برونسازی و شخصی سازی تعصبات در هذیان های آزار و شکنجه: رابطه با بینش ضعیف و نظریه ذهن

The presence of externalizing bias (EB) for negative events together with personalizing bias (PB) (a bias to blame others rather than circumstances) may jointly constitute a vulnerability to develop persecutory delusions (PDs). Whereas EB purportedly serves to defend a vulnerable self-concept by avoiding negative self-attributions and might therefore exacerbate poor insight, PB may reflect cognitive deficits, including theory-of-mind impairment. We investigated these proposals in 34 schizophrenic patients with a history of PDs and 21 healthy controls. Patients with moderate- to severe-PDs and patients without a current PD showed excessive EB which was, surprisingly, absent in patients with mild persecutory delusions (mild-PDs). That EB might wax and wane with fluctuating delusional intensity was interpreted in accord with a new dynamic model of attribution self-representation cycles [Bentall et al., 2000. PDs: A review and theoretical integration. Clinical Psychology Review, 21, 1143–1192]. As predicted, EB exacerbated poor insight. However, counter to predictions, theory-of-mind impairment did not increase PB, which was marked in all participants, whether clinical or non-clinical; instead, theory-of-mind impairment was also correlated with poor insight. Our findings indicate multiple pathways to poor insight, one of which is a theory-of-mind difficulty, impairing the capacity to simulate other perspectives for the purpose of critically evaluating one's own beliefs and circumstances.

Advocates of a two-factor explanatory framework for delusions propose a distinction between the factors that trigger an initial implausible thought (and thus contribute towards explaining the thematic content of a particular delusion) and the factors that explain the uncritical adoption of an implausible thought as a delusional belief (Davies, Coltheart, Langdon, & Breen, 2001; Langdon & Coltheart, 2000). The recent Garety, Kuipers, Fowler, Freeman, and Bebbington (2001) model of psychosis is generally consistent with this approach. For example, these authors propose that the route to psychosis begins with a precipitating environmental and/or neurobiological event that disrupts cognitive processes1 leading to anomalous experiences. These anomalous experiences trigger a search for explanatory hypotheses (e.g. the experience of hearing voices prompts the initial thought “some invisible entity is speaking to me”) and the full psychotic break with reality occurs when sufferers generate and adopt as a delusional belief an externalizing misattribution (e.g. “God is speaking to me”) rather than self-correcting with an internal attribution (e.g. “something wrong with my mind is causing me to hear voices”). It has been proposed that an extreme self-serving attributional bias may contribute towards a failure of this type in persecutory-deluded (PD) people. The earliest version of this account was the defensive attribution model of persecutory delusions (PDs; Bentall, Kinderman, & Kaney, 1994). According to this account, people who are prone to develop PDs harbour latent negative self-beliefs and subconsciously defend against the activation of such beliefs by avoiding self-blame for negative events. The cost is an externalizing misattribution concerning the cause of negative experiences and a paranoid world-view of other people as perpetrators of harm. More recently, Bentall (2001) has noted that a defensive avoidance of self-blame for negative events is not sufficient to explain the generation of other-blaming delusions, since not blaming anyone, or simply putting events down to unknown circumstances or chance, would satisfy the same defensive function. Bentall et al. (Bentall et al., 2001; Kinderman & Bentall, 2000) have therefore proposed that two distinct attributional biases—an externalizing bias (EB) (blaming others or circumstances rather than oneself when negative events occur) and a personalizing bias (PB) (blaming others rather than circumstances)—need both be present to constitute a vulnerability to PDs. A substantial literature supports the role of an excessive EB in PD formation. For example, numerous studies have used the Attributional Style Questionnaire (ASQ: Peterson et al., 1982) to demonstrate that depressive individuals attribute the cause of negative events to themselves and the cause of positive events to other people, whereas PD people excessively externalize the cause of negative events (Candido & Romney, 1990; Kaney & Bentall, 1989; Lyon, Kaney, & Bentall, 1994). That a PB might also contribute towards PD formation has been supported more recently by studies that have used the Internal Personal and Situational Attributions Questionnaire (IPSAQ: Kinderman & Bentall, 1996). The IPSAQ is similar to the ASQ in distinguishing between self and non-self causes of positive and negative events but also distinguishes between non-self causes that attribute blame to other people vs. situational factors. For example, Kinderman and Bentall (1997) used the IPSAQ to demonstrate that paranoid people are more inclined than depressed and healthy controls (HCs) to blame other people, rather than circumstances, for negative events that have been attributed to non-self causes. The EB and the PB are conceived as qualitatively distinct attributional biases that contribute independently to the generation of PDs. Whereas the EB purportedly functions to defend a vulnerable self-concept by avoiding negative self-attributions, the PB may reflect cognitive deficits, including theory-of-mind impairment ( Bentall, 2001; Bentall et al., 2001; Kinderman & Bentall, 2000). The latter is the idea that blaming other people is the easy default and that a more effortful cognitive search is needed to generate situational explanations. Cognitive deficits, including theory-of-mind impairment, may compromise this cognitive search. Theory-of-mind refers to the capacity to infer mental states in order to explain and/or predict people's behaviour. Frith (1992) was the first to propose a link between theory-of-mind impairment and PDs. He suggested that paranoid schizophrenic patients, unlike autistic individuals, know that other people have minds, but have lost the full capacity to make appropriate inferences concerning the contents of other people's minds (see Frith, 2004; Langdon, 2003, for more on this distinction). This incapacity breeds suspicion about other people's hidden intentions and leads to unfounded beliefs that other people are conspiring against the patient. However, despite extensive evidence of performance deficits on theory-of-mind tasks in people with schizophrenia (Langdon, Coltheart, Ward, & Catts, 2001a; Pinkham, Penn, Perkins, & Lieberman, 2003), the findings are equivocal concerning a specific link between theory-of-mind impairment and PDs (Bentall et al., 2001; Harrington, Langdon, Siegert, & McClure, 2005; Langdon et al., 2001a). Bentall et al. (Bentall, 2001; Bentall et al., 2001; Kinderman & Bentall, 2000) have therefore proposed that theory-of-mind impairment might contribute towards PD formation via a slightly different mechanism to that proposed by Frith. On their alternative account, a difficulty with appreciating the contents of another person's mind, including a difficulty in understanding why that other person might act in a certain way, would make situational attributions relatively inaccessible. This impaired capacity to appreciate the situational determinants of another person's behaviour would then combine with the defensive avoidance of self-blame for negative events in order to cause the generation of other-blaming delusions. Support for a link between theory-of-mind impairment and PB comes from evidence that, amongst non-clinical subjects, those who perform poorly on a task requiring the inferring of mental states of characters in complex social situations are also more likely to blame other people rather than circumstances for negative events that have been attributed to non-self causes (Kinderman, Dunbar, & Bentall, 1998). The present study aimed to demonstrate that the EB and the PB are qualitatively distinct and that the joint presence of both biases is necessary to constitute a vulnerability to PDs. That these two biases are qualitatively distinct would be supported by evidence that each bias is associated with a distinctive factor: theory-of-mind impairment in the case of the PB and a lack of awareness of mental illness (or poor insight) in the case of the EB. That theory-of-mind deficits might contribute towards PB has already been discussed; that poor insight might be associated with EB accords with non-deficit accounts of poor insight. Insight is a multifaceted construct, incorporating awareness of mental illness, the capacity to recognize psychotic symptoms as abnormal, and medication compliance (David, 1990). A deficit approach to poor insight conceives of the underlying cause as cognitive impairment (e.g. executive deficits: Amador, Strauss, Yale, & Gorman, 1991); in other words, patients who lack insight simply lack the cognitive capacity to recognize their own mental illness. In support of this view is the evidence of significant correlations between insight scores and measures of IQ and neuropsychological function in people with schizophrenia ( David et al., 1995; Rossell, Coakes, Shapleske, Woodruff, & David, 2003). However, not all studies find significant correlations of this type and even when these are found the size of the correlation coefficients is generally small ( Lysaker, Lancaster, Davis, & Clements, 2003). The implication here is that deficit accounts of poor insight leave much to be explained. The alternative non-deficit approach conceives of poor insight as the result of psychological defence mechanisms ( Lysaker et al., 2003; Lysaker, Bryson, Marks, Greig, & Bell, 2004); in other words, patients with poor insight have the cognitive capacity to recognize their own mental illness but are subconsciously motivated to deny or avoid doing so because of the distress this will cause. In support of this alternative approach is the evidence that poor insight is associated with (a) a greater reliance on denial as a coping strategy ( Lysaker et al., 2003) and (b) a self-deceptive tendency to generate positively biased self-reports ( Lysaker et al., 2004; Moore, Cassidy, Carr, & O’Callaghan, 1999). Implicit in the non-deficit approach is the idea that poor insight serves a defensive function (i.e. patients are motivated to avoid the threat of endorsing negative self-concepts such as having a mental illness: Moore et al., 1999). The EB in people who are prone to develop PDs serves a similar defensive function—i.e. according to the defensive attribution model of PD formation, the EB is the product of a defensive avoidance of activating implicit negative self-schemata (Combs, Penn, & Matthews, 2003). As such, the defensive attribution model of PDs seems to predict that the negative experience of being diagnosed with a mental illness and all the stigma this entails should activate a defensive EB, if present in people who are prone to develop PDs, in order to avoid/deny the negative self-attribution that a patient's own psychological problems are contributing to their mental illness. To put it another way, according to the defensive attribution model of PDs and a non-deficit approach to poor insight, one might expect a defensive EB to exacerbate poor insight. Such a view is generally consistent with the views of Freeman and colleagues (Freeman & Garety, 2003; Freeman, Garety, Kuipers, Fowler, & Bebbington, 2002; Garety et al., 2001), who have referred to poor insight as a “second-order” externalizing misattribution motivated by a need to defend a vulnerable self-concept against the threat of accepting ownership of a stigmatizing mental illness. Preliminary support for a link between EB and poor insight comes from evidence that EB is associated with a difficulty in recognizing psychotic symptoms as abnormal in early psychosis patients, many of whom will be paranoid (Drake et al., 2004). Our approach was to compare the attributional styles of three groups: PD schizophrenic patients, persecutory-prone but not currently persecutory-deluded (non-PD) schizophrenic patients, and HCs. Past history of PDs was used as the initial indicator of a propensity to develop PDs. Levels of general paranoid ideation (i.e. not necessarily delusional ideation) were then assessed as part of the experimental protocol in order to corroborate that the patients with a history of PDs, although not currently PD, were still paranoid and thus prone to develop a PD. It was hypothesized that if EB and PB jointly constitute a vulnerability to develop PDs, the two persecutory-prone patient groups will show extreme biases of both types when compared to HCs. We further suspected that levels of bias (of both types) would be most extreme in patients with a current PD. It was also hypothesized that the EB would be associated with poor insight, whereas the PB would be associated with theory-of-mind impairment.

Nineteen patients with a SAPS rating of two or more for PDs were classified as PD.6 The remaining 15 were classified as non-PD; nine had no delusions and six had mild non-PDs (five with loss of boundary delusions and one with religious delusions). There were no group differences in years of age or formal education, or NART-estimated IQ, p’s>.30p’s>.30 (see Table 1). Levels of paranoia (assessed using the PS) were similar in the two clinical groups with both clinical groups being significantly more paranoid than HCs, p=.0005p=.0005. Levels of depression were also similar in the two clinical groups, and, although both clinical groups tended to be more depressed than HCs, this difference was non-significant, p>.10p>.10. The two clinical groups did not differ in age of illness onset, duration of illness or type of medication, p’s>.15p’s>.15. Delusions and hallucinations were more severe in the PD group, whereas ratings of thought disorder, bizarre behaviour and negative symptoms were similar in the two clinical groups.