APA 2014-3: Intuitions in philosophy, pro and con

trusting_your_intuitionby Massimo Pigliucci

My series of reports from the Eastern Division of the American Philosophical Association meetings continues with an installation on the role of intuitions in philosophy, a topic that has seen much controversy recently. The chair of the session was Joshua Schechter (Brown University), and the two speakers were Herman Cappelen (University of St Andrews, Scotland), arguing that — if we look carefully — philosophers don’t actually use intuitions, or at the least not in the way alleged by critics, and John Bengson (University of Wisconsin-Madison), who argued instead that philosophers do use intuitions, and defended the practice. I will summarize the two talks in sequence, with the usual interspersed comments. I must say upfront, however, that I very much agree with Cappelen’s take, and in fact I do cite some of his work in a book currently under review by Chicago Press, on the concept of progress in philosophy, hopefully out by early 2016.

Cappelen wrote an entire book on this topic, aptly entitled Philosophy Without Intuitions [1]. He maintains that a close examination of what philosophers actually do reveals that by and large they do not rely on intuitions as data to test philosophical hypotheses or conjectures, which, incidentally, calls into question the whole initial impetus for the experimental philosophy (XPhi) program [2]. Cappelen noted that, interestingly, it isn’t just critics of philosophy, or supporters of XPhi who make that mistaken claim, so do too, on the other side, so-called philosophical rationalists, except that they think this is a good thing (I assume that the second speaker, Bengson, belongs to the latter category).

Cappelen proposed that any criticism of philosophical intuitions ought to provide a good working definition of intuitions and how they are (allegedly) deployed by philosophers, as well as empirical evidence that philosophers actually do use intuitions in said manner. Little has been done on the first count, and next to nothing on the second one, not even by XPhi practitioners, who are usually more interested in showing how laypeople’s intuitions differ from those of professional philosophers (and which I find just as interesting as showing that laypeople’s intuitions about quantum mechanics or natural selection differ from those of professional physicists or biologists).

Philosophy Without Intuitions, then, explores philosophers’ actual use of the vocabulary of intuition, empirically finding that they by and large do not in fact use intuitions as data to test their hypotheses. The book also investigates the practice of thought experiments and how they work, since this is the area of philosophical inquiry that most often is brought up in the context of the use or misuse of intuitions. Cappelen observed that this is a messy and difficult type of exploration, as the literature is very large, spanning a broad range of philosophical topics, stretching for literally more than 2000 years. Again, however, he found very little support for the alleged use of intuitions-as-data in philosophy.

David Chalmers (he of the p-zombies [3]) has countered that there are many judgments in philosophy that are justifiable and yet not broadly inferential in nature — those would be the  “intuitions” in question. Paul Boghossian has also been a critic of Cappelen, claiming that almost everyone accepts, for instance the famous twin-earth judgment arguing for semantic externalism [4], and yet that it is not inferential, it is “intuitive.”

But Cappelen thinks that it is unwarranted to label something “intuitive” just because it is not broadly inferential. Another possibility, for instance, is that sometimes people think they are justified (by reason, not intuition) in arriving at a judgment (say, about p-zombies), but they are really not. However, in the case of the twin-earth thought experiment proposed by Hilary Putnam, Cappelen maintains that there is in fact a broadly inferential (again, as opposed to intuitive) justification of the conclusion of the thought experiment, which explains why most philosophers consider Putnam’s judgment correct.

At this point Cappelen introduced his concept of Socratic knowledge [5]: this is implicit knowledge that people recognize as correct when presented with it. For instance, all people in the conference room had a vast knowledge base about reference, because they use it constantly. And yet, most of them would not be able to articulate it on the spot, certainly not in any sophisticated manner. The same is true for agency, belief attribution, knowledge attribution, and so on — just in virtue of being agents with beliefs, knowledge, etc.

Now, making tacit knowledge explicit and generalizing about it can be surprisingly hard. But the idea is that lots of cases like the twin-earth thought experiment involve the deployment of people’s tacit knowledge, which means that they are (justifiable or not) inferences, contra Chalmers.

Cappelen proposed — correctly I think — that the idea of Socratic knowledge so described is not even in the neighborhood of the kind of things that either philosophers or psychologists call intuitive, contra to what is maintained by some XPhi-ers. And of course these judgments are also not un-empirical, as tacit knowledge often involves quite a bit of empirical input. Intuitions (in the philosophical sense) are also not “spontaneous,” coming about without reflection on the part of the agent (as it is, in fact, the case for intuitions in the psychological sense).

