<%BANNER%>

Hierarchical Organization of Abstract Nouns

Permanent Link: http://ufdc.ufl.edu/UFE0044710/00001

Material Information

Title: Hierarchical Organization of Abstract Nouns Implications for Neurolinguistic Theory
Physical Description: 1 online resource (48 p.)
Language: english
Creator: Troche, Joshua E
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: abstract -- cognition -- emotion -- semantics
Speech, Language and Hearing Sciences -- Dissertations, Academic -- UF
Genre: Communication Sciences and Disorders thesis, M.A.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The organization and neural representation of concretewords has long been an intense area of interest in neurolinguistics.  Many theories stress the importance ofhierarchical organization of the lexical networks of concrete words (e.g.,labrador–dog-animal).  Hierarchicallexical organization maps well to the structure of the brain and provides acompelling account of the graceful degradation of naming seen in manyneurological disorders.  Very littleremains known about abstract words (e.g., truth); however, converging evidencesuggests that concrete and abstract words are unique in their neuralrepresentation.  One possibility is thatabstract words show a “loose” or non-hierarchical organization relative toconcrete words.  We investigatedclustering of 400 highly abstract and concrete words in multi-dimensionalspace. Using a 7-pt Likert scale, participants (N=365) rated each target wordon the following 12 dimensions: sensation, morality, ease-of-teaching,ease-of-modifying, action, thought, emotion, social interaction, time, space,quantity, polarity. Data reduction using factor analysis revealed three latentfactors, corresponding roughly to: concreteness, emotion/social cognition, andmagnitude.  We then plotted similaritiesin 3-dimensional space using hierarchical cluster analysis.  These analyses showed that abstract words docluster in hierarchies, but that these hierarchies are qualitatively distinctfrom concrete words.   At the most superordinate levels,emotion/social cognition are important grouping factors while at the mostsubordinate it is magnitude. These putative “clusters” encompass cognitivedimensions that are potentially represented in unique, distributed regions ofthe human brain (e.g., magnitude as parietally mediated, emotion as right hemisphereor amygdala mediated).  We discussimplications for theory of abstract and concrete word representation.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Joshua E Troche.
Thesis: Thesis (M.A.)--University of Florida, 2012.
Local: Adviser: Reilly, Jamie Joseph.
Local: Co-adviser: Edmonds, Lisa Anna.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044710:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044710/00001

Material Information

Title: Hierarchical Organization of Abstract Nouns Implications for Neurolinguistic Theory
Physical Description: 1 online resource (48 p.)
Language: english
Creator: Troche, Joshua E
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: abstract -- cognition -- emotion -- semantics
Speech, Language and Hearing Sciences -- Dissertations, Academic -- UF
Genre: Communication Sciences and Disorders thesis, M.A.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The organization and neural representation of concretewords has long been an intense area of interest in neurolinguistics.  Many theories stress the importance ofhierarchical organization of the lexical networks of concrete words (e.g.,labrador–dog-animal).  Hierarchicallexical organization maps well to the structure of the brain and provides acompelling account of the graceful degradation of naming seen in manyneurological disorders.  Very littleremains known about abstract words (e.g., truth); however, converging evidencesuggests that concrete and abstract words are unique in their neuralrepresentation.  One possibility is thatabstract words show a “loose” or non-hierarchical organization relative toconcrete words.  We investigatedclustering of 400 highly abstract and concrete words in multi-dimensionalspace. Using a 7-pt Likert scale, participants (N=365) rated each target wordon the following 12 dimensions: sensation, morality, ease-of-teaching,ease-of-modifying, action, thought, emotion, social interaction, time, space,quantity, polarity. Data reduction using factor analysis revealed three latentfactors, corresponding roughly to: concreteness, emotion/social cognition, andmagnitude.  We then plotted similaritiesin 3-dimensional space using hierarchical cluster analysis.  These analyses showed that abstract words docluster in hierarchies, but that these hierarchies are qualitatively distinctfrom concrete words.   At the most superordinate levels,emotion/social cognition are important grouping factors while at the mostsubordinate it is magnitude. These putative “clusters” encompass cognitivedimensions that are potentially represented in unique, distributed regions ofthe human brain (e.g., magnitude as parietally mediated, emotion as right hemisphereor amygdala mediated).  We discussimplications for theory of abstract and concrete word representation.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Joshua E Troche.
Thesis: Thesis (M.A.)--University of Florida, 2012.
Local: Adviser: Reilly, Jamie Joseph.
Local: Co-adviser: Edmonds, Lisa Anna.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044710:00001


This item has the following downloads:


Full Text

PAGE 1

1 HIERARCHICAL ORGANIZATION OF ABSTRACT NOUNS: IMPLICATIONS FOR NEUROLINGUISTIC THEORY By JOSHUA E. TROCHE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Joshua E. Troche

PAGE 3

3 To my wife, son, sister, and parents

PAGE 4

4 ACKNOWLEDGMENTS This has been a year full of extreme sadness and extreme joy. Thank s go out to everyone who has been there to lend support during the sad times but also those same people who have been there to celebrate the joyous times. Also I extend enormous thanks to my committee, Dr. Edmonds and Dr. Reilly. Your patience and guidance throughout this process have all owed me to grow immensely as a researcher and writer

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 ABSTRACT ................................ ................................ ................................ ..................... 8 CHAPTER 1 HIERARCHICAL ORGANIZATION OF ABSTRACT NOUNS ................................ 10 Differences between Abstract and Concrete ................................ ........................... 11 Theories of Abstract Structure and Organization ................................ .................... 12 2 METHOD ................................ ................................ ................................ ................ 15 Overview ................................ ................................ ................................ ................. 15 Participants ................................ ................................ ................................ ............. 15 Material and Procedure ................................ ................................ ........................... 15 Data Collection ................................ ................................ ................................ ....... 17 Data Analyses ................................ ................................ ................................ ......... 17 3 RESULTS ................................ ................................ ................................ ............... 19 Data Trimming ................................ ................................ ................................ ........ 19 Exploratory Factor Analysis ................................ ................................ .................... 19 Hierarchical Clust er Analysis ................................ ................................ .................. 20 Cluster Group Differences ................................ ................................ ...................... 20 4 DISCUSSION ................................ ................................ ................................ ......... 28 Gene ral Discussion ................................ ................................ ................................ 28 Conclusion ................................ ................................ ................................ .............. 31 APPENDIX A PARAMETER DESCRIPTION ................................ ................................ ................ 33 B LIKERT SCALE ANCHOR POINTS ................................ ................................ ........ 34 C CLUSTER MEMBERSHIP ................................ ................................ ...................... 35 LIST OF REFERENCES ................................ ................................ ............................... 45 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 48

PAGE 6

6 LIST OF TABLES Table page 3 1 Factor analysis/component matrix for dimensions ................................ .............. 22 3 2 Psycholinguistic and Factor Score Correlation Matrix ................................ ........ 22 3 3 Means and mean differences of clusters ................................ ............................ 27

PAGE 7

7 LIST OF FIGURES Figure page 3 1 Three dimensional representation of factor scores ................................ ............. 23 3 2 Hierarchical cluster analysis with cluster descriptions and numbers .................. 24 3 3 Left side of the hierarchical cluster analysis with cluster descriptions and numbers. ................................ ................................ ................................ ............ 25 3 4 R ight side of the hierarchical cluster analysis with cluster descriptions and numbers ................................ ................................ ................................ ............. 26

PAGE 8

8 Abstract of T hesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Re quirements for the Degree of Master of Arts HIERARCHICAL ORGANIZATION OF ABSTRACT NOUNS: IMPLICATIONS FOR NEUROLINGUISTIC THEORY By Joshua E. Troche August 2012 Chair: Jamie Reilly Co C hair: Lisa Edmonds Major: Communication Sciences and Disorders The organization and neural representation of concrete words has long been an intense area of interest in neurolinguistics. Many theories stress the importance of hierarchical organization of the lexical networks of concrete words (e.g., labrador dog animal) Hierarchical lexical organization maps well to the structure of the brain and provides a compelling account of the graceful degradation of naming seen in many neurological disorders. Very little remains known about abstract words (e.g., truth); however converging evidence suggests that concrete and abstract words are unique in hierarchical organization relative to concrete words. We investigated clustering of 40 0 highly abstract and concrete words in multi dimensional space. Using a 7 pt Likert scale, participants (N=365) rated each target word on the following 12 dimensions: sensation, morality, ease of teaching, ease of modifying, action, thought, emotion, soci al interaction, time, space, quantity, polarity. Data reduction using factor analysis revealed three latent factors, corresponding roughly to: concreteness, emotion/social cognition, and magnitude. We then plotted similarities in 3 dimensional space usin g

PAGE 9

9 hierarchical cluster analysis. These analyses showed that abstract words do cluster in hierarchies, but that these hierarchies are qualitatively distinct from concrete words. At the most superordinate levels, emotion/social cognition are important gr ouping encompass cognitive dimensions that are potentially represented in unique distributed regions of the human brain (e.g., magnitude as parietally mediated, emotion as ri ght hemisphere or amygdala mediated). We discuss implications for theory of abstract and concrete word representation.

