Holmes again, Holmes again, jiggety jig

Sun 23 Jul 2006 02:42 AM

Brian Weatherson links to a recent episode of The Philosopher's Zone, an Australian radio program. They have it on-line both as audio and transcribed.

The host, Alan Saunders, is interviewing Greg Restall. They are discussing the fact that, in classical logic, a contradiction entails everything but, in ordinary reasoning, it does not. Saunders considers an example that I have discussed here recently:

[We] actually allow for the inconsistencies. I mean, let's say it's a Sherlock Holmes story. Now we know that Sherlock Holmes was not the centre of Conan Doyle's interest, he wrote these stories very, very quickly, so if I found an inconsistency like that in a Sherlock Holmes story, I would sort of mentally dissolve it, because I'd think, Oh well, he just wasn't paying attention. If, on the other hand, it were an historical account of a murder and I found [an inconsistency], then I'd say, Well, the evidence isn't all in yet, I'm going to have to suspend judgment on this.
By 'mentally dissolving' the contradiction, we refuse to treat it as evidence about Holmes at all. It provides no constraint on the set of Holmes-worlds. When I discussed inconsistencies in Holmes stories, I suggested something similar.

In reading a historical account, Saunders suggests that we should take opposing claims seriously as evidence-- albeit conflicting evidence-- of how things actually happened. When considering the past, we know that there was some one way that the world was. There is no such assumption about Holmes' world. However, I am not sure that this difference is sufficient to justify Saunders' different handling of these two cases. We can take contradictions in stories as signs that there is some strangeness afoot in them. Why don't we?