Emotion in the Criminal Psychopath: Startle Reflex Modulation
Total Page:16
File Type:pdf, Size:1020Kb
Journal of Abnormal Psychology Copyright 1993 by the American Psychological Association, Inc. 1993, Vol. 102, No. 1.82-92 0021-843X/93/$3.00 Emotion in the Criminal Psychopath: Startle Reflex Modulation Christopher J. Patrick, Margaret M. Bradley, and Peter J. Lang Startle-elicited blinks were measured during presentation of affective slides to test hypotheses concerning emotional responding in psychopaths. Subjects were 54 incarcerated sexual offenders divided into nonpsychopathic, psychopathic, and mixed groups based on file and interview data. Consistent with findings for normal college students, nonpsychopaths and mixed subjects showed a significant linear relationship between slide valence and startle magnitude, with startle responses largest during unpleasant slides and smallest during pleasant slides. This effect was absent in psychopaths. Group differences in startle modulation were related to affective features of psychop- athy, but not to antisocial behavior per se. Psychopathy had no effect on autonomic or self-report responses to slides. These results suggest an abnormality in the processing of emotional stimuli by psychopaths that manifests itself independently of affective report. Abnormal or deficient emotional responding is considered learn to inhibit punished responses (see reviews by Hare, 1978a, to be a hallmark of psychopathy. Cleckley's (1955) classic diag- and Siddle & Trasler, 1981; see also Newman, Widom, & nostic criteria for psychopathy include absence of nervousness, Nathan, 1985). These data have led to speculation that psychop- lack of remorse or shame, egocentricity and incapacity for love, athy is related to a specific deficit in neurophysiological systems and general poverty in major affective reactions. He believed modulating fear behavior (Fowles, 1980; Gray, 1971). that a discordance between linguistic and experiential compo- One problem in this research area has been the absence of a nents of emotion, a condition he termed semantic dementia, consensus regarding diagnostic criteria for psychopathy. Many defined the essence of psychopathy. From this standpoint, the studies have used psychometric data (e.g., Minnesota Multipha- psychopath knows the "words" of emotion, but not the "music" sic Personality Inventory scores) to classify subjects, whereas (cf. Johns & Quay, 1962). others have equated psychopathy with criminality. In fact, a Lykken (1957) provided empirical evidence that psychopaths criticism of the Diagnostic and Statistical Manual of Mental defined by Cleckley's criteria are deficient in their capacity to Disorders (rev. 3rd ed.; American Psychiatric Association, 1987) develop anxiety responses. Subsequent empirical studies of criteria for antisocial personality disorder (a category that was emotion in psychopaths have focused largely on the responses intended to capture the essential features of psychopathy) has of these individuals in punishing or threatening situations. The been that criminal behaviors are overemphasized to the exclu- most consistent findings have been that psychopaths show re- sion of affective and interpersonal features believed to be patho- duced electrodermal response during anticipation of a noxious gnomonic of psychopathy (Hare, 1985; Harpur, Hare, & Haks- event and poor passive avoidance learning, that is, a failure to tian, 1989). A second difficulty of interpretation in psychophysiological Christopher J. Patrick, Margaret M. Bradley, and Peter J. Lang, De- studies of aversive responding in psychopaths arises from their partment of Clinical and Health Psychology and The Center for Re- near exclusive reliance on electrodermal and cardiovascular search in Psychophysiology, University of Florida. measures. Recent research suggests that neither measure in- This research was supported by Grants MH37757, MH41950, and dexes fear reliably. Electrodermal activity tends to increase as MH43975 from the National Institute of Mental Health to Peter J. emotional arousal increases, whether the eliciting stimulus is Lang. Preliminary findings were presented at the annual meeting of pleasant or aversive (Greenwald, Cook, & Lang, 1989). Heart the Society for Psychophysiological Research, October 1990, Boston. rate, on the other hand, appears to be sensitive to the task de- We are grateful to the staff and residents of the North Florida Evalua- mands of an experimental situation; the heart rate response to tion and Treatment Center (NFETC), who made this research possi- ble. We are especially indebted to Ted Shaw and Drew King, who an aversive stimulus, for example, may be either acceleratory or provided front-line assistance throughout the project, and to the direc- deceleratory, depending on the method used