Reflectivity and Learning From Aversive Events: Toward a Psychological Mechanism for the Syndromes of Disinhibition

Transcription

1 Psychological Review 1993, Vol. 100, No. 4, Copyright 1993 by the American Psychological Association, Inc X/93/S3.00 Reflectivity and Learning From Aversive Events: Toward a Psychological Mechanism for the Syndromes of Disinhibition C. Mark Patterson and Joseph P. Newman Gorenstein and Newman (1980) proposed that poorly modulated responding for reward is the common diathesis underlying disinhibited behavior in several traditionally distinct person categories: psychopathy, hysteria, early onset alcoholism, childhood hyperactivity, and nonpathological impulsivity (e.g., extraversion). The authors extend this proposal by theorizing a psychological mechanism that highlights relations among disinhibition, reflection, and failures to learn from aversive feedback. The hypothesized mechanism is presented as 4 generic stages of response modulation: the dominant response set, the reaction to an aversive event, the subsequent behavioral adaptation, and the immediate and long-term consequences of reflection, or the lack thereof. The mechanism has implications for disinhibited individuals' impulsivity and provides a point of departure to study factors responsible for similarities and differences among these syndromes. Animals with septal lesions do not freeze when punished (Blatt, 1976). They are poor at passive avoidance (McCleary, 1966) and at delaying gratification (Newman, Gorenstein, & Kelsey, 1983). They often appear impulsive, and by some accounts they are fearless (see Gray & McNaughton, 1983). Their poor performance on differential reinforcement of low rates schedules suggests a deficit in forming associations between reward-seeking responses and frustrative nonreward (Ellen, Wilson, & Powell, 1964). Septal lesions facilitate rewarded responding, active avoidance, and resistance to extinction (Gray, 1982). Empirical phenomena like these were the basis of Gorenstein and Newman's (1980) proposal that the observable effects of septal dysfunction in animals are quite analogous to the disinhibited behavior of psychopaths, hysterics, early onset alcoholics, hyperactive children, and nonpathological impulsive personalities exemplified by the Eysenckian (e.g., H. J. Eysenck, 1967)extravert. During the past two decades there has been a confluence of theoretical writing and empirical research on relations between a possible neuropsychological anomaly in the limbic system and human syndromes of disinhibition (e.g., Elliott, 1978; Fowles, 1980; Gray, 1987a; Hare, 1970; Mawson & Mawson, 1977; Ro- C. Mark Patterson and Joseph P. Newman, Department of Psychology, University of Wisconsin-Madison. Preparation of this article was supported in part by National Institute of Mental Health Grant MH to Joseph P. Newman. We wish to express appreciation to Jo-Anne Bachorowski, Eric Howland, David Kosson, Sharon Nichols, and Debra Pearce-McCall for their contributions. We also wish to acknowledge our anonymous reviewers for their thoughtful comments. Correspondence concerning this article should be addressed to C. Mark Patterson, who is now at the Department of Veterans Affairs Medical Center, 3801 Miranda Avenue, Psychology Service, 116B1, Palo Alto, California 94304, or to Joseph P. Newman, Department of Psychology, 1202 West Johnson Street, University of Wisconsin, Madison, Wisconsin senthal& Allen, 1978; Tarter, 1978;Trasler, 1978). Familial and genetic connections among these syndromes have also been proposed (see Cloninger, Reich, & Guze, 1978; H. J. Eysenck, 1981; Lilienfeld & Waldman, 1990; Tarter, 1983). Gorenstein and Newman's (1980) septal lesion model has been a springboard for the theoretical and empirical work on the psychological mechanism underlying human disinhibition presented in this article. Consider the psychopath. Poor judgment is a prominent feature of this syndrome. The impulsive quality of their hostile and hedonistic behaviors suggests that they often fail to use available information to anticipate risks. Thus, although they might gratify their impulse, they commonly incur unanticipated adverse consequences, such as social alienation, incarceration, or bodily harm. Well-adjusted extraverts, in contrast, though less prone to maladaptive judgment, are typified by relatively uninhibited, sometimes impertinent sociability, which also seems related to a lack of anticipation of social reproach (H. J. Eysenck, 1967). Both groups tend toward unchecked action. We argue that their failures to learn from aversive experience are caused by a psychological process that links impulsive and disinhibited behavior. Specifically, given certain motivational circumstances, disinhibited individuals are less apt to pause and reflect on the situation, thereby diminishing their ability to anticipate similar aversive events in the future. This curtailed reflectivity (cf. Kagan, 1966) is central to the proposed psychological diathesis. The research on human disinhibition that has been prompted by the septal lesion model has developed enough to serve as the basis for generating hypotheses about core psychological processes that underlie the phenotypic commonalities among the syndromes. As the septal lesion literature grows, the model may continue to influence research with humans. However, rather than updating the animal model, the main purpose of this article is to present a theoretical perspective on human disinhibition that is based primarily on human research. To introduce the theory, we review briefly its origins in the septal lesion model. 716

2 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 717 The Septal Lesion Model The model derives from the cluster of behavioral traits known as the septal syndrome caused by a septal lesion. Because the model is not intended to elucidate a neuronal mechanism for disinhibition, it does not rely on anatomical homologies to aid theory construction. Instead, the model relies on extensive face valid parallels between the behavior of animals with septal lesions and disinhibited humans. Its utility lies in its heuristic potential to guide elaboration of the nomological network around the core processes of human disinhibition. Even though the major thrust of this article is psychological, the role of biological substrates cannot be ignored. As research and theory building continue, we expect a more balanced biopsychosocial view of human disinhibition to emerge. An attractive feature of the septal model is that it permits careful scrutiny of a set of seemingly diverse disorders through one cross-sectional lens. Such an approach has the potential to identify core features of human disinhibition. The model also intends to avoid theoretical distortions that could arise from focusing on a prominent symptom from one syndrome as the hallmark of a common psychological denominator. For example, psychopathic behavior has been attributed to reduced sensitivity to punishment (see Hare, 1970, 1978). This hypothetical trait might play a role in psychopaths' disinhibition but be insignificant for the other syndromes. The animal model also has limitations. First, some aspects of human disinhibition may not have compelling analogs in animals with septal lesions. Second, and conversely, not all features of disordered septal function are likely to have valid human parallels. Third, no single universally accepted characterization of septal dysfunction exists. Indeed, the septum is multifunctional. Thus, as the nomological network around human disinhibition becomes better articulated, the model's heuristic power could become more limited as continuing research relies more on findings from the human literature. Gorenstein and Newman's (1980) specification of a unifying hypothetical construct for human disinhibition targeted what may be the septal syndrome's hallmark: a marked tendency toward response perseveration. McCleary (1966) described perseveration as an exaggerated tendency to emit a dominant response despite punishment, extinction, or contingency reversal. Emphasizing the importance of motivational factors, he noted that suppressive failures of animals with septal lesions occur when a dominant approach response set exists amidst inhibitory cues. This view implies two prerequisites for perseveration: a reward incentive to respond coupled with a punishment incentive to inhibit that same response. Although failures of punishment to deter approach behavior might indicate "insensitivity to punishment," Gorenstein and Newman suggested instead that hypersensitivity to reward might lead humans to perseverate, resulting in failures to avoid punishment. Response perseveration is epitomized by poor passive avoidance in both septal lesion animals (e.g., McCleary, 1966) and disinhibited humans (e.g., Trasler, 1978). In humans, this deficit is perhaps best illustrated by the repeated failures to learn from experience and the resulting chronic recidivism of criminal psychopaths. However, response perseveration is simply an observable sign of disinhibition, more descriptive than explicative. Effective theory building requires hypothetico-deductive methods to explain relations among psychological constructs, such as conditioning and disinhibition, and observable signs of those constructs. Passive avoidance means withholding a response (often an appetitively motivated one) to avoid punishment by learning cues that predict it- H. J. Eysenck (1967) described the process of conscience development as one of learning from authority figures which socially inappropriate behaviors must be suppressed to avoid punishment. Given this important role in human socialization, passive avoidance learning merits a central role in theorization about human disinhibition for two reasons. First, the clinical syndromes of disinhibition are relatively uninclined to suppress socially prohibited behaviors (e.g., Aronfreed, 1969; Blackburn, 1983). Second, the passive avoidance paradigm is a convenient experimental microcosm wherein key aspects of a psychological mechanism of perseveration and, therefore, human disinhibition may be described and refined theoretically. The paradigm readily incorporates rewards, punishments, the need to adapt ongoing behavior according to changes in the hedonic valence of environmental contingencies, and opportunities to learn from aversive events. The primary question we address with the paradigm is how does disinhibition contribute to failures to learn from experience? Once a proximal mechanism of these failures to learn is characterized in one syndrome, the passive avoidance paradigm can be used to study its generalizability vis-a-vis disinhibited behavior in the others. Using this paradigm is not an end in itself, however, for the theory addresses failures to learn from aversive events in a variety of situations. The theory focuses on response modulation (see McCleary, 1966). Response modulation is a complex process involving temporary suspension of a dominant response set and a brief concurrent shift of attention from the organization and implementation of goal-directed responding to its evaluation. Adaptive response modulation also involves making any degree of adjustment to the response set once the corrective information is accommodated, including response inhibition, selection of an alternative response strategy, or making no adjustment if the response is judged to be appropriate. Thus, the process may also involve comparing the current stimulus context with past associations to evaluate the adaptiveness of available response options. Inhibition and disinhibition are possible outcomes of this process. Moreover, the adaptiveness of both can vary greatly depending on the nature of response contingencies. Typically, disinhibition is maladaptive insofar as it involves a failure to pause and accommodate corrective feedback. In the remainder of this article, we present a generic psychological mechanism hypothesized to underlie the disinhibited response style. To this end, we begin by discussing a series of experiments with extraverts, focusing on the passive avoidance paradigm to illustrate the mechanism of disinhibition. Following a detailed explication of the process of disinhibition, we discuss further relations among disinhibition, reflectivity, and impulsivity. Then we consider the important influence of Jeffrey Gray's neuropsychological theory of anxiety on our proposals about disinhibition. Next, we state specific hypotheses about

3 718 C. MARK PATTERSON AND JOSEPH P. NEWMAN how the process of disinhibition differs for extraverts and psychopaths and then comment on the proposed mechanism's applicability to the remaining syndromes. Lastly, we discuss alternative explanations for disinhibitory psychopathology. Extraversion, Empirical Findings, and the Evolution of a Mechanism Of all the syndromes, extraversion offers perhaps the easiest access to the mechanism of disinhibition because the behavior of extraverts is the least distorted by negative life events and comorbidities (e.g., substance abuse) that befall the more pathological syndromes. Moreover, their social disinhibition and impulsivity or spontaneity are often esteemed and adaptive (Wilson, 1981). In contrast, psychopaths' often troubled childhoods proclivity for incarceration, superficial relationships, and substance abuse might modify their expression of the diathesis. Most extraverted college students have not experienced the untoward effects of an antisocial life-style. Thus, although their lapses in judgment and failures to learn from experience may be relatively mild, they offer a purer expression of the mechanism. We do not consider extraverts to be at elevated risk for psychopathology. Our view is similar to H. J. Eysenck's (1967) in that a high level of extraversion (or disinhibition proneness, in our proposal) predisposes to psychopathology only when potentiated by an independent risk factor, such as high neuroticism. 1 In addition to these theoretical advantages, there are practical reasons to study extraverts and their nondisinhibited counterparts: introverts. Well-adjusted extraverts comprise a large proportion of the general population, making them readily available for study; they are relatively easy to identify using paper-and-pencil measures (S. B. G. Eysenck & Eysenck, 1975), unlike the other syndromes (see Hare & Cox, 1978); and they may be more likely to volunteer for research than are other disinhibited persons. The research we describe next considers both intraindividual and situational factors that appear to influence extraverts' disinhibited and impulsive behavior. We proceed by addressing specific research questions about several facets of passive avoidance learning. Does Availability of Reward Contribute to Disinhibited Individuals' Passive Avoidance Deficit? Newman, Widom, and Nathan (1985) demonstrated the context-specific nature of extraverts' passive avoidance learning deficit. They examined the combined effects of reward and punishment incentives in a successive go-no-go learning task (see also Moses, Ratliff, & Ratliff, 1979). The structure of the task follows from the septal model: Rats with lesions tend to perform at least as well as controls on avoidance learning tasks without reward contingencies but are usually deficient when avoidance requires inhibition of a dominant approach response set (McCleary, 1966). Using eight two-digit numbers as discriminative stimuli, subjects were required to learn, by trial and error, to respond in the presence of one of the four "go" cues for reward (money) and to inhibit that response when one of the four "no-go" cues for punishment (loss of money) appeared. According to Newman et al. (1985), on the one hand, disinhibited individuals' dominant response set to approach for reward is "resistant to interruption or alteration by cues that are not associated with... current goal-directed behavior" (p. 1318). On the other hand, without a strong dominant response set for reward, punishment should sustain learning and have its usual suppressive effects on behavior. Hence, the authors used a comparison condition involving the same passive avoidance task, but using only punishment incentives, such that subjects lost money if they responded to a no-go cue or if they did not respond to a go cue. Extraverts committed significantly more passive avoidance errors (i.e., responded to more no-go cues) than introverts when rewarded for correct responses and punished for incorrect ones, whereas in the comparison condition they learned to inhibit responding at least as well as introverts when omission of punishment was the only incentive for correct responding. These results exposed an important interaction of situational and individual difference variables. We suggest that the availability of reward served to establish a dominant approach response set in all subjects but that extraverts were less inclined to modulate responding to accommodate the passive avoidance contingency. Several hypotheses can be proposed to understand extraverts' disinhibition in this situation. First, extraverts may be hypersensitive to prospects of reward because such cues gain the lion's share of their attention, rendering them less sensitive to less salient events, such as aversive cues. Second, they may be less sensitive to the motivational effects of punishment and other aversive stimuli, again making them less likely to learn cues that predict aversive experiences and more likely to continue to approach without interruption. Third, they may have a response modulation deficit involving a lesser ability to switch from one motivational set to another in adaptation to fluctuations in the likelihood of approach and avoidance incentives. Each hypothesis connotes a link between a motivational-cognitive bias and their behavioral style. These three hypotheses are not necessarily mutually exclusive, and Newman et al.'s (1985) results do not favor any one as the best account of the perseverative deficit. Yet, the results are consistent with the notion that poor response modulation is situation dependent. Extraverts were proficient at learning from punishment when there was no competing incentive to respond for reward. Thus, this finding counters theories that view extraverts as less conditionable with respect to punishment (Gray, 1971, 1987b) or less conditionable in general (H. J. Eysenck, 1967). What Is Unique About Disinhibited Individuals' Behavior in the Mixed-Incentive Situation That May Interfere With Learning From Experience? Nichols and Newman (1986) studied extraverts' and introverts' behavioral reactions to punishment and reward to explore 1 In the remainder of this article, for sake of brevity, we often substitute "disinhibited" for "disinhibition prone."

