There was an interesting post over at Certain Doubts a while back by Michael Huemer on holistic coherentism. Huemer argues that holistic coherentism doesn’t offer a genuine fifth option to the regress problem. Thus, coherentism in all its varieties is saddled with having to endorse circular reasoning (a very bad thing to have to endorse!). The regress problem concerns the structure of reasons. To put it abstractly, suppose the reason for p is q. Take the ancestral of the reason of relation. (The ancestral of a relation, R, is a relation, call it R*, such that if there’s an R-chain from x to z then x stands in R* to z. For instance, ancestor of is the ancestral of parent of.). Either the ancestral of reason of for q includes (i) some reason such that nothing stands in the reason of relation to it [Foundationalism], (ii) some reason such that it stands in the ancestral of the reason of relation to itself [Coherentism], (iii) there are a (potentially) infinite number of reasons [Infinitism], or (iv) nothing at all [Skepticism].
Huemer argues that in end a holistic coherentist ends up endorsing (ii). I think, however, that this isn’t right. The holistic coherentism denies the account of reasons operative in the regress argument. In particular, the holistic coherentism denies that the “reason of” relation is a singular relation. Let me explain. (more…)