PAGE 10

10 CHAPTER 1 HIERARCHICAL ORGANIZATION OF ABSTRACT NOUNS Much is known about the structure and organization of concrete concepts (e.g., dog, desk, cup) and their representation in the human brain. However, there remain fundamental gaps in our understanding of how abstract words such as love, truth, and happiness are represented. These gaps in theory have produced an incomplete picture of l anguage and cognition. Abstract concepts comprise a significant portion of adult language. Therefore, a more comprehensive understanding of the underlying structure of abstract words is necessary before we can truly understand the nature of language in a holistic manner. One of the fundamental principles of neuroscience is that the brain is optimized for hierarchical processing. Although most research on hierarchical processing has focused on sensation and perception (e.g., cortical visual processing), th ere exists a wide body of literature to support the claim that some elements of language (i.e., word meaning) are also organized hierarchically. Rosch and colleagues (1981; 1975; 1976) performed pioneering work in t he domain of hierarchical organization of concrete concepts, postulating that objects are naturally categorized into basic level categories. Objects in these basic level categories share many common attributes, similar motor programs, and similar shapes. T hese basic level categories combine (superordinate categories) and diverge (subordinate categories) creating a hierarchical organization of concrete concept knowledge. Whereas hierarchical organization of concrete concepts is widely accepted, it remains u nclear whether abstract concepts display this same hierarchical organization. An understanding of this hierarchical organization in abstract concepts, if present, would represent a significant step forward in determining a complete theory of

PAGE 11

11 abstract conce pt structure. Therefore, our aim in this study was to determine whether abstract concepts displayed a hierarchical organization. Differences between Abstract and Concrete Theories of the organizational structure of abstract concepts are motivated by a vari ety of differences between concrete and abstract concepts. Concrete concepts have been operationally defined as being spatiotemporally bound to a definite time and place in space (Hale, 1988) In contrast, abstract concepts are operationally defined by the absence of spatiotemporal binding, existing without ties to a specific time or place in space (Hale, 1988) The differences between abstract and concrete con cepts extend past their operational definition as they have been found to posses differences both in their psycholinguistic attributes as well as the richness of their association in the brain. Psycholinguistically, abstract concepts have a greater incide nce of polysemy (i.e., multiple meanings that are semantically related) and homonymy (i.e., multiple meanings that are not semantically related ; Crutch & Jackson, 2011) The i ncreased incidence of polysemy weakens the relationship between word and concept due to the semantically related meanings of a word occupying a similar semantic space. Abstract concepts, therefore, are considered to be more difficult to learn than concrete concepts. This finding has received converging support from developmental studies of learning in young children and also in retrospective analyses of age of acquisition (Gilhooly & Logie, 1980; Reed & Dick, 1968) A consistent finding in both domains is that abstract words tend to be acquired much later than concrete words (Bonner, et al., 2009) Although psycholinguistic differences reveal some differences in the structure and organization of abstract and concrete concepts, a phenomenon k nown as the word concreteness effect has unveiled additional evidence for differences in the structure and

PAGE 12

12 organization of concrete and abstract concepts. The word concreteness effect is a term describing the collective advantage for concrete over abstract words in a variety of domains, including recall accuracy (Walker & Hulme, 1999) age of acquisition (Gilhooly & Logie, 1980) word list memory (Allen & Hulme, 2006) naming latency (Bleasdale, 1987) word recognition (Schwanenflugel, Harnishfeger, & Stowe, 1988) and resilience to neurological injury (Franklin, Howard, & Patterson, 1995; Katz & Goodglass, 1990; Martin & Saffran, 1 992; Roeltgen, Sevush, & Heilman, 1983; Warrington, 1975) Theories of Abstract Structure and Organization Many theories have emerged in an attempt to explain the word concreteness effect. There are two theories, however, that have been particularly infl uential, the Context Availability Model (Schwanenflugel, et al., 1988; Schwanenflugel & Shoben, 1983) and the Dual Coding Theory (Paivio, 1991) The Context Availability Model proposes the dissociation bet ween abstract and concrete concepts is due to differences in verbal context. In the Context Availability Model, verbal context can be understood as information (supplied by discourse or the between concepts. Increased association is postulated to lead to a richer representation of the concept in the brain (Schwanenflugel, et al., 1988; Schwanenflugel & Shoben, 1983) Concrete concepts, therefore, have greater availability to contextual information due to greater availability of contextual information. Stronger availability of contextual information for concrete concepts can be illustrated by comparing a concrete concept such as dog with an abstract co ncept such as magnitude. While associated information for a concept such as dog is readily available, the associated information for a concept such as magnitude is less available.

PAGE 13

13 The Dual Coding Theory postulates that differences in processing between con crete and abstract concepts arise from the presence of both linguistic information and experiential (i.e., visual and verbal codes) information, whereas abstract concepts are only supported by linguistic information (Paivio, 1991) The Dual Codin g Theory predicts that the concreteness effect is the result of dual, redundant representation for concrete words in two parallel semantic systems (i.e., visual and verbal). Dual Coding Theory has consequently been described as representing a class of mul tiple semantics theories, premised on the idea that conceptual knowledge is supported by dissociable semantic memory subsystems. Both the Dual Coding Theory and Context Availability Model postulate that inherent differences between concrete and abstract co ncepts have led to differences in the structure and organization of concrete and abstract concepts. A more recent theory on the structure and organization of semantic memory by Kousta and collegues (2011; 2009) arg ues that all concepts, whether concrete or abstract display an embodied structure and organization. The word concreteness effect has often been postulated to be incongruent with an embodied structure and organization for abstract concepts. Kousta and colle agues ( 2011) however, were able to eliminate the word concreteness effect by controlling for context availability and imageability. They explain that unlike concrete concepts, which are mostly supported by sensorimotor information, abstract concepts are mo stly supported by affective information. Although the theory by Kousta and colleagues contributes an important piece to our overall understanding of abstract concept structure and organization, there are still important questions that remain unanswered. I t is still wholly unclear whether abstract

PAGE 14

14 concepts display a hierarchical organization as is seen in concrete concepts. For this question to be answered, however, it was necessary to determine the features of abstract concepts as the features of a concept are an important aid in their organization. As previously discussed, sensorimotor information has been found to play a role in the structure and organization of concrete concepts (Paivio, 1991; Schwanenflugel, et al., 1988) We hypothesized, however, that generating features for abstract concepts, due to their low imageability, would likely involve a broader set of cognitive domains such as affective information. In this study, we attempted a novel approach to analyzing the features of abstract words that considered this broader set of cognitive domains. With this expanded set of domains and featur es we then attempted to determine if abstract concepts displayed a hierarchical organization as is seen in concrete words. Our hypothesis was that abstract concepts would display a hierarchical organization as is seen in concrete concepts.

