Deviant sexual thoughts: mental control and the treatment of sexual offenders.

The clinical literature has long illustrated the paradoxical findings that deliberate attempts to suppress particular thoughts actually increase their occurrence. These unwanted, often intrusive, thoughts that are a major feature of obsessive disorders, depression, sleep disorders, and a range of other disturbances are of particular clinical concern. The exploration of psychological factors associated with cognitive control is, then, clinically relevant. In the current article we consider the role of mental control, especially thought suppression, in explaining the occurrence of unwanted thoughts, specifically in relation to deviant sexual thoughts. Many features of sexual offending, such as the effects of stress or strong affective states on offending and the rapid escalation in severity and frequency of sexual offending, reported by both therapists and researchers, can be explained by the mental control literature. In addition, the role of suppression in therapy for sexual offenders and its implications for relapse are considered. We argue that the use of suppression techniques by therapists is not sufficient to prevent the occurrence of sexually deviant thoughts and the recurrence of sexual offenses. Therapists also must teach offenders to manage stress effectively and to develop appropriate beliefs about what is controllable. Making suppression techniques automatic and avoiding high-risk situations for offending are also important skills for the offender to learn in therapy.

The cognitive and emotional demands of modern life mean that it has become increasingly important to learn how to manage effectively our mental processes and behavior. In fact, the pursuit of happiness, knowledge, and a successful career depend to a large extent on the achievement of mental control (Wegner & Erber, 1993). Controlling one's thoughts and their expression is a vital component of everyday life. You may discuss some thoughts with your partner, but it would be unwise to discuss them with your boss; sexist thoughts about women may be shared with rugby club mates but not with work colleagues. We hesitate to share some thoughts with anyone, and we may try to banish them from our minds. Mental control has recently become the focus of a flurry of theoretical and empirical research and is moving beyond its folk psychological status (Wegner, 1994). A number of different psychological phenomena associated either directly or indirectly with mental control processes, for example, are attention, self-regulation, and defence mechanisms (Wegner & Erber, 1993).

As far back as Freud (1915/1957) it was proposed that people attempt to exclude negative material from awareness to preserve a sense of well-being. Indeed, Erdelyi and Goldberg (1978) found that 99% of nonclinical participants reported having tried on occasion to suppress disturbing thoughts to avoid the emotions associated with those thoughts. Freud proposed that this suppression of unwanted thoughts could occur both through unconscious processes and deliberate effort. More recently, Salkovskis and Campbell (1994) emphasized the need to consider differences between instructed suppression (thoughts supplied by the experimenter) and self-motivated suppression (thoughts supplied by the participant). Focus in the clinical literature has been on self-motivated suppression, especially failure to suppress successfully unwanted intrusive thoughts and its links to clinical disorders such as obsessive-compulsive disorder (Clark, Ball, & Pape, 1991; Salkovskis, 1989) and depression (Rachman & Hodgson, 1980; Wenzlaff, Wegner, & Roper, 1988).

The emphasis in this article is on thought suppression and its implications for relapse, and therapy in sexual offenders. In addition to the impact of self-imposed thought suppression by sexual offenders, the usefulness of instructed suppression in therapy, particularly cognitive behavioral therapy, is considered. Clinicians using recent cognitive-behavioral treatment approaches in sexual offending, such as relapse prevention, attempt to enhance offenders' self-management skills to maintain the initial behavior change induced by therapists (Laws, 1989). Effective mental control is a vital prerequisite to the development of such skills; self-management requires mental control.

The conceptual underpinning of current self-management, mental control, or thought-suppression techniques in therapy for sexual offenders is unclear and suffers from a lack of theoretical depth (Ward, Hudson, & Siegert, 1995). Simply instructing people to suppress unwanted thoughts is not sufficient to avoid those thoughts and behaviors (see Wegner, 1994, for a review). We argue that therapists need a more thorough understanding of the complexities of thought suppression before it can become an effective treatment tool. We suggest that an integration of the mental control literature with current views of sexual offender treatment can provide that theoretical depth and, hence, guide effective treatment programs. An understanding of the mechanisms underlying suppression may help therapists avoid certain pitfalls in the treatment of sex offenders. It also clarifies why certain internal strategies are effective and why some critical psychological processes in relapse, for example, the abstinence violation effect, occur following the return of unwanted thoughts or feelings.

The suppression of inappropriate sexual thoughts is likely to be of particular importance in understanding and minimising relapse. It is reasonable to assume that the loss of cognitive or behavioral control is partially responsible for recidivism in those men committed to maintaining changes made in treatment programs (Laws, 1989; Pithers, 1990). Inappropriate sexual behaviors are typically preceded by sexual thoughts or fantasies (Pithers, 1990). We argue that incorporating the teaching of suppression techniques into therapy will both reduce the incidence of inappropriate sexual thoughts and the likelihood of those thoughts leading to inappropriate sexual behaviors. The early Christians adopted some extreme measures in an attempt to prevent their unwanted sexual thoughts from becoming behavior, from taking frequent cold showers to jumping into beds of stinging nettles. We argue that thought suppression can be both an easier and a more effective method of preventing inappropriate sexual thoughts and fantasies from becoming action.

In this article we describe a recent influential model of mental control (Wegner, 1994) and consider some major research findings on thought suppression and their implications for sexual offending. At issue is whether teaching offenders to suppress inappropriate sexual desires or fantasies is a useful tool in the reduction of recidivism. We argue that effective suppression of unwanted sexual thoughts and fantasies may reduce the likelihood that those thoughts will result in deviant sexual behaviors. Finally, future directions for research and therapy are considered.

