You are on page 1of 17

Journal of Cognitive Psychotherapy: An International Quarterly

Volume 21, Number 3 2007

The Menace Within: Obsessions and the Self

Frederick Aardema, PhD Kieron OConnor, PhD, MPhil
Centre de Recherche Fernand-Seguin Montreal, Quebec, Canada

The current article attempts to provide a theoretical account of obsessions about blasphemy, sexuality, and aggression from an inference-based perspective. It is argued that self-evaluative and self-representational dimensions in obsessions need to be taken into account to allow for the misrepresentation of mental states. A persuasive narrative containing rhetorical devices informs the misrepresentation of mental states and gives credibility to the obsession. These narrative devices seem to originate from a distrust of the self where the person overinvests in a sense of self-as-could-be rather than a sense of self-as-is, which consequently gives rise to a discordance between the person and the obsession (i.e., egodystonicity). The article concludes with some theoretical and clinical implications for cognitive models of obsessive-compulsive disorder.

Keywords: obsessive-compulsive-disorder; self; reasoning; inference-based approach

My conscience hath a thousand several tongues, And every tongue brings in a several tale, And every tale condemns me for a villain. William Shakespeare, Richard III, Act 5, Scene iii.

hile common usage of the term obsession generally refers to an excessive preoccupation with a particular idea or activity (e.g., obsessed with work, obsessed with sex), the clinical use of the term is reserved for obsessive-compulsive disorder (OCD), where it refers to recurrent and persistent thoughts, impulses, or images that are experienced, at some time during the disturbance, as intrusive and inappropriate and that cause marked anxiety or distress (American Psychiatric Association, 2000). The most well-known subtypes of OCD revolve around washing and checking, and the person is plagued by frequent thoughts that, for example, he or she may be contaminated or has forgotten to lock the door. Subsequently, the person engages in frequent washing or checking to alleviate the distress caused by the obsession. However, one of the largest OCD subgroups (estimates are as high as 50% to 60%) experience obsessions without overt compulsions (Weisman et al., 1994). Typically, these obsessions revolve around thoughts of blasphemy, aggression, and sexuality, including thoughts of performing embarrassing acts, sexual assault, child molestation, acts of violence, murder, or anything else that is considered a forbidden act of thought and/or behavior. These types of obsessions, especially in severe cases, present themselves to a persons awareness with great intensity and have a

2007 Springer Publishing Company

The Menace Within


very strong reality value (i.e., they feel real). For example, a person with the obsessions that he or she might be a child molester or a murderous lunatic experiences this idea with a force that makes it almost impossible to dismiss. However, obsessions are not solely defined by their persistence and intensity. The main distinguishing characteristic in the clinical use of the term obsession lies in its unwanted nature (Rachman & Hodgson, 1980). The unwelcome entry of obsessions differentiates OCD from other disorders, such as the paraphilias, where the person happily invites persistent thoughts into awareness. A related aspect to the unwantedness of obsessions is that they are usually experienced as unreasonable thoughts. People with OCD will often exclaim, I know it makes no sense, but I cannot stop thinking about it, Its unrealistic, but I cannot do anything else other than do the [ritual], or Its stronger than me. This incongruence between the person and the obsession has often been referred to as the egodystonic nature of obsessions, where the person experiences the thought as unreasonable but at the same time is unable to distance himor herself from the thoughts. However, despite egodystonic experience, the person with OCD generally recognizes obsessive thoughts as coming from within and in one way or another as their own thoughts. Insight into the origin of the obsession as coming from within, as well as insight into its irrationality, is largely responsible for current conceptualizations of OCD as an anxiety disorder rather than a belief disorder. However, the extent that the person evaluates the obsession as incongruent with his or her actual character (i.e., the degree of egodystonicity) may vary between individuals and across different situations. For example, egodystonicity may be stronger in the therapists office when the person quietly reflects on his or her obsessional ruminations, whereas the same obsessions may be experienced as completely egosyntonic during a full-blown obsessional episode. This distinction is quantifiable as the degree of belief in the obsession in and out of the OCD context (Grenier & OConnor, 2007). In some cases, the obsession may be evaluated as an entirely realistic state of affairs where, for example, the person believes he or she actually could murder someone or actually is a child molester. Hence, egodystonicity likely lies on a continuum where people with OCD differ in the extent that they consider the obsession to be in discordance with their actual character, tendencies, and motivations. We will refer to this type of discordance as subjective discordance in order to highlight its subjective nature and its dependency on the individuals personal evaluation. However, discordance between the obsession and the persons character also exists on a more fundamental level. This type of discordance refers to the falsehood of the obsession independent of the persons own evaluation and could be termed objective discordance. In other words, the obsession is always false and incongruent with the persons actual character, intent, and behavior, whether or not the person subjectively evaluates it as incongruent. There are numerous examples in the literature of objective discordance where the content of the obsession stands in sharp contrast with the persons actual desires, inclinations, and beliefs. Rachman and Hodgson (1980) cite a poignant phrase noting the apparent discrepancy between the person and the obsession: They repent when they have not sinned, and accuse themselves without form or matter; their virtues make them tremble and in their innocence they are afraid (p. 24). How do people come to seriously consider the possibility of being capable of terrible acts? How does someone come to doubt his or her sexual orientation? How does a responsible mother arrive at the belief she might suffocate her children? The current article attempts to provide a theoretical account of obsessions about blasphemy, sexuality, and aggression from an inference-based perspective on obsessions (Aardema & OConnor, 2003; OConnor, Aardema, & Plissier, 2005; OConnor & Robillard, 1995, 1999). We will start with a brief historical overview of unwanted mental states, followed by a critical discussion of current cognitive models of obsessions. Next, it will be argued that self-evaluative and self-representational dimensions in obsessions need to be taken into account that allow for the misrepresentation of mental states.


Aardema and OConnor

In turn, it will be shown how a persuasive narrative containing rhetorical devices informs the misrepresentation of mental states and gives credibility to the obsession. We will relate these imaginative and inference processes to figureground relationships in self-as-is and the self-as-could-be leading to an illusionary sense of discordance between the obsession and the self (i.e., egodystonicity). Finally, we will conclude with some theoretical and clinical implications for cognitive models of OCD.


