Tuesday, May 10, 2011

General Philosophy of Science: What is it Now?

Further to my comment earlier about Scepticism and Philosophy of Science, some further thoughts about General Philosophy of Science:

General Philosophy of Science was almost entirely a twentieth century phenomenon. Before the 20th century, there was scientifically informed metaphysics and epistemology—I am thinking of Descartes, Locke, Kant and their intellectual heirs, philosophers who believed that science could teach us something about the structure of reality. The central problem of twentieth century GPOS, however, was new. What is the status of unobservable theoretical entities? What are they? And how can we know them? The problem was urgent because theoretical entities were increasingly a feature of nineteenth century (and later) physical science--atoms, fields, etc.

The idea that drove GPOS was that theoretical entities could be constructed from (or alternatively eliminated in favour of) sense data. This idea, which was also the founding idea of analytic philosophy, began its long decline in the 1960s, when Hilary Putnam introduced ways of talking about unobservables that did not rely on these constructive techniques. Putnam’s work was particularly attractive to philosophers because it showed a way out of the incommensurability problem that arose from Kuhn’s work, for incommensurability sounded to many like a reductio of the whole analytic programme. Now, a decade into the 21st century, the problem of unobservables is no longer at the centre of analytic philosophy, and anti-sense-datum realism has become the norm. (In the PhilPapers survey, 82% of respondents leaned toward non-sceptical realism, and only 4% to idealism; 75% to scientific realism, and 75% to correspondence or deflationary theories of truth. The percentages don't change much when you look at self-identified GPOSers.)

Accordingly, GPOS has declined. How many top-twenty departments have hired GPOSers in the last decade? Very few, I venture: the assistant and associate professors who list themselves as philosophers of science are either philosophers-of-X (POXers?), GPOSers having given way to formal epistemologists or analytic metaphysicians.

GPOS does, however, live on, mostly in satellite departments devoted to History and Philosophy of Science—there are such units in Cambridge, Sydney, Pittsburgh, Chicago, Paris, Toronto—and also in a few philosophy departments that have chosen to develop specializations in philosophy of science—Bristol, the University of Washington, the University of British Columbia are examples that come to mind. In these departments, there are usually POXers devoted to physics and biology and social science, and a GPOSer who is viewed as providing support for the foundations. But the question that I think has not been resolved is how GPOS lives in an HPS environment. Has GPOS become, roughly, Kuhnian? Feminist? Is there a viable path for development that traces back to the Viennese origins of GPOS?

(Cross-posted at NewAPPS)


  1. Hi Mohan,

    As I have already argued a few times on this blog, I think the PhiSci (as opposed to PoX) has a future only insofar as it interacts and communicates with "neighbouring" philosophical fields (e.g. epistemology, logic, philosophy of language, metaphysics, etc. personally, I would add, political philosophy). I would argue that that was the way it was meant to be in pre-Khunian times.

    Once I thought I knew why PhiSci-ers had stopped practicing PhiSci in that spirit but I no longer think I know.

    If I get your point (and please feel free to correct me if I'm wrong), you seem to be wondering if that project is still viable, but I'm not clear what are the reasons to think it's not, especially after the Khunian spell has been broken for a few decades.

  2. Hi Gabriele,

    I don't hold the Kuhnian spell responsible for the decline of GPOS. In my view, Kuhn and classic GPOS shared the idea that theoretical terms have to be reduced to (?) (It's not clear to me what ontology Kuhn wanted to reduce theoretical entities to.) In any case, Kuhn thought that since theory change leads to different methods of reduction, so different theories are incommensurable.

    Anyway, for what it is worth, I think classic GPOS is not worth pursuing. This is because I reject the ontology of sense-data. I don't think THAT project is viable, but I wonder whether there is a replacement.

  3. I would broaden the scope of GPOS, surely for example the demarcation problem (along with the broader claims in the demarcation debate about what the basis of science falsifcationism vs. inductivism vs. bayesianism and so on) was a central line that defined camps for at least a time in 20th century phil. of science. Of course the demarcation problem is arguably a dead issue, whereas at least the realism vs. anti-realism about unobservables has some life. Although, Bayesians seem to me to be keeping alive the idea of a general method of science (as opposed to method of x).

    Also, all the non-key issues in GPOS like reduction, confirmation, modelling etc., still have a unique GPOS character (versus philosophy of mind or general epistemology discussions of similar topics) and are not exhausted by the positions one might take in the central debate(s) of GPOS. Also some such fields are strong and general. So I'm skeptical that GPOS needs a "central debate" to remain a going concern.

    I suspect the apparent decline of general philosophy of science is due to a shift from questions that assumed or looked for common elements in science to ones that end up focusing on each sub-discipline's particular methods and results. The attempt to understand and categorize science as it is actually practiced would be one characteristic of much more recent philosophy of science. While some skepticism about the lack of unity is new, I think the old questions often depended on much more coarse grained kind of unity.

    Doing a random glance at old Philosophy of Science articles I sort of suspect philosophy of x has always been the dominant subject (used to be philosophy of logic, math or physics, now biology is more popular).

  4. Hi Mohan,

    I must have misunderstood what you meant then. What threw me off is that you seem to be identifying GPoS with a very specific package of views within it (one which, btw, is very unpopular today). At least since the publication of van Fraassen's The Scientific Image in 1980 the debate seems to have have shifted form unobservable entities to theoretical terms and all sides seem to have given up on reduction to sense-data. However, the scientific realism debate, just to take one example, seems to be alive and well. So, if you are interested in seeing how GPoS lives in an HPS environment, I would say that you need to look no further than Toronto's IHPST--Anjan Chakravartty is one of the best examples of how to do GPoS within a HPS environment without being Kuhnian, feminist, or committed to sense-data. (If your question is more about how GPoS-ers interact with HoS, on the other hand, you could look at Michela Massimi at UCL or Hasok Chang at Cambridge). I take it though that your question was rhetorical so I suspect I'm still misunderstanding you.