Abstract

The nature of self has been discussed for centuries, with myriad theories specifying propositions of the form ‘The self is X’. Recently, psychology and neuroscience have added further such propositions and have sought to specify neural correlates for X. In this thesis, theories leading to all such propositions are subjected to methodological criticism. Specifically targeted are those theories that construct metaphysical, essentialist propositions on the nature of the self, and all other abstract concepts, more generally. On this point, it is concluded that theories of this type would benefit from taking into account the nature of language and the role it plays in the development of a theory. Theories that fail to consider language, run the risk of producing theoretical work affected by linguistic biases, those inherent to the language faculty. In this thesis, the biases of interest are referred to as ‘noun phrase reification’ and ‘clausal reification’, and an awareness of these is important for they can subvert the meanings of propositions, rendering them, and the theories built upon them, true by definition. In consideration of this methodological critique, a new theoretical approach to the self and to other abstract concepts is argued for. This new way of theorising combines aspects of basic scientific methodology with statistical modelling and linguistic theory, in which, the triangulation between two or more people, engaged in the act of naming objects and behaviours plays a formative role. Building on this new method in the context of the self, two experiments are performed using the electroencephalography neuroimaging technique; first in neurotypical adults, then in patients with a diagnosis of schizophrenia, a disorder commonly associated with an attenuated selfhood. It is hoped that this body of theoretical and empirical work will facilitate and catalyse progress on abstract concepts and their associated problems.