Thought suppression

Thought suppression is the process of deliberately trying to stop thinking about certain thoughts (Wegner, 1989). It is often associated with obsessive-compulsive disorder, in which a sufferer will repeatedly (usually unsuccessfully) attempt to prevent or "neutralize" intrusive distressing thoughts centered around one or more obsessions. It is also related to work on memory inhibition.

Thought suppression is different from Freud's (1955) concept of repression, which is unconscious and automatic and has relatively little empirical support (see Eysenck, 1985; Holmes, 1990 for a review). Over thirty-five experiments to date have found evidence for thought suppression and its effectiveness. It can produce paradoxical effects for personally irrelevant and relevant thoughts at both a mental and a behavioral level.

Empirical work
In order for thought suppression and its effectiveness to be studied researchers have had to find ways of tapping the processes going on in the mind so that they may be described. One such paradigm by Wegner, Schneider, Carter & White (1987) was to ask people not to think of a target (e.g. "white bear") for five minutes but if they did to ring a bell. After this, participants were told to think about the target for five minutes more. Compared to those who had not used suppression there was evidence for unwanted thoughts being immediately enhanced during suppression and, furthermore, a higher frequency of target thoughts during the second stage, dubbed the rebound effect (Wegner, 1989). This effect has been replicated with different targets (Lavy & Van den Hout, 1990) and even implausible targets like "green rabbit" (Clark, Ball, & Pape, 1991). As a result, Wegner (1994) suggested the "Ironic Process Theory" where two sometimes opposing mechanisms may be at work; one trying to think of something else (for example, something other than the bear) and the other, checking in sporadically to see if the thought in question has stopped.

While there is some good evidence for thought suppression causing increased immediate and/or delayed target thoughts, several critical points can be raised. Firstly, typical thought suppression may not involve simple targets like colored animals but socially more complex and personal thoughts. Secondly, the time frame used in these studies is only representative of thought suppression in short spaces of time, which may not accurately mirror typical human behavior where longer term suppression (like trying not to think about a recent ex-partner) may be manifest. Thirdly, the paradoxical effects could be elicited by the act of ringing the bell alone. Therefore, although there is good laboratory evidence for the poor effectiveness of thought suppression, confidently projecting such findings onto naturalistic behaviors is conceivably problematic.

Improved methodology
To better elucidate the findings of thought suppression, several studies have changed the target thought from a personally irrelevant to relevant one. Roemer and Borkovec (1994) found that participants who suppressed anxious or depressing personal thoughts showed a significant rebound effect compared to those who expressed the thoughts from the outset. Furthermore, Wenzlaff, Wegner, & Roper (1988) demonstrated that anxious or depressed subjects were less able to suppress negative unwanted thoughts. Despite Rassin, Merkelbach and Muris (2000) reporting that this finding is moderately robust in the literature some studies were unable to replicate results (e.g. Smári, Sigurjónsdóttir, & Sæmundsdóttir, 1994; Kelly & Kahn, 1994; Wegner, Quillian, & Houston, 1996). However, this may be explained by a consideration of individual differences.

Recent research by Geraerts, Merckelbach, Jelicic, & Smeets (2006) found that for individuals with low anxiety and high desirability traits (repressors) suppressed anxious autobiographical events intruded fewer times than in other (low, high and high defensive anxious) groups initially but showed more intrusions after one week. This difference in coping style may account for the disparities within the literature. That said, the problem remains that the cause of the paradoxical effect may be in the thought tapping measures used (e.g. bell ringing). Evidence from Brown (1990) that showed participants were very sensitive to frequency information prompted Clarke, Ball and Pape (1991) to obtain participants' aposterio estimates of the number of intrusive target thoughts and found the same pattern of paradoxical results. However, even though such a method appears to overcome the problem, it and all the other methodologies use self-report as the primary form of data-collection. This may be problematic because of response distortion, where participants may lower their reported frequencies so as to avoid the risk of being pejoratively labelled.

Behavioral domain
A reaction to this has been to explore the effects of thought suppression using more reliable measures, like behavior. Macrae, Bodenhausen, Milne, and Jetten (1994) found that when asked not to think about the stereotype of a certain group (e.g. a "skinhead"), individuals' written descriptions of a group member's typical day contained less stereotypical thoughts than that of controls. However, when told they were going to meet such an individual those in the suppression condition sat significantly further away from the seat the "skinhead" had evidently occupied moments earlier (by virtue of his clothes being present). These results show that even though there may have been an initial enhancement of the stereotype participants were able to prevent this being communicated in writing but not in their behavior.

Further experiments have documented similar findings (e.g. Cioffi and Holloway, 1993; Wegner, Shortt, Blake, and Page, 1990). In other experiments, participants were given cognitively demanding concurrent tasks and the results showed a paradoxical higher frequency of target thoughts than controls (Wegner & Erber, 1992; Wegner, Erber & Zanakos, 1993). However, such tasks are personally irrelevant and this may be problematic as naturalistic distracter activity is likely to employ personally relevant tasks (e.g. phoning a friend when trying not to think of an ex-partner). Another criticism has been that all these experiments may not be accounting for the plausible strategy of naturalistic thought suppression to find distracters.

Some studies have shown that when test subjects are under what Wegner refers to as a "cognitive load" (for instance, using multiple external distractions to try and suppress a target thought), the effectiveness of thought suppression appears to be reduced. However, in other studies in which focused distraction is used, long term effectiveness may improve. That is, successful suppression may involve less distractors. For example, Wegner, Schneider, Carter & White (1987) found that a single, pre-determined distracter (e.g., a red Volkswagen) was sufficient to eliminate the paradoxical effect post-testing. Evidence from Bowers and Woody (1996) is supportive of the finding that hypnotized individuals produce no paradoxical effects. This rests on the assumption that deliberate "distracter activity" is bypassed in such an activity. However, it may be also be that in real life, the functions of natural thought suppression in wake of unpleasant or threatening thoughts, isn't similar to how multiple or single distracters are used during experimental conditions in which participants are deliberately being instructed, under time constraints, and responding to investigator expectations.

Cognitive dynamics
Thought suppression is typically ineffective with activities causing an increase in the to-be-suppressed thought, which is exacerbated when the cognitive load is increased. The studies are unable to find this effect for emotional thoughts in hypnotized individuals, and when one distracter is used. In an attempt to account for these findings a number of theorists have produced cognitive models of thought suppression. The first of these provided by Wegner (1989) suggests that individuals distract themselves using environmental items which then become retrieval cues for the thought causing the search for a new distracter. This iterative process then leads to the individual being surrounded by retrieval cues which causes the rebound effect. Certainly the evidence for multiple distracters is supportive but it cannot explain the initial thought enhancement, or the single distracter results. Wegner hypothesizes that multiple retrieval cues not being forged explains, in part, the effectiveness of single distracters (i.e., a reduction of mental load). This is because there may be an ideal balance between the two processes with the cognitive demand not being too great as to let the monitoring process supersede it.

Wegner also proposes that increased cognitive load may overtax conscious effort at effective control, allowing the greater influence of the ironic thought process. However, while his theories may explain why suppression leads to increased frequency of intrusive thoughts, they cannot do so in a way that is completely satisfactory as some studies do not find supportive evidence. As recent research suggests (e.g. Geraerts et al., in press) there may be an important role of individual differences that may be able to account for this, however.

Recently thought suppression is been seen as a form of "experiential avoidance". Experiential avoidance refers to attempts to suppress, change or control unwanted internal experiences (thoughts, feelings, bodily sensations, memories, etc.) (Hayes ea, 1996; Kashdan ea, 2006). This line of thinking supports on Relational Frame Theory.

Summary
As time has progressed, experiments have become more elaborate and better able to extend their findings to thought suppression in its various guises. The results of these studies have demonstrated that trying to suppress impersonal and, sometimes, personal thoughts under cognitive load is often ineffective, as the frequency of those thoughts increases during suppression and after it. Importantly, while the evidence shows that we can control these thoughts from being translated into behavior when self-monitoring is high such control is not observable in normal, automatic behaviors (e.g. skinhead scenario). In addition, this phenomenon is made paradoxically worse by increasing the amount of distractions a person has, although the experiments in this area can be criticized for using impersonal concurrent tasks which may not properly reflect natural processes or individual differences. Under experimental conditions, the use of one distracter (or less cognitive load) suggests that thought suppression, and specifically its aftermath (i.e. the "rebound effect") can be accomplished.

Wegner's (1994) "Ironic Process Theory" represents one cognitive model which attempts to explain these results. However, given the mixed evidence and commensurate with the latest research, it is suggested that such a model needs to account for individual differences, as well as possible neurobiologic dynamics (see, for instance, Obsessive Compulsive Disorder), to be considered robust.

In conclusion, thought suppression is a real phenomenon with observable effects, and direct attempts to suppress unwanted thoughts---particularly when involving multiple distracters and/or extra cognitive loads---have been shown largely ineffective or difficult to achieve, in the laboratory at least.