The concept of unwanted mental states as existing independently from behavior is today a phenomenon taken for granted, but this may not always have been the case. In particular, various ancient cultures were far more ambivalent about separating mental states from observable behavior and even lacked a comprehensive vocabulary to talk about inner mental states independent from observable behavior. For example, in the ancient Hebrew language, the phrase slow to anger in the sentence The Lord is compassionate and gracious, slow to anger, abounding in love is literally translated as slow to nose, referring to the flaring of the nostrils when one is angry (Benner, 2006). The tendency of the ancient Hebrews to emphasize observable behavior rather than abstract thought is also apparent throughout the Old Testament with rules and regulations dealing primarily with observable action rather than inner mental states. The Ten Commandments serve as a particularly good example in this regard. Up until the ninth commandment, all rules and regulations are behavioral in nature (e.g., thou shalt not kill, thou shalt not steal). However, the 10th commandment seems to refer to an inner state of mind where one should not covet the possessions of others. However, because meaning in the Hebrew language is heavily dependent on context, scholars are still debating whether the meaning of the verb to covet refers to observable steps toward taking an forbidden action or an actual forbidden mental state or sin of intention (see von Rad, 1966, p. 59). Whether the term includes the intentional state, as well as taking actual steps toward an action, is not a question that concerns us here. What is particularly relevant, however, is the fact that there is confusion over the term in the first place and how the unity between thought and action came under attack in later historical movements. The influence of the Greco-Roman culture became ascendant around 200 b.c., and far more explicit rules of the heart and mind began to emerge as separate and distinct from behavior. The Greek culture was based on the intellect more than on action, and virtues and personal cultivation of the individual took center stage. Unlike the Hebrew language, the Greek language contained a more extensive vocabulary, allowing for a clear reference to mental states independently of observable behavior. In particular, the Greek New Testament contained a greater distinction between thought and action than the Old Testament, and unwanted mental states were addressed explicitly. In the words of Jesus Christ (Matthew 5:28), But I say to you, everyone who looks at a woman with lust has already committed adultery with her in his heart. This type of moral fusion between thought and action (see Safran, Thordarson, & Rachman, 1996) would dictate thinking on forbidden thoughts for centuries to come. Subsequent interpretations of the Old Testament as involving rules and matters of the mind also prevailed up to more recent times, giving rise to an increased complexity of biblical doctrine. This was well recognized by Rabbi Bahya Ibn Pakuda in his 11th-century work Duties of the Heart (see Weiman-Kelman, 1999):
it is easier to obey duties of the limbs, those obligations and prohibitions that involve outward behavior, and which thus involve social approbation and disapprobation, than it is to fulfill the duties of the heart, since only God knows if we have fulfilled the latter, not our neighbors, or, often, not even ourselves.

The Menace Within


An example that well illustrates how the 10th commandment Thou shall not covet led to complications, once it became understood as referring to an inner state of mind rather than behavior, is the following story of Rabbi Robert Klein (from Weissbard, 1997):
A gentleman with a fair reputation as a respectable citizen was caught stealing. When hauled into court he was asked by the judge whether he knew that stealing was forbidden in the Ten Commandments. The accused replied that when he first saw the object he stole, he sensed immediately that he would have to break one of the Ten Commandments. He would have to transgress either the Eighth Commandment, Thou shall not steal, or the Tenth Commandment, Thou shall not covet. After weighing in his mind the gravity of each transgression, he decided to violate the Eighth Commandment. He reasoned this way: If he stole the object, he would commit a crime once and the sin would have run its course. If he didnt steal the object, it would lead to the violations of the Tenth Commandment again and again, for each time he beheld it he would covet it.

Like the Greeks, the early Christians emphasized personal responsibility for thought, and a serious effort was made to curtail unwanted mental states. In the words of Paul (2 Corinthians 10:5), We demolish arguments and every pretension that sets itself up against the knowledge of God, and we take captive every thought to make it obedient to Christ. It should be noted that it took a long time before such ideas became popularized in the superstitious minds of Europe whose attitudes ran in strong opposition to that of the early Christians. Neither the Romans nor the pagans in the rest of Europe were particularly inclined toward virtue. The free attitudes of the common people in Europe and the lack of internal bounds on unwanted mental states and behaviors persisted all the way until the Renaissance. However, during the 16th century, as Protestants and Catholics warred with one another, Satan grew in stature. The grand status of the devil as responsible for all evils in the world did not fully emerge until the mid-16th century, and both Catholics and Protestants believed they could see the internal abyss opening up before their eyes and the devil seizing every opportunity to invade their being (Muchembled, 2003, p. 109). Not surprisingly, at the same time, the term obsession came to refer quite literally to the hostile action of the devil or an evil spirit (see Harper, 2001). In other words, the term obsession came to be associated with a very literal type of egodystonic experience as well as forbidden and unwanted states of mind. However, over time, the literal image of the devil was supplanted by that of the devil within, a demonic inner spirit that lies in all people and that needed to be resisted (Muchembled, 2003). Increasingly, people were expected to confess not only their behavior but also their inner life. For example, confessing sexual misconduct had long been an important part of religious confession, but whereas confession was previously focused on sexual deeds alone, the range of confession became much wider with the inclusion of desires and fantasies (Foucault, 1980). Not surprisingly, unwanted mental states came to be taken very seriously by the devout, with some of these accounts being very reminiscent of obsessive thinking as we understand it today (e.g., Bunyan, 1998). As the church confirmed its monopoly on what constituted forbidden thoughts, the scope of unwanted states of mind widened. In particular, people during the Victorian age of suppression in the 19th century feared the involuntary revelation of the inner self to others. As McAdams (1993) wrote, While you might not be able to attain conscious insight into the deep secrets of your own mind, the Victorians believed, there was always the danger of revealing your hidden self to others, as objective observers might come to know you better than yourselves (p. 121). Interestingly, the story of Dr. Jekyll and Mr. Hyde was a grand success at the time, and the Victorian life of rectitude and responsibility required vigilance, lest the demons from within erupt violently or lustfully onto the public scene (p. 121). Hence, by the end of the 19th century, the notion of unwanted mental states going against the will and volition of the individual had firmly established itself in peoples minds without any direct reference to religion. The devil had gone underground, finding refuge in a divided self.


Aardema and OConnor

In light of previous historical developments, it is not surprising that early psychological theories of obsessions came to rely on notions of a divided self to explain the egodystonic nature of obsessions. One such theory was proposed by Dagonet (1870), who considered OCD to be a form of folie impulsive where irresistible violent impulses overcame the will of the person. The notion of a divided self, however, found its most systematic application in the writings of Freud, who relegated all that was negative to the dark unconscious. Ultimately, the mind was driven by unacceptable, unconscious sexual or aggressive impulses that had to be accommodated/transformed/suppressed by the conscious mind. In fact, even if our mental states revealed no particular negative content, such an observation was meaningless since there was an unconscious boiling underneath, casting its shadow on innocent thought. Not surprisingly, there is an uncanny resemblance between Freudian psychic architecture (i.e., superego, ego, id) and the various levels of the afterlife (i.e., heaven, purgatory, hell) portrayed by platforms and ladders in medieval Mummers Plays, with devils and demons running up and down, an allusion not lost on the Victorian mind. The Freudian model of mental illness continued to be highly influential until well past the second half of the 20th century. However, around the end of the 1960s, new models of psychological disorders started to emerge, based initially on behavioral conceptions and, later, cognitive-behavioral models of mind.


As much as we may feel we have escaped the superstitious leanings of the 17th century, they still structure our thinking in insidious ways. In particular, current cognitive models of OCD still hold the view that intrusions, or obsessions, represent an aspect of the self that can never be totally avoided. Although there is general consensus among theorists that people with OCD virtually never act out their obsessions, the obsessions or intrusions related to particular aversive and depraved acts are still considered to occur as part of a normal stream of consciousness. Indeed, some have explicitly referred to obsessive intrusions as resulting from the spirit of the perverse and as reflecting a universal tendency to engage in bad thoughts (Baer, 2001). However, at the same time, this spirit of the perverse is considered to be a normal part of us that simply exists as a fact of life. Hence, a part of the current appraisal solution to the occurrence of these thoughts sometimes consists of embracing and accepting the thought and at other times of encouraging people with OCD to no longer give these thoughts any significance and importance. The claim of most modern cognitive models of OCD that obsessions find their origin in intrusive cognitions ultimately resembles earlier historical movements that considered obsessions the result of devilish influences, the devil within, or a divided self. The term intrusive implies a thought coming from elsewhere, as if it were a spontaneous, noninferential phenomenon for which there is no rational explanation. Granted, the origin of obsessions has been located within the self but outside of the realm of the reasoning self by viewing them as the result of automatic processes (Salkovskis, 1989) or as mental flotsam (Rachman, 1993). Of course, references to an internal or external devil are no longer given any explanatory value in contemporary works on OCD except in metaphor. However, in all cases, the mind is viewed as subject to unwanted influences that, even though recognized as coming from within, are ultimately outside the persons control. Yet are intrusive cognitions a given phenomena, occurring as if they were external, neutral events that require no further explanation? Or is it time to change metaphors and investigate the structure of intrusions? Because several studies have shown intrusive cognitions to be similar in content in both OCD and the general population, suggesting that intrusions are a universal, normal phenomenon (Rachman & De Silva, 1978; Salkovskis & Harrison, 1984; for a review, see Julien, OConnor, & Aardema, 2007), any intrinsic value attached to the content of the thought has been dismissed

The Menace Within


from the equation. Yet if intrusions do not arrive haphazardly, independently of internal and external triggers, then is the term intrusion an adequate or even accurate one for obsessions? As noted by OConnor (2002),
initial thoughts of God, of sex, of violence may be internal percepts reflective of a current religious, sexual or violent attitude, equivalent to the environmental trigger of seeing a door on leaving the house. Also, the client may mix up percept with reflection on the percept and report the obsessional thoughts as being about God or sex rather than a specific doubt concerning what thoughts of God or sex might imply. In both cases, the intrusion develops subsequent to the registration of a percept and as such may be seen as part of a continued reflection on the percept. So, is it accurate even in pure obsession to speak of the thought intruding? Does the thought intrude into the stream of consciousness or is it a continuation of information processing by other means? (p. 40)

An inference-based approach to OCD (see OConnor, Aardema, & Plissier, 2005) argues that an obsession is an inference in continuity with current internal or external events (e.g., perhaps the oven is left on, maybe my hands are dirty). Further, it may frequently be an inference of doubt, which, unlike intrusions, is already emotionally charged and leads to a spiraling chain of negative secondary possibilities. In the case of obsessions with overt compulsions, the inference is logically implied in the thoughtaction sequence of obsessional behavior (Clark & OConnor, 2004). If Im scrubbing a toilet clean, then there must be a thought or inference concerning the possibility of the toilet being unclean that motivates the cleaning. This possibility is not merely a passing thought whose meaning is derived from its appraisal; rather, the inferences or the initial intrusive ideas are thematic and person specific (OConnor, 2002), and regardless of the normal content of intrusions, in the obsessional case, their arrival on the scene may very well be abnormal. Indeed, recent findings suggest the content of obsessional intrusions may be generated in more inappropriate contexts than the intrusions of a non-OCD population (OConnor, Julien, & Aardema, 2006). Likewise, experimental research indicates that people with OCD tend to doubt more while relying on more remote information when reaching an inference about a possible state of affairs (Plissier & OConnor, 2002a, 2002b). In these circumstances, the generation of obsessional inferences constitutes an inferential confusion where the person mistakes a remote hypothetical possibility for a real probability (Aardema & OConnor, 2003; Aardema, OConnor, Emmelkamp, Marchand, & Todorov, 2005; OConnor & Robillard, 1995, 1999). In this model, there is no such phenomenon as an intrusion arriving noninferentially. Rather, obsessions are primary inferences that are the result of prior reasoning, which invests meaning into the obsessional doubt. Indeed, recent evidence suggests that such reasoning processes take precedence over the appraisal occurring in the aftermath of a primary inference (Aardema, OConnor, & Emmelkamp, 2006; Aardema, Radomsky, OConnor, & Julien, 2007) and that OCD should perhaps be reconceptualized as a belief disorder rather than an anxiety disorder (Aardema, Emmelkamp, & OConnor, 2005; Aardema, Kleijer, Trihey, OConnor, & Emmelkamp, 2006). Moreover, if the obsession or primary inference in OCD is inherently emotionally charged and meaningful to the person and therefore not requiring appraisal in order to be experienced as significant, then it must contain representational and evaluative dimensions as the result of prior reasoning.