Back to Boghossian’s objection about twin-earth, which is representative of a lot of contemporary thought experiments in philosophy. That too, according to Cappelen, is a case of deployment of Socratic knowledge. He thinks the reason so many people seem to agree on the central judgment for Putnam’s twin-earth judgment (whatever one takes that to be) is not merely sociological (i.e., Putnam was a highly influential philosopher, people wanted jobs in analytic departments, etc.), but based on the fact that the thought experiment prompts people to reflect on their Socratic knowledge of reference, a reflection that (usually) prompts agreement with Putnam’s conclusions.

But aren’t cases like twin-earth and p-zombies not so “far out,” i.e., artificial, that they cannot possibly be cases of Socratic knowledge? Don’t they require a (highly controversial) epistemology of counterfactuals [6] in order to be judged? Cappelen rejects that need, which would in turn require some type of appropriate modal logic, and thinks that the category of “far out” is simply not at all helpful. All there is are cases that are difficult to judge, some of which are “far out” and some are actually not; moreover, one can easily construe very ordinary hypotheticals about which it is also difficult to render a judgment, for instance because they present us with reasonable, and yet conflicting, epistemic or ethical judgments (in other words, far-outness doesn’t track the difficulty of such cases).

As I said at the beginning, I found Cappelen’s presentation thoroughly persuasive, and having read some of his work I am actually convinced that the case of philosophical intuitions has been vastly overstated, ironically both by supporters and detractors of the practice.

Nonetheless, let us briefly take a look at Bengson’s talk in turn. According to him, intuition may play a role in philosophical understanding, an idea perhaps first pursued by Aristotle. Specifically, intuition may play a role as “sense maker,” one that cannot be fully accounted for by reason. (No need to read anything mystical in this, by the way. In psychology, for instance, intuitions are the result of sub-conscious  processing of information, and as such they are not necessarily, or entirely, open to scrutiny by reason.)

Intuition for Bengson results in a non sensory, conscious state of mind in which “it strikes one that p,” e.g., it strikes me that killing an innocent is wrong, or that identity is transitive. People sometimes speak of these cases as “just seeing that.”

This, of course, does not mean that one cannot argue against the validity of such states. Nonetheless, it strikes me that both examples are far from optimal: one may perhaps begin with the sort of strong pull described by Bengson, but if challenged one would have to explain oneself, which immediately moves us away from the alleged intuition as determinative of the judgment, and closer to Cappelen’s idea of Socratic knowledge.

Bengson suggested that understanding is a mental state in which someone grasps something. “The house burned down because of faulty wiring,” for instance, or “the second Newtonian law is F=ma.” It is the actual grasp (epistemic), not just the sense of grasping (psychological), that counts as understanding, since one could have the impression of understanding, but it may turn out that that impression was incorrect.

Understanding is — according to Bengson — more demanding than knowledge, since knowledge can be achieved by simple rote memorization. I may know that quantum mechanics tells us about virtual particles coming in and out of existence, for instance, but I don’t necessarily understand how that is possible. Understanding, then has a kind of intelligence and robustness that simple knowledge lacks. (Here clearly the author was not using the Platonic meaning of knowledge as justified true belief.)

The question then, is what are the sources of one’s understanding, again from an epistemic, not a psychological, perspective. The idea is that intuition may be one such source, although of course a story is still needed as to how, exactly, intuition may lead to understanding, as obviously understanding can come in a variety of ways, including acquisition of new knowledge, memory recall of previous knowledge, or even direct physical manipulation of things (e.g., understanding how to use a hammer by, in fact, using it).

Toward the end of his talk, Bengson gave the example of someone who has to count the total number of fruits (say, a mix of apples and oranges) in a basket, and initially thinks that the order of the counting is important (e.g., first apples, then oranges). But she then immediately sees (intuits) that, in fact, the order is irrelevant, that the addition will yield the same result regardless. Bengson also gave a couple of examples from metaphysics, but since those are inherently controversial, and to say that “we have the intuition that” in metaphysics comes perilously close to begging the question, I will skip them.

Again, although Bengson’s presentation was interesting, I did not find his arguments compelling. Take his example of the counting of apples and oranges: is this really an instance of intuition? Why is it not rather the outcome of a quick reflection on the nature of addition, perhaps drawing on the subject’s Socratic knowledge of what addition is and how it works?


Massimo Pigliucci is a biologist and philosopher at the City University of New York. His main interests are in the philosophy of science and pseudoscience. He is the editor-in-chief of Scientia Salon, and his latest book (co-edited with Maarten Boudry) is Philosophy of Pseudoscience: Reconsidering the Demarcation Problem (Chicago Press).

[1] Philosophy without Intuitions, by H. Cappelen, Oxford University Press, 2014.

[2] Experimental philosophy is not an elephant, by M. Pigliucci, Rationally Speaking, 15 March 2013.

[3] The zombification of philosophy (of mind), by M. Pigliucci, Rationally Speaking, 29 July 2008.

[4] Twin Earth thought experiment, Wiki entry.

[5] The term is obviously inspired by one of the Platonic dialogues, the Meno, in which Socrates poses a mathematical puzzle to a slave, showing that the slave knew the answer even though he had not actually studied mathematics and was not aware of his knowledge. Of course Plato interpreted this as the result of recollections from previous existences, while Cappelen thinks Socratic knowledge is a type of implicit knowledge acquired through practice.

[6] The logic of conditionals, by Horacio Arlo-Costa, Stanford Encyclopedia of Philosophy, 2007.


Categories: commentary

Tags: , ,

57 replies

  1. Joshua, I’ll read it, but the very first sentence of the abstract is problematic:

    “The debate between compatibilists and incompatibilists depends in large part on what ordinary people mean by ‘free will’”

    Why? I don’t think it does, at all. That would be like saying that the debate in physics between different interpretations of quantum mechanics depends in large part on what ordinary people mean by “quanta.” No, it doesn’t.


  2. SciSal wrote:

    Joshua, I’ll read it, but the very first sentence of the abstract is problematic:

    “The debate between compatibilists and incompatibilists depends in large part on what ordinary people mean by ‘free will’”

    Why? I don’t think it does, at all. That would be like saying that the debate in physics between different interpretations of quantum mechanics depends in large part on what ordinary people mean by “quanta.” No, it doesn’t.


    There’s the further problem that surveying people is not the way one determines what the meanings of words are or even what their correct common usage consists of. Internally represented stereotypes–as reported to a questioner–are, at best, *one* element in what Putnam referred to as the “meaning vector.”

    I frankly find the whole X-Phi adventure Quixotic — when are philosophers going to give up the fantasy that we are some kind of rarefied scientists?


  3. Why to call a type of knowledge according to a Socrates who died more than 24 centuries ago? Socrates was the first to count fruits? Really? Did he invent vegetables, too? Is it not a bit similar in spirit to claiming that Jesus invented love? A form of celebritism?

    Did not human beings evolve thinking over millions of years, and achieved modern performance at least 50,000 years ago? (Or better: brain size has decreased recently, maybe from life becoming too easy from abundance of food one does not have to chase down.)

    Those who don’t need intuition, and deduct everything, are they not like gods? Or the average fully convinced Jihadist?

    What’s thinking? There are roughly two opposite extremes: thinking as when running down a mountain side, bounding from rock to rock, when the mind is on the edge of life. Or thinking, spilled in an armchair, as in pondering new, delicate thoughts in domains human beings never considered before.

    As it is the same brain doing both, the same basic mechanisms have to be at work.

    And what does thinking involve? Neuronal networks.

    Neuronal networks are the embodiment of logic. Early computer work was made in analogy with what was known from neuroanatomy at the time. Given some inputs in a network, computation can occur, and a neuronal output, say to motor neurons, comes predictably out. That’s how we know how to walk, swim, eat with a fork.

    The question then arises of how the neuronal networks are built and evolve. Or how, in the heat of the moment, a neuronal network branch is selected over another (say when playing chess, or when bounding from that rock to that rolling and slipping scree, with a deadline, a literal deadline, measured in hundreds of a second, not a leisurely academic deadline measured in weeks.) .
    This processes cover what is commonly called “intuition” (or wild guessing, the exact term is of no import).

    Brain science has progressed much. We know dendrites, oligodendrocytes, and other glial cells (which build their own –emotional?- networks), and complicated geometrodynamics inside synapses are all involved in the elaboration of neuronal networks. That is, in the elaboration of logic.
    The scientific method does not differ from elaborating ever more successful networks for walking.

    Except that, instead of just walking in 3D, one learns to walk all the paths we discover in reality’s various realms.

    The philosophical method is what prompts to experiment, it is what happens after an imaginary network has been created in the brain, suggesting what the world COULD be in some ways. Philosophy is the epitome of intuition: it grabs a sparkle in the world, and makes a universe out of it.

    From what philosophy suggests, and seduces with, experiments, confrontations of one’s imaginary networks with reality, can be attempted. The results feedback on the imaginary networks, and intuition, that is, modifying such imaginary networks, will be in turn attempted.

    At this point, the Quantum, which computes teleologically, may well spur on nanostructures.


  4. Hi SciSal,

    Why? I don’t think it does, at all. That would be like saying that the debate in physics between different interpretations of quantum mechanics depends in large part on what ordinary people mean by “quanta.” No, it doesn’t.

    That is an obviously false analogy (and a preposterous compliment to those who opine on free will).

    Take the debate between Dennett and Harris on this subject. They both take the beliefs of what they term “the Average Joe” or “Everyday Folk” as their starting point. Their one area of agreement is that this alleged belief is wrong. (Yes, I know – Harris is not a philosopher, but Dennett is).

