Muddy Paste

Next Monday is Veteran's Day. There are no classes.

Next class, we start on a new text: Collins and Pinch

Godfrey-Smith proposes a frappe of empiricism, naturalism, and scientific realism. He misdescribes his own view, because his view has a scoop of social structure of science in it, which isn't mentioned in his description.

There is a lot to like in his view. In fact, his view includes almost every desirable characteristic of every view. That isn't hard, since his view is almost just the conjunction of every view. What's wrong with that? Why have people taken their views to be rivals? Because they are incompatible with each other. They contradict each other. I'm not sure that Godfrey-Smith's view literally embraces contradictions, in part because I'm not sure what his view is. But that is not the worst of it. That, after all, may just be brevity at the end of an introductory text. What is really worrisome is that the view seems to explain everything. What's wrong with that? Popper's "demarcation criterion" said that a theory is not scientific if it can explain everything. A scientific theory must take the risk of being wrong, of being falsifiable. A theory that can "explain" everything explains nothing. Godfrey-Smith only makes matters worse by embracing various pluralisms, for example, about explanation.

Some philosophers who give theories of this type say, quite reasonably, "But I'm a philosopher. I wasn't trying to give a scientific theory, but a philosophical analysis of ... ." Godfrey-Smith isn't entitled to saying that because he wants to claim the mantles of empiricism and naturalism. Empiricists typically take all knowledge to be like scientific knowledge in fundamental respects. Naturalists take all knowledge to be on the same footing as scientific knowledge, and take science to be the paradigm of knowledge. If you want to be an empiricist and naturalist, you can't also claim that your theory is philosophical knowledge in some sense that makes that different in kind from scientific knowledge.

Probably the most important insight about science that has arisen from criticism of the logical empiricists is that part of what makes science what it is is the social systems and arrangements by which science is organized.

Questions about what constitutes enough skepticism and enough trust, and about what theories are worth pursuing and which should be dropped haven't gotten even plausible, let alone good, answers when philosophers have tried to specify rules or principles to answer them. However, the social structure of science makes sense of a certain balance in these dichotomies which is good for advancing knowledge and which is such that we can explain why it is good for advancing knowledge.

There is an older point about science, which is actually probably the most important, and one which the logical positivists and Godfrey-Smith were trying to capture with their versions of "empiricism." Science values making disputes resolvable by public, impersonal means. (Remember Boyle vs Hobbes) Part of doing that is looking for and structuring theories in such a way that reproducible experiments play a role in evaluating them . Part of that is making arguments publically on minimal common ground. That, I submit, is the real core of why philosophers of science have advocated empiricism: What turns out to be good science has little or nothing to do with authority and with prior beliefs, and everything to do with using means of persuasion that are public, reproducible, and convincing to all parties to the extent possible. Why is this important?

Let me contrast scientific attacks on problems to religious and political ones:

Hunger. The dominant religious approach is to emphasize the virtue of charity. You shouldn't let anyone starve, you should help, ... . It appeals to people with a particular belief system to help in particular charitable ways. The dominant political approach is to institute mechanisms for the redistribution of wealth. Taxes used for food-relief programs; land reform, and the like. The scientific approach is to find better and more efficient ways of producing food. Better strains of crops, better ways of controlling pests, better ways of fertilizing.

Abortion. Religion, to the extent it has an approach, has tended to oppose abortion, not only for those who practice the religion, but for everyone else too. This is a straightforward attempt to broaden religious authority. Political approaches have included restricting information and availability of abortion and, on the other side, preventing and eliminating such restrictions. Science responded to the problem by inventing effective, inexpensive methods of birth control.

Science, unlike religion or political process, is designed to produce consensus in a diverse group, and it is therefore a touchstone of the possibility of liberal democracy.

Do scientific theories really tell us what is true? Do they have some special claim to that? If what is true is supposed to be true of a common, public world, independent of one's particular culture and belief system, then science is the only method of attempting to find out the truth that is directed at finding out the truth about that world. Of course, particular scientific theories have different degrees of reliability, evidence, utility, and so forth, but they are all empirical in this sense: the control and predictive abilities they confer are available everywhere and to everyone.

Criticism of Godfrey-Smith's Conclusion

What is empiricism? Roughly, a view that holds that experience is fundamental for knowledge.

What does it mean to say that something is fundamental? That it is the base on which the rest rests. So far as I can see, the bedrock feature of empiricism is that it takes experience to be prior to and a basis for the rest of our knowledge. But that is false because of the theory ladenness of observation. Godfrey-Smith minimizes the importance of theory ladenness, and claims to be an empiricist, but even he concedes that

I, in fact, think there is even more to the theory ladenness of observation than that, but that is enough to leave it terribly hard to see what empiricism could still amount to.

Why does Godfrey-Smith want to be an empiricist? The central one is that our theories are in important part determined by interaction with the world that they are about. But that is true, for example, on the view I advocated above, which is not empiricist. That is an important feature of an adequate view of science, but it does not require one to be an empiricist.

Realism: It seems to me that whether or not one is a realist about scientific theories, paradigms, whatever, can be thought of as a psychological or personal matter outside of science: van Fraassen locates the difference between realism and his antirealism in whether one takes theories, paradigms, whatever, to be TRUE or just empirically adequate. One could take either attitude, and still do the same things in the lab and publish the same papers. So I'm not sure that this is even an interesting question within the philosophy of science.

Godfrey-Smith wants to naturalize his empiricism and his realism. What does that mean? It means that he wants to rely on scientific evidence about experience and about theoretical entities to decide what to think about them. That seems to me to be an interesting program. But it has little to do with EMPIRICISM or REALISM: Empiricism wants to take experience to somehow be fundamental. Work in perceptual psychology, for example, is at least as theory laden as any other scientific work, and so it can't license taking experience to be fundamental. At best, it might do something like what Godfrey-Smith says: it might show experience to be reliable. That is interesting and important work, and it might even someday be helpful in designing experiments that are from a broader common basis, but it doesn't help in supporting EMPIRICISM. What about REALISM? REALISM is, by definition, about things that are not observed, and so it is hard to see what scientific experience could do that is relevant to the issue. Well, things that used to be unobservable become observable, and so we can look at our track record. When we become able to observe things, are they as we theorized? Whatever the outcome of such studies, one could take them to show that

When we become able to observe things that were previously unobservable, our theories of them are usually empirically adequate.

If I am right about EMPIRICISM and REALISM, then they are not interesting theories about science. Godfrey-Smith has described some interesting things we might try to learn about that have no names. Well, in this, as in everything else, it is a good idea to recycle. We might name these things, empiricism and realism.

-- ShaughanLavine - 09 Nov 2005 - 05 Nov 2007