Thought Suppression

A Model of Mental Control

People have been shown to have the ability to suppress a whole variety of thoughts, at least in the short term, across many different domains. People can suppress sexist comments (Wegner, Erber, & Bowman, 1993), thoughts about white bears (Wegner, Schneider, Carter, & White, 1987), taped stories (Clark et al., 1991; Clark, Winton, & Thynn, 1993), affect-laden stories (Muris, Merckelbach, van den Hout, & de Jong, 1992; Wenzlaff et al., 1988), stereotypes (Macrae, Bodenhausen, Milne, & Jetten, 1994), test success or failure (Conway, Howell, & Giannopoulos, 1991), pain (Cioffi & Holloway, 1993), emotional expression (Gross & Leverson, 1993), and, importantly, personally relevant self-generated intrusive thoughts, both positive and negative (Howell & Conway, 1992; Kelly & Kahn, 1994; Parkinson & Rachman, 1981; Salkovskis & Campbell, 1994; Sutherland, Newman, & Rachman, 1982). People given suppression instructions report fewer unwanted thoughts than those given no specific suppression instructions. Importantly, research has shown some sexual offenders to be able to inhibit their penile erections to sex-related stimuli in the laboratory (Freund, Watson, & Rienzo, 1988; Hall, 1989). The exact nature of that inhibition and the mechanism by which arousal was suppressed is not clear, however.

Wegner's (1994) model of mental control proposed a two-process system to explain how individuals first identify unwanted thoughts (the monitoring process) and then replace those unwanted thoughts with more acceptable ones (the operating process). For mental control to be possible, individuals must have a desired mental state, which can be self-imposed or imposed by a therapist. During the monitoring process, thoughts inconsistent with the desired mental state are continually sought. This search for unwanted thoughts can take two forms: It can either be a generalized search for all thoughts inconsistent with a desired mood state (a feature-negative search) or a more focused search for specific unwanted thoughts (a feature-positive search). The distinction between these two types of searches is an important one for thought suppression, as feature-positive, specific searches are far easier and less effortful, than generalized feature-negative searches (Newman, Wolff, & Hearst, 1980), although for the identification of unwanted thoughts, both types of search require few cognitive resources. To illustrate this distinction, consider two desired goals: a state of happiness and a state of not having unwanted sexual thoughts. In the case of a state of happiness, all thoughts inconsistent with such a state (e.g., sad or angry thoughts) would be sought (a feature-negative or generalized search). In the case of wanting to avoid having some specific sexual thoughts, only those specific unwanted thoughts need be identified (a feature-positive search).

To achieve mental control successfully, each time unwanted thoughts are identified they must be replaced by more desired or appropriate, goal-consistent thoughts (i.e., with distracters)--Wegner's operating process. Again, this replacement of unwanted thoughts can take two forms. When a person desires to achieve a certain mental state (e.g., happiness), the unwanted, goal-inconsistent thoughts can be replaced by items consistent with the goal (a feature-positive process). When mental control involves the desire to avoid a particular mental state (e.g., having inappropriate sexual thoughts), the unwanted thoughts can be replaced by any distracter thoughts that are not the unwanted thoughts (a generalized, feature-negative process).

The idea behind this model of mental control, therefore, is that a goal is established and sought. Thoughts inconsistent with that state are identified and replaced by goal-consistent ones. Wegner proposed that the identification and replacement of unwanted thoughts need not be a conscious process. If the identification and replacement of unwanted thoughts proceed without conscious awareness, individuals are not aware of the presence of unwanted thoughts. If, however, the replacement of the unwanted thoughts occurs only after they have reached conscious awareness, then individuals are aware of the presence of the unwanted thoughts and their subsequent attempts to replace them. Importantly, as noted earlier, the identification and replacement of the unwanted thoughts can take two forms as a function of the specificity of the searches. If specific thoughts (either seeking specific unwanted thoughts or specific replacement thoughts) are being sought, then the search is a feature-positive one. On the other hand, if the search is a generalized one for any relevant thoughts (either any thoughts inconsistent with the goal or any appropriate replacement thoughts), the search is a feature-negative one.

The ability to suppress unwanted thoughts successfully is a function of the cognitive resources available. The replacement of unwanted thoughts by goal-consistent thoughts is a demanding process, and when cognitive resources are scarce (e.g., under high cognitive load, emotional or physical stress, time pressure, or high arousal), this replacement process is disrupted. When the process of replacement fails, the most active or accessible thoughts are usually the unwanted thoughts, inadvertently activated through their identification as being unwanted. This results in lapses in suppression and the intrusion of the unwanted thoughts (e.g., Wegner et al., 1993). The processes inherent in the exercise of mental control may actually undermine achievement of that control (Macrae et al., 1994; Wegner, 1994). Indeed, the achievement of mental control may be impeded by attempts to do so, the ironic consequences of attempted thought suppression.

Some types of mental control are simply easier to achieve than others. Distinctions between focused searches for specific thoughts (feature-positive searches) and generalized searches (feature-negative searches) were drawn previously. Specific, feature-positive searches are cognitively easier than generalized, feature-negative ones (Newman et al., 1980). In trying to achieve a goal, the identification of unwanted thoughts (the monitoring process) and their replacement by goal-consistent thoughts (the operating process) work in a complementary fashion. If the identification process involves a specific, feature-positive search, then the replacement process will involve a generalized, feature-negative search, and vice versa. Because the replacement of unwanted thoughts is the resource-demanding aspect of mental control, more effortful than the identification of unwanted thoughts, the ideal situation is for that process to involve cognitively less demanding, specific searches. Mental control processes that involve the identification of specific unwanted thoughts (e.g., sexually deviant thoughts) are especially difficult because the search for replacement, distracter thoughts is a generalized, feature-negative search (looking for all non-deviant sexual thoughts or non-sexual thoughts). For this reason, we argue that therapists will achieve more success if they have sexual offenders set a goal of sexual adjustment rather than a goal of the suppression of unwanted sexual thoughts in their attempts at mental control.

The Ironic Effects of Suppression

Ironically, failures in suppression may actually result in a worse situation than when no attempts at suppression were made, that is, hyperaccessibility of the undesired thoughts. The unwanted thoughts become more accessible than if no attempts had been made to suppress those thoughts, known as an ironic or rebound effect (Macrae et al., 1994; Wegner, 1994). This hyperaccessibility can result from a lack of cognitive resources to suppress the unwanted thoughts (e.g., Wegner et al., 1993) or from a removal of suppression instructions or motivations (Macrae et al., 1994). If there are insufficient cognitive resources available for the replacement of unwanted thoughts by goal-consistent ones, then the unwanted thoughts become accessible as a result of their previous activation through identification as unwanted thoughts. These unwanted thoughts become more accessible than if no attempts at suppression (and hence no activation through identification) have been made. Ironic effects can also occur in situations of low cognitive load. Successful suppression of thoughts in one task (in response to instructions) leads to a greater use of those thoughts in a second, even allegedly unrelated task. Macrae et al. (1994) showed that this rebound effect could also extend to subsequent behavior. Participants instructed to suppress their stereotypic thoughts in describing a skinhead target in Task 1 showed greater avoidance of (sat further away from) another skinhead target in Task 2 than did participants who had no suppression instructions. Hyperaccessibility (or "rebound") effects can be explained by considering the activation levels of relevant thoughts. By virtue of identification as unwanted thoughts, the activity level of those thoughts is increased. When the active suppression is stopped, therefore, unwanted thoughts have high levels of activity and so are likely to be used in subsequent judgment situations. The rate of decay of the activated unwanted thoughts is as yet unknown, but some effects of priming can last up to 16 months after the initial priming episode (e.g., Jacoby, 1983; Smith, Stewart, & Buttram, 1992).

Successful suppression of sexually deviant thoughts in one situation, then, could lead to increased likelihood of sexually deviant behavior if the motivation for suppression is decreased or there is a change in situational cues. Motivation for suppression could be reduced by, for example, feelings of sexual arousal, anger, or depression. The situational cues for suppression could change as the result of, for example, the arrival of a potential victim or the removal of external constraints (e.g., another adult).

Key Features of Thought Suppression and Their Implications for the Treatment of Sexual Offenders

Suppression techniques may already be incorporated into cognitive behavioral therapy (CBT) for sexual offenders. The self-management focus of CBT directly involves the self-monitoring and suppression of sexually deviant thoughts and feelings Withers, 1990). The emphasis in self-regulation is on individuals actively controlling their actions through the operation of regulatory mechanisms and feedback from the environment. Many components of CBT require individuals to suppress maladaptive thoughts or feelings and replace them with more adaptive ones. The challenging of offenders' cognitive distortions requires that they learn to observe their thoughts and counter them in some way. When an offender notices the re-emergence of distorted attitudes about women or children, this thought is suppressed and replaced with a more appropriate alternative. We argue that the role of such suppression techniques needs to be formalized and emphasized. An understanding of the processes underlying suppression and rebound also allows for therapy to go beyond the initial mental control, or suppression, of unwanted thoughts. In this section we consider key features of thought suppression and their implications for therapists introducing thought-suppression techniques to sexual offenders.

Adopting Thought-Suppression Techniques

The first step in the use of self-regulation interventions, such as relapse prevention for sexual offenders, is for offenders to see their sexual thoughts and fantasies as inappropriate or unwanted. Cognitive distortions are very common and take a number of forms, for example, inaccurate interpretations of a victim's behavior or elaborate rationalizations that absolve the self of responsibility for the sexual offense. The first step in therapy must be to get clients to acknowledge their thoughts as distorted. The challenging of cognitive distortions is therefore a priority in the treatment of any sexual offenders, and most comprehensive treatment programs include modules explicitly dealing with this issue (Marshall, in press). Victim empathy training, for example (Marshall, in press), requires that offenders learn to appreciate what their victims experienced and how it has adversely affected them. Such an approach demonstrates to offenders the inappropriateness of their sexual thoughts, fantasies, or desires. This may be a difficult task, given the pleasure and sexual arousal often derived from such sexual thoughts (Ward, Hudson, & Marshall, 1994).

A loss of mental control (or more accurately, the disengagement of self-regulatory processes) has also been hypothesized to be a desirable state for some sex offenders and one way of avoiding self-censure. Offenders have been described as suspending self-regulation during the offense chain or as using a processing style of cognitive deconstruction (Baumeister, 1991; Bliss & Larson, 1985; Ward, Hudson, & Marshall, 1995), that is, of deliberately relaxing mental control. Both cognitive deconstruction. (Baumeister, 1991) and suspension of self-regulation are typified by offenders failing to consider long-term consequences of their behavior, focusing on immediate gratification and, as a consequence, not experiencing any incongruity between their behavior and their self-image. The deconstructed state, or suspension of self-regulation, is seen as an attempt to reduce the negative implications of self-awareness, which would lead to the behavior being stopped (during the offense) or to feelings of guilt and compunction (after the offense). Ward et al. (1995) argued that cognitive deconstruction provides an explanatory framework for the precursors as well as the consequences of offenses. Such a state both predisposes offending and perpetuates such behavior. The cognitive deconstructed state is similar to the situation of the breakdown of mental control described previously, increasing the likelihood of people violating their usual ethical and personal values (i.e., offending). That offenders already seek the state of suppression breakdown will increase therapists' difficulties in installing inhibition as an active control strategy in offenders. Prior to entering therapy, sexual offenders often suppress adaptive (e.g., that sex with your child is wrong) or self-depreciating (e.g., self as weak) thoughts to justify or rationalize their offending. In therapy, offenders are directed toward suppressing inappropriate sexual thoughts or fantasies.

The Inevitability of Failure: Motivation Is Not Enough

Having the motivation to suppress unwanted sexual thoughts may not be sufficient for self-regulation to succeed. Even when offenders accept their thoughts as inappropriate or change their beliefs in response to therapists, unwanted thoughts still recur. Suppression is a difficult state of mental control to maintain, as is shown by the intrusion of unwanted thoughts, even under suppression instructions. Suppression is never completely successful; there is always some intrusion of the thoughts being suppressed. However, the incidence of those thoughts is greatly reduced relative to control or thought-expression conditions (Salkovskis & Campbell, 1994). Successful suppression is likely to be especially difficult, and lapses in suppression especially frequent, for offense- or addiction-related thoughts. Such thoughts are chronically accessible and will demand more cognitive resources for suppression than less accessible thoughts, simply because of their frequency of occurrence and hence identification by the monitoring process. Thought suppression provides a useful mechanism for decreasing the incidence of unwanted or inappropriate thoughts. By itself, however, it is not sufficient to prevent completely the occurrence of deviant sexual thoughts and fantasies. In addition to encouraging offenders to adopt such mental control strategies, therapists must also emphasize the inevitability of lapses and provide offenders with the ability to identify high-risk situations, under which lapses are most likely to occur, and strategies to cope with such lapses, such as the use of specific distracters, and, especially, their affective consequences.

The Resource-Demanding Nature of Thought Suppression

Thought suppression is resource demanding, and if sufficient cognitive resources are not available, suppression may falter (Wegner et al., 1993). The more demanding the situation, the more likely offenders are to rely on inferential shortcuts (e.g., Fiske & Taylor, 1991; Hamilton & Trolier, 1986) such as well-learned behavioral scripts (e.g., offending behaviors). That is, demanding situations increase the likelihood of offending behaviors. The situations likely to produce cognitive busyness and hence hamper suppression are just those situations or circumstances described in the literature as facilitating the likelihood of (re-)offending, factors such as high stress levels or strong affective responses (Pithers, 1990; Ward et al., 1995). Several researchers have found that sexual offending is often precipitated by the experience of high levels of stress (e.g., Pithers, Kashima, Cumming, & Beal, 1988). This effect can be explained if offenders are trying to suppress, or control, their sexual thoughts. High levels of stress reduce the available cognitive resources for thought suppression. As a consequence, it will be difficult to suppress deviant sex-related thoughts. Unwanted thoughts or feelings will become accessible as a result of ironic effects. Further attempts to manage the intrusive thoughts, in conjunction with the distress caused by their recurrence, will further increase the individual's level of stress and result in greater impairment of the operating process and weakened cognitive control. It is hard for individuals caught up in this cycle to implement effective coping strategies, despite having learned them while in therapy. This may explain the puzzling finding of research into relapse that participants often fail to implement coping strategies that they already possess (Saunders & Allsop, 1987).

The relapse-prevention treatment model is essentially built around the assumption that offenders need to develop coping responses to risk factors at different points in the offense chain (Laws, 1989). Offenders need to be able to identify high-risk factors and avoid them. In addition to encouraging the process of inhibition, then, therapists need to make offenders aware of, and responsive to, the high-risk factors that will make that inhibition process especially difficult. Obviously, a complete avoidance of high cognitive demands is impossible in everyday life, but offenders need to be aware of what constitute high-risk situations so that they can predict situations under which they will be most vulnerable to lapses in mental control. It is for these situations that offenders need to establish specific distracter, or substitute, thoughts.

Relapse-prevention interventions require that offenders learn to respond to high-risk situations and lapses with rapid, well-rehearsed strategies. Suppression, or replacement, strategies must, then, become well-learned default strategies if self-regulation is to be an effective therapy. The success of suppression can be increased by providing specific distracter or substitute thoughts. Whenever the intrusive thought appears, it is replaced by the specific distracter thought. Not only does this increase the successful suppression of unwanted thoughts, but it also reduces the discomfort produced by failed suppression (Salkovskis & Campbell, 1994). For example, a person trying to give up smoking may try to replace thoughts of a cigarette with thoughts about particular foods or drinks. For sexual offenders, inappropriate sexual thoughts could be replaced either by specific, appropriate sex-related thoughts or thoughts unrelated to sex. The use of distracters becomes increasingly effective with practice (Kelly & Kahn, 1994), so persisting with self-regulatory strategies will make them more effective. The use of specific distracters in suppression reduces the scope of feature-negative searches. Instead of searching for any potential substitute thoughts--any items that are not the particular target (e.g., all appropriate sex-related thoughts or thoughts unrelated to sex)--the operating process only need search for the specific distracter thoughts; the generalized, feature-negative search becomes a specific, feature-positive search. In turn, this process is less cognitively demanding, and so there are fewer lapses in suppression. The use of role playing in therapy sessions will assist in the effectiveness of distracters. Repetition or practice of replacement strategies will result in the operating process becoming more automated (Wegner, 1994), such that it becomes a progressively less conscious, less effortful process. Indeed, the use of programmed coping responses in relapse prevention is based upon the assumption that greater automatisation of coping responses leads to better therapy outcome (Pithers, 1990). As this occurs, of course, the number of high-risk situations may diminish as fewer resources are needed for successful suppression.

The use of specific distracters in suppression needs to be separated from the concept of distraction. The use of specific distracters focuses the mind on specific alternative thoughts rather than simply diverting attention (like distraction) and without imposing the same high-cognitive demands. It is important, however, that the proposed distracters are sufficiently relevant and available; the more positive these distracters, the better. Smokers trying to give up would likely find particular pleasurable foods or drinks far better distracters than, say, white bears. Similarly, sexual offenders would find appropriate sexual thoughts better distracters than thoughts unrelated to sexual behavior, especially if they are still preoccupied with sex.

The cognitive load of therapy interventions can also be reduced through appropriate structuring of the client's goal. Suppression is most successful when the operating process involves feature-positive searches and the monitoring process feature-negative searches. Rather than instructing clients to suppress deviant fantasies, therapists should encourage offenders to seek a state of sexual adjustment. Offenders using the monitoring system, then, would conduct a feature-negative search, identifying any thoughts inconsistent with the goal of sexual adjustment (this would include deviant sexual thoughts), and the operating system would replace those thoughts with goal-consistent thoughts (e.g., thoughts about appropriate sexual liaisons such as those between consenting adults).