4 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 719 differences between them as a function of the situation. They designed a pattern-matching task involving noncontingent response feedback (50% success and 50% failure) given under three different incentive conditions: punishment-only, rewardonly, and a mixed-incentive condition, which combines reward and punishment. Response latencies were measured on trials following feedback to assess subjects' immediate reactions to feedback and, in particular, the extent to which subjects interrupted responding in reaction to punishment. As predicted, extraverts' and introverts' reactions to feedback differed in the mixed-incentive condition. Extraverts responded faster after punishment than after reward. Introverts showed the Feverse pattern of slower response latencies after punishment than after reward. This result suggests that introverts are more likely to pause after punishment, even while seeking reward, and that extraverts tend to form approach response sets that resist interruption or may even be potentiated by punishment. In addition, group differences in pausing after punishment occurred without an overall difference in absolute response speed. However, although no significant group differences in response latency appeared in the mixed-incentive condition, extraverts responded more quickly in the reward-only condition, perhaps indicating that they are more activated than introverts by the prospect of reward, a speculation consistent with Gray's (1981) theory of extraversion-introversion (see also Derryberry, 1987). These results suggest that extraverts are less inclined than introverts to modulate reward-seeking behavior when given aversive feedback and that they are apt to show paradoxical facilitation of responding, rather than response interruption or suppression following punishment. However, because Nichols and Newman (1986) used noncontingent reinforcement to ensure uniform administration of reward and punishment, the impact of individual differences in response modulation on learning could not be assessed. The authors suggested that extraverts' disinhibited reaction to punishment might interfere with aversive conditioning in tasks with response-contingent feedback, as in the passive avoidance learning paradigm. That is, the extent of pausing after aversive feedback might be related to the degree of learning about the effects of one's responses. Does Disinhibited Individuals' Reaction to Punishment Contribute to Their Failure to Learn From Mistakes? Patterson, Kosson, and Newman (1987) reported on two experiments that addressed this question in the passive avoidance paradigm. The first experiment was designed to determine whether extraverts would exhibit the paradoxical response style under conditions involving both response-contingent reward and punishment and whether this reaction would," in turn, interfere with learning inhibition. Experiment 1 used a computerized version of the Newman et al. (1985) successive go-no-go discrimination task to enable recording of response times on trials immediately following feedback to assess the extent to which subjects paused after punishment and reward. Again, extraverts committed more passive avoidance errors in a mixed-incentive situation, corroborating earlier results (Newman et al., 1985). Unlike introverts, who responded more slowly after punishment than after reward, extraverts did not slow down following punishment. Indeed, extraverts responded faster after punishment than after reward, a result very similar to Nichols and Newman's (1986) finding. In addition, examination of the relation between passive avoidance errors and response times following punishment showed that shorter pauses after punishment were associated with poorer passive avoidance learning for all subject groups, whereas response times following reward exhibited a negligible relation to learning. The results of Experiment 1 together with those of Nichols and Newman (1986) formed the basis for hypothesizing about the cognitive processes associated with extraverts' response style. To assess the extent to which extraverts' failures to inhibit punished responses stem from a lack of reflection on cues that predict punishment, Experiment 2 used a similar go-no-go discrimination task with a new measure of reaction to feedback. During the feedback interval before the onset of the next trial, subjects were informed whether or not their response was correct while the cue remained in view to facilitate its association with the feedback. Subjects controlled feedback duration (5-s maximum) by pressing a button to end the feedback display and begin the next trial. This variable time period operationalized the amount of reflection on feedback. By having subjects terminate feedback before the onset of the next trial, feedback duration was not confounded by anticipatory processes associated with the presence of the next cue. In Experiment 1, by contrast, response time after feedback was measured as the latency of response to the next stimulus. The second (comparison) condition of Experiment 2 used a postresponse interval lasting a full 5 s to enable subjects to pause and view feedback for a fixed period. The results of Experiment 2 yielded similar information about extraverts' impulsive response style. As predicted, extraverts committed more passive avoidance errors and reflected less on punishment feedback (in the variable feedback condition), although it is important to point out that their poor passive avoidance and relative lack of reflectivity was due primarily to neurotic extraverts. Again, a general relation between learning and pausing appeared for all groups: More time spent reflecting on punishment feedback predicted fewer passive avoidance errors. Subjects in the fixed-duration-feedback condition committed fewer passive avoidance errors than those in the variable-duration-feedback condition, perhaps owing to the requirement to pause for the full 5 s, but the enforced delay did not eliminate group differences in passive avoidance errors. In summary, the combined results suggest that greater reflection following punishment leads to better learning from punishment for extraverts and introverts alike. However, extraverts, particularly neurotic ones, paused less after punishment, suggesting that when motivated to respond for reward their reaction to punishment interferes with processing punished errors and contributes to their more general propensity for impulsive, nonreflective action. An alternative interpretation relates to possible differences in disinhibited subjects' performance strategy. That is, they might make more errors because their response speed, both anticipatory and after punishment, is faster overall. However, no relation between anticipatory response speeds and errors was found (i.e., no speed-accuracy trade-off). The cognitive processes in-

5 720 C. MARK PATTERSON AND JOSEPH P. NEWMAN volved in retrospective reflection (consideration of past self-environment transactions) may differ from those involved in prospective reflection: consideration of how to act based on learning or, more generally, weighing response alternatives under conditions of uncertainty (see Kagan, 1966). It appears more likely that extraverts' poorer passive avoidance derives from their failure to pause after punishment to reflect on the warning cues that preceded punishment, rather than from a proclivity for rapid responding per se. Indeed, neurotic extraverts' anticipatory response times tended to be slowest (Patterson et al., 1987, Experiment 2; see Newman, Patterson, Howland, & Nichols, 1990, for similar results with low-anxious psychopaths). Extraverts' and introverts' reactions to punishment in the mixed-incentive context have also been examined from a psychophysiological perspective (Howland & Newman, 1985, 1987). Following Fowles's (1980) suggestion that heart rate (HR) acceleration indicates arousal of a behavioral activation system, HR acceleration was measured as one index of reaction to feedback in a passive avoidance paradigm quite similar to Patterson et al.'s (1987) mixed-incentive, fixed-feedback condition. The feedback interval was extended to 8 s to allow for artifact-free recording of psychophysiology. As predicted, punishment did not reduce HR acceleration for extraverts as much as it did for introverts. Specifically, unlike introverts, extraverts' HR acceleration was no less after punishment than after reward. Neurotic extraverts' reaction was particularly distinctive: Their HR stabilized at a higher level than all other subjects' HR during the last half of the 8-s aversive feedback interval. Interestingly, neurotic extraverts also had a large electrodermal response to aversive feedback suggesting that, for them, an increase in electrodermal activity does not demand a corresponding decrease in cardiovascular activity-behavioral activation, as one might expect under these conditions (see Fowles, 1980). As noted earlier, differences between neurotic and stable extraverts were also apparent in the passive avoidance paradigm used by Patterson et al. (1987). Their respective failures to pause and learn following punished errors appears to vary as a function of the situation, as well. Stable extraverts' deficit appeared when punished errors were followed immediately by another opportunity to respond for reward (Experiment 1) but not when the response to feedback was assessed in the absence of immediate cues to respond again (Experiment 2). In contrast, neurotic extraverts' passive avoidance deficit appeared regardless of whether the next trial began immediately or after a brief delay. While responding for reward, neurotic extraverts appear more aroused by punishment and, hence, less inclined to modulate responding. We hypothesize that their more pronounced physiological reaction to punishment (seen as motoric activation) competes with their ability to pause and reflect on information related to their mistakes. This research review provides substance for elaboration of the mechanism of disinhibition. The theory proposed later has three aims. The first is to integrate the aforementioned findings on extraverts, with the expectation that the theory will evolve as research continues. Thus, a second purpose is to create a deductive framework to inspire further experimentation that will be able to accommodate findings on each of the syndromes. Finally, it sets the stage for a more microanalytic view of situationspecific processes underlying disinhibited individuals' often troublesome failures to learn from experience. The Mechanism of Disinhibition In essence, our theory states that disinhibited individuals' impulsive behavioral style stems partly from their active, nonreflective reaction to punishment or frustration. In particular, when engaged in reward-seeking behavior, they have a strong biopsychological reaction to events that block, interrupt, or punish their action. Their typical reaction is one of "forging ahead" rather than "stopping to check it out": Such an immediate, active response can readily become maladaptive to the extent that it leads to further untoward consequences and precludes pausing to learn from experience. We present the mechanism as four conceptually separable, but interdependent and temporally overlapping, stages. Each stage includes intraindividual and situational factors to varying degrees. The relative contribution of each of these stages to the disinhibited and impulsive behaviors of each syndrome will vary. The First Stage Given an opportunity for reward, disinhibited and nondisinhibited individuals alike may adopt a dominant response set to approach, meaning that as long as an appetitive motivational state exists, goal-directed behaviors are most likely to be emitted. The key cognitive aspects of this response set are an effortful allocation of attention to goal-relevant environmental stimuli and an expectation that reward is likely. When the set is very strong, an overfocusing of attention on the goal may occur, akin to premature closure (see Maccoby, 1983; Messer, 1976), which restricts information gathering and consideration of response alternatives, and may engender unrealistic optimism. That is, overfocusing may cause discounting or neglect of cues for punishment or frustration (see Newman et al., 1983; Newman, Patterson, & Kosson, 1987; Siegel, 1978). Approach behavior continues as long as the focus on reward continues, despite changed response contingencies. Physiologically, the approach set manifests as an HR increase, at least partly independent of somatic coupling (see Fowles, 1980, 1988), with concurrent involvement of the central neurophysiological substrate of the behavior activation system. The control of motor activation and output has been attributed largely to the mesolimbic dopamine (DA) system (Iversen, 1977). In addition, it appears that both approach and active avoidance behaviors, related to reward learning and learned relief, respectively, also rely on this DA transmission (Beninger, 1989). The individual differences represented by Stage 1 are how readily approach response sets are formed and how intensely their activation is maintained. Individuals who form approach response sets more readily and maintain them more intensely are more likely to be disinhibited. As such, their approach behavior occurs with more force and fewer micromomentary shifts of attention to assess environmental changes that could

6 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 721 lead to adaptive response modulation (see Gray, 1982, for a discussion of the comparator function). The apparent exaggerated responsiveness to reward in disinhibited humans has several analogs in the septal lesion literature (Gorenstein & Newman, 1980; Gray & McNaughton, 1983; Newman etal., 1983). The Second Stage The dominant response continues until the goal is reached or an obstacle stalls or prevents further action. The key scenario for disinhibition involves an aversive event that disrupts the approach response set leading to two important consequences. First, there is an automatic call to process the unexpected disruption (see Ohman, 1979). Second, because violated expectancies lead to increased arousal (Epstein, 1972), there is for all persons an immediate arousal increment. This arousal underlies an effortful, emotional reaction in light of the mismatch between expectation and reality. The reaction is emotional because it amplifies (see Tomkins, 1984) whatever form the behavioral adaptation takes (in Stage 3) to the aversive event and effortful because of the added central and peripheral exertion needed to fuel the imminent adaptation. 2 The individual difference variable described at this stage is a dimension of reactivity to aversive events, akin to the dimensions of temperament called emotionality (Rowe & Plomin, 1977) or neuroticism (H. J. Eysenck, 1967; see also Wallace, Bachorowski, & Newman, 1991). The strength of a given reaction has both unconditioned and learned determinants such as temperament, learning history with the controllability and predictability of reinforcers, the intensity of the aversive event itself, and the intensity of the ongoing goal-directed action associated with Stage 1. The arousal state at Stage 2 is largely a function of one's reactivity to aversive events. Our view is that the disposition for aversive reactivity is independent of the strength of one's bias to adopt appetitive motivational sets. In this article, we assume that disinhibited persons do not differ uniformly from nondisinhibited persons in their basic reactivity to aversive events. The pronounced disinhibition displayed by neurotic extraverts, however, is suggestive of greater reactivity at Stage 2 (see Wallace et al., 1991). Eysenck's dimension of neuroticism, for instance, is typically regarded as a nonspecific amplifier of behavior. Moreover, neuroticism has been linked to sensitivity to stimuli for reward and punishment as well as to the strength of a person's automatic motor and attentional responses to significant or potentially significant events (see Gray 1981, 1982). In contrast with neurotic introverts, neurotic extraverts' psychophysiological response may resemble the classical defensive response more so than the classical orienting response, meaning that they would be more reactive to the intensity features of novel, unexpected, or aversive events, particularly when they are approaching reward. When combined with extraversion (or a bias toward active responding), reactivity of this nature enhances response output rather than stimulus input (see Graham, 1979) and should therefore contribute to impulsivity. Further research is needed to determine whether neurotic extraverts are unique in this regard among the syndromes of disinhibition. 3 The Third Stage Immediately following the unexpected aversive event and the momentary disruption of the dominant behavior, a behavioral coping response occurs. The arousal increment (Stage 2) enables an effortful, adaptive switch to a passive, informationgathering set. However, disinhibition-prone persons are less likely to perform the switch: Their responding is apt to be disinhibited, such that the dominant response is facilitated rather than suppressed, especially if reward cues remain in the immediate stimulus context. Facilitation of their appetitive responding renders alternative (particularly passive) responses improbable. They continue to seek reward, or they actively seek relief from the aversive event. Active pursuit of reward or relief denotes a bias to establish controllability over aversive events through action rather than predictability of those events through reflection (see Mineka & Kihlstrom, 1978). In contrast, nondisinhibited individuals are apt to direct effort toward inhibition of ongoing behavior and initiation of information processing to adjust their course of action and expectations after accommodating the unanticipated feedback. At the physiological level, Gray (1987a) has noted that serotonin (5-HT) regulation is critical to such activity. Also, specific 5-HT pathways have been isolated that inhibit DA pathways underlying approach behavior and are necessary for behavioral inhibition in the presence of aversive cues (Thiebot, Hamon, & Soubrie, 1984). Nondisinhibited individuals answer the call (Stage 2) to process the feedback, shifting from automatic processing, under the reward expectancy, to controlled processing. An orienting response indicates allocation of central processing resources (Ohman, 1979). Such processing involves conscious rehearsal, (re)evaluation of the situation (Atkinson & Shiffrin, 1968), and language mediation (Martin & Levey, 1987). In a context of recurring aversive events, this mode of processing is necessary to establish predictability and minimize response uncertainty (see Martin & Levey, 1987). Response modulation bias is the individual difference variable described at this stage. That which is modulated is the overt goal-directed behavior and, hence, its underlying response set. Disinhibited persons often fail to alter their response set in accordance with changing environmental events and contingencies: They do not pause, process, and then go on. In contrast, nondisinhibited persons are biased toward passive coping: Their response is reflective. They readily switch their attentional focus and motivational set to accommodate feedback. As we discuss further in the following paragraphs, disinhibition contributes to impulsive behavior in two ways. First, the 2 For the purposes of this article, emotion is considered to be a multidetermined event. Determinants include the individual's temperament, coping style, learning history, and situation. The experience of emotion includes a set of autonomic, perceptual, motor, and central responses. 3 Whether or not disinhibited persons are more reactive may depend largely on how reactivity is measured. If it is measured as interruptibility, for example, they seem to be less reactive to punishment. Alternatively, if one measures their motor behavior after punishment, greater reactivity is suggested. Adding psychophysiological measurements to behavioral observations may help resolve this question.

