The Problem with Experimental Philosophy

24 06 2011

Let me preface this post by saying that I heartily applaud the injection of empiricism into philosophy. If one adopts the philosophical starting point of investigating the natural world, one needs to start with reliable observations of the natural world to spark off and inform philosophical enquiry.

However, while I also applaud the intentions of experimental philosophy (x-phi), I fear that it can often be conducted in a manner that is arse-backwards. This not only undermines x-phi and its results, but it foregoes an alternative approach to integrating the philosophical with the empirical that might be even more fruitful than the much of the current x-phi research.

Let me begin by going back a couple of steps and drawing a (somewhat coarse) distinction for you, one I’ve drawn before. There are two types of philosophers in the world: pure and applied.

Pure philosophers are wound up and set loose on any question the mind can conjure. They’re unbounded by earthly concerns and can roam the full breadth of the conceptual world unimpeded. They might occasionally tackle real-world questions or stumble upon real-world answers, but that’s not their primary pursuit.

These are our metaphysicians, epistemologists, logicians and our metaethicists, amongst others. Even much of moral and political philosophy is, in my opinion, pure philosophy. It might start with a problem that sprung forth from the real world, but from that starting point they’re free to explore where the thinking takes them – e.g. what would moral properties look like; what are the limits of moral responsibility; what is justice; what is sovereignty? But in doing their conceptual wrangling they develop also new tools of thought.

Then we have applied philosophers (precious few, I might add). These are explicitly engaged in searching for answers to real-world questions, using the tools of thought developed by pure philosophers and, crucially, embracing empirical evidence to guide their gaze. Sure, they will develop and adapt their own tools as well, but many of their key tools are derived from pure philosophy.

So applied philosophers start with empirical information about real world issues then apply the tools of thought developed in pure philosophy to understand, characterise, carve up, tear down and wrangle ideas in order to make better sense of real world problems.

Now we come to x-phi, which if you look at it through the lens of this distinction, is often a somewhat strange beast. For much x-phi doesn’t seek to use the tools of pure philosophy to help understand empirical issues. Instead much x-phi uses empirical tools to better understand philosophical issues. And I wonder how enlightening this can be.

Take the research conducted by Eddy Nahmias & Dylan Murray on lay intuitions about free will. In this paper, Nahmias and Murray interpret x-phi research done into people’s intuitions about free will and moral responsibility, concluding that incompatabilist intuitions are problematic, and suggesting this amounts to a challenge to incompatabilism in general.

To me, this is arse-backwards.

Experimental studies can, indeed, tell us many useful things about our intuitions concerning many curly and counter-intuitive things. It might even find that many of our intuitions are problematic in themselves. We might find that our intuitions are contradictory, or implications that we wouldn’t be happy with, or even lead to absurdities. These are philosophical reflections on real intuitions.

But I’m sceptical about whether intuitions can’t tell us terribly much about what is the correct answer to questions about free will. In fact, I’d suggest the Nahmias and Murray study simply reveals that our intuitions are often unreliable and that philosophical intuitionism in general is deeply problematic. But didn’t we know this already?

It seems to me that such x-phi, which seeks to use empirical tools to study philosophical questions, is of limited utility unless we think our intuitions about philosophical questions are good grounds for holding certain philosophical conclusions. And I’d suggest that most philosophers would (or should) be wary of such bald intuitionism; for example, we know most people have intuitions about time and space that don’t square with our best theories about how time and space actually work.

Rather, I’d like to see x-phi turn around and function in the opposite way: taking the tools of philosophy and applying them to real world, empirical problems. So instead of probing our intuitions to learn something about free will, we can use philosophical ruminations about free will to tease apart what people say and think about free will, perhaps highlighting problems in their intuitive reasoning. That, at least, would be arse-forwards.




5 responses

25 06 2011

Very much enjoyed the article (that distinction needs more advertising!) up to the last paragraph where you lost me at the suggestions bit. Perhaps an example?

26 06 2011
Dan Hicks

I agree with your criticism of x-phi, but not your version of the pure/applied distinction. (I look at several different versions of the pure/applied distinction for science over the course of my dissertation, plus I’m most certainly in the camp of applied philosophers, so I Have Opinions on this.) The phrase `using the tools of thought developed by pure philosophers’, plus the overall structure of the post (pure philosophers are listed first, and not characterized by reference to applied philosophers) suggests that pure philosophy has a certain kind of at-least-quasi-epistemic priority over applied philosophy. Roughly: first the pure philosophers do their thing, then after they’re finished the applied philosophers pick up the results and straightforwardly apply them. Or, a little more precisely, the work of applied philosophers does not (significantly) influence the work of pure philosophers; influence runs only one way, from pure to applied.

The parallel claim is spectacularly false with respect to both natural and social sciences; look, for example, at the steam engine and thermodynamics, or much of the history of medicine. In the history of our discipline, work that we would consider both pure and applied are deeply intertwined in, for example, Descartes, Locke, Hume, Kant, Mill, and even to an extent Carnap and Hempel (not to mention Dewey and Neurath). More recently, I know several applied philosophers of science who are very dismissive of recent work in pure philosophy of science precisely because it (the pure phil sci) is all but useless for tackling the kinds of problems the applied folks are interested in: Bayesian formal epistemology and the metaphysics of QED aren’t all that useful for developing guidelines for science in public policy. It’s reasonable to think that some of the grad students of these applied philosophers will be assigned to write dissertations developing more applicable pure philosophy. (I suppose that is more-or-less what I ended up doing.) The result would be pure philosophy heavily influenced by applied philosophy.

26 06 2011

1) First, a one-liner about distinguishing between theory and application: I’m reminded of the idealist comment, “Well, even if works in practice, does it work in theory?”

2) You’re right that the x-phi method (as currently practiced) tells us more about people’s intuitions or attitudes towards certain philosophical questions that it does to actually answer the questions. It’s difficult to see how (as currently practiced) studying, tabulating and quantifying attitudes/intuitions resolves — or can resolve — any philosophical debate. Even if all seven billion people on the planet agreed “there is no free will” (or the opposite), we could all be wrong. And even if we realized that, the tabulation would not tell us anything about where the error is. X-phi proponents hope, in such a case, this might tell us something deep about human nature — such as “we are incapable of believing in free will.” But that would be a useless conclusion, especially if we were all wrong.

3) The X-phi movement in this decade, as I understand it, centered first around the issue of free will and that is the focus of Nahmias & Murray cited above. X-phi was conceived as a way to bring psychological science into the philosophical debate — first about free will, then gradually on consciousness in general. The psychological approach built into the x-phi method practically precharges the discipline towards epistemology. To me, the jury is still out — way out — on whether x-phi psychological study can tell philosophers anything we don’t already know (or disagree about). Tabulating folk wisdom in itself is interesting, but it resolves few classic debates. X-phi needs more depth.

Now back to the distinction between philosophies, pure and applied. In Dan’s comment above, I see the mirror of a shadow of an outline in his analysis which I’d like to question a bit. C.P. Snow once drew a distinction between “pure science” and “applied” (ie, engineering), but saw not much practical difference in them, since both were iterations of scientific method. Dan seems to be using that pure vs applied science model in his comment (thus my shadow metaphor). But — just musing here — I doubt that the relationship between pure and applied philosophy actually mirrors the relationship between science and engineering. The label terms are the same, but the difference between theory and practice across the (ahem) magisteria just seems to me to be comparing different types of things. In philosophy, it’s all ideas and the physical realm (if you can call it that) ultimately is policy, more than objects.

Or maybe I’m just being philosophical.

1 07 2011
James Gray

I think many philosophical experimentation is attempted to be used to debunk intuitionism. I think that is also a problematic use considering that the prejudice of people in general is not supposed to be equivalent to intuitions in philosophy.

12 06 2013
Is Experimental Philosophy Bad Science? | The Partially Examined Life Philosophy Podcast | A Philosophy Podcast and Blog

[…] not going to criticize X-Phi’s approach to intuitions.  If you want that, check this out.  I want to look at its experimental approach.  You don’t have to be a scientist to […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: