COHESION AND COHERENCE IN THE SPEECH OF PSYCHOPATHIC CRIMINALS by SHERRIE ELLEN WILLIAMSON B.Sc.(Hon)., University of V i c t o r i a , 1982 M.A., University of B r i t i s h Columbia, 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Psychology) We accept t h i s thesis as conforming to the required standard. UNIVERSITY OF BRITISH COLUMBIA September 1991 (c) Sherrie Williamson, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date DE-6 (2/88) - i i -Abstract This study was designed to examine the hypothesis that the speech of criminal psychopaths i s poorly integrated. Measures of cohesion ( l e x i c a l , r e f e r e n t i a l , conjunctive) and coherence (plot-units) were used to assess the degree to which independent clauses were linked together i n the personal narratives of criminal psychopaths and criminal nonpsychopaths. General deviance in communication, as measured by the Scale for Thought, Language, and Communication Disorders (Andreasen, 1980), was also assessed. A s i g n i f i c a n t number of psychopaths produced disordered communications. These communications f a i l e d on a number of l e v e l s : Psychopaths used r e l a t i v e l y few cohesive li n k s between sentences, f a i l e d to provide appropriate referents i n discourse, f a i l e d to l i n k action and r e s o l u t i o n i n s t o r i e s , and showed s i g n i f i c a n t c l i n i c a l impairment i n t h e i r a b i l i t y to communicate. The r e s u l t s suggest that e f f e c t i v e connections among speech units i n psychopaths' discourse are not as numerous as those found i n nonpsychopaths. In addition, psychopaths may suffer from a more general impairment i n communication that i s related to, among other things, discourse which has a tendancy to s l i p off track and a f a i l u r e to d i r e c t l y answer a li s t e n e r ' s questions. - i i i -Table of Contents Abstract i i Table of Contents i i i L i s t of Tables v Acknowledgements v i Introduction 1 Psychopathy 8 The Assessment of Psychopathy 8 Psychopathy's Relationship to other Constucts 9 Psychopathy and Language 10 Psychopathy and Affective Processing 16 Language and Aff e c t in Psychopaths 17 Thought Disorder 20 Cohesion: Meaning and Measurement 23 The Register of the Narratives 31 Reference Patterns 33 Coherence: Meaning and Measurement 36 Plot-unit Analysis 36 Experimental Hypotheses 40 Method 40 Subjects 40 Procedure 42 - i v -Speech Samples 42 SADS-L 44 Cohesion and Reference Ratings 46 Plot-unit Ratings 47 Thought Disorder Ratings 48 Results 52 Thought Disorder 52 Af f e c t Manipulation Check 56 Discourse Variables 57 Cohesion 62 Incompetent References 67 Coherence 68 Discussion 68 References 84 -v-L i s t of Tables Table I. Categories of Cohesion 25 Table I I . Plot-units and Their Def i n i t i o n s 49 Table I I I . Summary of Information for Thought Disorder (TD) Categories 54 Table IV. Means and Standard Deviations of the Discourse Variables for the A f f e c t i v e and Neutral Stories 59 Table V. Correlations of the Dependent Variables with Total Words, Clauses and MLU for the A f f e c t i v e and Neutral Stories 60 Table VI. Correlations of the PCL-R with the Dependent Variables for the Neutral and Affective Stories 63 Table VII. Means and Standard Deviations of the Dependent Variables for the Affective and Neutral Stories for Psychopaths and Non-psychopaths ... 65 - v i -Aknowledgements I would l i k e to express my thanks to the many people who helped me to complete t h i s research. F i r s t , I would l i k e to thank my research supervisor, Robert D. Hare, who has always been generous with h i s support and help. His ideas have formed the basis f o r much of the research I have been involved i n during my graduate career. Second, I would l i k e to thank my other two committee members, Judith Johnston, who offered invaluable suggestions during the i n i t i a l stages of t h i s project and who helped me learn about the subject of discourse, and D i m i t r i Papageorgis, who f i r s t taught me about psychopatholgy and whose h e l p f u l c r i t i s i s m s led to the f i n a l form of t h i s t h e s i s . To the r e l i a b i l i t y coders, Theresa Newlove, V a l e r i e Patterson, and Katie Strachan, my sincerest thanks for a l l the work and time you have put into t h i s project. I would also l i k e to thank Mike Laycock and Sherrie Green who c a r r i e d out the psychopathy assessments. I would also l i k e Adelle Forth to know how much I appreciated her help in organizing and arranging the l o g i s t i c s of not only t h i s study, but almost a l l of the other research I have been involved i n . Most importantly, I thank Tim. -1-Introduction The concept of psychopathy has a long and r e l a t i v e l y controversial history. The d e f i n i t i o n and conceptualization of the construct have undergone a number of changes over the l a s t two hundred years (see Pichot, 1978 for a review). I n i t i a l l y , both personality and behavioural variables were included i n the description of psychopathy; however recent d e f i n i t i o n s have often focused on i t s behavioural manifestations. Common to most c l i n i c a l descriptions are the following c h a r a c t e r i s t i c s : a lack of empathy, a f a i l u r e to form enduring r e l a t i o n s h i p s with others (lovelessness), shallow affect, a lack of remorse or g u i l t , a f a i l u r e to learn from experience, a lack of anxiety, and persistent a n t i s o c i a l behaviour that usually begins i n childhood (Cleckley, 1976; Craft, 1965; Grant, 1977; Hare, 1970; McCord & McCord, 1956). Recent investigations involving psychopathic subjects have moved away from the realm of c l i n i c a l d e s c r i p t i o n to empirical studies of the cognitive and behavioural processes that are thought to underlie the psychopathic personality. Most relevant to the present research are investigations of language processing. This work suggests that there are differences between psychopaths and nonpsychopaths i n the way they perceive and process language (Hare & McPherson, 1984; Hare & J u t a i , 1988; Hare, Williamson, & Harpur, 1988; Williamson, Harpur, & Hare, 1990; 1991) and i n the organization of t h e i r speech (Hare - 2 -& Gillstrom, 1991/ Gillstrom & Hare, 1988). I t has been proposed that these processing and production differences may be p a r t i c u l a r l y apparent for language s t i m u l i requiring deep semantic analysis or conveying a f f e c t i v e information (Hare & Gillstrom, 1991; Hare & J u t a i , 1988; Hare et a l . , 1988; Williamson et. a l , 1991, 1990). Speculation concerning these language abnormalities has led to the suggestion that the thoughts and concepts underlying psychopaths' speech may be poorly integrated or not well connected to each other (Gillstrom & Hare, 1988; Hare & Gillstrom, 1991) . One way of investigating t h i s hypothesized d e f i c i t i s through the analysis of cohesion and coherence. The present investigation examined narrative speech i n psychopaths through measures of text cohesion and coherence. These measures have been used to examine other c l i n i c a l groups that ex h i b i t communication d e f i c i t s or disorders of thought (e.g. Harvey, 1983; Rochester and Martin, 1979; Rochester, Martin, & Thurston, 1977). Measures of cohesion (Halliday & Hasan, 1976) assess the degree to which words and phrases i n a sentence form relationships with other words and phrases i n the sentence or i n other sentences to create a text. In the present study four measures of text cohesion were obtained. The f i r s t , l e x i c a l cohesion, ref e r s to the r e i t e r a t i o n of a word or phrase (e.g. I saw a dog - Then the dog saw me; a word or phrase i s repeated - 3 -i n some form). The second, r e f e r e n t i a l cohesion, r e f l e c t s the rela t i o n s h i p between a word and a previous speech unit because i t i s a pronoun, a demonstrative, or a comparative term (e.g. I saw a dog - Then the dog saw me; the i s a demonstrative r e f e r r i n g back to a dog). The t h i r d , conjunctive cohesion, lin k s two clauses together by conjunctions (e.g. I saw a dog -Then the dog saw me; then forms a temporal conjunction). The fourth cohesion variable measured i n the present study was incompetent references. This variable a c t u a l l y r e f l e c t s a f a i l u r e i n the use of r e f e r e n t i a l cohesive t i e s and also measures a f a i l u r e i n the use of the r e f e r e n t i a l " p h o r i c i t y " system. P h o r i c i t y systems refer to the structuring of utterances on the basis of what speakers assume t h e i r l i s t e n e r s know. They involve speech units that reguire previously presented information for t h e i r interpretation. For instance, consider the two independent clauses and the r e f e r e n t i a l cohesion that exists between them; 1. Jack went up the h i l l , 2. He f e l l down. It i s cle a r that he refers to Jack and that the speaker i s safe in assuming that the l i s t e n e r w i l l know that t h i s i s the case. However, sometimes speakers f a i l to provide adeguate p r i o r information for the l i s t e n e r to be able to i n f e r exactly which pri o r speech unit a pronoun, demonstrative, or comparative term i s r e f e r r i n g to. For example, 1. Jack and John went up the h i l l , 2. He f e l l down. In t h i s case i t i s unclear who i t i s - 4 -that he r e f e r s to. Individuals with language and communication d e f i c i t s often produce a r e l a t i v e l y high number of these incompetent references (e.g. Docherty, Schnur, & Harvey, 1988; Harvey, 1983) . Past studies have also shown that these communication-disordered individuals may e x h i b i t less cohesion i n t h e i r speech, or may use more r e i t e r a t i v e ( l e x i c a l cohesion) than semantic cohesive strategies (e.g. Rochester & Martin, 1979). I f psychopaths do i n fact have d i f f i c u l t y i n l i n k i n g speech u n i t s , then they might have d i f f i c u l t y i n forming cohesive texts and i n using the r e f e r e n t i a l p h o r i c i t y system e f f e c t i v e l y . Coherence, unlike cohesion, i s concerned with defining the meaning r e l a t i o n s h i p s that exist among events or propositions described within a text (e.g. Reiser & Black, 1982). The measure of coherence used in the present study i s p l o t - u n i t analysis (e.g. Botvin & Sutton-Smith, 1977; Gillam, 1989). Plot-unit analysis assesses the course of actions or events that makes up a narrative. For example, an i n i t i a l state of threat should be followed by an action to remove or deal with the threat, and then some type of resolution (successful or not successful). If the i n i t i a l state of threat i s an important part of the story being told, and i f i t i s not referred back to, then the story w i l l seem less coherent. Consider the following short narratives that were generated s p e c i f i c a l l y to the prompt, "I want you to t e l l me - 5 -about a time you f e l t r e a l l y angry. Try to remember how you f e l t then and then t e l l me what happened" (pauses and hesitations not included): 1. I was at home fe e l i n g lousy [negative f e e l i n g state] -because my f r i e n d J. ratted on me [ v i l l a i n y ] - I figured I'd get even by paying a v i s i t to h i s o l d lady [plan] - So I went to her house and scared her a b i t [plan c a r r i e d out] - She t o l d J. - and I was glad [negative f e e l i n g state n u l l i f i e d ] - cause he deserved i t - He didn't think I had i t in me - He's a f r a i d of me now [ v i l l a i n y n u l l i f i e d ] The p l o t units are: a. negative f e e l i n g state/negative feeling state n u l l i f i e d b. v i l l a i n y / v i l l a i n y n u l l i f i e d c. plan/plan ca r r i e d out 2. Angry - I was - I was angry - [negative f e e l i n g state] These tough guys were bugging me [threat] at the bar. My g i r l - f r i e n d was there. She wanted me to go with her to get a hose or something [plan]. These guys kept bugging me. My f r i e n d P. was there. He's complaining away cause I was supposed to be putting up t h i s job. They wouldn't stop so I beat on t h i s guy [attack] Then I went shopping with K. ( g i r l - f r i e n d ) [plan carried out] In t h i s narrative the plot units are: a. negative f e e l i n g state/ ? (possible -6-n u l l i f i c a t i o n ) b. threat/ ? (possible n u l l i f i c a t i o n ) c. plan/plan c a r r i e d out d. attack/? (possible counterattack, wound, k i l l , flee) If the two texts above are compared i t can be seen that the f i r s t e x hibits a more coherent des c r i p t i o n (1. i s from a nonpsychopath; 2. i s from a psychopath). This i s p a r t l y due to a difference between the texts i n the amount of closure of plot units. I f a great deal of inference i s applied to the second narrative, a reader might guess that the protagonist i n the story might have had his negative feelings n u l l i f i e d or he would not have gone shopping, or that the e n t i r e threat (these guys) was somehow n u l l i f i e d because he was even able to go shopping. A great deal more inference i s required for the second narrative because concepts have not been s p e c i f i c a l l y linked together. Consequently, the second narrative i s more vague or lacks the same degree of coherence as the f i r s t . I f psychopaths do have d i f f i c u l t y in l i n k i n g together concepts i n th e i r speech, then psychopathy should be associated with f a i l u r e s i n making these t i e s . In addition to the question of how well speech units may be linked together i n psychopaths' language there i s the larger issue of how well psychopaths actually communicate. Observers in our lab have suggested that psychopaths have d i f f i c u l t y i n -7-adhering to one t r a i n of thought, use excessive jargon, and sometimes use words they have made up. For t h i s reason dyadic speech was assessed i n psychopaths by r a t i n g t h e i r conversational language on the Scale for the Assessment of Thought, Language, and Communication (TLC; Andreasen, 1979a; 1979b). Although t h i s instrument was developed as a measure of formal thought disorder, i t s author has described i t mostly as a scale to assess language and communication d e f i c i t s from which a disorder of thought may or may not be i n f e r r e d . In the present study the TLC ratings were taken to represent a d e f i c i t i n language or communication and not thought disorder per se. For instance, the TLC includes a category of derailment. A high r a t i n g on t h i s category suggests that an i n d i v i d u a l produces speech where ideas are only obliguely r e l a t e d to previously presented ones. This may be representative of some underlying disturbance of thought, but more importantly, i t impairs communication since a li s t e n e r w i l l f i n d d e r a i l e d discourse d i f f i c u l t to understand. To be consistent with previous studies that have used the TLC in assessing communication impairments i n various psychopathological groups, the term thought disorder w i l l be used with reference to the TLC. However, thought disorder w i l l be taken to mean a d e f i c i t i n language or communication measured by c l i n i c a l r a t i n g s cales. Because high TLC scores are evident i n c e r t a i n pathological groups, a structured diagnostic interview, the -8-Schedule for A f f e c t i v e Disorders and Schizophrenia - Lifetime Version (SADS-L; Spitzer, Endicott, & Robins, 1979) was administered to diagnose functional p s y c h i a t r i c disorders which may be r e l a t e d to thought disorder. Diagnoses were made using the t h i r d e d i t i o n of the Research Diagnostic C r i t e r i a (RDC; Spitzer, Endicott, & Robins, 1989). This ensured that a l l subjects were normal with respect to the functional p s y c h i a t r i c disorders characterized by thought disorder. By analyzing the narrative speech of psychopaths through cohesion and coherence measures t h i s study was intended to t e s t the hypotheses that psychopathy i s associated with the production of speech units that are poorly connected or integrated. In addition, by obtaining an o v e r a l l measure of communication impairment in the dyadic speech of psychopaths, the hypothesis was tested that psychopaths s u f f e r from a more general d e f i c i t i n communication. Psychopathy The Assessment of Psychopathy For a number of reasons psychopaths have usually been studied i n a forensic environment. F i r s t , psychopaths are easiest to f i n d i n prisons and j a i l s . Second, assessment devices have often focused on the commission of a n t i s o c i a l acts as opposed to the i d e n t i f i c a t i o n of personality v a r i a b l e s . This trend i s r e f l e c t e d i n the d e f i n i t i o n of a n t i s o c i a l personality disorder described i n the current edition of The Diagnostic and -9-S t a t i s t i c a l Manual of Mental Disorders (DSM-III-R; American Psychiatric Association, 1987). For the research proposed here the Revised Psychopathy Checklist (PCL-R: Hare, 1991; Hare et a l . , 1990) was used to assess psychopathy. Extensive evidence att e s t s to the r e l i a b i l i t y and v a l i d i t y of the Psychopathy Checkl i s t for the assessment of psychopathy i n criminal populations (Hare, 1980, 1983, 1985b, 1990; Hare et a l . , 1990; Harpur, Hakstian, & Hare, 1988; Harpur, Hare, & Hakstian, 1989; Newman & Kosson, 1986; Wong, 1984). Unlike the DSM-III-R, the PCL-R measures two correlated constructs that are both considered important elements of the psychopathic personality (Harpur et a l . , 1989). Factor 1 includes items measuring an egocentric, c a l l o u s and remorseless use of others and defines a personality construct that c l o s e l y resembles the c l a s s i c a l c l i n i c a l description of the psychopath (e.g. Cleckley, 1976). Factor 2 encompasses items related to a chron i c a l l y unstable and a n t i s o c i a l l i f e s t y l e and resembles the diagnosis of a n t i s o c i a l personality disorder defined i n DSM-III-R. Unless otherwise noted, studies c i t e d i n the following review w i l l have used the PCL-R (or i t s o r i g i n a l version) to assess psychopathy. Psychopathy's Relationship to other Constructs Psychopathy has been found to be p o s i t i v e l y correlated -10-with scores on measures of impulsivity, machiavellianism, narcissism, and sensation seeking, but negatively related to measures of s o c i a l i z a t i o n (Hare, 1991; Harpur et a l . , 1989). Foreman (1988) found that psychopathy was p o s i t i v e l y correlated with an interpersonal s t y l e of dominance, but negatively correlated with nurturance, as measured by the Interpersonal Adjective Scales (Wiggins, P h i l l i p s , & Trapnell, 1990) . Rorschach responses made by psychopaths have been found to correlate with psychodynamic measures related to narcissism, egocentricity, low anxiety and emotional detachment (Gacono, 1990; Gacono & Meloy, 1988). Psychopathy i s also associated with diagnoses or ratings of substance abuse disorder, h i s t r i o n i c personality disorder, n a r c i s s i s t i c personality disorder, and a n t i s o c i a l personality disorder (Hart and Hare, 1989; Smith and Newman, 1990; Hemphill, Hart, & Hare, 1990). Generally, psychopathy does not overlap with mental disorders t y p i c a l l y associated with psychosis or thought disorder. Hart and Hare (1989) found that i n a forensic p s y c h i a t r i c population, patients diagnosed as psychopathic were less l i k e l y than other patients to receive a DSM-III Axis I diagnosis. Furthermore, psychopathy was negatively correlated with p r o t o t y p i c a l i t y ratings of schizophrenia and had an odds-r a t i o of l e s s than one for the diagnosis of schizophrenia. Psychopathy and Language -11-C l i n i c a l l y , i t has often been noted that psychopaths appear to be highly verbal (Cleckley, 1976). Many can be disarmingly charming and t e l l impressive s e l f - s e r v i n g s t o r i e s that lead to a r t f u l swindles (Millon, 1981). Their glibness, i n s i n c e r i t y , and a b i l i t y to deceive a l l appear to involve an e f f e c t i v e use of language (Hare et a l . , 1988). However, observations made by Hare and his colleagues suggest a d i f f e r e n t picture of psychopaths' language (Hare & Gillstrom, 1991; Hare, Forth & Hart, 1989). After viewing hundreds of hours of videotaped interviews with psychopathic and nonpsychopathic criminals, and conducting such interviews themselves, observers i n Hare's lab have often suggested that psychopaths appear to use excessive jargon, poorly integrated phrases, and that they seem to have d i f f i c u l t y adhering to one t r a i n of thought. This has led to the suggestion that there i s a poor integration of the thoughts and concepts underlying the speech of psychopaths (Hare and Gillstrom, 1991). Currently, there are no studies which d i r e c t l y address these issues, but there are some which offe r results suggestive of differences between psychopaths and others in the perception and production of speech. E i c h l e r (1965) analyzed the speech of young adult males who met the c r i t e r i a for the American P s y c h i a t r i c Associations' (1952) category of sociopathic disorder. He found that sociopaths scored higher than did normals on the categories of - 1 2 -negation, r e t r a c t i o n , q u a l i f i e r s , and evaluation. Q u a l i f i e r s (suppose, more or less, etc.) tend to add ambiguity to statements. Retractors - a word or phrase that p a r t i a l l y or t o t a l l y detracts from the statement preceding i t - also reduce language c l a r i t y . E i c h l e r ' s r e s u l t s suggest that there may be a basic conceptual inconsistency and i l l o g i c a l i t y i n sociopathic thought. For example, to the question, "Did you ever s t e a l from them {foster family}?", a psychopath studied i n our lab re p l i e d , " I t wasn't r e a l l y too bad. Not too often. Once i n a while I'd take some pillows or something. But I wouldn't be ste a l i n g . I'd just take them and use them and lose them or something. 1 1 In a previous study, Weintraub and Aronson (1962) found that i n the speech of normal young adult males, evaluators (a value judgement) and q u a l i f i e r s , r e t r a c t o r s , and q u a l i f i e r s , and negators (no, not, nothing, etc.) and r e t r a c t o r s were highly correlated. Therefore t h e i r co-occurrence i n the speech of psychopaths would not be considered unusual. However, excessive use of q u a l i f i e r s , retractors and negators would re s u l t i n speech f i l l e d with the following type of statement, "I guess that was good, but i t r e a l l y wasn't." This would provide less than optimal information for the l i s t e n e r as i t i s conceptually inconsistent and ambiguous i n meaning. A second i n d i r e c t method of assessing the a b i l i t y of -13-psychopaths to li n k conceptual units i n t h e i r speech i s to study the hand gestures they make while speaking. The rationale for t h i s research i s that gestures may have t h e i r basis i n the same i n t e r n a l processes as speech, and may be viewed as an external representation of these processes (e.g. Butterworth & Beattie, 1976; Cicone, Wapner, F o l d i , Z u r i f , & Gardner, 1979; Gillstrom & Hare, 1988; McNeill, 1985). Gi l l s t r o m and Hare (1988) assessed the degree to which psychopaths made use of diff e r e n t types of hand gestures while involved i n conversation. They found that psychopaths, r e l a t i v e to nonpsychopaths, made inordinate use of a p a r t i c u l a r type of hand gesture, referred to as a beat. Beats are small rapid hand movements that occur during speech or pauses i n speech. Beats may r e f l e c t the degree to which discourse i s broken down into functionally discreet units (McNeill & Levy, 1982), and may mark meta-linguistic points in the breakdown of speech processes, possibly r e f l e c t i n g an attempt to r e i n s t a t e speech flow (McNeill, 1985). Gillstrom and Hare (1988) suggested that the overuse of beats may mean that the cen t r a l language processes of psychopaths are organized into r e l a t i v e l y small conceptual units. This would be r e f l e c t e d i n speech which i s made up of short, poorly integrated phrases. Although there have been no empirical studies that have investigated the speech output of psychopaths, a number of studies have looked at the way i n which psychopaths d i f f e r from -14-nonpsychopaths i n the processing of simple verbal input. These investigations have been summarized i n a recent review a r t i c l e (Hare et a l . , 1988), but w i l l be b r i e f l y described here since they suggest that psychopaths may perceive or process language d i f f e r e n t l y than do nonpsychopaths. In general, these investigations involved the study of perceptual asymmetries with divided v i s u a l f i e l d or d i c h o t i c l i s t e n i n g techniques. Typically, when single words are presented to both ears simultaneously, or a l t e r n a t e l y to the l e f t or r i g h t v i s u a l f i e l d , most right-handed i n d i v i d u a l s exhibit a r i g h t ear or a right v i s u a l f i e l d advantage for the report of the s t i m u l i they have heard or seen. This i s due to the neurophysiology of the vis u a l and auditory systems coupled with control of language functions by the l e f t hemisphere i n most i n d i v i d u a l s . In the auditory system c o n t r a l a t e r a l projections are dominant, with the r e s u l t that information presented to the r i g h t ear projects most strongly to the l e f t hemisphere. Information presented to the r i g h t v i s u a l f i e l d projects to the l e f t hemisphere from the r i g h t nasal hemiretina and the l e f t temporal hemiretina. Information transmitted i n i t i a l l y to the l e f t hemisphere i s better perceived and reported, presumably r e f l e c t i n g more e f f i c i e n t or p r e f e r e n t i a l processing of l i n g u i s t i c materials by that hemisphere. Using a standard dichotic l i s t e n i n g paradigm, Hare and McPherson (1984) found that psychopaths showed a smaller r i g h t --15-ear advantage (less l a t e r a l i z a t i o n ) than did nonpsychopaths. Despite t h i s reduced l a t e r a l i t y the o v e r a l l performance of psychopaths was as good as that of nonpsychopaths. In a divided v i s u a l f i e l d task where subjects saw common words flashed b r i e f l y to either the l e f t or r i g h t v i s u a l f i e l d , psychopaths and nonpsychopaths both exhibited a r i g h t v i s u a l f i e l d - l e f t hemisphere advantage (Hare, 1979). However, i n a divided v i s u a l - f i e l d task that involved d i f f e r i n g l e v e l s of semantic categorization, l a t e r a l i z a t i o n differences between psychopaths and nonpsychopaths were obtained (Hare & J u t a i , 1988). S p e c i f i c a l l y , when words were required t o be matched on the basis of morphological features, or whether or not they were a member of a concrete category, psychopaths and nonpsychopaths both exhibited the fewest errors for s t i m u l i presented to the r i g h t v i s u a l - f i e l d ( l e f t hemisphere). When the subjects had to match words to a superordinate abstract category, which presumably required a greater degree of semantic processing, psychopaths made the fewest errors for sti m u l i presented to the l e f t v i s u a l - f i e l d ( r i g h t hemisphere) whereas nonpsychopaths continued to perform i n the expected manner. The most general conclusions that can be drawn from these studies i s that the l e f t hemisphere of psychopaths i s not strongly s p e c i a l i z e d for language or that the resources for language processing i n the l e f t hemisphere are r e l a t i v e l y -16-limited i n psychopaths (Hare et a l . , 1988). More s p e c i f i c a l l y , psychopaths may have d i f f i c u l t y in the i n t e r p r e t a t i o n of l e x i c a l items within the l e f t hemisphere. None of the tasks in the reviewed studies could be considered to be d i f f i c u l t r e l a t i v e to the requirement that an individual produce a cohesive and coherent narrative of some l i f e event. In the research presented here psychopaths were required to produce such a text. Psychopaths, when compared to nonpsychopaths, were expected to r e l y more on simple l e x i c a l relationships than semantic and l o g i c a l ones i n producing cohesive texts. Psychopathy and A f f e c t i v e Processing C l i n i c i a n s have long noted that psychopaths seem to suffer from a general poverty of a f f e c t . This defect has generally been described as a shallowness of f e e l i n g (e.g. Cleckley, 1976; Grant, 1977), r e f l e c t e d behaviourally i n l a b i l e and short-lived expressions of emotion. "But mature whole hearted anger...deep joy, and genuine despair are reactions not l i k e l y to be found within t h i s scale" (Cleckley, 1976, p. 212). Using experimental paradigms Hare and h i s colleagues (e.g. Hare, 1978; Hare & Craigen, 1974) have demonstrated that psychopaths show unusual autonomic responses following a sig n a l of an impending aversive event. Psychopaths co n s i s t e n t l y show smaller than usual skin conductance responses but larger than usual heart rate acceleration. Hare (1978, 1982) has suggested -17-that t h i s may r e f l e c t an adaptive coping mechanism whereby the heart rate acceleration helps psychopaths to reduce the impact of the forthcoming aversive events. The small electrodermal responses are seen as evidence for the success of t h i s coping strategy. Patrick and Lang (1989) found that psychopathic sex offenders gave smaller autonomic responses during imagery of f e a r f u l material than did other sex offenders. Psychopaths also f a i l e d to exhibit a reduction i n the b l i n k - s t a r t l e r e f l e x while viewing s l i d e s with p o s i t i v e a f f e c t i v e content and also f a i l e d to exhibit appropriate f a c i a l muscle responses to s l i d e s with negative content. Generally, these studies would seem to support the notion that psychopaths, at least autonomically, show anomalous or reduced responses to a f f e c t i v e information. Language and Aff e c t i n Psychopaths Cleckley (1976) suggested that psychopaths suffer from a deep-seated semantic disorder which he termed semantic dementia. At i t s core lay a complete lack of meaning-related elaborative a f f e c t i v e processes. This d e f i c i t was then "masked" by well functioning expressive and receptive processes. "Here i s the spectacle of a person who uses a l l the words that would be used by someone who understands, and who could define a l l the words but who i s s t i l l blind to the meaning." (Cleckley, 1976, p.214). Others have described the psychopath i n a s i m i l a r way: "... ideas of mutuality of sharing and of i d e n t i f i c a t i o n -18-are beyond h i s understanding in an emotional sense; he knows only the book meaning of words" (Grant, 1977. p. 50). Cleckley's description of semantic dementia suggests that psychopaths suffer from an i n a b i l i t y to r e l a t e r e a l feelings to words. Although there have been a variety of studies examining language processing i n psychopaths, r e l a t i v e l y few investigations have e x p l i c i t l y examined the processing of a f f e c t i v e l i n g u i s t i c information. One recent study (Williamson, et a l . , 1990) examined the responses of psychopaths to emotional and neutral words during a divided v i s u a l f i e l d l e x i c a l decision task. The e x p l i c i t task involved distinguishing between words and pronounceable nonwords, but the experimental manipulation of interest was whether or not the decision time and event-related brain p o t e n t i a l s (ERPs) would be d i f f e r e n t for emotional and neutral words. Consistent with the l i t e r a t u r e , nonpsychopaths responded more quickly to the emotional than to the nonemotional words, and amplitudes of s p e c i f i c ERP components were larger in response to emotional words than to neutral words. Psychopaths f a i l e d to show these eff e c t s , suggesting that they carried out less semantic elaboration of the a f f e c t i v e words than did the nonpsychopaths. Whether t h i s r e f l e c t e d a f a i l u r e to integrate a f f e c t i v e and l i n g u i s t i c processes, or a general f a i l u r e of elaborative processes could not be determined from t h i s study. -19-Another recent investigation into the processing of l i n g u i s t i c a l l y based a f f e c t i v e information provides some evidence that i t i s the integration functions that are impaired i n psychopaths (Williamson et a l . , 1991). Experiment 1 investigated the basis for the way i n which psychopaths group words. Using a method described by Brownell, Potter, and Michelow (1984) , subjects were required to i n d i c a t e which two out of three words were most similar i n meaning. Psychopaths were less l i k e l y to match words on the basis of emotional p o l a r i t y (both words have a positive or negative a f f e c t i v e tone) than were nonpsychopaths. Osgood (Osgood, May, & Miron, 1975) has shown that the evaluative or good-bad dimension accounts for most of the variance individuals use when rating words. Psychopaths appeared to make less use of t h i s basic component of word meaning than did nonpsychopaths. Taking t h i s observations into account, i n the second experiment of Williamson et a l . (1990), i t was hypothesized that i f psychopaths have a d e f i c i t i n making use of information concerning emotional p o l a r i t y then they would tend to make errors suggestive of a confusion of emotional valence. Other psychopathological groups, most notably depressives (Williamson, Crockett, Hurwitz, & Remick, i n press) and schizophrenics (Cramer, Weegman, & O'Neil, 1989), have previously been shown to confuse emotional valence. -20-Tests were constructed which required subjects to match clauses or pictures on the basis of inf e r r e d emotional information. Errors could be made in a number of ways including the matching of emotions of opposite valence. A s i m i l a r l y constructed p i c t o r i a l task was included to assess the s p e c i f i c i t y of t h i s e f f e c t to l i n g u i s t i c information. An analysis of the types of errors made supported the hypothesis that psychopaths show a tendency to confuse emotions of opposite p o l a r i t y , p a r t i c u l a r l y for the l i n g u i s t i c task. The authors suggested that the results across the p i c t u r e and sentence tasks may have indicated that psychopaths have d i f f i c u l t y i n integrating information across l i n g u i s t i c units only when a f f e c t i v e processing i n the formation of conceptual relationships i s required. For t h i s reason, i n the present study, subjects were asked to produce two s t o r i e s , both of which were drawn from the subjects' own experience. One concerned a putative emotional topic and the other a neutral topic. I t was predicted that psychopathy would be most strongly associated with dependent measures i n the analysis of the a f f e c t i v e rather than the neutral narrative. Thought Disorder Formal thought disorder has been described as a disruption in the l o g i c a l r e l a t i o n s h i p among ideas (Bleuler, 1950). Concept boundaries may become f l u i d so that incongruous ideas -21-and objects are seen as related (e.g. Meehl, 1962). In addition, disorders r e l a t i n g to the content of thought, rather than i t s form, have also been included under the d e f i n i t i o n of thought disorder. For instance, responses to s t i m u l i may be bizarre or i d i o s y n c r a t i c (e.g. Harrow, Grossman, Marshall, S i l v e r s t e i n , & Meltzer, 1982). I n i t i a l l y , c l i n i c a l observation led to the b e l i e f that thought disorder was pathognomic of schizophrenia (Bleuler, 1950); however, recent studies suggest that thought disorder may be present i n nonschizophrenics (e.g. Andreasen & Powers, 1974; Harvey & Brault, 1986). Some investigators have suggested that disordered thinking f i t s along a continuum with normal thinking (Harrow & Quinlan, 1977). Harrow and Quinlan (1977) found that schizophrenics d i f f e r e d from other patient groups on measures of severe deviant thinking, but not on measures of milder thought disorder. Andreasen and Grove (1986) suggested that i n fact normal indivi d u a l s may exhibit mild thought disorder. One category of thought disorder - derailment - i s found i n normal individuals at s i g n i f i c a n t levels. Twenty to t h i r t y percent of normals have been reported to exhibit derailment at greater than mild l e v e l s (Andreasen & Grove, 1986; Oltmanns, Murphy, Berenbaum, & Dunlop, 1985). Andreasen (1979a, 1979b) developed a set of d e f i n i t i o n s that many researchers have used in studying disordered speech. These d e f i n i t i o n s cover eighteen categories of thought disorder - 2 2 -such as derailment, tan g e n t i a l i t y (replying to a question i n an oblique or irrelevant manner), and poverty of content of speech (an adequate amount of speech containing l i t t l e information). The scale that the categories form i s known as the Scale for the Assessment of Thought, Language, and Communication Disorders (TLC; Andreasen, 1980). This scale emphasizes the d i r e c t observation of language behaviour i n evaluating thought disorder. The author makes the point that thought disorder i s most often assessed from speech, and not from an i n d i v i d u a l ' s thoughts; therefore "disorganized speech" rather than "thought disorder" i s a more accurate term. As previously stated, the present study conceptualized ratings on the TLC as being indi c a t i v e of disordered communication or language. The TLC has been shown to be both a r e l i a b l e and v a l i d instrument (e.g. Andreasen, 1979a, 1979b; Andreasen & Grove, 1986; Docherty, Schnur, & Harvey, 1988; Harvey, 1983; Harvey, Earle-Boyer, & Levinson, 1988; Harvey, Earle-Boyer, & Weilgus, 1984) . Past research has found that TLC-rated thought disorder i s made up of several components that have d i f f e r e n t prognostic significance and are sensitive to patient diagnosis. For instance, poverty of speech appears to be an enduring feature of chronic schizophrenic i l l n e s s (Andreasen, Hoffman, & Grove, 1985; Docherty, Schnur, & Harvey, 1988; Pogue-Geile & Harrow, 1984) and to predict poor outcome (Harvey et a l . , 1988). Andreasen and Grove (1986) found that the negative signs of - 2 3 -thought disorder, poverty of speech and poverty of content of speech (adequate amount of speech conveying l i t t l e information), were most useful in predicting continued impairment amongst ps y c h i a t r i c patients. In manic patients, pos i t i v e rather than negative signs of thought disorder have been found to be most stable in the acute stage of i l l n e s s (Harvey et a l . , 1984). In the present study, the scores for two subscales based on TLC categories, as well as an o v e r a l l measure of thought disorder, were calculated (Andreasen, 1979a). P o s i t i v e formal thought disorder encompasses pressure of speech, t a n g e n t i a l i t y , derailment, incoherence, and i l l o g i c a l i t y . Negative formal thought disorder includes only poverty of speech and poverty of content of speech. The o v e r a l l score was the t o t a l of a l l thought disorder categories. It was predicted that increasing psychopathy would be associated with higher scores on p o s i t i v e thought disorder, since the scale contains categories related to a previous c l i n i c a l observation, namely that i n psychopathic speech concepts are not well connected. Cohesion: Meaning and Measurement Cohesion i s defined by Halliday and Hasan (1976) as a semantic relationship. Cohesion i s based on the meaning existing between an element in a text and some other element that i s c r u c i a l to i t s interpretation. The r e l a t i o n between elements can be systematized by c l a s s i f y i n g them i n t o a small -24-number of d i s t i n c t categories: reference, conjunction, l e x i c a l , substitution, and e l l i p s e s . The l a t t e r two w i l l not be discussed further as they are not included i n the present research. Their occurrence i n the speech of subjects i n similar studies has been shown to be of very low frequency (e.g. Rochester & Martin, 1979; Wykes & Leff, 1982). I t should be noted that since cohesion i s text based, t i e s may occur between clauses that are not immediately adjacent to one another. The various types of cohesion measured i n the present study can be seen i n Table I. Referential cohesion ref e r s to how an in d i v i d u a l i s able to interpret the meaning of one text element because of i t s rel a t i o n s h i p to another text element. I t i s interpretable either because i t i s i d e n t i f i e d with the referent (pronomial, demonstrative) or because i t i s compared with the referent (comparative). Conjunction i s d i f f e r e n t from other cohesive relationships in that i t expresses the l o g i c a l r elationships e x i s t i n g between clauses. These can be additive, adversative, temporal, causal, or continuative i n nature. The conjunction w i l l stand i n some type of r e l a t i o n s h i p to the sentence preceding i t , l i n k i n g i t to the one that follows. Lexical cohesion i s the r e i t e r a t i o n of a previous text element. This can occur through straight r e p e t i t i o n (same roo t ) , a synonym, a superordinate, or a general item. I t may involve the r e p e t i t i o n of words, phrases, or whole clauses. -25-Table I Categories of Cohesion Category Subcategory Examples Reference Conjunction 1. Pronomial 2. Demonstrative 3. Comparative 1. Additive 2. Adversative When we were walking out of the bar one night a guy grabs her // So I just about got i n a f i g h t with him "And when the door opens I'm going to get you" // I just t o l d them that It could take two months to save the money // That's a long time I knew the bartender there // And he always makes nice strong drinks I just took o f f // But i t was the only time (table continues) -26-Le x i c a l 3. Causal 4. Temporal 5. Continuative 1. Same root 2. Synonym (or Co-hyponyms) 3. Superordinate 4. General item We went out drinking //So I had a l o t to drink She went out that night // But then I went out with my buddies He never taught before // Actually the second year I had my f i r s t year teacher They put me i n to see the doctor // I seen the doctor And I despised i t // Cause I hate that c i t y with a passion We were s i t t i n g i n Alph a v i l l e // This c i t y i s dead It was a puzzle that was given to me // I can get that thing together Note. Categories are adapted from Rochester and Martin (1979) . Quotations are from subjects in the present study. Presuming -27-items and t h e i r referent are underlined. // indicates independent clause boundaries. Halliday and Hasan have argued that the cohesive relations that are established by conjunction and reference are semantic i n nature while those established by l e x i s are lexicogrammatical i n nature. However the expression of the r e l a t i o n s requires both the semantic and lexicogrammatical systems. This implies that simple word r e p e t i t i o n establishes some type of semantic relationship. However, when a speaker repeatedly uses some form of a word i n a text l e x i c a l cohesion i s coded i r r e s p e c t i v e of the meaning of that word. Simple l e x i c a l r e p e t i t i o n may not require the semantic encoding of actual word meanings to act as a form of text cohesion. This i s p a r t i c u l a r l y evident when the same l e x i c a l items are used repeatedly across clauses or sentences, a speech p e c u l i a r i t y that has been related to empty speech (Nicholas, Obler, Albert, & Helm-Easterbrooks, 1985) and "schizophasic" speech (Chaika, 1974). Chaika (1974) suggested that schizophrenics may not be able to match semantic features to words i n the lexicon. Fine (1978) has suggested that conjunction and reference should be viewed as tending to establish meaning relationships, whereas l e x i c a l t i e s should be viewed as e s t a b l i s h i n g more formal l i n k s . Generally, simple r e p e t i t i o n has been viewed as carrying the least amount of meaningful information (Fine, -28-1978; Rochester & Martin, 1979). Past studies have investigated whether"or not thought disordered i n d i v i d u a l s r e l y more on r e i t e r a t i o n than on the semantic r e l a t i o n s h i p s expressed by r e f e r e n t i a l cohesion (e.g., Rochester and Martin, 1979). However, the r e s u l t s of these studies have been mixed. Rochester and Martin (1979) , i n t h e i r study of schizophrenic speakers, examined cohesive r e l a t i o n s h i p s as a means of describing how thought disordered speakers might d i f f e r from non-thought disordered speakers. They hypothesized that the vague and tangential speech often seen i n thought disordered speakers may be due to a f a i l u r e to l i n k clauses to p r i o r discourse. They predicted that t h i s might be r e a l i z e d i n fewer cohesive t i e s of a l l types, or i n a preponderance of l e x i c a l t i e s . The authors hypothesized that r e i t e r a t i o n seems to require le s s extensive integration of textual information than do other forms of cohesion. The authors' hypotheses proved to be p a r t i a l l y correct. Overall, both thought disordered and non-thought disordered schizophrenics used le s s cohesion than the normal group. However, as predicted, the thought disordered group used more l e x i c a l cohesion than did the non-thought disordered group. Thought disordered speakers tended to r e l y very heavily on l e x i c a l cohesion, es p e c i a l l y simple r e p e t i t i o n as opposed to synonymy or general categories. The investigators concluded that thought disordered speakers were a c t i v e l y - 2 9 -p r o c e s s i n g a t a l e x i c o g r a m m a t i c a l l e v e l , t h e r e b y f o r m i n g l e x i c a l c h a i n s , but t h e y o f t e n d i d not form t h e more m e a n i n g f u l semantic t i e s . S i m i l a r r e s u l t s were o b t a i n e d f o r t h e o f f s p r i n g o f s c h i z o p h r e n i c s i n a s t u d y of t h e speech o f c h i l d r e n v u l n e r a b l e t o p s y c h o p a t h o l o g y (Harvey, Weintraub, & N e a l e , 1982). These c h i l d r e n used more l e x i c a l t i e s t h a n d i d normal c h i l d r e n . C h i l d r e n o f s c h i z o p h r e n i c s a l s o used fewer c o h e s i v e t i e s o v e r a l l . The a u t h o r s suggested t h a t t h e i r r e s u l t s were i n d i c a t i v e o f p o o r l y l i n k e d speech u n i t s i n t h e c h i l d r e n o f s c h i z o p h r e n i c p a r e n t s . R o c h e s t e r and M a r t i n (1979) found t h a t t h o u g h t d i s o r d e r e d speakers used more l e x i c a l c o h e s i o n t h a n d i d n o r m a l s p e a k e r s . Other r e s e a r c h e r s have found no d i f f e r e n c e s between normal a d u l t s and p a t h o l o g i c a l groups i n t h e i r use o f l e x i c a l t i e s (Harvey, 1983; R a g i n & Oltmanns, 1986). However, R o c h e s t e r and M a r t i n (1979) used as t h e i r dependent measure t h e p r o p o r t i o n o f l e x i c a l c o h e s i o n t o o t h e r t y p e s o f c o h e s i o n , whereas Harvey (1983) and R a g i n and Oltmanns (1986) used number o f t i e s p e r c l a u s e . Harvey r e p o r t e d t h a t i f R o c h e s t e r and M a r t i n s ' method had been used i n h i s s t u d y , then t h e r e l a t i v e p r o p o r t i o n s o f l e x i c a l c o h e s i o n would have d i s t i n g u i s h e d n o r m a l s from p a t i e n t s . R a g i n and Oltmann's study o f l e x i c a l c o h e s i o n and thought d i s o r d e r s u f f e r e d from a number o f m e t h o d o l o g i c a l problems w h i c h makes i t d i f f i c u l t t o e v a l u a t e . I n comparing -30-normals, schizophrenics, schizoaffectives, and manics the authors used an average of less than ten subjects per group and based t h e i r analysis on only f i f t e e n clauses per subject. Not s u r p r i s i n g l y , they found few group differences. Those that did prove s i g n i f i c a n t were related to within-clause l e x i c a l cohesion (the r a t i o n a l e for t h i s measurement was not made clear) and changes i n cohesion over time. The lack of within-clause l e x i c a l cohesion was related to thought disorder and not c l i n i c a l diagnosis. As the pathological subjects became less thought disordered they used less l e x i c a l cohesion. A number of studies have looked at cohesive t i e s i n r e l a t i o n to thought disorder and d i f f e r e n t types of psychopathology (e.g. Harvey, 1983; Ragin & Oltmanns, 1986; Wykes & Leff, 1982). This research has grown out of the debate about whether or not manics, l i k e schizophrenics, can be considered thought disordered. Harvey (1983), found that thought disordered schizophrenics and manics used less reference and conjunctive t i e s than did normal speakers. The non-thought disordered speech segments of thought disordered patients did not d i f f e r from those of normals. What i s most important about t h i s study i s that i t established a r e l a t i o n s h i p not between disordered communication and diagnosis, but between disordered communication and c e r t a i n patterns of cohesive t i e s . Wykes and Leff (1982) suggest that cohesion analysis may uncover whether or not ideas and phrases -31-are appropriately linked. If they are not, the authors f e l t that such a d e f i c i t was most closely related to the thought disorder category of derailment. In the present study cohesion analysis allowed the assessment of the extent to which psychopathy was associated with a tendency to produce poorly developed texts through an inadequate use of cohesive t i e s , or perhaps through the inordinate use of l e x i c a l cohesion. It was hypothesized that i f psychopaths have a general d e f i c i t in integrating semantic information to form cohesive texts, then psychopathy would be associated with narratives f i l l e d with l e x i c a l t i e s at the expense of other cohesive t i e s . An alternative p r e d i c t i o n was that i f psychopathy i s related to a generalized d e f i c i t i n the formation of a text through cohesive re l a t i o n s h i p s , then psychopathy would be associated with fewer t i e s of a l l types. If psychopathy i s related s p e c i f i c a l l y to a verbal a f f e c t i v e d e f i c i t that involves a d i f f i c u l t y in carrying out the semantic integration necessary for combining a f f e c t i v e and l i n g u i s t i c information, then i t was predicted that these d e f i c i t s would be apparent only for a f f e c t i v e texts. The Register of the Narratives Halliday and Hasan (1976) stress the fact that r e g i s t e r , along with cohesion (e.g. E l l i s & Ure, 1969; Hymes, 1964) define a text. The concept of register has been described i n various ways but usually includes information concerning the -32-topic or f i e l d of a text, i t s mode, i t s purpose or message-form, and the rel a t i o n s h i p existing between l i s t e n e r and speaker. Register re l a t e s to the context of a s i t u a t i o n and the fact that a text i s a communication. The narratives produced in the present research were texts of planned discourse. Subjects were asked to produce a story about a past event i n t h e i r l i v e s based on a p a r t i c u l a r topic. They were given a few minutes to think about what to say and how to say i t . They were given some di r e c t i v e s on what form i t should take. The topics of the discourses were r e l a t i v e l y unrestricted i n that subjects could t e l l a story about any time they f e l t angry or had d i f f i c u l t y i n doing something. The relationship e x i s t i n g between the l i s t e n e r and speaker was f a i r l y formal; however the li s t e n e r was not perceived as being a member of the prison bureaucracy. The mode of the s i t u a t i o n was oral with an orientation to relate past events that were personal to the speaker. From the l i s t e n e r ' s viewpoint, the purpose of the communication was for the speaker to produce a planned discourse i n story form. Ochs (1979) has pointed out that there are often differences between planned and unplanned discourse. E s s e n t i a l l y , discourse that i s planned makes r e l a t i v e l y more use of adult sy n t a c t i c a l constructions than i s seen i n unplanned discourse. For instance, the syntax w i l l make a semantic l i n k e x p l i c i t and s p e c i f i c so that conjunctions such as "because" or "so" are expressed (e.g. "I'm so t i r e d - I went -33-running today - I think I ' l l r est" versus "I'm so t i r e d because I went running today so I think I ' l l r e s t " ) . This may be because i t takes more planning to form s p e c i f i c semantic relationships. There i s no reason to believe that psychopathy might be associated with a f a i l u r e to be s e n s i t i v e to the r e g i s t e r of a discourse. I f i n fact psychopaths can be charming and are good manipulators, and yet produce speech that i s not well connected, then one might expect them to be e s p e c i a l l y sensitive to the demands of the l i s t e n e r at a pragmatic l e v e l . It should be noted that other authors have had a less rigorous view of what constitutes textness. P e t o f i (1985) defines a text as a verbal object which i s i d e n t i f i e d as a text by any one interpreter at any p a r t i c u l a r time. Brown and Yule (1983) simply state that a text " i s the verbal record of a communicative event (p. 190)." From t h i s viewpoint, nothing defines textness except for an individual's d e c i s i o n that a given object i s a whole i n the communication context. Coherence i s not necessary to form a text. For example, a disorganized monologue produced by a thought disordered speaker may form a text since i t can be viewed as a sample of incoherent speech. Reference Patterns Unlike l e x i c a l and conjunctive cohesion, r e f e r e n t i a l cohesion i s part of one of the p h o r i c i t y systems i n English. Phoricity systems r e f e r to the structuring of utterances on the - 3 4 -basis of what speakers assume the i r l i s t e n e r s know. They involve speech units that require previously presented information for t h e i r interpretation. The noun phrase or nominal group i s one such speech unit. Generally, there are two types of nominal groups. The f i r s t simply presents new information and the second presumes previous information so that i t may be understood by the l i s t e n e r . Information can be presumed from previous verbal information as i n r e f e r e n t i a l cohesion, e i t h e r e x p l i c i t l y (e.g., So I was i n the c i t y and I didn't l i k e i t ) or i m p l i c i t l y (e.g., I went to the c i t y and robbed the bank). Information can also be presumed from the immediate environment (e.g., Look at that guard), or general knowledge (e.g. The p o l i c e are bad). Sometimes, the references contained i n a text are ambiguous. One example of t h i s type of reference would be: I was d r i v i n g i n my car with my g i r l f r i e n d and my ex- g i r l f r i e n d Then she said that she didn't believe me Since i t i s unclear who i t i s that she r e f e r s to, the l i s t e n e r i s unable to interpret the second clause. References can also be unclear. For instance a speaker may make a pronomial reference when there i s no previous referent. If an individual's utterances are f u l l of unclear or ambiguous references then the l i s t e n e r i s unable to e f f e c t i v e l y understand what the speaker means. - 3 5 -Rochester and Martin (1979) found that thought disordered speech samples contained more unclear and ambiguous references than did non-thought disordered samples. This f i n d i n g has been consistent across a number of studies. Harvey (1983) found that unclear and ambiguous references d i f f e r e n t i a t e d between the speech of thought disordered and nonthought disordered manics and schizophrenics. Harvey et a l . (1982) also reported that the speech of children of schizophrenics contained more of these types of references than did the children of normals and other at r i s k groups. Based on a study that assessed thought disorder and reference d i f f i c u l t i e s at a six month i n t e r v a l , Docherty et a l . (1988) suggested that the frequency of incompetent references (ambiguous and unclear combined) may be a stable t r a i t of schizophrenics. In the present study the unclear and ambiguous references categories were coded and combined into the category of incompetent references (Docherty et a l . , 1988). This category was then correlated with PCL-R scores. I t was predicted that i f increasingly high scores on the PCL-R are associated with a f a i l u r e to develop texts on a semantic l e v e l , then they should also be associated with a r e l a t i v e l y high degree of incompetent references. A l t e r n a t i v e l y , i f psychopathy i s associated with p a r t i c u l a r d i f f i c u l t i e s i n the integration of a f f e c t and language, then incompetent references would be mostly associated with the a f f e c t i v e narrative. -36 -C o h e r e n c e : M e a n i n g a n d M e a s u r e m e n t A l t h o u g h c o h e s i o n c a n b e u s e d t o i d e n t i f y t h e u n i t y o f a t e x t i t d o e s n o t t h e g u a r a n t e e t h a t a d i s c o u r s e w i l l b e u n d e r s t a n d a b l e o r m e a n i n g f u l . C o h e s i o n c a n b e s c o r e d i r r e s p e c t i v e o f m e a n i n g r e l a t i o n s h i p s a n d i s t h e r e f o r e n o t a n a d e q u a t e a s s e s s m e n t o f t h e o v e r a l l u n d e r s t a n d a b i l i t y o f d i s c o u r s e . I n a s s e s s i n g t h e c o h e r e n c e o f s p e e c h t h e r e l a t i o n s h i p s e x i s t i n g b e t w e e n e v e n t s w i t h i n a d i s c o u r s e a r e t a k e n i n t o a c c o u n t . I f c o n n e c t i o n s b e t w e e n p r o p o s i t i o n s a r e w e a k , t h e n t h e o v e r a l l t e x t w i l l b e r e l a t i v e l y l e s s c o h e r e n t o r u n d e r s t a n d a b l e . I n a n a l y z i n g a t e x t i t i s i m p o r t a n t t o u n d e r s t a n d how a s e r i e s o f c o n c e p t s l i n k u p t o f o r m m o r e c o m p l e x m e a n i n g s . T h i s c a n b e d o n e by e x a m i n i n g c o h e r e n c e . I n c o n t r a s t t o c o h e s i o n , c o h e r e n c e i s more c o n c e r n e d w i t h d e l i n e a t i n g t h e r e l a t i o n s h i p s t h a t e x i s t b e t w e e n e v e n t s d e s c r i b e d w i t h i n a t e x t . T h e r e a r e a number o f s y s t e m s f o r e x a m i n i n g t h e s e r e l a t i o n s h i p s ( e . g . G u l i c h & Q u a s t h o f f , 1 9 8 2 ; O m a n s o n , 1 9 8 2 ; R e i s e r & B l a c k , 1 9 8 2 ) ; h o w e v e r t h e y h a v e n o t o f t e n b e e n u s e d t o i n v e s t i g a t e p s y c h o p a t h o l o g i c a l g r o u p s . One s t u d y h a s shown t h a t t e x t - b a s e d p r e s u p p o s i t i o n a l r e l a t i o n s h i p s b e t w e e n p r o p o s i t i o n s t e n d t o b r e a k down i n s c h i z o p h r e n i c s ( H o f f m a n , K i r s t e n , S t o p e k , & C i c c h e t t i , 1982) . P l o t - u n i t A n a l y s i s I n t h e p r e s e n t s t u d y , t h e n a r r a t i v e s s u b j e c t s p r o d u c e s h o u l d c o n t a i n some s t r u c t u r i n g o f s e m a n t i c r e l a t i o n s h i p s s o -37-that they form a meaningful text. The story should begin at the beginning, move ahead reasonably smoothly i n time, and evaluate states and events so that i t i s possible to i n f e r what the story i s about (Polanyi, 1985). Subject narratives should meet these conditions i f t h e i r discourse i s to be seen as coherent and meaningful. One way of describing these relationships i s through the use of story grammars. Story grammars were constructed for the description of the structure of a genre of s t o r i e s known as f o l k - t a l e s . Very s p e c i f i c and invariant rules were developed to describe the ordering of events within these s t o r i e s . However these rules may not be appropriate for the analysis of narratives that may be structured d i f f e r e n t l y (Mandler, 1984, p. 17). For t h i s reason, the use of story grammars would not seem to be appropriate f o r the present purposes. Another way to approach the problem i s to look at the way i n which expectancies are set up in the l i s t e n e r by information given i n the narrative (Black & Reiser, 1982). Generally, i f a problem i s mentioned that i s central to the point of the story, then information concerning some type of re s o l u t i o n or outcome i s expected to be provided by the narrator. I f the narrator i s able to structure the story appropriately then both problem and resolution or r e s u l t w i l l be provided. For instance, i f the event involves anger, then the narrator should attempt to explain what led to the anger and i t s outcome. - 3 8 -One way to assess t h i s i s through the use of p l o t - u n i t analysis (Botvin & Sutton-Smith, 1977; Sutton-Smith, Botvin, & Mahony, 1976). Compared to story grammars there i s more emphasis on meaning in p l o t - u n i t analysis. Instead of consisting of a schematic structure within which elements are placed, p l o t - u n i t analysis i s directed at discovering how the text explains the actions of the main characters. P l o t - u n i t s are verbal noun dyads indicating action (or the p o t e n t i a l for action) and resolution (e.g. injury - recovery or death). As well as being sensitive to the structural complexity of a story (Botvin & Sutton-Smith, 1977), these dyads are also useful i n assessing the extent to which one concept i s linked to another within a text. For instance, i f the occurrence of an injury i s central to a story then some type of resolution or r e s u l t should occur with regard to the injury. I f i t does not, then the discourse may seem unconnected or vague. Secondary dyads should not interrupt the flow of the story. If present, they should be embedded within primary dyads (Botvin & Sutton-Smith, 1977). As previously indicated, psychopaths may have d i f f i c u l t y i n l i n k i n g conceptual units to form a coherent text. Plot-unit analysis allowed the assessment of t h i s hypothesized d e f i c i t through the i d e n t i f i c a t i o n of l o g i c a l conjugates which should be present i n p a i r s . I t was predicted that psychopathy would be associated with a tendency to introduce the f i r s t part of a -39-plo t - u n i t , but not to complete i t . The dependent measure i n t h i s case was the proportion of closed to open p l o t u n i t s . -40-Experimental Hypotheses It has been suggested that psychopaths' concepts, as expressed i n t h e i r speech, are not well connected to one another. This d e f i c i t may take one of a number of forms. Based on c l i n i c a l observation and the research reviewed above, i t was hypothesized that: 1. Psychopathy i s associated with an increase i n l e x i c a l cohesion or a decrease i n the use of a l l types of cohesion. 2. Psychopathy i s associated with an increase i n the use of incompetent references. 3. Psychopathy i s associated with a f a i l u r e to open and close the same plot-unit. 4. Psychopathy i s associated with p o s i t i v e , but not negative measures of thought disorder. These hypotheses were modified by the caveat that, i f abnormalities i n speech production are a f f e c t i v e l y based i n psychopaths, then these predictions would be r e a l i z e d for the a f f e c t i v e , but not the neutral s t o r i e s . - Method Subjects Subjects were 40 male inmates from a Canadian Federal prison near Vancouver, B.C. who had volunteered to p a r t i c i p a t e i n an ongoing research project. Psychopathy was assessed using the PCL-R, a 20 item scale of proven r e l i a b i l i t y and v a l i d i t y (e.g., Hare, 1980, 1985, 1991; Hare et a l . , 1990). I t measures -41-both behavioural and personality variables on the basis of an extensive interview with the inmate and a review of h i s i n s t i t u t i o n a l f i l e s . The summed score (maximum of 40) provides a global measure of psychopathy. The i n t r a c l a s s c o r r e l a t i o n c o e f f i c i e n t for the PCL-R i n the population from which t h i s sample was drawn was .84. The interview and ratings were ca r r i e d out by individuals trained in i t s use (not the experimenter). When ava i l a b l e the means of double r a t i n g s were used to increase r e l i a b i l i t y . Past research has demonstrated that those indiv i d u a l s with low and high PCL-R scores do not d i f f e r on l e v e l s of education, parents' s o c i a l c l a s s , I.Q. or other neuropsychological variables (e.g. Hare, F r a z e l l e , Bus, & J u t a i , 1980; Harpur et. a l . , 1989; Hart, Forth, & Hare, 1990). In most studies, subjects have been divided into high (PCL-R 30 or greater) or low (less than or equal to 20 on the PCL-R) psychopathy groups for comparative research. For the primary analysis i n the present study, however, psychopathy was treated as a continuous variable. Some ad d i t i o n a l analyses were ca r r i e d out using subjects c l a s s i f i e d as psychopaths (PCL-R score 30 or greater; n = 21) or as nonpsychopaths (PCL-R score less than 30; n = 15). For a l l analyses, 4 subjects who received diagnoses of eit h e r schizophrenia or s c h i z o a f f e c t i v e disorder were omitted as described below. The i n d i v i d u a l who performed the i n i t i a l PCL-R assessments re c r u i t e d subjects to par t i c i p a t e in the present research i f - 4 2 -they had learned English as their f i r s t language and i f they were between the ages of 18 and 60. The mean age of the 36 subjects was 29.6 years (S.D. = 9.1; range = 19 to 56 years). The Mean years of education was 8.9 (S.D. = 1.9; range = 4 to 12) . Neither education nor age was found to c o r r e l a t e with PCL-R ratings nor any of the dependent variables. The mean PCL-R score was 29.4 (S.D. = 8.2; range = 7.5 to 40.0). Procedure Two s t o r i e s , one involving an angry incident, and one involving a topic of personal d i f f i c u l t y , were c o l l e c t e d from the subjects at the beginning of the session. This was followed by the administration of the SADS-L interview. C o l l e c t i o n of the speech samples and the administration of the SADS-L interview were carried out by the author, a Ph.D. l e v e l c l i n i c a l psychology graduate student, who was b l i n d to PCL-R ratings. Subjects were paid $7.00 for the session which lasted from 1 to 2.5 hours. Subjects were tested i n d i v i d u a l l y . Speech Samples The subject was asked to produce two types of s t o r i e s that were to be based on personal experience. This task required a subject to r e c a l l a time when he was angry and a time when he had d i f f i c u l t y in doing or learning something. The d i r e c t i v e f o r the angry ( d i f f i c u l t y ) story was: "I want you to t e l l me a story about a time you f e l t r e a l l y angry (had d i f f i c u l t y i n doing or learning something). Like most s t o r i e s , i t should have - 4 3 -a beginning, a middle, and an end. T e l l the story as i f you are there. Describe what happened and what you did. I ' l l give you a couple of minutes to think about i t . Then I want you to t e l l me what happened l i k e i t was a story. You should speak f o r a few minutes. Are there any questions?" The interviewer then answered any questions that the subject had concerning the instructions and t o l d him not to talk to the interviewer while he was t e l l i n g the story. I f a subject asked f o r s p e c i f i c advice about what to speak about he was encouraged to continue to t r y to r e c a l l an angry ( d i f f i c u l t ) time. Two subjects (both nonpsychopaths) could not think of a angry time so they were asked to speak about a sad incident. Some subjects were not sure what was meant by the topic " d i f f i c u l t y i n doing or learning". The experimenter gave the example of " t r y i n g to figure out a math problem". None of the subjects used t h i s example i n creating a story. After questions were answered the interviewer repeated the instructions and the subject began. A couple of the subjects asked the interviewer questions about story content af t e r they had begun t h e i r narration. For example, one subject asked i f someone that he was describing was known to the interviewer. In these cases the interviewer responded with "mmm" and the subject continued. The emotional topic occurred f i r s t or second on a random basis. There was 20 minutes between each story. A f t e r the t e l l i n g of each story the experimenter asked questions about -44 -i t s content for about 10 minutes. The narratives were recorded on audio tape with high q u a l i t y head-held microphones. SADS-L After c o l l e c t i n g the speech samples the SADS-L structured interview was administered. The f i r s t 15 to 30 minutes were spent gathering demographic information from the subject. General questions about the subject's past p s y c h i a t r i c history were also asked. The subject was then t o l d that he would be asked a set of standard questions about how he f e l t and thought. The interview took from 30 minutes to 1.5 hours and was videotaped. Before making RDC diagnoses both the experimenter and the r e l i a b i l i t y coder reviewed the t r a i n i n g materials provided by the authors. The r e l i a b i l i t y coder was a M.A. l e v e l graduate student i n c l i n i c a l psychology who had a great deal of previous experience i n the administration of structured interviews and t h e i r scoring. The experimenter had been previously trained i n the use of the SADS in a p s y c h i a t r i c h o s p i t a l . Both the SADS-L videotapes and prison f i l e s were used to make diagnoses. The second rater made diagnoses for a randomly selected set of 20 subjects based on the same information. Subjects met the c r i t e r i a for a variety of diagnoses. However, for the present study only the subjects who met the c r i t e r i a for disorders found to have been related to thought disorder were of i n t e r e s t . Two subjects met the c r i t e r i a for -45-schizoaffective disorder (both depressed type), and two subjects met the c r i t e r i a for schizophrenia (one paranoid and one disorganized). These subjects were excluded from the study. The two subjects who were rated as s c h i z o a f f e c t i v e by the experimenter were given the same diagnoses by the student. The two raters also agreed on a past instance of s c h i z o a f f e c t i v e disorder (depressed type) for one of the subjects. Neither r a t e r gave any other current diagnosis of schizophrenia, schizoaffective disorder, mania, or depression. Agreement was almost perfect on other diagnoses not relevant to the current study. One of the two subjects diagnosed as schizophrenic had been rated i n the psychopathic range on the PCL-R with a score of 30. The other subjects excluded from the study had PCL-R scores of 24, 23, and 5. I t was surprising to f i n d that 10% of the i n i t i a l sample suffered from serious psychopathology. However, the prison from which the inmates were drawn was at the time being used to house the most dangerous and disturbed offenders i n i t s geographic area. The mean PCL-R score obtained i n the present study i s i n fact s l i g h t l y higher than has been found i n past research, as i s the proportion of inmates defined as psychopathic by the PCL-R. Both the elevated PCL-R scores and the serious pathology that was found may have been a r e f l e c t i o n of prison p o l i c y . At the time, offenders were being screened -46-upon a r r i v a l at the i n s t i t u t i o n , and the less v i o l e n t and less dangerous inmates were being sent to other i n s t i t u t i o n s . Cohesion and Reference Ratings The experimenter transcribed a l l of the 80 s t o r i e s from the audio tapes, including pauses and word r e p e t i t i o n s . The st o r i e s were then broken down into independent clauses based on the d e f i n i t i o n of Rochester and Martin (1979): an independent clause i s a unit which stands by i t s e l f as a de c l a r a t i v e , interrogative, imperative, or exclamatory structure. Relative clauses, adverbial clauses, and fact or report complements were treated as part of t h i s basic unit as were sentence modifiers. Number of words and number of clauses were counted. Cohesion and incompetent references were scored according to a manual that was created for t h i s study based on the book Cohesion i n English (Halliday & Hasan, 1976) and an unpublished manual provided by Harvey (1983). Hanging clauses were included at the beginning of a new clause. Unless these clauses were necessary for the interpretation of subsequent information they were not scored for l i n g u i s t i c variables nor were they included i n the word count. Repetitions of words within a clause were also not scored (e.g., "They - they went to the store"; only the second "they" was coded for l i n g u i s t i c variables and included i n the word count). A masters l e v e l graduate student from speech and hearing sciences who was f a m i l i a r with the coding of cohesion, -47-but knew nothing about psychopaths, then performed the same analysis on 30 (15 a f f e c t i v e and 15 neutral) of the 80 s t o r i e s using the manual provided by the experimenter. Previous investigators have found the r e l i a b i l i t y for coding cohesion to be high. Harvey (1983) obtained Kappas of .79 for l e x i c a l cohesion, .87 for r e f e r e n t i a l cohesion, and .94 fo r conjunction. Ragin and Oltmanns (1986) obtained an i n t e r r a t e r r e l i a b i l i t y of .95 for the amount of l e x i c a l cohesion per t r a n s c r i p t . Rochester and Martin (1979) computed Kappas of .73 to 1.00 for selecting cohesive items and .96 for se l e c t i n g referents. In the present study i n t e r r a t e r r e l i a b i l i t y (Pearson product moment correlation) f o r the neutral (affective) story for number of cohesive t i e s per t r a n s c r i p t was .81 (.84) f o r l e x i c a l cohesion, .82 (.88) for r e f e r e n t i a l cohesion, and .92 (.90) for conjunction. For the i d e n t i f i c a t i o n of incompetent references i t was .91 (.90). Interrater r e l i a b i l i t y for the number of words and clauses ranged from .95 to .98 for the a f f e c t i v e and neutral s t o r i e s . Plot-Unit Ratings There i s no set of standard plot-units for the analysis of narratives of r e a l l i f e events. For t h i s reason, p l o t - u n i t s were selected that had proven useful i n a previous story structure analysis (Gillam, 1989). In addition, new p l o t - u n i t s were created based on a subset of the present n a r r a t i v e s . The author and the same coder who had scored cohesion f o r -48-r e l i a b i l i t y j o i n t l y reviewed the 30 narratives (15 a f f e c t i v e and 15 neutral) that had previously been coded by the second r a t e r . They i d e n t i f i e d and discussed new p l o t u n i t s , i n addition to those given i n Gillam (1989), necessary f o r coding the narratives. The basic unit of analysis was the independent clause. The p l o t units that were used, both from Gillam (1989) and newly created, are shown i n Table I I . A l l of the plot-units were then used by both r a t e r s to independently analyze 5 s t o r i e s selected from a book of short s t o r i e s . R e l i a b i l i t y for these ratings was high, and any discrepancies were discussed by the raters. F i n a l l y , t h i r t y new narratives from t h i s study were selected ( f i f t e e n a f f e c t i v e and f i f t e e n neutral s t o r i e s ) , and were coded by both r a t e r s . The i n t e r r a t e r r e l i a b i l i t y i n the present study for the neutral (affective) s t o r i e s was .88 (.89) for the i d e n t i f i c a t i o n of p l o t - u n i t s (the same plot u n i t opened and closed). This value compares favourably with those reported by Botvin-and Sutton-Smith (1977), who reported interrater agreement of .81 for the i d e n t i f i c a t i o n of p l o t - u n i t s using graduate students with three hours of t r a i n i n g , and by Gillam (1989), who reported an i n t e r r a t e r r e l i a b i l i t y of .91. Using the same p l o t - u n i t s , the author then rated the remaining 42 narratives. Thought Disorder Ratings The TLC was scored according to the manual provided by i t s author (Andreasen, 1980) and the revisions proposed by Oltmanns - 4 9 -Table II Plot-units and Their Definitions Problem Result Negative Feeling State N u l l i f i e d / n o t N u l l i f i e d Description of s e l f as having negative feelings. V i l l a i n y N u l l i f i e d / n o t N u l l i f i e d Injurious or malevolent act committed on the speaker. Deception Revealed/not revealed A misleading act or statement that i s intended to cause a p a r t i c i p a n t i n the narrative to act or think wrongly. Threat N u l l i f i e d / n o t n u l l i f i e d A p o t e n t i a l source of danger, harm, or d i s t r e s s . (table continues) -50-Table II continued Problem Result Plan A scheme (thought, intention) f o r doing something. Attack To attack a participant with intent to harm (can be verbal). Carried Out/not Carried Out Counterattack/Wound/Flee Injury Recovery/Death Physical harm or damage to a par t i c i p a n t . Pursue Capture/Escape/Release To chase i n order to capture or harm. Search Find/not f i n d To look for something. (table continues) -51-Table II continued Problem Result Compete Win/Lose To compete with another par t i c i p a n t for something. Command Obeyed/Not obeyed To be ordered to do something. Aid Accepted/Not accepted To o f f e r aid to a p a r t i c i p a n t . Promise Kept/Not kept To promise a p a r t i c i p a n t something. Problem Resolved/not Resolved A problem that i s hard for a participant to contend with. Note. Adapted from Gillam (1989). - 5 2 -et a l . (1985) based on the f i r s t 30 minutes of the SADS-L interview, with the appropriate correction for time as instructed i n the manual. Each TLC category i s scored on a 0 to 3 or 4 point scale. In the present study, the experimenter and a PhD graduate student i n forensic psychology, not involved i n any of the previous ratings, f i r s t read the manual, and scored 5 t r a i n i n g tapes to discuss discrepancies i n scoring. Then the experimenter rated a l l the subjects on the TLC and the student rated 50% of the subjects to assess r e l i a b i l i t y . T y p i c a l l y , for any category a score of 0 i s considered to be i n d i c a t i v e of no thought disorder, a score of 1 to be representative of mild thought disorder, and 2 taken to indicate the d e f i n i t e presence of thought disorder. Kappa c o e f f i c i e n t s of agreement for the presence (2 or greater) and absence (less than 2) of thought disorder for the frequently occurring categories of t h i s scale are given i n Table I I I . They range from .58 to .82 and are quite similar to those obtained by other researchers (e.g. Andreasen, 1979a; Harvey, 1983; Oltmans et a l . , 1985). For categories having a base-rate of le s s than 20 percent, Kappas were not calculated; however the absolute agreement between raters for these categories was high (at least 95 percent). Results Thought Disorder The thought disorder ratings were highly negatively skewed - 5 3 -r e s u l t i n g i n a non-linear relationship with the PCL-R. The thought disorder ratings were therefore transformed by t h e i r square root (Cohen & Cohen, 1983) . This produced an approximately normal d i s t r i b u t i o n for thought disorder and an approximately linear relationship between thought disorder and the PCL-R. The PCL-R was found to correlate .35 (p_ < -03) with the transformed scores. This suggests that increasing psychopathy i s accompanied by an increased tendency to produce disordered communications. R e l i a b i l i t i e s , means, and standard deviations f o r thought disorder categories, broken down by psychopathy group, are presented i n Table I I I . Only categories for which at l e a s t one i n d i v i d u a l scored a 1 are included. A score of 1 or below on the TLC i s considered to be i n d i c a t i v e of mild thought disorder or no thought disorder. Thought disorder appears to have occurred with r e l a t i v e l y greater frequency among subjects r e l a t i v e to a sample of normal ind i v i d u a l s assessed by Andreasen and Grove (1986). The percentage of individuals exhibiting scores of 2 or greater for that sample and the present sample can be compared i n Table I I I . A Chi-Square analysis was performed, with subjects c l a s s i f i e d by the presence or absence of thought disorder and by the presence or absence of psychopathy. The analysis was highly s i g n i f i c a n t (Chi-sguare(1) = 13.02, p < .001). Twenty -54-Table III Summary of Information for Thought Disorder (TD) Categories Psychopaths Nonpsychopaths (n = 21) (n = 15) Category M ( SD ) p a M ( SD ) p a pb Kappa Poverty of Speech .81 (1.0) 29 .07 ( -3) 0 5 .58 Poverty of Content . 10 ( -3) 0 .07 ( -3) 0 1 -Pressure of Speech .33 ( -9) 14 .00 ( .0) 0 6 -Tangentiality .76 ( .9) 29 .46 ( -9) 13 2 .82 Derailment 1.24 (1.2) 48 .40 ( .6) 17 32 .82 I l l o g i c a l i t y .09 ( .4) 5 .00 ( -0) 0 0 -Incoherence .23 ( -4) 9 . 00 ( .0) 0 0 -D i s t r a c t i b l e Speech . 10 ( -3) 0 .00 ( -0) 0 3 -Circumstantiality 1.20 (1.0) 48 .80 (1.0) 26 6 .78 Loss of Goal 1.29 (1.0) 53 .73 (1.1) 23 18 .80 R e l i a b i l i t y P o s i t i v e TD 2.67 (3.1) 57 1.00 (1.9) 13 .91 Negative TD .81 (1.0) 29 . 13 ( -4) 0 .87 Total TD 6.14 (3.7) 95 2.67 (3.9) 33 .86 (table continues) -55-Table III continued Note. Only categories for which at least one subject scored a one or more are included. aThe percentage of individuals scoring greater than 2 or greater on the variable. ^The percentage of 94 normal in d i v i d u a l s scoring 2 or greater on the variable as reported by Andreasen and Grove (1986). -56-out of 21 subjects i n the psychopathic range of the PCL-R were c l a s s i f i e d as thought disordered, whereas only 5 of the 15 subjects i n the nonpsychopathic range were so c l a s s i f i e d . Similar analyses were carr i e d out separately f o r p o s i t i v e and negative thought disorder. Neither positive (r = .16, p > .34) nor negative (r = .29, p_ > .09) thought disorder c o r r e l a t e d s i g n i f i c a n t l y with psychopathy. This apparent lack of rel a t i o n s h i p may have been due to the highly skewed nature of the thought disorder d i s t r i b u t i o n s that were not amenable to correction. However, there was a relationship between PCL-R group and po s i t i v e thought disorder for the Chi-Square analysis, (Chi-square (1) = 7.07, p_ < .01). The most frequently occurring categories of p o s i t i v e thought disorder were derailment and tang e n t i a l i t y . Group differences i n negative thought disorder were not s i g n i f i c a n t after the Yates Correction, (Chi-square(1) = 3.29, p > .07), with 6 psychopaths and none of the nonpsychopaths scoring 2 or greater. A l l of the 6 psychopaths' scores of 2 or greater were due to poverty of speech and not poverty of content of speech. The mean values for each group for positive, negative, and t o t a l thought disorder can be found i n Table III. Aff e c t Manipulation Check Whether or not the two story types d i f f e r e d on emotionality was assessed by two raters, not involved i n any other ratings, who rated each story on two 7-point Lykert -57-scales. One of the raters was a graduate student i n c l i n i c a l psychology and the other had just completed an undergraduate degree i n psychology. For the f i r s t scale, ratings were made to the question "How emotional i s t h i s story?" with scale anchors of "Extremely" (7) and "Not at a l l " (1). For the second scale, ratings were made to the question "How much interpersonal c o n f l i c t i s there i n t h i s story?" with scale anchors of "A l o t " (7) and "None at a l l " (1). The intra c l a s s c o r r e l a t i o n f o r the average of two raters was .69 for emotionality and .79 f o r c o n f l i c t . Paired sample t - t e s t s were carried out on the resultant mean scores. On the emotionality dimension the two story types d i f f e r e d (t = 2.59, p < .01; mean f o r the d i f f i c u l t y story = 3.9, SD = 1.7; mean for the angry story = 4.8, SD = 1.3). The s t o r i e s also d i f f e r e d on the interpersonal c o n f l i c t dimension (t = 3.99, p < .001; mean f o r the d i f f i c u l t y story = 3.2, SD = 1.6; mean for the angry story = 4.8, SD = 1.3). These results suggest that the two s t o r i e s d i f f e r e d on the rated dimension of emotionality; however they may al s o have d i f f e r e d on other dimensions. In subsequent discussions the angry story w i l l be referred to as af f e c t i v e and the d i f f i c u l t y story w i l l be referred to as neutral. However, even though the "neutral" story was rated as les s emotional then the angry one, there i s no evidence that i t was not emotional. Discourse Variables The Means and S.D.s for t o t a l number of words, t o t a l - 5 8 -number of clauses, mean length of utterance (MLU; mean number of words per independent clause) and open and closed p l o t - u n i t s for each story type can be found i n Table IV. No s i g n i f i c a n t differences were found between high and low psychopathy groups for any of these variables. As can be seen i n Table IV, the number of clauses and words used was sensitive to whether or not subjects related a neutral or an a f f e c t i v e story. Subjects t o l d longer s t o r i e s for the a f f e c t i v e topic; however MLU d i d not d i f f e r for the two story types. Subjects also produced more open and closed plot-units for the a f f e c t i v e story suggesting i t was more complex than the neutral story. In order to assess the r e l a t i o n s h i p between the dependent variables and the discourse variables, correlations were calculated between the two sets of variables f o r a l l subjects. These correlations are shown i n Table V. The number of open and closed plot-units was included on the dependent variable l i s t so as to assess the r e l a t i o n s h i p between story length and story complexity. The c o r r e l a t i o n s show that the t o t a l number of clauses was highly r e l a t e d to the number of open and closed p l o t - u n i t s . This suggests that story complexity, as assessed by the number of p l o t - u n i t s , was r e l a t e d to story length. However, story length was not s i g n i f i c a n t l y associated with the degree to which subjects opened and closed the same plot- u n i t s , as indicated by the nonsignificant c o r r e l a t i o n , suggesting that coherence was not -59-Table IV Means and Standard Deviations of the Discourse Variables f o r the A f f e c t i v e and Neutral Stories Neutral A f f e c t i v e Story Story Variable M (SD) M (SD) £(34) Open Plo t Units 3.94 (1.97) 5. 67 (3.57) 3.18** Closed Plot Units 2.75 (1.75) 4.36 (2.98) 3.50*** Number of Words 395 (205) 537 (384) 2.75** Number of Clauses 45 ( 26) 61 ( 46) 2.50* MLUa 9. 07 (1.73) 9.00 (2.20) .24 a MLU length of utterance which i s the number of words per independent clause. * E < .05. * * p_ < .01. *** E < .001. -60-Table V Correlations of the Dependent Variables with Total Words, Clauses and MLU for the A f f e c t i v e and Neutral Stories Variable Total Words Total Clauses MLUa Neutral Story L e x i c a l Cohesion .07 .23 . 35* Referential Cohesion .07 -.01 . 32* Conjunctive Cohesion -.08 .02 -.01 Total Cohesion .07 -.06 . 35* Incompetent References .27 * .33 -.12 Closed/Open Plot-units .05 . 05 .02 Open Plo t Units .63*** . 7 2 * * * -.19 Closed Plot Units . 54 .63*** -.15 (table continues) Table V continued -61-Variable Total Words Total Clauses MLUa A f f e c t i v e Story Le x i c a l Cohesion .07 .23 .28 Refer e n t i a l Cohesion .07 -.01 .39* Conjunctive Cohesion -.08 . 02 -.02 Total Cohesion .07 -.06 AC** . 46 Incompetent References . 14 . 12 . 08 Closed/Open plot-units .02 . 05 . 02 Open Pl o t Units .68*** * * * .82 -.20 Closed P l o t Units .58*** .70*** -.14 Note. Cohesion variables and incompetent references are occurrences per clause. a Mean length of utterance. * p < .05. ** p < .01. -62-r e l a t e d to story length. MLU was related to t o t a l cohesion for both story types. Rochester and Martin (1979) demonstrated that for non-thought disordered individuals the longer the MLU the greater the number of cohesive t i e s . However, the authors also found that t h i s relationship did not hold for thought disordered individuals. Cohesion Table VI displays the correlations between the PCL-R and the dependent variables. These correlations reveal a negative association between the PCL-R and the cohesion variables f o r the neutral, but not the a f f e c t i v e , s t o r i e s . This appears to be due to a general trend for psychopaths to use l e s s cohesion than the nonpsychopaths i n the neutral story. In the a f f e c t i v e story the groups used similar amounts of cohesion. The relationship between psychopathy and the e f f e c t of an emotional topic on story t e l l i n g was assessed by p a r t i a l i n g the variance of the neutral story variables out of the c o r r e l a t i o n s between psychopathy and the a f f e c t i v e dependent variab l e s . The p a r t i a l correlations between psychopathy and the cohesion variables for the a f f e c t i v e story, with neutral story variance removed, are shown in Table VI. None of these c o r r e l a t i o n s were s i g n i f i c a n t , suggesting that the relationship between psychopathy and cohesion was not p a r t i c u l a r to an emotional story t o p i c . -63-Table VI Correlations of the PCL-R with the Dependent Variables for the Neutral and A f f e c t i v e Stories Neutral A f f e c t i v e A f f e c t i v e Story Story Story (neutral variance Variable partialed) Lexical Cohesion -.41** .07 .20 Referential Cohesion -.30 -.20 -.07 Conjunctive Cohesion * -.37 -.01 -.07 Total Cohesion -.47** -.08 .20 Incompetent References . 23 .43** . oo Closed/Open plot-units A A * * -. 44 -.40* .29 Note. Cohesion variables and incompetent references are occurrences per clause. * p_ < .05. * * E < .01. * * * E < -001. -64-The Means and S.D.s for the cohesion variables f o r each a f f e c t type are presented i n Table VII for psychopaths and nonpsychopaths. Due to the number of t tests that were performed the Bonferonni correction was used across each story type. The familywise Type I error rate was held at .05 by t e s t i n g each comparison at p_ =.008 (.05/6 v a r i a b l e s ) . Total cohesion was about 2.48 t i e s per clause across the d i f f e r e n t types of s t o r i e s . This figure compares favorably with those previously found for interviews and the r e t e l l i n g of a story j u s t heard, which have ranged from 2.04 to 3.86 i n normal in d i v i d u a l s (Harvey, 1983; Rochester and Martin, 1979). Referential and conjunctive cohesion are also i n the same range as that previously found. Harvey (1983) reports values of .97 for reference and .45 for conjunction i n normals based on an open top i c interview. Close i n value to those figures are those found i n the present study: .88 (neutral story) and .92 ( a f f e c t i v e story) for reference and .41 (neutral story) and .43 ( a f f e c t i v e story) for conjunction. What i s d i f f e r e n t from previously reported values are the figures for l e x i c a l cohesion. For both the a f f e c t i v e and neutral story r e l a t i v e l y high amounts of l e x i c a l cohesion were found: 1.18 t i e s per clause for the neutral story and 1.14 t i e s per clause f o r the a f f e c t i v e story. Harvey (1983) reported .66 t i e s per clause and Ragin and Oltmanns reported -65-Table VII Means and Standard Deviations of the Dependent Variables for the Affective and Neutral Stories for Psychopaths and Non-psychopaths Psychopaths Nonpsychopaths (n = 21) (n = 15) Variable M (SD) M (SD) t(34) Neutral . Story L e x i c a l Cohesion 1.11 ( -19) 1.29 ( .29) 2.31 Re f e r e n t i a l Cohesion .78 ( -27) 1.01 ( .23) 2.43 Conjunctive Cohesion . 38 ( -09) .43 ( .08) 1.74 Total Cohesion 2.28 ( .43) 2.73 ( . 39) 3. 19* Incompetent References .08 ( .07) .04 ( .06) 1.69 Closed/Open Plot-units .60 ( .34) •89 ( .20) 2.99* (table continues) -66-Table VII continued Psychopaths Nonpsychopaths (n = 21) (n = 15) Variable M (SD) M (SD) t(34) Affective s Story Lex i c a l Cohesion 1.13 ( -28) 1.15 ( .29) . 17 Ref e r e n t i a l Cohesion .84 ( -31) 1.03 ( -22) 2.00 Conjunctive Cohesion .43 ( .15) .43 ( .10) .04 Total Cohesion 2.40 ( -49) 2.61 ( .34) 1.40 Incompetent References . 11 ( -06) .04 ( .06) 2.98* Closed/Open Plot-units .67 (. -16) .91 ( -16) 3.29* Note. Cohesion variables and reference f a i l u r e s are occurrences per clause. The familywise Type I error rate was held at .05 by te s t i n g each comparison at p_ = .06/6 = .008. * E < .008. -67-.71. However, Rochester and Martin (1979) found that the proportion of cohesion to t o t a l cohesion i s se n s i t i v e to context i n that for interviews, where topics were based on past experience, subjects used l e x i c a l cohesion to a greater degree than they d i d other types of cohesion. For the r e t e l l i n g of a story just heard subjects tended to use reference to a greater degree than other types of cohesion. I t may be that the p a r t i c u l a r context of the present study led to the p a r t i c u l a r amounts of cohesion that were found, or i t may be that inmates tend to use more l e x i c a l cohesion i n t h e i r speech than noninmates. Group differences were apparent for t o t a l cohesion i n the neutral story. Psychopaths used less cohesion over a l l than the nonpsychopathic group. There were also trends for group differences on l e x i c a l (p = .03) and r e f e r e n t i a l (p_ = .02) cohesion. Nonpsychopaths tended to use both types of cohesion more often than dis psychopaths. For the a f f e c t i v e story, no group differences were found, although once again there was a trend for nonpsychopaths to use more r e f e r e n t i a l cohesion than the psychopaths (p = .05). Incompetent References As can be seen i n Table VI, increasing psychopathy was related to a tendency to make incompetent references. For the a f f e c t i v e story psychopathy added a s i g n i f i c a n t amount of variance to t h i s relationship, over and above that contributed - 6 8 -by the neutral story. The mean number of incompetent references for the a f f e c t i v e and neutral s t o r i e s can be seen i n Table VII. Amounts of .06 (neutral story) and .08 (a f f e c t i v e story) per clause are higher than the zero values previously reported for normals (Harvey, 1983; Rochester & Martin, 1979). In the present study, psychopaths, when compared to nonpsychopaths, were found to use more incompetent references for the a f f e c t i v e , but not the neutral, story. A value of .11 i s f a r above that previously reported f o r normal individuals, but below .19, a value reported f o r TD schizophrenics (Harvey & Brault, 1986). Coherence Correlations shown in Table VI suggest that increasing psychopathy was associated with a f a i l u r e i n p l o t - u n i t closure fo r both the affect i v e and neutral s t o r i e s . As i s shown i n Table VII, f o r both story types psychopaths produced fewer completed plot-units than nonpsychopaths, about 65 versus 90 percent. These results suggest that the s t o r i e s produced by psychopaths were less coherent than those produced by nonpsychopaths. Discussion The most general conclusion that can be drawn from the res u l t s of t h i s study i s that psychopathy i s associated with a tendency to produce disordered communications. These communications appear to f a i l on a number of l e v e l s . F i r s t , -69-psychopathy i s associated with the use of few cohesive l i n k s between sentences. Secondly, psychopaths sometimes f a i l to provide the appropriate referent for what they are t a l k i n g about. T h i r d l y , they frequently introduce information that sets up expectations i n li s t e n e r s about what they might hear next, and then f a i l to provide that piece of information. F i n a l l y , psychopathy appears to be related to c l i n i c a l l y - r a t e d d e f i c i t s i n communication, as measured by the TLC. The l e v e l of impaired communication found i n the psychopathic subjects was somewhat unexpected. I t was thought that psychopaths would score higher than nonpsychopaths on p o s i t i v e thought disorder. This would be consistent with the c l i n i c a l observation that t h e i r speech tends to s l i p o f f track. The f i n d i n g that almost 30 percent of the psychopaths exhibited negative formal thought disorder, and that t h i s was due to poverty of speech and not poverty of content of speech, was su r p r i s i n g . Poverty of content of speech might have been expected, based on the c l i n i c a l impression that psychopaths speak a l o t but impart l i t t l e information. That they exhibited poverty of speech could be an a r t i f a c t of prison l i f e , i n that an inmate often does not discuss or divulge personal information. Alternatively, i t could be representative of more serious pathology. Andreasen and Grove (1986) found that negative thought disorder was associated with more long term impairment i n functioning than was p o s i t i v e thought disorder. - 7 0 -Negative thought disorder also appears to be a stable t r a i t of schizophrenia, whereas posi t i v e thought disorder does not (Docherty et a l . , 1988). A l l but one of the psychopaths in the present study exhibited a s i g n i f i c a n t amount of disorganized communication as measured by the TLC. This needs to be reconciled with some of the previous c l i n i c a l descriptions of psychopathy, and with the c l i n i c a l presentation of other thought disordered groups. Psychopaths have been described both as g l i b con a r t i s t s who can t a l k almost anyone into anything, and as i n d i v i d u a l s who produce poorly connected speech units . The findings of the present study would seem to support the l a t t e r . Reduced communicability, as measured by the TLC, was higher i n psychopaths than in nonpsychopaths. It i s possible that the pattern of the subcategories of thought disorder found f o r psychopaths may d i f f e r from that found for other p s y c h i a t r i c groups. For t h i s reason, psychopathic speech may communicate more information than that of, say, schizophrenics. A l t e r n a t i v e l y , i t could be that Hare's recent observations concerning psychopathic speech are correct. Rather than being the g l i b con a r t i s t s that they are commonly portrayed as, i t may simply be that t h e i r lack of s o c i a l anxiety coupled with a desire to be dominant in most s o c i a l s i t u a t i o n s makes them i n i t i a l l y a t t r a c t i v e to others. This would allow them to take advantage of some of the people some of the time, even i f , , i n -71-r e a l i t y , t h e i r a b i l i t y to communicate ve r b a l l y i s somewhat impaired. In the present study s p e c i f i c patterns of thought disorder, other than positive and negative, were not investigated. I t was predicted that psychopathy would be associated with positive signs, which i t was. The frequency of derailment and tangentiality that was found i n the speech of psychopaths would suggest that they frequently s l i p o f f top i c when they are speaking, and that they often f a i l to answer d i r e c t questions. Psychopaths exhibiting p o s i t i v e thought disorder might be expected to produce speech that i s poorly connected, but fluent. For the neutral story they did produce speech that was lower in cohesion than that of nonpsychopaths. They also produced fewer completed plo t - u n i t s for both story types. Both of these findings are i n d i c a t i v e of a r e l a t i v e l y high l e v e l of unconnected discourse i n psychopaths. However, fluent speech that changes topic quickly may not always sound strange to l i s t e n e r s ; without changes i n t o p i c , speech would become very boring. Schizophrenics or manics presenting with s i g n i f i c a n t thought disorder are usually immediately i d e n t i f i a b l e , without the aid of a c l i n i c a l r a t i n g scale. Psychopaths are not generally recognized as thought disordered. One possible reason for t h i s discrepancy i s that there may be differences i n the content of speech when comparing psychopaths and other thought disordered groups. The -72-TLC focuses on the form of speech and not i t s content. Often, both the form and content of schizophrenic speech are odd and bizarre. Idiosyncratic speech i s common i n thought disordered schizophrenics (Harrow et a l . , 1982). In psychopaths t h i s has not been observed. If a psychopath r e p l i e s tangentially but without strange content to a speaker's question, the speaker may simply assume that individual i s evasive or g l i b . Consider the following two responses to the question, "Do your moods go up and down?" 1. "I'm ju s t such a - uh - believer that - uh - that l i f e i s so short and that we're here for such a short time and so - so we're a l l going to die anyways at one stage so then - uh - you - we pass on into a t o t a l l y new s t r a t a and a l l the problems of t h i s world for us are solved and then we have a new set of problems and a new set of joys -whichever one - uh - i t s not something I claim to understand". 2. "Uh - up and down ? - well you know - some people worry and I don't worry about much". Both responses are tangential r e p l i e s to the question. However, the second response, which was made by a psychopath, would probably pass without comment i n a regular conversation, despite the f a c t that i t f a i l s to answer the question. The f i r s t response was produced by an i n d i v i d u a l that was excluded from the present study due to an RDC diagnosis of -73-schizophrenia. His checklist score was i n the nonpsychopathic range. His response was not only tangential but contained somewhat overly abstract,if not strange content. Most l i s t e n e r s would immediately be aware that there i s something odd about t h i s i n d i v i d u a l ' s speech and also would f i n d i t d i f f i c u l t to understand. I t i s possible that although the form of psychopaths' speech would be considered to be i n d i c a t i v e of some type of communication disorder, i t s content i s ordinary enough that l i s t e n e r s can e a s i l y i n f e r a meaning. Future research might focus more on l i s t e n e r s and how i t i s that they i n f e r meanings from the speech of psychopaths. For instance, psychopaths may not produce unusual content i n t h e i r speech, but instead exhibit d e f i c i t s i n the form of t h e i r discourse. I t i s possible that i t i s easier for a l i s t e n e r to i n f e r what an i n d i v i d u a l means i f form, rather than content, i s impaired. Rosenberg and Tucker (1979) found that the content of schizophrenic speech was d i s t i n c t from that of nonpsychiatric patients. The authors suggested that schizophrenics v i o l a t e the semantic baselines established by members of a language community. These violations do not have to be i n the form of bizarre content, but merely involve deviations from expected sex and s o c i a l roles (e.g. age or f a m i l i a r i t y ) . These types of deviations are related to the previously discussed concept of r e g i s t e r . In the present study there was no indication that any of the subjects f a i l e d to take into account the topic of the - 7 4 -story or, more generally, that they were to t e l l a story. Neither did the raters of the story t r a n s c r i p t s encounter bizarre or strange content. However they did sometimes encounter information that seemed confused or did not make sense. Consider the following story about being angry: "My most emotional experience was - uh - one of them - the most cl e a r e s t one was i n here when I was the - i n the middle of l i f e s k i l l class when somebody pushed the panic button by accident - I never had on a white t e e - s h i r t - I had a blue tee-shirt on - And - uh - they pulled me out there and centred me out i n front of c l - c l o - the crowd - Cause they had no r e a l reason to take anybody away -Because i t was by accident - So they decided to centre me out - Because the copper i n M9 didn't say that he didn't l i k e me - And he gave me a hassle - And said to centre me out - And made me go back to the u n i t and change -I didn't say anything - He was doing a l l the t a l k i n g - He asked me why I didn't have a blue t e e - s h i r t on - I answered h i s question by saying that a l l my white t e e - s h i r t s were i n the wash - He said "well go to SIS" - "And grab a green s h i r t then" - I said why" - "I'm already here" - "This -the c l a s s i s almost over" - And - uh - he said "too bad" -"Go back and do i t " in front of about 25 guys" (the panic button i s pushed when a guard i s i n trouble; SIS i s stores) . - 7 5 -Th i s story, the shortest produced, was t o l d by a psychopath who scored in the moderate range on poverty of speech and al s o exhibited derailment and loss of goal. This i n d i v i d u a l ' s confusion over t e e - s h i r t colours i s obvious. He also seemed to have d i f f i c u l t y i n producing a story with a clea r beginning and end. His i n t e r j e c t i o n of information concerning t e e - s h i r t s before i t was relevant suggests that he may have had d i f f i c u l t y in either organizing the information into a coherent account or that he f a i l e d to take into account the l i s t e n e r ' s needs. Bernstein (1966) has made a d i s t i n c t i o n between elaborated and r e s t r i c t e d speech codes. Restricted speech codes are those found among c u l t u r a l groups and the intent of the speaker i s conveyed not through verbal selections, but through changes i n gesture, physical set, and other such devices. When using a r e s t r i c t e d code, speech sequences are disju n c t i v e and concrete; propositions may not be f u l l y developed and gaps in logic can be present; meanings are discontinuous, but lis t e n e r s are able to i n f e r what the speaker means by attending to extraverbal channels. This type of speech i s often used between people who know each other well. Elaborate speech codes r e l y on the s e l f - e d i t i n g of information so that i t takes into account the speaker's r o l e i n a communication. The preparation and imparting of e x p l i c i t \ - 7 6 -information i s the major purpose of the elaborated code. I t i s possible that some subjects i n the present study produced s t o r i e s i n r e s t r i c t e d code form, perhaps that belonging to prison inmates. This would mean that the l i s t e n e r would have to make inferences about meanings based on information that i s c u l t u r a l l y foreign, since the speaker has f a i l e d to be sen s i t i v e to the register of the s i t u a t i o n . If t h i s were so, the story quoted above would make more sense to another inmate than i t did to the raters. This argument could be used to suggest that psychopaths are somehow in s e n s i t i v e to a l i s t e n e r ' s needs and therefore produce r e l a t i v e l y unconnected restricted-code speech regardless of the s o c i a l context. This would allow a psychopath to communicate with those i n a si m i l a r c u l t u r a l group without having to produce e x p l i c i t meanings. However, communication with those outside t h i s c u l t u r a l group would be impaired. In the present experiment, psychopathy was associated with a f a i l u r e to present story information that was coherent. Psychopaths presented information to open a pl o t - u n i t , and then d i d not include enough additional information i n subsequent discourse to l e t the l i s t e n e r know what happened. This suggests that expectancies were set up i n the l i s t e n e r to which the speaker was insensitive. I t also suggests that psychopathy was associated with a f a i l u r e to l i n k actions and resolutions within t h e i r narratives. This would make i t - 7 7 -d i f f i c u l t for the listener to understand what the speaker i s t r y i n g to say. Black and Reiser (1982) , i n t h e i r discussion of the s t r u c t u r a l models of comprehension, have suggested" that a model of text comprehension should explicate how knowledge structures are used to guide the r e t r i e v a l of information from a representation. They suggested that p l o t - u n i t s are useful i n understanding a story because they allow the l i s t e n e r to predict future story events. For example, i f a competition p l o t - u n i t i s i d e n t i f i e d , then the l i s t e n e r can i n f e r and expect that someone w i l l win and someone w i l l lose, based on previous knowledge structures. Black and Reiser (1982) also suggested that p l o t - u n i t constituents have another function i n text comprehension. They f a c i l i t a t e the r e c a l l of information. For example, an in d i v i d u a l may r e c a l l that there was a competition i n a story. In r e t r i e v i n g information about the competition, a strong expectation w i l l be set up as to the competition outcome. Additionally, the information i n other p l o t - u n i t s may become activated. I f an individual remembers that someone won the competition, then the expectancy for a rematch may be set up and memory may be searched for such information. In t h i s way, memory r e t r i e v a l from a representation begins with knowledge structures that organize information. In psychopaths i t i s possible that there i s some breakdown i n t h i s process, whereby - 7 8 -assessing one part of a plot-unit does not lead to the a c t i v a t i o n of i t s corresponding constituent, or related p l o t -u n i t s . I t i s also possible that the knowledge structures themselves could be faulty, whereby the p l o t - u n i t based expectations that psychopaths have are somehow d i f f e r e n t than those of other individuals. In the present study, t o t a l cohesion, l e x i c a l cohesion, and conjunctive cohesion i n the neutral story decreased as psychopathy increased. This was i n accordance with the p r e d i c t i o n that psychopaths would use l i t t l e cohesion o v e r a l l . I t suggests that the independent clauses i n the texts of psychopaths are not as well-linked as those of nonpsychopaths. For the a f f e c t i v e story t h i s r e l a t i o n s h i p did not hold. Although i t appeared that cohesion measures were topi c s e n s i t i v e i n t h e i r relationship to psychopathy, the actual r e l a t i o n s h i p seems to be one where both groups changed l e v e l s of cohesion across the two s t o r i e s , with the psychopaths s l i g h t l y increasing and the nonpsychopaths s l i g h t l y decreasing cohesion l e v e l s . This resulted i n no r e l a t i o n s h i p with the PCL-R for l e x i c a l cohesion i n the a f f e c t i v e story. The f i n d i n g that cohesion measures may be sensitive to t o p i c should be useful knowledge for future studies of text cohesion. Perhaps the most interesting finding i n the present study was that psychopaths have a tendency to produce incompetent references. This tendency did not d i s t i n g u i s h psychopaths from -79-nonpsychopaths i n the neutral story, but i t did i n the a f f e c t i v e story. Subjects tended to produce longer s t o r i e s , with more p l o t - u n i t s , in the a f f e c t i v e s t o r i e s than i n the neutral ones. I t i s possible that the increased l i n g u i s t i c processing demands required to produce the longer and more complex s t o r i e s , and to integrate more p l o t - u n i t s , resulted i n less resources being available for the e d i t i n g of inappropriate references. Cohen (1978) has demonstrated that reference f a i l u r e s increase i n normals, and to a larger extent in schizophrenics, as the task of communicating a referent becomes more d i f f i c u l t . Within the context of an emotional story, the task of communicating may have been more d i f f i c u l t for psychopaths than for nonpsychopaths. After reviewing the l i t e r a t u r e covering information processing and attentional functioning i n the developmental course of schizophrenic disorders, Nuechterlein and Dawson (1984) suggested that r e f e r e n t i a l communication d e f i c i t s (among others) are l i k e l y to be the r e s u l t of a reduction i n the processing capacity that i s available f o r task-relevant cognitive operation i n persons vulnerable to schizophrenia. In fact, the authors suggested that r e f e r e n t i a l communication d e f i c i t s may represent a v u l n e r a b i l i t y factor for schizophrenic disorders. Perhaps i n thought disordered schizophrenics a genetic predisposition operates (Harvey et a l , 1988) to l i m i t language-based resources. A similar -80-mechanism might underlie psychopaths' reference f a i l u r e s . Hare et a l . (1988), have suggested that the resources for processing language may be limited i n psychopaths. Hare and Gillstrom (1990) have suggested that psychopaths may produce short and poorly integrated sentences. In the present study, psychopathy was not associated with the MLU, or the number of clauses and words produced. This would suggest that psychopaths do not produce shorter utterances, or sentences, at l e a s t when t e l l i n g a story, but only that the utterances they do produce are poorly connected. There are a number of limitations to generalization of the findings of the present study. The f i r s t has to do with r e g i s t e r . In t h i s study subjects were asked to produce a monologue i n story form. This form of communication i s much d i f f e r e n t than, say, a two-way dialogue, or g i v i n g a prepared speech. I t i s not known i f the results obtained here would be reproduced across di f f e r e n t contexts and tasks. The r e l a t i v e proportions of the di f f e r e n t cohesion v a r i a b l e s to t o t a l cohesion are s e n s i t i v e to task (Rochester & Martin, 1979), and task complexity may affect the degree to which individuals produce unclear references (Cohen, 1978). A second l i m i t a t i o n has to do with the basic unit of the analysis. There were only about f i f t y clauses per story on which to base the present conclusions. This represents only a few minutes of speech. Although the e f f e c t s found i n these -81-samples were f a i r l y strong, the degree to which they would generalize to other stories and subjects i s not clear. Another l i m i t a t i o n of the current study has to do with how b l i n d the raters were concerning group membership. A l l of the r e l i a b i l i t y raters, except for the i n d i v i d u a l who scored cohesion and coherence and the two i n d i v i d u a l s who scored story a f f e c t i v i t y , had previously interviewed psychopaths. This may or may not have had an e f f e c t on t h e i r ratings. Another d i f f i c u l t y with interpreting the present r e s u l t s i s that thought disorder ratings were made on conversational speech patterns, whereas the dependent measures were derived from a story monologue. Some authors have i d e n t i f i e d s p e c i f i c passages of disordered speech for which cohesion and reference ratings were performed (Harvey, 1983). These passages contained s i g n i f i c a n t l y more incompetent references than did samples of non-disordered passages. In the present study the relati o n s h i p between psychopathy and the reference measures, and between psychopathy and thought disorder, might have been better c l a r i f i e d by looking s p e c i f i c a l l y at disordered speech samples. The coding of cohesion and reference f a i l u r e s i s a f a i r l y straightforward enterprise. However, to capture a l l of the information contained in a story, measures a d d i t i o n a l to p l o t -units should be used. Information about s e t t i n g , characters, and time could be included. Single-meaning constituents can be -82-used to assess information concerning actions (e.g., reactions, a c t i v i t i e s , announcements, narrator elaborations), and contextual constituents (e.g., location, contextual events, time) can be used to assess story background (Gillam, 1989). Future research might examine the degree to which these other aspects of psychopaths' s t o r i e s deviate from those of nonpsychopaths. For instance, meaning may be e a s i l y derivable from psychopaths' s t o r i e s because the stoies include large amounts of t h i s extra information. A d d i t i o n a l l y , the s t r u c t u r a l complexity of psychopaths' s t o r i e s could be evaluated by examining the relationships among pl o t - u n i t s within a story (Botvin & Sutton-Smith, 1977). F i n a l l y , the nature of the sample i n the present study was somewhat d i f f e r e n t from that found i n other studies of criminals. I t was not expected that four subjects would be l o s t due to severe psychopathology. The sample was also s l i g h t l y unusual i n the number of psychopaths i t contained. Usually they form about a t h i r d of most volunteer samples i n a prison population. As suggested above, the p o l i c y i n the prison at the time the sample was c o l l e c t e d may have affected i t s composition. Despite these caveats, the present r e s u l t s s t i l l suggest that a s i g n i f i c a n t number of psychopaths produce disordered communications. Psychopathy was associated with the use of r e l a t i v e l y few cohesive links between sentences, f a i l u r e to - 8 3 -provide appropriate referents in discourse, f a i l u r e to l i n k actions and resolutions in stories, and a s i g n i f i c a n t c l i n i c a l impairment i n the a b i l i t y to communicate. Generally, these re s u l t s suggest that e f f e c t i v e connections among speech units in psychopaths' discourse are not as numerous as those found i n nonpsychopaths. In addition, psychopaths l i k e l y suffer from a more general impairment in communication, r e l a t e d to, among other things, discourse that has a tendency to s l i p o f f track and that often f a i l s to d i r e c t l y answer a l i s t e n e r ' s questions. The r e l a t i o n s h i p between the communication d e f i c i t s exhibited by psychopaths and other psychopathological groups may be of in t e r e s t to future investigators. Whether or not these d e f i c i t s share a common developmental h i s t o r y , or underlying mechanism, may be of some importance to those researchers engaged i n the task of looking f o r markers of mental disorders. Future research on psychopaths might also investigate the rel a t i o n s h i p between d e f i c i t s i n language processing and production; a d e f i c i t i n one does not necessarily explain a d e f i c i t in the other. -84-References American Psychiatric Association. (1952). Diagnostic and s t a t i s t i c a l manual: Mental disorders. Washington, DC: Author. American Psychiatric Association. (1987). Diagnostic and s t a t i s t i c a l manual of mental disorders (3rd ed. rev.). Washington, DC: Author. Andreasen, N. C. (1979a). Thought, language and communication disorders: I. C l i n i c a l assessment, d e f i n i t i o n of terms, and evaluation of the i r r e l i a b i l i t y . Archives of General Psychiatry, 36, 1315-1321. Andreasen, N. C. (1979b). Thought, language and communication disorders: I I . Diagnostic significance. Archives of General Psychiatry, 36, 1321-1330. Andreasen, N. C. (1980). Manual for use with the scale for Thought. Language and Communication Disorders. Iowa City: University of Iowa. Andreasen, N. C., & Grove, W. M. (1986). Thought, language and communication i n schizophrenia: Diagnosis and prognosis. Schizophrenia B u l l e t i n , 12, 348-359. Andreasen, N. C., & Powers, P. S. (1975). C r e a t i v i t y and psychosis: An examination of conceptual s t y l e . Archives of General Psychiatry, 32., 70-73. Andreasen, N. C , Hoffman, R. E., & Grove (1985). Mapping abnormalities i n language and cognition. In M. Alpert (Ed.). Controversies in schizophrenia. New York: Guilford Press. Berstein, B. (1966). Elaborated and r e s t r i c t e d codes: Their s o c i a l origins and some consequences. In A. G. Smith (Ed.), Communication and culture: Readings i n the codes of human inte r a c t i o n . New York: Holt, Rhienhart and Winston. Bleuler, E. (1911). Dementia praecox or the group of the schizophrenias. Translated (1950) by J. Zinkin. New York: International Universities Press. Botvin, G. J., & Sutton-Smith, B. (1977). The development of st r u c t u r a l complexity in children's fantasy narratives. Developmental Psychology, 13, 1-388. -85-Brown, G. , & Yule, G. (1983). Discourse analysis. New York: Cambridge University Press. Brownell, H. H., Potter, H. H., & Michelow, D. (1984). S e n s i t i v i t y to l e x i c a l denotation and connotation i n brain-damaged patients: A double diss o c i a t i o n ? Brain and Language, 22., 253-265. Butterworth, B., & Beattie, G. (1976). Gestures and silence as indicators of planning in speech. In R. N. Campbell and P. T. Smith (Eds.). Recent advances i n the psychology of language. New York: Plenum Press. Chaika, E. (1974). A l i n g u i s t looks at "schizophrenic" language. Brain and Language, 1, 257-276. Cicone, M., Wapner, W. , F o l d i , N., Zurif, E., & Gardner, H. (1979). The r e l a t i o n between gesture and language i n aphasia communication. Brain and Language. 8, 324-349. Cleckley, H. (1976). The mask of sanity (5tn ed.). St. Louis, MO: Mosby. Cohen, B. D. (1978). Referent communication disturbances i n schizophrenia. In S. Schwartz (Ed.), Language and cognition i n schizophrenia. Lawrence Erlbaum: H i l l d a l e , New Jersey. Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioural sciences (2nd ed.). H i l l s d a l e , New Jersey: Lawrence, Erlbaum. Craft, M. J. (1965). Ten studies i n the psychopathic personality. Bristol:Wright. Cramer, P., Weegman, M., & O'Neil, M. (1989). Schizophrenia and the perception of emotions: How accurately do schizophrenics judge the emotional states of others? B r i t i s h Journal of Psychiatry, 155, 225-228. Docherty, N., Schnur, M., & Harvey, P.D. (1988). Reference performance,and p o s i t i v e and negative thought disorder: A follow-up study of manics and schizophrenics. Journal of Abnormal Psychology, 97. 437-422. Eich l e r , M. (1965). The application of verbal behaviour analysis to the study of psychological defense mechanisms: Speech patterns associated with sociopathic behaviour. The -86-Journal of Nervous and Mental Disease, 141, 658-663. E l l i s , J . , & lire, J. (1969). Registers. In A. R. Meetham (Ed.), Encylcopedia of l i n g u i s t i c s : Information and control. Oxford: Pergamon. Foreman, M. (1988). Psychopathy and interpersonal behavior. Unpublished doctoral dissertation, U n i v e r s i t y of B r i t i s h Columbia, Vancouver, Canada. Fine, J. (1978). Conversation, cohesive and thematic patterning in children's dialogues. Discourse Processes, 1, 247-266. Gacono, C B . (1990). An empirical study of object r e l a t i o n s and defensive structure in a n t i s o c i a l personality disorder. Journal of Personality Assessment, 54, 589-600. Gacono, C.B., & Meloy, J.R. (1988). The r e l a t i o n s h i p between cognitive s t y l e and defensive process i n the psychopath. Criminal Justice and Behavior, 15, 472-483. Gillstrom, B. J. & Hare, R. D. (1988). Language r e l a t e d hand gestures i n psychopaths. Journal of Personality Disorders. 2, 21-27. Gillam, R. (1989). An investigation of the o r a l language, reading, and written language competencies of language impaired and normally achieving school-age children. Unpublished doctoral dissertation, U n i v e r s i t y of B r i t i s h Columbia, Vancouver. Grant, V. (1977). The menacing stranger. New York: Dabor Science. Gulich, E. & Quasthoff, U. M. (1982). Narrative analysis. In T. A. Van Dijk (Ed.), Handbook of discourse analysis (vol. 2). Orlando, F l o r i d a : Academic Press. Halliday, M. A. K., & Hasan, R. (1976). Cohesion i n English. London: Longman. Hare, R. D. (1970). Psychopathy: Theory and research. New York: Wiley. Hare, R. D. (1978) . Electrodermal and cardiovascular correlates of psychopathy. In R. D. Hare and D. S c h a l l i n g (Eds.), Psychopathic behaviour: Approaches to research (pp. 107-143) . London: John Wiley & Sons. Hare, R. D. (1979) . Psychopathy and l a t e r a l i t y of cerebral -87-function. Journal of Abnormal Psychology, 88, 605-610. Hare, R. D. (1980). A research scale for the assessment of psychopathy in criminal populations. Personality and Individual Differences, 1, 111-119. Hare, R. D. (1982). Psychopathy and ph y s i o l o g i c a l a c t i v i t y during a n t i c i p a t i o n of an aversive stimulus i n a d i s t r a c t i o n paradigm. Psychophysiology. 19, 266-271. Hare, R. D. (1983). Diagnosis of a n t i s o c i a l p e rsonality disorder i n two prison populations. American Journal of Psychiatry, 140, 887-890. Hare, R. D. (1985). Comparison of procedures f o r the assessment of psychopathy. Journal of Consulting and C l i n i c a l Psychology, 53, 7-16. Hare, R.D. (1991) . The Hare Psychopathy Checklist - Revised. Toronto, Ontario: Multi-Health Systems. Hare, R. D. & Craigen, D. (1974). Psychopathy and physiological a c t i v i t y i n a mixed-motive game s i t u a t i o n . Psychophysiology, 11, 197-206. Hare, R. D., Frazelle, J ., Bus, J., & J u t a i , J . W. (1980). Psychopathy and the structure of primary mental a b i l i t i e s . Journal of Behavioural Assessment. 2, 77-88. Hare, R. D., & Gillstrom, B. J. (1991). Hand gestures and speech encoding d i f f i c u l t i e s in psychopaths. Manuscript submitted for publication. Hare, R. D., Harpur, T. J., Hakstian, A. R., Forth, A. E., Hart, S. D., & Newman, J. P. (1990). The re v i s e d psychopathy c h e c k l i s t : R e l i a b i l i t y and f a c t o r structure. Psychological Assessment: A Journal of Consulting and C l i n i c a l Psychology. 2, 338-341.. Hare, R. D. & J u t a i , J. W. (1988) . Psychopathy and cerebral asymmetry i n semantic processing. Personality and Individual Differences. 9, 329-337. Hare, R. D. , & McPherson, L. M. (1984). Psychopathy and perceptual asymmetry during verbal d i c h o t i c l i s t e n i n g . Journal of Abnormal Psychology. 93., 141-149. Hare, R. D., Williamson, S., & Harpur, T. J . (1988). Psychopathy and language. In T. E. Moffit & S. A. Mednik (Eds.), B i o l o g i c a l contributions to crime causation (pp. -88-68-92). Dordrecht, Netherlands: Nijhoff Martinus. Harpur, T. J . , Hakstian, A. R., & Hare, R. D. (1988). The factor structure of the Psychopathy Ch e c k l i s t . Journal of Consulting and C l i n i c a l Psychology, 56, 741-747. Harpur, T. J . , Hare, R. D., & Hakstian, A. R. (1989). Two-factor conceptualization of psychopathy: Construct v a l i d a t i o n and assessment implications. Psychological Assessment: A Journal of Consulting and C l i n i c a l Psychology, l , 6-17. Harrow, M., Grossman, L. S., S i l v e r s t e i n , M. L. , & Meltzer, H. Y. (1982). Thought pathology in manic and schizophrenic patients: Occurrence at hospital admission and seven weeks l a t e r . Archives of General Psychiatry. 39, 665-671. Harrow, M., & Quinlan, D. (1977). Is disordered thinking unique to schizophrenia? Archives of General Psychiatry, 34., 15-21. Hart, S.D., Forth, A.E., & Hare, R.D. (1990). Performance of male psychopaths on selected neuropsychological t e s t s . Journal of Abnormal Psychology, 99, 374-379. Harvey, P. D. (1983) . Speech competence i n manic and schizophrenic psychosis: The association between c l i n i c a l l y rated thought disorder and cohesion and reference performance. The Journal of Abnormal Psychology. 92, 368-377. Harvey, P.D., & Brault, J. (1986). Speech performance i n mania and schizophrenia: The association of p o s i t i v e and negative thought disorders and reference f a i l u r e s . Journal of Communication Disorders, 19, 161-173. Harvey, P. D., Earle-Boyer, E. A., & Levinson, J . C. (1988). Cognitive d e f i c i t s and thought disorder: A r e t e s t study. Schizophrenia B u l l e t i n . 14, 57-66. Harvey, P. D., Earle-Boyer, E. A., & Weilgus, M. S. (1984). The consistency of thought disorder i n mania and schizophrenia. The Journal of Nervous and Mental Disease, 172. 458-463. Harvey, P. D., Weintraub, S., & Neale, J . M. (1982). Speech competence of children vulnerable to psychopathology. Journal of Abnormal Child Psychology. 10, 373-388. Hemphill, J . F., Hart, S. D., & Hare, R. D. (1990). Substance -89-abuse by criminal psychopaths. Manuscript submitted for publication. Hoffman, R. E., Kirsten, L., Stopek, S., & C i c c h e t t i , D. V. (1982). Apprehending schizophrenic discourse: A s t r u c t u r a l analysis of the l i s t e n e r ' s task. Brain and Language. 15, 207-233. Hymes, D. (1964). Toward ethnographies of communicative events. In P. P. G i g l i o l i (Ed.), Language and s o c i a l context. Hammondsworth, England: Penguin. Mandler, J. M. (1984). Stories, s c r i p t s , and scenes: Aspects of schema theory. London: Lawrence Erlbaum. McCord, W. & McCord, J. (1956). Psychopathy and delinguency. New York: Grune & Stratton. McCutchen, D. & P e r f e t t i , C. A. (1982). Coherence and connectedness i n the development of discourse production. Text. 2, 113-139. McNeill, D. (1985). So you think gestures are nonverbal? Psychological Review, 92, 350-371. McNeill, D., & Levy, E. (1982). Conceptual representations i n language a c t i v i t y and gesture. In R. J. J a r v e l l a and W. Klein (Eds.), Speech, place and action: Studies i n dyslexia and re l a t e d topics. Chichester, England: Wiley. Meehl, P. E. (1962). Schizotaxia, Schizotypy, Schizophrenia. American Psychologist, 17, 827-838. Newman, J. P. & Kosson, D. S. (1986). Passive avoidance learning i n psychopathic and nonpsychopathic offenders. Journal of Abnormal Psychology, 95, 252-256. Millon, T. (1981). Disorders of personality DSM I I I : Axis I I . New York: John Wiley & Sons. Neuchterlein, K. H., & Dawson, M. E. (1984). Information processing and attention functioning i n the developmental course of schizophrenic disorders. Schizophrenia B u l l e t i n , 10, 160-198. Nicholas, M., Obler, L. K. , Albert, M. L., & Helm-Easterbrooks, N. (1985). Empty speech i n alzheimer's disease and fluent aphasia. Journal of Speech and Hearing Research. 28, 405-410. -90-Ochs, E. (1979) . Planned and unplanned discourse. In T. Givon (Ed.), Syntax and semantics (vol. 12, p. 51-80). New York: Academic Press. Oltmanns, T. F., Murphy, R. , Berenbaum, H., & Dunlop, S. R. (1985). Rating Verbal communication impairment i n schizophrenia and a f f e c t i v e disorders. Schizophrenia B u l l e t i n . 2, 292-299. Omanson, R. C. (1982). An analysis of narratives: Identifying central, supportive, and d i s t r a c t i n g content. Discourse Processes, 5, 195-224. Osgood, C. E., May, W.H., & Miron, M. S. (1975). Cross-cultural Universals of A f f e c t i v e Meaning. Chicago, Univ. I l l i n o i s Press. Patrick, C.J., & Lang, P.J. (1989). Psychopathy and emotion i n a forensic population. Paper presented at the Meeting of the Society for Psychophysiological Research, New Orleans. Pogue-Geile, M. F., & Harrow, M. (1984). Negative and p o s i t i v e symptoms i n schizophrenia and depression: A followup. Schizophrenia B u l l e t i n . 10, 371-387. Ragin, A. B. & Oltmanns, T. F. (1986). Lexical cohesion and formal thought disorder during and af t e r psychotic episodes. Journal of Abnormal Psychology. 95, 181-183. Pichot, P. (1978). Psychopathic behaviour: A h i s t o r i c a l overview. In R. D. Hare and D. Schalling (Eds.), Psychopathic behaviour: Approaches to research (pp. 55-70). London: John Wiley & Sons. Polanyi, L. (1985). Conversational s t o r y t e l l i n g . In T. A. Van Dijk (Ed.), Handbook of discourse analysis ( v o l . 3). Orlando, F l o r i d a : Academic Press. Ragin, A. B., & Oltmanns, T. F. (1983). P r e d i c t a b i l i t y as an index of impaired verbal communication i n schizophrenic and a f f e c t i v e disorders. B r i t i s h Journal of Psychiatry, 143, 578-583. Reiser, B. J . , & Black, J . B. (1982). Processing and s t r u c t u r a l models of comprehension. Text, 2, 225-252. Rochester, S. R., & Martin, J. R., (1979). Crazy t a l k : A study of the discourse of schizophrenic speakers. New York: Plenum. -91-Rochester, S. R., Martin, J. R., & Thurston, S. (1977). Thought-process disorder in schizophrenia: The l i s t e n e r ' s task. Brain and Language, 4, 95-114. Rosenberg, S.D., & Tucker, G.J. (1979). Verbal behaviour i n schizophrenia: The semantic dimension. Archives of General Psychiatry. 36. 1331-1337. Smith, S.S., & Newman, J.P. (1990). Alcohol and drug abuse/dependence disorders i n psychopathic and nonpsychopathic criminal offenders. Journal of Abnormal Psychology, 99, 430-439. Spitzer, R. L., Endicott, J., & Robins, L. (1979). The schedule for a f f e c t i v e disorders and schizophrenia - Lifetime Version (3rd. ed.). New York, Biometrics Research. Spitzer, R. L., Endicott, J., & Robins, L. (1989). Research diagnostic c r i t e r i a for a selected group of functional disorders (3rd. ed.). New York, Biometrics Research. Sutton-Smith, B., Botvin, G., & Mahony, D. (1976). Developmental structures in fantasy narratives. Human Development, 19, 1-13. van Dijk, T., A. (1985). Semantic discourse a n a l y s i s . In T. van Dijk (Ed.), Dimensions of discourse (vol. 2, pp. 103-136). London: Academic Press Weintraub, W., & Aronson, H. (1962). The a p p l i c a t i o n of verbal behaviour analysis to the study of psychological defense mechanisms. The Journal of Nervous and Mental Disease, 139. 169-181. Wiggins, J . S., P h i l l i p s , N., & Trapnell, P. (i n press). C i r c u l a r reasoning about interpersonal behavior: Evidence concerning some untested assumptions underlying diagnostic c l a s s i f i c a t i o n . Journal of Personality and S o c i a l Psychology. Williamson, S., Crockett, D., Hurwitz, T., & Remick, R. (in press). The processing of emotional information by depressed and organic patients. Archives of C l i n i c a l Neuropsychology. Williamson, S., Harpur, T. J., & Hare, R. D. (1990). S e n s i t i v i t y to emotional po l a r i t y i n psychopaths. Paper presented at the Meeting of the American Psychological Association, Boston. -92-Williamson, S., Harpur, T. J., & Hare, R. D. (1991). Abnormal processing of a f f e c t i v e words by psychopaths. Psychophysiology, 28, 260-273. Wong, S. (1984). Criminal and i n s t i t u t i o n a l behaviors of psychopaths. In Programs Branch Users Report, M i n i s t r y of the S o l i c i t o r General of Canada, Ottawa, Ontario. Wykes, T. & Leff, J. (1982) . Disordered speech: Differences between manics and schizophrenics. Brain and Language, 15 117-124.