7 722 C. MARK PATTERSON AND JOSEPH P. NEWMAN disinhibited response is immediate, allowing minimal information gathering or planful cognition, and maladaptive insofar as failure to pause causes further aversive experiences in the new environment. Disinhibition typically represents perseveration of the dominant response set. This behavior, although impulsive, may remain instrumental and, as noted earlier (see Stage 2), is colored by emotion, often anger or a similar affective response to the aversive event. Interestingly, anger and aggression, also a presumed manifestation of disinhibition in many cases, have been linked to activity in the DA system (Hamburg, Hamburg, & Barchas, 1975). The second path to impulsive behavior and cognition involves the associative consequences of Stage 3 discussed later (see Stage 4). The septal lesion literature merits reference here. Dickinson (1975) wrote that "septal damage... appears to produce a shift in the unconditioned reactions to primary aversive stimuli from suppression to activation" (p. 860). Similarly, Dickinson (1974) argued that septal lesions "decrease the suppressive property of aversive stimuli while augmenting their facilitative or drive-enhancing property" (p. 444). These remarks combine the ideas of the first three stages: A poorly modulated reaction to an aversive event facilitates motor responding and is likely to preclude reflection. Notably, septal lesions appear to increase shock-elicited aggression (see Gray & McNaughton, 1983). The Fourth Stage A disinhibited response bias incurs an associative deficit that is fundamental to the impulsive cognitive style. Specifically, to the extent that the approach response is disinhibited after an aversive event, there is a failure to pause and process the cues for punishment or frustration necessary for learned modulation of behavior and sound judgment. That is, disinhibition interferes with retrospective reflection on stimuli and behaviors that warn of aversive outcomes. Retrospective reflection is the process whereby causal associations are formed between behaviors and their consequences, as well as the stimuli that predicted those consequences. It is the cognitive process underlying the development of one's sense of predictability of the environment. Baker and Mercier (1989) argued that variations in the performance and learning of animals and humans are due to variations in the degree of their passive information gathering about relations among stimuli to form a "temporal cognitive map" (p. 93) and their retrospective processing of that information before renewed action. It follows that disinhibition works against such memorybased prospective reflection. Because disinhibited individuals form relatively fewer inhibitory associations involving cues that predict aversive events, their responding is likely to be swayed more by expectations of reward than by the associative products of retrospective reflection in mixed-incentive situations. In other words, relative to nondisinhibited individuals, they are less apt to consider the effects of an action sequence in the presence of legitimate cautionary cues because neither the action sequence nor those cues signify danger to them. This lack of prospective reflection is the proximal source of their "poor judgment" and is the second pathway to impulsivity for disinhibited individuals. Furthermore, disinhibited individuals' response repertoires are apt to be skewed such that dominant behaviors are more often active goal-directed ones, partly because of their lack of consideration of passive alternatives that are based on retrospective reflection, which causes them to appear to lack cognitive breadth. In this way, disinhibition fosters a vulnerability to a self-perpetuating cycle of impulsive active coping. Note that this associative deficit does not mean that disinhibited individuals cannot learn from their mistakes; at least three studies have demonstrated that their performance improves when they are forced to pause and reflect on response feedback (e.g., Arnett, Howland, Smith, & Newman, 1993; Newman et al., 1987; Patterson et al., 1987). Nevertheless, as noted by Patterson et al., if disinhibited individuals remain set to respond for reward following an aversive event, then their information processing is likely to be suboptimal even when they are forced to pause just after the event. Synopsis When an opportunity for reward exists, disinhibited individuals may adopt a strong approach set. If, while that set is dominant, punishment, omission of reward, or delayed gratification occurs unexpectedly, there is an immediate increment in arousal. The arousal boost subserves expenditure of effort for either (a) a behavioral stop and reflection or (b) response facilitation in accordance with the original motivational set or another highly prepared active response (e.g., aggression). Negative affect often colors either outcome. Facilitated responding typifies disinhibited individuals. Disinhibited individuals' lack of retrospective reflection is central to their enduring impulsive style: To the extent that they fail to shift from an active set to a passive, information-gathering set, they do not learn to anticipate when a passive response is the most adaptive way to avoid an aversive event. Although disinhibited persons' bias toward active responding is thought to boost their sense of controllability, it could also result in a relatively impoverished associative memory network for cues and responses that predict aversive events. A reflective bias, in contrast, not only supports elaboration of such a predictive memory network but it also enhances the person's capacity to consider the relative merits of active versus passive response alternatives, which, in turn, renders impulsive behavior all the less likely. Relations Among Disinhibition, Reflectivity, and Impulsivity As discussed earlier, impulsive behavior is a likely by-product of the process of disinhibition and the resulting lack of reflection. Cognition that is nonreflective or lacking in controlled and analytic processing of information is, by definition, impulsive (D. Shapiro, 1965). When disinhibited persons are asked to explain their impulsive behavior, common self-observations are "I do things on the spur of the moment" or "I just felt like it." They are prone to act without benefit of affective associations that might inhibit their behavioral intentions. Their uninhibited action is directly related to a lack of prospective reflection or, in other words, a lack of planful thought and sound judgment. For example, impulsive individuals are more likely to take an im-

8 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 723 mediate and small reward than wait for a larger one (Mischel, 1983). Conversely, they are more likely to avoid an immediate and small punishment and to accept a larger future one (Hare, 1970). These examples epitomize their living-in-the-moment, sometimes negligent life-style. Such uninhibited action suggests a lack of deterrent cognition across a variety of situations. Disinhibited action, a type of uninhibited action, represents response facilitation under aversive conditions. One important implication of the four-stage model is that disinhibition proneness engenders an active coping learning bias that may be more or less adaptive, depending on situational demands. That is, insofar as this response bias fosters controllability over aversive events, it may be adaptive. However, this bias may be maladaptive to the extent that it limits the formation of inhibitory associations and so undermines the predictability of the environment. The poverty of inhibitory associations, in turn, limits prospective reflection and then fosters impulsive behavior. Two Tasks of Aversive Learning Modern learning theory (e.g., Mackintosh, 1983; Rescorla, 1978; see also Mineka & Kihlstrom, 1978) suggests that learning about aversive events involves two key tasks: (a) information processing to identify temporal and spatial contiguities and contingencies among stimuli and reinforcers to enhance predictability and (b) learning motoric instrumental responses to enhance controllability. Each task requires a response set, one primarily attentional (stimulus analysis), the other primarily motoric (response activation). These two sets are distinct but complementary insofar as they require opposing neural (Levy, 1969) and psychophysiological (Graham, 1979) organizations to support dissimilar ways of coping (viz., passive vs. active) with environmental challenges. Some current theories of the relation between information processing and motoric activation favor the view that the two processes can occur in parallel in some situations (e.g., Eriksen & Schultz, 1979) such that inhibitory and excitatory motor systems compete (e.g., Logan & Cowan, 1984). On this view, the overt response, or lack thereof, is thought to represent the outcome of a competition between response systems: inhibition and activation. Stimulus context influences this outcome. Hence, when signaled or actual aversive events are involved, different individuals tend to emphasize one system over the other according to their affective response dispositions (e.g., Hodes, Cook, & Lang, 1985). Information processing coupled with this covert response competition is probably preattentive, although it can access stored stimulus and motor associations to past aversive experiences, on the basis of retrospective reflection, to the extent that they match a current threat. Presumably, because impulsive individuals recruit fewer associations to support inhibitory response alternatives when threatened (D. Shapiro, 1965), and because less competition among alternative responses manifests as a shorter anticipatory reaction time (Grice, Nullmeyer, & Spiker, 1982; cf. Wallace et al., 1991), their dominant responses should be emitted more rapidly. Chronometric analyses indicate impulsive individuals' preference for response activation and execution over response inhibition and stimulus analysis in judgment tasks (Dickman & Meyer, 1988). Impulsive individuals also appear to prefer active over passive, reflective, response strategies in situations involving contemporaneous approach and passive avoidance contingencies (Newman et al., 1985; Patterson et al., 1987). Similarly, Brebner and Cooper (1974) proposed that extraverts are "geared to respond" and introverts are "geared to inspect." Guided by this model, Brebner and Cooper (1978) compared the two groups and found that extraverts engaged in less stimulus analysis before responding, made more responses at a faster rate, and produced more clusters of successively accelerated responses. Also, in a simple reaction time task, extraverts committed more errors on catch trials (Brebner & Flavel, 1978). Importantly, such response activation effects seem to be enhanced in extraverts in appetitive contexts, whereas introverts' stimulus analysis appears enhanced in the presence of aversive cues (Derryberry, 1987; see also Bachorowski & Newman, 1990; Wallace & Newman, 1990). The degree of reflection bears a temporally proximal causal relation to learning. Dispositional variables also influence learning and affect, as research on temperament suggests (e.g., Fulker, 1981; Rowe & Plomin, 1977). In this regard, there is a theoretical tradition that posits two basic mutually inhibitory central motivational control systems, one being behavior-inhibiting and input-enhancing to support strategic (i.e., controlled) information processing, the other being output-enhancing and response-activating to support performance (e.g., Fowles, 1988; Graham, 1979; Gray, 1975; Konorski, 1948; Pribram & Mc- Guiness, 1975; Routtenberg, 1968). Evidence suggests that reliance on one system more than the other underlies basic personality dimensions, such as introversion-extraversion (Tucker & Williamson, 1984). Note also the functional relation between these systems and the two tasks of aversive learning: Pausing to assimilate strategic information is linked to the orienting response (OR) system, whereas motoric learning is supported more directly by the response activation system (Graham, 1979). 4 Although the body of research on the relation of personality to the OR system presents a complex interpretive problem, Stelmack's (1981) review has indicated greater electrodermal activity (EDA) in introverts, including more frequent nonspecific skin conductance responses and higher skin conductance levels, and faster OR habituation for extraverts. 5 Orienting appears to 4 Graham (1979) suggested that the term activation be used to refer to output-enhancing operations rather than the traditional term defensive response. She suggested that the latter is appropriate only at painful levels of stimulation. 5 Evidence suggests that individuals who devote more controlled processing capacity to external stimuli are large electrodermal responders (Dawson & Schell, 1985). Extraverts' apparent emphasis on parallel rather than on controlled information processing coincides with their habituation bias, a tendency to minimize processing of recurring events (Tucker & Williamson, 1984). However, it is important to distinguish between short- and long-term habituation. Fast short-term habituation is associated with less controlled processing, whereas long-term habituation is facilitated by more controlled processing (Ohman, 1979). This view suggests extraverts' habituation bias is limited to short-term effects and agrees with proposals that they and other disinhibited persons are

9 724 C. MARK PATTERSON AND JOSEPH P. NEWMAN indicate the onset of memory updating to accommodate temporal or causal changes in relations among previously associated events (Donchin, 1981; Graham, 1979). This updating counters impulsive behavior by making a reliable predictor or predictors of aversive events (e.g., conditioned stimuli that elicit fear or signal frustration) available for future response modulation by means of concerted operation of the two central control systems. The behavior-inhibiting system reacts to the informativeness of aversive stimuli, whereas the response activation system reacts specifically to increases in the intensity of the stimulation (Graham, 1979). A response keyed to intensity rather than to informativeness leads to HR acceleration in preparation for active coping (see Graham, 1979; Hare, 1978;Obrist, 1981), as we suggest occurs fordisinhibited persons. HR deceleration is associated with orienting, perhaps to dampen somatic feedback and enhance processing of external events (Graham, 1979; Kahneman, 1973). Evidence of cardiovascular differences between extraverts and introverts is inconclusive (Geen, 1983), yet if increased HR indicates appetitive motivation as well as active coping (Fowles, 1988), then Howland and Newman's (1987) finding of sustained HR increases in neurotic extraverts in the mixed-incentive passive avoidance paradigm is consistent with our proposal that disinhibited individuals' approach behavior is facilitated by punishment, which, in turn, impedes reflective processing. 6 Memory and Affective Response Dispositions Emotional information is thought to be coded in memory as propositions organized in associative networks (e.g., Anderson & Bower, 1974; Bower, 1981;Kieras, 1978) for semantic knowledge about stimulus and response relations (e.g., the meaning of a threat and coping options) and procedural knowledge about motor and psychophysiological acts (e.g., facial, visceral, or skeletal). These networks are linked to production systems (e.g., Newell, 1973) containing both information analysis and response generation programs whose activation is the occurrence and expression of emotion (Lang, 1985). Hence, it is reasonable to assume that response biases that are based on either of the two central control systems will foster and will also be enhanced by corresponding associative biases. In our view, disinhibited individuals' response bias involves a memory bias for controllability and activation of response generation programs rather than a bias for predictability and inhibitory programs that promote semantic depth and differentiation by means of reflection. Evidence reviewed by Rodin, Rennert, and Solomon (1980) suggests that motivation for controllability comes from a belief that it ensures positive outcomes. Controllable outcomes necessarily yield a degree of predictability, although predictability does not necessarily guarantee controllability (see Mineka & novelty seekers (Douglas & Peters, 1979; H. J. Eysenck, 1967; Quay, 1965; Zuckerman, 1978). Novelty seeking seems to be another activation-based behavior subserved by the mesolimbic DA system (Iversen, 1977). Evidence that increased EDA is linked to slow short-term habituation, controlled processing, and more efficient associative learning (Ohman, 1979) supports this view. Kihlstrom, 1978). Controllability requires action; predictability does not. The apparent general relation between controllability and behaviors motivated by positive outcomes fits with our view of disinhibition proneness as a particular response bias that is based, in part, on actively seeking reward or relief (see also Gray, 1987b). Following this reasoning further, the disinhibited individual's affective association network is likely to have more hope- and relief-based propositions and expectancies because these affects are related to persistent or increased active responding. In contrast, we hypothesize that fear- and frustration-based propositions and expectancies are more readily formed in those with biases for reflectivity and passive or decreased responding under aversive conditions. Related to this view, Mayo (1983) demonstrated superior retrieval of positive versus negative memories in extraverts. Gray (1981) predicted that extraverts are more sensitive to appetitive stimuli and that introverts are more sensitive to aversive stimuli; evidence indicates that extraverts learn better than introverts with appetitive stimuli, whereas the converse obtains under aversive conditions (Boddy, Carver, & Rowley, 1986; Gupta &Nagpal, 1978; Gupta &Shukla, 1989; Kantorowitz, 1978a, 1978b; Seunath, 1975). It follows that introverts are more cautious than extraverts. For example, introverts use more stringent response criteria than do extraverts in signaldetection paradigms (e.g., Gillespie & Eysenck, 1980; Harkins & Geen, 1975). Moreover, extraverts tend to commit more false alarms on tasks requiring vigilance (M. W. Eysenck, 1981; Newman et al., 1985; Patterson et al., 1987), and vigilance correlates positively with EDA (Gange, Geen, & Harkins, 1979; Krupski, Raskin, & Bakan, 1971). Similarly, Derryberry (1987) showed that introverts appear to allocate more attention to negative cues than do extraverts. Also, because learning punishment cues appears to depend much more heavily on retrospective reflection than learning reward cues, at least in the passive 6 Our studies of the relation between reflectivity and passive avoidance learning have emphasized response latency after punishment and frustration. Ohman (1979) argued that the associative strength between an aversive event and its contextual cues depends on the amount of controlled processing. Because this kind of processing is serial or linear, its amount is perhaps best measured in a time base, like our reflectivity variable (Patterson et al., 1987). However, a host of other variables exist that could broaden the study of reflectivity and response activation, such as P300 latency as a marker of memory revision (Donchin, 1981), EDA and HR (Howland & Newman, 1987), response force and electromyographic activity to assess subthreshold overt response activation (Coles, Gratton, Bashore, Eriksen, & Donchin, 1985), verbal memory for learned events, and subjective verbal report of aifective experience. The psychophysiological findings described here for neurotic extraverts stand in contrast with those for psychopaths (Arnett et al., 1993). In a mixed-incentive passive avoidance paradigm like the one used by Howland and Newman (1987), psychopaths' (especially low-anxious ones) HR was lower following punishment than following reward feedback compared with controls. Similarly, their skin conductance responses (SCRs) following punishment tended to be less frequent, whereas their SCRs after reward were more frequent, than controls'. These results were interpreted to mean that psychopaths allocate less attention to punishment than reward feedback in approach-avoidance conflicts. Also, the results are an example of how two syndromes of disinhibition may differ somewhat in their pathways to disinhibited behavior.

10 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 725 avoidance paradigm (see Patterson et al., 1987), it stands to reason that reflective individuals will be better at learning from aversive events. Given that successful learning from aversive events relies initially on a pattern of orienting, behavioral inhibition, increased EDA, and decreased HR, disinhibited individuals appear handicapped in this regard. However, their response bias is apt to be maladaptive only when there is a demand for passive coping in mixed-incentive situations (e.g., Newman et al., 1985, 1990). Indeed, not all learning requires sustained reflection and behavioral inhibition (see Martin & Levey, 1987). Thus, with their bias for active coping, disinhibited individuals might be better prepared to learn active responses as a means of establishing control over aversive events, as suggested by the adage "when the going gets tough, the tough get going." Consider, for example, an experiment involving a learned helplessness induction with extraverts and introverts (Tiggeman, Winefield, & Brebner, 1982): Extraverts were much more resistant to the effects of uncontrollable aversive outcomes as indexed by measures of response persistence. This resistance again suggests the bias toward enhancing controllability rather than predictability of aversive events. Also, Pearce-McCall and Newman (1986) showed that extraverts' success expectations increased in a betting task involving an opportunity to win money following noncontingent punishment, whereas introverts' expectations declined. Extraverts also showed an undiminished level of perceived control following punishment, in contrast with introverts' decreased sense of control. These observations suggest that extraverts persist in the face of uncertainty, threat, or even actual punishment and fit with our characterization of extraverts as having an active response bias supported by hope- and relief-based expectations. The notion that extraverts possess an optimistic cognitive style (e.g., H. J. Eysenck, 1967) has empirical support (Wilson, 1981). Relation to Gray's Neuropsychological Model of Anxiety Our proposals concerning the process of disinhibition and its relation to reflectivity and impulsivity have been derived primarily from our empirical findings presented earlier. In addition, McCleary's (1966) writing on the response modulation function of the limbic system (see Gorenstein & Newman, 1980) and more recently Gray's comprehensive and detailed analysis of septo-hippocampal system (SHS) functioning (Gray, 1971, 1982, 1987a) have had a substantial impact on our proposals. Gray's theory has been influenced by the work of H. J. Eysenck (e.g., H. J. Eysenck, 1967) on conditioning and introversion-extraversion. Eysenck proposed that introverts condition better than extraverts, which renders the former more conscientious and the latter more sensation seeking. He attributed the difference to extraverts' lower cortical arousal. Gray (e.g., Gray, 1981) modified Eysenck's view by positing a more restricted conditioning superiority for introverts: They are said to be more sensitive to cues for punishment and frustrative nonreward, and hence more conditionable when such cues are operative, whereas extraverts are better at learning about reward and active avoidance because of their greater sensitivity to cues for reward and the omission of punishment. Differences between Eysenck and Gray reflect their divergent assumptions about the physiological basis of these individual differences. Eysenck's theorizing has emphasized the ascending reticular activating system. In contrast, Gray, like Gorenstein and Newman (1980), has focused on the SHS (Gray, 1982,1987a). According to Gray (1982), the SHS may be conceptualized as a hypothetical behavioral inhibition system (BIS). The BIS is activated by punishment cues and novel stimuli. Once activated, the BIS functions (a) to interrupt ongoing behavior that may lead to aversive consequences (though Gray posits that the BIS does not mediate unconditioned reactions to aversive events), (b) to augment arousal that intensifies behavior immediately following an interruption, and (c) to recruit attention for careful analysis of the environment, especially its novel features. Because of the similarities between BIS activity and anxiety, and because anxiolytic drugs dampen BIS activity, Gray has proposed that the SHS represents the neural underpinnings of anxiety. Thus, the BIS governs orienting and its correlates such as heightened EDA and the motoric quiescence that promotes passive avoidance. Gray (e.g., Gray, 1981) also proposed a modification of Eysenck's personality theory so that the basic trait dimensions map directly onto specific physiological structures. Gray suggested that Eysenck's introversion-extraversion and neuroticism dimensions result from interactions between two primary and opposing constructs: impulsivity and anxiety. The degree of trait anxiety corresponds to the strength of the BIS, whereas impulsivity corresponds to the strength of a hypothetical reward system (seegray, Owen, Davis, &Tsaltas, 1983). Fowles(1980) described this reward system as a behavioral activation system (BAS) and proposed that its activity is associated with HR increases and response activation to promote approach behavior or active avoidance. Although Gray's (e.g., Gray, 1982) writings provide a comprehensive model of the SHS and are particularly detailed regarding the relation of SHS functioning to anxiety, the implications of his model for disinhibited behavior are less well developed. Gray has written comparatively little about the other systems that interact with the BIS to influence response modulation. In addition, applications of Gray's model to psychopathology have been restricted to hypotheses regarding the absolute strength or relative dominance of the BIS. Fowles (1980, 1988), for example, has proposed a weak BIS model of psychopathy. Although it shares the major components of Gray's model, the four-stage model represents a slightly different view of SHS functioning. Our view is guided by our empirical findings and our aim to state specific hypotheses about the psychological processes mediating disinhibited behavior. Each stage of the model represents a potential locus for the poor response modulation displayed by disinhibited individuals. The processes described in Stage 1 of our model resemble functions attributed to the BAS. That is, individuals react to reward opportunities by focusing attention on relevant cues and by initiating goal-directed behavior. Poor response modulation cannot occur in the absence of a dominant approach response set (McCleary, 1966). As noted earlier, excessive focus on re-

11 726 C. MARK PATTERSON AND JOSEPH P. NEWMAN ward may reduce the probability that response inhibition and reflectivity will occur when unexpected, negative events are encountered. Stage 2 of our model comprises certain functions of Gray's BIS. Stage 2 begins when an unexpected negative event is encountered. Unexpected negative events are hypothesized to increase nonspecific arousal (see Gray, 1987a), which, in turn, promotes stimulus processing or vigorous motor behavior. These functions are analogous to those of the BIS. However, unlike Gray's model, which limits BIS activity to conditioned aversive stimuli, an individual's unconditioned responses to aversive events play a central role in Stage 2 of our model. In many cases, it is the unconditioned response to aversive events (to stop and reflect or to go more vigorously) that becomes the prototype of subsequent responses in the presence of stimuli that signaled the original aversive event. According to Wallace et al. (1991), bypassing response modulation to facilitate rapid responding in emergent situations may be a vital function of the nonspecific arousal system (NAS). Consequently, sudden increases in NAS activity, by frequently fostering an active response, may contribute to the development of disinhibited as opposed to reflective reactions to motivationally significant situations. Thus, Stage 2 arousal, if too great, also presents a potential locus for the breakdown of response modulation. The behavioral consequences of arousal are described in Stage 3 of our model. Following Gray (1987a), we have proposed that nonspecific arousal increases the intensity of whatever behavior subsequently occurs. That is, general arousal fuels the reaction to overcome the perceived difficulty, whether the person's coping bias is active (i.e., motoric) or passive (i.e., cognitive). Thus, these reactions at Stage 3 are quite similar to the two primary outputs of Gray's BAS and BIS, respectively: goaldirected motor behavior or cognitive processing (allocation of attention to stimuli that result from transactions with the environment). For Gray (1987a), the response selected depends on the relative strengths of BAS and BIS activity, which, in turn, reflects the relative strengths of a person's sensitivities to reward and punishment cues. Although the behavioral outputs at Stage 3 of our model parallel those described by Gray, we regard individual differences at this stage (i.e., differences in response modulation) as due substantially to one's response bias (active vs. passive) and reactivity (Stage 2). Both influence the modulation of one's dominant response. Finally, Stage 4 involves the consequences of response modulation. To the extent that individuals pause and reflect following aversive events, we propose that they are more likely to encode stimuli and responses associated with those events. This retrospective reflection plays a fundamental role in associative learning and facilitates anticipation of aversive outcomes and adaptive behavioral inhibition to avoid those outcomes. Active responses to punishment or frustration may result in weaker internalization of discriminative cues for such events and, thus, limit future passive coping options. Alternatively, individuals may instead come to associate those cues with active coping responses that might sometimes be more effective in attaining one's goals. Gray has not emphasized these associative aspects of learning but has described a process called tagging that is relevant to Stage 4 of our model. For Gray (1982), the motor programme which was in the course of execution at the time of the mismatch... is tagged with an indication which, in English, might read 'faulty, needs checking'. This has two consequences. On future occasions the relevant programme is executed with greater restraint... (and it) is given enhanced attention, (p. 279) This proposal is consistent with our statements about reflectivity and learning. In summary, as might be expected from models that are based on similar neurophysiological systems, our four-stage model has much in common with Gray's formulations. Despite important differences in the role accorded to the NAS (see Wallace et al., 1991) and our attention to both unconditioned and conditioned punishment stimuli, research on the four-stage model is largely compatible with the broader framework provided by Gray's model (see Gray, 1987a). Certainly, we do not regard the four-stage model as an alternative explanation of SHS functioning. Rather, our model represents an attempt to integrate current theorizing about the SHS, including Gray's significant contribution to this realm, with recent research on disinhibited individuals to provide a deductive framework that will advance our understanding of disinhibited behavior. To date, explanations that rely on Gray's model to account for different psychopathologies are based largely on the notion of cross-situational dominance of the BIS over the BAS, or vice versa (e.g., Fowles, 1988; Gray, 1982). The four-stage model suggests alternative explanations for the response modulation problems seen in disinhibitory psychopathology. After reviewing pertinent laboratory findings on the syndromes of disinhibition, we discuss in more detail this issue of alternative mechanisms for disinhibition. Disinhibitory Psychopathology There is ample evidence (e.g., Newman et al., 1985; S. K. Shapiro, Quay, Hogan, & Schwartz, 1988; Siegel, 1978) that disinhibited individuals learn passive avoidance less efficiently in mixed-incentive situations. These findings suggest poorly modulated active coping responses. The four-stage mechanism describes direct links between certain situational events, impulsive behaviors, and the mental processes, most notably poor reflectivity, that underlie those coping behaviors. Can one mechanism be so central to the range of disinhibitory syndromes? Essentially, we propose that the syndromes share a final common pathway to impulsive behavior: situationally specific poor response modulation and a lack of reflective cognition. However, some of the factors that lead to such learning failures may differ among the syndromes. At the same time, most, if not all, of the syndromes may also share a psychobiological diathesis for disinhibition (see Gorenstein & Newman, 1980) that is variously modified by coexisting inheritances and environmental influences through the course of development to form a family of syndromes. In this section, we review briefly theory and evidence related to similarities and differences among the syn-

