2.10.2013

Metaphysics, models, and pragmatism

My response a draft of Jon's:

The more I read it, the more I hate this paper. It is actually a stunning example of historically ignorant and completely unsatisfying metaphysics. It is also just bad philosophy.

I wrote you a drunken text about it last night, and let me continue to be belligerent about it here. This whole issue was gone over quite thoroughly in post-positivistic philosophy of science, sometimes as a discussion of reduction (which I know you are familiar with) but also as a discussion of the unity of the sciences, or the autonomy of the special sciences. There's even an SEP article about it:

http://plato.stanford.edu/entries/scientific-unity/

It mentions, among other things, about philosophy of science pre-1990 (which might never have happened from the look of your paper), and Ian Hacking's critique of the unity of the sciences from the early 90s, which is worth knowing, relevant to this discussion, and also completely absent. You'll also see that a range of pluralistic positions are described in the encyclopedia entry, none of which make an appearance in your paper.

I see no substantive progress being made by your work here that isn't already in the literature available in the field. Instead, I see a paper written by someone who has yet to realize that the field exists and has some homework to do.  

This is one of the major problems with philosophical practice, especially in metaphysics, today: some kids who know nothing of the history of ideas have an idea that they think is novel, but has actually been studied carefully for years; but because they know nothing of history also know nothing of the problems with various formulations of the views, and therefore carelessly recapitulate so many of the mistakes that so many people already worked so hard to avoid. Then their peers read their misguided regurgitation of history and use it to calibrate their own interpretations. The effect is more like a bull in a china shop than an example of careful and clearheaded philosophy. 

Just for instance, you begin the discussion of "real patterns" with a distinction between the actual world and the set of all possible worlds. Am I to think that what follows in this discussion requires the metaphysics of possible worlds to articulate? If so, you never use this machinery again in the article-- a machinery that is well studied in both its applications and limits that you just blithely skip over. If I have any ontological objections to this machinery being employed, they slip off your paper without any friction. Insofar as your paper is meant to clarify ontological disputes in the sciences, this lack of engagement with the literature renders it utterly useless. 

Here's another way of characterizing metaphysicians today: "Hey, if I argue for a stupid position using these dumbed-down and obvious words, it sounds like my position is novel and compelling!" This is what you are doing. Stop it, it makes things worse. 

Look, here's a sentence you even italicized because you think it is that important to your point:

_science is the business of identifying genuine patterns in how the world changes over time. _

This is a meaningless claim. Any philosophical work being done here is being done exclusively by the word "genuine", which utterly begs the question. You say projectability is the standard for judging a pattern to be genuine, and you have a nice example, but it is far from clear that "projectibility" can be made any more clear than "genuine". I'd refer you to Goodman here I don't think that'll do any good. I'll just say you might think about reading some philosophy before you graduate, it might be good for you. 

In any case, you've loaded us up with your own proprietary ontology (bitmaps and whatever) which has no expressive advantage over existing ontologies, and the whole argument leaves me no closer to understanding the possibility of pluralism in science any better than Mitchell's own work. 

I'm left feeling like this isn't an article, these are notes for yourself in prelude to actually doing research on this problem. If you want to write an actual article, I want to see some concrete cases (in climate science, for instance) where there are two different models, each making different predictions and suggesting different policies, and then I want you to show me explicitly how a pluralist approach actually yields a more satisfying scientific answer than other meta-scientific approaches. In other words, if your views have any substance you should be able to show how your metaphysics are more useful than other metaphysical popsitions, and that's not what you've done here. You've shown why you like the view better, but you haven't shown that it can do more, and the latter is all that really matters. 

Frankly I don't think you can do it; in other words, I think the views here are completely without substance. Faced with a case of a plurality of models, your arguments here are no better equipped than any other view in terms of making the science more effective. If science is about identifying real patterns that are effective in prediction, and your view here has no predictive or effective advantages, then I'm struggling to see how you've identified a real pattern. In other words, you aren't doing science. You are doing metaphysics. Stop it. 

The idea that such sloppy work would suffice to ground the naive realism of laws and actual worlds is just embarrassing. The fact that you'll get a job off this shit is literally the cause of my depression and insomnia. 

For what it's worth, I don't think Mitchell's view is useless, but I don't take her to be offering a metaphysical view. All this metaphysics completely misses the point of pluralism. The point is in your paper, but it isn't the focus, and you are so distracted by the metaphysics that you almost miss it yourself. The point is that these issues don't get sorted by metaphysics at all, but only by scientific practice itself, and scientific practice is empirically pluralistic. 

But scientific practice flourishes as a multiplicity not just because of metaphysical facts about the world (though that's involved), but also because of all sorts of reasons involving the complexities of the human culture and technological situation in which the practice of science is embedded. The distribution of research grants, for instance, can easily produce a pluralism of scientific perspectives that are genuinely responsive to real patterns, but they are patterns like "the concentration of religious conservatives in the south" and "the frequency of space shuttle disasters", which may have just as much influence over the practice of science as any "natural laws". Restricting your discussion of the practice of science to the metaphysics of convergence with laws already gives up the game. 

And the problem, of course, is that the patterns that guide the practice of science can diverge arbitrarily from the world itself, or focus narrowly on some aspects to its great detriment. Humorism (http://goo.gl/ioFOM) is one of my favorite examples of such a view, but pick your own if you'd like. I don't see anything in the pluralism you are advocating that would suggest a way to recognize when we are holding a bad theories, how to avoid adopting them, or how to remove them once we have them. In other words, the pluralism you are advocating doesn't give us an empistemological advantage in picking better scientific models, and therefore I see no reason to think it provides a promising metaphysical approach either. 

But whatever. Fuck you.