Well, I'm just back from the 3rd Queensland Biohumanities Conference, convened by Paul Griffiths, which was titled "Idealization, mechanism and reduction: New Directions in the Philosophy of Proximal Biology". Speakers were Bill Bechtel (UC San Diego), Alex Rosenberg (Duke), Marcel Weber (Basel), Ingo Brigandt (Alberta), Mark Colyvan (Sydney), Stephen Downes (Utah), Karola Stotz (Indiana), James Tabery (Pittsburgh) and Rasmus Winther (UNAM, Mexico City), plus a host of people from around Australia.
It was, I expected, going to be more of the same old boring stuff on reductionism in biology, but it turned out to be a most entertaining and enlightening session and I got to meet some folk who I have been reading for a while. This seems to be the year for that - at PSA in Vancouver I met Bill Wimsatt and Jim Griesemer, who I hadn't previously met but who had been influential on my thinking. Now Bill and Alex. Good year...
The issues turned around the nature of reduction - whether it is an explanatory issue or a causal/ontological one. Alex thinks it is fundamentally about causation; Bill (Bechtel) that it is about explanation. I tend to want to be an ontological reductionist, or a physicalist, who thinks that all biology resolves down to physics and that biology has no laws of its own (although it clearly has generalisations), but I think that the issue is explanatory, with Bill. He rejects the notion of "levels", claiming instead that the explanatory power lies in setting up a domain needing explanation. The example used a lot in the conference, given Bill's plenary talk, was the discovery of proton transfer in the Krebs cyclephosphorylation. What called for explanation is how protons could be transported across the nuclear membrane. So the protons in the cycle are part of the explanatory domain, while the protons in the membrane molecules were not (they were insensitive to the explanation given). This is called explanatory relativity (see the excellent book by Alan Garfinkel, Forms of Explanation, 1981) or contextualism. But I remain committed to the view that all that has causal efficacy in biology, as in any science, is the physics, even if, as was argued in the final Q&A, causality in (the philosophy of) physics is a highly questionable notion.
The talks (well, most of them - Karola's, Stephen's and Ingo's failed to record; my bad) will be up on the Biohumanities Podcast page in a few weeks, once I've had a chance to edit them, along with PDFs of the presentations.
Most of the participants went to Byron Bay for surfing and camping, the bastards. I get to stay in Brisbane and look after my son and write and revise papers, oh joy.
Late note: Larry Moran picked up my mistake - oxidative phosphorylation (properly called "membrane-associated electron transport", he tells me), not Krebs... my bad recollection.
- Log in to post comments
But I remain committed to the view that all that has causal efficacy in biology, as in any science, is the physics, even if, as was argued in the final Q&A, causality in (the philosophy of) physics is a highly questionable notion.
I think similar problems are found in computer science. In an abstract mathematical sense the computer (by which I mean the mathematical contrivance of A.M. Turing and Alonzo Church, not an actual physical implementation) reduces to the bit. But while bits have to be taken into account when studying computational complexity, they are almost useless when analyzing the growth-rates/efficiency algorithms, and in software engineering it is virtually impossible (especially in modern development environments) to work on the object/machine code level.
That is why I think a general reductionism/holism dichotomy is rather useless. What level you need to work on depends on the explanatory power of doing so. Science should only be concerned with explanatory power (I think this is called instrumentalism).
Correction: There are some formal models of computation that do not involve bits, and Turing and Church are noted for different formalisms, but in general my point should be clear.