directed Sensory-Motor Stimulation on Mean Listening Comprehension Test Score Source df SS MS F Previous educational experience (C0V2) 1 232.76 232.76 8.89* Level of stimulation (Group) 1 557.51 557.51 21.30** Error 79 2067.39 26.17 *Significant at alpha < .01. *Significant at alpha < .0001.

PAGE 88

77 Table 12 contains the adjusted least squares means for listening comprehension test scores for students who are provided a higher level of teacher-directed sensory-motor stimulation and those provided a lower level of such stimulation. Table 12 Adjusted Least Squares Means for Listening Comprehension Test Scores by Groups Group HEAR INT 21.30 TRAD 15.91 Null Hypothesis 2b . The mean listening comprehension score of students who are provided a higher level of teacher-directed sensorymotor stimulation does not differ significantly from that of students provided a lower level. Inspection of Table 11 indicates that the F ratio of 21.30 was statistically significant (alpha < .05) for differences between groups. Therefore, the second subsection of the second hypothesis is rejected. The means listed in Table 12 show that students who were provided a higher level of teacher-directed sensory-motor stimulation achieved the higher mean listening comprehension test score. Tables 13 and 14 represent the covariance analysis results for the effect of level of teacher-directed sensory-motor stimulation on the mean student speaking proficiency test score.

PAGE 89

78 Table 13 Summary of Covariance Analysis for Effect of Level of Teacher-Directed Sensory-Motor Stimulation on the Mean Speaking Proficiency Test Score Source df SS MS Previous educational experience {C0V2) 1 511.58 511.58 13.24* Level of stimulation (Group) 1 1819.78 1819.78 47.09** Error 79 3053.09 38.65 *Signif icance at alpha < .01. **Significance at alpha < .0001 Table 14 reflects the adjusted least squares means for student speaking proficiency test scores for students who are provided a higher and a lower level of teacher-directed sensory-motor stimulation. Table 14 Adjusted Least Squares Means for Speaking Proficiency Test Scores Group SPEAK INT TRAD 24.93 15.19

PAGE 90

79 Null Hypothesis 2c . The mean speaking proficiency test score of students who are provided a higher level of teacher-directed sensorymotor stimulation does not differ significantly from that of students provided a lower level. Inspection of Table 13 indicates that the F ratio of 47.09 was statistically significant (alpha < .05) for differences between the groups. Therefore, the third subsection of the second null hypothesis is rejected. The means listed in Table 14 show that students who were provided a higher level of teacher-directed sensory-motor stimulation achieved the higher mean speaking proficiency test score. Independent Chi-Square Test Results In order to test for significant (alpha < .05) differences between groups in the third and fourth null hypotheses, data from rating scales were analyzed as frequencies in contingency tables. The 50 subjects in the intensive language program who were provided a higher level of teacher-directed sensory-motor stimulation were designated as Group INT. The 32 subjects in the traditional language program provided a lower level were designated as Group TRAD. These two groups constituted one variable in the 2x5 contingency tables used originally. The other variables were (a) self-assessed ability to use the German language as measured by . .5 points on the Performance Rating Scale and (b) course satisfaction as measured by 5 points on the Satisfaction Rating Scale. Both rating scales are items on the Student Foreign Language Data Sheet and Rating Form included as Appendix D. Collapsed Variables . Initial chi-square analyses were done with the original 5-point variables; however, in order to reduce the number

PAGE 91

80 of low frequency cells, subsequent analyses reported in Tables 15 through 19 were done with collapsed variables. Points on the Performance Rating Scales with values above 2 were collapsed to a value of 3 and points with values below 2 were collapsed to a value of 1 in the 2x3 contingency tables used to test the third hypothesis. The combined points of previous and current course satisfaction ratings in the 2x3 contingency tables used to test the fourth hypothesis were collapsed as follows: (a) combined points with values above 5 were collapsed to a value of 3; (b) combined points with values of 3, 4, or 5 were collapsed to a value of 2; (c) combined points with values below 3 were collapsed to a value of 1. Null Hypothesis 3a . The self-assessed reading comprehension ratings of students who are provided a higher level of teacher-directed sensory-motor stimulation do not differ significantly from those of students provided a lower level. Table 15 represents the variables in Null Hypothesis 3a arranged in a 2 X 3 contingency table. Group INT was provided a higher level of teacher-directed sensory-motor stimulation than was Group TRAD. Inspection of Table 15 indicates that the chi-square value of .345 was not statistically significant at alpha < .05. Therefore, Null Hypothesis 3a is not rejected on the basis of the frequency with which the two groups of students self-assessed their reading comprehension skills according to three nominal ratings. Null Hypothesis 3b . The self-assessed listening comprehension ratings of students who are provided higher level of teacher-directed sensory-motor stimulation do not differ significantly from those of students provided a lower level.

PAGE 92

81 Table 15 Self -Assessed Reading Comprehension Ratings by Student Groups Student group Self-assessed reading comprehension rating Poor 1 Fair 2 Excellent 3 INT Frequency Percent Row percent Column percent TRAD Frequency Percent Row percent Column percent Chi square = 0.345* 2 2.44 4.00 66.67 1 1.22 3.13 33.33 20 24.39 40.00 64.52 11 13.41 34.33 35.48 28 34.15 56.00 58.33 20 24.39 62.50 41.67 *Alpha < .8416. Table 16 represents the variables in Null Hypothesis 3b arranged in a 2 X 3 contingency table. Group INT was provided a higher level of teacher-directed sensory-motor stimulation than was Group TRAD. Inspection of Table 16 indicates that the chi-square value of 3.435 was not statistically significant at alpha < .05. Therefore Null Hypothesis 3b is not rejected on the basis of the frequencies with which

PAGE 93

82 Table 16 Self-Asssessed Listening Comprehension Ratings by Student Groups Student group Self-assessed listening comprehension rating Poor Fair Excellent 1 2 3 INT Frequency Percent Row percent Column percent 3 3.66 6.00 42.86 15 18.29 30.00 51.72 32 39.02 64.00 69.57 TRAD Frequency Percent Row percent Column percent 4 4.88 12.50 57.14 14 17.07 43.75 48.28 14 17.07 43.75 30.43 Chi square = 3.435* *Alpha < .1795.

PAGE 94

83 the two groups of students self-assessed their listening comprehension skills according to three nominal ratings. Null Hypothesis 3c . The self-assessed writing proficiency ratings of students who are provided a higher level of teacher-directed sensorymotor stimulation do not differ significantly from those of students provided a lower level. Table 17 represents the variables in Null Hypothesis 3c arranged in a 2 X 3 contingency table. Group INT was provided a higher level of teacher-directed sensory-motor stimulation than was Group TRAD. Inspection of Table 17 indicates that the chi-square value of 5.329 was not statistically significant at alpha < .05. Therefore, Null Hypothesis 3c is not rejected on the basis of the frequencies with which the two groups of students self-assessed their writing skills according to three nominal ratings. Null Hypothesis 3d . The self-assessed speaking proficiency ratings of students who are provided a higher level of teacher-directed sensorymotor stimulation do not differ significantly from those of students provided a lower level. Table 18 represents the variables in Null Hypothesis 3d arranged in a 2 X 3 contingency table. Group INT was provided a higher level of teacher-directed sensory-motor stimulation than was Group TRAD. Inspection of Table 18 indicates that the chi-square value of 9.254 was statistically significant at alpha < .05. Therefore, Null Hypothesis 3d is rejected. Students who were provided a higher level of teacher-directed sensory-motor stimulation self-assessed their speaking proficiency higher than did students provided a lower level.

PAGE 95

84 Table 17 Self-Assessed Writing Proficiency Ratings by Student Groups Student group Self -assessed writing proficiency rating Poor 1 Fair 2 Excellent 3 INT Frequency Percent Row percent Column percent 11 13.41 22.00 84.62 20 24.39 40.00 50.00 19 23.17 38.00 65.52 TRAD Frequency Percent Row percent Column percent 2 2.44 6.25 15.38 20 24.39 62.50 50.00 10 12.20 31.25 34.48 Chi square = 5.329" Alpha < .0696.

PAGE 96

85 Table 18 Self -Assessed Speaking Proficiency Ratings by Student Groups Student group Self-assessed speaking proficienty rating Poor 1 Fair 2 Excellent 3 INT Frequency Percent Row percent Column percent 8 9.76 16.00 40.00 13 15.85 26.00 52.00 29 35.37 58.00 78.38 TRAD Frequency Percent Row percent Column percent 12 14.63 37.50 60.00 12 14.63 37.50 48.00 8 9.76 25.00 21.62 Chi square = 9.254* *Alpha < .0098. Null Hypothesis 4 . The course satisfaction ratings of students who provided a higher level of teacher-directed sensory-motor stimulation do not differ significantly from those of students who are provided a lower level .

PAGE 97

86 Table 19 represents the variables in Null Hypothesis 4 arranged in a 2 X 3 contingency table. Group INT was provided a higher level of teacher-directed sensory-motor stimulation than was Group TRAD. Table 19 Foreign Language Course Satisfaction Ratings by Student Groups Student group Low 1 Course satisfaction rating Moderate 2 High 3 INT Frequency Percent Row percent Column percent 2 2.44 4.00 33.33 5 6.10 10.00 26.32 43 52.44 86.00 75.44 TRAD Frequency Percent Row percent Column percent 4 4.88 12.50 66.67 14 17.07 43.75 73.68 14 17.07 43.75 24.56 Chi square = 16.529^ *Alpha < .0003.

PAGE 98

87 Inspection of Table 19 indicates that the chi-square value of 16.529 was statistically significant at alpha < .05. Therefore, Null Hypothesis 4 is rejected. Students who were provided a higher level of teacher-directed sensory-motor stimulation rated their satisfaction with foreign language courses higher than did students provided a lower level Discussion of Descriptive Findings In the preliminary phase of this study, the investigator sought documentation for the character and duration of teacher-directed sensory motor stimulation provided for students in the intensive and traditional beginning German programs at the University of Florida. Twelve videotapes provided a means of charting and counting durations of four welldefined kinesic and vocal language behaviors that constituted the operational definition of teacher-directed sensory-motor stimulation. The subjects were six teachers conducting two 50-minute classes each. Rejection of Null Hypotheses la, lb, Ic, and Id The results of counting seconds per contact hour for each of four well-defined kinesic and vocal language behaviors showed that teachers who used the intensive language model directed (a) 5.62 times as many seconds of gross body movement, (b) 2.97 times as many seconds of hand gestures, and (c) 3.88 times as many seconds of spoken German toward students as did traditional language teachers during each contact hour. Traditional language teachers directed 582.49 times as many seconds of spoken English toward students as did intensive language teachers during each contact hour.

PAGE 99

88 These quantitative findings were consistent with qualitative findings described in the literature by Byrd (1980), Johnston (1980), and Rassias (1970) and with informal observations made by the investigator in beginning German classes during 1980-1982. Instructors who used the intensive language model spent high levels of energy in attempting to create optimum learning environments of dynamic teacher/student interaction. These teachers were constantly moving about the classroom, usually using a choreographed rhythm and pace of verbal and nonverbal cues prescribed by Rassias to reduce inhibitions to communication. The cues were in the form of chorals, finger-snaps, pointing gestures, acknowledging expressions, and large body movements on three axes of movement directed toward student— side to side, up and down, and back to front. The omission of spoken English prescribed by Rassias and noted during observations in intensive language classrooms was confirmed by the finding that only 14 seconds of English were spoken in the total of 16,200 seconds counted for teachers who used the intensive language program. This finding contrasted sharply with the finding that the traditional language teachers spoke English 50% of the counted time. Except in one case (traditional teacher B, day 2), the six teacher subjects behaved as they had been observed informally by the investigator, in a manner students agreed was usual behavior during the entire semester. On the second day of videotaping, traditional teacher B departed from normal teaching behavior and from the scheduled grammar lesson by substituting a dramatic lecture about East Germany, including the singing of its national anthem in German. This admittedly atypical behavior was verified by students who commented verbally and

PAGE 100

89 in writing on the Foreign Language Student Data Sheet and Rating Form to inform the investigator of an increase in the activity of the teacher when observed or videotaped. The extreme data from this teacher's Video Behavioral Observation Chart were deleted from the data used in this study. Had these data been included, however, the results would have been the same with respect to the rejection of the first null hypotheses. Rejection of Null Hypothesis 1 The rejection of the main section of the first null hypothesis indicated that instructors who use the intensive language model provide more teacher-directed sensory-motor stimulation than do traditional language teachers. This descriptive finding was consistent with quantitative findings from the rejection of the first subsectional null hypotheses. One of the assumptions of this study was that teachers direct sensory-motor stimulation toward students via their gross body movements, hand gestures, and vocalizations. It was further assumed in this study that these teaching behaviors have neurophysiological implications for learning to communicate in a language. Although the neurophysiological implications of teaching behaviors are not addressed as such, comments made by Rassias in the film Effective Language Teaching (1972) reflect such implications for teaching and learning through multisensory stimulation and sensory-integrative brain processes. Rassias greets teachers who attend workshops with the statement that they will have their vitality and enthusiasm judged by their "ability to meet the class and to flow with it and get out of it maximum performance." He emphasizes the need for reality and for commitment

PAGE 101

90 to communication, stating, "If it is to have any validity today, the entire process of communication must shape our awareness of ourselves and our awareness of others. Nothing is real unless it touches me and I am aware of it" ( Effective Language Teaching . 1972). Without referring to teacher-directed sensory-motor stimulation as such, Johnston (1980) described the vitality of teachers and tutors in an intensive language program as "the enthusiasm and objectified 'hard work' of instructors" who motivate the student to learn (p. 99). Johnston further described teaching behaviors prescribed by Rassias as economical because they effectively eliminate the need for expensive audio-visual equipment. He stated. The intensive language model relies on interpersonal contact between teacher and language learner. . . . The Master and Apprentice Teachers serve as models for the learner; they drill by means of a four-part technique, including (a) voice modulation, (b) snapping the fingers, (c) pointing at designated responders, and (d) intense eye contact. Through "surprise" and personal involvement, teachers motivate students to share in the unique experiences a foreign language can transmit, (p. 103) Byrd (1980) reiterated a similar theme in reporting the successful implementation of the Dartmouth Intensive Language Model that "exceeds by far that of traditional techniques previously employed at Emory and Henry" (pp. 297-298). He gave credit to the stimulation provided by apprentice teachers (tutors) when he stated. Using techniques developed by John Rassias (which require the drill leaders to move about constantly, never follow any set pattern, and punctuate the drill cues with a lively finger snap), the apprentice teachers conduct classes so that students remain alert, never knowing when they will be called on to respond. Not only are the techniques effective, but the apprentice teachers' enthusiasm and animation also turn out to be highly contagious; all at once the students begin to enjoy the fruits of their effort and, therefore, to profit from an al 1 -important intrinsic motivation, (p. 299)

PAGE 102

91 The terms used by Rassias, Johnston, and Byrd in reference to such concepts and behaviors as vitality, animation, enthusiasm, motivation, awareness, eye contact, movement, voice modulation, finger-snapping, and finger-pointing are analogous to the "high and low inference" teaching concepts and behaviors hypothesized to be related to student achievement by Rosenshine (1970, p. 510). In the present study, the application of the neurophysiological implications of these teaching behaviors to the term teacherdirected sensory-motor stimulatio n provided the theoretical framework for the quasi-experiment designed to determine the effects of teacher-directed sensory-motor stimulation on student foreign language communication effectiveness. Interpretations of the results of this study were based on the significance of inferential statistical analyses. Discussion of Inferential Statistical Findings In the second phase of the study, the independent variable was computed as levels of teacher-directed sensory-motor stimulation. The results of computations showed that whereas the combined teacher and tutor hours in two required semesters of the intensive language program were 1.6 times as many as in three required semesters of the traditional program, the level of teacher-directed sensory-motor stimulation provided for students in the intensive language program was 5.5 times as high as that provided in the traditional language program. These results were interpreted as documentation that teachers who used selected kinesic and vocal language behaviors prescribed in the Dartmouth Intensive Language Model provide a higher level of sensory-motor stimulation for students than do traditional language teachers, regardless of contact time.

PAGE 103

92 Rejection of Null Hypotheses 2, 2a, 2b, and 2c The rejection of these hypotheses at highly significant (alpha < .0001) probability levels verified the second assumption that students who are provided a higher level of teacher-directed sensory-motor stimulation achieve a higher level of foreign language communication effectiveness. This was the major finding in the present study to determine the effects of teacher-directed sensory-motor stimulation on student foreign language communication effectiveness. One of the assumptions based on the second main assumption of the study suggested that students in the intensive language program would have lower scores on the grammatical structure-vocabulary test than would students whose instructors used traditional teaching methods and textbooks emphasizing English translation and grammatical analysis of the structure of German. Analysis of data pointed in the opposite direction. Students in the intensive language program, where grammar was taught only incidently through the use_ of language, scored significantly higher on the grammar test than did students in the traditional language program. A further assumption of the study was that students in the intensive language program would have higher listening and speaking test scores than would students who were provided a lower level of teacherdirected sensory-motor stimulation in the traditional language program. Analysis of data confirmed this. The mean listening comprehension test score for the intensive (experimental) group was 21.30, whereas the mean for the traditional (control) group was 15.91. Even more indicatively, the mean speaking proficiency test score for the intensive group was 24.93, whereas the mean for the traditional group was 15.19.

PAGE 104

93 The rejection of the second main and subsectional null hypotheses substantiates findings noted by Hosenfeld (1979) to indicate the inadequacy of learning through passive reading-grammar tasks. Additionally, the rejected hypotheses in the present study support the assumption of Rassias (1970) that students learn to communicate effectively in foreign languages through us£ of the language, not through analysis of the language. These findings suggest congruence with empirical findings in occupational therapy (e.g.. Angel o, 1980; Ayres, 1972a, 1981) that showed academic and language performances improving following sensory-motor stimulation. Failure to Reject Null Hypotheses 3a, 3b, and 3c Results of chi -square analyses of data from rating scales showed no significant (alpha < .05) difference between the traditional and intensive groups on their self-assessments of their reading comprehension, listening comprehension, and writing proficiency. The frequencies with which the two groups of students self-assessed their reading comprehension skills were fairly proportional. The frequencies with which the two groups self-assessed their listening comprehension skills indicated that a higher, not significant, percentage (64%) of students in the intensive group than in the traditional group (43.75%) felt confident about their listening comprehension of German. The frequencies with which the two groups self-assessed their writing proficiency skills showed that a higher, not significant, percentage (22%) of intensive language students than of traditional language students (6.25%) felt their writing skills to be nominally poor.

PAGE 105

94 These findings were consistent with previous reports of beginning German student attitudes by Von Gal (1974) and Johnston (1980). Von Gal's large-scale study revealed that students in the established traditional beginning German program felt less than satisfied with their ability to use the language functionally in real-life situations. Johnston reported that students in the experimental intensive language program, when asked for suggestions to improve the program, suggested that they be given more reading and writing assignments. The failure to reject hypotheses formulated to test the significance of differences between the groups of students on the selfassessments of their reading, listening, and writing skills showed that the majority of students in both groups felt that their ability to use the German language in these communication modes was at least "fair." In the case of reading comprehension, the majority of students in both groups felt their ability to perform this skill to be "excellent." In the case of listening comprehension, the majority of students in the intensive group self-rated this skill as "excellent." Writing appeared to be the only skill in which the majority in neither group felt highly proficient. These results reflect the objectivity with which the student subjects self -assessed their skills—students appeared to try to use the rating scale conscientiously, so as to provide the investigator with undistorted, nonsubjecti ve data. Rejection of Null Hypothesis 3d The results of chi -square analysis of the frequency with which the two groups of students rated their speaking proficiency showed that a significantly (alpha < .01) higher percentage {58%) of students in

PAGE 106

95 the intensive group than in the traditional group (25%) rated their ability to speak the German language as "excellent." Conversely, a significantly (alpha < .01) higher percentage (37.5%) of students in the traditional group than in the intensive group (16%) rated their speaking proficiency as "poor." These results might have been anticipated from numerous accounts of the success of the Dartmouth Intensive Language Model in increasing students' oral communication confidence (e.g., Byrd, 1980; Effective Language Teaching , 1972; Johnston, 1980; Luxenberg, 1978; Rassias, 1970, 1972). Observations of students who were provided a higher level of teacher-directed sensory-motor stimulation in intensive beginning German classrooms showed the spontaneity and exuberance with which these students discoursed in a foreign language. Anecdotal evidence from intensive language students at the University of Florida and at Dartmouth attests to the phenomenon of beginning language students speaking a foreign language without inhibition within a few days after enrollment in an intensive language program. In addition, the rejection of Null Hypothesis 3d was consistent with the rejection of Null Hypothesis 2c that indicated the significantly higher ability of students in the intensive language program to speak proficiently as measured by test scores. Rejection of Null Hypothesis 4 The fourth null hypothesis was concerned with students' satisfaction with their beginning German courses at the University of Florida. The results of chi-square analysis of data from the Satisfaction Rating Scale showed that a significant (alpha < .001) percentage (86%) of

PAGE 107

96 intensive language students assigned a high value to their satisfaction with their courses in the intensive language program. This finding was consistent with the statistically significant lower percentage (4%) of intensive language students than of traditional language students (12.50%) who rated their courses in the lowest category. Observations of students in intensive classrooms indicated that students enjoyed the dynamic teacher/student interaction generated by teachers who used the intensive language model. Anecdotal evidence revealed that students were enthusiastic about the intensive language program and appreciative of the energy expended by teachers and tutors to stimulate learning through all of the senses. The rejection of the fourth null hypothesis in the present study substantiated informally observed affective responses of students in the intensive language program and supplemented evidence in the literature (e.g., Byrd, 1980; Johnston, 1980; Schulz, 1978; Stansfield & Hornor, 1981). Further substantiation was found in student comments on the Student Foreign Language Data Sheet and Rating Form. Both at the pilot testing and the later collection of data, students voluntarily wrote favorable comments regarding the intensive language model and instruction in the intensive language program. Although an analysis of these comments was not included in the scope of the study, a count of 129 forms showed that 37 out of 69 intensive language students voluntarily wrote in a section designated for "comments." Of these 37 writings, 29 were favorable comments about the instructors and the model. In contrast, of the 12 out of 60 traditional language students who voluntarily wrote a message on the form, only 3 commented favorably about the study of German.

PAGE 108

97 Another indication of the high regard students displayed for the efforts of instructors who used the intensive language model to help them learn to use and enjoy the German language was the announcement ("Three Graduate Students," 1982) that intensive language teacher A had been selected by a "University-wide selection committee" as one of three "outstanding teaching assistants for 1981-1982" (p. 14). The finding that students who were provided a higher level of teacher-directed sensory-motor stimulation were more satisfied with their intensive language instruction than were students who were provided a lower level of such stimulation was consistent with findings of Esp (1978) and Wyckoff (1973). Both Esp and Wyckoff reported that the nonverbal mobility and gesturing behaviors of teachers positively affected student attitude toward instruction.

PAGE 109

CHAPTER V CONCLUSIONS AND RECOMMENDATIONS This study arose from a need for educational research to improve the teaching of foreign languages in the United States, It was based on the belief that an understanding of how brain processes interact with language learning may help to explain how languages can be taught more effectively. The study was designed to determine if the neurophysiological implications of teaching behaviors prescribed in the Dartmouth Intensive Language Model could provide a basis for effective teaching. Conclusions Within the limitations stated in Chapter II, and on the basis of data collected and analyzed for this study, the following conclusions evolved: 1. Instructors who use the intensive language model provide more teacher-directed sensory-motor stimulation than do teachers who use traditional language instructional methods in beginning German classrooms at the University of Florida. This conclusion was based on descriptive findings from data collected from videotapes to measure the total mean seconds per contact hour of three selected kinesic and vocal language behaviors: (a) gross body movement, (b) hand gestures, and (c) spoken German. Instructors who used the intensive language model directed a mean of 4447.67 seconds per contact hour of these behaviors toward 98

PAGE 110

99 students, whereas traditional language instructors directed a mean of 1505,83 seconds per contact hour, at most, toward students. 2. Beginning German students who were provided a higher level of teacher-directed sensory-motor stimulation achieved a higher level of foreign language communication effectiveness than did students provided a lower level of stimulation. This conclusion was based on evidence from scores on three test forms of the Zertifikat Deutsch als Fremdsprache administered to two groups of student subjects. Students in the final semester of the intensive language program who comprised the experimental group achieved a significantly (alpha < .0001) higher mean total test score of 60.80 than did students in the traditional program, who comprised the control group, and who had a mean of 42.31. 3. Beginning German students who were provided a high level of teacher-directed sensory-motor stimulation in the intensive language program were more confident of their ability to communicate in German through speech than were students who completed the traditional language program. This conclusion was positively supported by the significantly (alpha < .01) higher percentage of students in the intensive language program who self-assessed their speaking proficiency as "excellent"; and was inversely supported by the significantly higher percentage of traditional language students who rated this ability as "poor." 4. Beginning German students who were provided a higher level of teacher-directed sensory-motor stimulation were significantly (alpha < .001) more satisfied with their foreign language courses than were students who were provided a lower level. This conclusion was suggested by the fact that 86% of students in the intensive group were highly

PAGE 111

100 satisfied with the efficacy of intensive language courses to enable them to communicate effectively in German. This evidence of "high" satisfaction was absent in the responses of 56.25% of students in the traditional language program. The conclusions based on inferential statistics are not claimed to have resulted solely from the treatment variable, teacher-directed sensory-motor stimulation, as defined in this study. Instead, conclusions are presented with the suggestion that they may have been caused, at least partially, by the teaching behaviors prescribed in the Dartmouth Intensive Language Model— behaviors that were assumed to have neurophysiological implications for language learning. Recommendations for further study of the problem posed by the need for a coherent theory of language learning on which to base language teaching behaviors were derived from consideration of the acknowledged limitations of the present study and from implications of the findings. Recommendations The present study grew out of a need for improved foreign language learning and more effective language teaching. It centered on the apparent effectiveness of the Dartmouth Intensive Language Model with prescribed teaching behaviors that appeared to stimulate students to learn more easily to communicate in a foreign language. A review of literature in related fields provided support for the neurophysiological implications for language learning found in teaching behaviors prescribed by Rassias to increase foreign language communication effectiveness.

PAGE 112

101 Evidence corroborated the theory that human beings learn through their sensory systems as a natural consequence of the processes of living. Support was found for the premise that those who wish to influence the learning behaviors of other human beings must somehow stimulate the senses of the learner. According to sensory integration theories developed by Ayres (1972b), the learning organ called the brain develops and functions through "the action of the environment upon the organism and the reaction of that organism upon the environment" (p. 22). Although Ayres developed her theories of the need for sensory-motor stimulation from her experiences as an occupational therapist working to mitigate learning dysfunctions in children, her concepts may be applied to all types of learning dysfunction— in the present study, the lack of facility with foreign languages. Thus the present study was based on the theory that the teacher is the reality of learning whenever the teacher behaves so as to interact with the student to elicit an adaptive sensory-motor response from the student. In an environment of dynamic teacher/student interaction, both teacher and student teach and learn through teacher-directed sensory-motor stimulation. One of the major limitations of the present study was the inability to measure all possible student and teacher behaviors that might indicate the extent of interaction observed in intensive beginning classrooms at the University of Florida. The three kinesic and vocal language behaviors that were measured for their effectiveness in promoting student achievements on three tests measuring knowledge of grammar, listening comprehension, and speaking proficiency in German do not tell the whole story of the phenomenon of learning that appeared to happen in these classrooms.

PAGE 113

102 Despite the highly significant results of this study, further research is required to substantiate the creative teacher's intuitive feelings about the art of teaching, and to elucidate the biological and neurophysiological implications of the relationship of effective teaching behavior to effective learning behavior that have been hinted at throughout history. Such scholars as Bacon, Comenius, Dewey, James, Montessori , Ogden, Pestalozzi, and Whitehead related the art of teaching to the art of developing human potential through the natural integrating learning processes of living, doing, and experiencing. The potential of multisensory stimulation described in the recent writings of Coppola (1970), Edwards (1979), Montague (1978), and Nelson (1977) must be studied. The need for dynamic teacher/student interaction for the development of the natural capacity to learn has been hypothesized recently by numerous concerned educators (e.g.. Chappie, 1981; Gorman, 1974; Holt, 1967, 1981; Hyman, 1974; Suzuki, 1969, 1973). Perhaps these hypotheses can be tested through research designed to reduce the acknowledged limitations to the conclusions of the present study, not only for the improvement of foreign language education, but for the benefit of education in general. Specific suggestions for such research include: 1. Follow-up studies designed to reduce limitations of (a) nonrandomization, (b) lack of information about the personality factors of students, and (c) constraints that prevent the measurement of all possible student and teacher behaviors that might indicate the extent of interaction observed in beginning German classrooms at the University of Florida.

PAGE 114

103 a. To remove limitations of nonrandomization, follow-up studies may take the form of true experiments in which students are randomly selected and randomly assigned to treatment groups that are differentiated by higher and lower levels of teacher-directed sensory-motor stimulation as measured in the present study. b. To reduce the influence of personality factors, followup studies may attempt to measure personality configurations and learning styles that tend to influence the behaviors of students (Llorens & Adams, 1976). The anecdotal evidence of positive comments made by students in the intensive language program suggested that students who self-selected this program had personalities and learning styles conducive to less passive, less verbal, less linear, and less literal instructional methods than are traditionally used in the American educational system (Rennels, 1976). Conversely, the rare negative reactions from students in the intensive language program pointed to learning styles more suited to what Rennels calls "education's emphasis on analytical/ linear functions considered desirable in Western society" (p. 471). According to Rothbart (1972), everyone is born with the potential for a creative, global learning style— a style he described as "the product of intense interaction between inborn abilities and propensities and favorable environmental conditions" (p. xi). Research is needed in general education, as in foreign language education, to confirm or to deny speculations that analytical/linear methods of instruction condition students to inhibit their natural neurophysiological capacity to gather information through their senses while actively participating, experiencing, doing, and communicating.

PAGE 115

104 c. To alleviate constraints to the measurement of all possible nonverbal and vocal language teaching behaviors, follow-up studies should allow more time and more sophisticated instruments and notational systems to record and count these behaviors. In future studies, more time and better instrumentation may be employed to measure facial expressions, voice tones, eye contact, and a myriad of other paralinguistic acts and communicative body movements defined most comprehensively by Key (1975, 1977) and discussed most intuitively by Sarles (1974, 1977). 2. Studies of a more naturalistic or ethnographic nature designed to capture the qualitative rather than the quantitative aspects of teaching behaviors, communication effectiveness, and dynamic teacher/ student interaction observed in beginning German classrooms. Such studies may amplify the use of videotape equipment used in the present study through the use of high speed or low speed cinematography for behavioral microanalysis of classroom interactions. The works of Condon (1966, 1974), Eibl -Eibesfeldt (1972, 1975), and Trevathen (1974, 1977) exemplify the work being done to discover the ways we human beings learn through communication, language, and the interactive experiences of life. Such work has been documented in a film (Wolff, 1971) showing the ways that babies naturally depend on their sensory systems to gather information from the environment through the phenomenon of the movement of bodies through time and space. A film such as this is needed to document the dependence of students on the movements of their instructors, not only in foreign language classes, but in all classes designed for the exchange of information from teacher to student.

PAGE 116

105 3. Studies designed to apply the teaching behaviors prescribed in the Dartmouth Intensive Language Model to other areas of teaching and learning. The major conclusion reached in the present study suggests that not only were the receptive and expressive language skills significantly improved by a high level of teacher-directed sensory-motor stimulation, but also the grammatical skills. Students who were taught grammar only incidently through the us£ of the language scored significantly (alpha < .0001) higher on a test of grammatical structurevocabulary than did students in a traditional program that emphasized structural analysis and English translation of syntax. This finding points to the need for more research to substantiate the effectiveness of subcortical sensory-integrative processes for the facilitation of learning. Such research is needed not only by foreign language educators, but by all instructors who understand that the effectiveness of a formal plan of curriculum and instruction is contingent on what Snyder (1970) called the more effective "hidden curriculum" that is taught and learned with subconscious ease through the medium of nonverbal communication. In summary, if the conclusions of the present study have to any extent determined that the neruophysiological implications of teaching behaviors prescribed by Rassias can provide teachers with a basis for effective language teaching, there is much need for continued research to test the theory that teacher-directed sensory-motor stimulation effects student learning.

PAGE 117

REFERENCES Adams, J. A. Issues for a closed-loop theory of motor learning. In G. E. Stelmach (Ed.), Motor control. Issues and trends . New York Academic Press, 1976. Adams, J. A. Theoretical issues for knowledge of results. In G. E. Stelmach (Ed.), Information Processing in motor control and learning . New York: Amsterdam Press, 1978. Angel 0, J. K. B. Effects of sensory integration treatment on the lowachieving college student. American Journal of Occupational Therapy , 1980, 34, 671-675. Asher, J. J. The total physical response approach to second language learning. Modern Language Journal , 1969, 5^, 3-17. Asher, J. J. Children's first language as a model for second language learning. Modern Language Journal . 1972, 56, 133-139. Asher, J. J. Learning another language through actions. The complete teacher's guidebook . Los Gatos, Calif.: Sky Oaks Publications, 1977. Asher, J. J. Fear of foreign languages. Psychology Today , 1981, 15, 52-59. ~ Ayres, A. J. Tactile functions. American Journal of Occupational Therapy , 1964, 18, 6-11. Ayres, A. J. Improving academic scores through sensory integration. Journal of Learning Disabilities , 1972, 5, 338-343. (a) Ayres, A. J. Sensory integration and learning disorders . Los Angeles: Western Psychological Services, 1972. (b) Ayres, A. J. Effect of sensory integration therapy on the coordination of children with choreoathetoid movement. American Journal of Occupational Therapy , 1977, 31^, 291-293. Ayres, A. J., & Mailloux, Z. Influence of sensory integration procedures on language development. American Journal of Occupational Therapy , 1981, 35, 383-390. Bacheller, F. Communication effectiveness as predicted by judgments of the severity of learner errors in dictations. In J. W. Oiler & K. Perkins (Eds.), Research in language testing . Rowley, Mass.: Newbury House, 1980. 106

PAGE 118

107 Bailey, N., Madden, C, & Krashen, S, D. Is there a "natural sequence" in second language learning? Language Learning , 1974, 24, 235244. Balke, F. H. An assessment of a German study abroad language program: An evaluation of the Oregon Summer Study Abroad (Doctoral dissertation. University of Oregon, 1980). Dissertation Abstracts Internationa! , 1980, 41, 1983A. (University Microfilms No. 8024841) Barsch, R. H. Achieving perceptual -motor efficiency (Vol. 1). Seattle, Wash.: Special Child Publications, 1967. Barsch, R. H. Achieving perceptual -motor efficiency (Vol. 2). Seattle, Wash.: Special Child Publications, 1968. Bechtereva, N. P. The neurophysiological aspects of human mental activity (2nd ed.). New York: Oxford University Press, 1978. Benseler, D. P., & Schulz, R. A. Methodological trends in college foreign language instruction. Modern Language Journal , 1980, 64^, 88-96. Birdwhistell , R. L. Introduction to Kinesics: An annotation system of body motion and gesture . Louisville, Kentucky: University of Louisville, 1952. Birdwhistell, R. L. Kinesics and context . Philadelphia: University of Pennsylvania Press, 1970. Birdwhistell, R. L. The language of the body: the natural environment of words. In A. Silverstein (Ed.), Human communication: Theoretical explorations . Hillsdale, N.J.: Lawrence Erlbaum, 1974. Birkmaier, E. M. Research on teaching foreign languages. In R. M. Travers (Ed.), Second handbook of teaching . Chicago: Rand McNally, 1973. Bred, R. I. (Ed.). Language study for the 1980s . Chicago: Modern Language Association of America, 1980. Bruner, J. S. On knowing . Cambridge, Mass.: Harvard University Press, 1962. Bruner, J. S. From communication to language: a psychological perspective. Cognition , 1975, 3, 255-287. (a) Bruner, J. S. The ontogenesis of speech acts. Journal of Child Language . 1975, 2, 1-19. (b) Bruner, J. S. Early social interaction and language acquisition. In H. R. Schaffer (Ed.), Studies in mother-infant interaction . London: Academic Press, 1977.