Representational dimensions are clearly apparent in obsessions without overt compulsions where the obsessional concern revolves around a possible cognitive or mental state of affairs, as in the case of obsessions about blasphemy, aggressions, and sexuality (e.g., I might be a blasphemer, I might suddenly hit someone, or I might be a child molester). For example, the person sees


Aardema and OConnor

a knife and then thinks I could kill someone. Here, the unwanted thought takes the form of an evaluation that the person believes himself or herself to be capable of killing someone. However, not only do these inferences contain evaluative dimensions, but they also appear to be self-representational and metacognitive in nature. That is, the inferences in obsessions without overt compulsions naturally imply a self-referential state. In recent years, metacognitive models have provided accounts of the maintenance of anxiety disorders (e.g., Wells, 2000). Traditionally, metacognition refers to the notion of thoughts about ones own thoughts and has been defined as knowledge and cognition about cognitive phenomena (Flavell, 1979). In a wider definition, however, metacognition may refer to any mental state that includes awareness of influences on thinking, such as emotion, desires, moods, and motivations. In particular, the role of metacognition in the generation of intrusive cognitions has recently been elaborated on by Aardema and OConnor (2003). They argue that the person who experiences obsessions does not experience aberrant thoughts as part of the normal stream of consciousness with motivated intent; rather, he or she is mostly preoccupied with the possibility of experiencing that thought. Clearly, a person who is vigilant about the occurrences in his or her mind finds that these thoughts light up in his or her awareness because of the mere act of selfmonitoring, cognitive evasion, or suppression (see Wegner, 1989). However, a crucial distinction has to be made between blasphemy with motivated intent as it may occur in the normal stream of consciousness and thoughts that come about as part of a dysfunctional metacognitive representation of thought. The person may start out with an ill-informed inference (e.g., I might be the type of person who offends God), followed by metacognitive operations in order to banish possible negative thoughts from ones mind (e.g., I should not think bad thoughts) with the inadvertent effect of thinking about the very thing the person attempts to avoid (e.g., I just had bad thoughts). In reality, however, the person is merely imagining the possibility of having the thought rather than actually having the thought. The person attempts not to have thoughts they have not had, but trying not to have a possible thought automatically implies that the thought is possible. In keeping with the tradition of various fusions reported in OCD, this process has been parsimoniously termed thought-thought fusion (i.e., thinking about the possibility of a thought is like having the thought) (Aardema & OConnor, 2003; OConnor & Aardema, 2003). However, the term fusion is a metaphor and is more accurately described as a confusion and failure to distinguish between cognitive and metacognitive states of mind. This formulation does not imply that the person may not experience very vivid scenarios and images. For example, one of our clients experienced frequent aversive images of someone cutting of his fingers. On further investigation, however, these images still found their origin in the self-representational inference: I might have been tortured in a previous life. In fact, the client purposely invited these images into his mind in order to (psychically) find out who tortured him in that previous life. In other cases, the inference might be of a less schizotypal nature, such as I could hurt someone, which then might be followed by a more involuntary aversive imagined scenario as if these tendencies are actually a real aspect of the person. However, the imaginary scenario of harm following from the self-evaluation that I could harm someone is confused with an impulse of motivated intent. The tendency of people with pure obsessions to engage in metacognitive thought and control is often quite striking, and clinically this tendency can pose a tremendous challenge. For example, if one informs a client that his or her tendency to control and suppress these thoughts only increases their (possible) occurrence, the person with OCD may very well engage in obsessive attempts to not suppress a particular thought. Other researchers before us have noted the excessive control exercised by OCD patients in regulating their mental states (Purdon & Clark, 2002). Similarly, we argue that the metacognitive operation of thought control is an important factor in the development and maintenance of OCD. However, this thought control is not postulated to be a preexistent isolated trait where people with OCD are in greater need of control than

The Menace Within


others. Rather, the need for control arises from a failure to distinguish between metacognitive and cognitive states of mind. That is, the belief that the person is really experiencing a motivated impulse while in fact this is not the case leads to obvious problems with a sense of control since the person attempts to remove motivated thoughts that are not there. This failure in distinguishing between an actual thought and a thought about a thought can become a threat in itself for the person with OCD where he or she becomes the victim of his or her own mind and increasingly perceives his or her own thoughts as uncontrollable (Clark, 2004). For example, people with OCD who attempt to give less importance to the thoughts in order to lessen their impact and frequency may attempt to do so while fearing at the same time that their thoughts will probably go out of control again, and hence they set themselves up to experience a situation where the thoughts do go out of control by merely imagining the possibility. In extreme cases, this failure may manifest itself in reaction to practically every therapeutic intervention where attempts to establish change quickly becomes absorbed into the obsessional process. When we explained to one of our clients that the tendency to suppress and avoid possible thoughts was not advisable because it only increased thinking about their possible occurrence, he engaged in obsessive efforts to not think the (possible) thoughts. The implication of an inferencebased approach to obsessions remains, however, that the person is inferring a state that is not thereand that he or she misrepresents this state as being an accurate reflection of self.


The rational argument of Descartes (1985) that one cannot have false beliefs about ones mental states has been discredited in the past century (see James, 1950). Rosenthal (2000a, 2000b) has proposed the concept of higher-order thoughts (HOTs), which provides the leeway for us to hold discordant self-representations and misrepresent states even to the degree that we can become aware of mental states that are in fact not there. According to the HOT theory of consciousness, we become conscious of mental states by having thoughts about their being present (HOTs), and these (metacognitive) thoughts are extrinsic and distinct from the conscious states themselves. In fact, in having an inaccurate HOT, one is conscious of oneself as being in a particular state under an inaccurate description (Rosenthal, 2005, p. 133). In other words, HOTs and metacognitive judgments may very well go beyond the actual states we judge ourselves to be in, subjectively indistinguishable from ones actual mental states themselves. It has been shown that people literally invent mental states to explain their own behavior (Nisbett & Wilson, 1977; Wilson & Nisbett, 1978) and that a fundamental limit on selfknowledge is the inaccessibility of nonconscious mental processing to conscious awareness. As noted by Wilson and Dunn (2003), there is now considerable evidence that people have limited access to the reasons for their evaluations and that the process of generating reasons can have negative consequences. Interestingly, Brewin (1996) made a similar point for psychologists attempting to infer deeper-lying cognitive structures, which are not based on direct observation. These dissociations between nonconscious and conscious awareness highlights the limitations of our awareness in coming to know ourselves, and, as pointed out by McAdams (1993), we are left to infer the nature of these nonconscious processes by taking what we do know (our conscious mental states) and filling in the gaps about what we dont know (nonconscious states) by constructing a coherent narrative about ourselves. So what are the (observable) narratives by which obsessive-compulsives come to hold discordant selfrepresentations and mental states? Consider the following narrative leading the person to engage in frequent attempts not to think about the devil:


Aardema and OConnor

I sometimes feel angry and depressed, and these are not the right feelings to have. I do not want such negativity inside of myself. These feelings are a bad thing to have and who knows how deep they go. They are the kind of feelings that are responsible for the ills in the world. They are the kind of feelings that the devil would have all the time. Having these kind of feelings is like thinking of the devil. I dont want to have these feelings or ever think about the devil, because it is a negative thing to do. If I try as hard as I can to not think about the devil then its like getting rid of all negativity inside of me. (OConnor, Aardema, & Plissier, 2005, p. 120)

As can be seen in this narrative, an actual mental state (depression and anger) is rendered insignificant because of a category error or irrelevant association (confusing thinking of the devil with having negative feelings). It removes the person away from the self-as-is and leads to experiencing thoughts about the devil and blasphemous thought while in fact these thoughts about the devil solely consist of attempts not to think of the devil. The narrative accounts for the obsession itself as it brings possible thoughts about the devil into focus. It is not a static appraisal of what it might mean to think about the devil that generates distress; rather, the obsession is inherently charged as the result of a narrative that brings possible thoughts about the devil into focus. Typically, obsessional narratives contain a variety of reasoning devices (see OConnor, Aardema, & Plissier, 2005, p. 117), giving credibility to the misrepresentation of self (e.g., I might be a crazy person), which subsequently leads to absorption in feared states of mind associated with negative self-representation. Typical rhetorical devices include category errors (e.g., I heard people who meditate have seen demons. Therefore, if I concentrate too hard I may be subject to evil influences, since meditating is about concentrating), purely imaginary sequences (e.g., I see myself cutting of all my fingers as if I could really do it), dismissal of evidence going against the inference (e.g., Even though I never hurt someone, their obsessions were never really the same, which means I could still be dangerous), and inverse inference (e.g., Psychopaths never feel guilty. Therefore, if I dont feel guilty, I may hurt my children). These rhetorical devices in OCD justifying the primary inference are always remote from reality and lead the person to arrive at an obsessional inference (of self) on a purely subjective basis while the person unwittingly confuses reality with possibility. Consequently, the person then acts as if the possible were a real probability (i.e., absorption into the inference I might be ) and becomes immersed into the feared states of mind that are associated with the misrepresentation of self by imagining the presence of blasphemous thoughts, aggressive impulses, and forbidden sexual thoughts. For example,
I cant go too far from home, or the city that I live in, because I dont know how panicky I might get. I might go really crazy and do something to myself. Who knows what is really wrong with me. One of my family members has schizophrenia, and I might have some serious disturbance also. I could be crazy enough to cut out my tongue. When Im anxious, I can see myself doing it, using a knife to cut it off. It feels I could actually do it. Then Ill be in the middle of nowhere without help. Then when Im found Ill be send off to a psychiatric hospital somewhere that I dont know, and theyll lock me up. (OConnor, Aardema, & Plissier, 2005, p. 121)

The crossover point where the person moves from reality into the imagination, often experienced as subjectively independent from conscious inference, is characterized by a complete, sudden shift into the obsessional world with varying amounts of dissociation (and hence experienced as intrusive) (OConnor & Aardema, 2003, 2005). However, the crossover point into the imagination is informed not by objective criteria but by a persuasive narrative leading the person to treat a possibility as a reality. Consequently, the intrusive phenomena characterizing OCD cannot be considered part of the self or even part of a divided self; rather, they are part of an imagined self.

The Menace Within



Obsessions often deal with threat in one way or another, and this article discusses the menacing character of obsessions that come from within. While the importance of threat in OCD has been noted before, it is generally considered not a unique characteristic of OCD but rather a general vulnerability factor for anxiety disorders. As noted by Aardema and OConnor (2003), a more specific theme in the narratives of people with OCD revolving around blasphemy, sexuality, and aggression appears to be a fear of who they could be or might become. In other words, people with blasphemous, sexual, and aggressive obsessions seem strongly invested in a sense of self-as-could-be as opposed to a sense of self-as-is (see Figure 1). A fear of self in OCD has recently found some empirical support in a study by Ferrier and Brewin (2005), who found that people with OCD have a tendency to make more negative inferences about themselves (on the basis of their intrusions). Similarly, Rachman (1997) has noted that people with OCD fear the significance of intrusive cognitions as to what they may reveal of their character. However, in terms of the inference-based model, it would be argued that a distrust of self provides the background context for the initial obsessional inference or intrusion to occur rather than merely being an appraisal after the fact. The person starts out with a figure ground relationship heavily geared toward experiencing the self and its associated mental states according to what it could be rather than experiencing the self-as-is. This may be the result of pathological imaginative processes, learning experiences, or other developmental factors that led the person to underdevelop his or her sense of self-as-is and define him- or herself strongly by what he or she could be or might become. An ambivalent sense of self (see Bahr, 2004), sensitive self-domains (see Doron, Kyrios, & Moulding, 2007), and sociotropy (Beck, 1983) may all relate to dysfunctional figureground relationships of the self in OCD. In particular, excessive concern with interpersonal relationships combined with a tendency to sacrifice personal needs and autonomy could potentially lead to an underinvestment in selfas-is (see Careau, OConnor, Freeston, & Turgeon, 2007). The result could be acquired self-doubt, lack of self-confidence, excessive self-monitoring, and a distrust of the self-as-is. Observable behavior and past behavior (which would safeguard the person to some extent from obsessional inferences about imagined harm) is given limited importance as a reference source in defining the self-as-is. This general self-doubt and lack of
A. Normal Sense of Self B. Obsessional Sense of Self

SELF-AS-COULD-BE Back ground

SELF-AS-IS Back ground

SELF-AS-IS Foreground


FIGURE 1. Normal and obsessional relation between self-as-is and self-as-could-be.


