What philosophy is "good for"

There is a very strange discussion popping up from time to time in some of the blogs that I read, where somebody will claim that philosophy is useless because it has not contributed anything to our understanding of the natural world or "to society" in recent times. Although I think that the charge of scientism - empirical science is all we need, every other field of scholarship is useless - is usually, mostly a straw man, it seems that there is a vocal minority of people who really think like that.

For starters, to the degree that this is about philosophy contributing to our understanding of the natural world this is clearly the wrong question to ask. What have bus drivers, as a profession, lately contributed to that endeavour? Nothing; but that does not mean the profession is useless, merely that it has a different job. Conversely, everybody who does contribute to our understanding of nature is by definition a scientist, so the claim that only scientists directly contribute to that understanding is true but trivially so.

The question could then rephrased more generally as: what do philosophers actually do? What is philosophy good for?

Now I am not a philosopher myself, and the question would perhaps be best answered by a member of that profession. But it so happens that just before I saw that remarkably nihilistic discussion about the value of philosophy I saw a use of philosophy outside of the academic context that provides a very good example of the kind of thing that the field is "good for".

In this post on his website Why Evolution Is True, Jerry Coyne had taken a completely consequentialist stand on the issue of punishment:

If you're a determinist about behavior and a consequentialist about punishment, as I am, then you punish people only if it's for the good of society. (My view is that at the moment of the slaughter, Gutierrez had no "choice" to not kill the birds.)

And there are three social goods to come from punishments like incarceration: deterrence of others, sequestration of someone who could be dangerous to society, and reformation of a criminal so he doesn't repeat his offense when freed. All three of these apply to Gutierrez: jailing him will probably deter others who want to kill wild animals, people who do that tend to be murderous psychopaths who could kill again (maybe people next time) and so need to be put away, but such people may be susceptible to reformation [...].

If none of these reasons obtain, there's no reason to imprison anyone; or can you give me one? But surely deterrence and sequestration apply in most cases--though not capital punishment, which data show isn't a deterrent. And if no social good results from imprisonment, in what sense would Gutierrez still "deserve" to be imprisoned? To satisfy a sense of vengefulness? That, to me, is not a good reason, for it caters to our baser instincts--the same instincts and feelings that make people favor executions. So, if Gutierrez can be reformed, poses a danger to society, or can be a deterrent to others, yes, he "deserves" punishment. But he doesn't deserve it just because he needs to be "paid back" for what he did.

In short, locking somebody up is to be justified (only) by good societal outcomes, while that person "deserving" to be locked up is not a just and reasonable concept (because JC believes that the existence of cause-and-effect is incompatible with personal responsibility). To this the commenter cjwinstead replied as follows:

Suppose we have strong justification to believe that punishing Gutierrez's mother will satisfy the goals of deterrence and reformation; and keeping her hostage would be as effective as sequestration (maybe he really cares a lot about his mother). If we have evidence that this will be more effective in those goals, is there any reason not to punish her? What if she gladly volunteers to receive the punishment on his behalf? I would say that Christian Gutierrez deserves to be the subject of punishment in a way that his mother does not. Proxy punishments do happen in our justice system, and they are arguably effective at deterrence and reformation. Should they be supported if they work?

This, right there, is one of the things that philosophy is "good for". This is not science, obviously, as no empirical data are involved in any but the most remote ways. What cjwinstead has done is propose a thought experiment - a classical method of analytic philosophy - to lay bare our instincts about something (here, that we would consider punishing the mother unjust), to start a conversation about where those instincts come from and what, if anything, they mean to us, and perhaps in particular to demonstrate the absurd consequences of a position (here, basing moral philosophy entirely on consequentialism).

Of course, you may disagree with cjwinstead in this instance. What is more, while his comment sparked a very long discussion, nearly all the people who replied to it missed its point in a way that is somewhere between spectacular and hilarious. But again, this is the kind of thing that is philosophy, and it is useful and necessary to hash out issues that cannot be adjudicated based on empirical studies alone.

We may do science to find out if deterrence and reformation work or not, but science alone cannot necessarily tell us if we should prefer consequentialism to deontology, for example. And even if some scientismist were to argue that it can, using analytic philosophy to point out an internal contradiction or absurdity in an argument still saves us the major investment of conducting a large scientific study to test it.