What do philosophers of mind actually do? Some quantitative data

There seems to be a widely shared sense these days that the philosophical study of mind has been undergoing some pretty dramatic changes. Back in the twentieth century, the field was dominated by a very specific sort of research program, but it seems like less and less work is being done within that traditional program, while there is an ever greater amount of work pursuing issues that have a completely different sort of character.

To get a better sense for precisely how the field has changed, I thought it might be helpful to collect some quantitative data. Specifically, I compared a sample of highly cited papers from the past five years (2009-2013) with a sample of highly cited papers from a period in the twentieth century (1960-1999). You can find all of the nitty gritty details in this forthcoming paper, but the basic results are pretty easy to summarize.

First, I looked at the methods employed in the papers from each time period. Each paper was coded into one of three categories:

  • A priori. Papers that do not make any use of results from systematic empirical studies
  • Experimental. Papers that report original experimental results
  • Relies on empirical. Papers that do not report original experimental results but do engage with the results of empirical studies from the existing literature
[Note that the phrase ‘a priori’ is being used here in a broader sense in which it characterizes all research that does not rely on the results of systematic empirical studies.]

In the twentieth century sample, the distribution of methods looks like this:


In other words, the majority of papers are a priori, with a minority relying on empirical work and none at all reporting original experimental studies.

Now take a look at the distribution in the papers from the past five years:


As the figure shows, the change in our field has been pretty radical. At this point, only a small minority of the highly cited papers are purely a priori. The majority engage in some way with empirical findings, and a surprisingly large percentage actually report original experimental results.

I then looked at the topics that the papers investigated. Within the twentieth century papers, the focus was clearly on questions about the metaphysics of mind. In particular, there were a huge number of papers on either the mind-body problem or the nature of content. These topics clearly dominated the philosophical discussion in that period.

Now suppose we switch over to the contemporary sample. There we find a surge of papers engaged in a very different sort of project. Specifically, there are a very large number of papers looking at the cognitive processes underlying people’s ordinary intuitions about philosophically relevant questions (morality, causation, knowledge, etc.). This work tends to be concerned not so much with metaphysical questions as with questions in cognitive science.

Within the contemporary sample, the number of papers pursuing this new research program is actually greater than the number pursuing all of the topics continuing within the twentieth century tradition put together. (For example, suppose we add up all the papers on the mind-body problem, the nature of content, consciousness, perception, the extended mind. The number of papers within this new research program still comes out to be higher than the number on all those topics combined .) So this new research program has at least a fair claim to be the dominant one in the contemporary literature.

I would be very interested to hear any thoughts that people might have either on this specific dataset or on recent developments in the philosophical study of mind more generally. All of the details are available within the published paper, but please do feel free to write in with comments even if you haven’t taken a look at the paper itself.

[Header image credit: TateMag]


  1. interesting paper, josh! it’s nice to see this sort of sociology of philosophy. the qualitative trend is plausible and unsurprising. but the quantitative results seem to be affected by some idiosyncratic coding choices.

    for example, your list of highly-cited “philosophy of mind” papers for 2009-13 includes 11 papers on knowledge attribution, 10 papers on moral cognition, 5 papers in intentional action judgments, five papers on free will judgments, four papers on joint action, and so on. standardly those would be regarded as topics in epistemology, ethics (moral psychology), and the philosophy of action, not philosophy of mind. looking at your supplementary materials, this seems to have happened because you have coded any paper reporting an original psychological result as philosophy of mind. that seems highly idiosyncratic. prima facie, to count any x-phi paper on any philosophical topic as philosophy of mind is to run together methodology with subject matter.

    looking further at the supplementary materials, it’s also striking that you classify any paper in the philosophy of cognitive science as philosophy of mind. that’s perhaps somewhat more defensible than the above, but it’s also not really standard practice.

    it’s also interesting to see that for a paper to count as “relies on empirical”, it suffices for it to cite a single paper from the empirical literature. one result of that choice is that a general trend toward citing more papers could be expected on its own to produce a qualitative trend toward “relies on empirical”. in addition, a general trend to at least mention empirical results even in largely a priori work will lead to some surprising classifications.

    for example, the first five papers listed on your 2009-13 list (papers in analysis on perception or consciousness by nanay, tye, rosenthal, block, and weisberg) all appear to use largely a priori methodology (they cite one or two empirical results along the way but these results aren’t focal and don’t drive the central discussion), but they come out as “relies on empirical” in your list. the trend toward papers of this sort is interesting in its own right, but it shouldn’t be overstated.

    that said, i’m sure that your qualitative results supporting a trend toward the empirical are robust enough to survive all sorts of recoding. it’s clear that there’s been a strong trend toward engaging empirical work in the philosophy of mind, and i think this is a very good thing. at the same time it’s also clear that there remains a robust body of work driven by relatively a priori considerations. i look forward to seeing more work in the sociology of philosophy to get at the quantitative trends in a more fine-grained way.

    • Thanks for all these helpful comments! I am so sorry to be late in getting back to you all. I happened to be away at a conference at the moment this post was put up, but I’m back now and very much looking forward to responding.


      Both of these points are very well-taken. Let me take them up in turn.

      You rightly point out that some of the papers that I classify as engaging with the empirical data are actually primarily concerned with more a priori issues and only discuss empirical results in, say, one specific section. Just as you suggest, the percentage of papers in the more contemporary sample that are listed as relying on empirical data would slightly decrease if we eliminated papers of this type from the category.

      Yet, at the same time, it should be noted that the *change* over time would become even more stark if we coded in this alternative way. Of the papers that do discuss empirical results but do so only in a very minor way, the vast majority are in the twentieth century sample. Thus, if we recoded those papers, the difference between the twentieth century sample and the contemporary sample would be far greater than it appears to be now.

      Second, you note that some of the papers included within the sample do not appear to be in the field of ‘philosophy of mind,’ narrowly construed. This point is also a very helpful one that is certainly worthy of further consideration. In the actual paper, I instead use the phrase ‘the philosophical study of mind.’ Clearly, philosophical work in cognitive science does constitute research in the philosophical study of mind, even if one might doubt that it falls into the area traditionally described as ‘philosophy of mind.’

  2. I always found it curious that every agrees philosophers of biology must know biology, philosophers of physics must know physics, but “armchair” philosophers of mind insist they need not know any psychology. It seems odd to insist that introspecting on one’s own mind is sufficient for developing theories about the mind, despite the fact that introspection typically yields wildly false conclusions concerning the nature and workings of the mind. Some of the best philosophy of mind, to my way of reckoning, has been done by philosophers who pay attention to developments in psychology, neuroscience, and cognitive science, and use those sciences to develop their theories. I also greatly admire philosophers who explain the implications of scientific discoveries to scientists. Scientists tend to be good at discovering things but often fail to realize their greater implications.

    • David Silverman

      There is a difference though between the kind of philosophy of mind that should be empirically informed, and the kind that needn’t be. I agree that falls into the former category, but fails to engage very much with relevant psychology and cognitive science, and that makes it weaker. But the idea that all philosophy of mind should engage with science seems wrong: unlike philosophy of biology etc., philosophy of mind doesn’t claim to be philosophy *of* a science. There is lots you can find out about the mind without empirical work.

    • I’m not sure this is quite fair to the philosophers of mind that do work that falls into the apriori category. The philosophers of mind that I read who approach their topics using traditional philosophical methods don’t do their work by simply introspecting and recording what they find.

      I’d think that there’s room for a number of different approaches, particularly because the people who pursue these different approaches often are out to address different questions. But I would be glad to hear more from you about how you think empirical work would improve the best work being done using traditional philosophical methods.

  3. James Genone

    It’s very interesting to see this comparison, as it illustrates a trend I personally welcome! However, I also think your methodology may overstate the decrease in emphasis on a priori philosophy of mind. In addition to the reasons Dave mentions, it strikes me that citation patterns for a priori papers may differ from those that are more empirically focused, with it taking a somewhat longer period of time for discussion of a priori papers to build relative to empirical topics, which seem to move more quickly (this is just an impression based on following the literature in philosophy of perception and ex-phi semantics, but I wonder if it might explain some of your data). Moreover, many empirical papers are published in journals with relatively quick times to publication, which might speed up the accumulation of citations in that area over short timeframes. I think the results would look different if you selected 2003-2013 as your comparison period for contemporary work. I did this quickly on Web of Science for Mind & Language and Mind, and it did seem to me at a quick glance that it produced a higher proportion of a priori papers than your list included.

    • James,

      Nice point. This certainly does seem possible, and I guess the only way to know for sure would be to wait a decade or so and then look back and see which of the 2009-13 papers were most highly cited then. Just as you say, it might be that we will find at that point that things have shifted a bit toward the a priori side.

      But I actually think there are some specific reasons to make the opposite prediction. Most of the change observed in the present dataset seems to be driven by a demographic shift. A whole bunch of truly great a priori philosophers of mind have recently retired, and a whole bunch of more empirically oriented philosophers have just entered the discipline.

      If this demographic trend continues, we should expect to find a corresponding trend in the citations of the 2009-2013 papers. As ever more a priori philosophers of mind retire and ever more empirically oriented philosophers enter the discipline, we should expect an even greater tendency in the future for higher rates of citation for the empirically oriented papers being written now.

      But of course, this itself is an empirical question, and the only way to know for sure is to wait a while and see what happens.

  4. Hi Josh, thanks for this post, and the really interesting paper. I had a worry similar to James Genone’s about how differences in citation practices among the various sub-fields could account for some of what you are reporting here. I’d love to hear what you think about this.

    I also wanted to observe something that’s more evident in the full paper than in your post, namely the increased focus on specific aspects of mind and cognition, i.e. things like perception, concepts, consciousness, etc. in contrast to the more general metaphysical issues. This shows up clearly in the relative distributions of topics that you report:

    For the twentieth century papers, the distribution of topics was as follows:

    The Mind–Body Problem (17 papers)
    The Nature of Content (14 papers)
    Self-Knowledge (6 papers)
    Perception (5 papers)
    Consciousness (4 papers)
    Intention (3 papers)
    Desire (2 papers)
    Qualia (2 papers)
    Memory (2 papers)
    Theory of Mind (2 papers)
    Emotion (2 papers)
    Fictional Emotions (2 papers)

    Aficionados of philosophy will recognize many of these topics as aspects of a closely related set of issues in the metaphysics of mind.

    The distribution of topics in the contemporary papers, however, is radically different:

    Perception (14 papers)
    Knowledge attribution (11 papers)
    Theory of mind and social cognition (11 papers)
    Moral cognition (10 papers)
    Extended mind (7 papers)
    Consciousness (7 papers)
    Mind perception (6 papers)
    Intentional action judgments (5 papers)
    Free will judgments (5 papers)
    Embodied cognition (4 papers)
    Joint action (4 papers)
    Psychology of language (4 papers)
    Causal judgments (4 papers)
    Foundational issues in neuroscience (3 papers)
    Enactivism (3 papers)
    Introspection (3 papers)
    Trust (3 papers)
    Attention (3 papers)
    Self-Knowledge (2 papers)
    Mental causation (2 papers)
    The Nature of Content (2 papers)
    Concepts (2 papers)
    Intention (2 papers)
    Phenomenal concepts (2 papers)
    Knowing how (2 papers)
    Memory (2 papers)

    As you go on to say, “Contemporary papers in the philosophical study of mind are concerned primarily with the attempt to investigate specific aspects of cognition. They tend not to be aimed at addressing metaphysical questions but instead to be engaged in an effort to use systematic empirical data to gain a better understanding of how these aspects of cognition actually work.”

    Partly for reasons similar to some of what Dave Chalmers observes, I’d suggest that this (very motley) “aspects of mind” program has at least as much claim as x-phi to be the dominant contemporary paradigm.

    • John,

      What you say here is clearly correct. Suppose we group together all of the papers that take up questions about specific aspects of cognition and how they actually work. Then this broad category seems to be indisputably the dominant force in the contemporary philosophical study of mind. Nothing else could even compete with it.

      The point I was trying to make in my original post was that, within this broad category, there is a specific subpart that seems to have been especially active over these past five years. But perhaps you are right in emphasizing the broader category rather than the specific subpart.

  5. Thanks Josh, very interesting. In response to Dave’s comment that some papers categorized here as ‘relying on empirical stuff’, including my Analysis paper on attention, only cite a couple of empirical findings and are really a priori papers: My paper was originally much more empirical, but then I thought that if I want a shot at publishing it in Analysis, I should take out all the empirical bits, so I did so. And I’m sure I’m not the only one who was/is thinking this way. So the trend you are reporting is very encouraging – maybe next time I send something to Analysis, I won’t take out the empirical stuff…

Comments are closed.

Back to Top
%d bloggers like this: