Inference to the Only Explanation

Philosophy and Phenomenological Research 74 (2007) 424-32

I propose that in some cases we may infer the truth of an hypothesis, since it is the only hypothesis left unrefuted by the evidence (a la Sherlock Holmes). Peter Lipton’s description of the Semmelweis case seems to provide an example of this. But he takes it to be a case of inference to the best (loveliest) explanation. I locate this source of difference of opinion in Lipton’s equation of evidence with (non-factive) observation. This equation gives us too little evidence; and it makes observation insufficient to refute a hypothesis. I contrast this with refutation by a known proposition.