Aardema and OConnor

confidence may feed into a distrust of self, which provides a fruitful ground for a wide variety of idiosyncratic obsessional narratives giving rise to imagined mental content (i.e. intrusions). Since the person is unaware of the falsehood of the obsession, he or she will engage in attempts to correct or safeguard the self. Clinically speaking, these attempts may manifest themselves in a tendency by some OCD patients to seek out and elaborate on their obsessions in a visual format. Instead of engaging in obvious attempts to avoid the thoughts, the person may repeatedly play out the entire obsessional scenario in his or her head while at the same time being extremely distressed by it. For example, the client with obsessions about cutting of his fingers was unable to find peace until he had mentally proceeded to cut off all fingers in an attempt to find out where the thoughts came from while, of course, not realizing that he was actively creating and imagining the scenario. Another client felt the compulsive urge to complete a scenario where he tortured his children in an (obsessive) attempt to face his fears. Yet another man, with fears of being homosexual, engaged in external tests by rubbing his penis in the presence of other men in order to establish whether he was attracted to them, thus reinforcing the reality of his fear. This self-exposure functions as a reinforcer of symptoms rather than alleviating anxiety. In fact, such self-exposure is the opposite of exposure in vivo and response prevention in a clinical setting since the tests function as a compulsive strategy to find resolution to the thoughts. An actual exposure regimen may be difficult to implement in these particular clients and, as noted by Rachman (1971), should include an awareness of covert neutralizing strategies maintaining the anxiety. Additionally, we would propose addressing the initial obsessional inference providing the driving force behind the compulsive self-tests and strategies. The final inference-based model of self-representation for obsessions of blasphemy, sexuality, and aggression is schematically presented in Figure 2.

We started this article with the dilemma of subjective and objective discordance in OCD, which led us into a historical analysis of obsessions and evaluative dimensions in intrusions or inferences. We suggested that obsessions involved the (mis)representation of mental states and that narratives and reasoning devices in obsessional thinking draw on imaginative processes leading to the imposition of remote possibilities onto reality and eventually into an obsessional construction of a remote self. But how does all this inform the egodystonic nature of obsessions? Cognitive models that locate the origin of obsessions in intrusive cognitions are not very clear about the issue of objective discordance in obsessions. While such models hold that people with OCD virtually never act on their obsessions, these obsessions are nevertheless to be considered real (albeit insignificant) aspects of the person since, after all, intrusive cognitions do occur. However, if the obsession (or intrusion) is the product of a faulty self-evaluation, then intrusions are not normal occurrences requiring appraisal in order to be experienced as significant. Rather, the obsession or intrusion is already emotionally charged and informed by a faulty self-theme. People may, of course, differ in terms of subjective discordance and the degree to which they hold their own obsessions to be realistic, which touches on the issue of overlap of OCD with delusional disorder. Not all obsessions are experienced as entirely egodystonic, and some individuals and subtypes (e.g., hoarders) may not consider the obsession to be entirely discrepant with their identity. Yet, whether or not people with OCD consider their obsessions to be realistic, they are always objectively discordant with the persons actual self because the obsession is the result of a false self-evaluation. Hence, the egodystonic experience of obsessions is an illusionary sense of discordance that results from the incongruence between the self-as-is and an imagined self-as-could-be. It is illusionary because the experience of discordance between the person and the obsession is based on the fear of a nonexistent possible self.

The Menace Within


SELF-AS-IS Background

Lack of confidence or self -doubt


Overinvestment in experiencing the self-as-could-be

Excessive self -monitoring

Constructed narrative and inference of what could be supplied by rhetorical devices

Reliance on outside irrelevant givens at the expense of true knowledge of self

Distrust of self as is

Self absorption in imaginary possibilities (intrusions )

Confusing farfetched events with actual events

Illusionary sense of discordance between self as -is and possible self (egodystonicity)

Distress and compulsive attempts to correct or safeguard self

FIGURE 2. Schematic representation of an inference-based approach to obsessions about blasphemy, sexuality, and aggression.

While indeed obsessions are more than just an intense mental state or excessive preoccupation, it appears likely that the intensity of obsessions also play a key role in the extent to which a mental state is experienced as subjectively discordant. For example, the phrase I am overcome with grief refers to the intense emotion as if it is an outside force stronger than the person. While historically the notion of inexplicably intense mental states may have been attributed to devilish influences and later came to refer to the devil within, an inference-based approach explains intensity in terms of degree of absorption in a faulty inferential confusion process where the person experiences the obsession as if it were a reality (or as if part of the self-as-is) while it resides only in their imagination.


Aardema and OConnor

Appraisal models of OCD have so far been unwilling to consider obsessions or intrusions as the result of a background context of more permanent cognitive structures; rather, they view these thoughts as relatively neutral isolated events to be separated from their consequent subjective appraisal. This is not surprising since the appraisal model of OCD found its origin in Becks model that emphasizes the appraisal and significance placed on objective, external events. However, the objectification of intrusive phenomena as if they are neutral and objective events has led to incoherence in cognitive accounts of OCD. Clearly, we are indebted to the original clinical observations of writers who have noticed the excessive self-scrutiny in OCD (Rachman & Hodgson, 1980). Yet it may not be any particular cognitive belief domain that causes OCD; rather, any theme that exemplifies a threat from within and makes ones actual mental content suspect could make a person more vulnerable to OCD. This distrust of self is informed by inherent figureground relationships in the construction of self where the person with OCD is overinvested in who they could be or might become. The result is a psychopathology of self-doubt, and in that lack of certainty, all seems possible and plausible. The inference-based approach to OCD anchors the occurrence of intrusive cognitions in a reasoning-based model where imagination plays a decisive role in rendering remote possibilities more probable. In the case of obsessions about blasphemy, sexuality, and aggression, the OCD client experiences imaginary mental events that inappropriately become the focus of selfscrutiny and neutralization behaviors. We gave an account of how such beliefs in nonexistent states emerge and how they should be differentiated from a conceptualization of intrusive thoughts that require no further explanation. That is, the context in which the person with OCD comes to infer the presence of unwanted inner states is quite different from the decontextualized occurrence presently ascribed to intrusive cognitions. The person with OCD doubts in spite of the presence of conflicting sense data (i.e., his or her actual behavior and self-as-is) and infers this possibility on the basis of irrelevant information. Hence, clinically speaking, an inference-based approach would not advise the person to tolerate the uncertainty regarding, for example, a clients obsessional doubts about his or her sexual orientation. Rather, since it is the inference of nonexistent mental states that characterizes obsessional inferences, it would explore the subjective reasoning leading up to these inferences, which render the person with OCD unable to dismiss these thoughts as irrelevant or insignificant occurrences. Hence, asking the person in therapy to dismiss these thoughts would be not be on the basis of the idea that these are normal thoughts and occur to everyone but rather on the basis that they are objectively discordant with their actual states. And this is a core difference between inference- and appraisal- or acceptance-based approaches to OCD. We have outlined some historical and societal influences that exemplify the theme of a threat from within to elucidate the present state of affairs in cognitive thinking on OCD, but it is unlikely that any of these factors bear any direct relationship to the development of OCD symptoms. While dogmatic religion may at times feed into a strong distrust to self and has been found to be associated with certain selective OCD symptoms (Sico, Novara, & Sanavio, 2002), it is unlikely to be a cause of OCD. Besides, a distrust of the self has become part of our thinking in a wide variety of nonreligious narratives, and people with a secular orientation are not immune to obsessions (except, perhaps, for blasphemous obsessions). Cultural differences likely apply as well where a distrust of self is informed by a different context and background than in Western societies. Hence, it would be a mistake to challenge the persons values in therapy since the target is not values but rather the process by which the person comes to infer the existence of nonexistent mental states. If the obsession is tied with moral scruples, the therapist might usefully question why one religious person experiences OCD symptoms while another does not. In our clinical experience,

