Friday, October 11, 2013

Philosophy of science and the demarcation problem

To: Mark s.
From: Geoffrey Klempner
Subject: Philosophy of science and the demarcation problem
Date: 30th December 2010 13:09

Dear Mark,

Thank you for your email of 14 December, with your essay for the University of London BA Methodology module, in response to the question, ''There is not a single distinction between science and non-science, but rather a variety of distinctions between various ways we have of finding things out.' Discuss.'

I found this very informative and also gripping. You have grasped the importance of the challenge of demarcation, and provided a useful sketch of much of the landscape around this problem. As your aim evidently was to be thorough, I did miss discussion of Feyerabend's view which you refer to as 'scepticism' at one point, and at another, 'relativism'. But you can't have everything.

You are right to ask at the beginning what follows from the decision to regard a particular activity as 'scientific' or not. Grant funding, for one thing. However, the focus of your question about consequences seems to be confined to the distinction between 'science' and 'pseudoscience' (e.g. astronomy versus astrology). What about the humanities? Why does it not matter that they are not 'scientific'? (You do at one point describe important similarities between the we one appraises the work of scientific researchers and researchers in the humanities.)

One could say that all academic disciplines share a certain broad methodology which involves all the norms described by Merton -- communalism, universality, disinterestedness, originality and scepticism (or at least four out of five, one might draw the line at universality) -- though modified to take into consideration the differences in the practices of research (e.g. setting up experiments in a laboratory vs poring over old texts in the Bodleian). Which in turn suggests that we need to have something here about the actual practice of science, what distinguishes the particular way that scientists go about doing responsible, peer evaluable, disinterested research which is different from the way that non-scientists do this.

This question grips me because I feel that I 'know' what science is, and, moreover, it seems to matter that we should know this and be able to state it. Not because scientific inquiry is the only responsible way to inquire, or the only true path to knowledge, but rather because this seems to be an important part of one's self-understanding as a 'scientist'. You are in an interesting position in regard to this question: your work as a clinician requires to to adhere to 'scientific' norms and principles, you may even be an active contributor to research programs (e.g. testing a particular drug or therapy), even though 'scientist' would not be the correct description of what you do.

An example which springs to mind which raises the question of the meaning of 'science' in an acute way is the difference between archeology and paleontology. Both the archeologist and the paleontologist are concerned with particular historical events. But arguably, only the paleontologist views this research in relation to the idea of scientific law (e.g. the theory of evolution). Is that the difference? that scientists put questions to nature, while non-scientists do not?

On this view, everything turns on the idea of who or what you are 'putting the question' to. The researcher in the Bodleian is asking questions of human beings (even though they are not in a position to answer back). There has always been human life (for as long as there has existed some form of inquiry) but there hasn't always existed the *idea of nature*. This was a momentous discovery. If one were inclined to scepticism, one could even imagine a scenario where, unknown to us, there is no 'nature', there only seems to be (a version of Descartes' 'evil demon' hypothesis). Surely, therefore, you can't ask the question, 'What is science?' without simultaneously asking what it is for there to be such a thing as 'nature' or a 'natural world'.

The fact that nature exists (or, at least, the fact that we must assume that nature exists) imposes a particular kind of discipline which is different from that in other areas of research. Testability, reproducibility assume a special importance (even though they already had a role -- you can test a historical hypothesis, for example). This in turn suggests to me a possible way to have one's cake and eat it. There is, after all, just one thing that characterizes science: its relation to nature. But this relation in turn generates or requires multiple and overlapping criteria for what does or does not count as being 'scientific'.

Although your essay is very thorough -- or perhaps because of it -- there did seem to be a slight structural flaw in your argument. Following the plan which you set out at the beginning, you discuss Kuhn's objection to Popper under 'Falsifiability is irrelevant to normal science' and respond on Popper's behalf. Then two subsections later in 'Falsification of an isolated hypothesis is impossible' you discuss Duhem and the problem of auxiliary hypotheses. What I missed here was recognition that this is precisely the problem that Kuhn picks up upon: that you can defend any favoured theory by means of suitable adjustments in the surrounding network of auxiliary theories which are considered less 'precious' from the point of what we believe to be the body of established scientific knowledge.

All the best for 2011!

Geoffrey