摘要：In the last half century the coherence theory of truth has largely fallen into disuse and disrepute. While there is now some flirting with coherence approaches, as each approved version of the majority position, the correspondence theory, duly founders, and holism gains in fashionability, still coherence has but few committed friends.1 Granted, it has had friends of a sort: most notably Rescher, who has made significant contributions, on which others may profitably build. But Rescher, while advocating what he calls a 'coherence theory' has twisted the theory into what it is not, a modified "self-evidence" theory, and has also warped it into a methodological pragmatism that would have made straight old-timers like Bradley and Blanshard blanch. As well the major virtues of the theory - if only it could be got to work, which unfortunately it can't - have been appreciated by isolated explorers of the wide truth terrain, such as Blackburn (see esp. his pp.237-8). The present exercise supplies one way of enabling the theory to work, without undue warping.2 That way does not pretend to be an authentic historical way, only an historically con trolled and informed way. For the primary purpose here is not historical explication; it lies rather in the development of coherence theory beyond its previous and varied historical settings3 , to render it somewhat more adequate and more coherent, and to begin to display some of its further virtues.