The Menace Within


the answer to that question lies in the reasoning processes that lead the person into a faulty obsessional inference rather than any general belief. The question to pose is why the person trusts himor herself in one situation (e.g., that he or she is not a bank robber) and not another (e.g., that he or she might be a child molester), thus exposing the vulnerable self-theme in the incongruence between both scenarios. The reasoning devices that lead the person to believe in an obsessional state of affairs are antagonistic to promoting a more reality-based trust of self. Therapy would rely primarily on psychoeducation and how the obsession or intrusion stems from a feared self that is not part of the person (not even as a given reality and therefore ultimately requiring acceptance). This approach has already found a fruitful application in many obsessions with overt compulsions where the person distrusts his or her own senses (OConnor, Aardema, Bouthillier, et al., 2005). It stands to reason that an increased trust of self-as-is would be equally beneficial for a person suffering from blasphemous, sexual, and aggressive obsessions.

Aardema, F., Emmelkamp, P. M. G., & OConnor, K. (2005). Inferential confusion, cognitive change and treatment outcome in obsessive-compulsive disorder. Clinical Psychology and Psychotherapy, 12, 337345. Aardema, F., Kleijer, T., Trihey, M. OConnor, K., & Emmelkamp, P. M. G. (2006). Processes of inference, schizotypal thinking, and obsessive-compulsive behavior in a normal sample. Psychological Reports, 99, 213220. Aardema, F., & OConnor, K. (2003). Seeing white bears that are not there: Inference processes in obsessions. Journal of Cognitive Psychotherapy, 17, 2337. Aardema, F., OConnor, K., & Emmelkamp, P. (2006). Inferential confusion and obsessive beliefs in obsessive-compulsive disorder. Cognitive Behaviour Therapy, 35, 138137. Aardema, F., OConnor, K., Emmelkamp, P. M. G., Marchand, A., & Todorov, C. (2005). Inferential confusion and obsessive-compulsive disorder: The Inferential Confusion Questionnaire. Behaviour Research and Therapy, 43, 293308. Aardema, F., Radomsky, A., OConnor, K. P., & Julien, D. (2007). Inferential confusion, obsessive beliefs and obsessive-compulsive symptoms: A multidimensional investigation of independent cognitive domains. Manuscript submitted for publication. American Psychiatric Association. (2000). Diagnostic and statistical manual of manual disorders (4th ed., Text Revision). Washington, DC: Author. Baer, L. (2001). The imp of the mind: Exploring the silent epidemic of obsessive bad thoughts. New York: Dutton. Bahr, S. S. (2004). Self-ambivalence in obsessive-compulsive disorder. Melbourne: University of Melbourne, Department of Psychology. Beck, A. T. (1983). Cognitive therapy of depression: New perspectives. In P. Clayton & J. E. Barrett (Eds.), Treatment of depression: Old controversies and new approaches (pp. 265290). New York: Raven Press. Benner, J. A. (2006). Hebrew thought. Retrieved April 15, 2006, from www. ancient_hebrew. org/12_thought. html Brewin, C. R. (1996). Theoretical foundations of cognitive-behavior therapy for anxiety and depression. Annual Review of Psychology, 47, 3357. Bunyan, J. (1998). Grace abounding to the chief of sinners: John Bunyans autobiography. Greenville: Emerald House Group. (Original work published 1666) Careau, Y., OConnor, K. P., Freeston, M. H., & Turgeon, L. (2007). Childhood experiences and adult beliefs in OCD: Testing a specific and a general aetiological model. Manuscript submitted for publication. Clark, D. A. (2004). Cognitive-behavioural therapy for OCD. New York: Guilford Press.


