Wednesday, April 02, 2008

Thought Substitutes

I've recently lamented how logical formalisms and inductive meta-arguments risk being misused as substitutes for careful thought. It's also a major worry I have about (some) "experimental philosophy". Consider, for example, the following:
Does common sense morality assume objectivity? According to a recent study by Goodwin and Darley, most folk actually don't believe that their moral judgments are objectively true.

It's not clear how this fact is any response to the question. The question, recall, is not whether most folk believe in moral objectivism (which is all the survey can tell us). That's quite irrelevant. The real question is whether our moral practices rationally commit us to objectivism, and that is not a purely empirical question. It's a normative question, so there is simply no way to answer it without actually doing philosophy.

I don't mean to bash all attempts at empirically informed philosophy. If we are concerned with analyzing actual moral practices, it's an empirical question just what those are, i.e. what moralizing behaviours people in our society engage in. It's entirely appropriate to use empirical data as a starting point for philosophical inquiry, especially if that data is precisely what we're wanting to analyze. My point is simply that empirical work cannot substitute for philosophical analysis.

Further, it will rarely be worthwhile to ask folk for their theoretical opinions. Surveys like the above merely tell us what pop-philosophical theories are most prevalent in our public culture at present. Many people will profess a belief in relativism, for example, even if further probing would eventually reveal that they don't actually accept the implications of this view. As R.M. Hare once wrote:
If we want to find out what ordinary people mean, it is seldom safe just to ask them. They will come out with a variety of answers, few of which, perhaps, will withstand a philosophical scrutiny or elenchus, conducted in the light of the ordinary people's own linguistic behaviour (for example what they treat as self-contradictory).

So, next time you come across a study reporting the philosophical beliefs of non-philosophers, just remind yourself of the classic Onion study:


  1. Isn't one standard line of thought something like this:

    1) Philosophical theories are better when they conserve common-sense than when they reject it. At least, the burden of proof is on those who reject it.
    2) Common-sense says moral realism is true.

    So: 3) If no-one can show why moral realism is false, then it is true.

    Presumptive arguments for moral realism are pretty standard so far as I am aware. Wouldn't it be at least mildly interesting if the second premise were false?


  2. The question, recall, is not whether most folk believe in moral objectivism (which is all the survey can tell us). That's quite irrelevant. The real question is whether our moral practices rationally commit us to objectivism, and that is not a purely empirical question.

    I disagree. The question 'Does common sense morality assume objectivity?' is a factual question about the moral views of ordinary folk rather than a normative question about our rational commitment to such views. And far from being irrelevant, this factual question must be answered before we can address the normative question. To take a parallel case about ‘folk epistemology’, a number of recent studies show that, contrary to what philosophers have been assuming for decades, subjects do not unanimously deny that people have knowledge in Gettier cases. For this reason, epistemologists cannot so easily use such cases as counterevidence to the view that knowledge is justified true belief. So not only is the question you consider a factual one that, as such, must be answered experimentally, but it is one whose answer has implications for the questions which, in your own opinion, have philosophical importance.

    Your worries about trying to resolve a dispute concerning what people believe by asking people to report their beliefs are legitimate, but they call for more care in designing an experimental study rather than for abandoning experiment altogether. Moreover, this is a worry that the authors of the study you cite seem to be aware of. The authors didn’t ask their subjects a general question about whether they believe moral judgments to be objective. Instead, they presented them with various particular moral statements, and asked them whether those statements express truths or merely subjective attitudes.

  3. Richard,

    That video is a fabulous illustration. I have to admit up front to being highly sceptical of the value of experimental philosophy, but I've yet to look closely at it. However, I think you misconceive the central use to which it is usually put. You say:

    "empirical work cannot substitute for philosophical analysis."

    That may be true, but that's not typically what the experimental philosophers are up to. I take it that the aim of these empirical studies is to temper philosophical analysis. For instance, Knobe mentions that case studies on linguistic intuitions reveal cultural bias, and with respect to Kripkean philosophy of language, he says that the result show "Asian people may not share many of the intuitions on which widely accepted philosophical theories have been based."

    If true, this is interesting. It suggests that philosophical analyses are radically culture-bound. Exactly what the intended effect is, I cannot say. Are we supposed to conclude there is no objective truth? Just one culture has things right? Maybe everyone is right, and there is more than one truth of the matter? I don't know what the ex-phiers think is supposed to follow.

    But my concern is more about the methodology than the philosophical application of the data. How do we test people's intuitions, especially (as per Hare) their informed or philosophicall relevent intuitions? How do we do this in a way that circumvents the biased interpretation of the experiementer who brings to the table a highly sophisticated philosophical perspective? I just don't see how the ex-phiers can make good on what they are trying to do in the first place, at least, I am so far unconvinced.

  4. Pablo, you say:

    The question 'Does common sense morality assume objectivity?' is a factual question about the moral views of ordinary folk rather than a normative question about our rational commitment to such views.

    This appears to assume that 'common sense morality' is simply a matter of the 'views of ordinary folk' and is entirely divorceable from actual practice. But surely this is an extraordinarily odd thing to think. (It is, in fact, like thinking that folk epistemology is exhausted by people's views of how the term 'knowledge' should be applied, rather than also including information about how people actually set about trying to know things.) Someone's moral views may not adequately fit the moral practices they think necessary, the moral institutions they would defend, etc.

    If we were to look at 'folk reasoning' it would be quite odd to just look at what sort of logical systems non-logicians, who have had very little time or resources for reflection on the subject, can spontaneously produce on the spot. That might be interesting. But what you're interested in is, as it were, the internal constitution of their reasoning, and this requires also seeing what is implied by their actual rational practice, which may be very different. Similarly, if you are interested in the internal constitution of folk morality, you can't just gather views of morality from people who haven't really devoted much attention to the subject at that level; that's not folk morality, but just spontaneous reflection about folk morality from people who don't usually think about it.

    A good example, I think, can be found in studies of folk physics. If you just look at how people would describe physical systems off the top of their heads, they are often very, very wrong. But you need to look at their practice; if you do, you find that their actual interactions with physical systems aren't always reflected in their descriptions, that they behave in ways for which their explanations cannot account, and that they sometimes will be using correct or at least approximately correct physics principles in one area of life (e.g., car mechanics) but just not be seeing that the principles generalize, or that they sometimes take folk principles that are excellent approximations in one domain with which they are familiar in practical life and falsely generalize them. And so forth. If you don't look at actual practice, you can miss this entirely and are in danger of saying stupid things about folk physics. Something similar could be said about folk psychology, folk economics, etc. You need to look both at what people say and what is implied by what they do.

  5. Colin - I didn't mean to suggest that x-phi is usually misused in this way. I'm just worried about the possibility (demonstrated in the quoted snippet), so it's worth noting this risk in order to better guard against it. (The worry in your last paragraph is interesting, though.)

    Alex (and Pablo) - it depends how you interpret claims about what 'common-sense says'. Does this just mean 'what most folk are willing to profess a belief in'? Then it is uninteresting. I instead take it to mean something more like 'what common-sense platitudes entail'. Empirical work may then be helping as a starting point, like I said, to discover what the folk platitudes in an area are. (And note that these may be revealed in action as well as thought, as Brandon helpfully points out.) But a significant gap remains that can only be filled by philosophical interpretation of said platitudes, i.e. to work out what the platitudes have to 'say' about objectivity.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)