From the Internal Lexicon to Delusional Belief

Max Coltheart

This is the first in a series of posts on the papers published in an issue of AVANT on Delusions. Here Max Coltheart summarises his paper  'From the Internal Lexicon to Delusional Belief'.

Ten years ago, in an article on the two-factor theory of delusion, I wrote: 

'Suppose that as we go about everyday life we use an internal model of the world (Gray 1995; Sokolov 1963) to continuously predict what we will experience next. These predictions will normally be fulfilled, but occasionally not: occasionally something not predicted by the internal model occurs. That event indicates that there is something wrong with the database of beliefs that the model uses to predict what will happen next in the world. So the database needs to be fixed (by modifying existing beliefs or adopting new ones) so that it becomes compatible with the unexpected event' (Coltheart 2005).

As I emphasize in my Avant article, this is the model of belief formation on which the two-factor theory of delusional belief has been based since its inception (Langdon and Coltheart 2000; Davies et al. 2002). The key point here is that in cases of delusion, the prediction error that leads to the adoption of some new belief – the delusional belief – occurs because of some defect of perceptual or affective processing: this defect is Factor 1 in the two-factor account. For example, people predict that when they encounter a person emotionally close to them – a spouse, for example – a response of the autonomic nervous system will occur (since that is what usually occurs when one encounters such a person). 

If a neuropsychological impairment is suffered which disconnects the face recognition system from the autonomic nervous system, this prediction will fail when the spouse is encountered. That prediction error prompts the idea 'That is not my wife, it is a stranger', since if the encountered person were indeed a stranger that would explain the absence of autonomic response. The belief here is the Capgras delusion. But the absence of autonomic response – Factor 1 - is not sufficient to cause Capgras delusion, since patients with ventromedial frontal damage also lack this autonomic response to familiar faces, yet are not delusional. So, the two-factor theorist argues, a second factor must also be present, a Factor 2, which is an impairment of the normal processes of belief formation – perhaps always associated with specific damage to right dorsolateral prefrontal cortex (Coltheart 2007).

We have proposed this type of analysis for a variety of monothematic delusions. For each delusion, we have:

(a) identified or proposed a particular neuropsychological impairment that would cause a prediction error that prompts the delusional idea;

(b) shown or argued that each impairment has been reported in patients who are not delusional;

(c) and argued that in all of these delusional conditions a Factor 2 which is impaired belief evaluation is also present which is responsible for the delusional idea being adopted as a delusional belief.

Hence two-factor theorists have always held the view that prediction error is a key component in the genesis of delusional belief, whilst arguing that this is not sufficient for the causation of such beliefs: two factors, not one, are necessary. Fletcher and Frith have agreed: 'two deficits are necessary to explain these delusions: a primary deficit (paralysis or memory loss) and a failure to suppress the implausible responses that result from this deficit. In the case of neurological patients, false beliefs seem to derive from the coincidence of damage in two locations, with the abnormal belief formation associated with damage to the prefrontal cortex' (2009: 50-1).

I know of no account of delusional belief other than the two-factor theory that has been successfully applied to the explanation of such a range of monothematic delusions.