Aardema and OConnor

Clark, D. A., & OConnor, K. (2004). Thinking is believing: Ego-dystonic intrusive thoughts in obsessive-compulsive disorder. In D. A. Clark (Ed.), Unwanted intrusive thoughts in clinical disorders (pp. 145174). New York: Guilford Press. Dagonet, H. (1870). Des impulsions dans la folie et de la folie impulsive. Annales mdico-psychologique, 4, 215259. Descartes, R. (1985). Meditations on first philosophy. In J. Cottingham, R. Stoothoff, & D. Murdoch (Trans.), The philosophical writings of Descartes (p. 150). Cambridge: Cambridge University Press. (Original work published 1641) Doron, G., Kyrios, M., & Moulding, R. (2007). Sensitive domains of self-concept in obsessive-compulsive disorder (OCD): Further evidence for a multidimensional model of OCD. Journal of Anxiety Disorders, 21, 433444. Ferrier, S., & Brewin, C. R. (2005). Feared identity and obsessive-compulsive disorder. Behaviour Research and Therapy, 43, 13631374. Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive-developmental inquiry. American Psychologist, 34, 335348. Foucault, M. (1980). The history of sexuality. New York: Knopf. Grenier, S., & OConnor, K. P. (2007). Clinical assessment of obsessional beliefs: A three dimensional approach to measuring insight. Manuscript submitted for publication. Harper, D. (2001). Online etymology dictionary. Retrieved April 26, 2006, from James, W. (1950). The principles of psychology. New York: Dover. (Original work published 1890) Julien, D., OConnor, K. P., & Aardema, F. (2007). Intrusive thoughts, obsessions and appraisals in obsessivecompulsive disorder: A critical review. Clinical Psychology Review, 27, 336383. McAdams, D. P. (1993). The stories we live by: Personal myths and the making of self. New York: Guilford Press. Muchembled, R. (2003). A history of the devil: From the middle ages to the present. Cambridge: Polity Press. Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84, 231259. OConnor, K. P. (2002). Intrusions and inferences in obsessive-compulsive disorder. Clinical Psychology and Psychotherapy, 9, 3846. OConnor, K., & Aardema, F. (2003). Fusion or confusion in obsessive-compulsive disorder. Psychological Reports, 93, 227232. OConnor, K. P., & Aardema, F. (2005). The imagination: Cognitive, pre-cognitive and meta-cognitive aspects. Consciousness and Cognition, 14, 233256. OConnor, K., Aardema F., Bouthillier D., Fournier, S., Guay, S., Robillard, S., et al. (2005). Evaluation of an inference based approach to treating obsessive-compulsive disorder. Cognitive Behaviour Therapy, 43, 116. OConnor, K., Aardema, F., & Plissier, M. C. (2005). Beyond reasonable doubt: Reasoning processes in obsessive-compulsive and related disorders. Chichester: Wiley. OConnor, K., Julien, D., & Aardema, F. (2006, September). Content and context of intrusions in obsessional and control populations. Paper presented at the Annual Congress of the European Association for Behavioural and Cognitive Therapy, Paris. OConnor, K. P., & Robillard, S. (1995). Inference processes in obsessive-compulsive disorder: Some clinical observations. Behaviour Research and Therapy, 33, 887896. OConnor, K. P., & Robillard, S. (1999). A cognitive approach to modifying primary inferences in obsessivecompulsive disorder. Journal of Cognitive Psychotherapy, 13, 117. Plissier, M. C., & OConnor, K. P. (2002a). Deductive and inductive reasoning in obsessive-compulsive disorder. British Journal of Clinical Psychology, 41, 1527. Plissier, M. C., & OConnor, K. P. (2002b, September). The role of inductive reasoning in obsessional doubts. Paper presented at the Annual Congress of the European Association for Behavioural and Cognitive Therapy, Maastricht.

The Menace Within


Purdon, C., & Clark, D. A. (2002). Control of thoughts. In R. O. Frost & G. Steketee (Eds.), Cognitive approaches to obsessions and compulsions: Theory, assessment, and treatment (p. 2943). Oxford: Elsevier Science. Rachman, S. J. (1971). Obsessional ruminations. Behaviour Research and Therapy, 9, 229235. Rachman, S. (1993). Obsessions, responsibility and guilt. Behaviour Research and Therapy, 31, 149154. Rachman, S. (1997). A cognitive theory of obsessions. Behaviour Research and Therapy, 35, 793802. Rachman, S., & Hodgson, R. (1980). Obsessions and compulsions. Englewood Cliffs, NJ: Prentice Hall. Rachman, S. J., & DeSilva, P. (1978). Abnormal and normal obsessions. Behaviour Research and Therapy, 16, 233248. Rosenthal, D. M. (2000a). Consciousness, content, and metacognitive judgements. Consciousness and Cognition, 9, 231242. Rosenthal, D. M. (2000a). Metacognition and higher-order thoughts. Consciousness and Cognition, 9, 203214. Rosenthal, D. M. (2005) Consciousness, interpretation and higher-order-thoughts. In P. Giampieri-Deutsch (Ed.), Psychoanalysis as an empirical, interdisciplinary science: Collected papers on contemporary psychoanalytic research (p. 119142). Vienna: Verlag der sterreichischen Akademie der Wissenschaften (Austrian Academy of Sciences Press). Safran, R., Thordarson, D. S., & Rachman, S. (1996). Thought action fusion in obsessive-compulsive disorder. Journal of Anxiety Disorders, 5, 379391. Salkovskis, P. M. (1989). Cognitive-behavioural factors and the persistence of intrusive thoughts in obsessional problems. Behaviour Research and Therapy, 27, 677682. Salkovskis, P. M., & Harrison, J. (1984). Abnormal and normal obsessions: A replication. Behaviour Research and Therapy, 22, 549552. Sico, C., Novara, C., & Sanavio, E. (2002). Religiousness and obsessive-compulsive cognitions and symptoms in an Italian population. Behaviour Research and Therapy, 40, 813823. von Rad, G. (1966) Deuteronomy: A commentary. Philadelphia: Westminster Press. Wegner, D. M. (1989). White bears and other unwanted thoughts. New York: Guilford Press. Weiman-Kelman, L. (1999). The sixth commandment. In R. S. Mikva (Ed.), Broken tablets: Restoring the ten commandments and ourselves. Woodstock, VT: Jewish Lights Publishing. Weisman, M. M., Bland, R. C., Canino, G. J., Greenwald, S., Hwu, H. G., Lee, C. K., et al. (1994). The cross national epidemiology of obsessive-compulsive disorder. Journal of Clinical Psychiatry, 55, 510. Weissbard, D. R. (1997). Coveting. Retrieved April 22, 2006, from Wells, A. (2000). Emotional disorders and metacognition: Innovative cognitive therapy. Chichester: Wiley. Wilson, T. D., & Dunn, E. W. (2003). Self-knowledge: Its limits, value, and potential for improvement. Annual Review of Psychology, 55, 17.117.26. Wilson, T. D., & Nisbett, R. E. (1978). The accuracy of verbal reports about the effects of stimuli on evaluations and behaviour. Social Psychology, 41, 118131. Correspondence regarding this article should be directed to Frederick Aardema, Centre de Recherche FernandSeguin, 7331 Hochelaga, Montral, Qubec H1N 3V2, Canada. E-mail: