Thursday, 4 July 2013

Egocentric Representation: a Positive Dimension to Abnormal Self-Experience in Schizophrenia?

Gregory Yates
In schizophrenia, the basic experience of existing as a “self” – as a subject whose thoughts, beliefs and actions coincide with what is regarded to be “self” – is disturbed. This has been hypothesised as the distinct phenotypic core of schizophrenia (Sass and Parnas, 2003; Parnas et al., 2005) and the central psychopathological trait marker of psychotic vulnerability (Nelson et al., 2008; Parnas et al., 2005) 'The clinical symptoms come and go,' describes one schizophrenia patient, 'but this nothingness of self is permanently there.' (Kean, 2009)

Anomalous self-experiences (ASEs) provide much of the material for delusional beliefs formed in acute schizophrenia (Stanghellini, 2012). In line with my last post on this blog, I would like to consider whether ASEs might also provide a ‘secondary gain’ or epistemic benefit (Graham, 2013). I will examine one delusion frequently reported by schizophrenia patients, and highly suggestive of an ASE: the delusion of reference.
Delusions of reference are ‘beliefs that a wide variety of neutral events have special significance and refer to the individual personally’ (see Startup and Startup, 2005) A typical example is recounted by Kendler (1999), whose patient R recalls entering a stadium full of people and becoming convinced that he was being watched and applauded by all of them. What underlies a delusion such as this is egocentricity – an ASE entailing a felt “link” between the self and the world inhabited by it (Parnas et al., 2005

As an abnormality of subjective orientation (Mishara, 2005; Gallagher,1985), egocentric representation (ER) of the self/world relation can be contrasted to allocentric representation (AR), which utilises visual scene context, rather than the self, as the frame of reference for conceptualising the self/world relation (Mishara, 2005
Delusions of reference demonstrate that ER is capable of “fooling” schizophrenia patients into forming maladaptive beliefs. But what these notorious delusions perhaps obscure, as I will now show, is that ER can also provide unique protections against being “fooled”.
Granted, AR is an extremely practical mode of self-world experience. Without AR, schizophrenia patients experience significant difficulties in using complex visual context to guide action, which results in impaired performance on a variety of tasks, such as those involving perceptual closure of incomplete objects (Doniger et al., 2001) and determination of relative relationships between visual objects (Mishara, 2005)
But AR is not infallible. In tasks where visual context is misleading, AR is disadvantageous. Take the classic Müller-Lyer illusion, pictured below:


The horizontal bar appears longer when the arrowheads are pointing inwards, but in reality both bars are of identical length:

Overcoming the Müller-Lyer illusion requires “filtering out” misleading context provided by the arrowheads – a cognitive strategy that does not come naturally to many. Yet, to schizophrenia patients, whose egocentric self-experience precludes use of this sort of context in the first place, this is no difficulty: they excel at Müller-Lyer tasks relative to “normals” (Parnas, 1999) Indeed, they excel to a degree similar to patients with stroke lesions damaging the ventral pathway, which has led some clinicians to suggest increased reliance on dorsal-stream processing as a cause of ASEs in schizophrenia (Mishara, 2005).
The epistemic superiority of ER to AR on tasks such as the Müller-Lyer challenge can be taken as further evidence that (1) cognitive abnormalities in schizophrenia are context-dependent and, (2) there exists the possibility of a significant positive dimension to psychosis.


  1. as someone diagnosed and treated for a psychotic disorder i found the perspective of this article very interesting. often the negative stigma associated with psychosis overshadows anything about it which could be seen as positive!

  2. Great stuff. Would ER be tied into deficiencies in social interaction in schizophrenia, such as the failure of patients to include the relevant cues necessary for the listener to establish the sequential theme of their discourse? If so how can ER translate into a ‘secondary gain’ in the domain of language production?

  3. That's really fascinating. But I'm not sure solving simple optical illusions is a "significant positive" and haven't thought enough (=at all) about the issue before- what kind of major beneficial aspects might this lead to?

    1. I think we might be understanding the word 'significant' in slightly different ways. What I'm after here is something closer to statistical 'significance': this effect is unlikely to have arisen by chance, and the magnitude of the relationship is sufficient to provoke clinical interest.

      However, I understand what you're getting at. I think it's worth clarifying exactly the sense I'm using 'secondary gain' in. It's a term (and one I think is really useful!) used by Graham in 'The Disordered Mind' (2013). It refers to an advantage entailed by mental disorder that cannot be taken as a "compensation" for the mental disorder - that is to say, it could not be seen as a "fair trade". An analogy to somatic disorder might be the enhancement of other sensory modalities entailed by blindness. It's an advantage, but it's not what most would see as "worth it".

      So, what I'm not saying is that egocentric representation (ER), or indeed the 'epistemic urgency' I discussed in my first post here, is a 'primary gain'. It's difficult to imagine anyone who would voluntarily undergo psychosis in order to acquire these advantages.

      That's the frame in which I am discussing the positive dimension to psychosis. Though, I suppose I ought to add, what I am saying is somewhat stronger than most treatments of this question.

      I am not, for example, putting forward what I call a 'threshold' hypothesis about psychosis. Generally, these hypotheses say something like this: "a little X is adaptive, but the excess of X in psychosis is maladaptive." 'Threshold' hypotheses are often put forward in questions of schizophrenia and creativity - a little out-of-the-box thinking is creative, but the excess in psychosis leads to delusional thinking.

      Instead, I'm advancing a 'context-dependency' hypothesis: "an excess of X in psychosis is more likely to be maladaptive in many environments" This is because we can conceive of situations (perhaps even real ones!) where 'secondary gains' such as ER and 'epistemic urgency' are consistently rewarded. In this case, it may well become a 'primary gain'. But this is not likely in the society we currently live in.

      This is something I'll be writing about in more depth in the future. I hope at least some of that answers your question!

  4. Nice post, thank you! I wonder if your argument would change if you took into account the fact that the Müller-Lyer illusion is highly culture-dependent (Segall, Campbell, Herskovits 1966). Is it an epistemic advantage not to fall for an illusion that is caused by children growing up in houses featuring carpentered angles? However, this objection can perhaps be accommodated by your 'context-dependency' hypothesis. What do you think?


Comments are moderated.