PAGE 15

15 CHAPTER 2 MET HOD Overview In the experiment to follow we examined Likert scale ratings for concrete and abstract nouns across a variety of cognitive constructs (e.g., emotional valence, social cognition). Our aim was to evaluate the presence of latent factors that orga nize the lexical networks of abstract and concrete words. We mathematically modeled the clustering of latent variables using standard data reduction procedures (i.e., factor analysis) with hierarchical cluster analysis. Participants Participants were recr uited through the online crowd sourcing program Mechanical Turk (Amazon Incorporated, 2012) Participants were nominally compensated for complet ing anywhere from 1 to all 12 surveys. Inclusion criteria required participants to be native English speaker s residing in the United States Participants assented informed consent in accord with the experimental protocol (IRB 02). Age ranged between 17 and 83 years of age with an average age of 40.77 and a level of education that ranged from 9 to 20 years with an average education of 15.35 years. Two hundred forty nine surveys were completed by females (68.2%) and 116 surveys were completed by males (31.8%). Material s and Procedure Four hundred nouns were c hosen from the Medical Research Council (MRC) Psycholinguistic Database (Coltheart, 1981) to be included in this study. Ratings of imageability, age of acquisition (AOA), concreteness, and familiarity were accumulated

PAGE 16

16 from the MRC database f or all 400 nouns while frequency ratings were gathered from the Brysbaert and New ( 2009) database These psycholinguistic variables were input into our analys es as dependent variables. Half of the nouns chosen were classified as concrete and the other half as abstract. Nouns were classified as concrete when imageability scores were above 500. Nouns were classified as abstract when imageability scores were below 450. The average imageability rating for concrete nouns was 611.42 while the average imageability rating for abstract nouns was 303.74. A broad set of cognitive domains were considered in order to determine the features of abstract concepts They are as follows: emotion (Vuilleumier, Armony, Dri ver, & Dolan, 2003) social cognition (Amodio & Frith, 2006) executive function (Stuss, Shallice, Alexander, & Picton, 1995) time (Walsh, 2003) and episodic memory (Moscovitch, Nadel, Winocur, Gilboa, & Rosenbaum, 2006; Squire, 1982) spatial cognition (Burgess, Maguire, & O'Keefe, 2002) and magnitude (Walsh, 2003) From these broad cognitive do mains we creat ed 12 finer grained parameters: sensation, morality, ease of teaching, ease of modifying, action, thought, emotion, social interaction, time, space, quantity, and polarity. A description of these parameters as presented to participants can be found in Appendix A. A survey was created for each of the 12 parameters (12 surveys). Each survey included all 400 nouns in a pseudorandomized order. Participants were asked to make judgments using a seven point Likert scale on whether they agreed or disa greed with a statement regarding the parameter in question (see Appendix B for anchor points). Participants were instructed to use the entire scale and to work quickly but not carelessly.

PAGE 17

17 Data Collection As stated earlier, participants were recruited throu gh the online program Mechanical Turk. Mechanical Turk represents an online pool of workers from around the globe who perform virtual tasks. Mechanical Turk has been utilized extensively in recent human factors research studies and has been formally evalua ted for validity and reliability (Buhrmester, Kwang, & Gosling, 2011) Utilizing this online survey tool allow ed us to increase the external validity and inferential power of our task through diverse and rapid sampling. In order to complete our task the participant had to complete an online survey created within Survey Monkey (SurveyMonkey.com LLC 2011) Data Analyses We excluded participant data that corresponded to one or more of the following conditions : 1) Taking less than 600 seconds (i.e. 10 minutes) to complete the survey, 2) Using less than ha lf of the seven point likert scale (i.e. 3 numbers or less), and 3) More than 20 of the same response in a row. A factor analysis was then performed in the Statistical Package for the Social Sciences (SPSS) version 18, in order to determine potential laten t factors across the 12 parameters chosen. Following the identification of latent factors the factor scores, which are a standard score output of the factor analysis, were correlated with the psycholinguistic variables of imageability, concreteness, famil iarity, AOA, and frequency. A hierarchical agglomerative cluster analysis was then performed in order to determine the cluster properties of the 400 nouns. Agglomerative clustering involves grouping observations with a bottom up approach until all observat ions become one

PAGE 18

18 efficient and resistant to outliers (Ward, 1963) In order to determine distance between observations, the Euclidean squared coefficient was used as it is the geometric distance in the multidimensional space. A method outlined by Aldenderfer and Bashfield (1984) was then implemented to determine what number cluster solution would be appropriate for the data. The method involved comparing cluster solutions with partitional k means solutions. The two determined, t tests were completed in order to determine differences between clusters for the chosen psycholinguisti c variables and the factor scores of the latent factors

PAGE 19

19 CHAPTER 3 RESULTS Data Trimming The investigator and a rater blind to the purpose of the study displayed a 99.3% inter rater agreement on surveys to be excluded. Of the original 545 surveys 180 (33 %) were eliminated leaving 365 surveys for final analysis. Exploratory Factor Analysis We conducted an Exploratory Factor Analysis using a principal components (PCA) method of extraction and three latent f actors were extracted based on their Eigenvalues ( model fit, R 2 =.81) We reduced the original set of 12 factors via a Var imax rotation method with Kaiser Normalization This procedure yielded a new, lower dimensionality space representing linear combinations of groups of factors (approximating latent vari ables). We saved these factor loadings as new variables (see Table 3 1). Factors agglomerated as follows: 1) Emotion, Polarity Social, Morality, Action, Thought ; 2) Ease of Teaching, Sensation, Ease of Modifying, Time ; 3) Space, Quantity. Factor scores which are a z score derivative of the variance, were computed by SPSS using the Anderson Rubin (1956) method. This method was chosen due to its creation of uncorrelated factor scores. The three new factors were analyzed in order to determine with which psycholinguistic factors they correlated (Table 3 2) Factor 1 was significantly correlated with concreteness ( .557), imageability ( .361), frequency (.255), AOA (.227), and familiarity (.163). Factor 2 was significantly correlated with AOA ( .844), concret eness (.784), imageability (.725), familiarity (.506), frequency (.401).

PAGE 20

20 Factor 3 was significantly correlated only with familiarity (.237). Each of the three new latent factors was given a title, which suggested a common theme displayed by the parameter m embers. The first factor, which contained the parameters of emotion, polarity social construct, morality, action and thought, was a factor representing affective association and social cognition. The second factor, which contained the parameters ease of t eaching, ease of modifying, sensation, and time, was highly correlated with AOA, concreteness, and imageability and was deemed a factor similar to concreteness. The third factor, which contained the space and quantity parameters, was not found to be highly correlated with any psycholinguistic variable and was de duced to be a factor of magnitude. Figure 3 1 displays the factor scores for each word represented in a three dimensional scatter plot. Concrete nouns are represented by orange circles and abstr act n Hierarchical Cluster Analysis Hierarchical cluster analysis revealed a possible 10, 11, 12 or 13 cluster solution. In order to d etermine whether this 10, 11, 12 or 13 cluster solution was a good fit for the data, the clust ers from the hierarchical cluster analysis were compared to the clusters created by a partitional k means it appa (Aldenderfer & Blashfield, 1984) T (.87) and was therefore considered the best solution. Cluster Group Differences The hierarchical cluster analysis revealed that the 12 clust ers combined upwards 5 levels until they merge d into one group (see Figure 3 2, 3 3, & 3 4). The membership of each cluster is presented in Appendix C T tests were performed to compare differences between neighboring clusters for psycholinguistic variable s and factor scores It should

PAGE 21

21 be noted that since the factor scores are z scores and come from data measured on the same scale the mean difference between the t wo clusters is synonymous with the C Table 3 3 displays the individual means of the psycholinguistic variables and factor scores for each cluster, while also displaying the mean difference for the closest cluster neighbor. Significant mean differences as indicated by the t tests are represented by asterisks. Ea ch cluster was then given a description and cluster number as seen in Figures 3 2, 3 3, and 3 4. The description of each cluster represents a common theme seen in the word members of each cluster while the cluster number represent the hierarchical level, from top to bottom, and order, from left to right. For instance, cluster C2.3 is a cluster in the second level and is third from the left.

