On the use and abuse of philosophical jargon 
I wrote my dissertation on the underdetermination of theory by data. I managed to turn material from my thesis into almost ten free standing papers, but despite some perfunctory efforts I never turned my work on underdetermination into a book. I thought I might finally write that book when I went to Pittsburgh in Fall 2010.

About the same time, I started working on natural kinds. I expected to write a couple of papers on the topic. Yet the project grew larger, and I ended up writing a book on natural kinds during my semester of sabbatical.

Both 'underdetermination' and 'natural kinds' are items of philosophical jargon which are usually presupposed to be well-defined and univocal. Yet both are also used in a tremendous variety of ways, and the literature includes authors' unwittingly using the terms in different ways.

'Underdetermination' is often taken to be a label for the so-called Duhem-Quine Problem but also for the Problem of Empirically Equivalent Rival Theories. It is also sometimes taken to include riddles of induction (Hume's, Goodman's, or both) or quotidian cases of curve-fitting. These are very different things. So one possible diagnosis is just that the word is used in a confused way.

In my thesis, I found a way to characterize underdetermination so that all of these different issues turned out to be varieties of the same thing. The basic idea is that a case of underdetermination involves a set of rival theories, a standard for what will count as responsible choice between them, and a scope of circumstances in which responsible choice is impossible. The disparate things travelling under the banner 'underdetermination' can be obtained by filling in these three parameters in different ways.

This was kind of a neat trick, but the problem was there was really no useful work done by having the big umbrella term. The analysis shows merely that there is a way to make sense of 'underdetermination' talk, not that you really ought to have it as a term in your vocabulary. Many specific varieties of underdetermination have important consequences, so I wrote papers about those. I could not see how to write one sensible book covering all the disparate stuff.

'Natural kind' is similarly used as a label for many different things: categories which support induction, categories with essences, categories that ought to appear in a scientific account of the world, categories which we rigidly designate, and so on. Again, these are very different things. So you might just throw up your hands.

Yet the conception of 'natural kind' which I defend in the book does not vindicate all these different presumptions. I do not treat natural kinds as needing to have essences, and I am neutral on the question of how natural kind terms refer. So what I offer is more an explication than an analysis.

I think that 'natural kind' in the sense which I defend in the book does important philosophical work. It is useful for framing questions about whether Pluto is really a planet, about the reality of species, and so on.

I had not even explicitly noticed that my work on underdetermination had headed into a cul-de-sac. It was only after I had completed the manuscript for my book that I realized that I really do not think that 'underdetermination' is an especially helpful term to have in our philosophical vocabulary but that 'natural kind' is.

To put the point concisely: 'Underdetermination' is not a natural kind for philosophy of science, but 'natural kind' is.

Even more concisely: Yay explication and 'natural kind'! Boo analysis and 'underdetermination'!

[ add comment ] ( 2349 views )   |  [ 0 trackbacks ]   |  permalink
Philosophy of science as it was taught to John Rawls 
[crossposted at It's Only a Theory]

My colleague Jon Mandle has been looking at John Rawls 1950 doctoral dissertation, A Study in The Grounds of Ethical Knowledge. Jon asked me about a section in which Rawls contrasts ethical theory and scientific theory. The philosophy of science that he presumes is really just background. Yet he discusses what is now often called the Duhem-Quine Problem, a couple of years before Quine's "Two Dogmas of Empiricism". So where did Rawls get it from?

I did not have a good answer to this, beyond the obvious suggestion. So I decided to share the interesting bit here.

Read More...

[ add comment ] ( 5234 views )   |  [ 0 trackbacks ]   |  permalink
Nature, don't think I won't cut you 
When I learned about the recent collection Carving nature at its joints: Natural kinds in metaphysics and science, I had to change the title of my forthcoming book. After much discussion here, on other blogs, and on Google+, and after some back and forth with the publisher, my book will be Scientific enquiry and natural kinds: From planets to mallards. (A better title, I think. So the rigamarole was worth it.)

I learned about the Carving collection when I was asked to review it for Notre Dame Philosophy Reviews. I love open-access journals in general and NDPR in particular, so I was glad to do it. My hatchet job review has now been published over on the NDPR site.

[ add comment ] ( 2103 views )   |  [ 0 trackbacks ]   |  permalink
Grounding metaphysics 
In a recent article,* Karen Bennett poses and attempts to answer a metaphysical dilemma about the relationship between ontological grounding and the fundamental. Grounding is a relation between basic and less basic facts or stuff. For example: a series of notes grounds a melody, the physical grounds the mental (according to physicalists), and so on. Fundamental things and facts are ones which are absolutely basic, i.e. ungrounded.

Is the grounding relation itself fundamental?

First, suppose the answer is yes. Then consider some grounding relation, like that physical states ground my sadness. It is ex hypothesi a fundamental fact. So there are fundamental facts that involve my sadness, even though sadness is not a fundamental notion. That can't be.

Second, suppose the answer is no. Then there must be some further fact or stuff which grounds physical states grounding my sadness. Regress ensues.

Bennett's own strategy is to argue that the regress is stopped by the nature of the grounding relation. I have little to say about that. Instead, I want to thnk about some different replies.

