Empirically-informed philosophy of logic?

In my previous post on implicit biases, I noted that such biases are a special case of a more general phenomenon, namely the fact that our actual reasoning and judging mechanisms seem to deviate significantly from the traditional picture of ‘human rationality’, as widely documented by research in psychology and behavioral economics, among other areas. As it turns out, our perceptual, cognitive and decision mechanisms seem to be far less accurate than we had taken them to be until this body of research started to emerge. The booming field of behavioral economics has been showing that our decision and choice mechanisms deviate significantly from the traditional construal of rationality (see for example this great TED lecture by Dan Ariely). Psychological research has been demonstrating that our perceptual capacities are not nearly as accurate as we take them to be (as e.g. the strange phenomenon of the invisible gorilla). When it comes to reasoning, the Piagetian idea that human cognition evolves naturally towards abstract, ‘logical’ reasoning (that is, following the canons of classical logic) is by now almost entirely discredited; our actual reasoning mechanisms do not in any way appear to follow the canons of logic as traditionally construed.

What does this all mean for philosophy? I submit that large portions of philosophical inquiry must be seriously reassessed in light of these findings. In epistemology, for example, skeptical arguments are often dismissed on the assumption that, in spite of the extreme cases where our senses, for example, do fool us, our perceptual capacities are largely reliable. But this is not what emerges from research on the psychology of perception; we are systematically making very rough approximations and thus essentially just ‘guessing’ when making perceptual judgments.

However, I’m no epistemologist, so I leave the work of reassessing traditional epistemology in light of these findings to more competent colleagues. My own research is on the philosophy of logic, and so I ask myself what the findings on the psychology of human reasoning may mean for logic as an enterprise. Of course, the easy answer is that they do not mean much, on the assumption that logic concerns correct, idealized reasoning, and thus not how people actually reason. Hadn’t Frege showed once and for all more than a century ago that any psychologistic account of logic is utterly flawed? Well, things are not so simple – see for example this great paper by Pelletier et al. on the current situation of psychologism in logic. In fact, just as naturalism is currently a very popular approach in philosophy, psychologism in logic has several recent proponents (Maddy, Sher), in particular some who adopt an evolutionary perspective (Cooper). The idea would be that ‘logical thinking’ would have emerged as a result of an evolutionary process in the human species, which would have increased its fitness. One of the appeals of such accounts is that they seem to offer a solution to the traditional realist vs. idealist divide; from an evolutionary point of view, logic would have both a realist grounding (in terms of adaptation to the environment) and an idealist nature (it would be hardwired in our cognitive apparatus).

It all sounds very nice. But what is the real import of a psychologistic account of logic, given the experimentally observed discrepancy between actual patterns of human reasoning and the canons of logic and deduction as traditionally construed? If logic is indeed ‘in our heads’, it is odd, to say the least, that we do not seem to be making systematic use of its principles in our reasoning practices. What the experimental data show is not so much that we are ‘irrational’, but rather that the game we play when reasoning is simply not the logic/deduction game as traditionally construed. Clearly, without empirical corroboration, such naturalistic-psychologistic accounts of the development of logic are no more than ‘just-so stories’. But then, where on earth does logic (as traditionally construed, in particular with respect to monotonicity) come from?

The point I wish to make here by discussing the specific case of psychologistic-naturalistic accounts of logic and logical reasoning is to argue in favor of an empirically-informed philosophy of logic. If as logic is to have any bearing at all on human reasoning and cognition (which is in itself already a substantial thesis), we cannot be content with idealized, armchair pictures of human reasoning and human rationality. There is already a substantial amount of data amassed by research in the psychology of reasoning, and although there is still quite a lot of controversy among the researchers working in the field, the data are there, waiting for us philosophers to incorporate them in our discussions. But notice that I am not claiming that philosophical issues (about logic or otherwise) should be reduced to empirical issues; rather, I am making the weaker claim that philosophical analysis should be informed by empirical data. Most likely, the data will underdetermine the solution to a given issue, so there is still plenty of room for philosophical theorizing. But those explanations that are outright incompatible with the data are to be excluded from the space of possible solutions.

Of course, I am not the first one to make such claims. G. Harman, in particular, has been arguing for years that the actual patterns of human reasoning and the canons of logic are in massive disagreement with each other. Still, mainstream philosophy of logic remains by and large oblivious to empirical data on human reasoning. Now, in an effort to make philosophy of logic a bit more empirically-informed, I’m hosting a workshop in December with the (lengthy) title ‘From cognitive science and psychology to an empirically-informed philosophy of logic’. (So yes, this post is essentially a plug for the workshop!) The idea is not only to discuss specific philosophical issues pertaining to logic on which empirical data about human cognition may have a bearing, but also to discuss the very idea of an empirically-informed philosophy of logic in this particular sense. Is it a fruitful enterprise? I think it is, but there is no doubt that much remains to be done before it becomes a serious contender in philosophical debates.