PAGE 22

22 Table 3 1. Factor analysis/c omponent matrix for dimensions Component Predictor Factor 1 Factor 2 Factor 3 Emot ion .905 .229 .027 Positive/Negative .880 .115 .235 Social .855 .280 .090 Morality .794 .479 .057 Action .722 .517 .169 Thought .719 .594 .094 Ease of Teaching .376 .880 .040 Sensation .447 .846 .026 Ease of Modifying .104 .736 .310 Time .350 .685 .319 Space .006 .208 .846 Quantity .273 .412 .691 Note: The above component matrix was derived using SPSS Varimax rotation with Kaiser normalization. The rotation converged after five iterations. Table 3 2. Psycholinguistic and Factor Score Correlation Matrix Imag AOA Frqy CNC Fam Emo Cnc/Tch Mag Imag 1 AOA .86* 1 Frqy .22* .44* 1 CNC .94* .85* .21* 1 Fam .29* .56* .44* .27* 1 Emo .36* .23* .26* .56* .16* 1 Cnc/Tch .73* .84* .40* .78* .51* 0 1 Mag .01 .03 .09 .06 .24* 0 0 1 Note: Imag=Imageability, AOA=Age of Acquisition, Frqy=Frequency, CNC=concreteness, Fam=Familiarity, Emo=Emotion/Social Cognition, Cnc/Tch=Concreteness /Ease of Teaching, Mag=Magnitude, *=p<.01

PAGE 23

23 Figure 3 1. Three dimensional representation of factor scores. Blue s represent abstract nouns while orange circles represent concrete nouns. CNC= concreteness.

PAGE 24

24 Figure 3 2. Hierarchical cluster analysis with c luste r descriptions and n umbers

PAGE 25

25 Figure 3 3 Left side of the h ierarchical c luster a nalysis with c luster d escriptions and n umbers

PAGE 26

26 Figure 3 4. Right side of the hierarchical cluster analysis with cluster descriptions and numbers

PAGE 27

27 Table 3 3 Means and mean d ifferences of c lusters C1.1 C1.2 MD C2.1 C2.2 MD C2.3 C2.4 MD C3.1 C3.2 MD Imag 600.12 349.54 250.58* 598.92 601.87 2.95 300.18 392.05 91.87* 595.54 618.53 22.99 AOA 279.68 461.80 182.12* 274.19 286.06 11.87 501.44 425.47 75.97* 280.07 249.86 30.21 Frqy 33.49 37.68 4.19 19.80 53.54 33.74* 7.72 64.84 57.12* 14.05 53.98 39.93* CNC 594.72 337.55 257.17* 599.72 587.46 12.26* 308.19 361.11 52.92* 598.67 605.60 6.93 Fam 536.10 518.45 17.65* 526.10 550.58 24.48* 492.15 540.3 2 48.17* 519.33 564.40 45.07* Emo 0.74 0.58 1.32 0.78 0.72 0.06 0.09 1.18 1.27 0.86 0.15 0.71 Cnc/Tch 0.69 0.54 1.23 0.64 0.76 0.12 1.21 0.07 1.28 0.54 1.25 0.71 Mag 0.01 0.01 0.02 0.81 1.78 2.59 0.20 0.19 0.39 0.85 0.55 0.30 C3.3 C3.4 MD C3.5 C3.6 MD C3.7 C3.8 MD C4.1 C4.2 MD Imag 587.05 608.22 21.17 340.81 283.56 57.25* 563.09 345.76 217.33* 240.38 289.52 49.14 AOA 311.50 270.00 41.5 500.53 501.92 1.39 221.00 466.36 245.36* 617.00 495.15 121.85* Frq y 38.85 59.71 20.86 6.14 8.49 2.35 223.05 26.53 196.52* 1.48 9.37 7.89 CNC 585.65 588.21 2.56 326.73 299.42 27.31* 574.48 305.10 269.38* 285.67 301.10 15.43 Fam 534.70 557.06 22.36 485.19 495.28 10.09 584.95 528.60 56.35* 424.43 505.00 80.57* E mo 0.88 0.65 0.23 0.80 0.25 1.05 1.23 1.17 0.06 0.34 0.33 0.67 Cnc/Tch 0.68 0.80 0.12 1.59 1.02 0.57 1.69 0.32 2.01 1.59 0.95 0.64 Mag 2.55 0.60 1.95 0.54 0.56 1.10 0.30 0.31 0.61 1.48 0.45 1.03 C4.3 C4.4 MD C5.1 C5.2 MD C5.3 C5.4 MD Imag 380.10 327.04 53.06* 370.13 383.73 13.60 362.67 309.70 52.97* AOA 474.94 461.94 13 512.50 463.38 49.12 442.67 478.00 35.33 Frqy 19.70 30.35 10.65 7.40 23.50 16.10 45.46 22.97 22.49 CNC 294.80 311.28 16. 48 272.25 303.00 30.75 312.53 310.64 1.89 Fam 512.30 538.38 26.08* 465.63 529.27 63.64* 557.59 528.48 29.11* Emo 1.59 0.94 0.65 1.34 1.67 0.33 1.62 0.60 1.02 Cnc/Tch 0.12 0.43 0.31 0.45 0.02 0.43 0.05 0.67 0.72 Mag 0.50 0. 76 1.26 1.60 0.17 1.43 0.99 0.65 0.34 Note: Imag=Imageability, AOA=Age of Acquisition, Frqy=Frequency, CNC=concreteness, Fam=Familiarity, Emo=Emotion/Social Cogni tion Cnc/Tch=Concreteness/Ease of Teaching, Mag=Magnitude, MD=Mean Difference, *=p<. 05

PAGE 28

28 CHAPTER 4 DISCUSSION General Discussion Much remains to be learned about the linguistic and neurological representation of abstract words. Advances in this domain can be aided by a vast body of research on the hierarchical representation of concrete words. We investigated clustering and conceptual features of abstract words. In order to determine if abstract concepts displayed a hierarchical organization it was necessary to determine the features of abstract concepts. Concrete concepts have salient s ensorimotor features which aid in their hierarchical organization while abstract concepts do not. It was, therefore, necessary to expand our search beyond sensorimotor features to include a range of alternate cognitive domains. The 12 parameters we chose, which represented several broad cognitive domains, were found to represent three distinct latent factors: emotion/social cognition, concreteness, and magnitude. These new factors were then input into our model to determine if abstract concepts displayed a hierarchical organization. It was revealed that abstract concepts do display a hierarchical organization. A level by level analysis of the hierarchy revealed several findings. At the most superordinate level the nouns separated into two clusters (C1.1 & C1 .2). Essentially these two clusters were concrete nouns (C1.1) and abstract nouns (C1.2). There were also comparable differences in the level of emotion/social cognition of the clusters with nouns low in emotion/social cognition being represented in C1.1 a nd nouns moderate to high in emotion/social cognition being represented in C1.2. While it is unsurprising that

PAGE 29

29 the nouns would cluster based on concreteness, it is unexpected that nouns would cluster relative to emotion/social cognition at the most superor dinate level. The role of emotional valence in the organization of concepts is a fairly recent development. Altarriba and colleagues (1999) have argued that highly emotiona lly valent concepts should occupy their own place in the semantic space separate from concrete and abstract concepts. More recently Kousta and colleagues (2011) have argued for an embodied theory of semantic memory, that unlike other embodied theories inclu des abstract concepts. Unlike Altarriba and colleagues, Kousta and colleagues argue that highly emotionally valent words should not occupy their own space, but that valence is an important factor in the organization of abstract concepts. Our findings suppo cognition factor was found to be an organizing factor at the superordinate levels of the hierarchy. In addition, our findings suggest that valence not only aids in the organization of abstract conc epts but also provides an organizational structure to concrete concepts. In Figure 3 4, take note of Cluster C3.7, which is comprised of predominately concrete nouns but is nested on the abstractly dense side of the hierarchy. This cluster, however, unlike the other clusters of concrete nouns, contains concrete nouns high in emotion/ social cognition (e.g., father and chocolate). According to this hierarchy, therefore, this cluster of nouns aligns closer to abstract concepts such as cowardice and freedom th an other concrete concepts such as fisherman and corn. One of the most novel findings of this study came from examination of the magnitude factor. To date, no studies have examined a magnitude factor as a possible factor of importance in the organization o f concepts. Our study revealed, however, that

PAGE 30

30 the magnitude factor does have a role in the organization of concepts at the most subordinate levels of the hierarchy. It is important to note that this factor was not found to correlate with any of the psycho linguistic variables including imageability and concreteness. These findings demonstrate that the magnitude factor can be differentiated from concreteness ; as i t could be assumed that as a factor of orientation and size, this factor would be related to con creteness. The lack of correlation with concreteness may be due to space and quantity being part of what Walsh (2003) theorized as a distinct magnitude system in the bra in. Walsh, however, grouped time in with space and quantity, which was not supported by our data. Some models of the factor analysis did place the time factor with the space and quantity factor, but these models were not the best fit model and were therefo re not applied. The fact that time was not grouped with space and quantity may have been due to our attempts to simplify parameters in order to facilitate understanding in participants. A number of caveats should be noted when considering the findings in t h is current study. One concern relates to the specificity of the domains being tested. The rating deliberately broad and non technical in order to ensure that participants ha d a clear understanding of and confidence in the judgments they were being asked to make. Using such lay terminology inevitably leads to uncontrolled variation in rating specificity. nced by many different social cognitive processes such as mental state attribution or empathy A second but related point concerns the number of rating variables selected. The list of broad cognitive domains selected was by no means exhaustive, and some do mains

PAGE 31

31 variable could have been subdivided into separate ratings for vision, audition, gustation, olfaction and different types of somatosensory information. Such a division mi ght provide much richer feature ratings particularly for concrete items, but these modalities were combined so as to limit the number of different ratings requested from each participant, and because the focus of the study was upon the conceptual structure of abstract words and the contribution of non sensorimotor domains. Conclusion Our findings d emonstrate that abstract concepts, like concrete concepts, exhibit a hierarchical organization. We also confirmed the important role of emotion/social cognition i n the organization of concepts. In addition, the distinct domain of magnitude was discovered to serve a purpose in the organization of concepts. These findings display a departure from widely held theories of abstract concepts such as the dual coding theor y and context availability model. These theories contend that abstract and concrete concepts are qualitatively different It is also possible to infer t hat these theories would suggest that abstract concepts would not display a hierarchical organization O ur findings, however, suggest that differences between abstract and concrete concepts may be more indicative of the types of information that support the concept (e.g. valence, magnitude) rather than the strength or availability of information as suggested in previous theories. These findings also suggest that the structure and organization of concrete concepts and abstract concepts may display more similarities than differences, which is a separation from previous thought on abstract concept structure. Our results also suggest the possibility of new neuroanatomical correlates for abstract concepts although our study cannot directly speak to this assertion.

PAGE 32

32 Overall we contend that these findings represent a new approach to the understanding of abstract co ncepts. Although these findings are a promising foundation for a new theory of abstract concept structure and organization, there remain too many unanswered questions to be able to consider the theory complete. As previously stated, the list of cognitive d omains tested was not exhaustive, and further research is necessary to determine the presence of additional factors that aid in the organization of abstract concepts. While our findings support the presence of a hierarchical organization of abstract concep ts, questions remain as to whether abstract concepts also display an embodied organization. Future research will investigate abstract concept structure through the effects of priming in both healthy and diseased populations as well as exploring lesion mode and corticobasal degeneration (CBD; visual spatial difficulties).

PAGE 33

33 APPENDIX A PARAMETER DESCRIPTION Parameter Definition Polarity I relate this word to positive or negative feelings in myself Sensation I relate this word to physical feelings like vision, hearing, smelling, etc. Action I relate this word to actions, doing, performing and influencing. Thought I relate this word to mental activity, ideas, opinions, and judgments. Emotion I r elate this word with human emotion. Social interaction I relate this word with relationships between people. Time I relate this word with time, order, or duration. Space I relate this word to position, place or direction. Quantity I relate this word to size, amount or scope. Morality I relate this word to morality, rules or any thing that governs my behavior. Ease of modifying I can easily choose an adjective for this word (the ugly truth, whole truth, etc.) Ease of teaching/learning This word could be easily taught to a person who does not speak English.

PAGE 34

34 APPENDIX B LIKERT SCALE ANCHOR POINTS

PAGE 35

35 APPENDIX C CLUSTER MEMBERSHIP Word 1st Level 2nd Level 3rd Level 4th Level 5th Level Abundance C1.2 C2.4 C3.8 C4.4 C5.4 Ambition C1.2 C2.4 C3.8 C4.4 C5. 4 Arrangement C1.2 C2.4 C3.8 C4.4 C5.4 Attention C1.2 C2.4 C3.8 C4.4 C5.4 Awareness C1.2 C2.4 C3.8 C4.4 C5.4 Comparison C1.2 C2.4 C3.8 C4.4 C5.4 Competence C1.2 C2.4 C3.8 C4.4 C5.4 Comprehension C1.2 C2.4 C3.8 C4.4 C5.4 Conclusion C1.2 C2.4 C3.8 C4. 4 C5.4 Consequence C1.2 C2.4 C3.8 C4.4 C5.4 Corporation C1.2 C2.4 C3.8 C4.4 C5.4 Debt C1.2 C2.4 C3.8 C4.4 C5.4 Decision C1.2 C2.4 C3.8 C4.4 C5.4 Democracy C1.2 C2.4 C3.8 C4.4 C5.4 Determination C1.2 C2.4 C3.8 C4.4 C5.4 Disaster C1.2 C2.4 C3.8 C4.4 C 5.4 Economy C1.2 C2.4 C3.8 C4.4 C5.4 Event C1.2 C2.4 C3.8 C4.4 C5.4 Evidence C1.2 C2.4 C3.8 C4.4 C5.4 Fact C1.2 C2.4 C3.8 C4.4 C5.4 Importance C1.2 C2.4 C3.8 C4.4 C5.4 Incentive C1.2 C2.4 C3.8 C4.4 C5.4 Intention C1.2 C2.4 C3.8 C4.4 C5.4 Introducti on C1.2 C2.4 C3.8 C4.4 C5.4 Logic C1.2 C2.4 C3.8 C4.4 C5.4 Miracle C1.2 C2.4 C3.8 C4.4 C5.4 Moment C1.2 C2.4 C3.8 C4.4 C5.4 Occupation C1.2 C2.4 C3.8 C4.4 C5.4 Outcome C1.2 C2.4 C3.8 C4.4 C5.4 Ownership C1.2 C2.4 C3.8 C4.4 C5.4 Perception C1.2 C2.4 C3.8 C4.4 C5.4 Possibility C1.2 C2.4 C3.8 C4.4 C5.4 Prediction C1.2 C2.4 C3.8 C4.4 C5.4 Presence C1.2 C2.4 C3.8 C4.4 C5.4 Production C1.2 C2.4 C3.8 C4.4 C5.4 Promotion C1.2 C2.4 C3.8 C4.4 C5.4 Proof C1.2 C2.4 C3.8 C4.4 C5.4 Quality C1.2 C2.4 C3.8 C4 .4 C5.4 Reality C1.2 C2.4 C3.8 C4.4 C5.4

PAGE 36

36 Significance C1.2 C2.4 C3.8 C4.4 C5.4 Vocabulary C1.2 C2.4 C3.8 C4.4 C5.4 Ability C1.2 C2.4 C3.8 C4.4 C5.3 Accomplishment C1.2 C2.4 C3.8 C4.4 C5.3 Belief C1.2 C2.4 C3.8 C4.4 C5.3 Brain C1.2 C2.4 C3.8 C4.4 C5. 3 Faith C1.2 C2.4 C3.8 C4.4 C5.3 Freedom C1.2 C2.4 C3.8 C4.4 C5.3 Holiday C1.2 C2.4 C3.8 C4.4 C5.3 Idea C1.2 C2.4 C3.8 C4.4 C5.3 Improvement C1.2 C2.4 C3.8 C4.4 C5.3 Information C1.2 C2.4 C3.8 C4.4 C5.3 Intelligence C1.2 C2.4 C3.8 C4.4 C5.3 Justice C1.2 C2.4 C3.8 C4.4 C5.3 Knowledge C1.2 C2.4 C3.8 C4.4 C5.3 Leadership C1.2 C2.4 C3.8 C4.4 C5.3 Memory C1.2 C2.4 C3.8 C4.4 C5.3 Opportunity C1.2 C2.4 C3.8 C4.4 C5.3 Reputation C1.2 C2.4 C3.8 C4.4 C5.3 Responsibility C1.2 C2.4 C3.8 C4.4 C5.3 Skill C 1.2 C2.4 C3.8 C4.4 C5.3 Wisdom C1.2 C2.4 C3.8 C4.4 C5.3 Admiration C1.2 C2.4 C3.8 C4.3 C5.2 Attitude C1.2 C2.4 C3.8 C4.3 C5.2 Behavior C1.2 C2.4 C3.8 C4.3 C5.2 Benefactor C1.2 C2.4 C3.8 C4.3 C5.2 Character C1.2 C2.4 C3.8 C4.3 C5.2 Charity C1.2 C2.4 C3.8 C4.3 C5.2 Confidence C1.2 C2.4 C3.8 C4.3 C5.2 Crisis C1.2 C2.4 C3.8 C4.3 C5.2 Danger C1.2 C2.4 C3.8 C4.3 C5.2 Duty C1.2 C2.4 C3.8 C4.3 C5.2 Equality C1.2 C2.4 C3.8 C4.3 C5.2 Expression C1.2 C2.4 C3.8 C4.3 C5.2 Fantasy C1.2 C2.4 C3.8 C4.3 C5.2 Honesty C1.2 C2.4 C3.8 C4.3 C5.2 Identity C1.2 C2.4 C3.8 C4.3 C5.2 Independence C1.2 C2.4 C3.8 C4.3 C5.2 Insight C1.2 C2.4 C3.8 C4.3 C5.2 Instinct C1.2 C2.4 C3.8 C4.3 C5.2 Integrity C1.2 C2.4 C3.8 C4.3 C5.2 Interaction C1.2 C2.4 C3.8 C4.3 C5.2 Kindn ess C1.2 C2.4 C3.8 C4.3 C5.2

PAGE 37

37 Opinion C1.2 C2.4 C3.8 C4.3 C5.2 Purpose C1.2 C2.4 C3.8 C4.3 C5.2 Satisfaction C1.2 C2.4 C3.8 C4.3 C5.2 Tradition C1.2 C2.4 C3.8 C4.3 C5.2 Truth C1.2 C2.4 C3.8 C4.3 C5.2 Cowardice C1.2 C2.4 C3.8 C4.3 C5.1 Criticism C1.2 C2.4 C3.8 C4.3 C5.1 Deceit C1.2 C2.4 C3.8 C4.3 C5.1 Disagreement C1.2 C2.4 C3.8 C4.3 C5.1 Hatred C1.2 C2.4 C3.8 C4.3 C5.1 Malice C1.2 C2.4 C3.8 C4.3 C5.1 Mercy C1.2 C2.4 C3.8 C4.3 C5.1 Revenge C1.2 C2.4 C3.8 C4.3 C5.1 Artist C1.2 C2.4 C3.7 Baby C 1.2 C2.4 C3.7 Body C1.2 C2.4 C3.7 Boy C1.2 C2.4 C3.7 Cat C1.2 C2.4 C3.7 Child C1.2 C2.4 C3.7 Chocolate C1.2 C2.4 C3.7 Christmas C1.2 C2.4 C3.7 Dad C1.2 C2.4 C3.7 Doctor C1.2 C2.4 C3.7 Father C1.2 C2.4 C3.7 Food C1.2 C2.4 C3.7 Girl C1.2 C2.4 C3.7 Grandmother C1.2 C2.4 C3.7 Heart C1.2 C2.4 C3.7 Kitten C1.2 C2.4 C3.7 Laughter C1.2 C2.4 C3.7 Mother C1.2 C2.4 C3.7 Photo C1.2 C2.4 C3.7 Puppy C1.2 C2.4 C3.7 Sister C1.2 C2.4 C3.7 Television C1.2 C2.4 C3.7 Woman C1.2 C2.4 C3.7 Advantage C1.2 C2.3 C3.6 C4.2 Adversity C1.2 C2.3 C3.6 C4.2 Announcement C1.2 C2.3 C3.6 C4.2 Assistance C1.2 C2.3 C3.6 C4.2 Assumption C1.2 C2.3 C3.6 C4.2 Circumstance C1.2 C2.3 C3.6 C4.2 Cognition C1.2 C2.3 C3.6 C4.2

PAGE 38

38 Coincidence C1.2 C2.3 C3.6 C4.2 Complication C1.2 C2.3 C3.6 C4.2 Consideration C1.2 C2.3 C3.6 C4.2 Consistency C1.2 C2.3 C3.6 C4.2 Context C1.2 C2.3 C3.6 C4.2 Definition C1.2 C2.3 C3.6 C4.2 Denial C1.2 C2.3 C3.6 C4.2 Description C1.2 C2.3 C3.6 C4.2 Destiny C1.2 C2.3 C3.6 C4.2 Development C1.2 C2.3 C3.6 C4.2 Difference C1.2 C2.3 C3.6 C4.2 Dilemma C1.2 C2.3 C3.6 C4.2 Distraction C1.2 C2.3 C3.6 C4.2 Diversity C1.2 C2.3 C3.6 C4.2 Error C1.2 C2.3 C3.6 C4.2 Example C1.2 C2.3 C3.6 C4.2 Exception C1.2 C2.3 C3.6 C4.2 Explanation C1.2 C2.3 C3.6 C4.2 Gender C1.2 C2.3 C3.6 C4.2 Identification C1.2 C2.3 C3.6 C4.2 Ignorance C1.2 C2.3 C3.6 C4.2 Immortality C1.2 C2.3 C3.6 C4.2 Impairment C1.2 C2.3 C3.6 C4.2 Incident C1.2 C2.3 C3.6 C4.2 Interruption C1.2 C2.3 C3.6 C4.2 Legality C1.2 C2.3 C3.6 C4.2 Limitation C1.2 C2.3 C3.6 C4.2 Mastery C1.2 C2.3 C3.6 C4.2 Method C1.2 C2.3 C3.6 C4.2 Mystery C1.2 C2.3 C3.6 C4.2 Myth C1.2 C2.3 C3.6 C4.2 Necessity C1.2 C2.3 C3.6 C4. 2 Opposition C1.2 C2.3 C3.6 C4.2 Originality C1.2 C2.3 C3.6 C4.2 Permission C1.2 C2.3 C3.6 C4.2 Phenomenon C1.2 C2.3 C3.6 C4.2 Philosophy C1.2 C2.3 C3.6 C4.2 Preference C1.2 C2.3 C3.6 C4.2 Preparation C1.2 C2.3 C3.6 C4.2 Recognition C1.2 C2 .3 C3.6 C4.2 Regulation C1.2 C2.3 C3.6 C4.2 Reinforcement C1.2 C2.3 C3.6 C4.2 Replacement C1.2 C2.3 C3.6 C4.2

PAGE 39

39 Requirement C1.2 C2.3 C3.6 C4.2 Response C1.2 C2.3 C3.6 C4.2 Selection C1.2 C2.3 C3.6 C4.2 Separation C1.2 C2.3 C3.6 C4.2 Situatio n C1.2 C2.3 C3.6 C4.2 Stimulus C1.2 C2.3 C3.6 C4.2 Stupidity C1.2 C2.3 C3.6 C4.2 Tendency C1.2 C2.3 C3.6 C4.2 Theory C1.2 C2.3 C3.6 C4.2 Topic C1.2 C2.3 C3.6 C4.2 Translation C1.2 C2.3 C3.6 C4.2 Uncertainty C1.2 C2.3 C3.6 C4.2 Unity C1.2 C2 .3 C3.6 C4.2 Willingness C1.2 C2.3 C3.6 C4.2 Exclusion C1.2 C2.3 C3.6 C4.1 Fallacy C1.2 C2.3 C3.6 C4.1 Heresy C1.2 C2.3 C3.6 C4.1 Idiom C1.2 C2.3 C3.6 C4.1 Impossibility C1.2 C2.3 C3.6 C4.1 Irony C1.2 C2.3 C3.6 C4.1 Metaphor C1.2 C2.3 C3.6 C4.1 Pretense C1.2 C2.3 C3.6 C4.1 Accumulation C1.2 C2.3 C3.5 Acquisition C1.2 C2.3 C3.5 Addition C1.2 C2.3 C3.5 Amplitude C1.2 C2.3 C3.5 Appointment C1.2 C2.3 C3.5 Aspect C1.2 C2.3 C3.5 Availability C1.2 C2.3 C3.5 Brevity C1.2 C2. 3 C3.5 Calculation C1.2 C2.3 C3.5 Capacity C1.2 C2.3 C3.5 Category C1.2 C2.3 C3.5 Clearance C1.2 C2.3 C3.5 Combination C1.2 C2.3 C3.5 Convergence C1.2 C2.3 C3.5 Deduction C1.2 C2.3 C3.5 Dimension C1.2 C2.3 C3.5 Distribution C1.2 C2.3 C3.5 Duration C1.2 C2.3 C3.5 Dynasty C1.2 C2.3 C3.5 Emergence C1.2 C2.3 C3.5 Episode C1.2 C2.3 C3.5

PAGE 40

40 Establishment C1.2 C2.3 C3.5 Extent C1.2 C2.3 C3.5 Hierarchy C1.2 C2.3 C3.5 Magnitude C1.2 C2.3 C3.5 Majority C1.2 C2.3 C3.5 Midnight C1.2 C2.3 C3.5 Occasion C1.2 C2.3 C3.5 Origin C1.2 C2.3 C3.5 Paradigm C1.2 C2.3 C3.5 Proportion C1.2 C2.3 C3.5 Reduction C1.2 C2.3 C3.5 Retention C1.2 C2.3 C3.5 Unit C1.2 C2.3 C3.5 Variety C1.2 C2.3 C3.5 Autumn C1.1 C2.2 C3.4 Ball C1.1 C2.2 C3.4 Bed C1.1 C2.2 C3.4 Beverage C1.1 C2.2 C3.4 Boat C1.1 C2.2 C3.4 Bottle C1.1 C2.2 C3.4 Boulder C1.1 C2.2 C3.4 Cake C1.1 C2.2 C3.4 Calendar C1.1 C2.2 C3.4 Car C1.1 C2.2 C3.4 Ceiling C1.1 C2.2 C3.4 Chair C1.1 C2.2 C3.4 Church C1.1 C2.2 C3.4 Clothing C1.1 C2.2 C3.4 Cottage C1.1 C2.2 C3.4 Desk C1.1 C2.2 C3.4 Diamond C1.1 C2.2 C3.4 Door C1.1 C2.2 C3.4 Elephant C1.1 C2.2 C3.4 Factory C1.1 C2.2 C3.4 Fireplace C1.1 C2.2 C3.4 Fog C1.1 C2.2 C3.4 Fountain C1.1 C2.2 C3.4 Grass C1.1 C2.2 C3.4 Hat C1.1 C2.2 C3.4 Hospital C1.1 C2.2 C3.4 Hurricane C1.1 C2.2 C3.4 Jacket C1.1 C2.2 C3.4 Lightning C1.1 C2.2 C3.4

PAGE 41

41 Mattress C1.1 C2.2 C3.4 Menu C1.1 C2.2 C3.4 Mon ey C1.1 C2.2 C3.4 Moonlight C1.1 C2.2 C3.4 Neck C1.1 C2.2 C3.4 Orchestra C1.1 C2.2 C3.4 Oven C1.1 C2.2 C3.4 Pie C1.1 C2.2 C3.4 Rainbow C1.1 C2.2 C3.4 Refrigerator C1.1 C2.2 C3.4 Rocket C1.1 C2.2 C3.4 Roof C1.1 C2.2 C3.4 Saloon C1.1 C2.2 C3.4 Shirt C1.1 C2.2 C3.4 Snow C1.1 C2.2 C3.4 Sofa C1.1 C2.2 C3.4 Tree C1.1 C2.2 C3.4 Truck C1.1 C2.2 C3.4 Wallet C1.1 C2.2 C3.4 Window C1.1 C2.2 C3.4 Winter C1.1 C2.2 C3.4 Bridge C1.1 C2.2 C3.3 City C1.1 C2.2 C3.3 Cliff C1.1 C2.2 C3.3 Forest C1.1 C2.2 C3.3 Hotel C1.1 C2.2 C3.3 Island C1.1 C2.2 C3.3 Lake C1.1 C2.2 C3.3 Landscape C1.1 C2.2 C3.3 Location C1.1 C2.2 C3.3 Mansion C1.1 C2.2 C3.3 Mountain C1.1 C2.2 C3.3 Ocean C1.1 C2.2 C3.3 Palace C1.1 C2.2 C3.3 Pond C1.1 C2.2 C3.3 Pyramid C1.1 C2.2 C3.3 River C1.1 C2.2 C3.3 Road C1.1 C2.2 C3.3 Sky C1.1 C2.2 C3.3 University C1.1 C2.2 C3.3 Volcano C1.1 C2.2 C3.3 Zoo C1.1 C2.2 C3.3 Apple C1.1 C2.1 C3.2

PAGE 42

42 Bird C1.1 C2 .1 C3.2 Blood C1.1 C2.1 C3.2 Butterfly C1.1 C2.1 C3.2 Candy C1.1 C2.1 C3.2 Coffee C1.1 C2.1 C3.2 Dentist C1.1 C2.1 C3.2 Eagle C1.1 C2.1 C3.2 Gun C1.1 C2.1 C3.2 Lip C1.1 C2.1 C3.2 Pillow C1.1 C2.1 C3.2 Policeman C1.1 C2.1 C3.2 Salad C1.1 C2.1 C3.2 Skin C1.1 C2.1 C3.2 Sugar C1.1 C2.1 C3.2 Alligator C1.1 C2.1 C3.1 Ambulance C1.1 C2.1 C3.1 Ankle C1.1 C2.1 C3.1 Arrow C1.1 C2.1 C3.1 Banana C1.1 C2.1 C3.1 Beaver C1.1 C2.1 C3.1 Beetle C1.1 C2.1 C3.1 Bel l C1.1 C2.1 C3.1 Blade C1.1 C2.1 C3.1 Bracelet C1.1 C2.1 C3.1 Bubble C1.1 C2.1 C3.1 Bullet C1.1 C2.1 C3.1 Butter C1.1 C2.1 C3.1 Caterpillar C1.1 C2.1 C3.1 Chicken C1.1 C2.1 C3.1 Chipmunk C1.1 C2.1 C3.1 Cocktail C1.1 C2.1 C3.1 Coffin C1.1 C2.1 C3.1 Corn C1.1 C2.1 C3.1 Corpse C1.1 C2.1 C3.1 Coupon C1.1 C2.1 C3.1 Cow C1.1 C2.1 C3.1 Cranberry C1.1 C2.1 C3.1 Crocodile C1.1 C2.1 C3.1 Crown C1.1 C2.1 C3.1 Cucumber C1.1 C2.1 C3.1 Cup C1.1 C2.1 C3.1 Darknes s C1.1 C2.1 C3.1 Dove C1.1 C2.1 C3.1

PAGE 43

43 Drum C1.1 C2.1 C3.1 Eyeball C1.1 C2.1 C3.1 Fisherman C1.1 C2.1 C3.1 Flask C1.1 C2.1 C3.1 Football C1.1 C2.1 C3.1 Frog C1.1 C2.1 C3.1 Gorilla C1.1 C2.1 C3.1 Grasshopper C1.1 C2.1 C3.1 Helmet C1.1 C2.1 C3.1 Horse C1.1 C2.1 C3.1 Item C1.1 C2.1 C3.1 Jewel C1.1 C2.1 C3.1 Key C1.1 C2.1 C3.1 Kite C1.1 C2.1 C3.1 Lamb C1.1 C2.1 C3.1 Lamp C1.1 C2.1 C3.1 Laundry C1.1 C2.1 C3.1 Lemon C1.1 C2.1 C3.1 Leopard C1.1 C2.1 C3.1 Lettuce C1.1 C2.1 C3.1 Lion C1.1 C2.1 C3.1 Lizard C1.1 C2.1 C3.1 Lobster C1.1 C2.1 C3.1 Macaroni C1.1 C2.1 C3.1 Microscope C1.1 C2.1 C3.1 Missile C1.1 C2.1 C3.1 Mosquito C1.1 C2.1 C3.1 Mustard C1.1 C2.1 C3.1 Necklace C1.1 C2.1 C 3.1 Newspaper C1.1 C2.1 C3.1 Onion C1.1 C2.1 C3.1 Opera C1.1 C2.1 C3.1 Orchid C1.1 C2.1 C3.1 Peach C1.1 C2.1 C3.1 Pig C1.1 C2.1 C3.1 Pigeon C1.1 C2.1 C3.1 Pimple C1.1 C2.1 C3.1 Potato C1.1 C2.1 C3.1 Propeller C1.1 C2.1 C3.1 Queen C1.1 C2.1 C3.1 Rabbit C1.1 C2.1 C3.1 Raspberry C1.1 C2.1 C3.1 Robin C1.1 C2.1 C3.1

PAGE 44

44 Sandal C1.1 C2.1 C3.1 Shark C1.1 C2.1 C3.1 Shrimp C1.1 C2.1 C3.1 Skull C1.1 C2.1 C3.1 Skunk C1.1 C2.1 C3.1 Snake C1.1 C2.1 C3.1 Spider C 1.1 C2.1 C3.1 Stapler C1.1 C2.1 C3.1 Tennis C1.1 C2.1 C3.1 Thorn C1.1 C2.1 C3.1 Toilet C1.1 C2.1 C3.1 Tomato C1.1 C2.1 C3.1 Tongue C1.1 C2.1 C3.1 Tool C1.1 C2.1 C3.1 Towel C1.1 C2.1 C3.1 Typewriter C1.1 C2.1 C3.1 Wolf C1.1 C2. 1 C3.1

PAGE 45

45 LIST OF REFERENCES Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster analysis Thousand Oaks: Sage Publications. Allen, R., & Hulme, C. (2006). Speech and language processing mechanisms in verbal serial recall. Journal of Memory and Lan guage, 55 (1), 64 88. Altarriba, J., Bauer, L., & Benvenuto, C. (1999). Concreteness, context availability, and imageability ratings and word associations for abstract, concrete, and emotion words. Behavior Research Methods, 31 (4), 578 602. Amazon Incorpora ted (2012). Mechanical Turk. www.mturk.com. Amodio, D. M., & Frith, C. D. (2006). Meeting of minds: the medial frontal cortex and social cognition. Nat ure Rev iews Neurosci ence 7 (4), 268 277. Anderson, T. W., & Rubin, H. (1956). Statistical inference in fa ctor analysis. Proceedings of the Berkeley Symposium on Mathematical Statistics and Probability, 5 111 150. Bleasdale, F. A. (1987). Concreteness dependent associative priming: Separate lexical organization for concrete and abstract words. Journal of Expe rimental Psychology: Learning, Memory, and Cognition, 13 (4), 582 594. Bonner, M. F., Vesely, L., Price, C., Anderson, C., Richmond, L., Farag, C., et al. (2009). Reversal of the concreteness effect in semantic dementia. Cognitive Neuropsychology, 26 (6), 56 8 579. evaluation of current word frequency norms and the introduction of a new and improved word frequency measure for American English. Behavior Research Methods, 41 (4), 977 99 0. Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk. Perspectives on Psychological Science, 6 (1), 3 5. Burgess, N., Maguire, E. A., & O'Keefe, J. (2002). The h uman h ippocampus and s patial and e pisodic m emory. Neuron, 35 (4), 625 641. Coltheart, M. (1981). The MRC psycholinguistic database. The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 33A (4), 497 505.

PAGE 46

46 Crutch, S. J., & Jackson, E. C. (2011). Contrasting graded effects of semantic similarity and association across the concreteness spectrum. The Quarterly Journal of Experimental Psychology, 64 (7), 1388 1408. Franklin, S., Howard, D., & Patterson, K. (1995). Abstract word anomia. Cognitive Neuropsychology, 12 (5), 549 566. Gilhooly, K., & Logie, R. ( 1980). Age of acquisition, imagery, concreteness, familiarity, and ambiguity measures for 1,944 words. Behavior Research Methods, 12 (4), 395 427. Hale, S. C. (1988). Spacetime and the a bstract/ c oncrete d istinction. Philosophical Studies: An International J ournal for Philosophy in the Analytic Tradition, 53 (1), 85 102. Katz, R. B., & Goodglass, H. (1990). Deep dysphasia: Analysis of a rare form of repetition disorder. Brain and Language, 39 (1), 153 185. Kousta, S. T., Vigliocco, G., Vinson, D. P., Andrews, M ., & Del Campo, E. (2011). The representation of abstract words: Why emotion matters. Journal of Experimental Psychology: General, 140 (1), 14 34. Kousta, S. T., Vinson, D. P., & Vigliocco, G. (2009). Emotion words, regardless of polarity, have a processing advantage over neutral words. Cognition, 112 (3), 473 481. Martin, N., & Saffran, E. M. (1992). A computational account of deep dysphasia: Evidence from a single case study. Brain and Language, 43 (2), 240 274. Mervis, C. B., & Rosch, E. (1981). Categorizat ion of Natural Objects. Annual Review of Psychology, 32 (1), 89 115. Moscovitch, M., Nadel, L., Winocur, G., Gilboa, A., & Rosenbaum, R. S. (2006). The cognitive neuroscience of remote episodic, semantic and spatial memory. Current Opinion in Neurobiology, 16 (2), 179 190. Paivio, A. (1991). Dual coding theory: Retrospect and current status. Canadian Journal of Psychology/Revue canadienne de psychologie, 45 (3), 255 287. Reed, H. B., & Dick, R. D. (1968). The learning and generalization of abstract and concret e concepts. Journal of Verbal Learning and Verbal Behavior, 7 (2), 486 490. Roeltgen, D. P., Sevush, S., & Heilman, K. M. (1983). Phonological agraphia: Writing by the lexical semantic route. Neurology, 33 (6), 755 765.

PAGE 47

47 Rosch, E. (1975). Cognitive representa tions of semantic categories. Journal of Experimental Psychology: General, 104 (3), 192 233. Rosch, E., Mervis, C. B., Gray, W. D., Johnson, D. M., & Boyes Braem, P. (1976). Basic objects in natural categories. Cognitive Psychology, 8 (3), 382 439. Schwanenf lugel, P. J., Harnishfeger, K. K., & Stowe, R. W. (1988). Context availability and lexical decisions for abstract and concrete words. Journal of Memory and Language, 27 (5), 499 520. Schwanenflugel, P. J., & Shoben, E. J. (1983). Differential context effect s in the comprehension of abstract and concrete verbal materials. Journal of Experimental Psychology: Learning, Memory, and Cognition, 9 (1), 82 102. Squire, L. R. (1982). The n europsychology of h uman m emory. Annual Review of Neuroscience, 5 (1), 241 273. St uss, D. T., Shallice, T., Alexander, M. P., & Picton, T. W. (1995). A m ultidisciplinary a pproach to a nterior a ttentional f unctions. Annals of the New York Academy of Sciences, 769 (1), 191 212. SurveyMonkey.com LLC. (2011). Survey Monkey. www.surveymonkey.c om. Vuilleumier, P., Armony, J. L., Driver, J., & Dolan, R. J. (2003). Distinct spatial frequency sensitivities for processing faces and emotional expressions. Nat ure Neurosci ence 6 (6), 624 631. Walker, I., & Hulme, C. (1999). Concrete words are easier to recall than abstract words: Evidence for a semantic contribution to short term serial recall. Journal of Experimental Psychology: Learning, Memory, and Cognition, 25 (5), 1256 1271. Walsh, V. (2003). A theory of magnitude: common cortical metrics of time, space and quantity. Trends in Cognitive Sciences, 7 (11), 483 488. Ward, J. H., Jr. (1963). Hierarchical g rouping to o ptimize an o bjective f unction. Journal of the American Statistica l Association, 58 (301), 236 244. Warrington, E. K. (1975). The selective impairment of semantic memory. The Quarterly Journal of Experimental Psychology, 27 (4), 635 657.

PAGE 48

48 BIOGRAPHICAL SKETCH Joshua Troche graduated in 2009 from the University of Florida w ith a B achelor of S cience in p sychology and minor in r eligion and c lassical s tudies. In 2009, h e began a combined M aster of A rts / Doctor of Philoso p h y program in the Department of Speech, Language and Hearing Sciences. His research interests include semantic memory, dementia and cognitive rehabilitati on.