PAGE 119

108 Bruner, J. S. Foreword. In A, Lock (Ed.), Action, gesture and symbol . London: Academic Press, 1978. Byrd, C. W. Intensive language instruction at a small liberal arts college: The Dartmouth Model at Emory and Henry. Modern Language Journal , 1980, 64, 297-302. Campbell, D. T,, & Stanley, J. C. Experimental and quasi-experimental designs for research on teaching. In N. L. Gage (Ed.), Handbook of research on teaching . Chicago: Rand McNally, 1963. Carroll, J. B. The foreign language attainments of language majors in a survey conducted in U.S. colleges and universities: Final report . Contract No. OE-4-14-048 U.S. Office of Education) Cambridge, Mass.: Harvard University, 1967. (ERIC Document Reproduction Service No. ED 013 434) Carroll, J. B. Promoting language skills: The role of instruction. In D. Klahr (Ed.), Cognition and instruction . Hillsdale, N.J.: Lawrence Erlbaum, 1976. Chall, J. S., & Mirskey, A. F. (Eds.). Education and the brain: The seventy-seventh yearbook of the National Society for the Study of Education (Pt. 2) . Chicago: The University of Chicago Press, 1978. Chappie, E. D. Movement and sound: The musical language of body rhythms in interaction. Teachers College Record , 1981, 82, 635-648. Chomsky,, N. Linguistic theory. In J. W. Oiler 8 J, C. Richards (Eds.), Focus on the learner . Rowley, Mass.: Newbury House, 1973. Christensen, C. B. Beyond the desk. In J. K. Phillips (Ed.). Building on experience: Building for success . Stokie, 111.: National Textbook, 1979. Chun, J. A survey of research in second language acquisition. Modern Language Journal , 1980, 64, 288-296. Clark, F. A., & Steingold, L. R. A potential relationship between occupational therapy and language acquisition. American Journal of Occupational Therapy , 1982, 36, 42-44. Coghill, G. E. Anatomy and the problem of behavior . Cambridge: University Press, 1929, Condon, W. S., & Ogston, W. D. Sound film analysis of normal and pathological behavior patterns. Journal of Nervous and Mental Diseases , 1966, 143, 338-347.

PAGE 120

109 1 Condon, W. S., & Sander, L.W. Neonate movement is synchronized with adult speech: Interactional participation and language acquisition. Science , 1974, 183, 99-101. Coppola, A. F. Reality of the haptic world. Phi Kappa Phi Journal . Winter 1970, 14-33. Cratty, B. J. Human behavior: Exploring educational processes . Wolf City, Texas: The University Press, 1971. Cratty, B. J. Psycho-motor behavior in education and sport . Springfield, 111.: Charles C. Thomas, 1974. Cratty, B. J. Movement behavior and motor learning {3rd ed.). Philadelphia: Lea "& Febiger, 1975. (originally publ i shed, 1973) Credell, C. F. Nonverbal behavior in teaching: Another dimension to classroom communication (Doctoral dissertation, Claremont Graduate School, 1977). Dissertation Abstracts International , 1977, 38 , 1848A. (University Microfilms No. 77-22464) Curran, C. A. Counsel ing-Learning: A whole person model for education . New York: Grune & Stratton, 1972. D'Anglejan, A. Language learning in and out of classrooms. In J. C. Richards (Ed,), Second and foreign language learning . Rowley, Mass.: Newbury House, 1978. Davis, M. Towards understanding the intrinsic in body movement . New York: Arno Press, 1975, de Quiros, J, B., & Schrager, 0. L. Neuropsychological fundamentals in learning , San Rafael, Calif.: Academic Therapy Publications, 1978. Deutscher Volkshochschul-Verband & Goethe-Institut. Das Zertifikat Deutsch als Fremdsprache (2nd ed.). Bonn-Bad Godesberg, Munchen: Deutscher Volkshochschul-Verband & Goethe Institute, 1977. Dohner, C. Rating scales in clinical evaluation. In D. M. Irby & M. K. Morgan (Eds.), Clinical evaluation: Alternatives for health related educators . Gainesville, Fla.: The Center for Allied Health Personnel , 1974. Dye, J. C. The use of body movement to facilitate second language learning for secondary school students: A teacher's guide (Doctoral Dissertation, New York University, 1977). Dissertation Abstracts International . 1978, 38, 6 080 A. (University Microfilms No. 78-03049) Eccles, J. C. (Ed.). Brain and conscious experience . New York: Springer-Verlag, 1966. Eccles, J. C. The understanding brain . New York: McGraw-Hill, 1977.

PAGE 121

no Eccles, J. C, Ito, M., & Szentagothai , J. The cerebellum as a neuronal machine . New York: Springer-Verlag, 1967. Eckman, F. R., & Hastings, A. J. (Eds.). Studies in first and second language acquisition . Rowley, Mass.: Newbury House, 1977, Edelman, G. M., & Mountcastle, V. B, The mindful brain . Cambridge, Mass. : The MIT Press, 1978. Edwards, B. Drawing on the right side of the brain . Los Angeles: J. P. Tarcher, 1979. Effective Language Teaching . New York: Association-Sterling Films, 1972. (Film) Eibl-Eibesfeldt, I. Similarities and differences between cultures in expressive movements. In R. A. Hinde (Ed.). Nonverbal communication . Cambridge: University Press, 1972. Eibl-Eibesfeldt, I. Ethology: The biology of behavior (2nd ed.). New York: Holt, Rinehart & Winston, 1975. Ekman, P., & Friesen, W. V. Constants across cultures in face and emo, tion. J ournal of Personality and Social Psychology , 1971, 17 , 124-129. Esp, B. The effects of teacher kinesic communication (Doctoral dissertation, Hofstra University, 1978) Dissertation Abstracts International , 1978, 39, 2828A. (University Microfilms No. 78-21251) Fanslow, J. F. The treatment of error in oral work. Foreign Language Annals , 1977, 10, 583-593. Gage, N. L. Theories of teaching. In E.R. Hilgard (Ed.), Theories of learning and instruction: The sixty-third yearbook of the National Society for the Study of Education (Pt. 2). Chicago: The University of Chicago Press, 1964. Gage, N. L., Belgard, M., Dell, D., Hi Her, J. E., Rosenshine, B., & Unruh, W. R. Explorations of the teacher's effectiveness in explaining . Technical Report No. 4, Stanford Center for Research and Development in Teaching. Stanford University, 1968. (ERIC Document Reproduction Service No. ED 028 147) Galloway, V. B. Perceptions of the communicative efforts of American students of Spanish. Modern Language Journal , 1980, 64, 428-433. Gallwey, W.T, The inner game of tennis . New York: Random House, 1974. Geschwind N. Disconnection syndromes on animals and man. Brain , 1965, 88, 237, 294. Geschwind N. The organization of language and the brain. Science , 1970, 170, 940-944.

PAGE 122

Ill Gibson, J. J. The senses considered as perceptual systems . Boston: Houghton Mifflin, 1966. Goodman, L. Symposium No. 14: Introduction. Journal of Special Education . 1979, 13, 7. Gorman, A. H. Teachers and learners (2nd ed.). Boston: Allyn & Bacon, 1974. Hall, E. T. The silent language . New York: Doubleday, 1959. Hall, E. T. The hidden dimension . New York: Doubleday, 1966. Hall, E. T. Proxemics. Current Anthropology , 1968, 9^, 83-108. Hebb, D. 0. The organization of behavior . New York: Wiley, 1949. Hebb, D. 0. Essay on mind . Hillside, N.J.: Lawrence Erlbaum, 1980. Heilman, K. M. Language and the brain: Relationship of localization of language function to the acquisition and loss of various aspects of language. In J. S. Chall & A. F. Mirskey (Eds.), Education and the brain: The seventy-seventh yearbook of the National Society for the Study of Education (Pt. 2). Chicago: The University of Chicago Press, 1978. Herrick, C. J. The evolution of human behavior . Austin: University of Texas Press, 1956. Hieke, A. E. Audio-lectal practice and fluency acquisition. Foreign Language Annals , 1981, 14, 189-194. Hilgard, E. R., & Bower, G. H. Theories of learning (3rd ed.). New York: Appleton-Century, 1966. Holt, J. How children learn . New York: Dell, 1967. Holt, J. Teach your own . New York: Delacorte Press/Seymour Lawrence, 1981. Hosenfeld, C. A learning-teaching view of second language instruction. Foreign Language Annals , 1979, 12, 51-54. Hubel , D. H. The brain. In The brain. A Scientific American book . San Francisco: W. H. Freeman, 1979. Huck, S. W., Cormier, W. H., & Bounds, W. G., Jr. Reading statistics and research . New York: Harper & Row, 1974. Hyman, R. T. (Ed.). Teaching: Vantage points for study . Philadelphia: J. B. Lippincott, 1974.

PAGE 123

112 Jackson, G. L. The effects of video-vs. audiotaped interviews on listening comprehension in third-quarter, beginning college French (Doctoral dissertation, Ohio State University, 1979). Dissertation Abstracts International , 1979, 40, 3842A. (University Microfilms No. 80-01750) Jacobson, M. Developmental neurobiology (2nd ed.). New York: Plenum Press, 1978. Jaffe, J. Parlimentary procedure and the brain. In A. W. Siegman & S. Feldstein (Eds.), Nonverbal behavior and communication . Hillsdale, N.J.: Lawrence Erlbaum, 1978. Jesperson, 0. How to teach a foreign language . (S. Yhlen-Olsen Bertelsen, trans.) London: George Allen & Unwin, 1904. Johnston, 0. W. Implementing the intensive language model: An experiment in German at the University of Florida. Foreign Language Annals . 1980, 13^, 99-106. Jones, R. L. Testing: A vital connection. In J. K. Phillips (Ed.), The language connection: From classroom to the world . Stokie, 111.: National Textbook, 1977. Jorstad, H. L. The education and reeducation of teachers. In F. M. Grittner (Ed.), Learning a second language: The seventy-ninth yearbook of the National Society for the Study of Education (Pt. 2). Chicago: University of Chicago Press, 1980. Keilstrup, D. V. Practical guidelines and activities for an advanced foreign language intensive program. Modern Language Journal , 1981, 65^, 377-382. Kendon, A. Movement coordination in social interaction; some examples described. Acta Psychol ogi ca , 1970, 32, 100-125. Kendon, A., Harris, R. M., & Key, M. R. (Eds.). Organization of behavior in face-to-face interaction . The Hague: Mouton, 1975. Kephart, N. C. The slow learner in the classroom . Columbus, Ohio: Charles E. Merrill Books, 1960. Key, M. R. Paralanguage and kinesics . Metuchen, N.J.: The Scarecrow Press, 1975. Key, M. R. Nonverbal communication: A research guide and bibliography . Metuchen, N. J.: The Scarecrow Press, 1977. Kimura, D. The asymmetry of the human brain. Scientific American , 1973, 228, 70-78. Krashen, S. D. Adult second-language acquisition as post-critical period learning. International Review of Applied Linguistics and Language Teaching , 1979, 43, 39-52.

PAGE 124

113 Lamendella, J. General principles of neurofunctional organization and their manifestation. Language Learning , 1977, 2J_, 155-196. Lapan, M. T, Acting on the realities of second language education. In J. K. Phillips (Ed.) J The new imperative: Expanding horizons of foreign language education . Stokie, 111.: National Textbooks, 1980. Lashley, K. S. Brain mechanisms and intelligence . Chicago: University of Chicago Press, 1929. Lashley, K. S. In search of the engram. In D. G. Stein & J. R. Rosen (Eds.), Learning and memory . New York: Macmillan, 1974. (Reprinted from Society of Experimental Biology Symposium No. 4 . Cambridge University Press, 1950.") Lenneberg, E. H. Biological foundations of language . New York: John Wiley & Sons, 1967. Le Winn, E. B. Human neurological organization . Springfield, 111.: Charles C. Thomas, 1969. Lieberman, P. A study of prosodic features. In T. A. Sebeok (Ed.), Current trends in linguistics (Vol. 12). The Hague: Mouton, 1974. Lieberman, P. The evolution of speech and language. In J. F. Kavanaugh & J. E. Cutting (Eds.), The role of speech in language . Cambridge, Mass.: The MIT Press, 1975. Llorens, L. A., & Adams, S. P. Entering behavior — student learning styles. In C. W. Ford & M. K. Morgan (Eds.), Teaching in the health professions . St. Louis: C. V. Mosby, 1976. Lock, A. (Ed.). Action, gesture and symbol . London: Academic Press, 1978. Luria, A. R. The working brain: An introduction to neuropsychology (B. Haigh, trans. ) . New York: Basic Books, 1973. Lutz, K. A. The implications of brain research for learning strategies and educational practice . Los Angeles: University of California, 1978. (ERIC Document Reproduction Service No. ED 163 068) Luxenberg, S. All the class is a stage. Change . 1978, 10, 30-33. Magrun, W, M. , Ottenbacher, M. S., McCue, S., & Keefe, R. Effects of vestibular stimulation on spontaneous use of verbal language in developmental ly delayed children. American Journal of Occupational Therapy , 1981, 35, 101-104.

PAGE 125

114 McCallum, R. S., & Glynn, S. M. The hemispheric specialization con struct: Developmental and instructional considerations for cre ative behavior . Paper presented at the meeting of the National Association of School Psychologists, San Diego, Calif., 1979 (ERIC Document Reproduction Service No. ED 172 492) McGuiness, D., & Pribram, K. The neuropsychology of attention. In M. C. Wittrock (Ed.), The brain and psychology . New York: Academic' Press, 1976. Mehrabian, A. Nonverbal communication . Chicago: Aldine, 1972, Mehrabian, A. Public places and private spaces . New York: Basic Books, 1976. Mehrabian, A., & Ferris, S. R. Inference of attitudes from nonverbal communication channels. Journal of Consulting Psychology , 1967, 31,, 248-252. (a) Mehrabian, A., & Weiner, M. Decoding of inconsistent communications. Journal of Personality and Social Psychology , 1967, 6^, 109-114. (b) Montague, A. Touching: The human significance of skin . New York: Harper & Row, 1978. Moskowitz, G. Caring and sharing in the foreign language classroom . Rowley, Mass.: Newbury House, 1978. Mountcastle, V. B. An organizing principle for cerebral function: The unit model of the distributing system. In G. M. Edelman & V. B. Mountcastle, The mindful brain . Cambridge, Mass.: The MIT Press, 1978. Nelson, G. How to see . Boston: Little, Browne, & Co., 1977. Office of Education, DHEW. Strength through wisdom: A critique of U.S . capability. A report to the President from the President's Commis sion on Foreign Language and International Studies . Washington, D. C.: U.S. Government Printing Office, 1979. Oiler, J. W., & Perkins, K. (Eds.). Research in language testing . Rowley, Mass.: Newbury House, 1980. Oiler, J. W., & Richards, J. C. (Eds.). Focus on the Learner . Rowley, Mass.: Newbury House, 1980. Olson, L., & Samuels, S. J. The relationship between age and accuracy of foreign language pronunciation. Journal of Educational Research , 1973, 66, 263-268. i

PAGE 126

115 Ong, W. J. The history and the future of verbal media. In A. Silverstein (Ed.), Human communication: Theoretical explorations . Hillsdale, N.J.: Lawrence Erlbaum, 1974. Paget, R. Human speech . New York: Harcourt & Brace, 1930. Perkins, J. A. Cover letter from the chairman. Strength through wisdom : A critique of U.S. capability. A report to the President from the President's Commission on Foreign Language and International Studies . (Office of Education, DHEW) Washington, D.C.: U7S. Government Printing Office, 1979. Petrovich, S. B. & Hess E. H. An introduction to animal communication. In A. W. Siegman & S. Feldstein (Eds.), Nonverbal behavior and communication . Hillsdale, N.J.: Lawrence Erlabaum, 1978. Piaget, J. The language and thought of the child (3rd ed., M. Gabain, trans. 7^ New York: Humanities Press, 1959. (Originally published, 1926.) Pribram, K. H. Neurological notes on the art of education. In E. R. Hilgard (Ed.), Theories of learning and instruction: The sixtythird yearbook of the National Society for the Study of Education (Pt. 1). Chicago: The University of Chicago Press, 1964. Pribram, K. H. Languages of the brain . Englewood Cliffs, N.J.: Prentice Hall, 1971. Pribram, K. H. Modes of central processing in human learning and remembering. In T. Teyler (Ed.), Brain and learning . Stanford, Conn.: Greylock, 1978. Primeau, J. K. The resurgence of foreign language study. Modern Language Journal , 1979, 53, 117-122. Racle, G. L. Can suggestopaedia revolutionize language teaching? Foreign Language Annals , 1979, 12, 39-49. Rassias, J. A. Report to the Esso Foundation on two years' experimentation in intensive language training at Dartmouth College . Hanover, N.H. : Dartmouth College, 1970. (ERIC Document Reproducti on Service No. ED 161 841) Rassias, J. A. Why we must change. Association of Departments of Foreign Languages Bulletin , March 1972, 3^, 9-13. Rassias, J. A. New dimensions in language training: The Dartmouth College experiment . Washington, D.C.: Department of Health, Education and Welfare, 1976. (ERIC Document Reproduction Service No. ED 046 286) Rennels, M. R. Cerebral symmetry: An urgent concern for education. Phi Delta Kappan , 1976, 57, 471-472.

PAGE 127

116 Richards, J. C. (Ed.). Understanding second and foreign language learning . Rowley, Mass.: Newbury House, 1978. Rivers, W. M. Students, teachers, and the future. Foreign Language Annals , 1975, 8, 23-32. Rosenshine, B. Objectively measured behavioral predictors of effectiveness in explaining. In N. L. Gage, M. Belgard, D. Dell, J. E. Hiller, B. Rosenshine, & W. R. Unruh, Explorations of the teacher's effectiveness in explaining . Technical Report No. 4 Stanford Center for Research and Development in Teaching. Stanford University, 1968. (ERIC Document Reproduction Service No. ED 028 147) Rosenshine B. Enthusiastic teaching: A research review. School Review , 1970, 78, 419-515. Rothbart, H. A. Cybernetic creativity . New York: Robert Speller & Sons, 1972. SAS Institute Incorporated. The SAS users guide (1979 ed.). Cary, N.C.: SAS Institute Incorporated, 1979. Sarles, H. B. Facial expressions and body movement. In T. A. Sebeok (Ed.), Current trends in linguistics (Vol. 12). The Hague: Mouton, 1974. Sarles, H. B. After metaphysics. Series: Studies in semiotics (Vol. 13). Lisse: The Peter De Ridder Press, 1977. Sarnat, H. B., & Netsky, M. Evolution of the nervous system . Oxford: Oxford University Press, 1974. Scheflen, A. A. Human communication: Behavioral programs and their integration in interaction. Behavioral Science , 1968, 13^, 44-55. Schiefelbush, R. L., & Hollis, J. H. (Eds.). Nonspeech language and communication . Baltimore: University Park Press, 1980. Schulz, R. A. Survey of successful undergraduate foreign language programs in U.S. institutions of higher education . Washington, D.C.: National Endowment for the Humanities, 1978. (ERIC Document Reproduction Service No. ED 159 924) Sheldon, A. Assumptions, methods and goals in language acguisition research. In F. R. Eckman & A. J. Hastings (Eds.), Studies in first and second language acquisition . Rowley, Mass.: Newbury House, 1977. Sherrington, C. The integrative action of the nervous system . New York: Arno Press, 1973. (Reprint of the 1948 reprinted edition, Cambridge: University Press) (First published in 1906.)

PAGE 128

117 Siegman, A. W., & Feldstein, S. (Eds.). Nonverbal behavior and communication . Hillsdale, N. J.: Lawrence Eribaum, 1978. Smith, J.L. Sensorimotor integration during motor programming. In G. E. Stelmach (Ed.), Information processing in motor control and learning . New York: Academic Press, 1978. Snow, C. E. Mother's speech to children learning language. Child Development , 1972, 43, 549-565. Snow, C. E. The development of conversation between mothers and babies. In V. Lee (Ed.), Language development . New York: John Wiley & Sons, 1979. Snow, C. E., & Hoefnagel -hTdhle. Age differences in the pronunciation of foreign sounds. Language and Speech , 1977, 20_, 357-365. Snyder, B. The hidden curriculum . New York: Alfred A. Knopf, 1970. Sperry, R.W. The great cerebral commissure. Scientific American , January 1964, 42-52. Sperry, R. W. Lateral specialization in the surgically separated hemispheres. In F. 0. Schmitt & F. G. Worden (Eds.), The neuro sciences (Vol. 3). Cambridge, Mass.: The MIT Press, 1974. Spitz, R. A., & Wolf, K. M. The smiling response. Genetic Psychology Monographs , 1946, 34, 59-125. (Reprinted in Facial Expressions in Children: Three Studies . New York: Arno Press, 1972) Stansfield, C, & Hornor, J. The Dartmouth-Rassias model of teaching foreign languages. Association of Departments of Foreign Language Bulletin , 1981, 12, 21-27. Stelmach, G. E. (Ed.). Motor control: Issues and trends . New York: Academic Press, 1976. Stelmach, G. E. (Ed.). Information processing in motor control and Learning . New York: Academic Press, 1978. Stelmach, G. E., & Requin, J. (Eds.). Tutorials in motor behavior . Amsterdam: North Holland Publishing, 1980. Stevick, E. W. Language instruction must do an about face. Modern Language Journal , 1974, 58, 379-384. (a) Stevick, E. W. The meaning of drills and exercises. Language Learning, 1974, 24, 1-23. (b) ,

PAGE 129

118 Stilwell, J. M., Crowe, T. K. , & McCallum, L. W. Postrotary nystagmus duration as a function of communication disorders. American Journal of Occupational Therapy , 1978, 32, 222-228. Strasheim, L. A. Creativity lies trippingly on the tongue. Modern Language Journal , 1971, 55^, 339-345. Studdert-Kennedy, M. The perception of speech. In T. A. Sebeok (Ed.), Current trends in linguistics (Vol. 12). The Hague: Mouton, 1974.. Suzuki, S. Nurtured by love: A new approach to education . New York: Exposition Press, 1969. Suzuki, S. Children can develop their ability to the highest standard. In E. Mills & Sr. T. C. Murphey (Eds.), The Suzuki concept: An introduction to a successful method for early music education . Berkely, Calif.: Diable Press, 1973. Szentagothai , J., & Arbib, M. A. Conceptual models of neural organization . Cambridge, Mass.: The MIT Press, 1975. Teyler, T. (Ed.). Brain and learning . Stanford, Conn.: Greylock, 1978. Thiel, F. Sights, sounds, and scents; the physical side of foreign study. Modern Language Journal , 1980, 64, 434-440. Three graduate students named best teaching assistants. The Independent Florida Alligator , April 22, 1982, p. 14. Tomkins, S., & McCarter, R. What and where are the primary affects? Perceptual and Motor Skills , 1964, 18, 119-158. Tomkins, S., & Izard, C. E. (Eds.). Affect, cognition and personality . New York: Springer, 1965. Trevathen, C. Conversations with a two-month-old. New Scientist , 1974, 62, 230-235. Trevathen, C. Descriptive analysis of infant communicative behavior. In H. R. Schaffer (Ed.), Studies in motherinfant interaction . London: Academic Press, 1977. Trevathen, C. Functional organization of the human brain. In M. C. Wittrock (Ed.). The brain and psychology . New York: Academic Press, 1980. U.S. Congress. House. Congressman E. F. Landgrebe commending Professor Rassias for outstanding work in the field of language instruction. Extension of Remarks, 93rd Cong., 2nd sess., March 29, 1974, Congressional Record , vol. 102 (44).

PAGE 130

119 Valette, R. M. Modern language testing (2nd ed.). New York: Harcourt, Brace, & Jovanovich, 1977. Valette, R. M. Evaluating the second-language learning program. In F. M. Grittner (Ed.), Learning a second language: The seventy-ninth yearbook of the National Society for the Study of Education (pt. 2). Chicago: The University of Chicago Press, 1980. Von Gal, C. A proposal for a systematic approach to the evaluation of an ongoing first-year college language program . Unpublished master's thesis. University of Florida, 1974. Walsh, T. M., & Diller, K. C. Neurol inguistic foundations to methods of teaching a second language. International Review of Applied Linguistics in Language Teaching , 1978, 16^, 1-14. Warriner, H. P. Foreign language training in the schools— 1979: Focus on methodology. In Background papers and studies. A report to the President from the President's Commission on Foreign Language and International Studies . Office of Education (DHEW). Washington, D.C.: U.S. Government Printing Office, 1979. Waterson, N., & Snow, C. E. (Eds.). The development of communication . New York: John Wiley & Sons, 1978. Williams, H. G. Neurological concepts and perceptual -motor behavior. In R. C. Brown & B. C. Cratty (Eds.), New perspectives of man in action . Englewood Cliffs, N.J.: Prentice-Hall, 1969. Wittrock, M. C. Education and the cognitive processes of the brain. In J. S. Chall & A. F. Mirskey (Eds.), Education and the brain: The seventy-seventh yearbook of the National Society for the Study of Education (pt. 2)~ Chicago: The University of Chicago Press, 1978. Wittrock, M. C. Learning and the brain. In M, C. Wittrock (Ed.), The brain and psychology . New York: Academic Press, J98Q, Wolff, L. (Producer). Rock-a-bye baby . New York: Time-Life Films, 1971 (Film) Woodford, P. E. Foreign language testing. Modern Language Journal , 1980, 64, 97-102. Wyckoff, W. L. The effect of stimulus variation on learning from lecture. The Journal of Experimental Education , 1973, 41, 85-90. Young, J. Z. An introduction to the study of man . Oxford: The Clarendon Press, 1971. Young, J. Z. Programs of the brain . Oxford: Oxford University Press, 1978.

PAGE 131

APPENDIX A SAMPLE PAGE FROM VIDEO BEHAVIORAL OBSERVATION CHART FOR ONE TEACHER

PAGE 132

a: LU CO o cm o CO o UJ o o $.-H U " I— o CU •!^ r— o cn a. c 00 LU ^ E O SCO O T3 +-> O C ^ (U E to O O SE CD +-> O to E O U (U to O) -D SC 3 03 !-> rc lo CO M o to o c o o O) oo o I/) o c: o (J o CO c o o QJ 00 to 3 o O O I O o C\J <— I 00 cvj CO c» ^ ro ^ I I I O CM o O ro CO 00 in to ^ o rH O I I I CO VD LT) O CM LT) «D VO CNJ to CM o cri 1— I Lft tn I I I I o CO 'do CM «3tn r— I i-H r-H 1— I ^ i— I i-H rocTicooto>— tt— icno orocM <— i.-iCMrocoLr)0<— iLD cm^lo I I I I I I I I I III cMCJDCMCTiLnoococri vocMr-i 1— i.-HCMCMrotnO'— ^^Lo 1— lO"*"— I"— li— ''— ICMLDi— 1 I— lOOtOOCMtni— 1^1 — LDt-H O"— 'cooo'^^'d-LnLnLnO'— I I I I I I I I I I I I I OCTlOCMCTli— l O •— I CM t3CO I I I I I CO CO ^ O CO O -I CM ^ CO o o CM I o o o CM o o o o o o • • • • CM CO CM 1 CM CM o 1 o O o o O • • r-H CM CO CM CM CM 121

PAGE 133

122 ^ I — O CD Q. C 1/1 UJ ra +-> o (/) CO cn o T3 t— 1 CO «^ o C o 1 1 1 o IT) CM ai OJ t— t CO CO o CO o c o o 1/5 (U -o iC 3 fO +-> 3r l/l O) rc M O o c o o 1 o c E 1/1 O) 1/1 > o o SE CD 03 O I/) T3 o o 00 l/> +-> C t-H t— I o ro <— I 00 CM •— I CVI •— I 1— I o r-N. en t-H CO O CM CM CO LD ID 00 Lf) I CO LO CO CO I CO U3 en LO 1— I r~. cT> IT) O CM O in CM CO CO t— I CO «ao o o o o o o o tn o LO CM CM CM CM O 1 1 o o 1 o O o o o o o o CM CM O o

PAGE 134

APPENDIX B INTEROBSERVER RELIABILITY CHARTS

PAGE 135

INTEROBSERVER RELIABILITY CHARTS Intensive language teacher A Observer UlOSS Douy movement Hand gestures Spoken German opuiscll Engl ish R.C.J. 129 104 106 0 F.L.W. 133 1 1 L lUdi 0 A N 1 1 OA 1 Uf 129 145 Disagreements 29 25 42 0 Total of agreements plus disagreements 300 300 300 300 Percentage of agreement 90% 92% 86% 100% Intensive language teacher B Observer Gross body movement Hand gestures Spoken German Spoken English R.C.J. 159 139 147 0 F.L.W. 166 152 146 0 A.N.L. 135 147 142 0 Disagreements 31 13 5 0 Total of agreements plus disagreements 300 300 300 300 Percentage of agreement 90% 96% 98% 100% 124

PAGE 136

125 Traditional language teacher A Observer Gross body movement Hand gestures Spoken German Spoken Engl ish R.C.J. 25 30 69 189 F.L.W. 23 26 70 197 A.N.L. 13 23 62 187 Disagreements 12 7 8 10 Total of agreements plus disagreements 300 300 300 300 Percentage of agreement 96% 98% 97% 97% Traditional language teacher B Observer Gross body movement Hand gestures Spoken German Spoken Engl ish R.C.J. 31 91 55 210 F.L.W. 36 106 52 193 A.N.L. 32 91 45 206 Disagreements 5 15 10 17 Total of agreements plus disagreements 300 300 300 300 Percentage of agreement 98% 95% 97% 94%

PAGE 137

APPENDIX C SAMPLE PAGE FROM VIDEO BEHAVIORAL OBSERVATION CHART FOR ONE TUTOR

PAGE 138

> o +J O E E (/) OJ V) > o o SE CO o +-> 3 •»-> o T3 c o u 1T3 -(-> o 10 c o o 0} 00 O c o (J CM — < i-H CO Ln ^ V£) Lf) 00 "— I CM O CM t-H CM CO ^ O I I I I I I I CO o 00 "— < 00 CM o r-H ro n ^31— ir-~cMLf)OLnovoo^i*>-o<— iLDOvocTiLni^ oorH.— icMCMnn^^^ooO"— 1<— ii— «*coloctii— i<^oo^cy»unr^ojcT> oo-— ii— ir-icMCvJcooo«a-«=rLnooOi— i>-^cnocMCOCMOLni-— <<;JOi— iCMrO";3-LnoO'— ICMCMCO^LO vf I— I i-H cr> •— t cvj Lo O •— I •— ' o o o >— I CO ^ Lo o I I I I I I I o Lo cyi 00 o en 1X3 O O O CM ^ tn cM'^a-cM'ia-in .-hooi— icolocovd>* o I CO cn CO CM r-. •— I c-j CO un I I I I I 1— I Ln I— I CO CM t— 4 CM CO CO Ln CMI^CMi— IC7lU3<=J-0 oOf-o I I I I I I I I <— l'=^rHOO«=fCOOOVO ooi— ii-HCM^-vfir) o o O o o o .— 1 CM CO CM CM CM OJ 1 1 1 +-> o o o o o o C •ro r— 1 CM s: CM CM CM CM LO CM CM CM CM CTi O 00 CM Ln O o o CO CO «^ un o I I I I I I I O ^ CO
PAGE 139

O CO o O o O) ra +-> o hO o 00 t-H CO <— I CM (NJ CO ocMr^ocMOuncM.— I «;j-tnLnoocM— ir-» ro^Lnunoooococo PO C^J I— I ro CO vo to o c o u O) 00 o CO o c: o u ID C/7 CO CO o IX) 1—1 o CO o CM CM CO ^ ^ tn i I I I I I I o CO «do o o o CM CO CO 'aur) CM >-H CM CO "^a i-H CO 1—1 IT) CO r«-. cTi CO o ^ uo 1—1 CO ^ o o o o o o un o LD CM CM CM 1 1 o o O o o O o o O CM O O

PAGE 140

APPENDIX D STUDENT FOREIGN LANGUAGE DATA SHEET AND RATING FORM

PAGE 141

FLW Dissertation Name Course # Date STUDENT FOREIGN LANGUAGE DATA SHEET AND RATING FORM Previous Exposure to German 1. Have you previously taken a GERMAN course at UF? Yes No If yes, check spaces that apply: INTENSIVE GERllZO 1121 1122 1130 TRADITIONAL GER1120 1121 1122 2. Have you previously taken a GERMAN course at another school? Yes No If yes, indicate number of school years* in spaces that apply: *0ne school year is defined as two quarters or semesters at the postsecondary level or one Sept.June year at the K-12 level. Indicate parts of school years as 1/4, 1/2, etc. Elementary or Middle School yrs. H.S. yrs. College yrs. Military yrs. Peace Corps yrs. 3. Have you previously had experience with GERMAN outside of school? Yes No If yes, indicate number of months in spaces that apply: Heard GERMAN spoken by native speakers in Germany mos. Heard GERMAN spoken by native speakers outside of Germany mos. Spoke GERMAN in Germany mos. Spoke GERMAN outside of Germany mos. Student Ratings 4. Student's assessment of ability to use the German language: Performance Rating Scale (circle numbers that apply) Excellent Fair Poor Reading Comprehension 4 ^3 2 ^1 0 Listening Comprehension 4 3 2 ^1 0 Writing Proficiency 4 ^3 2 1 0 Speaking Proficiency 4 3 2 ^1 0 130

PAGE 142

131 5. Student's satisfaction with UF courses to increase German communication skills: Satisfaction Rating Scale (circle numbers that apply) High Moderate Low Current Course 4 ^3 2— 1 0 Previous INTENSIVE 4 ^3 2 1 0 Previous TRADITIONAL 4 ^3 2 ^1 0 Student Data: Male Female Reason for taking German: Required Not Required Comments :

PAGE 143

BIOGRAPHICAL SKETCH Florence L. Walters, n^e Reinhard, began life in New York City on May 12, 1923. She attended the city's public schools and graduated from Walton High School in 1940. On the basis of New York State Regents Examinations, she was admitted to an honors program at Hunter College where she pursued lower division courses in the arts and sciences to qualify for admission to the Philadelphia School of Occupational Therapy, affiliated with the University of Pennsylvania, Philadelphia. She was graduated with a diploma in occupational therapy in 1944. After passing the registration examination administered by the American Occupational Therapy Association, she served in two Army hospitals, Crile General at Cleveland and Walter Reed in Washington, D.C. Following her marriage to James H. Walters in 1946, she accompanied her husband to Gainesville, where he entered the University of Florida, graduating with degrees in electrical and industrial engineering. Between 1946 and 1958 they became the parents of four children. During this time, Florence began the first professional department at the Florida Farm Colony— later named Sunland Training Center at Gainesville— and from 1950 to 1967 administered the Department of Occupational Therapy at Sunland, except for brief leaves to take care of family responsibilities. During 1957-1958, Florence also served as consultant to Dr. Darrel Mase, the first dean of the College of Health Related Professions, who 132

PAGE 144

133 planned the Occupational Therapy Department at the J, Hill is Miller Health Center, University of Florida. After some part-time work, providing occupational therapy for patients on the psychiatric unit of Shands Teaching Hospital, she began the first official occupational therapy program at Alachua General Hospital in 1968. She left this growing program in 1973 to become a work evaluator for the Association for Retarded Citizens. In this capacity, she administered a program for the Department of Vocational Rehabilitation, contracting to provide counselors with comprehensive evaluations of their clients' capacity for work-related industries. This position led to a similar position at Exceptional Industries, a sheltered workshop for the handicapped in Gainesville. In 1978, after her four children had obtained higher education degrees, Florence entered the University of Florida to obtain a Bachelor of Health Science degree, a goal she accomplished in December of that year. A year later she obtained the degree of Master of Education in a program designed for health-related educators. She has since been engaged in meeting requirements for the degree of Doctor of Philosophy within the Division of Curriculum and Instruction, College of Education. Her future plans are to continue studying the sensory integration implications of nonverbal communication toward the goal of developing a coherent theory of learning on which educators can base their teaching.

PAGE 145

I certifiy that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. This dissertation was submitted to the Graduate Faculty of the Division of Curriculum and Instruction in the College of Education and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1982 Margaret 1^ Morgart,' Chairman] Professor of Instructional readership and Support ilames wV Hensel Professor of Instructional Leadership and Support Dora A. Hicks Professor Emeritus of Health Education and Safety Dean for Graduate Studies and Research