Tag: new talk

I gave a talk at the EPSA meeting in Düsseldorf on what I call the argument from the good lot (pdf). It uses a case study from genetics in the first half of the 20th century – how was DNA shown to bear hereditary information? – to argue that scientists work hard to avoid one of the great pitfalls of inferences to the best explanation: the bad lot.

I’ve uploaded the slides (pdf) to a talk I gave last week at a workshop in Oulu in Finland. This is my latest attempt to explicate the methodology of integrated history and philosophy of science (or at least of one type of such work). The workshop’s topic was “Testing philosophical theories against the history of science”, and the full program is well worth a look – it was a stimulating event.

I have uploaded the slides from my second Pittsburgh lunchtime talk. This is an initial presentation of a current paper project. Here is the question: If science proceeds by (1) proposing a number of candidate explanations for a phenomenon, (2) ranking these explanations by explanatory power and (3) accepting the most highly ranked of the candidates, then why should we expect science to arrive at truth? After all, it is always possible that we simply failed to consider the true hypothesis in the first place. This would explain why so many successful — that is, highly ranked — past theories were later abandoned. In recent years this issue has been vigorously pursued by Kyle Stanford, who speaks of the “problem of unconceived alternatives”. In my talk I develop an account of why the problem of unconceived alternatives is not acute in much of the life sciences. More to follow.

I’ve uploaded the slides to my Pittsburgh Center for Philosophy of Science Lunchtime Talk. It’s a relatively mature version of the work I have been doing with Tim Räz on the methodology of integrated history and philosophy of science.

I have uploaded the slides to my recent talk “A Tale of Two Puzzles.” It presents the best current version of my thinking about causal inference and modeling as distinct scientific practices. In the last part of the talk, I show how the CI/modeling distinction can perhaps do work in the debate about scientific realism.