to evoke emotion tor of NFETC, A. Dennis Geis, and his assistant, Brenda Ormsbee, for (e.g., visual stimulus vs. imaginal; Lang, Bradley, & Cuthbert, their administrative support. Robert Bruner and Kevin Lay played 1990; Vrana & Lang, 1990). vital roles in both the diagnostic and testing phases of the study. We Psychophysiological studies have generally found reduced also thank Bruce Cuthbert for his technical and methodological con- skin conductance responses in psychopaths during anticipation tributions. of punishment. With regard to heart rate, some experiments Correspondence concerning this article should be addressed either have not found differences between psychopathic and nonpsy- to Christopher J. Patrick, who is now at the Department of Psychology, Florida State University, Tallahassee, Florida 32306, or to Peter J. chopathic subjects in threatening situations (e.g., Hare & Quinn, Lang, Center for Research in Psychophysiology, Box J-165, JHMHC, 1971; Tharp, Maltzman, Syndulko, & Ziskind, 1980), but Department of Clinical and Health Psychology, University of Florida, others have reported increased heart rate acceleration along Gainesville, Florida 32610. with decreased electrodermal activity (Hare & Craigen, 1974; 82 PSYCHOPATHY AND STARTLE 83 Hare, Frazelle, & Cox, 1978; Ogloff & Wong, 1990). Hare vides a direct index of the valence disposition of the organism (1978a) attempted to explain this discordant pattern of electro- (appetitive or defensive) in relation to putatively emotional dermal and cardiovascular responding by postulating that psy- stimuli. chopaths adopt an active coping stance in the face of threat that There is independent evidence linking larger startle re- reduces anxiety (see also Fowles, 1980). Siddle and Trasler sponses to fear states. Studies of classical conditioning in ani- (1981) cited several problems with this interpretation and ques- mals and humans have demonstrated augmented acoustic star- tioned whether the impact of an impending aversive stimulus is tle during presentation of a conditioned stimulus previously attenuated by psychopaths in this fashion. What is needed to paired with shock (Brown, Kalish, & Farber, 1951; Greenwald, resolve this issue is a response measure that indexes aversive Hamm, Bradley, & Lang, 1988). In related research with ani- emotional activation independently of arousal and experimen- mals, Davis (1979,1986) demonstrated that this effect can be tal task demands. blocked by administration of antianxiety drugs and that activ- ity in the region of the amygdala probably plays a mediating role. The startle response thus provides a potential solution to Emotional Valence and the Startle Response the problem of fear measurement in psychophysiological re- Lang (1985) outlined a dimensional theory of affect in which search. emotional response is conceptualized fundamentally as an ac- tion disposition. Emotional responses to stimuli or situations are denned in terms of two orthogonal dimensions: affective Research Questions valence and arousal. Valence refers to the directionality of the Affective Modulation of Startle in a Criminal Population elicited action disposition, varying from avoidance to ap- proach: arousal refers to the intensity of activation of the re- To date, all investigations of the affect-startle phenomenon sponse. The emotion of fear, for example, may be conceptual- have been conducted with normal college student subjects. The ized as a state of low valence (involving preparation for avoid- present study examined emotional responses to pleasant, neu- ance or withdrawal) and high arousal or activation. tral, and unpleasant visual stimuli in a sample of convicted As noted earlier, electrodermal response provides a reliable criminal offenders. Measures included physiological responses measure of emotional arousal, but not of valence. Changes in (startle, corrugator electromyographic [EMG] activity, heart facial activity appear to index valence (Greenwald et al., 1989) rate, and skin conductance), affective reports (pleasantness, but can be problematic because they are subject to voluntary arousal, and interest), and behavior (stimulus-viewing time). We distortion (Craig, Hyde, & Patrick, 1991). Recent research indi- were interested in how patterns of responses in this sample of cates that the probe-startle response may provide an alterna- criminal offenders would compare with the responses of nor- tive measure of emotional valence with unique advantages mal subjects. To permit a direct comparison, stimulus mate- (Lang et al., 1990; Vrana, Spence, & Lang, 1988). Studies with rials and experimental procedures in the present study paral- normal subjects have repeatedly shown that reflex eyeblink re- leled those used in earlier research with college students (Brad- sponses to a sudden, intense