Progress in understanding and remediating disorders of thought has been hampered by the lack of an actual theory of thought. Neurocognitive variables such as memory, executive functioning or ‘theory of mind’ have formed the bread and butter of clinical testing in cognitive neuropsychiatry. However, several decades of studies have persistently revealed difficulties in the attempt to causally link deficits in such cognitive domains to clinical symptoms such as formal thought disorder or delusions. Are we missing out on an important element in the human thought system that is critical to normal neurocognitive functioning and potentially explanatory for some of its forms of decline?
In this presentation I firstly review some formal aspects of referentiality as an inherent aspect of all human language use. Referentiality empirically and cross-linguistically exhibits a number of forms ranging from maximally generic and indefinite to definite, rigid, deictic (indexical), and finally personal (1st or 2nd) ones. These forms are hierarchically ordered in the sense that an indefinite form such as a green car functions referentially through the mediation of a lexical description (being a CAR and GREEN), while no such description appears in forms of reference such as that or it. Present evidence supports, as a heuristic hypothesis, that (i) all of these forms of reference bi-uniquely correspond to specific grammatical configurations, (ii) grammatical organization in language never corresponds to anything other than specific forms of reference in one of three domains (the nominal, verbal, or clausal), (iii) lexical organization of meaning as such exhibits none of these forms, and (iv) these forms directly correspond to the fact that all meaning in language has a formal ontology in the sense that whenever we refer, the referents are formally individuated as properties, masses, objects, persons, events, propositions, facts, etc. I argue that if grammar is the organizational and cognitive principle behind these forms, and the lexicon (semantic memory) feeds in lexicalized concepts (categories), it is not clear what else is needed for a theory of thought, making an independent category of thought (and a ‘language of thought’) potentially redundant. As language is intrinsically social and shared, and social cognition in humans is intrinsically linguistic, it is not clear that ‘social cognition’ can be an independent variable either.
This makes the prediction that disorders of thought should have specific grammatical correlates (insofar as thought of the relevant type is inherently referential). In the second part of the talk, I review evidence collected over several decades in the study of language in schizophrenia and autism, as well as recent evidence from our own lab, to test this prediction. The evidence suggests that there is a clinical dimension to the hierarchy of reference above, insofar as the higher regions of the hierarchy (definites and above) are more impaired across disorders than the lower ones, with grammatical Person in particular as a particularly vulnerable dimension in the cognitive and linguistic profiles of both autism and schizophrenia.