## Words about words

Tue 12 Nov 2013 10:17 PM

Miles MacLeod has a nice review of my book over at Metascience. You can see the first two pages for free, which are the ones before he starts raising objections.

I updated the on-line drafts of my work-in-progress papers about what Nelson Goodman would say and what John Stuart Mill would say.

## Comments

from: Greg

Wed 13 Nov 2013 12:02 PM

First, I must confess that I still haven't read the book. I apologize!

But I wanted to ask about one thing.

"a kind K is a natural kind for a domain D if [and only if? -GFA] (1) it achieves inductive and explanatory success for D and (2) no alternative taxonomy without K would do so."

Does your account include the 'only if' direction? (I see that Matt Slater thinks so, in his NDPR review.)

If so, then I wonder whether we can get a counter-example from the different formulations of classical mechanics. In these cases, there are (what I consider) intuitive natural kinds -- e.g., FORCE in Newtonian mechanics -- which (1) achieve inductive and explanatory success, but (~2) there IS an alternative taxonomy without FORCE that also achieves inductive and explanatory success (specifically, Hamiltonian and Lagrangian formulations of classical mechanics, which say nothing about forces, and couch laws in terms of energy or action).

So, summarizing: (i) Do you accept the 'only if' direction?

(ii) If you accept the 'only if' direction, what do you think about the case of FORCE?

from: P.D.

Fri 15 Nov 2013 10:59 AM

Sorry for the delay in approving your comment. SimplePHPblog is supposed to send me an e-mail when there are new comments, but sometimes it just doesn't. Every time it glitches, I briefly consider migrating to Wordpress.

As for you question, let's treat it as "if and only if". What, then, about FORCE? It is dispensable in Hamiltonian and Lagrangrian formulations, so we can do comparably successful science without it. It doesn't meet the requirement of the restriction clause, and so it's not a natural kind.

One could deny this directly, by arguing that FORCE is somehow really real. On my account, any arguments for that would have to show that there really is some cost to doing without FORCE after all and this deny the alleged dispensability. I haven't thought enough about the details to rule that out.

Another possibility is to accept that FORCE is not a flat-footedly natural kind, but to insist that it is nevertheless more natural than earlier kinds like IMPETUS. That treats naturalness as a matter of degree, something of the form "kind K1 is more natural than kind K2 for domain D". This can be captured in terms of relative success and dispensability.

from: Greg

Sat 16 Nov 2013 05:30 AM

No problem -- I never really thought of the philosophical blogosphere as a time-sensitive operation.

If you argue that forces are really real, then will you also argue that energy and action are real? Or were you thinking that there would be just one best formulation of classical mechanics, and the kinds it appeals to are the (most) natural ones?

from: P.D.

Sat 16 Nov 2013 02:37 PM

My hunch is that if force were flat-footedly real, then action wouldn't be. But I really don't know enough about the details to even have a confident hunch.

Regardless - the case is a bit of a misfit for my account, because my account is about natural kinds. If particular instances of force are real, then I suspect that FORCE would turn out to be a natural kind. But my account doesn't say how to figure out whether particular instances of force are real.