Why coherence is not truth‐conducive

Tomoji Shogenji (1999) argues that beliefs that are more coherent need not thereby be more likely to be true. This would strongly suggest that coherence (in the sense of mutual support) is not truth conducive, i.e., that the traditional philosophical problem of whether coherence implies truth has been given a negative solution. But Shogenji resists this conclusion, maintaining that the notion of truth conduciveness is inadequately captured by the formula ‘more coherence implies a higher likelihood of truth’. Further, he goes on to show that coherence is truth conducive on another, supposedly more adequate, account of the problematic concept. In this paper I shall argue that while Shogenji’s observation of a lack of correlation between coherence and truth is correct, his reasons for not drawing the natural philosophical conclusion are not. Shogenji has drawn attention to the following problem of coherence and specificity. He asks us to imagine ‘an epistemically ultraconservative agent who only holds a few extremely unspecific beliefs say, some rocks are heavier than others; some animals sleep sometimes; and someone is humming some tune somewhere’ (1999: 342). In this case it is, Shogenji writes, ‘very likely that her beliefs are all truth even though they do not hang together’ (ibid.). Meanwhile ‘a huge collection of highly specific beliefs such as the entire body of medical science almost certainly contains errors even though they tightly hang together’ (ibid.). Hence, more coherence does not imply a higher joint likelihood of truth. This informal argument can be made formally precise as follows. Following Shogenji we first define the degree of coherence of a pair, A and B, by