Comments

Tuesday, January 6, 2015

For SM (scientific method) fans

Here's a couple of things that I ran across lately that tackles SM issues.  I know that many of you are into this sort of thing (and my view is that this is ok so long as it is among consenting adults). So here is a piece by Noah Smith where I first ran across the discussion and here is a piece that he links to. The discussion was prompted by a piece in Nature (here) by two important physicists complaining yet again about string theory's lack of visible empirical support and a reply by Brian Greene and a tweet by Sean Carroll defending the virtues of "elegance" and "explanatory" power  as good arbiters of scientific worthiness.

This is all very interesting and good fun, especially when viewed from the linguistic sidelines. However, I find the whole discussion of more sociological than scientific or philosophical interest.

First note how the piece in Nature pulls out the whole Popper falsification card to poop on what it doesn't like. I think I've noted before (here) that this maneuver is generally used as a rhetorical slap in the face and that the people who use it have a none too subtle view of what Popper had in mind. At any rate, I agree with Carroll that falsification is a pretty blunt critical tool and that it cannot carry the weight that its wielders seem to believe that it has. So, whether Popper meant falsification in this way or not, that's its current usage as the Nature piece demonstrates yet again.

Second, nobody is suggesting relying exclusively on elegance or explanatory power as a metric for theory evaluation.  Everyone agrees that it would be nice were there some novel data as well. What's at stake is what to make of theories that seem to have little chance of being empirically tested, where facing the tribunal of experiment seems like a far fetched hope.  The answer obviously is that there will be lots of disagreement and that nothing profound can be concluded.  Elegance and explanatoriness do matter and among extant theoretical options some are better or worse along these dimensions and that is a good thing to know. But, nobody thinks that this is enough. The question then is what to do when one is in such a bind and what I mean here is that the question is what research should be funded. Physics is not cheap and so it is reasonable to ask whether elegance and explanatoriness are sufficient to keep pumping the big bucks into the field.  I can imagine different people concluding differently. And, IMO, I believe that this is what most of the fight is about, at least sotto voce. It's one thing to pour tons of cash into a field with obvious empirical pizzaz (think Hubel pix) and technological benefits. It's another to support elegance and explanatoriness. Scientists have staked their prestige on not being theologians. They are publicly the hard core realists: just the facts m'am sorts of people. This is science's special virtue (at least public relations wise) and what lends it its prestige. The debate about SM in the pieces here touches the core of this conceit. And that is a good thing, for whatever SM is (and it is not one thing at all) it's not that. SO the prestige gained by citing it endlessly is attained under false premises. Not a good thing for real scientists to do.

Third, linguists would benefit from thinking about the issues raised here in physics. That's not because linguistics is like physics. It isn't. Physics is the most successful scientific domain (recall I have lots of physics envy, as should you) and linguistics is still a very young field with modest accomplishments (though there are some of a non-trivial nature as I've argued endlessly). But that's why the debate is worth our while. Because to its supreme success, physics is where large methodological issues should be hashed out, not in underdeveloped domains like ours. Of course, once hashed out there, there may be take home messages, but don't count on it.

So take a look. The pieces are short and combative. They all make reasonable points. The debate is also inconclusive. There is no one size fits all SM. There are techniques that have proved useful in some places at some times. If we are lucky they may prove to be useful again. We  muddle through and do our best. SM is the injunction to use your smarts creatively. Sadly, there is no recipe for that.


No comments:

Post a Comment