The high resource demands of suppression have direct implications for therapists, because therapy may provide a large cognitive load for clients (e.g., learning new coping mechanisms and behaviors). Lapses in suppression are particularly likely in conditions of high cognitive demand, and this could lead to offenders experiencing an increased number of unwanted thoughts during therapy. Unless adequately dealt with by therapists, this, in turn, could lead to negative evaluations of the self and/or the therapy and hence reduce the effectiveness of that therapy. Equally, suppression success in a low-load therapy situation may not generalize to more demanding everyday situations. Clients may be able to suppress unwanted thoughts successfully in the supportive therapy situation where there are few cognitive demands or external factors. However, in everyday situations the pressures of additional cognitive and emotional demands may result in failure of the previously successful suppression techniques. Failure to implement coping strategies learned in therapy has puzzled researchers (Saunders & Allsop, 1987); mental control strategies may explain this failure.

Negative Affect in Response to Failed Suppression

Failed suppression often produces a negative emotional response that may reduce the success of thought suppression. Intrusive thoughts are simply harder to dismiss in a dysphoric than a happy mood state (Sutherland et al., 1982). The negative response to failed suppression attempts may also exacerbate the problem through its impact on both cognitive and affective processes.

Affective states reduce mental capacity, reducing the availability of cognitive resources for subsequent suppression of unwanted thoughts. This reduction in the resources available consequently reduces the success of such suppression. Such a response can lead to the development of a vicious cycle (Wegner, 1994). A negative response to an initial lapse in suppression results in a reduction of available resources for subsequent suppression. As a consequence, lapses become more frequent, resulting in a stronger negative response, which further reduces the available resources, and so on. A vicious self-depreciating cognitive-affective cycle is underway. Such a cycle could develop for offenders desperate to suppress inappropriate sexual thoughts. This cycle may offer an explanation for an increase in severity and frequency of sexually deviant behavior over time (e.g., Marshall & Barbaree, 1989). Successful suppression also can lead to heightened arousal levels (e.g., Gross & Leverson, 1993; Wegner, Shortt, Blake, & Page, 1990), which also reduce the available cognitive resources and, hence, may impede the suppression process. The emotional response to successful suppression may, therefore, reduce the likelihood of further successful suppression. The impact of arousal effects is further enhanced for sexual offenders as a result of the frequent arousal response to sexually deviant thoughts or fantasies (Darke, 1990; Pithers et al., 1988).

In addition to the impact of negative affective states on cognitive processing, individuals' responses to failed suppression also may influence subsequent motivation to employ such techniques. Of particular importance are the possible feelings of helplessness and loss of self control. It is easy to see how feelings of helplessness or not being in control could result from failed suppression. The individual is trying to suppress unwanted thoughts, yet they keep reappearing; apparently, then, cognitive activity is not under the individual's control. As a consequence, individuals may be less motivated to use suppression techniques again, even in situations where they would be effective, and also may have reduced confidence in both their therapy and therapists. Such responses need to be avoided through the introduction of appropriate affective coping strategies in response to inevitable lapses in suppression. For self-regulation to be an effective treatment strategy, a necessary prerequisite is that offenders still feel in control of their mental state but wish to change or control that state. If people do not feel they have control over the occurrence of their thoughts and behaviors, then attempts at mental control or suppression are doomed to failure. This is not to suggest that offenders should be encouraged not to feel a negative response to their lapses, especially to reoffending. Rather, the nature of that response is critical, especially in relation to feelings of self-control. Negative emotional reactions to lapses in suppression and to rebound effects should be not be at the expense of a loss of-perceived self-control.

Consider the differential consequences of feelings of guilt and shame to lapses. When people feel guilt, they see themselves as having had control over the cause of a negative outcome (Graham & Weiner, 1991). A guilty response is characterized by people feeling badly about the specific act but generally okay about themselves. Feelings of shame, however, may reduce the ability to suppress unwanted thoughts. Shame reactions are characterized by uncontrollable causal beliefs (Weiner, 1986), subjective distress, and withdrawal but, paradoxically, increases in anger, hostility, and externalising behavior (Tangney, 1991). Data linking the degree of shame felt by sexual offenders and the extent to which they blame and victimize others are needed. However, it is possible, from the findings with other populations (Graham & Weiner, 1991; Tangney, 1991; Weiner, 1986), to speculate that feelings of guilt after inappropriate sex-related thoughts/behaviors will reduce the likelihood of their recurrence, whereas feelings of shame at their offending will lead offenders to blame and victimize others and increase the likelihood of relapse.

This effect can be linked to the abstinence violation effect (AVE; Marlatt & Gordon, 1985), which considers the impact of lapses (initial violations of an abstinence rule or the occurrence of some undesired activity) on self-regulation and the subsequent consequences for full-blown relapse (i.e., resumption of problematic levels of the addictive behavior). This effect has been documented in child molesters (Ward et al., 1994), with lapses in self-regulation being defined as voluntarily induced high-risk thoughts (e.g., deviant sexual fantasising) and relapse as actual offenses. Response to abstinence violation includes two reciprocal components: a cognitive attribution of cause and an affective reaction. If the lapse is attributed to external, unstable specific factors, then the negative affective response is minimal. If the lapse is attributed to internal, dispositional factors, then the negative affective response is high. Greater negative affect is more likely to yield an AVE leading to full-blown relapse. Encouraging a low AVE response to such lapses reduces the risk of lapses turning into relapse, that is, reduces the risk of deviant sexual fantasies becoming inappropriate sexual behaviors. A minimal AVE response is incompatible with feelings of uncontrollability, emphasising the point made earlier that self-regulation strategies will only be effective in situations where offenders feel in control of their mental state. Given the inevitability of lapses in suppression, it is important that offenders are prepared for the occurrence of lapses so that they do not despair, or feel out of control, as a consequence.

Ironic Effects as a Consequence of Successful Suppression

In addition to learning to identify high-risk situations for lapses in suppression, offenders should also learn to avoid potential rebound situations. Rebound effects can occur even after successful suppression, simply as a result of a change in suppression cues (Macrae et al., 1994). Such rebound effects, in part, may help to explain why sexual offenders show high rates of recidivism (e.g., Marshall, in press). The high-risk situations for rebound effects are similar to those for lapses in suppression, that is, in situations of high cognitive load, when performing concurrent tasks, when under time pressure, when there are affective and stress-related preoccupations (Gilbert, 1991; Logan, 1979), and in states of alcohol- or drug-induced impaired attention (Steele & Josephs, 1990). Such high-risk situations need to be avoided, therefore, even after successful suppression.

Similar to lapses in suppression, the more chronically accessible the unwanted thoughts, the greater the ironic consequences of suppression. The more chronically accessible, the more these thoughts will be identified during the monitoring process and the more resources will be needed by the operating process to suppress them. This increases the likelihood that insufficient resources will be available and, hence, increases the likelihood of suppression faltering. Similarly, those thoughts will be more accessible on subsequent tasks. Of course, offense- or addiction-related thoughts will be chronically accessible and consequently especially hard to suppress and prone to rebound effects. Like lapses, rebound effects are reduced when specific distracters are employed in suppression (Wegner et al., 1987). Mention of distracters is negatively correlated with the intrusion of unwanted thoughts (Kelly & Kahn, 1994), and providing participants with a specific distracter, a test, reduced the occurrence of rebound of self-generated thoughts (Salkovskis & Campbell, 1994). The presence of specific distracters also reduces the negative affect and feelings of a loss of control that accompany rebound effects (e.g., Kelly & Kahn, 1994). Employing specific distracters increased the success of suppression attempts and reduced the occurrence of rebound effects.

Offender Types

The success of the use of suppression techniques in therapy may be related to the type of sexual offender. Different types of sexual offenders have been identified in the literature. Although such categorisations are oversimplifications and do not capture the complexities of offender typology (Knight & Prentky, 1990; Simon, Sales, Kaszniak, & Kahn, 1992), two types of offenders are consistently identified, the preferential and situational offenders (Lanyon, 1986). These types of offenders are likely to be differentially suited to self-regulation therapies.

Situational offenders' first offense is usually in adulthood, and these offenders often experience higher levels of guilt and remorse than preferential offenders whose deviant sexual interest usually begins earlier (Ward, Louden, Hudson, & Marshall, 1995). Situational offenders are likely to accept their sexual thoughts and fantasies as unwanted and, therefore, be willing to attempt to suppress or control such thoughts. Such offenders may have already tried to instigate self-imposed suppression of their sexually deviant thoughts. This urge to suppress deviant thoughts, however, may restrict offenders' abilities to suppress those thoughts effectively, suppression being impaired by a negative response to lapses in suppression. The use of suppression techniques with situational offenders must, therefore, emphasize the inevitability of lapses in mental control and incorporate techniques to cope with those lapses, both cognitively (e.g., use of specific distracters) and affectively (e.g., avoiding feelings of helplessness). Incorporating techniques to deal with the consequences of suppression of unwanted thoughts will result in offenders' desire to suppress their unwanted sexual thoughts having a positive outcome (reducing the occurrence of those thoughts) rather than a negative outcome (persistence of the disorder).

Preferential offenders are less likely than situational offenders to be motivated to suppress deviant sexual thoughts. Suppression of deviant sexual thoughts will likely be highly cognitively demanding for such offenders; deviant sexual thoughts are chronically accessible for these individuals. Their extensive cognitive networks of sexually inappropriate memories, attitudes, and beliefs also may make them prone to strong rebound effects as these thoughts flood into consciousness following lapses in suppression. The AVE framework is inapplicable for some offenders (usually preferential offenders). For those offenders who respond positively to offending, there will be no search for causal explanations, or excuses, for lapses. Self-regulation will only be an effective therapy for those who acknowledge their offending as undesirable.

Implications for Clinical Practice

Guidelines from the mental control literature are summarised next. The use of these guidelines in therapy for sexual offenders remains to be tested. First, offenders need to acknowledge their sexually deviant thoughts and behaviours as unwanted. Therapists can then develop with offenders a goal of sexual adjustment. Mental control will be more successful if the goal is one of sexual adjustment rather than one of banishing unwanted sexual thoughts. Therapists also need to bear in mind at this stage that the use of mental control techniques may be applicable only for certain types of sexual offenders (e.g., situational offenders). Second, after identifying the appropriate goal, the offender should be encouraged to practice the mental control techniques of identification of unwanted sexual thoughts and the replacement of those thoughts by appropriate distracter thoughts. The more practice offenders have (e.g., through role play), the more automated and effective the mental control will become and the fewer lapses in mental control and intrusions of unwanted thoughts will occur. Third, in addition to teaching mental control techniques to offenders, therapists also need to prepare offenders for inevitable lapses in mental control and the intrusion of unwanted thoughts. In particular, the affective reaction of offenders to these lapses needs to be developed so that feelings of helplessness and a loss of control are avoided. Fourth, offenders need to be taught to identify, and avoid where possible, high-risk situations for lapses in mental control and rebound effects after successful mental control. Offenders need to be able to identify situations that are stressful and produce a high cognitive load.

Conclusions

We argued that thought suppression is applicable to sexual offending. The processes underlying thought suppression and rebound effects provide a theoretical basis for a number of effects reported by researchers and therapists. The mental control model, with its two-process system, accounts for the breakdown in cognitive control that underlies many features of sexual offending. For example, the impact of stress, alcohol, or strong affective states on the replacement of unwanted thoughts ironically leads to the return of unwanted thoughts, images, and feelings. This fact, alongside the tendency for ironic processes at times to escalate rapidly (a self-loading ironic system), also can account for the rapid increase in severity and frequency of sexual offending.

Concerning treatment, the three features mentioned in the literature on mental control as ways of countering susceptibility to ironic effects are useful general guides in therapy. Teaching offenders to manage stress effectively and to develop appropriate beliefs about what is controllable and making the operating process automatic are important. At a more specific level, the implications of thought suppression for different treatment strategies, for example, the importance of specific, feature-positive searches, are of practical use to therapists. Viewing therapy as constituting a cognitive load in its own right means that in some situations it may be necessary to adjust the pace of treatment accordingly. To ignore this variable may result in a paradoxical increase in unwanted thoughts or feelings and an erosion of an offender's confidence that therapy will prove effective.

Researchers should further investigate the application of the whole mental control model to sexual offending, rather than just thought suppression. Such work may clarify some relevant factors underlying relapse in different types of offenders. Identification of individuals who find it difficult to suppress unwanted thoughts may predict successful versus unsuccessful treatment outcome and also help to guide intervention choice. Those men who experience problems monitoring mental states, or engaging in other metacognitive activities, may not respond optimally to CBT. The inability to suppress dysfunctional thoughts or feelings effectively would mean it is unlikely that a treatment approach such as relapse prevention would be beneficial. The underlying theoretical basis for making such treatment choices has not been clearly formulated and has typically been based on judgments concerning the client's degree of motivation. The mental control model spells out the mechanisms generating the dispositional characteristics associated with client suitableness for CBT Wegner and Zanakos (1993) described a trait of thought suppression. The greater an individual's ability to suppress thoughts, the more applicable will be CBT.

In addition to the implications of mental control for the treatment of sexual offenders, the cognitive processes of suppression also may be involved in the development of sex offending. For example, some rapists' offending appears to be causally related to an excessively strong need to control and dominate women (Darke, 1990; Scully, 1988). Unwanted thoughts in this context would revolve around the issue of personal inadequacy, perceiving the self as weak, despised, and vulnerable. Either suppression of these thoughts of inadequacy or seeking a mental state where the self is seen as powerful and dominant may lead to ironic effects and, as a consequence, strong affective responses and impaired cognitive control processes. This may contribute to a vulnerable individual being increasingly preoccupied with unwanted self-deprecating thoughts and ultimately acting in a sexually aggressive manner. The possible role of thought suppression in the development of sexual crimes needs further attention.

The development of comprehensive therapy programs for sexual offenders has made rapid progress in recent years, and there are now grounds for optimism concerning the treatability of these men. We argue that an integration of the mental control literature from clinical and social domains (e.g., Johnston & Ward, 1996) may provide a useful perspective on the treatment of sexual offenders, in addition to suggesting a potentially promising future research program.

References

Baumeister, R. F. (1991). Escaping the self New York: Basic Books.

Bliss, E. L., & Larson, E. M. (1985). Sexual criminality and hypnotizability. Journal of Nervous and Mental Disease, 173, 522-526.

Cioffi, D., & Holloway, J. (1993). Delayed costs of suppressed pain. Journal of Personality and Social Psychology, 64, 274-282.

Clark, D. M., Ball, S., & Pape, D. (1991). An experimental investigation of thought suppression. Behavior Research and Therapy, 29, 253-257.

Clark, D. M., Winton, E., & Thynn, L. (1993). A further experimental investigation of thought suppression. Behavior Research and Therapy, 31, 207-210.

Conway, M., Howell, A., & Giannopoulos, D. (1991). Dysphoria and thought suppression. Cognitive Research and Therapy, 15, 153-166.

Darke, J. L. (1990). Sexual aggression: Achieving power through humiliation. In W. L. Marshall, D. R. Laws, & H. E. Barbaree (Eds.), Handbook of sexual assault: Issues, theories and treatment of the offender (pp. 55-72). New York: Plenum.

Erdelyi, M. H., & Goldberg, B. (1978). Let's not sweep repression under the rug. Toward a cognitive psychology of repression. In J. F. Kihlstrom & F. J. Evans (Eds.), Functional disorders of memory (pp. 355-402). Hillsdale, NJ: Erlbaum.

Fiske, S. T., & Taylor, S. E. (1991). Social cognition (2nd ed.). New York: McGraw-Hill.

Freud, S. (1957). Repression. In J. Strachey (Ed. & Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 14, pp. 146-158). London: Hogarth. (Original work published 1915).

Freund, K., Watson, R., & Rienzo, D. (1988). Signs of feigning in the phallometric test. Behaviour Research and Therapy, 26, 105-112.

Gilbert, D. T. (1991). How mental systems believe. American Psychologist, 46, 107-119.

Graham, S., & Weiner, B. (1991). Testing judgements about attribution-emotion-action linkages: A life-span approach. Social Cognition, 9, 254-276.

Gross, J. J., & Leverson, R. W. (1993). Emotional suppression: Physiology, self-report and expressive behavior. Journal of Personality and Social Psychology, 64, 970-986.

Hall, G. C. N. (1989). Sexual arousal and arousability in a sexual offender population. Journal of Abnormal Psychology, 98, 145-149.

Hamilton, D. L., & Trolier, T. K. (1986). Stereotypes and stereotyping: An overview of the cognitive approach. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination and racism (pp. 127-163). Orlando, FL: Academic Press.

Howell, A., & Conway, M. (1992). Mood and the suppression of positive and negative self-referent thoughts. Cognitive Therapy and Research, 16, 535-555.

Jacoby, L. L. (1983). Remembering the data: Analyzing interactive processes in reading. Journal of Verbal Learning and Verbal Behavior, 22, 485-508.

Johnston, L., & Ward, T. (1996). Social cognition and sexual offending: A theoretical framework. Sexual Abuse: A Journal of Research and Treatment, 8, 55-80.

Kelly, A. E., & Kahn, J. H. (1994). Effects of suppression of personal intrusive thoughts. Journal of Personality and Social Psychology, 66, 998-1006.

Knight, R. A., & Prentky, R. A. (1990). In W. L. Marshall, D. R. Laws, & H. E. Barbaree (Eds.), Handbook of sexual assaults: Issues, theories and treatment of the offender (pp. 23-52). New York: Plenum.

Lanyon, R. I. (1986). Theory and treatment in child molestation. Journal of Consulting and Clinical Psychology, 54, 176-182.

Laws, D. R. (1989). (Ed.). Relapse prevention with sexual offenders. New York: Guilford.

Logan, G. D. (1979). On the use of concurrent memory load to measure attention and automaticity. Journal of Experimental Psychology: Human Perception and Performance, 5, 189-207.

Macrae, C. N., Bodenhausen, G. V., Milne, A. B., & Jetten, J. (1994). Out of mind but back in sight: Stereotypes on the rebound. Journal of Personality and Social Psychology, 67, 808-817.

Marlatt, G. A., & Gordon, J. R. (1985). Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. New York: Wiley.

Marshall, W. L. (in press). Assessment, treatment, and theorizing about sex offenders: Developments over the past 20 years and future directions. Criminal Justice and Behavior.

Marshall, W. L., & Barbaree, H. E. (1989). Sexual violence. In K. Howells & C. R. Hollin (Eds.), Clinical approaches to violence (pp. 205-246). New York: Wiley.

Muris, P., Merckelbach, H., van den Hout, M., & de Jong, P. (1992). Suppression of emotional and neutral material. Behavior Research and Therapy, 30, 639-642.

Newman, J. P., Wolff, W. T., & Hearst, E. (1980). The feature-positive effect in adult human subjects. Journal of Experimental Psychology: Human Learning and Memory, 6, 630-650.

Parkinson, L., & Rachman, S. (1981). The nature of intrusive thoughts. Advances in Behavior Research and Therapy, 3, 101-110.

Pithers, W. D. (1990). Relapse prevention with sexual aggressors: A method for maintaining therapeutic gain and enhancing external supervision. In W. L. Marshall, D. R. Laws, & H. E. Barbaree (Eds.), Handbook of sexual assault: Issues, theories and treatment of the offender (pp. 343-361). New York: Plenum.

Pithers, W. D., Kashima, K. M., Cumming, G. F., & Beal, L. S. (1988). In A. C. Salter (Ed.), Treating child sex offenders and victims: A practical guide (pp. 131-170). Newbury Park, CA: Sage.

Rachman, S., & Hodgson, R. (1980). Obsessions and compulsions. Englewood Cliffs, NJ: Prentice Hall.

Salkovskis, P. M. (1989). Cognitive-behavioral factors and the persistence of intrusive thoughts in obsessional problems. Behavior Research and Therapy, 27, 677-682.

Salkovskis, P. M., & Campbell, P. (1994). Thought suppression induces intrusion in naturally occurring negative intrusive thoughts. Behavior Research and Therapy, 32, 1-8.

Saunders, W. A., & Allsop, S. (1987). Relapse: A psychological perspective. British Journal of Addiction, 82, 417-429.

Scully, D. (1988). Convicted rapists' perceptions of self and victim: Role taking and emotions. Gender & Society, 2, 200-213.

Simon, L. M. J., Sales, B., Kaszniak, A., & Kahn, M. (1992). Characteristics of child molesters: Implications for the fixated-regressed dichotomy. Journal of Interpersonal Violence, 7, 211-225.

Smith, E. R., Stewart, T. L., & Buttram, R. T. (1992). Inferring a trait from a behavior has long-term, highly specific effects. Journal of Personality and Social Psychology, 62, 753-759.

Steele, C. M., & Josephs, R. A. (1990). Alcohol myopia: Its prized and dangerous effects. American Psychologist, 45, 921-933.

Sutherland, G., Newman, B., & Rachman, S. (1982). Experimental investigations of the relations between mood and intrusive unwanted cognitions. British Journal of Medical Psychology, 55, 127-138.

Tangney, J. P. (1991). Moral affect: The good, the bad and the ugly. Journal of Personality and Social Psychology, 61, 598-607.

Ward, T., Hudson, S. M., & Marshall, W. L. (1994). The abstinence violation effect in child molesters. Behavior Research & Therapy, 32, 431-437.

Ward, T., Hudson, S. M., & Marshall, W. L. (1995). Cognitive distortions and affective deficits in sex offenders: A cognitive deconstructionist interpretation. Sexual Abuse: A Journal of Research and Treatment, 7, 67-83.

Ward, T., Hudson, S. M., & Siegert, R. J. (1995). A critical comment on Pithers' relapse prevention model. Sexual Abuse: A Journal of Research and Treatment, 7, 167-175.

Ward, T., Louden, K., Hudson, S. M., & Marshall, W. L. (1995). A descriptive model of the offense chain for child molesters. Journal of Interpersonal Violence, 10, 452-472.

Wegner, D. M. (1994). Ironic processes of mental control. Psychological Review, 101, 34-52.

Wegner, D. M., & Erber, R. (1993). Social foundations of mental control. In D. M. Wegner & J. W. Pennebaker (Eds.), Handbook of mental control (pp. 36-56). Englewood Cliffs, NJ: Prentice Hall.

Wegner, D. M., Erber, R. & Bowman, R. E. (1993). On trying not to be sexist. Unpublished manuscript.

Wegner, D. M., Schneider, D. J., Carter, S. Ill., & White, L. (1987). Paradoxical effects of thought suppression. Journal of Personality and Social Psychology, 53, 5-13.

Wegner, D. M., Shortt, J. W., Blake, A. W., & Page, M. S. (1990). The suppression of exciting thoughts. Journal of Personality and Social Psychology, 58, 409-418.

Wegner, D. M., & Zanakos, S. (1993). People who try not to think: The trait of thought suppression. Manuscript submitted for publication.

Weiner, B. (1986). An attributional theory of motivation and emotion. New York: Springer-Verlag.

Wenzlaff, R., Wegner, D. M., & Roper, D. (1988). Depression and mental control: The resurgence of unwanted negative thoughts. Journal of Personality and Social Psychology, 55, 882-892.

Manuscript accepted May 20, 1996

Lucy Johnston Tony Ward Stephen M. Hudson University of Canterbury

Address correspondence to Lucy Johnston, Ph.D., Department of Psychology, University of Canterbury, Private Bag 4800, Christ-church, New Zealand. E-mail: [e-mail].