Anti-foundationalism and social ontology : towards a realist sociology
My concern in this thesis is with the transcendental question concerning the condition of possibility for social science. I argue that for social scientific knowledge to obtain one must: (1) have a conception of knowledge formation as theoretically mediated and fallible; and (2), social scientific knowledge claims must be about an object of study which conceptualises social structure as an enablement as well as an external constraint upon agency. This means: (1) arguing for an anti-foundational epistemology, which avoids becoming truth-relativism, by being complemented with a metaphysical realist ontology (giving us the position of 'realist anti-foundationalism'); and (2), using a social realist meta-theory of emergent properties to explain how methodology (i. e. the construction of specific theories and empirical research) has a conceptually mediated and fallible access to social reality. Developing a critical (i. e. transcendental) examination of the presuppositions of social scientific knowledge also means, afortiori, using realism as an underlabourer. The negative underlabouring role is to proscribe theories based on some form of epistemic immediacy, or being-knowing identity. It therefore means rejecting positivist, empiricist and essentialist versions of social science. The form of essentialism dealt with is called the sociological logic of immediacy, and this pertains to definitive ontologies of social structures or human being. Whereas the use of positivist and empiricist epistemology as a positive underlabourer produces a methodology that conflates the real into the 'actual' (i.e. decontextualised empirical 'facts'), the use of an essentialist ontology makes methodology either redundant (as the ontology mirrors all the essential properties which determine human behaviour), or an exercise in arbitrary verificationism. Against this, realist anti-foundationalism can act as apositive underlabourer for the social sciences if it is complemented by a social realist ontology of emergent properties, to act as a metatheory which guides methodology. In developing this argument my chief concern is to show that realism (as developed by Archer and Bhaskar) is a more adequate position than post-Wittgensteinian positions which focus on 'practices' and how people 'go on' in 'forms of life'. 'Adequacy' in this sense pertains to epistemological discussions about the status of knowledge, together with ramifications of post-Wittgensteinianism for knowledge of the socio-political realm. This means providing a critique of Rorty and Giddens, after dealing with the issue of empiricism. Although Rorty's critique of 'postmodernism' as essentialist is accepted. Whereas realism can explain how we have a conceptually mediated and fallible knowledge of reality, including social reality, post-Wittgensteinian positions fall into truth-relativism and essentialist conceptions of human being and social structures.