12 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 727 dromes. Most of the evidence relevant to the four-stage mechanism concerns extraverts and psychopaths. Evidence for the disinhibitory response style in the other syndromes is also discussed. Extroversion Extraverts are hypothesized to possess an exaggerated tendency to approach appetitive stimuli. This hypothesis localizes the source of their disinhibition to Stage 1 of the mechanism: Relative to nondisinhibited individuals, extraverts more readily adopt approach response sets. Because appetitive stimuli are more salient for extraverts, cues for punishment and frustration are apt to be neglected, which then can interfere with motor inhibition. Gray's (e.g., Gray, 1981) hypothesis that extraverts are more sensitive to reward is a more general statement of this view. A variety of evidence supports our hypothesis about extraverts' exaggerated appetitive responding. Nichols and Newman (1986), for example, showed that extraverts respond faster than introverts to reward cues. Similarly, Derryberry (1987) demonstrated that extraverts respond impulsively to reward stimuli. Studies cited earlier demonstrated that extraverts learn better than introverts under appetitive conditions (e.g., Boddy et al., 1986; Gupta &Shukla, 1989;Seunath, 1975), whereas the converse has been demonstrated with introverts under aversive conditions (e.g., Gupta & Nagpal, 1978). Evidence concerning the effects of neuroticism adds some complexity to these basic findings. Neurotic extraverts appear particularly prone to display disinhibited approach responding. Neuroticism amplifies the speed, force, or both of extraverts' approach responding, particularly in mixed-incentive contexts (see Wallace et al., 1991). In a simple motor task, Wallace and Newman (1990) found that neurotic extraverts responded in an inappropriately rapid manner when reward stimuli, but not when punishment stimuli, were present. Neuroticism is not always a liability for extraverts (or introverts) with respect to learning, however. For example, in a verbal conditioning procedure, Gupta (1990) found that reward facilitated learning among neurotic extraverts relative to stable introverts, whereas punishment facilitated learning for neurotic introverts relative to stable extraverts. Wallace et al.'s (1991) view of neuroticism as a manifestation of the reactivity of Gray's (1987a) NAS is consistent with H. J. Eysenck's (1967) view that neuroticism increases the risk of maladaptive expressions of extraverts' inherent response tendencies, such as their pursuit of reward. Thus, according to Wallace et al. (1991), whereas all extraverts are more responsive to reward than introverts (Stage 1), neurotic extraverts are more liable to react maladaptively to an aversive interruption of their goal-directed action because the increased speed, force, and attentional focus associated with neuroticism (see Stage 2) make response modulation less likely. 7 This added nonspecific arousal among neurotic extraverts further diminishes their ability to reflect prospectively or retrospectively on cues for aversive events. Several studies using various tasks have shown that neurotic extraverts are more vulnerable to acting impulsively (e.g., Bachorowski & Newman, 1990; Nichols & Newman, 1986; Patterson etal., 1987; Wallace & Newman, 1990). Psychopathy Clinical and empirical conceptions of psychopathy focus on their impoverished affective processes and impulsive, antisocial life-style (see Cleckley, 1976; Harpur, Hakstian, & Hare, 1988; McCord & McCord, 1964). Psychopaths' notorious recidivism (Hart, Kropp, & Hare, 1988; Robins, 1966) has been linked to their failures to learn from experience and poor judgment. Poor passive avoidance epitomizes this liability (e.g., Blackburn, 1983; Lykken, 1957; Newman et al., 1987; Schmauk, 1970). Collectively, however, research shows that psychopaths do not have a generalized learning deficit (see Brantley & Sutker, 1984). Like extraverts, psychopaths' passive avoidance deficit seems to be limited to mixed-incentive situations such that they perseverate responding for reward despite the presence of information warning against continued approach (Newman & Kosson, 1986; Newman etal., 1987,1990). However, several studies also suggest that somewhat different mechanisms underlie extraverts' and psychopaths' response modulation difficulties despite clear similarities in their behavior. Behavioral evidence. In a set of three experiments, Newman et al. (1990) presented evidence that low-anxious psychopaths' uninhibited responding for reward despite punishment is linked to low reflectivity, which, in turn, probably underlies their poor passive avoidance learning. 8 One of these experiments involved a pattern-matching reaction time task. The task was identical to the one described earlier in relation to extraversion (Nichols & Newman, 1986). In one condition, only noncontingent reward incentives were used: Psychopaths showed no evidence of greater activation (response speed) in the presence of reward than nonpsychopaths. Also, using the same task with mixed incentives, psychopaths showed no less pausing after punishment than did nonpsychopaths. These results suggest that for psychopaths, unlike extraverts, it is not exaggerated responding for reward per se that disrupts response modulation. Like their passive avoidance deficit, psychopaths' inability to delay gratification appears to be limited to conditions that include both approach and avoidance response contingencies. Specifically, Newman, Kosson, and Patterson (1992) found that psychopaths were relatively unwilling to delay approach responding when their response choices sometimes resulted in monetary punishment instead of monetary reward. However, in the same situation without punishment incentives (reward only), psychopaths delayed their responding somewhat (but not significantly) more than nonpsychopaths, providing further ev- 7 Neurotic introverts, too, can behave impulsively in some situations, as when they are engaged in an active motor response under aversive conditions (Wallace & Newman, 1990). 8 In addition to this study, studies by Newman, Kosson, and Patterson (1992); Arnett, Smith, and Newman (1991); Smith, Arnett, and Newman (1992); and other studies cited by Newman et al. (1990) have highlighted the importance of differentiating between high and low trait anxiety when assessing response modulation in psychopaths.

13 728 C. MARK PATTERSON AND JOSEPH P. NEWMAN idence that psychopaths' deficit does not depend simply on reward availability. Although alternative interpretations are possible (see Newman et al., 1992), the delay of gratification data suggest that psychopaths are not insensitive to (or underaroused by) punishment: Their behavior was disinhibited when punishment was involved but not when reward was the only incentive. Newman et al. (1990) also showed that psychopaths are no less sensitive to punishment than are nonpsychopaths and are perhaps more sensitive under some conditions. In their experiment using the pattern-matching task, Newman et al. (1990) found that psychopaths responded faster than controls in the punishment-only condition and, as noted earlier, were no less likely than controls to pause following punishment in the mixed-incentive condition. Finally, psychopaths learn passive avoidance as well as do nonpsychopaths when only punishment contingencies are involved (Newman & Kosson, 1986). Each of these studies suggests that explanations of psychopaths' passive avoidance deficit that appeal to notions of "insensitivity to punishment," a weak BIS, or a too-strong BAS are inadequate. Psychopaths' poor passive avoidance learning with monetary incentives is evident when laboratory tasks use punishment contingencies that do not become salient or apparent until after the dominant approach response is established. Once an approach response is dominant, psychopaths appear less able than nonpsychopaths to suspend that behavior and adjust their attentional resources to process the adverse implications of their responding. This lack of processing causes psychopaths to accrue fewer, if any, inhibitory associations to the approach response and thus renders them vulnerable to maladaptive behavior in the future when similar punishment contingencies exist. 9 Several findings suggest that psychopaths are adept at focusing attentional capacity but are poor at reallocating it once it has been focused (Harpur & Hare, 1989; Jutai & Hare, 1983; Kosson & Newman, 1986, 1989). This attentional rigidity can manifest as perseverated, or poorly modulated, responding. Howland, Kosson, Patterson, and Newman (1993), for example, presented evidence that when presented with inhibitory signals, psychopaths had difficulty suppressing a very simple but wellpracticed dominant response in a task with no demands for learning or memory. Together, the studies discussed earlier suggest that although psychopaths' inhibitory control system can function adequately in some situations, their self-regulatory failures are apt to occur when they must switch attention to accommodate less salient information that is important for guiding ongoing action. Presumably, inhibition promotes this sort of accommodation and learning of alternative adaptive responses. Psychopaths' paradoxical persistence despite frustration or punishment (see also Ross & Doody, 1973) resembles extraverts' resistance to learned helplessness (Tiggeman et al., 1982). Ross (cited in Ross & Doody, 1973) posited that punishment cues may elicit hope in psychopaths and thus preclude anxiety, a proposal that fits with our view of disinhibited persons as biased toward maintenance of controllability through instrumental (often appetitively motivated) action. In conjunction with the lack of evidence regarding hypersensitivity to reward (viz., a Stage 1 deficit), the psychopath's overuse of active coping strategies implies a response modulation deficit localized at Stage 3. Biopsychological evidence. Studies of psychopaths'psychophysiology support the active coping hypothesis, although the predominant view in the literature has been that they are psychophysiologically unresponsive to punishment. Their relatively weak SCRs in anticipation of aversive stimuli are often cited in support of this view (e.g., Fowles, 1980; Hare, 1965). In contrast with their weaker EDA, psychopaths exhibit greater HR increases (Hare, 1978) when threatened by punishment (Siddle, 1977). In accordance with these findings, Fowles (1980) extended Gray's (1977) anxiety theory to argue that psychopaths' weak EDA and low anxiety are due to a weak BIS, which leads to poor passive avoidance, and that their HR reactivity indicates an intact BAS, which, without the anxiety-based restraint of the BIS, leads to unsocialized approach and active avoidance behaviors such as impulsive self-gratification, lying, and aggression. Hare's (1978) view is that HR mechanisms prevail over EDA in the psychopath's "defensive" coping style such that warning cues for punishment are rendered less salient and hence less aversive. Although it is apparent that the patterning of HR and EDA in a given motivational context is crucial to understanding psychopaths' coping style, some recent research does not fully support Hare's (1978) analysis. For example, consistent with Hare's view, Ogloff and Wong (1990) reported that psychopaths showed less EDA and greater HR in anticipation of aversive events. However, psychopaths did not appear to experience those events as less aversive, nor did psychopaths' HR acceleration appear to be causally linked to their lower EDA. Newman and Wallace (1993) have offered an alternative to the standard "insensitivity to punishment" explanation: Psychopaths' lower anticipatory EDA represents less response uncertainty (see Schalling, 1978). That is, they engage in less prospective information processing to cope with the response conflict inherent in a mixed-incentive context (see Waid & Orne, 1982). However, their HR acceleration represents motor readiness for active coping. This pattern of directional fractionation (HR increase, low EDA; see also Steinberg & Schwartz, 1976) can be interpreted as an indicator of their bias for instrumental action to control aversive events instead of inhibition and reflective processing to learn to predict them. We propose that psychopaths are characterized by an anomalous dissociation between the activation (HR-mediated) and inhibition (EDA-mediated) components of their reactions to aversive events such that information that would be accommodated through reflective cognition, particularly in mixed-incen- 9 One frequently cited study, Schmauk (1970), reported findings that psychopaths and nonpsychopaths did not differ in their ability to learn a passive avoidance contingency when monetary incentives were used. However, the several more recent studies cited here contradict Schmauk's findings. Newman et al. (1990) proposed that psychopaths will perform worse than controls when they are required to alter an established approach response set in the context of a mixed-incentive paradigm where the punishment is initially less salient than reward. In Schmauk's study, punishment was salient from the outset and there was no competing reward contingency.

14 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 729 live contexts, does not serve to modulate motor planning as it does in nonpsychopaths. Mawson and Mawson's (1977) theory that psychopaths exhibit poor regulation of the central physiological control systems for behavioral activation and inhibition (i.e., dopaminergic, noradrenergic, or both vs. serotonergic, cholinergic, or both systems) is similar to our view of psychopaths' dissociation between these two systems. Mawson and Mawson's view was that this dysregulation is associated with greater asymmetry in resting levels between the two systems as well as with shifts that are faster in rate and greater in magnitude: Faster shifts to activation cause impulsivity and higher levels of activation cause hyperactivity or aggression. In summary, it now seems clear that psychopaths are sensitive, in the behavioral and psychophysiological senses, to both cued and, in particular, actual aversive experiences. Moreover, psychopaths appear no less able than others to avoid aversive events when that goal is the focus of their attention. Indeed, some studies indicate that allocation of attentional resources to salient events seems to be relatively exaggerated in psychopaths (Harpur & Hare, 1991; Newman & Wallace, 1993). Yet, there is no solid evidence that psychopaths are impaired neuropsychologically despite similarities between their behavior and that of individuals with actual frontal lobe anomalies (e.g., Blumer & Benson, 1975; Hart, Forth, & Hare, 1990; Kandel & Freed, 1989; Schalling, 1978; Smith, Arnett, & Newman, 1992). Nevertheless, psychopaths appear to process environmental stimuli, ones that are initially less salient, incompletely or not at all when their dominant response involves instrumental action. Also, when aversive contingencies that were initially less salient become manifest, psychopaths appear to have difficulty switching their attention to them. Several studies reviewed by Newman and Wallace (1993) support their proposal that psychopaths' difficulty switching attention (i.e., refocusing their controlled information processing) in this way reflects an automatic processing deficit. That is, with a learning history deficient in pausing and forming inhibitory associations, psychopaths are less likely to develop a repertoire of automatic attentional responses to stimuli predictive of aversive events. Hysteria, Early Onset Alcoholism, and Childhood Hyperactivity Gorenstein and Newman (1980) proposed that the syndromes of disinhibition may share a genetic, neurophysiological anomaly expressed phenotypically according to a diathesisstress model. We offer a brief sampler of findings that add to this perspective with respect to the remaining syndromes. 10 Evidence suggests genetic contributions to psychopathy (e.g., Crowe, 1983), alcoholism (e.g., Cloninger, Bohman, & Sigvardsson, 1981), somatization disorder (e.g., Bohman, Cloninger, von Knorring, & Sigvardsson, 1984), childhood hyperactivity (e.g., Tarter & Hegedus, 1983), and extraversion (e.g., Fulker, 1981). Some research also suggests that there is shared etiology among these syndromes (e.g., Bohman et al., 1984; Cloninger et al., 1978; Lilienfeld & Waldman, 1990; Pihl, Peterson, & Finn, 1990; Satterfield, 1978; Tarter, 1978; cf. Cadoret, O'Gorman, Troughton, & Heywood, 1985). Cantwell (1972) demonstrated a familial relation between hyperactivity in childhood and hysteria, psychopathy, and alcoholism in adulthood. Hyperactive children also show a cluster of cognitive and behavioral deficits that suggest a lack of reflectivity. Satterfield (1978) proposed that hyperactive children's lack of central inhibitory control (e.g., low cortical arousal or weak EDA) results in excessive motor activity, impulsivity, and difficulty sustaining attention. There is evidence, too, that hyperactives are vulnerable to poorly modulated shifts between motivational states (Douglas & Peters, 1979), much like Mawson and Mawson's (1977) theory of psychopathy. For example, they appear excessively reactive to the effects of reward and frustration (Douglas, 1983), such that they have trouble delaying responses for reward to avoid its omission (Gordon, 1979). Hyperactives also exhibit a paradoxical response to punishment. That is, in contrast with nonhyperactives, they do not slow down, which would allow more time to process the situation and then respond more cautiously following an aversive experience; instead, they respond impulsively and commit more errors (Sergeant & vandermeere, 1988). Poor inhibitory control is also an important feature of the hyperactive child syndrome. Schachar and Logan (1990), for example, reported that hyperactive children took longer to stop their ongoing behavior when presented with a signal to inhibit. These authors hypothesized that hyperactives' failure to interrupt their responding reflects deficient reallocation of attentional capacity from response processes to inhibitory processes. Whalen, Henker, Collins, McAuliffe, and Vaux (1979) reached a similar conclusion in reporting that hyperactive children are less likely to inhibit and shift their ongoing behavior in accordance with external demands to adopt role-appropriate behaviors. Hyperactive children perform poorly on cognitive tasks requiring vigilance, rule learning, and planning (Douglas, 1983). Douglas also argued that insofar as hyperactives are nonreflective, their experiences will be less well elaborated in memory, thus limiting future learning. O'Dougherty, Nuechterlein, and Drew (1984) found hyperactives to be impulsive on a sustained attention task. Freeman (1978) found them to be deficient on Lykken's (1957) passive avoidance learning task using positive feedback for correct responses and punishment for errors, as we have found for extraverts and psychopaths. Current evidence suggestive of related psychological processes in early onset alcoholics and hysterics is consistent with our integration. Early onset alcoholics have trouble sustaining 10 The names of clinical diagnostic categories included in our family of syndromes are largely obsolete with respect to the current nosology (Diagnostic and Statistical Manual of Mental Disorders, 3rd ed., rev.; American Psychiatric Association, 1987). Psychopathy is most closely related to antisocial personality disorder, although the overlap is not complete (e.g., Hare, 1985). Similarly, early onset alcoholism is not an official category but is a topic of research interest, as are hyperactivity in childhood (attention-deficit hyperactivity disorder) and the disorders that were once subsumed under the rubric of hysteria. Regarding the latter, and for the purposes of this article, somatization disorder is the official category with the firmest empirical connections to antisocial personality, antisocial behavior, and alcoholism (see Lilienfeld & Waldman, 1990).

