THESE-i This is to certify that the dissertation entitled Qualitative Differences in the Word Definition Responses of Aphasic and Non-aphasic Adults presented by Silvia A. Graves has been accepted towards fulfillment of the requirements for Ph.D. degree in Audiology § Speech Sciences . saw Major professor Date June 8, 1983 MS U is an Affirmative Action/Equal Opportunity Institution 0— 12771 MSU LIBRARIES ._:,—_ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. :5. a. .— ,_ i ’19.} :l.- .r. 4.1!!! ”bl-1..» fr '5. '5“ "' v ‘ ' ' ’ ~ ‘ a I a: s W aw 2. l; 2.2:, that. ~%‘i USE (my /§// -- air 7;? QUALITATIVE DIFFERENCES IN THE WORD DEFINITION RESPONSES OF APHASIC AND NONAPHASIC ADULTS BY Silvia Ann Graves A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Audiology and Speech Sciences 1983 Qualitative Differences in the Word Definition Responses of Aphasic and Nonaphasic Adults By Silvia Ann Graves ABSTRACT Previous studies support the contention that apha- sic individuals show a loss of the ability to abstract and an increased tendency toward concreteness in verbal and non-verbal behavior. The purpose of this study was to investigate the aphasic individual's verbalization of word meaning in terms of a concrete-abstract dimen- sion through a qualitative analysis of their verbatim responses to a word definition task. The sample consisted of 20 nonaphasic subjects be- tween the ages of AS and 70, and 20 aphasic subjects between the ages of NS and 83 who passed the criterion tests of word recognition and the PICA. Both subject groups were equated on the variables of age, education, and sex. The subjects were required to orally define 2O noun words on high and low frequency levels. Using a qualitative analysis system, their audiotape recorded definition responses were evaluated on a five-point classification system: Synonym, Explanation, Use and Description, Inferior, and Error types. The MANOCOVA and 3ftest techniques were employed to test the signifi- cance of the differences between mean scores for the two groups. The analysis led to the conclusion that the apha- sic, like the nonaphasic subjects, used both concrete and abstract definition categories but used them in varying degrees. Whereas the aphasic individuals tend to use more concrete (Use and Description and Inferior- type) and Error definitions, the nonaphasic individuals used more of the abstract definitions (Synonym and Ex- planation). This suggests that the aphasic subjects may not have complete impairment in abstraction; rather the aphasic subject can be impaired along a concrete— 'abstract continuum of language. Although the frequency ’of word occurrence (high or low) did not affect the rate of error response for the two groups, it did affect the type of definition response given. For all subjects, age had no effect on type of definition response; however, the higher the level of education, the more likely were abstract definitions offered. On the basis of the current study, it is likely that the speech—language pathologist may find a continu- um model rather than the categorical model of the ab- stract-concrete dimension of verbal language the most appropriate framework for evaluating aphasic individual's verbal ability. ACKNOWLEDGEMENTS Like every human product, this dissertation is a heur- istic endeavor with a history of indebtedness too long to be listed, too complex and subtle ever to be known by the writer. However, for their special prominence, special thanks goes to: Dr. Leo V. Deal, who was the writer's personal student advisor and chairperson of both the writer's advisory and dissertation committees, for his encouragement, advice, assistance, and for providing the means for following out graduate study; Dr. (Sister) Maryjfionora Kroger, for the influence she has had upon this study of this student. Her friendship, encouragement, advice, and sensitive criticism have been a profound help in completing this dissertation; Dr. Ida Stockman, for advising the dissertation to its completion and for significantly contributing editorial work well beyond normal expectations; The directors of the speech and hearing research facil- ities, Ms. Vidalia McTeer and Ms. Jan Jones; the director and assistant director of the Tri-County Office on Aging (Region 6), Ms. Roxanne O'Connor and Ms. Dorothy Payne, respectively; and to all subjects who gave so freely of their time without whose help this study may not have been completed; Ms. Vidalia McTeer, who shared the experiences of the research from start to finish and provided, apart from intellectually rewarding comradarie, the enjoyment of friendly sisterhood; and My husband, Vincent Rivera, for encouragement and 'vital support throughout the final phase of this doctor— al program. TABLE OF CONTENTS List of Tables . . . . . . . . . . . . . . . . . . List Of Figures 0 O O O O O O O O O O O O O O O 0 CHAPTER I: BACKGROUND AND STATEMENT OF THE PROBLEM 7, . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . Statement of the Proble . . . . . . . . . . Purpose of the Study . . . CHAPTER II: BACKGROUND AND LITERATURE REVIEW . . Related Theories of Aphasia which Suggest an Abstractness-Concreteness Dimension . . Pr0positional theory . . . . . . . Abstract-concrete theory . . . . . Divergent semantic theory . . . . . Unidimensional theory . . . . . . . . . . Regression theory . . . . . . . . . . . . Summary of related theories . . . . . . . Related Studies on Abstract-Concrete Dimension of Word Definitions . . . . . . . . Word definition performance in normal children . . . . . . . . . . . . . . . . Word definition performance in normal adults . . . . . . . . . . . . . . . . . Word definition performance in aphasic individuals . . . . . . . . . . . . . . . Stimulus Variable Affecting Word Definition Performance in Aphasic Adults . . . . . . . . Frequency of word occurrence . . . . . . Summary of Related Literature . . . . . . CHAPTER III: METHODS . . . . . . . . . . . . . . Operational Definition of Terms . . . . . . . Subjects and Subject Selection Criteria . . . General criteria . . . . . . . . . . . . Aphasic subjects selection: group I . . iv vii viii 45 45 47 47 47 Nonaphasic subject selection: group II Procedures for Subject Selection . . . - Description and Administration of Criterion Tasks . . . . . . . . . . . . . The Porch Index of Communicative Ability (PICA) . . . . . . . . . . The Word Recognition Screening Task Description of Experimental Task . . . Selection of stimulus words . . . . Procedures for Administering the Experimental Task . . . . . . Testing procedures . . . Scoring of responses . . Statistical analysis . . CHAPTER IV: RESULTS . . . . . . . . . . . . . Statistical Treatment of the Data . . Adjusted Scores for the aphasic and nonaphasic groups. . . . . . . . . . Aphasic and Nonaphasic Between-Group Differences. . . . . . Aphasic and Nonaphasic Within-Group Differences. . . . . . . . . . . Analysis of the mean scores for the aphasic group. . . . . . . . . . . . Analysis of the mean scores for the nonaphasic group . . . . . . . . . . Noun Stimulus Differences . . . . . . . . Supplementary Analysis . . . . . . . . . CHAPTER V: DISCUSSION . . . . . . . . . . . . Theoretical Implications . . . Clinical Implications . . . . . Diagnostic implications . . Therapeutic implications . Limitations of the Study . . . Subjects . . . . . . . . Method of scoring definition goon... spo a CHAPTER VI: SUMMARY, CONCLUSIONS, AND IMPLICATIONS FOR FURTHER RESEARCH . . . . . . Summary . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . Implications for Further Research 8 o o o o o o 48 49 50 50 51 51 52 55 57 59 62 64 64 66 69 70 71 72 79 81 82 85 86 87 89 89 90 91 91 95 96 APPENDICES O O O O O O O O O O O 0 O O O O O ‘ O O O O O 99 A. Description of Research Project for the Speech and Hearing Facility . . . . . . . . . . . . . . . 99 B. Description of Research Project for the Tri- County Office on Aging . . . . . . . . . . . . . . 100 C. Letter Requesting Permission for Release of Subjects' Telephone Numbers . . . . . . . . . . . 101 D. Interview Form . . . . . . . . . . . . . . . . . . 102 E. Informed Consent Contract . . . . . . . . . . . . 105 F. Description of the PICA Subtests Used . . . . . . 106 G. Response Score Sheet . . . . . . . . . . . . . . . 107 H. Total Definition Response Scores for Each Subject in the Aphasic and Nonaphasic Groups for the Five Quality Levels of Word Definitions . . . . . 109 BIBLIOGRAPHY . O O O O O O O O O O O O O O O O O O O O 111 vi 10. LIST OF TABLES Distribution of Sex, Age, and Education of the Aphasic and Nonaphasic Group . . . . . . . The 20 Stimulus Words Classified According to Level of Abstraction and Frequency of Occurrence in the English Language . . . . . . . Qualitative Classification System of Definition Responses . . . . . . . . . . . . . . Means and Standard Deviations of Definition Responses for the Aphas ic and Nonaphasic Groups on Each of Five Definition Response Categories Pooled across High- and Low-Frequency Words . . . Significance of Difference on Definition Response Scores on a Word Stimuli Task Between Nonaphasic and Aphasic Groups . . . . . . . . . . Correlation Between Dependent and Canonical Variables and Percent of Variance Listed in Order Specifying the Relative Variable Contribution to Group Separation . . . . . . . . Mean Differences Between Word Definition Scores by Aphasic and Nonaphasic Subjects Pooled Across High- and Low-Frequency Words . . . . . . . . . . Mean Values for Each of the Five Quality Levels of Word Definitions Given by the Aphasic and Nonaphasic Groups on Words of Low and High Levels of Frequency . . . . . . . . A Comparison of Word Definition Responses of Aphasic and Nonaphasic Groups on Words of Low and High Levels of Frequency . . . . . . . . Significance of the Relationship Between Word Definition Re5ponse Performance, Age, and Education for Aphasic and Nonaphasic Groups . vii 48 53 58 63 67 69 71 73 74 LIST OF FIGURES Distribution of Synonym Definition Response Scores for the Aphasic and Nonaphasic Groups Distribution of Explanation Definition Response Scores for the Aphasic and Non- aphasic Groups . . . . . . . . . . . . . . . Distribution of Use and Description Definition Response Scores for the Aphasic and Nonaphasic Groups . . . . . . . . . . . . Distribution of Demonstration, Illustration, Repetition, Inferior, Explanation Definition Response Scores for the Aphasic and Non- aphasic Groups . . . . . . . . . . . . . . . Distribution of Error Definition Response‘ Scores for the Aphasic and Nonaphasic Groups viii 76 77 77 78 78 CHAPTER I BACKGROUND AND STATEMENT OF THE PROBLEM Introduction Word definition responses, i.e., the ability to use words to represent meaning (Litowitz, 1977, 1981), have been of interest to many aphasiologists because of their assumed importance to language functioning (Bayles & Boone, 1982; Davis, Myers, Adamovitch, & Yorkston, 1981; Goldman- Eisler, 1964; Muma, 1975) and to human communication (Chapey, 1981; Krauss, 1972; Lesser, 1978; Ulatowski, Macaluso-Haynes, & North, 1981). The concept of "knowing" a word may appear obvious, yet there are complex behaviors involved; and knowledge of even simple words is a cumulative process that results from gradual develOpment (Litowitz, 1981). For the adult, knowe ing a word implies a conscious or unconscious understanding of the critical attributes included inia dictionary defini- tion such as the part of speech, how it can be combined in a sentence (selectional features), and which words have Sim? ilar meanings (synonyms) and opposite meanings (antonyms). Therefore, to define a word verbally involves memory, speech behavior, motivation, a knowledge of sentence struc- ture, and the ability to select and organize words in a meaningful way (Harrington & Erhmann, 1954). Further, it is this verbal information which is sought and examined in the definitional sections of most vocabulary tests. The importance of word definition performance becomes apparent when we consider that word definitions imply that a mental representation of the word is.the basis 0 for all responses to a given word. For example, in the de- velopment of word meaning, the individual learns to repre- sent the world of objects internally (Jackson, 1978; Lenne- berg, 1967). He/She cannot place the objects inside of . his/her brain, so he/she deals with real objects through mental representations. It is implicit in this process that the individual developsaimental substitute for the real object. Correspondingly, once the individual creates mental substitutions for objects, he/she begins to use‘ words or phrases in his/her language to verbalize what she/he means by the object. Thus, the mental representa- tion or content of ideas (i.e., concept) formed in the in- dividual's mind contains the information to be expressed linguistically about the object in order to make possible the sharing of experience necessary to the existence of every social group (Litowitz, 1977, 1981). . One view of the representational system underlying an object's meaning is that the mental representation can take either an abstract or a concrete conceptual form (Bruner, Goodnow, & Austin, 1956). That is, the input (which is of- ten a word) is translated into a conceptual form containing either abstract elements and their relations or concrete elements and their perceptual attributes (Nelson, 1974, 1975). In this sense,concepts (represented by words or phrases when making a word definition) can be described as either concrete or abstract according to the kind of con- ceptualization formed in response to the stimulus word. A characteristically concrete definition response, then, ex- presses the meaning of the referent in terms of an observ- able context-dependent relation (e.g., donkey--"it has four legs“). A characteristically abstract word definition re-. sponse expresses the meaning of the referent in terms of a generic class or category (e.g., donkey--"an animal") (Anglin, 1970, 1977). Studies cf word meaning as an index of concept forma- tion among normal subjects have employed a form of analysis related directly to word definition performance (Feifel, 1949; Feifel & Lorge, 1950; Green, 1931; Storck & Looft, 1973» These studies involving normal children and adults typically have required subjects to give definitions which were scored according to their definition levels (that is, level of conceptualization). They found a developmental progression in conceptualization from concrete and action- oriented definitions in early childhood to more abstract formalizations in later adolescence and adulthood. It is noteworthy that although it has been long asserted that aphasic individuals demonstrate difficulty associating words with their respective meanings (Doehring & Swisher, 1972; Gardner & Denes, 1973; Hanson, 1976; Myers & Line- baugh, 1981; Muma & McNeil, 1981; Osgood & Miron, 1963; Schuell, Jenkins, & Landis, 1961), this kind of investiga- tion has not been systematically carried out in the aphasic population. The use of word meaning as an index of concept forma- tion among adults with language problems has been studied but primarily in the context of word association tasks (Grossman, 1981; Goldfarb & Halpern, 1981; Milberg & Blum- stein, 1981; Kelter, 1976). These authors analyzed the free speech responses of their subjects and concluded that a "conceptual" deficit of some sort plagues the aphasic individual's ability to refer with a word. Studies employing word association tasks relative to word definition response performance have also examined conceptualization in aphasia by relying on lexical judg- ments (Biasschi, Denes, & Semenza, 1976; Whitehouse, Zurif, '& Caramazza, 1978; Zurif, Caramazza, Myerson, & Galvin, 1974). In order to test the defining properties of words in aphasia, these authors asked aphasic subjects to judge which two of three items best go together. Zurif et a1. (1974), in particular, concluded that aphasic individuals barely make the supraordinate category differentiation, thus supposing a common abstract conceptual relation; but, rather, they perform better on the thematic category in the sense of forming context-dependent or concrete relations. Others who have regarded aphasia as a shift in the level of 'conceptualization include Bouman and Grunbaum (1925), Lubinski et a1. (1980), and Milton, Wertz, Katz, and Prut- ting (1981). The research studies cited thus far on testing for word meaning impairment in aphasia suggest that there are characteristic patterns of difficulty in the word defini- tion response performance of aphasic individuals. Aside from word meaning, investigators have attempted to deter— mine whether other aspects of words produce inherently dif- ficult situations for the aphasic individual. The para- meters explored have included (1) the frequency of word occurrence in English usage (Bricker, Schuell, & Jenkins, 1964; William & Canter, 1981; Filby, Edwards, & Seacat, 1963; Goodglass, Hyde, & Blumstein, 1969; Goldfarb & Hal- pern, 1981; Howes, 1964, 1967; and Siegel, 1959): (2) part of speech (Eisenson, 1954; Siegel, 1959; Weisenberg & McBride, 1951; Wepman, 1951): (3) word length (Bricker et al., 1964; Filby et al., 1963; Goldstein, 1948; Jenkins, Miminez-Palon, Shaw, & Sefer, 1975; Siegel, 1951; Wepman, 1951; Wepman & Jones, 1961); and (4) abstract versus con- crete nouns (Brown, 1956; Goldfarb & Halpern, 1981; Spreen, 1968). Concrete noun words refer to those words whose ref- erence to objects, to material, to sources of sensation is relatively direct. Abstract nouns refer to those words whose reference to objects, to material, to sources of sen- sation is relatively indirect (Gorman, 1961). The patterns of difficulty with word tasks have been explained in terms of a broad theoretical framework, con- cerning the entire problem of aphasia (Chapey, 1977, 1981; Eisenson, 1954, 1971; Goldstein, 1942, 1948; Head, 1915; Kreindler, Gheorgita, & Voineseu, 1971; Schuell, 1974; Wepman, 1951, 1972). These theories indicate that the in- ability of an aphasic individual to produce a number and variety of responses is due to a fundamental problem of shifting from the concrete to the abstract conceptual per- spective in performing verbal and nonverbal tasks. Goldstein (1948) stated that normally a person . . . combines both (the concrete and ab- stract) attitudes and is capable of shifting from one to the other at will according to the demands of the situation. Some tasks can be performed only by virtue of the abstract at- titude. For others, the concrete attitude is sufficient. But he maintained that the aphasic individual's nonverbal and verbal behavior is concrete and she/he is unable to make the shift to the abstract attitude. More recently, Chapey (1977) stressed the fact that aphasic individuals often retain highly concrete verbal responses even when less concrete and more abstract re- sponses are not retained. In a later analysis, Chapey (1981) explained that the inability to generate a number of semantic responses (of which word definition is one aspect), to change direction of one's responses, or to use words flexibly to express a variety of relationships is a central component of the aphasic complex. Thus, the views of Chapey and Goldstein seem to be typical of Werner's (1965) proposal that the aphasic individual regresses from a more abstract, verbal conceptual level to a level that is concrete, general, and more tied to perceptual attributes. The term regression as used by Werner refers to the struc- tural re-emergence of developmentally lower levels of func- tioning when the more advanced and more recently developed levels of functioning are disorganized. Aphasia is thus. seen as a regression in cognitive processes. That is, it is conceived as a reversal of those patterns of conceptualiza- tion which are encountered in the normal course of develop- ment. Statement of the Problem A common theoretical claim is that aphasia, in general, is not simply an impairment in language but also a reduc- tion in verbal behavior which manifests itself in a more concrete manner of thinking (Goldstein, 1942, 1948). While many sources (Chapey, 1977, 1981; Lubinski et al., 1980; Osgood, 1953; Osgood & Miron, 1963; Siegel, 1959; Werner, 1965) have accepted and corroborated Goldstein's contribu- tion concerning abstract and concrete behavior, other in- vestigators have taken issue with his views, procedures, and results. One such author, Brown (1955), criticized Goldstein's failure to obtain normal control data in his work and his tendency to generalize from extensive studies of individual subjects. Another issue is the methodology used in later studies (Brown, 1955; Lubinski et al., 1980; Siegel, l959)in examining abstract behavior in aphasia. For instance, the Goldstein-Scheerer test (1941) used by Brown (1955) is nonquantitative and requires a rating of concrete- ness by the experimenter. Furthermore, in Brown's study, an adequate control group was absent. The Kendler reversal- nonreversal paradigm used by Lubinski et a1. (1980) is a nonverbal, pictorial task of common objects necessitating a pointing response by the subject. The word stimuli task by Siegel requires an utterance by the subject to each stimu- lus word as the card upon which it is typed is presented. Therefore, because of the inconsistency in the methodology of these studies, the results of these studies appear inconclusive. Furthermore, other authors (Brown, 1955; Eisenson, 1954; Wepman, 1951) indicate that the loss of ability to abstract was not a general characteristic of the aphasic individuals under their observation. Although research studies have suggested that defini— tions can reflect two different levels of conceptualization (i.e., the abstract and the concrete) little attention seems to have been given to the complexity of the response required in the word definition response performance of aphasic individuals. For example, the research studies cited to this point have used, in the main, either a match- ing task, a picture recognition task, or a multiple choice word list in which the mere selection of the response was considered most important. Selecting a correct definition from a number of possible definition alternatives may be a far much less complex task than that which requires spon- taneous verbal definition. While a word recognition task involves motivation, perception, and memory, the selection and organization of words into a meaningful sequence is not required by the subject. All that is required of the sub- jects is to recognize the word and make a choice limited to recognized possibilities. It would appear, then, that the question of whether the abstract solutions to language tasks are no longer available to an aphasic individual is deserving of more systematic study. Therefore, the present study attempts to investigate, qualitatively, the use of language in the ability of aphasic individuals to engage in abstract concep- tualization as inferred from their performance on a word definition task. The word definition task permits us to rate the performance of a subject according to all the variety of different responses given to all words and is therefore sensitive to whether a response is abstract, concrete, or even somewhat peripherally relevant; The assessment of an individual's performance in this way,- may have potential diagnostic value of clinical and theo- retical importance- For example, the traditional, standardized aphasia tests of vocabulary afford primarily a measurement of the number of words used infrequently in conversation. One may, however, measure other 10 aspects of vocabulary such as the subject's (l) ability to form concepts and (2) conceptual level of meaning fOrmed with familiar words. These variables are neglected by the usual standardized aphasia test of vocabulary which re- quires only that one meaning be given for each word and does not differentiate among the different shades, quality, and range of possible meanings. A qualitative analysis of word definition responses by aphasic individuals to familiar words may thus com- pliment the traditional aphasia test of vocabulary in two ways: (1) it requires a degrees of flexibility since the subject must shift from the meaning of one word to another one. Chapey (1971) has stressed the importance of flexi- bility for more articulated use of language; and (2) it may be more indicative of an easier solution by simple associative processes when the'mechanisms involved in the more difficult conceptual organization underlying word meaning are no longer avaliable. Thus, a qualitative analysis of word definition responses may be used as an index of measurement in the amount of change from the pre-impaired level of conceptual ability in word meanings. Purpose of the Study. The purpose of the present study is to investigate the quality of aphasic individuals' verbalizations of word meaning in terms of a concrete- abstract dimension. Using a qualitative analysis of the definition responses to selected noun stimuli by equated 11 groups of aphasic and nonaphasic subjects, this study at- tempts to answer the following questions: 1. Are there qualitative differences in the verbal definitions of selected nouns be- tween aphasic and nonaphasic groups? What types of word definitions are ob- served within the aphasic group and With- in the nonaphasic group? Are there differences between aphasic and nonaphasic groups in their word definition responses to high- and low- frequency nouns? CHAPTER II BACKGROUND AND LITERATURE REVIEW There is common agreement in the literature that apha- sia refers to known focal damage to a cortical area of the brain which typically involves the extensive central zone ixitheleft hemisphere (Lesser, 1978). Although it is wide- ly agreed that language disturbance can result from corti- cal insult, there is less global agreement on the nature of aphasia and exactly what language deficits are to be in- cluded. However, there appears to be a growing awareness that the general language disturbance can be viewed along an abstract-concrete dimension (Chapey, 1977; Goldstein, 1942, 1948). In this sense, aphasic individuals are said to be able to respond only to the immediate sense impression of an object and are unable to produce verbal responses con- ceptually from their experience of a word in terms of gen- eral information. This means, then, that the aphasic indi- vidual's language disturbance is at the abstract level. If this is the case, then it is reasonable to assume that an abstract impairment of language could be reflected in any aspect of the aphasic individual's language usage, in- cluding the making of a word definition. The first section of this chapter presents the theories of aphasia that seem to support a concrete-abstract 1') 13 dimension in aphasia. The second section summarizes empiri- cal studies related to the use of word definition responses in terms of the abstract-concrete dimension by normal children and adults as well as by aphasic individuals. Related Theories of Aphasia which Suggest an Abstractness- Concreteness Dimension Propositional theory. The propositional theory is associated with Hughling Jackson (1948, 1978) who consi- dered the physical basis of a spoken word as a sensori- motor and an audio—articulatory nervous arrangement. Jack- son pointed out that the utterance of any number of words would not constitute speech, for human beings do not think or speak in words alone but in words or signs referring to one another in a particular manner. These necessary rela— tionships inherent in meaningful language he termed "pro- positions." He maintained that a proposition implied a relationship of two images, internal (symbolic) in the sense that each of them is related to all other images organized in individuals and external (object) in the sense that it is also related to things in the environment. The two propositions taken together imply an internal and ex- ternal relationship; i.e., words are not only related to one another, but also to the situation in which they are used. In other words, propositional language consists of words relating to one another in a particular symbolic way wherein the words assume a syntactic relationship and, in doing so, modify the meaning of each other (Head, 1915, 14 1926). Thus, the unit of speech is not the individual word but, rather, the proposition formed by the interaCtion of all the words used (Weisenberg & McBride, 1935). It appears, then, that prOpositional speech is con- cerned with the expression of the sign (word) and object. Inherent within this interpretation is the realization that pr0positional speech represents abstract language be- cause Jackson contrasted "superior, intellectual speech" and automatic speech (such as emotional utterances, direct associational naming, highly frequent utterances) in which propositionality or rationality is minimal. Among those who share Jackson's propositional theory are Head (1915, 1926), Schuell (1950), Goldstein (1963), and Osgood and Miron (1963). With particular reference to aphasia, Hughlings Jack- son (1926) considered aphasia a loss of voluntary actions with conservation_of the more automatic. He indicated that aphasic individuals had not lost language but that they had lost the ability to use it voluntarily as a means of expressing relationships between sign (word) and object previously organized by his/her nervous sytem and those _ currently being registered from the environment. Jackson maintained that the language disturbance was part of a general impairment of the volitional ability to proposi- tionalize on expresed relationships such as in the making of a word definition. For example, the making of a word definition (like any complete language act) consists of 1.5 central processes and a response. The stimulus or input is registered and is related to "other images or associations organized in us" (Brown, 1956, 1958). The model of re- sponse may be graphic, oral, or gestural. The content of response is attributable both to characteristics of the stimulus and to the interpreting and mediating central pro- cesses. Consequently, if Jackson's prOpositional theory holds, it seems plausible that in defining a word an indivi- dual may assume either a concrete or abstract manner of thinking in his/her attempt to propositionalize language. In summary, Hughlings Jackson envisioned aphasia as a propositional thought disorder which may or may not in- volve concrete language. The aphasic individual is im- paired in the ability to synthesize verbal prOpositions, that is, to express numerous and varied word relationships. (Chapey, 1977). The impairment, in turn, may reflect an inability to abstract (Duffy, 1981). Abstract-concrete theory. Goldstein (1942) stated that the human organism was guided by conceptualization of categories, classes, and general meanings. (Goldstein fur- ther believed that human action was guided by what we "think" about what we see, hear, taste, smell, and touch. He observed that the ability to react to things in a con- ceptual manner necessitated the use of an abstract attitude. 'In this manner Goldstein differentiates between an attitude 16 necessary for the formulation of abstract concepts and sen— sory or concrete impressions of individual objects. An abstract attitude is utilized to express relation- ships between objects and events in the world. Under- standably, the expressed relationships never represent ob- jects themselves but rather the concepts which the mind has formed of the objects as a result of that autonomous activity by which the mind creates language. In contrast, in the concrete attitude, the individual is said to be stimulus-bound, i.e., capable of responding only to the immediate sense impression of an object or situation with an inability to go beyond that impression to a larger category of associated ideas or to shift to another impression. To support his theory, Goldstein (1948) used a non— verbal sorting task as a reflection of conceptual thinking or abstract thought and found that the aphasic indivi- dual's difficulty in abstraction was due to a failure to shift to new principles of sorting, by rigid adherence to grouping by physical cues like color and size and also by an inability to classify objects by nonphysical concepts such as function. While defining a word represents a ver- bal task as a reflection of conceptual thinking or abstract thought, later research (Brown, 1955) shows that there are, indeed, similar mechanisms of thought in verbal and nonverbal conceptual tasks. In particular, Brown (1955) provided support for Goldstein's (1948) finding in 17 that he also observed that the aphasic individual demon- strated difficulty in abstraction not only in nonverbal but also in verbal behavior. He concluded that both the aphasic individual's nonverbal and verbal behavior is concrete, whereas the normal person's behavior is either concrete or abstract, as determined by the requirements of the stimulus situation. According to Goldstein and Scheerer (1941), the be- havior exhibited by impaired abstract attitude is not a specific linguistic disorder separate from the individual who has it but a change in more basic orientation to language. Along these lines, Goldstein (1948) indicated that the behavior exhibited by impaired abstract attitude represented an inability to find meaning and to find words as names for a variety of objects. This means that im- pairment in abstract language may judiciously be reflected in propositional language. For instance, if one cannot abstract, one cannot symbolize or incorporate symbols into a number of different contexts and relationships. Impair- ment in abstract attitude may further inhibit the selection of choice and the ability to shift from the concrete to the abstract. It would appear, then, that the impairment in ab- stract attitude in aphasic adults may not be a loss of words but, rather, a loss of the use of words as symbols in certain generalized or more varied, conceptual situa~ tions with retention of the ability to emit the same sym- 18 bol as a'specific, concrete response to immediate "here and now" conditioning stimuli. Thus, the aphasic indivi- dual may be able to name objects in a highly structured context when the object is present because he has asso- ciated the name and the object but may be unable to de- scribe it, to discuss its use, or to identify the generic character of the word. In other words, the aphasic indi- vidual cannot use words in the categorical sense but only in relationship to one particular referent.- Goldstein also attributed impairment or loss of the abstract attitude in aphasia to a total personality change which is brought about by the cerebral insult. As a result of the impairment of the abstract attitude in aphasia, Goldstein stated the individual will have a prob- lem of integrating inner experiences and in separation of ego from the outer world. Therefore, the ability to think symbolically, without specific stimulus relationships, and to consider things which are only possibilities rather than actualities will be disturbed. Goldstein (1948) further postulated that impairment of the abstract attitude is manifested along a continuum of deficiencies ranging from complete absence of communi- cation by any means to a slightly abnormal, uneven, Spora- dic impairment. Further, the disability, aside from a change in total personality, may result in a change in volitional activity, such as that involved in conversation, in formulating code rules, in structure, and in dealing 19 with previous experience in building concepts. In con- trast, however, Eisenson (1954) and Wepman (1951) felt that this impairment of the abstract attitude was due not to a total personality change but rather to the aphasic indivi- dual's disinclination to assume the abstract attitude. Wepman and Eisenson felt that the concrete behavior of the aphasic individual was, therefore, amenable to change. In summary, Goldstein maintained that the abstract component of language in the aphasia symptom complex is impaired. Consequently, the impairment in abstract atti- tude will, in turn, damage the variety, quantity, and relevance of the aphasic individual's communication or his/her abstract verbal abilities. Divergent semantic theory. Chapey's (1977) concept of aphasia involves the linkage of thinking and memory to ver- bal production. In essence, the divergent semantic theory postulates that the critical component of verbal behavior involves a relationship to higher cognitive processes: thinking and memory. Thinking, according to Chapey, is the act or process of reasoning or of conceiving ideas; memory is the power, act, or process of remembering. Divergent semantic or word production, therefore, involves the generation of logical alternatives from given information where emphasis is upon variety, quantity, and relevance of output from the same source. It is concerned with the generation of logical possibilities, with the 20 ready flow of ideas and with the readiness to change the direction of one's responses. It involves providing ideas in situations where a proliferation of ideas occurs as a series of discrete lexical responses to a particular stimulus such as in defining a word. Such behavior neces- sitates the use_of a broad search of memory storage and the production of multiple or divergent responses to a single stimulus. Divergent semantic behavior, then, is directed toward new responses--new in the sense that the thinker was not aware of the response before he began the particular line of thought (Gowan, Demos, & Torrance, 1967). According to Chapey, aphasia is a divergent semantic impairment which involves a decrease in the aphasic indi- vidual's ability to generate logical semantic alternatives where emphasis is on variety, quantity, and relevance of output from the same source. The client's ability to be fluent or to produce a number of relevant ideas with words in response to given information is lowered. Thus, when a person with aphasia is asked to respond to a divergent task (such as "Can you list many different uses fora pen?"), the individual will demonstrate an impairment in ability to produce a number and variety of semantic responses. Therefore, divergent impairment in a spontaneous language context may be reflected by an inability to retrieve a variety of words and concepts. 21 In summary, Chapey (1977, 1981) envisioned aphasia as a deficiency in divergent thinking which is required for the production of spontaneous language. The aphasic indi- vidual's inability to generate a number and variety of semantic responses (of which word definition is one as- pect) may be attributed to an impairment in abstract at- titude. Unidimensional theory. Hildred Schuell (1964) popu- larized the concept of aphasia as a unified disorder or unidimensional phenomenon instead of resorting to all sorts of overlapping categories within the disorder. Schuell's theory of unidimensionality of aphasic di- sturbance is grounded in test and clinical data obtained from a factor analysis of the Minnesota battery of aphasia tests of 157 aphasic patients. Schuell identified five factors which could account for the pattern of aphasic be- havior. However, factor one--1anguage behavior--was the most significant factor of the five in that it accounted for 41% of the test variance. This factor, Schuell proposed, represented a general dimension of language be- havior that crosses all language modalities since 45 of the 69 tests from every section in the test battery were heavily dependent on just Factor One. Furthermore, the tests which loaded on Factor One involved kinds of inte— grations which could not be attributed to organization of motor responses or to events in outgoing pathways. Rather, 22 they involved utilization of a broad ability which was de- pendent upon higher level organizations. Consequently, it was this factor which provided evidence of aphasia as a general, rather than a modality specific, reduction of language which cuts across various language modalities such as comprehension of spoken language, speech, reading, and writing on which the various perceptual,-sensori- motor, or motor deficits may or may not be superimposed. Correspondingly, Schuell's concept of the cause of this general language breakdown reflected a broad and dynamic view of the language process and appears to en- compass far more than a language mechanism which can only generate highly learned semantic responses. She hypothe- sized that an individual is aphasic because of brain dam- age or a lesion in hiS/her brain which interferes with processing of verbal messages, that is,the analysis and integration of verbal messages. All aphasic individuals, Schuell said, show some impairment of vocabulary and of verbal retention span, with a proportionate amount of difficulty in formulating and responding to messages at some level of complexity. It is apparent that a concrete as well as an abstract language component can be inferred from the unidimensional model. Specifically, the aphasic individual's inability to retrieve and use language, to communicate a variety of ideas may be highly suggestive of a reduction in linguistic abstraction. 23 Thought progg§§_;hgg;y. Wepman (1972, 1976) took "a new look" at aphasia treatment and advocated a "nonlanguage, content-centered discussion therapy" for aphasic individuals. Wepman noted that aphasic individuals often substitute words which are associated with the actual words that they are trying to produce and that the remainder of the communi- cation effort often relates to the approximated rather than to the intended word. Wepman, therefore, seemed to suggest that limitations in abstract thought process may be the ba- sis for aphasic language impairment. Wepman viewed thought processing as the ability " . . . to symbolize both ver- bally and nonverbally at both the levels of concrete and abstract operations; to retain and recall past associae tions to presenting stimuli . . . " (p. 131). In his earlier work (1951), he indicated that the major aspect of thinking disorders in aphasia was the individual's inabil- ity to form abstract concepts. As a result, Wepman urged that remediation activities of language should not be on names (words) or other specific language activities. Rather, emphasis should be placed on content and ideas. Furthermore, because the aphasic individual displays a paucity of ideas, concreteness, and limited associations, Wepman advocated the stimulation of the aphasic subject's thought process by indirectly encouraging him/her to talk about things of interest to him/her. To this extent, .Wepman's thought-centered theory represents a reflection of 24 a proclaimed low level of abstract thinking or inability to abstract in aphasia. Regression theory. Roman Jakobson's (1968) concept of aphasia involves the assertion that aphasia represents a regression of language to a stage of language development, a kind of development in reverse. According to Whitehouse, Caramazza, and Zurif (1978), a clear statement of this relar tion between stages of language and thought and language breakdown appeared in Jakobson's Child Lapgpage, Apha- sia, and Language Universals (1968). The thesis of Jakob- son's theory is that the pattern of language dissolution in aphasics is similar, but in reverse order, to the pattern of language acquisition in children. Those aspects of lan- guage competence acquired last or, more precisely, those that are most dependent on other linguistic developments,. are likely to be the first disrupted consequent to brain damage, whereas those aspects of language competence that are acquired earliest and are thus "independent" of later developments are likely to be most resistent to the ef- fects of brain damage (p. 145). Acceptance of this regression hypothesis implies that language knowledge consists of a hierarchical or layered sequence of mental structures: at the "bottom” are the first acquired and more primitive; at the "top" the last acquired and those currently in use. This view of a tran- sitional or hierarchical relationship in language behavior 25 (i.e., the microgenesis of language) permits a number of brain-behavior interpretations. For instance, Wepman (1951), commenting on the regressive linguistic phenome- non, indicated that "cortical impairment reduces man to a more primitive level of thought, from the abstract to the concrete" (p. 191). Wepman and Jones (1961) prOposed that the varieties of aphasia could be scaled so as to corre- spond to the stages of development of speech in a child; that isithe major aphasic syndromes could be regarded as steps in a scale of regression. Rochford and William (1962, 1965) tested whether aphasic individuals' misnamings could be related to the age at which children learned the words. They found a close parallel between the number of correct responses given by the aphasic adults to the word and the age at which children had learned the names. It seems that the names first learned in childhood are the names last lost in aphasia. Rochford and William reported that the similarities between the performances of children and aphasic adults are so close that it almost seems pos- sible to speak of a "naming age" in aphasia. According to this model, the major principle of apha- sia therapy is that recovery proceeds in distinct develop- mental steps. This process should ideally demonstrate the developmental steps of word finding found in microgenetic experiments with normals. That is, therapy should start with the use of letter combinations as cues, followed by a stage where the semantic sphere emerges without being 26 able to advance toward verbalization, and, finally, the semantic sphere emerges to a more general categorical stage of the correct word. Regarding therapy in aphasia, Werner (1965) stated that . . . whereas paraphasic performance emerges through microgenetic derailment, re-educational gu1dance toward healthy language seems to be con- s1stent with deve10pmental stages observable under normal conditions of microgenesis (p. 354). Summary of related theories. The theories of aphasia-- including Jackson's propositional theory, Goldstein's abstract-concrete theory, Chapey's semantic theory, Schuell's unidimensional theory, Wepman's thought-process theory, and Jakobson's regression theory--have indicated that there are characteristic patterns of difficulty in the language of aphasic individuals. Further, these theories reveal that aphasic individuals demonstrate specific impair- ment of vocabulary performance as part of the aphasia symp- tom complex. More importantly, these theories not only provide strong support for the existence of a vocabulary deficit among aphasic individuals but also provide strong support for aphasia as a specific impairment of abstract- verbal language. An outgrowth of these theories is the contention that aphasic individuals typically retain spe- cific, concrete responses even when the responses that act as symbols for abstractions are not retained. The present study affords an Opportunity to test the abstract-concrete 27 dimension of verbal language with reference to word defi- nitions. Studies related to the abstractness-concreteness di- mension in word definitions of normal and aphasic indivi- duals are reviewed in the succeeding sections. Related Studies on Abstract-Concrete Dimen- sion of Word Definitions Word definition performance in normal children. A num- ber of studies have explored some of the possible forms verbal definitions might take in the normally intact brain. For the most part, such studies have incorporated the no- tion of qualitative definitional forms. For example, Anglin (1970) states that "it is a commonplace saying that the mind of a child is relatively 'concrete' and the mind of an adult 'abstract'" (p. 10). The words concrete and abstract are sometimes used in the sense of subordinate and superordinate. In this sense a relatively concrete mind would speak with subordinate categories, e.g., "apple,” and an abstract mind with superordinate categor- ies, e.g., "fruit." With respect to children, qualitative studies have included the work of Chambers (1904) who tested 2,922 children and young people ranging in age from five to 27 years. He asked them to define and tell what they meant by selected words; and the results showed that the defi- nitions could be classified under four headings: (1) no 28 answer, signifying absence of content; (2) wholly wrong answers; (3) vaguely right answers, i.e., having one or more correct features; and (4) correct answers. He sum- marized his findings by indicating that in the early years of life, there is an accurate knowledge of only those things which are most immediate and familiar. He also drew attention to the fact that a correct definition in the early years is "a mere outline, a framework of bare essen- tials which in later years is filled with various details" (p. 37). Kirkpatrick (1938) also indicated growth in the char- acter of the definitions attached to words. As a result of studying the word definitions of children in primary grades through college, he concluded that Descriptions which are so common in the high school and college papers are rarely or never given by children in the primary grades. The same is true of definitions by synonyms and inclusions under large terms. The younger children always define by mention of specific incident, e.g., "a chair is to sit on," "baby stands up in a chair," "a bee goes around a piazza and makes a noise." What anything can do, or what can be done to it, or with it, is of most importance in early knowledge of all things; hence we find the definitions of chil- dren expressing action and use more than any- thing else (p. 17). Marx (1928) carried out a qualitative study of the first 50 words of the Stanford-Binet (Binet, 1915) vocabu- lary on a fairly large group of children and adults and re- ported that the highest quality types of definitions in re- lation to chronological age were those of the synonym and 29 genus variety. Lowest on the scale were those definitions using illustration, use, and repetition as responses. This work confirms the earlier work by Dolch (1927). Green (1931) also qualitatively analyzed the re- sponses of 718 school children and 110 adults in 50 vocab-h ulary words (45 of which later became the Form L Vocabu- lary Test of the 1937 Revised Stanford-Binet). She worked out a method of weighting scores for each word in accor- dance with the relation between the quality of response and the developmental level of the subject. She found that children ages six and seven were characterized by "use” definitions and that young children perceived words as concrete ideas. She, therefore, concluded that the power of generalization was not yet fully developed sufficiently to make it possible for them to define orange as a "fruit" but, rather, as ”you eat it." Even though this work used a somewhat different classification from Marx (1928), Green's (1931) results followed the same trend of Marx. Feifel and Lorge (1950) gave the Binet vocabulary to 900 children aged 6-14 and analyzed their responses by means of a five-fold classification system. They found that younger children most often employed the use, descrip- tion, illustration, demonstration, inferior-explanation, and repetition types of responses, whereas older children most often used the synonym and explanation types of. 30 response. Younger children defined words more concretely, and older children more abstractly. Storck and Looft (1973) reported similar findings on word definition performance. This study replicated the research of Feifel and Lorge's (1950) into the adult and aged portions of the life span. They observed a develop- mental trend in the character of word definitions that was similar to the findings of Feifel and Lorge. The develop— mental trend was one of progression from concrete and action-oriented definitions in early childhood to more abstract and conceptual formalizations in later childhood and adolescence; apparently little change occurred in the structuring of word definitions throughout the vast remain- der of the life span. Thus, their findings indicated, as in earlier studies, that definitional performance parallels development in intellectual processes are outlined by qual- itative theories of cognition such as that of Piaget (1926, 1929, 1970). Implied in the foregoing findings is that words and quality of the definition in a well-developed language could be arranged in a hierarchy of ascending levels of increasing generality, with those serving "as labels for concrete particular things at the lowest level and those having reference to the most universal and abstract of concepts at the highest" (Watt, 1944, p. 53). This finding seems to provide a basis for Gerstein's (1949) theory that concrete or descriptive definitions of 31 words represent the lowest level of intelligence. Hence, definitions by functional attributes are replaced by ab- stract responses as the child gets older. Additional evidence for the development of more ab- stract conceptualization of words may be found in more recently reported research. Palermo and Molfese (1972), for example, reported that their results may be inter— preted as an indication of the relation between semantic development and general conceptual development of chil- dren. They stated that there are, indeed, "indications of moving from the ability to conceptualize in terms of con- crete operations to more abstract levels at about 11 or 12 years of age" (p. 424). Al-Issa (1969), Gerstein (1949), Gray and Holmes (1938), Reichard, Schneider, and Rapapport (1949), and Wolman and Barker (1965)also examined this pos- sibility and found that concrete and functional definitions of words give way to abstract definitions at about 10 to 12 years of age. Word definition performance in normal adults. Regard- ing level of word definitions in normal, older—aged adults, research has included the work of Fox (1947) who found no significant change in the qualitative responses of the 70 year old group as contrasted with the 40 year old group of subjects. Similarly, Feifel (l949)-—using groups aged 15-29, 30-49, and 50-79—-found no evidence of a decline in the quality of responses with increased age comparable to 32 the rise in quality in children. He concluded that "on the whole no clear differences appear to exist between the younger and older normals in their choice of types of re- sponses." This finding confirmed an earlier study by Green (1931) in which she administered 50 words (45 of which be- came the 1937 Terman vocabulary) to a group of 110 adults ranging in age from 19 to 84. She discovered that there was no great change in vocabulary score after the early twenties. Storck and Looft (1973) reported similar findings on vocabulary or word definition performance. They repli- cated an earlier study by Feifel and Lorge (1950) in which the latter involved 6-14 year olds by extending the age range to 66+. Like Feifel and Lorge (1950), Storck and Looft (1973) found a developmental progression, with age, occurring in the quality of word definitions as well as- in the range of vocabulary. Specifically, they found that synonyms increased in frequency through childhood and were the predominant form of response throughout adulthood; that use and description forms were rare, but were some- what more common in the youngest subjects; that explana- tiggg were slightly more frequent in the adult years; that demonstration and illustration forms were rare and indicated no age trend; and that error rate decreased through the childhood and adolescent years. Swaard (1945) matched a pOpulation of 45 university professors aged 60 to 80 against a control group of 45 33 younger professors and instructors aged 25 to 35. He con- cluded that in word knowledge on general vocabulary, the older aged are uniformly superior to the younger aged. David (1960) investigated vocabulary ability and the difficulty of the associates to be learned. He divided his elderly subjects (over 60 years of age) and his younger subjects (19 to 35 years) into three groups based upon their WAIS vocabulary scores (high, medium, and low). Then be classified his paired associates into two groups: hard and easy. David found that the more difficulty of the two groups of paired associates made for a larger decrement in performance with age than did the easier group of associates. The subjects who had higher vocabu- lary abilities performed better than those who had lower vocabulary abilities. However, the magnitude of the dif- ference between performance was similar for the older and the younger groups, i.e., the role of vocabulary level was the same. Thorndike and Gallup (1947) administered an untimed, multiple-choice vocabulary test made up of items taken from an intelligence scale to a sample of American adult voting public. They reported that age differences for the test were almost nonexistent and that it was not until the "over sixty" age group that any substantial drop in vocab- ulary score was observed. The experiemental evidence above seems to substanti- ate the contention that vocabulary performance follows a 34 progressive developmental trend not only in the vocabulary range but in the character of the definition as well from childhood to adulthood and declines little, if at all, with increasing age in normal individuals. Furthermore, analysis of the qualitative studies indicates that there. are many possible methods of approach used in defining a word, some of which are easy, others comparatively diffi- cult. Thus, the character and quality of the word defini- tion given by an individual can throw light on the concep- tual level or mode of thinking of the individual. Word definition performance in aphasic individuals. Unlike the study of word definitions in normal individuals, only a few research attempts have been made to systemati- cally study word meaning from a qualitative perspective in aphasia. One such investigation was the study of connota- tive meaning and use among aphasic individuals as it re- lates to denotative meaning (Gardner and Denes, 1973). Their purpose was to determine the aphasic individual's sensitiv- ity to the connotative or expressive meaning of abstract and concrete words. They administered a modified version of the semantic differential to 12 aphasic subjects and found that a subject's performance on the test of connota- tive comprehension correlated highly with his/her perfor-' mance on tests of denotative comprehension. They also con- cluded that the aphasic subject's comprehension of abstract nouns was as well preserved as his/her comprehension of 35 concrete nouns. This finding seems to contradict the gen- eral belief about the effects of abstract and concrete stimuli on aphasic individuals' performances (Goodglass, Hyde, & Blumstein, 1967; Gardner, 1973; Halpern, 1965; Oldfield & Wingfield, 1965; Rochford & William, 1965). In general these studies have advanced the hypothesis that the lower (i.e., concrete words) the level of word abstraction, the better the aphasic individual performed on it. Okado and Sachio (1980) investigated the relationship between anomic aphasic symptoms and word meaning. They analyzed the spoken expressions of aphasic subjects and found that repetitious speech usually occurs in word defi- tion response performance in aphasia and that some of the subjects repeated phrases again and again without paying any attention to what was said to them. ,This finding, they concluded, revealed that such word meaning disturbances in adults represented an impairment in cognitivefintellectual ability. In word definition performance, this kind of im- pairment will generally reflect an increase in Error type responses in the aphasic individual. ._ Gil (1980) also looked at word definition response performance in 41 aphasic subjects. His results revealed that the restriction of the Semantic field in the course of aphasia is greater for the connections that require a pre- vous transfer of meaning (in the case of connotative mean- ing) than it is for the semantic relations that preserve the word's pr0per (denotative) meaning. 36 Wilcox, Davis, and Leonard (1978) examined the abil- ity of 18 aphasic and normal individuals to comprehend utterances when correct interpretation required integra- tions of extra-linguistic cues in a natural setting. They used videotaped situations in which the correct interpre- tation of an utterance was not the denotative interpreta- _ tion but was the meaning conveyed by the request in a par- ticular context (connotative). The aphasic individual's performance with connotative meaning was superior to his/ her performance on denotative meaning as determined from a battery of auditory comprehension tests. They concluded that standardized tests of auditory language comprehen- sion, therefore, offer a measure of aphasic breakdown in linguistic processing but do not adequately reflect apha- sic individuals' preceptive abilities in natural communi- cative settings. This finding suggested that the aphasic individual performed better in response to a contextual- dependent conveyed (concrete) stimulus than a denotative conveyed (cognitive) stimulus. In contrast, Ammon, Moer- man, and Guleac (l977),in a study dealing with the question of whether in aphasia connotative meaning is disturbed, found that aphasic subjects have a disturbed perception of connotative meaning. They tested 26 German-speaking apha- sic subjects, 23 Dutch aphasic subjects, and 19 French- speaking aphasic subjects, ages 45-48 and age-matched con- trols. .The method used consisted of matching words to meaningless figures. It was proven that aphasic subjects 37 from different countries have a disturbed perception of connotative meaning. .In a study of word meaning by Woll, Cohen, and Kelter (1979), four groups (20 patients in each) with the follow- ing syndromes were tested and compared: Broca aphasics, Wernicke aphasics, brain-damaged nonaphasics, and chronic schizophrenics. The task was to select two out of three pictures that referred to the same object in two series. In one series, two of the three pictures showed parts of the objects. In the second series the objects appeared with typical situation backgrounds. In each series the third picture had a conspicuous physical or semantic fea- ture in common with one of the other two pictures. Aphasic and nonaphasic brain-damaged patients performed equally well when identifying objects from common situation back- grounds. However, when the pictures showed parts of ob- jects, both of the aphasic groups performed significantly worse than the brain-damaged nonaphasics. Schizophrenics did not differ significantly from any of the other three groups. It was concluded that aphasic subjects possess only limited components of concepts in making choices or interpretations. Whitehouse, Caramazza, and Zurif (1978) found dissimi- lar results. They explored the word retrieval or naming difficulties of Broca aphasics and anomic aphasics in re- lation to a recently develOped model of the normal mental lexicon that stresses (a) the importance of integrating 38 perceptual and functional information in the act of naming and (b) the inherent vagueness of conceptual categories based on such information. Ten 50-65 year old subjects were shown line drawings of various food containers varying in physical features such as height and weight. They were required to select a name for the object from a multiple- choice list (cup, bowl, glass). The Broca aphasics showed relatively normal naming profiles. In contrast, the pos- terior patients were unable to integrate perceptual and functional information and were insensitive to the fuzzy boundries between conceptual categories. These authors concluded that the results for the posterior patients re- flect an impairment in the underlying conceptual organiza- tion rather than retrieval difficulties. Another study of auditory comprehension and word mean- ing was that of Pizzamaglio and Appificiafuoco (1971). They constructed a 30 item multiple-choice test designed to evaluate the ability of aphasic subjects to understand the meaning of words in clusters of semantically similar alter- nates. Four-picture clusters containing pictures such as "hand," "foot," "leg," and "finger" were presented visually and subejcts were asked to point to the picture having the highest degree of "associative overlapping." Aphasic sub- jects made significantly lower scores than nonaphasic sub- jects; however, neither Broca and Wernicke aphasics nor Broca and amnesic subjects who were matched for severity of 39 overall comprehension impairment performed differently on the test. A set of investigations has examined the word meaning performance of aphasic individuals in a related task of word association response. One such study is that of Goldfarb and Halpern (1981). They presented a word asso- ' ciation test to a group of 32 aphasic adults and 32 nor- mal adults similar in age, sex, and education. They found that for the aphasic subjects, the percentage of categori- cal or paradigmatic word association responses increased as the level of word abstraction decreased. They con- cluded that this finding appeared to be an exclusively aphasic response, because no clear pattern of abstraction level preference emerged in the responses of the normal control group. Semenza, Denes, Luchese, and Bisacchi (1980) investi- gated conceptual structures in aphasia using a nonverbal test given to 14 Broca and Wersicke's aphasics and seven control subjects. They found that the understanding of categorical or class and thematic or contextual relation- ships was selectively impaired in aphasic individuals. Broca's aphasic subjects showed difficulty in evaluating thematic relationships and did not differ from normal con- trols as far as class relationships were concerned. The opposite occurred in Wernicke's aphasic subjects whose de- fect seemed to be selectively restricted to class relation- ships. 40 Zurif, Carmazza, Foldi, and Gardner (1979) looked at eight aphasic and four normal subjects on a word recogni- tion task. On only one trial, some words were instances on the superordinate category (thus presupposing a common abstract conceptual feature) and words could be linked thematically in the sense of forming context-dependent re- lations. Their results showed that the verbal memory lim- itations in aphasic individuals, as indicated by their response either in terms of superordinate categories or in terms of concrete, thematic relations, are only in part limited to constraints on the abstract-concrete conceptual features. With respect to aphasic individuals, given the fore- going research studies of word definition response per- formance in aphasia, it is impossible to explicate the absolute nature of the conceptual change, if any, in the quality of their response. However, a number of the studies cited did yield experimental evidence to support the contention that aphasic individuals display difficulty in their ability to communicate, to form concepts, and to reason symbolically with words. When we inquire about what it is that makes the use of verbal language, as manifested in a word definition, in aphasia more difficult in some situations than in others, it is wise to consider a whole set of influences (Lesser, 1978), some of which go beyond the realm of the word's definition. The next section ad- dresses one such influence, frequency of word occurrence, 41 which is expected to affect the aphasic individual's use of language. Stimulus Variable Affect- ing Word Definition Per- formance in Aphasic Adults Frequency of word occurrence. Word frequency is the frequency of occurrence of a word in the English language estimated on the basis of the Thorndike-Lorge word count. "Man" and "hat" are examples of frequently used words. Research findings vary in the reported difficulty of apha- sic individuals in their use and understanding of words of low and high levels of frequency. For instance, Filby, Edwards, and Seacat (1964) reported that aphasic subjects demonstrated equal difficulty on words of frequent and in- frequent occurrence. In contrast, Siegel (1959) found that when his 31 subjects were asked to respond to word stimu- li, their errors increased as word frequency level de- creased. Similarly, Gardner (1973, l974),in testing for the influence of word frequency on recognition of certain objects in 11 aphasic subjects,found that word frequency is a variable in aphasic individual's language performance. Other studies have also revealed that word frequency does not create a differential effect in aphasic indivi- duals' language performance. Marshall and Brown (1971) particularly examined the effects of word frequency upon verbal retention of aphasic adults. They constructed an experimental task of picturable nouns based on high, 42 moderate, and low frequency levels of occurrence from the Thorndike-Lorge (1964) word list. Their results revealed that degree of word frequency did not differentially af- fect aphasic subjects' vocabulary performances. Based on this finding, they concluded, like Schuell (1964), that aphasia reflects a unidimensional degree of impairment rather than a processing deficit on a particular semantic level. Lohman and Prescott (1978) and Klatt (1978) also examined the effects of word frequency on word perfor- mance of aphasic individuals and, like Marshall and Brown and others, they found that word frequency did not create a differential effect in aphasic individuals' vocabulary per- formance. Lohman and Prescott particularly stressed that -word frequency was not a variable in the performance of aphasic adults when word frequency, size, shape, color, and array of stimuli on the Revised Token Test (Vignolo & Derenzi, 1962) were equated. Klatt (1978) looked at word frequency as a variable in the definition of anomia. 'Eighteen adult mild-aphasic individuals were required to read list of words that were controlled with respect to length, frequency of occurrence, and part of speech. Results showed that part of speech. category had a significant influence on readability, whereas word frequency did not. Frequency of word occurrence in English’has also been verified in the context of word association response 43 performance of aphasic subjects. For example, Goldfarb and Halpern (1981) examined 32 aphasic subjects' perfor- mance on a word association task and found that frequent words resulted in more paradigmatic (class) responses than 'did infrequent words. This result occurred, according to Goldfarb and Halpern, because frequent words, which are learned earlier in life, are closer to the automatic level of speech (e.g., naming days of the week, naming months of the year, using profanities, etc.) and are more pre- served in aphasia. The experimental studies cited above have examined frequency of word occurrence as a contributing variable in the ease of langauge use by aphasic individuals. In gen- eral, these studies found inconclusive results relative to the effect of frequency of word occurrence on aphasic in- dividuals' language performances Thus, there appears a need for further research onfrequency of word occurrence as a variable of study in aphasia (Gardner, 1974; Lohman & Prescott, 1978; Siegel, 1959). Summary of Related Literature. The frequency of a word has been shown to vary in its effect upon aphasic indivi- duals' language performances. Such data, according to Duffy (1981), are largely the result of basic clinical and experimental research. Furthermore, Holland (1975) and Tikofsky (1971) have suggested that such data are invalu- AU able because one of the strategies,in designing aphasia treatment is to follow "leads" provided by research. The relationship between the production of a defini- tional response and language was explained in this review and related to a hierarchical or developmental progression from concrete to abstract language. Presumably, the making of a word definition by an invidual provides a measure of that subject's ability to verbalize what he means by the words tested (Moran, 1953; Cottrell, 1980). When the qua- lity of the verbalization is examined, it can be found that everyday concepts vary along a continuum ranging from concrete to abstract. At the concrete end of the continu- um lie those concepts that are defined in terms of their physical characteristics, like a particular kind of flower or make of car. Concepts that lie toward the abstract end of the concrete-abstract continuum are thought to be characteristic of idiosyncratic or mental representations.- Therefore, the definitional test provides support relative to an interesting speculation: the concrete level of functioning of aphasic individuals should be reflected in definitions of a lower conceptual level than definitions given by nornal subjects. The retrieval or expression of a word definition, then, reflects the use of words as conceptual instruments, i.e., "the ability to communicate, to form concepts, or to reason symbolically with words" (Moran, 1953). In this respect, definition of a word appears to be an intrinsic dimension of language which can be impaired in aphasia. CHAPTER III METHODS The general goal of this study was to investigate qualitative differences in the verbal definitions of selected nouns offered by aphasic and nonaphasic adults. The chapter describes the procedures that were used to achieve this purpose. Operational Defini- tion of Terms In order to assure a common frame of reference, the following terms are defined as they will be used through- out this study. Aphasia refers to a language impairment due to brain injury. Language refers to the symbolization or codification of concepts used to express ideas about the world (objects, events, and relationships) which is governed by the use of an arbitrary system of signals agreed upon by a community of language users (Bloom & Lahey, 1978). Word definition respon§g_refers to the ability to select and organize words in a meaningful way. It involves memory, motor behavior, perception, motivation, and a know- ledge of sentence strutture (Harrington and Ehrmann, 1954). 45 46 Abstract word definitions refer to concepts in terms of categories or classes that are generalizations of what one means about the word (Botwinick, 1967). The choice definition is a synonym or substitute term for the stimulus word, e.g., "car--auto" (Papania, 1954). Concrete word definitions refer to concepts in terms of Use and Description, Inferior Explanation, Repetition, and Demonstration responses that do not incorporate what one means by the object into a broader context of that which can be expressed linguistically. The concepts formed are not much more than literal recognition responses to the specific object (Botwinick, 1967; Papania, 1954). Word frequency refers to the number of times a word occurs in a sample of a fixed size (Howes & Geschwind, 1964). Attitude refers to a disposition to evaluate certain objects, actions, and situations in certain ways (Bruner, Goodnow, & Austin, 1956); attitude is a way of conceiving an object and is the mental counterpart of an object (Paris 1931). Concept refers to all the knowledge an individual possesses about a category of events or objects denoted by that word (Anglin, 1970; Litowitz, 1981); the association between a common response (often verbal) and a word; the cluster of different meanings and associations surrounding common words and ideas (Osgood, 1953). 47 Subjects and Subject Selection Criteria General criteria. The subjects of two groups of male and female adults (20 subjects per group): a group of aphasic and a group of nonaphasic individuals. There were no restrictions on race, occupation, or socio-economic status. All subjects had to be at least 45 years of age to participate in the study, and the sub- jects had to have at least a seventh grade education. Nonaphasic subjects, however, were matched as closely as possible to the aphasic subjects on the basis of age, sex, and education. The distributions of aphasic and nonapha- sic subjects with respect to age, sex, and education are summarized in Table 1. All aphasic and nonaphasic indi- viduals included in the study had spoken English as their native language since the third grade. Aphasic subjects election: group I. In addition to the above criteria, aphasic subjects were required to meet other criteria. At the time of the study, each subject (1) was or had been hospitalized for a cerebral vascular acci- dent, (2) was in at least the third post-traumatic month, and (3) had to have been diagnosed by a neurologist as hav- ing suffered left hemisphere brain damage with right hemi— plegia. To ensure that each aphasic subject was functioning at a minimal level of language skills, selected subtests of the Porch Index of Communicative Ability (PICA) (Porch, 48 Table 1 Distribution of Sex, Age, and Education of the Aphasic and Nonaphasic Group Mean Age Mean Range a Age Range Years of Group 11 Sex (Years) (Years) Educ. Educ. Aphasic 20 F (10) 64.6 . 45-83 13.4 7-21 M (10) Non- M (10) aphasic 20 F (10) 62.6 45-70 13.4 7-21 NOTE: 21 years = Ph.D., 18 years = Master's degree, 16 years = Bachelor's degree, 12 years = high school diploma, 7-11 years = actual number of years of school attendance prior to high school graduation. aNumbers in parentheses indicated the number of sub- jects for each gender. 1973, 1967) were administered. Subjects were expedted to achieve a response score of seven or higher on each sub- test. In addition, aphasic subjects, like nonaphasic subjects, were presented with and expected to recognize a picture of each of the stimulus words of the experimental task used in this study (see section on criterion task, p. 50). Nonaphasic subject selection: group II. Nonaphasic subjects were included in the present study if (1) they had no history 3’ W fi’iu W plate telephone airplane lamp cap map knife scale clock bell concrete \qumwxl 14 14‘ 12 12 overalls ambulance necklace valentine: microscope revolver bulb volcano bacon vase plate telephone NOTE: Numbers in parentheses indicate the frequency of the word's occurrence: an index of 7-17 indicates a frequency of 7-17 times per one million words; an index of "A" indicates a frequency of 50-100 times per one million words (Thorndike & Lorge, 1944). aGorman, 1961. b1 38- cannot be experienced by the senses. With respect to level of abstraction, any noun that refers to objects, materials, or persons should receive a concrete rating; any noun that refers to a concept that cannot be experi- enced by the senses should receive abstract rating (Gorman, 1961). The word ghaig can be experienced by the,senses and, therefore, is rated as concrete; the word indepen- ggggg cannot be experienced by the senses and, therefore, is rated abstract. However, for the purpose of the pre- sent study, only concrete or picturable nouns were in- cluded in the experimental word list. This constraint was necessary to reduce the possible confounding effect of lack of stimulus recognition on participants' perfor- mance S . Regarding grammatical class and ease of production, Halpern (1965a, 1965b) and Buckley and N011 (1981) are among those who have noted that aphasic individuals process nouns more easily than words in other grammatical classes, e.g., adjectives and verbs. Finally, with respect to frequency, the occurrence of the stimulus words in the English languge was checked against the Thorndike-Large (1944) word list. The levels of the frequency factor were operationally defined by Gor- man (1961) and used in this study as follows: low--a11 nouns having Thorndike—Lorge indices 7-17 occurring be- tween 7 and 17 times per one million words in the English 55 language; high--a11 nouns having Thorndike-Lorge index "A" occurring between 50 and 100 times per one million words in the English language (Gorman, 1961). The stimuli comprising the experimental word list, then, were chosen by selecting 10 concrete nouns at each of two levels of the frequency factor--1ow and high; thus, there were two treatment combinations: concrete-low (CL) and concrete-high (CH). The experimental word list used in this study can be found in Table 2 on page 53. Procedures for Administering the Experimental Task Testing procedures. Each subject was interviewed in- dividually. Prior to the experimental task, all of the subjects received the two pretests, the Porch Index of Com- municative Ability (PICA) (Porch, 1973) and the Word Recog- nition Screening Task (Schuell & Jenkins, 1961),to deter- mine his/her eligibility for participation in the study. Aphasic subjects received the criterion tasks prior to and on a different day of the experimental task. Typi- cally, the experimental task was given on the day follow- ing the administration of the criterion tasks. All sub- jects (aphasic and nonaphasic) received the experimental task in a single sitting. All subjects were tested by the investigator, a certified speech-language pathologist, using the same equipment and materials. Each subject was tested individually. Aphasic subjects were tested at home or in the various participating hospitals or clinics. Nonaphasic subjects were tested in a quiet room within the Tri-County Office on Aging or in their homes. During the testing procedure, the subject and exami- ner were seated.side-by-side. A Windsor tape recorder (model number 2244) was placed on the table, and an audio recording was made of all responses. In addition, sub- jects' re3ponses were simultaneously noted by the examiner on prepared test forms (see Appendix G). When feasible, distracting visual and auditory stimuli were removed from the testing site. In obtaining a definitional response, the verbal directions to the Experimental Task were as follows: Listen; and when I say a word, you tell me what it means. Are you ready? Apple. What is an apple? (etc.) Each of the 20 task words presented in Table 2 was given singly in a random order. There was no limit on the number of times the instructions could be given. However, if the examiner sensed that the subject, after receiving the instructions, was not fully aware of the nature and requirements of the task, the examiner still repeated the instructions and added, "Guess even if you are not sure." Thus, the actual presentation of the Experimental Task did not begin until the subjects had demonstrated that they were aware of what was expected of them and that they were ready to begin. Each response ended immediately following the subject's last utterance; that is, when the subject was no longer speaking. 57 Once the subject demonstrated awareness of what was expected, she/he was given 20 seconds to imitate.the re- sponse orally or in writing to each stimulus word. After this period, the experimenter made the transition to another stimulus word by saying, "All right, let us try something else." Scoring of regponses. The subject's audiotape- recorded definition responses were later replayed and eval- uated by the investigator and one independent observer who had no prior knowledge of the subject's background or group status. Each definition response was assigned to one of five categories as devised by Green (1931) and subsequently applied by Feifel (1949) to a qualitative analysis of "ab- normals” with the Terman Vocabulary List and Storck and Looft (1973) in their qualitative analysis of "normals" with the Stanford-Binet (Form L-M) Vocabulary Subtest. The qualitative classification system provided for five categories of definition "quality" as illustrated in Table 3. The first category is that of Synonym. The se- cond category of quality of definition provides for those definitions which give a somewhat lengthy, but essentially accurate, explanation of what the given word symbolizes. The third category provides for definitions that either describe the function of or describe the physical attri- butes of that for which the word stands. These two types of definitions are lumped together in the Use and Descrip- 58 Table 3 Qualitative Classification System of Definition Responses Quality Levels of Word Definitions Examples Synonym Category I Synonym: Join = connect Orange = a fruit Explanation Category II Explanation: Priceless = It's worth a lot of- money. Skill = being able to do some- thing well. Use, Description, and Use a and Description Category III Use: Orange = You eat it. Description: Orange = It's round. Use and Description: Bicycle = It has wheels and a handle bar. Illustration, Inferior Explanation, Repetition, and Demonstration Category IVb Illustration: Priceless = a gem Demonstration: Eyelash = subject pointed to lash Inferior Explanation: Scotch = hot Repetition: Puddle = puddle of water Error Category Wrong Definition: Orange = a vegetable Misinterpretation Regard = protects something Omits: when the word is left out Clang Association: Roar = raw; skill = skillet Repetition Without Explanation: . Puddle = puddle Incorrect Demonstration: Eyelash= points to eyebrow a and bUse and Description have been combined in the final scoring as have Demonstration and Repe- tition in keeping with Feifel's (1949) and Storck and Looft's (1973) methods. 59 tion category. The fourth category provides for essen- tially Inferior definitions. These may consist of simple Illustrations, Demonstrations, Inferior Explantions, and Repetitions of the stimulus word. The fifth category is that of Error. This may consist of Wrong Definitions, Repetition Without Explanation, Misinterpretations, Omits, Clang Associations, and Inferior Explanations. Table 3 presents a list of the categories and sample responses in each. All word meanings to stimulus words recognized by standard dictionaries were acceptable and scored 1, 2, 3, 4, 5 according to the category to which each belonged. A tally was made of as many categories given for each word by each subject. Each subject was then represented in the data by the number of definitions she/he produced and by the distribution of these responses over the categories (see Appendix H). One other judge independently classified each verba- tim word definition into one or as many of the five quali- tative categories. Any disagreements in categorization were subsequently resolved by discussion among the raters. Statistical Analysis The Statistical Package. for the Social Sciences (SPSS) was employed on the CDC Cyber 750 computer at the Michigan State University Computer Center. The frequency of re- sponses from all subjects for each of the five definitional 60 categories (Synonym, Use and Description, Inferior-type, Explanation, and Error) in each group was tallied across categories. Then the means and standard deviations for each of the five qualitative categories for each of the two groups were computed and used for further statistical analyses. The multivariate analysis of covariance LMANCOVA) was used to control for the effects of age and education on the mean scores for each of the two groups on each of the five categories of word definition. It was further used to investigate overall between-group differences on the five word noun definition categories. In order to identify the definition response variables responsible for the between- group differences that occurred, the univariate ANOVA statistical tests were applied. Follow-up analyses included the discriminant analy- sis and the Tukey t-test for pairwise multiple comparison of means. The discriminant analysis was used to find the linear combination of the five original dependent vari- ables that "best" separated the two groups by maximizing the between-group variance. The canonical variates, de- rived from the discriminant analysis, represented the relative contribution of each of the five dependent vari- ables to the separation of the two groups (Bray & Maxwell, 1982). 61 The multiple pairwise contrasts were used to test for significant within-group effects on each of the five defi- nitional categories. In order to investigate how the aphasic and nonapha- sic groups' error rates increased as a function of word frequency, a series of MANCOVA was used (Nie, Hall, & Jen- kins, 1972). Specifically, each subject's definition re- sponses were recorded over the five definitional response categories for each of the two frequency subclassifications, high and low. Consequently, one to five repeated observa- tions were obtained for the high- and low-frequency stimu- lus words separately. In all the analyses, .05 alpha level was regarded as the minimal criterion for statistical significance. The results follow in Chapter IV. CHAPTER IV RESULTS The general goal of the research was to document if and how aphasic and nonaphasic adults differ qualitatively in their definition responses to selected noun words. To accomplish this objective, the three major research ques- tions posed were: 1. Are there qualitative differences in the verbal definitions of selected nouns between aphasic and nonaphasic adults? 2. What type of definitions are observed in the aphasic group and within the nonaphasic group? 3. Are there differences between aphasic and nonaphasic groups in their word definition responses to high- and low-frequency nouns? The raw scores obtained for each of the five definitional categories (Synonym, Explanation, Use and Description, Inferior Explanation, and Error) were averaged and their means and standard deviations computed. A range of statis- tical procedures, which included multivariate analysis of variance and covariance and multiple comparison t-tests, were then applied to answer three research questions. 62 63 Table 4 Means and Standard Deviations of Definition Responses for the Aphasic and Nonaphasic Groups on Each of Five Defini- Response Categories Pooled across High- and Low-Frequency Words Aphasic Group Nonaphasic Group N = 20 N = 20 Category i SD . i SD Synonym 3.70 2.89 9.50 4.41 Explanation 3.40 4.34 5.80 4.56 Use and Description 9.55 5.30 5.25 4.67 Inferior Typea 2.85 3.34 .85 1.41 Errorb 1.20 2.36 .05 .22 NOTE: Values listed in this table are observed combined means for low- and high-frequency words. Inferior Explanation, Repetition, Demonstration, and Illustration aInferior Type bError = Wrong Definition, Repetition Without Explanation, Misinterpretation, In- correct Demonstration, Omits, Clang Association Table 4 presents the means and standard deviations for the five qualitative response categories for each group. The data show that for the 20 aphasic subjects, mean scores ranged from 1.20 on the Error category to 9.55 on the Use and Description category. For the 20 nonaphasic subjects, mean scores ranged from .05 on the Error category to 9.50 on the Synonym category. The data reveal that the Use and 64 Description types of responses were given more often than any-other type of definition by the aphasic subjects. On the other hand, nonaphasic subjects gave Synonym and Expla- nation types of responses most often. A comparison of the mean scores for the two groups re- vealed that mean scores for the aphasic group were lower _ than those of the nonaphasic group on just two definition categories, Synonym and Explanation. On the remaining cate- gories, Use and Description, Inferior-type, and Error, the mean scores for the aphasic group were higher than those of the nonaphasic group. Overall, it appears that the major distribution of definition response scores for the aphasic subjects is in the concrete category (Use and Description), whereas the major distribution of definition response scores for the nonaphasic group is in the abstract category (Synonym). Statistical Treat- ment of the Data Adjusted scores for the aphasic and nonaphasic groups. In order to reveal whether observed differences presented in Table 4 with regard to the types of qualitative defini- tions given by the aphasic and nonaphasic subjects were statistically significant, the data were subjected to the multivariate analysis of covariance (MANCOVA). 'The MANCOVA was used since observed variation in the word definition performance of the aphasic and nonaphasic groups could have been confounded by known subject differences in age and 65 education. Table 1 (see Chapter ILL p. 48) reveals that both groups differed in the range of chronological ages and years of formal education. .The age range was 45-83 for the aphasic subject and 45 to 70 for the nonaphasic subjects. The range of education level was 7 to 21 years for both the aphasic and nonaphasic subjects.' Age and education have been shown to affect the definition response performance of children (Cottrell et al., 1980; Papania, 1954) and adults (Feifel, 1949; Storck & Looft, 1973). Therefore, the sta- tistical tests applied involved the use of age and educa- tion adjusted scores. The method used to determine the extent to which the difference between word definition response performance for the two groups could be attributed to age and education involved the construction of regression equation for each variable and the prediction of a set of scores for each subject based on their correlation with age and education. These predicted scores were then subtracted from the raw scores, and the resulting group differences (i.e., the adjusted scores) on each category were used as the basis for data analysis of the aphasic and nonaphasic definition response performance in the present study (Nie et al., 1972). Differences in education, but not for age, were observed; but when the MANCOVA was repeated with education adjusted scores, differences between the aphasic and non- aphasic groups' definition responses were essentially 66 identical to those obtained with the original raw scores. Therefore, the variables of age and education could not account for or explain all differences found between or within the aphasic and nonaphasic groups. The succeeding sections describe the outcome of the specific statistical procedures that were applied to the data using adjusted scores to answer each of the three main research questions. Aphasic and Nonaphasic;Eetween§Group Differences Question one addressed group differences on the five definition response categories (Synonym, Explanation, Use and Description, Inferior-type, and Error), which have been used as an index of the abstract-concrete conceptual levels of definitions evoked by a verbal stimulus. The definitions given by each subject were scored for one of five quality levels described earlier: Synonym, Use and Description, Explanation, Inferior-type, or Error (see means and standard deviations for each group in Table 4). The research question posed was: Are there qualitative differences in the verbal definitions of selected nouns be- . tween aphasic and nonaphasic adults? The overall analysis, MANCOVA, was statistically sig- nificant: F (5,36) = 8.97; p«(.01, indicating significant differences between mean scores for the aphasic and nonapha- sic groups (see Table 5). To locate the definition 67 Table 5 Significance of Difference on Definition Response Scores on a Word Stimuli Task Between Nonaphasic and Aphasic Groups Source of Adjusteda Test Variance F-ratio p Multivariate Group Membership 8.97 .001 Analysis of Covariance Definitional Levels Univariates --Synonym 34.48 .01 --Explanation 3.52 -- --Use and Description 13.88 .01 --Inferior Typeb 7.06 .05 --Errorc 5.99 .05 aMultivariate analysis of covariance was used to control for the effects of education and age bInferior Type = Inferior Explanation, Repetition, Demonstration, and Illustration CError = Wrong Definition, Repetition Without Explanation, Misinterpretation, Incorrect Demon- stration, Omits, Clang Association response variables responsible for the significant overall effect, examination of the univariate F-ratio was necessary. Table 5 presents the results of the MANCOVA that was applied to test for significant differences between sub- ject groups' means on each response category. The data represent the combined results for high? and low-frequency words. As can be seen from Table 5, the univariate F- ratios, which were used as follow-up tests to explain group differences, revealed statistically significant 68 group effects for all categories of responses except the category Explanation. The application of the discriminant analysis as a follow-up to the MANCOVA provided further assistance in the interpretation of the significant means' results. The discriminant analysis characterized the relative contribu- tion of each of the five dependent variables to the sig- nificant differences between group means. Given discrimi- nant function coefficient weights, discriminant scores were computed for each subject to reveal the extent and direction of differences reflected by the statistical test (Bray and Maxwell, 1979) The different order in which the different conceptual levels of word definitions entered the discriminant analysis are found in Table 6. It shows that the Synonym definition responses (-.84) exhibited the greatest relationship with the canonical variables (discriminant scores); therefore, Synonym definition re- sponses were ranked as the "best" discriminator between the aphasic and nonaphasic groups and, therefore, ac- counted for the largest percent of score variance (36.1) between groups. The correlation between the original de- pendent and canonical variables and their corresponding percentage variance were ranked from high to low in the order of Use and Description, Inferior-type, Error, and Explanation (see Table 6). 69 Table 6 Correlation Between Dependent and Canonical Variables and Percent of Variance Listed in Order Specifying the Relative Variable Contribution to Group Separation Correlation Between Dependent and Percent Variable Canonical Variables Variance Synonym ' - .84994 36.1 Use and Description .52425 22.2 Inferior Typea .37396 15.9 Errorb .34445 14.6 Explanation - .26419 11.2 Note: Minus values indicate that the difference is in favor of nonaphasic group. aInferior Type = Inferior Explanation, Repetition, - Demonstration, Illustration, Definition Responses. b Error Type = Wrong Definition, Repetition Without Explanation, Misinterpretation, Incorrect Demon- stration, Omits, Clang Association Definition Re- sponses. Aphasic and Nonaphasic Within-Group Differences The second research question involved word definition responses given by each subject group without considering how the definition responses compared across the two sub- Ject groups. Specifically, the research question posed was: What types of word definitions are ob— served within the aphasic group and within the nonaphasic group? 70 Obtained differences between mean scores for each of the five definition categories served as the basis for analysis of within-group differences. Mean scores for each of the two subject groups were evaluated separately using the Tukey t-test for pair-wiSe multiple comparison of means. The experiment-wise error rate associated with mul- tiple pair-wise comparisons was controlled by using the analysis of variance prior to the Tukey t-test for signif- icance of means (Bray & Maxwell, 1979). Each subsequent t-test that reached the specified alpha level of .05 was, therefore, considered reliably significant. The t-test results, summarized in Table 7, reveal significant within-group differences among the definition categories for the aphasic and nonaphasic groups. These differences are discussed in the succeeding section. Analysis of the mean scores for the aphasic group. :n: the aphasic group the reader will recall that Use and De- scription definitions were most frequently produced; Syno- nym, Explanation, Inferior-type, and Error followed in that order (Table 4). As can be seen in Table 7, applica- tion of the t-test revealed that there are significant differences (p4<.05) between the mean frequencies of production of Use and Description and every other type. It is noteworthy that the frequencies with which Explanation and Synonym type definitions were produced did not differ. 71 Table 7 Mean Differences Between Word Definition Scores by Aphasic and Nonaphasic Subjects Pooled Across High- and Low-Frequen- cy Words Aphasic, N=20 Nonaphasic, N=20 Contrasts Difference ‘b Difference ‘t Synonym with: Explanation .30 .25 3.70 2.61** Use and Descrip- tion -5.85 -4.87** 4.25 2.96** Inferior-type .85 .70 8.65 8.36** Error 2.50 2.08* 9.45 9.57** Explanation with: Use and Descrip- tion -6.15 5.12** .55 .38 Inferior-type .55 .49 4.95 4.64** Error 2.20 1.83* 5.75 5.63** Use and Description with: Inferior-type 6.71 5.58** 4.40 4.03** Error 8.35 6.96** 5.20 4.97** Inferior-type with: Error 1.65 1.38 2.51** .80 NOTE: for significance are: 1.65; 2.33. *significant at .05 level of confidence **significant at .01 level of confidence For 95 degrees of freedom, t-values required 5% level of confidence, 1% level of confidence, Yet both were produced significantly less frequently than the Use and Description type responses, as were the frequen- cies with which the remaining two categories were produced. Analysis of the mean scores for the nonaphasic group. In the nonaphasic group, the reader will recall that syno- nym definitions were most frequently produced: Explanation, 72 Use and Description, Inferior-type, and Error followed in that order (Table 4). As can be observed in Table 7, appli- cation of the t-test revealed significant differences (p .05) between the mean frequencies of each definition type and every other type with one exception: the frequen- cies with which Explanation and Use and Description were produced did not differ significantly from each other. Noun Stimulus Differences The third and final research question of this study focused on rate of error and the conceptual level of word definitions by aphasic and nonaphasic groups on words of low and high levels of frequency. Specifically, the following research question was posed: Are there differences between aphasic and nonaphasic groups in their verbal definition responses to high- and low-frequency nouns? The means computed separately for the definition re— sponses on words of low- and high-frequency levels are pre- sented in Table 8. A test of the differences for the over- all mean scores for the two frequency levels was made using a two factor analysis of variance as found in Table 9. The overall mean score on Use and Description responses to high- frequency words (9.35) was significantly higher than the mean score of low-frequency words (5.45), revealing that Use and Description response tends to increase on words of high-frequency. On the other hand, for Synonym responses, 73 nuoocH .cowumumumnmucwmwz .soflumcmmem usonuflz QOwuwummmm .:0auflsflmmo moonz u gonna .cofiumcmmem ocm .coflumuumsHHH .cowumcmamxm HOMwasH .coflumuumsofimo .coflDMHUOmmé macaw .muwao .coaunuumcoemo noon .UonumE Amnmav m.umooa can xououm 6:0 Amvmav n “womanhoHHmMch m.wmuoq pom Hmmwom nufi3 msfimmwx :w ooswnaoo coma m>mn coaumfiuommo can own "meoz mm; mm. 3. as; mm; 348.: mm.m 36 2.1m 36 3.48.3 me; 2.x wagon low u no mo. 0.9 mo. mm. mm. om. m~.m mm.m oq.a om.m mm.m mm.m om.m om.m oo.m oammnmmcoz low u as o~.H mm. mm. mm.~ om.a mm.a mm.m om.m mo.v oe.m mm.a mm.a os.m mm. ms.~ unmmsm< Hmuoa roam 30a Hmuoe non: son Hmuoe roam son Hmuoe room 30a qmuoa swam zoo ozone .HOHH n m mmmhulquuomcH newumwuommo can mm: cofluscmammm ENcocxm, m m H m 0 O m 9 ¢ 0 aocmskum mo mam>mq so“: was 30; mo mouoz so mmsouo Dammnmmcoz was owmmnmm on» ho cm>wo mcofiuwcwmmo ouoz mo mam>mq wufiamso m>wm may no comm MOM mosam> com: m magma 74 Table 9 A Comparison of Word Definition Responses of Aphasic and Nonaphasic Groups on Words of Low and High Levels of Fre- quency df MS F P Synonym Between Groups 1 168.20 47.40 .001*** Between Treatments 1 26.45 26.05 .001*** Two-Way Interaction Group X Treatment 1 2.45 .69 .409 Explana- Between Groups 1 28.80 5.67 .069 tion Between Treatments 1 5.00 .98 .324 Two-Way Interaction Group X Treatment 1 .80 .15 .693 Use and Between Groups 1 92.45 16.37 .001*** Descrip- Between Treatments 1 76.05 13.46 .001*** tion Two-Way Interaction Group X Treatment 1 5.00 .86 .350 Inferior- Between Groups 1 20.00 11.96 .001*** (Demon- Between Treatments 1 .00 .00 .999 stration, Two-Way Interaction Illustra- Group X Treatment 1 .27 .27 .605 tion, Repetition, Inferior Explanation) Error Between Groups 1 6.61 9.23 .01** (Wrong Between Treatments 1 .11 .15 .693 Defini- Two-Way Interaction tion, Group X Treatment 1 .01 .01 .895 Clang As— sociation, Omits, In- correct De- monstration, Misinterpre- tation, Repe- tition With- out Explana- tion) NOTE: A separate two-factor analysis of variance with re- peated measures on one factor was performed for each of the five definition categories. **significant at .01 level ***significant at .05 level 75 a reverse effect for frequency was found. The mean score on Synonym responSe of low-frequency words (8.75) was com- pared with the mean synonym of high—frequency words (4.45), revealing that synonym responses tend to increase on low- frequency words. For the remaining categories, Explanation, Inferior, and Error type, the main effect of frequency proved to be nonsignificant at the 5% level, indicating that no differ- ences existed between words of high- and low-frequency levels in relation to these categories. Overall, then, it seems that the frequency variable did interact with the type of definitional response given but not with group mem- bership. For both the aphasic and nonaphasic groups, there was no systematic tendency for the subjects to give defini- tions in a particular category in respeCt to low- and high- frequency words. In order to reveal whether there was a statistical interaction between treatment (high and low-frequency) and group membership (aphasic and nonaphasic) the two-factor analysis of variance with repeated measures on one factor 'was separately applied to each definitional category. The results of the two-factor analysis of variance are summarized ianable 9. It is apparent from Table 9 that interaction between levels of frequency of occurrence and the two groups for all five definition cate- gories was not significant at the .05 level. This apparent lack of interaction between frequency of occurrence and the 76 two groups for the five definition categories is more clearly seen in Figures 1-5 which show the distribution of responses to high- and low-frequency words for each defini- tional category. Thus, the low-and high-frequency levels of words do not create a differential effect for the rate of error response by the two groups. A significant difference between low- and high-fre- quency words, however, was found for just two of the five categories. Specifically, the effect of frequency for Synonym and for Use and Description was significant (p‘<.001), as indicated in Table 9. Legend: 5.5 0---0 .0 aphasic . 0———0 nonaphasic E O O O U m ' Q‘ 2 . '\ g C \\\ “\. ' ‘\ \ O \\V Low Frequency 'High Frequency Nouns Nouns Figure 1. Distribution of Synonym Definition Response Scores for the Aphasic and Nonaphasic Groups. MEAN SCORES MEAN SCORES x... ‘i 6.0 5.5 5.0 4.5 4.0 3.5 3.5 .————‘___—_———""‘ 2.5 2.0 ‘____43 °______”_.__ 1.5 1.0 0.5 0.0 Low Frequency High Frequency Nouns Nouns Figure 2. Distribution of Explanation Definition Response Scores for the Aphasic and Nonaphasic Groups. . .,/’p I 5.0 //’ . o”’,/” 2.0 0.0 Low Frequency High Frequency Nouns Nouns Figure 3. Distribution of Use and Description Definition Response Scores for the Aphasic and Nonaphasic Groups. MEAN SCORES MEAN SCORES ____,._....—-O o———-’“""' .h— __1 Low Frequency High Frequency Nouns Nouns Figure 4. Distribution of Demonstration, Illustra- tion, Repetition, Inferior, Explanation Definition Response Scores for the Aphasic and Nonaphasic Groups. h..-—--_-* E ~—o Low Frequency High Frequency Nouns Nouns Figure 5. Distribution of Error Definition Response Scores for the Aphasic and Nonaphasic Groups. 79 Supplementary Analysis Several additional findings merit description. It appears that there is some linear relationship between the variable of education and type of word definition responses given (see Table 10). As can be seen from Table 10, there Table 10 Significance of the Relationship Between Word Definition Response Performance, Age, and Education for Aphasic and Nonaphasic Groups Covariate Definitional Level r t-value p Education Synonym .37 2.43 .05 Explanation .45 3.07 .01 Use and Description -.53 -3.94 .001 Inferior Typea -.14 - .889 -- Error -.18 -l.l4 -- Age Synonym -.10 - .655 -- Explanation -.07 - .538 7- Use and Description .21 1.55 -- Inferior Typea -.11 - .651 -- Errorb .01 .084 -- NOTE: Values listed in this table were computed before scores were corrected for age and education. aInferior Type = Demonstration, Repetition, Illus- tration, and Inferior Explanation. bError = Wrong Definition, Clang Association, Omits, Incorrect Demonstration, Misinterpretation, Repeti- tion Without Explanation. 80 is a moderate relationship between Education and Use and Description performance (r = -.53, p .001), albeit a nega- tive one. The two remaining significant variables, Synonym (r = .37) and Explanation (r = .45), are similarly related at the .05 and .01 levels, respectively. In particular, there was a significant and definite trend in both the non- aphasic and aphasic groups for the proportion of Synonym and Explanation type definitions to increase and of Use and Description type definitions to decrease as one moved up through the educational levels, i.e., from eighth grade to. college graduate and above. Inferior-type and Error defi- nition responses were not found to be linearly-related to education within the present study. Regarding age, none of the five conceptual levels of word definition was found to be significant. CHAPTER V DISCUSSION Previous studies dealing with aphasia tend to claim that the language deficit in aphasic individuals includes a loss of the ability to abstract and an increased tendency toward concreteness in verbal and nonverbal behavior (Gold- stein, 1942, 1948; Chapey, 1977, 1981). The purpose of this study was to investigate, qualitatively, aphasic indi- viduals' definitional level (i.e., conceptualization) to a word definition task relative to a concrete-abstract dimen- sion of word meaning. To achieve this objective, we used a five-fold classification system of word definitions: Syno- nyms, Explanation, Use and Description, Inferior, and Error- type. I A major finding of this study was that whereas aphasic and nonaphasic subjects used both abstract and concrete levels of conceptualization in their word definitions, the aphasic subjects used more of the concrete definitions (Use and Description, Demonstration, Illustration, Inferior Ex- planation, Repetition) and Error responses. On the other hand, the nonaphasic individuals also used the concrete de- finitions, although their most frequent responses were of the abstract type (Synonym and Explanation). The mean frequency with which Explanation-type responses was 81 82 produced by the two groups did not differ significantly. . —-—— --- For all subjects, however, low- and high—frequency words did create a differential effect in the frequency of response on two of the definition types (Synonym and Use and Description). The nature of this effect was indicated by a significant increase of Synonym responses to low- frequency words and by a significant decrease of Use and Description responses to high-frequency words. With respect to the rate of Error response, it was noteworthy that while the aphasic subjects made signifi- cantly more Error responses than the nonaphasic subjects, neither group's rate of Error responses increased significantly as a function of frequency of word occurrence. It appears that there is some linear relationship be- tween the variable education and type of responses given. In particular, the higher the level of education, the more likely were Synonym definitions offered. The theroretical and clinical implications of these ob- servations are discussed in the following sections. Theoretical’Implications The finding in the present study that aphasic subjects used both abstract and concrete levels of conceptualization in their word definitions suggests that aphasic individuals may not be sharply dichotomized in terms of the quality of their verbal ability. That is, the aphasic indi- viduals' responses appear to be distributed along a continuum. The finding in general throws doubt on the va- lidity of Goldstein's hypothesis of impairment relative to 83 degree of abstraction in aphasia. The current finding em- pirically supports the conclusions of Brown (1955), Eisen- son (1954), Siegel (1959), and Wepman (1951) that the ab- solute loss of ability to abstract was not a general charac- teristic of the aphasic individual. Even though a continuum was observed in the aphasic individual's level of conceptualization in their word definition performance, there was a tendency to use more concrete responses. This finding suggests that despite the ability to shift from one level to another, there seems to be a preferred conceptual mode . Conse- quently, we cannot dismiss the notion that there may be a concrete-abstract dimension in language performance. Further findings revealed that aphasic and nonaphasic individuals' frequency counts of Explanation type responses did not differ significantly, even though Explanation re- sponses have been traditionally classified as abstract re- sponses (Paparia, 1954). The lack of significant group differences with respect to the frequency of production of explanation definitions may be related to the stimulus. word. It has been noted in the literature (Gray & Holmes, 1938; Green, 1931; Litowitz, 1977, 1981; Terman, 1937) that a word which lends itself readily to a definition by a Synonym, an Explanation response will not be given by adults; but if it is a word for which a Synonym is inade— quate, an Explanation will be the most common type of adult response. If this were so, it suggests that aphasic 84 individuals, as well as nonaphasic individuals, might mani- fest Explanation type responses as an alternative verbal definition. Even when our findings suggests that aphasic individuals demonstrated a preferred conceptual mode in their word definition response performance, it appears that word definition response performance is, nevertheless, a complex issue. For example, no systematic frequency of pro- duction in word definition responses were observed for either group for concrete or abstract levels of conceptualization. This observation suggests that whether an individual pro— vides a certain type of definition response may depend on linguistic factors such as word frequency and extra- 1inguistic factors such as age and education. Thus, it may be difficult to classify an individual into a neat, dicho- tomous category relative to abstract or concrete dimensions of language. The lack of an increase in the rate of error responses by the two groups as a function of frequency does not lend support to Siegel's (1959) conclusion that infrequent words produced more errors than frequent words. This contradic- tory finding that infrequent words did not produce more errors than frequent words might be related to some aspect of the stimulus words that were used in this study. One such possibility is the function of the relative differ- ences in the degree of abstractness between the two frequen- cy levels of the stimulus words. Although the investigator 85 selected the frequent and infrequent words with a substan-' tial buffer zone between the two frequency levels, there is a possibility that in specifying concrete words, one is not dealing with concrete words per se but, rather, with 923- crete ideas. For example, Halpern (1965) reported that al- though names of family members are polysyllabic and fairly uncommon, they are quite easily grasped because kinship ex- periences provide the concretizing factor. Our word list, similarly, represented familiar words with numerous refer- ents. Thus, the frequency of the referents may have pro- vided the concretizing factors for the subjects of this study. Spreen (1968) demonstrated that familiar words, indeed, have more meanings than unfamiliar ones and that the number of different meanings is directly propotional to the word's frequency of usage. Therefore, it seems likely that the aphasic subject's knowledge of the meanings of the concrete words used in this study and concrete words, in general, is more likely preserved in aphasia, irrespec- tive of their frequency of occurrence. Clinical Implications On the basis of the current study, it is likely that the speech-language pathologist may find a continuum model rather than the categorical model of the abstract-concrete dimension of verbal language the most appropriate frame- work for evaluating aphasic individuals' verbal ability. This type of analysis may not only foster our general 86 understanding of language and its disorders, but it may also permit a more precise focusing of diagnosis and treatment of aphasia. The diagnostic and therapeutic im- plications that arise from the finding that a continuum rather than a categorical model of the abstract-concrete dimension in the verbal ability of aphasic individuals seems more appropriate are discussed in the following sec- tions. Diagnostic implications. Since the findings of this study support the observation that aphasic individuals are impaired in their ability to symbolize verbally at both abstract and concrete levels, it may be diagnostically ad- visable to perform a more systematic evaluation of the quality of their responses. Such an analysis may be more informative in differentiating concrete versus abstract language deficits in aphasia than traditional standardized approaches which employ the individual's vocabulary test score alone. Beyond this, generalizing from the results of the cur- rent study, speech-language pathologists may expect to find aphasic individuals relying on Use and Description attri- butes when retrieving the word's meaning in language tasks. Because this definition type implies knowledge of words, their occurrence should indicate to the clinician that the individual comprehends the target word but uses a lower level of conceptualization in expressing its meaning. 87 The use of specific cues to aid language retrievals has been stressed as an important intervention technique (Davis, 1983; Marshall, 1976). Thus, identifying the defi- nition types used by a particular individual could provide guidelines for choosing appropriate cueing and organizing techniques for that individual. Further, abstract verbal language evaluative procedures could be used to evaluate the effectiveness of an abstract verbal language therapeu- tic effort. Therapeutic implications. If aphasia is an abstract and a concrete verbal language impairment, as the results of this study seem to suggest, therapeutic effort should be directed toward minimizing both components of the impair- ment and stimulating the disrupted processes to function maximally. For example, whereas evaluation of langauge abilities might involve the discrimination between abstract and concrete dimensions of verbal language within the resi- dual speech of the aphasic individual and his/her ability to respond categorically, intervention might be directed toward facilitating a hierarchy of concept usage skills, from concrete responses to awareness of categories, to sim- ple categorization, to exhaustive categorization within groups. Along these same lines, remedial strategies could con- sist of cues directly related to the complexity of the re- sponse required. For example, a clinician, in order to 88 facilitate abstract verbal responses, may instruct an adult to think of a word that means the same as "car." Similarly frequent use of the concrete attribute category by aphasic individuals may suggest that providing subjects with func- tional attributes of the target verbal response could aid their retrieval of those words. In other words, a clini- cian could establish strategies particular to an aphasic individual's own retrieval which may aid the individual's verbal language retrieval or usage. The influence of education relative to word definition performance may provide additional insight into the design of therapy for aphasic subjects. In particular, Synonym and Explanation-type responses were observed to be posi- tively related to education, whereas Use and Description responses were negatively related to education. This ob- servation was shared by Feifel (1949) who showed that con- trol over Synonym responses grows increasingly greater dur- ing higher grades in school. Clinically, this suggests that additional factors such as the educational background of the aphasic individual should be considered as a vari- able in the desired conceptual level of the response during therapeutic assessment of aphasic individual's vocabulary performance. The findings revealed that high and low frequency words did create a differential effect for two of the defi- nition types. The nature of this effect was indicated by a significant increase of Synonym responses to low frequency 89 words and by a significant decrease of Use and Description responses to high-frequency words. Clinically, this would seem to suggest that aphasic individuals are more likely to manifest Synonym type responses when presented with low- frequency words and Use and Description type responses when presented with high-frequency words. In conclusion, the results of the present study sug- gest that a.word definition response task can be a. useful paradigm for the investigation of verbal language ability and especially of conceptual strategies used by aphasic persons. However, a note of caution is necessary, ina,word definition response performance is a very complex issue as a range of factors may interact to influence the giving of an abstract and/or concrete response at any one time. Limitations of the Study Two limitations in particular influenced the general- izability of the findings of this study: (1) the subjects selected and (2) the study's method of scoring definitional responses. These limitations are discussed below. Subjects. The criteria established for subject selec- tion (for example, the minimal level of behavioral perfor- mance required on the PICA, etc.) limited the type of apha- sic subjeCt that was included in the present study. There- fore, only a narrow segment of the aphasic pOpulation was sampled. Consequently, the results can only be generalized to subjects with similar characteristics. 90 Method of scoring definitional responses. Because of limitations of the materials which are currently available to assess abstract and concrete verbal behavior, this study was unable to categorically define aphasia as an abstract rather than a concrete verbal disorder. To establish this, one would need to construct abstract and concrete tasks calling for the production of identical final (verbal) pro- ducts. For example, hypothetically, one would need a lan- guage task that looks at gply concrete verbal definitions or only abstract verbal definitions. If subjects were capable of producing specific verbal responses using a concrete verbal thought process and unable to generate the same responses using an abstract thought process, only then could one conclude that the individual possessed an abstract verbal disorder as opposed to a concrete verbal disorder. Rather, this study explored subject ability to produce some components of abstract verbal behavior and compared these behaviors with the concrete verbal behaviors of the same subjects. As we did make judgments of sub- jects' levels of definition (i.e., conceptualization), the present investigation was also limited in that it did not evaluate subject ability to learn or acquire abstract verb- a1 behavior. The summary, conclusions, and implications for further :research are discussed in Chapter VI. CHAPTER VI SUMMARY, CONCLUSIONS, AND IMPLICATIONS FOR FURTHER RESEARCH Summary The purpose of this study was to investigate, using a qualitative scoring system, aphasic individual's level of verbalization to selected noun stimuli relative to a concrete-abstract dimension of word meaning. Three re- search questions concerning the aphasic individual's use of five qualitative types of definition responses (Synonym, Explanation, Use and Description, Inferior-type, and Error) were examined. The research questions were derived pri- marily from the hypotheses of Chapey (1977) and Goldstein (1942, 1948) that aphasia is accompanied by a loss of the ability to conceptualize on an abstract level and by an increased tendency toward the use of concrete forms of con- ceptualization. Specifically, the research questions posed were as follows: 1. Are there qualitative differences in the verbal definitions of selected nouns between aphasic and nonaphasic adults? 2. What types of word definitions are ob- served within the aphasic group 80d With' in the nonaphasic group? 91 92 3. Are there differences between aphasic and nonaphasic groups in their word definition responses to high- and low-frequency nouns? The sample consisted of 20 aphasic and 20 nonaphasic subjects ranging in age from 45 to 83 years. Each group included an equal number of males and females. The aphasic and nonaphasic subjects were equated on the variables of age, education, and sex. The aphasic individuals were se- cured from three speech and hearing facilities. The non- aphasic individuals came essentially from the Lansing, Michigan, Tri-County Office on Aging. The stimulus material for the study consisted of a 20- item experimental word list selected from Gorman's (1961) list of single nouns which has been scaled according to their parts of speech (noun), level of word abstractness (concrete and abstract), and frequency levels (high and low). The verbatim verbal definition response of each sub- ject to each of the 20 stimulus words were audiotape re- corded. The responses were replayed and evaluated by the investigator and one independent observer who had no prior knowledge of the subject's background or group status. Each definition response was analyzed, qualitatively, using a five-fold qualitative category system. The responses of the subjects were further subjected to statistical analysis using the Multivariate Analysis of Covariance (MANCOVA) and p-tests. 93 With respect to this study's research findings, re- search question one was answered affirmatively (p <.05) by comparison of between-group differences in four of the five conceptual levels of definition responses (Synonym, Use and Description, Inferior-type, and Error). The mean numbers of Explanation definitions produced by the two groups did not differ significantly from each other. Research question two was answered by examining within-group differences observed between the mean scores . for each of the two groups. We were concerned with deter- mining the order of preference for the definition types for aphasic and nonaphasic subjects. Results revealed that there were significant differences between the mean frequencies of production of each definition type for each of the two groups. It is noteworthy that examination of the’results for each of the two groups revealed generic similarities in quality but specific differences in quan- tity of the two groups' responses. With respect to research question three, the results revealed significant differences in the number of verbal definitions produced by the two groups for the two levels of frequency in the definition responses, revealing that specific responses either increased or decreased with an increase or’decrease in frequency of word occurrence. How- ever, with respect to frequency of occurrence and error response rate, the low-frequency words did not produce significantly more verbal error responses than high- 94 frequency words for either the aphasic or the nonaphasic group. Although not systematically investigated in this study, a significant linear relationship was found for education and word definition response but not for age. Conclusions On the basis of the results obtained in this study, the following conclusions are drawn: 1. From the theoretical point of view, the findings appeared to cast some doubts on the previous contention that a complete loss of abstraction is a component or apha- sia (Goldstein 1942, 1948). The aphasic individuals, like the nonaphasic individuals, operated on both concrete and abstract levels. However, whereas the aphasic individual's preferred conceptual mode was concrete, the nonaphasic in— dividual's preferred conceptual mode was abstract. 2. The clinical value of the findings of this study lies in the observation that a continuum model, rather than a categorical model, along the abstract-concrete dimension of language must be considered in evaluating an aphasic individual's level of conceptualization in verbal language. In other words, a qualitative analysis using word definition tasks may be a more informative approach than standardized vocabulary tests in focusing diagnostic and therapeutic techniques in aphasia. ' 3.. The use of definition responses to classify an 95 individual into a neat, dichotomous category relative to abstract or concrete dimensions of language has been found to be a complex issue. Whether an individual pro- vides a certain type of definition response may depend on linguistic factors such as word frequency and extra- 1inguistic factors such as age and education. 4. The word definition task, as used in this study, seems to be a useful paradigm in the qualitative analysis of aphasic individuals' verbal responses. It appears to have both diagnostic and therapeutic value in the con- sideration of aphasia. Implications for Further Research It is believed that subsequent studies of abstract and concrete verbal ability in aphasia should include aphasic subjects who display a greater range of verbal im- pairment, because generalization of our results is limited only to subjects who would have met subject selection cri- teria of this study. The results of such a study could then possibly be generalized to the aphasic population in general. 96 In addition, future research using the word defini- tion paradigm might focus on the relationship of word def— inition performance and site of lesion, e.g., frontal ver- sus posterior lesions and in left versus right hemisphere brain damaged individuals. The results of this study revealed that aphasic individuals with left hemisphere lesion gave significantly fewer abstract definitions and significantly more concrete definitions than equated non- aphasics. To investigate the specific nature of this con- nection in other samples, e.g., frontal, posterior, and right brain damaged individuals, may be a fruitful task for future research. Finally, future studies on the complexity of the re- sponse required in the vocabulary performance of aphasic individuals can be investigated as a function of other word parameters: Ovarious parts of speech, concrete and abstract levels of word abstraction and length--a11 Of which were not investigated in the present study. It seems that for the clinician employing word definition tasks to plan therapy must do so carefully, taking into consideration the notion that task performance may be re- lated to other factors besides productive language ability. For example, evidence from the present study suggests that the type of definition response given may be heavily de- pendent on education but not age. ‘Further investigation is necessary to enlighten us further regarding the nature 97 of the interaction between related factors as those cited above and spontaneous language production. APPENDICES APPENDIX A ' DESCRIPTION OF RESEARCH PROJECT FOR THE SPEECH AND HEARING FACILITY MICHIGAN STATE UNIVERSITY DEPARTMENT OF AUDIOLOGY AND SPEECH SCIENCES EAST LANSING ‘ MICHIGAN ' 48824 Description of Research Project for the Speech and Hearinngacility NAME OF FACILITY ADDRESS TELEPHONE NAME OF INVESTIGATOR IN CHARGE OF RESEARCH S. Graves Your assistance is requested in the identification of 15 aphasic individuals known to your facility for participation in a scien- tific study being conducted in the Department of Audiology and Speech Sciences at Michigan State University. This project is concerned specifically with speech and language of persons with aphasia. The results of the study will help to provide implica- tions for clinical management and potentially contribute to under- standing the basic nature of the language disordered behavior in aphasia. I have met with the Speech Pathologist of your facility and she has expressed a willingness to aid us. Our intent is to acquire _‘a list of names, addresses, and telephone numbers of individuals who have suffered left hemispheric pathology and have been judged to be aphasic. The rights of subjects will be observed and respected. In obtain- ing informed consent, we realize that it will be necessary to safeguard the rights of the human research participants. We will (a) obtain permission from the subject prior to the start of the research project, (b) obtain an agreement from the subject that s/he has voluntarily expressed a willingness to participate, and (c) protect the confidentiality of research data. We will be pleased to send you a c0py of the.research results if you desire. Thank you for your cooperation. Signed Title Date 99 MS” is n- Affimh’n‘ Arnh- [Frugal (‘AA 4 ' APPENDIX B DESCRIPTION OF RESEARCH PROJECT FOR THE TRI-COUNTY OFFICE ON AGING MICHIGAN STATE UNIVERSITY DEPARTMENT OF AUDIOLOGY AND SPEECH SCIENCES EAST LANSING ° MICHIGAN ° 48824 Description of Research Project for the Tri-County Office on Aging NAME OF FACILITY ADDRESS TELEPHONE NAME OF INVESTIGATOR IN CHARGE OF RESEARCH S. Graves Your assistance is requested in the identification of 15 normal, older individuals to serve as a control group in a scientific study being conducted in the Department of Audiology and Speech Sciences at Michigan State University. The project is concerned specifically with speech and langauge of persons with aphasia. The results of the study will help to provide implications for clinical management and potentially contribute to understanding the basic nature of the language disordered behavior in aphasia. I have met with the director of your facility, and she has inci- cated a willingness to aid us. Our intent is to acquire a list of names, addresses, and telephone numbers of individuals who are healthy, older adults. The rights of subjects will be ob- served and respected. In.obtaining informed consent, we realize that it will be necessary to safeguard the rights of the human research participants. We will (a) obtain permission from the subject prior to the start of the research project, (b) obtain an agreement from the subject that s/he has voluntarily expressed a willingness to participate, and (c) protect the confidentiality of research data. We will be pleased to send you a c0py of the research results if you desire. Thank you for your cooperation. Signed Title Date 100 APPENDIX C LETTER REQUESTING PERMISSION FOR RELEASE OF SUBJECTS' TELEPHONE NUMBERS MICHIGAN STATE UNIVERSITY DEPARTMENT OF AUDIOLOGY AND SPEECH SCIENCES EAST LANSING ° MICHIGAN ' 48824 Letter Requesting Permission for Release of Subjects' Telephone Numbers November 11, 1982 Dear : I am requesting permission to contact you by telephone to arrange a time to give you several selected tests of language usage. These tests will take approximately 45 minutes to an hour of your time. The results of the tests will be used in a study entitled "A Qualitative Analysis of the WOrd Definitions of Aphasic and Nonaphasic Subjects," which is being done as a doctoral disserta- tion in the Department of Audiology and Speech Sciences at Michi- gan State University. Your name will be excluded from the study. If you elect to provide consent to release your telephone number to me, you will be contacted to discuss this study further and to set a day and time that you can be tested. You will also be asked to sign an Informed Consent Form which permits the testing at your home or at our speech and hearing clinic. This study i3°seen to be a worthwhile project and has been re- viewed and approved by the Research Review Committee of Michigan State University. If you wish to be contacted for participation in this study, please call Ms. Silvia Graves no later than November 30, 1982. You may call 353-9206 between 9:00 am and 5:00 pm or 353-7938 after 6:30 pm. I greatly appreciate your c00peration in this study. Sincerely, Silvia A. Graves cc: Ms. Jan Jones, Speech Pathologist Sparrow Hospital, Lansing, Michigan 101 APPENDIX D I NTERVIEW _ FORM Study # Interview Form PART I: BIOGRAPHICAL DATA Name Date Residence *TIs this at home? (Circle one.) YES NO Date of birth . Sex (circle one): Male Female (mo.-day-year) - Information given by Race (check one): white non-white Education How far did you go in school? (Fill out below.) High school: grade completed Diploma? YES NO College: No. years completed Degree? BS MS PhD BA MA Spouse's Education How far did your spouse go in school? High school: grade completed Diploma? YES NO College: No. years completed Degree? BS MS PhD BA MA Occupational History What is or was your occupation? What were your dutues? Do you expect to return to work? Spouse's Occupational History What is your spouse's occupation? What are your spouse's duties? 102 103 PART II: LANGUAGE BACKGROUND Cirlce one: English only Bilingual (If English, how long have you spoken English? Age Grade If bilingual, brief language history: Describe your speech (good? normal? abnormal?) PART III: HEALTH HISTORY Have you had any history of illness such as stroke, accident or brain injury? If so, please fill in below. Nature and duration of illness How long ago was illness? Date of illness Cause of the illness? What type of lesion exists (vascular, traumatic, tumor, em- bolism, thrombosis) (please explain) What diagnostic tests were performed to determine this in- formation? By whom? (name of neurologist/physician) Hospital Address Hemiplegia (circle one): RIGHT LEFT RECOVERED ABSENT Which hand did you prefer (circle one): RIGHT LEFT Do you see normally? If not, state the reason: Do you hear normally? if not, state the reason: 104 PART IV: HISTORY OF SPEECH THERAPY Have you had any speech and language therapy since the onset of aphasia? (Circle one.) YES NO If so, give the name of therapist Address of therapist Address of therapist Type of therapy APPENDIX E INFORMED CONSENT CONTRACT Informed Consent Contract Description of Project You are asked to give your permisiion to be tested as part of a scientific research study. You will be given a test of language usage and a task of identifying and defining 20 words. I have freely consented to take part in a scien- tific study being conducted in the Department of Audiology and Speech Sciences at Michigan State University. The study has been explained to me, and I under- stand the explanation that has been given. I agree to attend all the testing sessions of the study. I understand that I am free to discontinue my par- ticipation in the study at any time. I understand that any results will be treated in strict confidence and that I will remain anonymous. Within these restrictions, I understand that general results may be presented at professional meetings and may appear in apprOpriate journals and other publications. Signed Date 105 APPENDIX F ’DESCRIPTION OF THE PICA SUBTESTS USED Subtest I IV IX XII II III VI VII VIII XI NOTE: aPorch, Alto, Description of the: PICA Subtests Used“ Output Verbal Verbal Verbal Verbal Gestural Gestural Gestural Gestural Gestural Gestural Gestural Gestural Task To discuss each test object, differentiating its primary characteristics To anme each object To say the name of each object To imitate the name of each - object To demonstrate the function of each object To demonstrate the function of each object as it is handed to subject To read each of ten cards and place it according to printed instructions near the object whose function is stated on the card To point to each object whose function is given verbally by the examiner To read each of ten cards and place it according to printed instructions near the object whose name is stated on the card To match a picture of each ob- ject with the appr0priate object To point to each object as it is named by the examiner - To match identical objects with each object Test objects include cigarette, comb, fork, key, knife, matches, pen, pencil, quarter, toothbrush. Each subtest is composed.of 10 tasks. Answers to each task are scored on a 16 point continuum accord- ing to the accuracy, respOnsiveness, completeness, promptness, and efficiency of a response. B. Porch index of communicative ability. Palo CA: Consulting Psychologists Press, 1967. 106 APPENDIX G RESPONSE SCORE SHEET Ammlvmv madam .HH Amonmmv mmm> .OH Ammummv aoomm .m Aomlovv osmoHo> .m Aweioev oasm .n Ammuvmv Hm>ao>mm .o ANmImmv mmoomouonz .m Ammammv mcflucwam> .v Aomumav momeomz .m Awaaoav mocmHsosm .m a mIv v mHkum>o .H c0wumao :oflumuumsHHH cowumfinommo sewn E>socxm one: msHszwm Iommd mamau .gOAHMHDmQOEoo was on: can Imcsamxm .soflmmfleo cowuwvmamm .dOADQAHome. .soflusuum sowumcmamxm .mmD Icoamo Down uoflummcH nuoocH .:0eu mm .mm umummdumucemes .mm .mm .ne .:0eum:medxm .sa .av .ee usonuwz soap .mm .mm .m twummmm .GOAD Iflcwmma msouz “msflmmfiz >. >H HHH HH H mcoflueocoo anme Hmcflmem Amie ANIHV * snsum muse wswz ammmm mmoum mmzommmm mwo u usage can .snmum u 6H no semen umeuuumamsuoz muo u unmet new .mnmue n he no semen umeuumoemmeda umooo 107 AnmIomo coeumosom Ammuvmv woe Ammo xom "meoz mmmoom 3 >H HHH HH H ”H03 mDHSEflUm mofluomoumo I Uwscflucoo ”Emmmm mmoow mmzommmm APPENDIX H TOTAL DEFINITION RESPONSE SCORES FOR EACH SUBJECT IN THE APHASIC AND NONAPHASIC GROUPS FOR THE FIVE QUALITY LEVELS OF WORD DEFINITIONS o o OH v m m 2 mm 0N o n h N v mH 2 cm mH H v wH o H NH m Nm NH O v m m m NH 2 we NH H H NH o H vH m an 0H 0 w n o v MH m Nb mH o o m OH m vH m mm «H o o h m m pH 2 mo mH o 0 HH m m NH 2 vm NH o o v mH v HN 2 mm HH 0 H NH N m NH m we OH N N m H m NH m vm m o H NH 0 N NH m Hm m m 0H v o o NH m on b m o m H m NH 2 mm o m m HH 0 H m 2 no m w v HH 0 H m m mm v o 0 NH H H NH 2 mm m o m mH N N NH m Hm N N o N HH m oH 2 Ho H onmnmN muHEO coHumnumsHHH coHu coHu E»: soHu xmm wm< muomnosm “:oHumHOOmm< mcmHo “:oHuHummmm ImHuommo In: nonhm Imo “coHusuumGOEmo “:oHumsmmem 0cm om: Immem loom uomuuoocH “coHumu HoHummcH CH ImumnmuchHz “coHu “coHumnumGOEmo unto» Imcmmem DDOEUHB “maze coHHHqumm “GOHUH: HOHHmmcH Imeo moons “Houum mmHmOUmado mcoHuHchmo whoa mo mHm>mH >UHHmsO m>Hm map How mmsouw OHmmcmwcoz can UHmmnm< on» pH pomnosm comm How mmuoom uncommom :oHuHsHmmo Hmuoe 110 OOOOOHOOOOOCOOOOOOOO noncommmu :oHuHchmo auHHmsg 0>Hm on» no Hm>mH moo.csnu muoe w>Hm oHsoo uncommon monsoon noncommmu oN can» whoa Hmuou msom H v m HH m m H m H H m OH O m 5 HH 0 N N 0H 0 m m OH H o m m o H mH N o m m «H o h m h o m v MH N m m m o m HH OH N HH m m H o m m o H 5 NH o w v OH o v H oH N NH o h H m m m OH 0 NH mH HN mH «H VH MH OH «H NH NH m NH mH NH HH NH NH hhhhSEhthEhZhSSEEEfi-I mo mv vm mm Hm mm Ho Ho mm mm me >0 mm on mm mm mm mm em on u MBOZ ov mm mm mm mm mm vm mm Nm om mN mN 5N 0N mN vN MN NN HN oemmema Icoz poocHusoo .mmuoom mucommmm :oHuHGHmmo Hmuoa BIBLIOGRAPHY BIBLIOGRAPHY Al-Issa, I. The development of word definition of chil- dren. Journal of Genetic Psychology, 1969, 114, 25-28. ‘ Ammon, K., Moerman, C., & Guleac, J. Aphasics' defective perception of connotative meaning of verbal items which have no denotative meaning. Cortex, 1977, 13, 453-457. Anglin, J. The growth of word meaning. Cambridge, MA: MIT, 1971. ' Anglin, J. Wprd, object, and conceptual development. New York: MIT, 1970. Bayles, K., & Boone, R. The potential for language tasks for identifying senile dementia. Journal of Speech and Hearing Disorders, 1982, 41, 218-228. Biasschi, P., Denes, G., & Semenza, C. Semantic field in aphasia: An experimental investigation on comprehen- sion of the relations of class and property. Disser- tation Abstracts International, 76:3290, l967--June, 1982. Binet, A., & Simon, T. The development of intelligence in children. Publicatiog of the training school at Vine- land, New Jersey, 1915, 316. Bloom, L., & Lahey, M. Language development and language disorders. New York: 'John Wiley and Sons, Inc., 1978. Botwinick, J. Cognitive procgsses in maturity and old age. New York: Springer, 1967. Bouman, L., & Grunbaum, A. A. Microgenesis and aphasia. In H. Weiner (Ed.), Journal of Abnormal and Social Psychology, 1925, 52, 347-353. Bray, J. H., & Maxwell, S. E. Analyzing and interpreting significant MANOVAs. Review of Educational Research, 1982, 52, 340-367; 111 112 Bricker, A., Schuell, H., & Jenkins, J. Effect of word frequency and word length on aphasic spelling errors. Journal of Speech and Hearing Research, 1964, 2, 183-192. Brown, I. Abstract and concrete behavior of dysphasic patients and normal subjects. Journal of Speech and Hearing Disorders, 1955, 29, 35-42. Brown, R. Language and categories. In J. Bruner, J. Goodnow, & G. Austin (Eds.) A study of thinking. New York: John Wiley and Sons, 1956. Brown, R. Words and things. Glencoe: Free Press, 1958. Bruner, J., Goodnow, J., & Austin, G. A stugy of thinking. New York: John Wiley and Sons, 1956. Buckley, C., & Noll, P. Lexical parameters affecting sen- tence recall by adult aphasic patients. In R. H. Brookshire (Ed.), Clinical aphasiology conference proceedings. Minneapolis: BRK Publishers, 1981. Chambers, W. How words get meaning. Pediatric Seminary, 1904, 22, 30-50. Chapey, R. Divergent semantic behavior in aphasia. Ph.D. dissertation, Teachers' College, Columbia University, 1977. Chapey, R. Language intervention strategies in adult aphasia. Baltimore: Williams and Wilkins, 1981. Cottrell, A., Montague, J., Farb, J., & Throne, J. An operant procedure for improving vocabulary definition performance in developmentally delayed children. Journal of Speech and Hearing Disorders, 1980, 55, 90-102. David, G. In S. Botwinick (Ed.), Cognitive processes in maturity and old age. New York: Springer, 1967. Davis, A. G. A survey of adult aphasia. New Jersey: Prentice-Hall, Inc., 1983. Davis, A., Myers, P., Adamovitch, B., & Yorkston, K. Treatment of right-hemisphere damaged patients. In R. H. Brookshire (Ed.), Clinical aphasiology confer- ence proceedings. Minneapolis: BRK Publishers, 1981. DeRenzi, E., & Vignolo, L. A. ~The token test: A sensi- tive test to detect receptive disturbances in apha- sics. Brain, 1962, 55, 665-678. 113 Doehring, D., & Swisher, L. Disturbances of connotative meaning in aphasia. Journal of Communication Disord- ers, 1972, 5, 251-258. Dolch, E. Reading and word meanings. Boston: Ginn, 1927. Duffy, J. Schuell's stimulation approach to rehabilita- tion. In R. Chapey (Ed.), Language intervention strategies in adult aphasia. Baltimore: Williams and Wilkins, 1981. Duffy, J., Watt, J., & Duffy, R. A strategy for investi- gating multivariate causal relationships in communi- cation disorders. Journal of Speech and Hearing Re- search, 1981, 22, 474-489. Dunn, L, M., Horton, K., & Smith, J. Peabody language development kits. Circle Pines, Minnesota: American Guidance Service, 1968. Eisenson, J. Aphasia in adults: Basic considerations. In L. Travis (Ed.), Handbook of speechgpathology. New York: Appleton-Century-Crofts, 1971. Eisenson, J. Examining for aphasia. New York: The Psy- chological Corporation, 1954. Eisenson, J. Therapeutic problems with aphasic adults. In L. Travis (Ed.), Handbook of speech pathology. New York: Appleton-Century-Crofts, 1971. Paris, E. The concept of social attitudes. ~In K. Young (Ed.), Social attitudes. New York: Holt, Rinehart, and Winston, 1931. Feifel, H. Qualitative differences in the vocabulary responses of normals and abnormals. Genetic Psychol- ogy Monographs, 1949, 55, 151-204. Feifel, H., & Lorge, I. Qualitative differences in the vocabulary responses of children. Journal of Educa- tional Psychology, 1950, 52, 1-18. Filby, Y., Edwards, A., & Seacat, G. Word length, fre- quency, and similarity in the discrimination behavior of aphasics. Journal of Speech and Hearing Research, 1963, 5, 255-261. Fox, C. Vocabulary ability in later maturity. Journal of Educational Psychology, 1947, 55, 482-492. 114 Gardner, H. The contribution of Operativity to naming capacity in aphasic patients. Neuropsychologia, 1973, 22, 213-220. Gardner, H. The naming of objects and symbols by children and aphasic patients. Journal of Psycholinguistic Research, 1974, 5, 133-149. Gardner, H., & Denes, G. Connotative judgments by aphasic patients on a pictorial adaptation of the semantic differential. Cortex, 1973, 5, 183-196. Gerstein, R. A. A suggested method of analyzing and ex- tending the use of Bellevue-Weschler vocabulary re- sponses. Journal of Consulting Psychology, 1949, 25, 366-374. Gil, R. Certain aspects of semantic disthrbances in apha- sia: The importance of the transfer of meaning and of phonological structures in the comprehension of the semantic field, Neurppsychologia, 1980, 25, 721- 725. Goldfarb, R., & Halpern, H. Word association of time altered auditory and visual stimuli in aphasia. Journal of Speech and Hearing Research, 1981, 25, 233-243. Goldman-Eisler, F. Hesitation, information, and levels of speech production. In deReuck, & O'Connor (Eds.), Disorders of language. Boston: Little, Brown and Company, 1964. Goldstein, K. Language and language disturbances. New York: Gune and Stratton, 1948. Goldstein,!L. The problem of the meaning of words based upon observation of aphasic patients. Journal of ngchology, 1942, 2, 301-316. Goldstein, K. In C. E. Osgood & M. S. Miron (Eds.), 5p- proaches to the study of aphasia. Urbana: Univer- ~sity of Illinois Press, 1963. Goldstein, K., & Scheerer, M. Abstradt and concrete behav- ior: An experimental study with special tests. Psyghological Monographs, 1941, 55. Goodglass, H., Hyde, M., & Blumstein, S. Frequency, pic- turability and availability of nouns in aphasia. 115 Gorman, A. M. Recognition memory for nouns as a function of abstractness and frequency. Journal of Experiment- al Psychology, 1961, 52, 23-29. Gowan, J., Demon, G., & Torrance, E. Creativity: Its educational implications. New York: Wiley, 1967. Gray, W. S., & Holmes, E. The development of meaning vo- cabularies in reading. Publications of the Labora- tory Schools of the University of Chicago, 1938, (6), 140. Green, H. J. A qualitative method for scoring the vocabu- lary test of the new revision of the Stanford-Binet. Master's thesis, Stanford University, 1931. Grossman, M. A bird is a bird: Making reference without and within superordinate categories. Brain and Lan- Halpern, H. Effect of stimulus variables on dysphasic verbal errors. Perceptual and Motor Skills, 1965b, _2_]-_' 291-2980 Halpern, H. Effect of stimulus variables on verbal per- severation of aphasic subjects. Perceptual and Motor Skills, 1965a, 25, 421-429. Hanson, B. R. Recall of sentence meaning in aphasic and nonaphasic adults. Journal of Communication Disor- ders, 1976, 5, 235-246. Harrington, R., & Ehrmann, J. C. Complexity of response as a factor in the vocabulary performance of schizo-‘ phrenics. Journal of Abnormal and Social Psychology, 1954, 55, 362-364. Head, H. Aphasia and kindred disorders of speech. London: Cambridge University Press, 1926. Head, H. Hughlings Jackson on aphasia and kindred affec- tions of speech. Brain, 1915, 55, 1-90. Holland, A. L. The effectiveness of treatment in aphasia. In R. H. Brookshire (Ed.), Clinical aphasiology con- ference proceedings. Minneapolis: BRK Publishers, 1975. Howes, D. H. Application of the word frequency concept to aphasia. In A. V. S. deReuck & M. O'Connor (Eds.), Disorders of language. London: Churchill, 1964. 116 Howes, D. H. Hypotheses concerning the functions of the language mechanism. In K. Salzinger & S. Salzinger (Eds.), Research in verbal behavior and some neurolo- gical implications. New York: Academic Press, 1967a. Howes, D. H. Some experimental investigations of language in aphasia. In K. Salzinger & S. Salzinger (Eds.), Research in verbal behavior and some neurological im- plications. New York: Academic Press, 1967b. Jackson, J. H. On affections of speech from disease of the brain. Brain, 1978, 2, 304-330. Jackson, J. H. On the nature of duality of the brain. In S. Taylor (Ed.), Selected writings of John Hughlinga Jackson. London: Staples Press, 1948. Jakobson, R. Child language, appasia, and phonological universals. The Hague, Netherlands: Mouton, 1968. Jenkins, J., Jiminez-Pabon, E., Shaw, R., & Sefer, J. Schuell's aphasia in adults: Diagnosis, prognosis, and treatment. Hagerstown, Maryland: Harper and Row, 1975. Kelter, S. Aphasic disorders in matching tasks involving conceptual analysis and covert naming. Cortex, 1976, 22, 383-394. .Kirkpatrick, E. A. A vocabulary test. Quoted in W. S. Gray and Eleanor Holmes, The development of meaning vocabularies in reading. Chicago: University of Chicago Press, 1938. ‘ Klatt, H. J. Readability of words as a function of their parts of speech in aphasico: A contribution to the definition of anomia. Dissertation Abstracts Interna- tional, 78:3290, 1967--June, 1982. Krauss, R. M. .Language as a symbolic process in communi- cation: A psychological perspective. In W. R. Looft (Ed.), Developmental psychologyig A book of readings. Illinois: The Dryden Press, 1972. Kreindler, A., Ghorgitta, N., & Voineseu, Il Analysis of verbal reception of a complex order with three ele- ments in aphasics. Brain, 1971, 55, 375-389. Lenneberg, E. Biological foundations of language. New York: Wiley, 1967. ' Lesser, R. Linguistic investigations of aphasia. London: Edward Arnold, 1978. 117 Litowitz, B. Hypothetical speech: A develOpmental per- spective. Journal of Psycholinguistic Research, 198L 100 Litowitz, B. Learning to make definitions. Journal of Child Language, 1977, 2, 289-304. Lohman, L., & Prescott, T. The effects of substituting "objects" for "forms" on the Revised Token Test (RTT) -performance of aphasic patients. In R. H. Brookshire (Ed.), Clinical aphasiology conference proceedings. Minneapolis: BRK Publishers, 1978. Lubinski, R. Chapey, R., & Tedesco, L. A. Concept forma- tion by aphasic adults using Kendler's reversal and nonreversal paradigm. Brip2sh Journal of Disorders of Communication, 1980, 25, 175-182. Marshall, R. C. WOrd retrieval behavior in aphasic adults. Journal of Speech and Hearing Disorders, 1976, 52, Marshall, R. C., & Brown, L. J. Effects of semantic re- latedness upon the verbal retention of aphasic indi- viduals. In R. H. Brookshire (Ed.), Clinical aphasi- ology conference proceedings. Minneapolis: BRK Pub- lishers, 1977. Marx, B. 1A°study of the first 50 words of the Stanford- Binet vocabulary. Master's thesis, Stanford Univer- sity, 1928. Milberg, J., & Blumstein, S. Lexical decision and aphasia: Evidence for semantic processing. Brain and Language, 1981, 25, 371-385. Milton, S., Wertz, R., Katz, P., & Prutting, C. Stimulus saliency in the sorting behavior of aphasic adults. In R. H. Brookshire (Ed.), Clinical aphasiology con- ference proceedings. Minneapolis: BRKqublishers, 1981. Moran, L. J. Vocabulary knowledge and usage among normal and schizophrenic subjects. Psychological Monographs, 1953, 67. Muma, J. R. Communication game: Dump and play. Journal of Speech and Hearing Disorders, 1975, 55, 296-309. Muma, J. R., & McNeil, M. R. Intervention in aphasia: Psycholinguistic perspectives. In R. Chapey (Ed.), Language intervention strategies in adult aphasia. Baltimore: Williams and Wilkins, 1981. 118 Myers, P., & Linebaugh, C. Perception of contextually conveyed relationships by right hemisphere damaged adults. Paper presented at the American Speech and Hearing Convention, Detroit, 1981. Nelson, K. Concept, word, and sentence: Interrelations in acquisition and development. Psychological Re- view, 1974, 52, 267-285. Nelson, K. The nominal shift in semantic-syntactic de- _velopment. Cognitive Psychology, 1975, 461-479. Nie, N. H., Hull, C. H., Jenkins, S. G., Steinbrenner, K., & Bent, D. H. Statistical package for the social sciences (2nd ed.). New York: McGraw-Hill, 1972. Okado, S., & Sacchio, K. The relationship between aukotic symptoms and word meaning disturbances. Journal of Child and Adolescent Psychiatry, 1980,22, 276-283. Osgood, C. E. Method and theory in experimental psychol- ogy. New York: Oxford Press, 1953. Osgood, C. E., & Miron, M. S. Approaches to the study of aphasia. Urbana: University of Illinois Press, 1963. Palermo, D. S., & Molfese, D. L. Language acquisition from age five onward. Psychological Bulletin, 1959, 25, 361-365. Papania, N. A qualitative analysis of the vocabulary re- sponses of institutionalized mentally retarded chil- dren. Journal of Clinical Psychology, 1954, 25, 361- 365. Piaget, J. Judgment and reasoning in the child. New York: Harcourt, 1929. Piaget, J. The language and thought of the child. New York: Harcourt, 1926. Piaget, J. Piaget's theory. In L. Carmichael (Ed.), Manual of ch22d psychology. New York: John Wiley and Sons, 1970, 703-732. Pizzamaglio, L., & Appicciafuoco, S. Semantic comprehen- sion in aphasia. Journal of Communication Disorders, 1971, 5, 280-288. 119 Porch, B. E. Porch index of communicative ability, vol- ume I: Theory and development. Palo Alto, CA: Con- sulting Psychologists Press, 1967. Porch, B. E. Porch index of communicative ability (PICA). Palo Alto, CA: Consulting Psychologists Press, 1973. Reichard, S., Schneider, M., & Rapaport, D. The develop- ment of concept formation in children. American Journal of Orthgpsychiatry, 1949, 25, 156-161. Rochford, G., & Williams, M. Studies in the development and breakdown of the use of names. I: The relation- ship between nominal dysphasia and the acquisition of vocabulary in childhood. Journal ngeurology, Neurosurgery, and Psychiatry, 1962, 25, 222-233. Rochford, G., & Williams, M. Studies in the development and breakdown of the use of names. VI: The effects of word frequency. Journal of NeugologyLNeurosur- gery, and Psychiatry, 1965, 25, 407-413. Santo Pietro, M. J., & Rigrodsky, S. The effects of tem- poral and semantic conditions on the occurrence of the error response and perseveration in adult aphasics. Journal of Speech and Hearing Research, 1982, 25, 184-192. Schuell, H. Clinical symptoms of aphasia. In L. Sies (Ed.), Aphasia theory andggerapy. Baltimore: Uni- versity Park Press, 1974, 87-101. - Schuell, H. Paraphasia and paralexia. Journal of Speech and Hearing Disorders, 1950, 25, 291-306. Schuell, H., & Jenkins, J. Reduction of vocabulary in Schuell, H., Jenkins, J., & Jiminez-Pabon, E. Aphasia in adults. New York: Harper and Row, 1964. Schuell, H., Jenkins, J., & Landis, L. Relationships be- tween auditory comprehension and word frequency in aphasia. Journal of Speech and Hearing Research, 1961, 5, 30-36. Semenza, C., Biasschi, P., & Lucchese, D. Selective defi- cit of conceptual structures in aphasia: Class ver- sus thematic relations. Brain and Language, 1980, 25, 243-248. 120 Siegel, G. Dysphasic speech responses to visual word stimuli. Journal of Speech and Hearing Research, 1959, 2, 152-160. Spreen, O. Psycholinguistic aspects of aphasia. Journal of Speech and Hearinquesearch, 1968, 22, 467-480. Storck, P., & Looft, W. Qualitative analysis of vocabu- lary responses from persons aged six to 60 plus. Journal of Educational Psychology, 1973, 55, 192-197. Swaard, K. Age and mental ability in superior men. Ameri- can Journal of Psychology, 1945, 55, 443-479. Terman, L. M., & Merrill, M. A. Measuring intelligence. Boston: Houghton Mifflin, 1937. Thorndike, E., & Lorge, I. The teachers' wordbook of 30,000 words. New York: Columbia University, 1944. Thorndike, R. L., & Gallup, G. H. Verbal intelligence of the American adult. Journal of Genetic Psychology, 1944, 10., 75-850 Tikofsky, R. Two studies of verbal learning of adult aphasics. Cortex, 1971, 2, 106-125. Ulatowski, H., Macaluso-Haynes, S., & North, A. Produc- tion of narrative and procedural discourse in aphasia. Brain and Languaga, 1981, 25, 345-371. Watts, A. F. The language and mental development of chil- ' dren. London: Harrap, 1944, 354. Weisenberg, T., & McBride, K. Aphasia. New York: Ronald Press, 1951. Wepman, J. Aphasia: Language without thought or thought without language. ASHA, 1976, 25, 131-136. Wepman, J. Aphasia therapy: A new look. Journal of Speech and Hearing Disorders, 1972, 52, 203-214. Wepman, J. Recovery from aphasia. New York: Ronald Press, 1951. Wepman, J., & Jones, L. V. Studies in aphasia: An ap- proach to testing. Chicago: University of Chicago Education-Industry Service, 1961. Werner, H. Microgenesis and aphasia. Journal of Abnormal and Social Psychology, 1965, 52, 347-353. 121 Whitehouse, P., Caramazza, A., & Zurif, E. Naming an aphasia: Interacting effects of form and function. Brain and Language, 1978, 5, 63-74. Wiggins, D. On sentence-sense, word sense, and difference of word sense. In D. Steinberg, & L. A. Sakobovitz (Eds.), Semantics: An interdisciplinary reader in philosophy, linguistics, and psychology. Cambridge: The University Press, 1971. Wilcox, M. J., Davis, C., & Leonard, L. Aphasics' compre- hension of contextually conveyed meaning. Brain and Williams, J., & Canter, G. On the assessment of naming disturbances in adult aphasia. In R. H. Brookshire (Ed.), Clinical aphasiology conference proceedings. .Minneapolis: BRK Publishers, 1981. Woll, G., Cohen, L., & Keeter, S. A study of concept discovery in aphasic patients. Dissertation Abstracts International, 4:5345, l967-l982. Wolman, R., & Barker, E. A developmental study of word definitions. Journal of Genetic Psychology, 1965, 107, 159-166. Zurif, E. B., & Caramazza, A. Psycholinguistic structures in aphasia. In H. Whitaker & H. A. Whitaker (Eds.), Perspectives in neurolinguistics and psycholinguis- tics: A series of monographs and treatises. New York: Academic Press, 1972. Zurif, E. B., Caramazza, A., Foldi, N., & Gardner, H. Lexical semantics and memory for words in aphasia. Journal of Speech and Hearing Research, 1979, 22, 456- 467. Zurif, E. B., Caramazza, A., Meyerson, R., & Galvin, J. Semantic feature representations for normal and apha- sic language. Brain and Language, 1974, 2, 167-188.