A strategy which Bennett considers but passes over is to reject grounding. This would mean that everything which exists is metaphysically fundamental. As she puts it, the world would be ontologically flat. She writes, "I have no knockdown argument against the claim that the world is flat. But every fiber of my being cries out in protest." Crying fibers are like incredulous stares, but I think it would be a bad turn to say that everything is necessarily fundamental. If that were so, it would not be clear what the hell 'fundamental' was supposed to mean. It must at least have a possible contrast to be sensible.

A related strategy is to reject fundamentality. Bennett considers this only in a footnote and objects:
I truly think it is near impossible - certainly a bad idea - to do away with fundamentality talk altogether. Everyone, even those who reject grounding, should be able to claim that some things are more fundamental than others. (Good luck doing philosophy if you can't.) [fn 9]

There is something funny about this reply. She is right to say that we ought to be able to say that "some things are more fundamental than others." For example: individuals are more fundamental than trios, subatomic particles are more fundamental than atoms, and so on. However, this comparative notion 'more fundamental than' does not require that we can say of anything that it is utterly fundamental. That notion of absolute fundamentality is one I can do without. I accept only the comparative notions of grounding and more-fundamental-than. This dissolves Bennett's dilemma, because without the utterly fundamental there is no way to pose the worry about grounding.

One may object to my suggestion: It is possible to define the monadic 'fundamental' in terms of the relations. Let 'A is fundamental' mean 'There does not exist an X such that X grounds A' or 'There does not exist an X such that X is more fundamental than A'.**

The problem with the objection is that the existential quantifier in the definition must be unrestricted. On the usual accounts, like Sider's, unrestricted quantification is itself a fundamental notion. By refusing to accept the notion of fundamentality, I am also refusing to countenance unrestricted quantification. So my suggestion eliminates the resources required for the objection.

Although the fibres of metaphysicians' being might cry out, rejecting fundamentality does not make it impossible for me to do philosophy. Insofar as I care about grounding or more-fundamental-than in the first place, it is to consider particular cases. For example: What's the relationship between the mental and the physical? the musical and the acoustic? kinds and individuals? These questions can be answered in terms of the relation. It seems to me, in fact, that a term 'utterly fundamental' does not help me with such questions whatsoever.


* 'By Our Bootstraps', Philosophical Perspectives, 25(1), 2011.
** This objection is readily available. Bennett suggests this argument in her essay. Ted Sider made it explicitly to a similar suggestion in the Q&A at the UAlbnay Grad Student Conference last Spring.

[ 3 comments ] ( 7163 views )   |  [ 0 trackbacks ]   |  permalink
What's the opposite of philosophically conservative? 
In summarizing his philosophical approach to the photographer Steve Pyke, David Lewis said, "I am philosophically conservative: I think philosophy cannot credibly challenge either the positive convictions of common sense or the established theses of the natural sciences and mathematics."*

This seems like an odd thing to say, and I suspect that it summarizes why David Lewis' work has always made me a little uneasy. He conceives of philosophy as doing something separate and outside ordinary or scientific enquiry. Such an approach makes metaphysics a matter of window dressing our beliefs, without any possible influence on what the main doxastic inventory is.

Of course, this kind of conservatism is not unique to Lewis. I gave a job talk once and, during the question and answer period, an epistemologist in the audience objected to my argument on the grounds that it might lead us to disagree with scientists about some things and (he said) he would not want to tell scientists that they were wrong. The best reply to such an objection: But what about when scientists are wrong? It would be perverse not to point that out.**

Diametrically opposed to Lewis' approach is a kind of eliminativist naturalism according to which responsible philosophy is just science that happens to be done in a department called 'Philosophy'. Quine is the posterboy for such an approach. This kind of eliminativism is conservative in its own way, because it means that there is nothing that philosophy as such can add to science. There ends up being no philosophy as such at all.

A natural middle position is to say that philosophers typically address different questions than scientists do. Moreover, the methods appropriate to those questions are not identical to methods appropriate to the natural sciences. There is no sharp boundary between the scientific and the philosophical (Quine is right about that) but there is sufficient difference on either side of the boundary that the existence of philosophy departments is not just as arbitrary administrative fact about universities. Yet the porous nature of the boundary means that the enquiries can have things to say to one another.

Philosophy of science must accept science as for the most part OK. If it yielded total, utter scepticism, then it would stop being philosophy of science and becomes something else. (Mysticism, maybe.) But the qualifier 'for the most part' is important. Philosophers can call into question parts of science. Philosophers of science might even challenge and overturn some canonical examples of good science; what they can't do is overturn all of them.


A further aside: It's odd that Lewis invokes the "convictions of common sense", as if common sense consists primarily of a paddock of inviolate beliefs. As I have argued elsewhere it is better to think of common sense as a commitment to giving prima facie trust to certain methods and inferences. For example, seeing x is prima facie reason to believe that x exists. The same holds for the sciences: They are in the first place a matter of method rather than a matter of conviction.


* HT: Steinblog.
** I do not recall what answer I actually gave. I recall being shocked by the objection, and I might just have said "Really?!?"

[ 3 comments ] ( 3339 views )   |  [ 0 trackbacks ]   |  permalink

<<First <Back | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | Next> Last>>