New versions of the coherence theory: Gadamer, Davidson, Foucault, and Putnam

This thesis explores four contemporary accounts of knowledge and understanding that highlight the importance of coherence as a major factor in justification, as a criterion of truth, and, in some cases, as a definition and ontology of truth. It seeks to explicate how these accounts advance the debate over coherentist epistemology by, principally, offering a new ontology of truth which demonstrates why coherence, if used as the criterion of truth, is a reliable means of achieving truth.;Michel Foucault's discursive theory of knowledge is then explored as an historical exemplification of Ian Hacking's "styles of reasoning." Foucault is found to hold, like Gadamer, a coherentist ontology of truth in his claim that discourses create the objects of which they speak. But Foucault's ontology is shown to be significantly different from Gadamer's on the question of the epistemic irreducibility of the subject's input to knowledge.;Next a summary of the debate over coherentist epistemology is given, with a focus on two criticisms previous coherentists could not sufficiently answer: (1) why coherence as a criterion of knowledge is truth-conducive, and (2) how a coherentist epistemology can avoid entailing a radical epistemological relativism. It is argued that the new coherentist ontology of truth developed by Gadamer and Foucault, and supplemented by Hilary Putnam's concept of internal realism, provides a new and stronger answer to (1).;Hans-Georg Gadamer's philosophical hermeneutics is shown to provide two arguments in defense of a coherence theory of justification. Further, Gadamer's ontology of inquiry is interpreted as a new coherentist conception of the ontology of truth. Donald Davidson's methodological arguments for a coherence theory of justification are then explored and shown to be similar to Gadamer's arguments. Davidson, however, retains a correspondence ontology of truth, and therefore must establish that as a criterion of truth coherence is truth-conducive. Davidson's agruments to this effect are evaluated and found to be weak.