15 730 C. MARK PATTERSON AND JOSEPH P. NEWMAN and switching attentional sets and are impulsive; they are often hyperactive and conduct disordered as children (see Pihl et al., 1990). Peterson and Pihl (1990) proposed that males at high risk for developing alcoholism are less able to use abstract cognition to foster inhibition in response to novel or threatening situations and are more likely to resort to active coping in those situations. Early onset alcoholics also seem to have a deficient serotonergic system that might predate their alcohol use and underlie their poor impulse control (Buydens-Branchey, Branchey, Noumair, & Lieber, 1989). Existing evidence is strongly suggestive of shared etiology between somatization disorder and antisocial personality disorder (Lilienfeld, 1992; Lilienfeld, VanValkenburg, Larntz, & Akiskal, 1986). To date, somatization disorder has not been studied as extensively as the other syndromes, although existing evidence suggests that so-called hysterical personalities have an impulsive cognitive style (D. Shapiro, 1965; Smokier & Shevrin, 1979). Bendefeldt, Miller, and Ludwig (1976), for example, found hysterics to be both impulsive on a vigilance task and field dependent. This limited review points up various connections among the syndromes. Obviously, more research is needed to test the utility of the four-stage mechanism as the basis of these proposed familial connections and as an explanation for their impulsive, nonreflective actions. Alternative Explanations for Disinhibitory Psychopathology From the perspective of Gray's and Fowles's work, the impulsive behavior of neurotic extraverts and psychopaths is due to a too-strong BAS and a too-weak BIS, respectively. That is, a dominant BAS results in hypersensitivity to reward, whereas a weak BIS causes insensitivity to punishment. Thus, impulsive behavior is due primarily to a trait-like imbalance in these individuals' motivation to approach rather than to inhibit (see Fowles, 1988). To account for psychopaths' lower anxiety, impulsivity, and failure to learn from punishment, Fowles suggested simply that their BIS is pathologically weak. Both Gray (1987a) and Fowles (1988) have interpreted our findings on passive avoidance learning in extraverts and psychopaths to be consistent with their perspective. In this section, we discuss alternative interpretations. The concept of perseveration borrowed from the literature on septal dysfunction (McCleary, 1966) is reflected in the fourstage model and highlights the temporal interaction between reward and punishment incentives. Newman et al. (1985) proposed that once a dominant response set to approach is adopted, alteration of that set in the face of punishment, extinction (nonreward), or contingency reversal is unlikely. Our investigations described earlier help clarify this point. Specifically, the tasks we have used to assess response modulation do not involve a forced trade-off between approach and avoidance contingencies. In the go-no-go passive avoidance task, for instance, it is possible to both maximize winnings and minimize losses through adequate response modulation. To optimize earnings, subjects must alternate between approach (responding to reward cues) and avoidance (withholding responses to punishment cues) sets. This task does not require subjects to favor one strategy over the other, but individuals with impaired response modulation are likely to show a response bias. If, as our theory proposes, difficulty modulating dominant responses explains disinhibited subjects' poor passive avoidance, such performance problems should be minimized by reducing the demand to alter the dominant response set. This has been the rationale behind using reward-only (Newman et al., 1985) and punishment-only (Newman & Rosson, 1986) versions of the go-no-go discrimination task. We argued that the symmetry of reinforcers for approach (i.e., go cues) and response inhibition (i.e., no-go cues) would lead subjects to attend to both aspects of the task from the outset, thereby obviating the need to alter an established response set (see Newman, 1987). Disinhibited individuals can inhibit punished responses as well as their inhibition-prone counterparts in these univalent incentive situations (Newman et al., 1985; Newman &Kosson, 1986). More recently, we redesigned the passive avoidance task to evaluate the importance of subjects' initial response set while including both reward and punishment incentives. The results suggested that disinhibited subjects perform at least as well as controls on tasks that require subjects to learn both approach and avoidance cues from the outset of the task (Newman et al., 1990). In addition, Arnett et al. (1993) reported that disinhibited subjects' passive avoidance learning was at least as good as controls' when the task involved a lengthy (at least 8 s) intertrial interval, presumably because the relatively long delay reduced the need for efficient response modulation. The apparent importance of time constraints suggests that individual differences in passive avoidance learning are not due simply to one motivational system (e.g., BAS or BIS) being inherently stronger (or weaker) than the other. Disinhibited subjects are adept at both avoidance and approach, but they more readily experience difficulty switching from an active to a passive set: Their performance remains intact as long as the requirements for response modulation are minimal. In other words, disinhibited subjects will appear less sensitive to cues that signal the need to revise their responding (e.g., Newman et al., 1987; Siegel, 1978) when tasks encourage them first to adopt a dominant approach set and then switch relatively quickly to an opposing response set. In highlighting four stages of the response modulation process, the model also points up four potential loci for the disinhibited person's difficulty learning from aversive events. Stage 1 focuses on the initial response set and attentional aspects of set formation. To the extent that a person's attention is focused too quickly or narrowly on cues related to their proximal goals, the probability that their goal-directed behavior will be influenced by other cues, such as cues for punishment or more distal goals, is reduced. This assumption resembles Gray's (1981) description of extraverts as differentially sensitive to reward as opposed to punishment cues. However, Wallace et al. (1991) have proposed that it is H. J. Eysenck's neuroticism dimension that underlies the deficient response modulation displayed by neurotic extraverts. According to these authors, it is the emotional reactivity associated with neuroticism that is responsible for the speed, force, and attentional focus of their goal-directed behavior. That is, individuals may initiate action before evaluating their response (see D. Shapiro, 1965), they may respond so

16 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 731 forcefully that interruption is difficult, or their attention may be so focused on an immediate goal that more temporally remote or perceptually less salient cues are not used (see Easterbrook, 1959). In a similar vein, several investigators have attempted to link the impulsive behavior of psychopaths with attentional overfocusing (e.g., Harpur & Hare, 1991; Jutai & Hare, 1983; Kosson & Newman, 1986; Newman et al., 1983). Stage 2 concerns the strength of a person's reaction to punishment and frustrative nonreward. A problem that may arise at this stage is that increases in nonspecific arousal could be insufficient to initiate response modulation (i.e., to interrupt an established response set). For example, it is often proposed that both psychopaths and extraverts do not experience sufficient fear arousal to sustain conditioning or to initiate avoidance learning (e.g., Chesno & Kilmann, 1975; H. J. Eysenck, 1967; Gray, 1971; Hare, 1970; Lykken, 1957). At the same time, it is possible that very large or rapid increases in arousal at Stage 2 may disrupt response modulation for other individuals. Hyperreactivity at Stage 2 may underlie the phenomenon of "anxious impulsivity" described by Wallace et al. (1991). In other words, just as emotional reactivity may confer an element of urgency and contribute to premature response initiation at Stage 1, excessive reactivity at Stage 2 could also bias an individual to act immediately, precluding response interruption. In this regard, Gray (e.g., Gray, 1982) has emphasized the necessity of distinguishing between conditioned and unconditioned punishment stimuli partly because the former tend to elicit response inhibition, whereas the latter tend to engender response facilitation. However, given that the psychophysiological impact of unconditioned punishment is likely to be both more rapid and intense than that of a conditioned stimulus, it may be that these features of unconditioned aversive stimuli necessarily increase the likelihood of response facilitation occurring. More broadly speaking, events of the first two stages may interact to decrease response modulation: The urgency to respond added at Stage 2 (because of increased NAS activity) compounds the preexisting readiness to enact the dominant response set. That is, the more urgent the need to respond and the more available a particular response, the less likely a person will be to pause to contemplate alternative responses. Conversely, response inhibition would be more likely under conditions involving less intense cues for punishment and response uncertainty. Stage 3 of our model is yet another potential locus for individual differences in response modulation. As noted earlier, one consequence of septal lesions involves a shift in the animal's reaction to punishment and nonreward from response inhibition to response facilitation. Similarly, cues for punishment appear to engender a unique, paradoxical facilitation of responding in psychopaths. In one study involving a computerized version of the Wisconsin Card Sorting Task (WCST) and monetary incentives for correct and incorrect responses, Howland and Newman (1993) reported that psychopaths and controls displayed opposite responses just after the first sorting rule was changed. The first rule in the WCST requires sorting cards by color. After 10 consecutive correct responses, the correct rule is changed (without informing subjects) to sorting cards by shape. Whereas the response latency of nonpsychopaths increased on the first occasion that a color-sorting response was punished during the shape-rule period, psychopaths responded more quickly. Very similar between-group differences in response latency were obtained by Arnett, Smith, and Newman (1991) with low-anxious psychopaths when, after a 1-min phase in which subjects responded for rewards only, a red light came on, indicating that a second phase in which money could be lost as well as won had begun. Finally, Newman (1991) reported that after training subjects to avoid punishment by inhibiting responses to the letter Q during a 150-trial pretreatment, psychopaths responded more quickly and controls more slowly when this letter was superimposed on a visual search, reaction time task. Even though the Q had no relevance to responding in the second part of the task, this cue for punishment elicited opposite reactions from psychopaths and controls. Stage 4 involves the information-processing consequences of retrospective reflection and behavioral disinhibition. We have proposed that greater retrospective reflection is associated with better learning from aversive events, and we provided preliminary evidence to support this proposal. However, we suspect that a person's tendency to learn from aversive events is influenced by other features as well as the degree to which prior experiences have been encoded through retrospective reflection. Analysis of immediate and delayed recall in introverts and extraverts, for example, suggests that introverts benefit from having a period of time to consolidate learning, whereas extraverts typically display better recall when tested immediately (M. W. Eysenck, 1981). Generally, individual differences in speed and depth of information processing may exert an impact on retrospective reflection that is independent of the influence of one's response bias. In summary, it is possible to explain many of the aforementioned findings in terms of Gray's "three arousal model" (see Fowles, 1980) by hypothesizing an overly strong BAS or tooweak BIS. However, these hypotheses do not fully account for important aspects of the observed phenomena, such as situation-dependent disinhibition, the importance of temporal factors, and the dissociation between arousal and inhibition. The four-stage model of disinhibition endeavors to be more specific in highlighting particular processes that mediate impulsive responding. Nevertheless, despite important differences in the role accorded to the NAS (see Wallace et al., 1991), our inclusion of unconditioned with conditioned stimuli as potential triggers of Stage 2, and our emphasis on disinhibition and reflectivity (as opposed to the relative sensitivities to reward and punishment cues), the results of studies that are based on the four-stage model may be integrated into the broader framework put forth by Gray and Fowles. Each stage of the model suggests a particular locus for disinhibition proneness. As such, the model permits more specific hypothesis generation regarding the different syndromes' disinhibited behavior, as we have done for both extraverts and psychopaths. Conclusions An impressive array of findings, primarily from lesion and anxiolytic drug studies, concerning the neurophysiology of the SHS and its relation to anxiety and behavioral inhibition has been coalesced and interpreted by Gray. The scope of his theory

17 732 C. MARK PATTERSON AND JOSEPH P. NEWMAN is quite broad (see Gray, 1982, 1987a). In contrast, our theory focuses on a more limited set of hypothetical constructs (e.g., response modulation, reflection, disinhibition, and impulsivity) joined by a single mechanism concerning disinhibition and how certain individuals fail to learn from (or at least learn differently about) aversive events. Moreover, Gray has devoted most of his eifort to the study of anxiety, whereas our thrust has been the study of disinhibited and impulsive behavior. As such, the two theories are more complementary than competing. In general, our theory focuses on problems in the modulation of a response bias whether it is active and disinhibited or passive and reflective, namely the self-perpetuating maladaptive consequences that occur without adequate modulation of either bias. One of our goals is to suggest a research framework within which important differences among the syndromes can be discerned more clearly. We proposed such a difference between extraverts and psychopaths and provided some supportive evidence. Although we have emphasized psychobiological processes in determining these differences, we recognize that environmental factors play a significant role. Indeed, within a diathesis-stress framework, there is a matrix of familial and nonfamilial environmental and social-learning influences that can modify an inherent neuropsychological bias. Although we suggest that somewhat different aberrations in the response modulation process may exist for the different syndromes, and that environmental factors contribute to the diverse expressions of the predisposition, we propose that these various differences ultimately converge on the final common pathway of behavioral disinhibition, diminished reflectivity, and poor learning from aversive events. A second goal is to stimulate further research by presenting several other domains of inquiry that are relevant to the fourstage mechanism. For example, the dynamics of prediction and control represent a complex research domain unto itself (see Mineka & Hendersen, 1985) that we believe can offer an important perspective on personality and psychopathology. A central theme in our arguments is that reflection promotes anticipatory coping, whereas disinhibition is reactionary. A passive, reflective coping bias tends to foster predictability, planning, and conscience; an active coping bias may support one's sense of controllability. Each style has its potential liabilities, however. Excessive disinhibition can lead to unrestrained efforts to gratify impulses without respect to social norms. Too much reflection can cause unwarranted anticipation of negative events and inaction. We have focused on the liabilities of disinhibition. There are also liabilities associated with a reflectivity bias. Together with a passive coping style, this bias may lead to distractibility or interruption in reaction to uncertainty or threat, thus fostering anxiety or perhaps helplessness depression. Gray's (1982) theory predicts that introverts, individuals with a dominant BIS, are more vulnerable to anxiety disorders. Introverts also appear more vulnerable to unipolar depressive disorders (Akiskal, Hirschfield, & Yerevanian, 1983). As the study of disinhibition becomes more grounded in research on neurophysiological substrates of behavior, individual differences in cerebral hemispheric specialization is another domain that may be applied profitably to understanding the processes of learning from and coping with aversive events. For'example, Hugdahl (1989) presented a methodology to study interactions between hemispheric specializations and asymmetries of conditioned responses. Although it is beyond the scope of this article, we believe that our proposals could be enriched further by reference to studies of the asymmetrical distribution of neurotransmitter systems as they relate to personality and behavior (Patterson & Newman, 1987; see Tucker & Williamson, 1984, for a review). We have presented in this article a primarily psychological theory of what is essentially a psychobiological diathesis. The biology and psychology of the process of disinhibition are ultimately inseparable. The other central constructs of the theory reflectivity, impulsivity, and response modulation are also psychobiological in nature. Thus, although biology and psychology can be studied separately, research must take both aspects into account to provide an integrative nomological network around the process of disinhibition. We are hopeful that our elaboration of the psychological processes involved in the breakdown of self-regulation will facilitate the ease with which advances in neuropsychology (e.g., Gray, 1987a, 1987b) and neurochemistry (e.g., Soubrie, 1986) can enrich our understanding of disinhibited behavior. Furthermore, such a nondualistic approach to the study of disinhibitory psychopathology should foster the development of more powerful clinical interventions for the various syndromes than reliance on either biological or psychological considerations alone would allow. References Akiskal, H. S., Hirschfield, R. M. A., & Yerevanian, B. I. (1983). The relationship between personality and affective disorders: A critical review. Archives of General Psychiatry, 40, American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. Anderson, J. R., & Bower, G. H. (1974). A propositional theory of recognition memory. Memory & Cognition, 2, Arnett, P. A., Howland, E. W., Smith, S. S., & Newman, J. P. (1993). Autonomic responsivity during passive avoidance in incarcerated psychopaths. Personality and Individual Differences, 14, Arnett, P. A., Smith, S. S., & Newman, J. P. (1991, October). Approach and avoidance motivation in incarcerated psychopaths. Paper presented at the thirty-first annual meeting of the Society for Psychophysiological Research, Chicago, IL. Aronfreed, J. (1969). Aversive control of socialization. In W. J. Arnold (Ed.), Nebraska Symposium on Motivation (pp ). Lincoln: University of Nebraska Press. Atkinson, R. C., &Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In K. W. Spence & J. T. Spence (Eds.), The psychology of learning and motivation: Advances in research and theory (Vol. 2, pp ). San Diego, CA: Academic Press. Bachorowski, J., & Newman, J. P. (1990). Impulsive motor behavior: The effects of personality and goal salience. Journal of Personality and Social Psychology, 58, Baker, A. G., & Mercier, P. (1989). Attention, retrospective processing and cognitive representations. In S. B. Klein & R. R. Mowrer (Eds.), Contemporary learning theories: Pavlovian conditioning and the status of traditional learning theory (pp ). Hillsdale, NJ: Erlbaum. Bendefeldt, F., Miller, L. L., & Ludwig, A. M. (1976). Cognitive performance of hysterics. Archives of General Psychiatry, 33, Beninger, R. J. (1989). The role of serotonin and dopamine in learning to avoid aversive stimuli. In T. Archer & L-G. Nilsson (Eds.), Aversion,

18 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 733 avoidance, and anxiety: Perspectives on aversively motivated behavior (pp ). Hillsdale, NJ: Erlbaum. Blackburn, R. (1983). Psychopathy, delinquency, and crime. In A. Gale & J. A. Edwards (Eds.), Physiological correlates of human behavior: Individual differences andpsychcpathology (Vol. 3, pp ). San Diego, CA: Academic Press. Blatt, R. C. (1976). Facilitation and nonfacilitation of active avoidance behavior of rats with septal lesions in the shuttlebox and running wheel. Journal of Comparative Physiological Psychology, 90, Blumer, D., & Benson, D. F. (1975). Personality changes with frontal and temporal lobe lesions. In D. F. Benson & D. Blumer (Eds.), Psychiatric aspects of neurologic disease. New York: Grune & Stratton. Boddy, J., Carver, A., & Rowley, K. (1986). Effects of positive and negative verbal reinforcement on performance as a function of extraversion-introversion: Some tests of Gray's theory. Personality and Individual Differences, 7, Bohman, M., Cloninger, R. C., von Knorring, A-L., & Sigvardsson, S. (1984). An adoption study of somatoform disorders: III. Cross-fostering analysis and genetic relationship to alcoholism and criminality. Archives of General Psychiatry, 41, Bower, G. H. (1981). Mood and memory. American Psychologist, 36, Brantley, P. J., & Sutker, P. B. (1984). Antisocial behavior disorders. In H. E. Adams & P. B. Sutker (Eds.), Comprehensive textbook of psychopathology(pp ). New York: Plenum Press. Brebner, J., & Cooper, C. (1974). The effect of low rate of regular signals upon the reaction times of introverts and extraverts. Journal of Research in Personality, 8, Brebner, J., & Cooper, C. (1978). Stimulus- or response-induced excitation: A comparison of the behavior of introverts and extraverts. Journal of Research in Personality, 12, Brebner, J., & Flavel, R. (1978). The effect of catch trials on speed and accuracy among introverts and extraverts in a simple RT task. British Journal of Psychology, 69, Buydens-Branchey, L., Branchey, M., Noumair, D., & Lieber, C. S. (1989). Age of alcoholism onset: II. Relationship to susceptibility to serotonin precursor availability. Archives of General Psychiatry, 46, Cadoret, R. J., O'Gorman, T. W., Troughton, E., & Heywood, E. (1985). Alcoholism and antisocial personality: Interrelationships, genetic and environmental factors. Archives of General Psychiatry, 42, Cantwell, D. (1972). Psychiatric illness in the families of hyperactive children. Archives of General Psychiatry, 27, Chesno, F. A., & Kilmann, P. R. (1975). Effects of stimulation intensity on sociopathic avoidance learning. Journal of Abnormal Psychology, 84, Cleckley, H. C. (1976). The mask of sanity (5th ed.). St. Louis, MO: Mosby. Cloninger, C. R., Bohman, M., & Sigvardsson, S. (1981). Inheritance of alcohol abuse: Cross-fostering analysis of adopted men. Archives of General Psychiatry, 38, Cloninger, C. R., Reich, T, & Guze, S. B. (1978). Genetic-environmental interactions and antisocial behaviour. In R. D. Hare & D. Schalling (Eds.), Psychopathic behaviour: Approaches to research (pp ). New York: Wiley. Coles, M. G. H, Gratton, G., Bashore, T., Eriksen, C. W., & Donchin, E. (1985). A psychophysiological investigation of the continuous flow model of human information processing. Journal of Experimental Psychology: Human Perception and Performance, 11, Crowe, R. R. (1983). Antisocial personality disorder. In R. E. Tarter (Ed.), The child at psychiatric risk (pp ). New York: Oxford University Press. Dawson, M. E., & Schell, A. M. (1985). Information processing and human autonomic classical conditioning. In P. K. Ackles, J. R. Jennings, & M. G. H. Coles (Eds.), Advances in psychophysiology (Vol. 1, pp ). Greenwich, CT: JAI Press. Derryberry, D. (1987). Incentive and feedback effects on target detection: A chronometric analysis of Gray's model of temperament. Personality and Individual Differences, 8, Dickinson, A. J. (1974). Response suppression and facilitation by aversive stimuli following septal lesions in rats: A review and model. Physiological Psychology, 2, Dickinson, A. J. (1975). Suppressive and enhancing effects of footshock on food-reinforced responding following septal lesions in rats. Journal of Comparative and Physiological Psychology, 88, Dickman, S. J., & Meyer, D. E. (1988). Impulsivity and speed-accuracy tradeoffs in information processing. Journal ofpersonality and Social Psychology, 54, Donchin, E. (1981). Surprise!... Surprise? Psychophysiology, 18, Douglas, V. I. (1983). Attentional and cognitive problems. In M. Rutter (Ed.), Developmental neuropsychiatry (pp ). New York: Guilford Press. Douglas, V. I., & Peters, K. G. (1979). Toward a clearer understanding of the attentional deficit of hyperactive children. In G. A. Hale & M. Lewis (Eds.), Attention and the development of cognitive skills (pp ). New York: Plenum Press. Easterbrook, J. A. (1959). The effect of emotion on cue utilization and the organization of behavior. Psychological Review, 66, Ellen, P., Wilson, A. S., & Powell, E. W. (1964). Septal inhibition and timing behavior in the rat. Journal of Comparative Neurology, 10, Elliott, F. (1978). Neurological aspects of antisocial behavior. In W. H. Reid (Ed.), The psychopath: A comprehensive study of antisocial disorders and behaviors (pp ). New York: Brunner/Mazel. Epstein, S. (1972). The nature of anxiety with emphasis upon its relationship to expectancy. In C. M. Spielberger (Ed.), Anxiety: Current trends in theory and research (pp ). San Diego, CA: Academic Press. Eriksen, C. W., & Schultz, D. W. (1979). Information processing in visual search: A continuous flow conception and experimental results. Perception and Psychophysics, 25, Eysenck, H. J. (1967). The biological basis of personality. Springfield, IL: Thomas. Eysenck, H. J. (Ed.). (1981). A model for personality. New York: Springer. Eysenck, M. W. (1981). Learning, memory and personality. In H. J. Eysenck (Ed.), A model for personality (pp ). New York: Springer. Eysenck, S. B. G., &Eysenck, H. J. (1975). Manual of the Eysenck Personality Questionnaire. San Diego, CA: ITS Publishers. Fowles, D. C. (1980). The three-arousal model: Implications of Gray's two-factor learning theory for heart rate, electrodermal activity, and psychopathy. Psychophysiology, 17, Fowles, D. C. (1988). Psychophysiology and psychopathology: A motivational approach. Psychophysiology, 25, Freeman, R. J. (1978). The effects of methylphenidate on avoidance learning and risk taking by hyperkinetic children. Unpublished doctoral dissertation, University of Waterloo, Ontario. Fulker, D. W. (1981). The genetic and environmental architecture of psychoticism, extraversion and neuroticism. In H. J. Eysenck (Ed.), A model for personality (pp ). Newark: Springer. Gange, J. J., Geen, R. G., & Harkins, S. G. (1979). Autonomic differences between extraverts and introverts during vigilance. Psychophysiology, 16, Geen, R. G. (1983). The psychophysiology of extraversion-introversion.

19 734 C. MARK PATTERSON AND JOSEPH P. NEWMAN In J. T. Cacioppo & R. E. Petty (Eds.), Social psychophysiology: A sourcebook (pp ). New York: Guilford Press. Gillespie, C. R., & Eysenck, M. W. (1980). Effects of introversion-extraversion on continuous recognition memory. Bulletin ofthepsychonomic Society, 15, Gordon, M. (1979). The assessment of impulsivity and mediating behaviors in hyperactive and nonhyperactive boys. Journal of Abnormal Child Psychology, 7, Gorenstein, E. E., & Newman, J. P. (1980). Disinhibitory psychopathology: A new perspective and a model for research. Psychological Review, 87, Graham, F. K. (1979). Distinguishing among orienting, defense, and startle reflexes. In H. D. Kimmel, E. H. van Oist, & J. F. Orlebke (Eds.), The orienting reflex in humans (pp ). Hillsdale, NJ: Erlbaum. Gray, J. A. (1971). The psychology of fear and stress. New York: McGraw-Hill. Gray, J. A. (1975). Elements of a two-process theory of learning. San Diego, CA: Academic Press. Gray, J. A. (1977). Drug effects on fear and frustration: Possible limbic site of action of minor tranquilizers. In L. L. Iversen, S. D. Iversen, & S. H. Snyder(Eds.), Handbook ofpsychopharmacology: Vol. 8. Drugs, neurotransmitters, and behavior (pp ). New \brk: Plenum Press. Gray, J. A. (1981). A critique of Eysenck's theory of personality. In H. J. Eysenck (Ed.), A model for personality (pp ). New \fork: Springer. Gray, J. A. (1982). The neuropsychology of anxiety. New York: Oxford University Press. Gray, J. A. (1987a). Perspectives on anxiety and impulsivity: A commentary. Journal of Research in Personality, 21, Gray, J. A. (1987b). The psychology of fear and stress. New York: McGraw-Hill. Gray, J. A., & McNaughton, N. (1983). Comparison between the behavioral effects of septal and hippocampal lesions: A review. Neuroscience and Biobehavioral Reviews, 7, Gray, J. A., Owen, S., Davis, N., & Tsaltas, E. (1983). Psychological and physiological relations between anxiety and impulsivity. In M. Zuckerman (Ed.), Biological bases of sensation seeking, impulsivity, and anxiety (pp ). Hillsdale, NJ: Erlbaum. Grice, G. R., Nullmeyer, R., & Spiker, V. A. (1982). Human reaction times: Toward a general theory. Journal of Experimental Psychology: General, 111, Gupta, B. S. (1990). Personality and reinforcement in verbal operant conditioning: A test of Gray's theory. Psychological Studies, 35, Gupta, B. S., & Nagpal, M. (1978). Impulsivity/sociability and reinforcement in verbal operant conditioning. British Journal of Psychology, 69, Gupta, B. S., & Shukla, A. P. (1989). Verbal operant conditioning as a function of extraversion and reinforcement. British Journal of Psychology, 80, Hamburg, D. A., Hamburg, B. A., & Barchas, J. D. (1975). Anger and depression in perspective of behavioral biology. In L. Levi (Ed.), Emotions: Their parameters and measurement (pp ). New 'Ybrk: Raven Press. Hare, R. D. (1965). Temporal gradient of fear arousal in psychopaths. Journal of Abnormal Psychology, 70, 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 & D. Schalling (Eds.), Psychopathic behavior: Approaches to research (pp ). New "York: Wiley. Hare, R. D. (1985). Comparison of procedures for the assessment of psychopathy. Journal of Consulting and Clinical Psychology, 53, Hare, R. D., & Cox, D. N. (1978). Clinical and empirical conceptions of psychopathy and the selection of subjects for research. In R. D. Hare & D. Schalling (Eds.), Psychopathic behavior: Approaches to research (pp. 1-22). New York: Wiley. Harkins, S., & Geen, R. G. (1975). Discriminability and criterion differences between extraverts and introverts during vigilance. Journal of Research in Personality, 9, Harpur, T. J., Hakstian, A. R., & Hare, R. D. (1988). Factor structure of the Psychopathy Checklist. Journal of Consulting and Clinical Psychology, 56,14\-l 41. Harpur, T. J., & Hare, R. D. (1989, June). Facilitation and inhibition of visual attention in psychopaths. Paper presented at the 4th meeting of the International Society for the Study of Individual Differences, Heidelberg, West Germany. Harpur, T. J., & Hare, R. D. (1991). Psychopathy and attention. In J. Enns (Ed.), The development of attention: Research and theory (pp ). Amsterdam: North-Holland. Hart, S. D., Forth, A. H., & Hare, R. D. (1990). Performance of criminal psychopaths on selected neuropsychological tests. Journal of Abnormal Psychology, 99, Hart, S. D., Kropp, P. R., & Hare, R. D. (1988). Performance of male psychopaths following conditional release from prison. Journal of Consulting and Clinical Psychology, 56, Hodes, R. L., Cook, E. W, & Lang, P. J. (1985). Individual differences in autonomic response: Conditioned association or conditioned fear? Psychophysiology, 22, Howland, E. W., Kosson, D. S., Patterson, C. M., & Newman, J. P. (1993). Altering a dominant response: Performance of psychopaths and low socialization college students on a cued reaction time task. Journal of Abnormal Psychology, 102, Howland, E. W., & Newman, J. P. (1985). Heart rate response following punishment feedback among extraverted and low-socialized college students. Psychophysiology. 22, 581. Howland, E. W, & Newman, J. P. (1987). Reaction to contingent punishment in extraverts: A psychophysiological analysis. Unpublished manuscript. Howland, E. W., & Newman, J. P. (1993). The effect of incentives on Wisconsin Card Sorting Task performance in psychopaths. Unpublished manuscript. Hugdahl, K. (1989). Human Pavlovian aversive conditioning: Effects of brain asymmetry and stimulus lateralization. In T. Archer & L-G. Nilsson (Eds.), Aversion, avoidance, and anxiety: Perspectives on aversively motivated behavior (pp ). Hillsdale, NJ: Erlbaum. Iversen, S. D. (1977). Brain dopamine systems and behavior. In L. L. Iversen, S. D. Iversen, & S. H. Snyder (Eds.), Handbook of psychopharmacology: Drugs, neurotransmitters, and behavior (Vol. 8, pp ). New York: Plenum Press. Jutai, J. W, & Hare, R, D. (1983). Psychopathy and selective attention during performance of a complex perceptual-motor task. Psychophysiology, 20, Kagan, J. (1966). Reflection-impulsivity: The generality of dynamics of conceptual tempo. Journal of Abnormal Psychology, 7, Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice Hall. Kandel, E., & Freed, D. (1989). Frontal lobe dysfunction and antisocial behavior: A review. Journal of Clinical Psychology, 45, Kantorowitz, D. A. (1978a). An experimental investigation of pre-orgasmic reconditioning and post-orgasmic deconditioning. Journal of Applied Behavioral Analysis, 11, Kantorowitz, D. A. (1978b). Personality and conditioning of tumes-

20 REFLECTIVITY IN SYNDROMES OF DISINHIBITION 735 cence and detumescence. Behavior Research and Therapy, 16, Kieras, D. (1978). Beyond pictures and words: Alternate information processing models for imagery effects in verbal memory. Psychological Bulletin, 5, Konorski, J. (1948). Conditioned reflexes and neuron organization. Cambridge, England: Cambridge University Press. Kosson, D. S., & Newman, J. P. (1986). Psychopathy and the allocation of attentional capacity in a divided attention situation. Journal of Abnormal Psychology, 95, Kosson, D. S., & Newman, J. P. (1989). Socialization and attentional deficits under focusing and divided attention conditions. Journal of Personality ami Social Psychology, 57, Krupski, A., Raskin, D. C, & Bakan, P. (1971). Psychophysiological and personality correlates of commission errors in an auditory vigilance task. Psychophysiology, 8, Lang, P. J. (1985). The cognitive psychophysiology of emotion: Fear and anxiety. In A. H. Tuma & J. D. Maser (Eds.), Anxiety and the anxiety disorders (pp ). Hillsdale, NJ: Erlbaum. Levy, J. (1969). Possible basis for the evolution of lateral specialization of the human brain. Nature, 224, Lilienfeld, S. O. (1992). The association between antisocial personality and somatization disorders: A review and integration of theoretical models. Clinical Psychology Review, 12, Lilienfeld, S. Q, VanValkenburg, C., Larntz, K., & Akiskal, H. S. (1986). The relationship of histrionic personality disorder to antisocial personality and somatization disorders. American Journal of Psychiatry, 143, Lilienfeld, S. O., & Waldman, I. D. (1990). The relation between childhood attention-deficit hyperactivity disorder and adult antisocial behavior reexamined: The problem of heterogeneity. Clinical Psychology Review, 70, Logan, G. D., & Cowan, W. B. (1984). On the ability to inhibit thought and action: A theory of an act of control. Psychological Review, 91, Lykken, D. T. (1957). A study of anxiety in the sociopathic personality. Journal of Abnormal and Social Psychology, 55, Maccoby, E. E. (1983). Social-emotional development and response to stressors. In N. Garmezy & M. Rutter (Eds.), Stress, coping, and development in children (pp ). New York: McGraw-Hill. Mackintosh, N. J. (1983). Conditioning and associative learning. New York: Oxford University Press. Martin, I., & Levey, A. B. (1987). Learning what will happen next: Conditioning, evaluation, and cognitive processes. In G. Davey (Ed.), Cognitive processes and Pavlovian conditioning in humans (pp ). New York: Wiley. Mawson, A. R., & Mawson, C. D. (1977). Psychopathy and arousal: A new interpretation of the psychophysiological literature. Biological Psychiatry, 12, Mayo, P. R. (1983). Personality traits and the retrieval of positive and negative memories. Personality and Individual Differences, 5, McCleary, R. A. (1966). Response modulating function of the limbic system: Initiation and suppression. In E. Stellar & J. M. Sprague (Eds.), Progress in physiological psychology (Vol. 1, pp ). San Diego, CA: Academic Press. McCord, W., & McCord, J. (1964). The psychopath: An essay on the criminal mind. Stanford, CA: Stanford University Press. Messer, S. (1976). Reflection-impulsivity: A review. Psychological Bulletin, 83, Mineka, S., & Hendersen, R. W. (1985). Controllability and predictability in acquired motivation. Annual Review of Psychology, 36, Mineka, S., & Kihlstrom, J. F. (1978). Unpredictable and uncontrollable events: A new perspective on experimental neurosis. Journal of Abnormal Psychology, 87, Mischel, W. (1983). Delay of gratification as process and as person variable in development. In D. Magnusson & V. L. Allen (Eds.), Human development: An interactional perspective (pp ). San Diego, CA: Academic Press. Moses, J. A., Jr., Ratliff, R. G., & Ratliff, A. R. (1979). Discrimination learning of delinquent boys as a function of reinforcement contingency and delinquent subtype. Journal of Abnormal Child Psychology, 7, Newell, A. (197 3). Production systems: Models of control structures. In W. G. Chase (Ed.), Visual information processing (pp ). San Diego, CA: Academic Press. Newman, J. P. (1987). Reaction to punishment in extraverts and psychopaths: Implications for the impulsive behavior of disinhibited individuals. Journal of Research in Personality, 21, Newman, J. P. (1991, October). Response modulation deficits in psychopaths. Paper presented at the meeting of the American Society of Criminology, San Francisco, CA. Newman, J. P., Gorenstein, E. E., & Kelsey, J. E. (1983). Failure to delay gratification following septal lesions in rats: Implications for an animal model of disinhibitory psychopathology. Personality and Individual Differences, 4, Newman, J. P., & Kosson, D. S. (1986). Passive avoidance learning in psychopathic and nonpsychopathic offenders. Journal of Abnormal Psychology, 95, Newman, J. P., Kosson, D. S., & Patterson, C. M. (1992). Delay of gratification in psychopathic and nonpsychopathic offenders. Journal of Abnormal Psychology, 101, Newman, J. P., Patterson, C. M., Rowland, E. W., & Nichols, S. L. (1990). Passive avoidance in psychopaths: The effects of reward. Personality and Individual Differences, 11, Newman, J. P., Patterson, C. M., & Kosson, D. S. (1987). Response perseveration in psychopaths. Journal of Abnormal Psychology, 96, Newman, J. P., & Wallace, J. (1993). Psychopathy and cognition. In K. S. Dobson & P. C. Kendall (Eds.), Psychopathology and cognition (pp ). San Diego, CA: Academic Press. Newman, J. P., Widom, C. S., & Nathan, S. (1985). Passive avoidance in syndromes of disinhibition: Psychopathy and extraversion. Journal of Personality and Social Psychology, 48, Nichols, S. L., & Newman, J. P. (1986). Effects of punishment on response latency in extraverts. Journal of Personality and Social Psychology, 50, Obrist, P. A. (1981). Cardiovascular psychophysiology: A perspective. New \brk: Plenum Press. O'Dougherty, M., Nuechterlein, K. H., & Drew, B. (1984). Hyperactive and hypoxic children: Signal detection, sustained attention, and behavior. Journal of Abnormal Psychology, 93, Ogloff, J. P. R., & Wong, S. (1990). Electrodermal and cardiovascular evidence of a coping response in psychopaths. Criminal Justice and Behavior, 17, Ohman, A. (1979). The orienting response, attention, and learning: An information-processing perspective. In H. D. Kimmel, E. H. van Oist, & J. F. Orlebke (Eds.), The orienting reflex in humans (pp ). Hillsdale, NJ: Erlbaum. Patterson, C. M., Kosson, D. S., & Newman, J. P. (1987). Reaction to punishment, reflectivity, and passive avoidance learning in extraverts. Journal of Personality and Social Psychology, 52, Patterson, C. M., & Newman, J. P. (1987). A neuropsychological perspective on response modulation in syndromes of disinhibition. Unpublished manuscript. Pearce-McCall, D. P., & Newman, J. P. (1986). Expectation of success

21 736 C. MARK PATTERSON AND JOSEPH P. NEWMAN following noncontingent punishment in introverts and extraverts. Journal of Personality and Social Psychology, 50, Peterson, J., & Pihl, R. O. (1990). Information processing, neuropsychological function, and the inherited predisposition to alcoholism. Neuropsychology Review, 1, Pihl, R. Q, Peterson, J., & Finn, P. (1990). Inherited predisposition to alcoholism: Characteristics of sons of male alcoholics. Journal of Abnormal Psychology, 99, Pribram, K. H., & McGuiness, D. (1975). Arousal, activation, and effort in the control of attention. Psychological Review, 82, Quay, H. C. (1965). Psychopathic personality as pathological stimulation seeking. American Journal of Psychiatry, 122, Rescorla, R. A. (1978). Some implications of a cognitive perspective on Pavlovian conditioning. In S. H. Hulje, H. Fowler, & W. K. Honig (Eds.), Cognitive processes in animal behavior (pp ). Hillsdale, NJ: Erlbaum. Robins, L. N. (1966). Deviant children grown up. Baltimore: Williams & Wilkins. Rodin, J., Rennert, K., & Solomon, S. K. (1980). Intrinsic motivation for control: Fact or fiction? In A. Baum & J. E. Singer (Eds.), Advances in environmental psychology: Applications of personal control (Vol. 2). Hillsdale, NJ: Erlbaum. Rosenthal, R. H., & Allen, T. W. (1978). An examination of attention, arousal, and learning dysfunctions of hyperkinetic children. Psychological Bulletin, 85, Ross, R. R., & Doody, K. F. (1973). Persistence in the psychopathic personality. Canadian Journal of Criminology and Corrections, 15, Routtenberg, A. (1968). The two-arousal hypothesis: Reticular formation and limbic system. Psychological Review, 75, Rowe, D., & Plomin, R. (1977). Temperament in early childhood. Journal of Personality Assessment, 41, Satterfield, J. H. (1978). The hyperactive child syndrome: A precursor to adult psychopathy? In R. D. Hare & D. Schalling (Eds.), Psychopathic behavior: Approaches to research (pp ). New York: Wiley. Schachar, R., & Logan, G. (1990). Impulsivity and inhibitory control in normal development and childhood psychopathology. Developmental Psychology, 26, Schalling, D. (1978). Psychopathy-related personality variables and the psychophysiology of socialization. In R. D. Hare & D. Schalling (Eds.), Psychopathic behavior: Approaches to research (pp ). New York: Wiley. Schmauk, F. J. (1970). Punishment, arousal, and avoidance learning in sociopaths. Journal of Abnormal Psychology, 76, Sergeant, J. A., & vandermeere, J. (1988). What happens after a hyperactive child commits an error? Psychiatry Research, 24, Seunath, O. M. (1975). Personality, reinforcement and learning. Perceptual and Motor Skills, 41, Shapiro, D. (1965). Neurotic styles. New York: Basic Books. Shapiro, S. K., Quay, H. C., Hogan, A. E., & Schwartz, K. P. (1988). Response perseveration and delayed responding in undersocialized aggressive conduct disorder. Journal of Abnormal Psychology, 97, Siddle. D. A. T. (1977). Electrodermal activity and psychopathy. In S. A. Mednick & K. O. Christiansen (Eds.), Biosocial bases of 'criminal behavior (pp ). New York: Gardner Press. Siegel, R. A. (1978). Probability of punishment and suppression of behavior in psychopathic and nonpsychopathic offenders. Journal of Abnormal Psychology, 87, Smokier, I. A., & Shevrin, H. (1979). Cerebral lateralization and personality style. Archives of General Psychiatry, 36, Smith, S. S., Arnett, P. A., & Newman, J. P. (1992). Neuropsychological differentiation of psychopathic and nonpsychopathic criminal offenders. Personality and Individual Differences, 13, Soubrie, P. (1986). Reconciling the role of central serotonin neurons in human and animal behavior. Behavioral and Brain Sciences, 9, Steinberg, E. P., & Schwartz, G. E. (1976). Biofeedback and electrodermal self-regulation in psychopathy. Journal of Abnormal Psychology, 85, Stelmack, R. M. (1981). The psychophysiology of extraversion and neuroticism. In H. J. Eysenck (Ed.), A model for personality (pp ). New York: Springer. Tarter, R. E. (1978). Etiology of alcoholism: Interdisciplinary integration. In P. E. Nathan, G. A. Marlatt, & T. Loberg (Eds.), Alcoholism: New directions in behavioral research and treatment (pp ). New York: Plenum Press. Tarter, R. E. (Ed.). (1983). The child at psychiatric risk. New York: Oxford University Press. Tarter, R. E., & Hegedus, A. M. (1983). Developmental disorders: Etiology and outcome. In R. E. Tarter (Ed.), The child at psychiatric risk (pp ). New York: Oxford University Press. Thiebot, M-H., Hamon, M.. & Soubrie, P. (1984). Serotonergic neurones and anxiety-related behavior in rats. In M. R. Trimble & E. Zarafian (Eds.), Psychopharmacology of the limbic system (pp ). New York: Oxford University Press. Tiggeman, M., Winefield, A. H., & Brebner, J. (1982). The role of extraversion in the development of learned helplessness. Personality and Individual Differences, 3, Tomkins, S. S. (1984). Affect theory. In K. R. Scherer & P. Ekman (Eds.), Approaches to emotion (pp ). Hillsdale, NJ: Erlbaum. Trasler, G. (1978). Relations between psychopathy and persistent criminality: Methodological and theoretical issues. In R. D. Hare & D. Schalling (Eds.), Psychopathic behavior: Approaches to research (pp ). New York: Wiley. Tucker, D. M., & Williamson, P. A. (1984). Asymmetric neural control systems in human self-regulation. Psychological Review, 91, Waid, W. M., & Orne, M. T. (1982). Reduced electrodermal response to conflict, failure to inhibit dominant behaviors, and delinquency proneness. Journal of Personality and Social Psychology, 43, Wallace, J., Bachorowski, J., & Newman, J. P. (1991). Failures of response modulation: Impulsive behavior in anxious and impulsive individuals. Journal of Research in Personality, 25, Wallace, J., & Newman, J. P. (1990). Differential effects of reward and punishment cues on response speed in anxious and impulsive individuals. Personality and Individual Differences, 11, Whalen, C. K., Henker, B., Collins, B. E., McAuliffe, S., & Vaux, A. (1979). Peer interaction in a structured communication task: Comparisons of normal and hyperactive boys and of methylphenidate (Ritalin) and placebo effects. Child Development, 50, Wilson, G. D. (1981). Personality and social behaviour. In H. J. Eysenck (Ed.), A model for personality (pp ). New York: Springer. Zuckerman, M. (1978). Sensation seeking and psychopathy. In R. D. Hare & D. Schalling (Eds.), Psychopathic behavior: Approaches to research (pp ). New York: Wiley. Received April 2, 1990 Revision received February 1, 1993 Accepted March 18, 1993