    If those beliefs are, as you say, irrelevant to the debate then why do they go on about it? Why does Dennett devote his first paragraph to something which is irrelevant to the debate?

    On the other hand, if they are saying that X is wrong, but nobody (or few people) believes X in the first place then a good deal of time has been spent arguing to no purpose.

    I doubt that you could point me to anything at all written about compatibilism or incompatibilism which does not take, as its starting point, an assumption about what most people believe about free will.

    So how did Dennett and Harris (and the others) gain their authoritative knowledge about what people like me believe about our own volition?

    Hi Aravis,

    There’s the further problem that surveying people is not the way one determines what the meanings of words are or even what their correct common usage consists of. Internally represented stereotypes–as reported to a questioner–are, at best, *one* element in what Putnam referred to as the “meaning vector.”

    But Dennett and Harris, for example, make very explicitly reference to what “the Average Joe” or “Everyday Folk” mean by free will. But it is not quite clear what they think that we “everyday folk” believe on the subject or how they came about this knowledge.

    The position is clear. Either the alleged belief of “everyday folk” is important to the point they are making, in which case they should be able to state what this belief is and how they know that “everyday folk” believe it. And I cannot see how you could know this other than by some sort of survey.

    Alternately this belief is not important and they should simply make their arguments without reference to it at all.

    I am quite happy to take both of you at your words and regard every philosopher’s work on compatibilism or incompatibilism as meaningless, if it depends upon an assumption about what “everyday folk” believe about our own volitions.

    Indeed that would save quite a good deal of time.


  5. Since there is not much feedback on what I wrote, let me spend some of my word allotment to answer Massimo directly on a (very subtle!) point, he interestingly brought up:

    “SciSal wrote:

    … the very first sentence of the abstract is problematic:

    “The debate between compatibilists and incompatibilists depends in large part on what ordinary people mean by ‘free will’”

    Why? I don’t think it does, at all. That would be like saying that the debate in physics between different interpretations of quantum mechanics depends in large part on what ordinary people mean by “quanta.” No, it doesn’t.”

    I have held my fire on this one, because it’s (apparently, only just apparently) far removed from the subject at hand (in truth, it’s the core of the subject at hand, but we will leave that under wraps, lest we suffer the cries of the Beotians, as Friedrich Gauss delicately put it, when he tried to justify why he did not talk about his ideas on two dimensional curvature).

    According to me, the problem with Quantum Physics is precisely what even extraordinary people mean by “Quanta”.

    Albert Einstein thought he knew better what “Quanta” were than “every Tom, Dick and Harry. Einstein got the Nobel in physics, precisely because he explained the photoelectric effect (discovered by a Frenchman nearly a century before) with “lichtquanten” (baptized “photons 23 years later, by others).

    The crux of Einstein’s reasoning was correct, and it’s surprising Planck did not think of it (because, given Planck, the argument is obvious).

    However, I think Einstein made a mistake: he supposed much more than he needed (it may be what riled up Planck, although he was probably irate not to have thought of the crux of the idea himself, as it was, as I said, obvious).

    By supposing he knew much more than he truly did, and something he did not need at all, Einstein got philosophical on us, and erroneously philosophical, unbeknownst to himself.

    Amusingly, he spent the rest of his life walloping in the Quantum mess he had himself created, like a mighty mammoth struggling in a tar pit he dug for himself.

    I claim that this philosophical mistake of Einstein, about the nature of “Quanta”, repeated ever since, is exactly what led to the multiverse derangement syndrome. (I am not going to give the details here, it’s off subject, and technical; but I have them, and the reasoning is airtight; it has to do with confusing the teleological with the actual.)

    So yes, it does matter what “Quanta” mean, it’s the core of a lot of problems, and philosophical intuition is paramount in physics. When one does not know what one is talking about, one should stay silent (as the other one said). Einstein talked too much in 1905, and the songbirds who followed him have sung the same song way too much. Time to change the music.

    Time, for some harsh philosophical critique. In the foundations of physics, deep philosophical intuition will bring new foundations.


  6. Aravis,

    I apologize for being presumptuous about your background, for whatever reason I read your comment as simple incredulity at the possible relevance of etiology to justification, rather than simply saying that the question of the epistemic status of intuitions is a distinct (though related) question. I’m not entirely sure how I got that impression, I may have missed a word or something.


  7. Robin,

    I actually think my analogy stands. Here is why. Dennett and Harris are arguing in the public arena, not with professional philosophers, or about what professional philosophers discuss when they tackle free will. The fact that you bring this up is, I’m afraid, a reflection of exactly the sort of damage that hacks like Harris have done to philosophy and public discourse. More specifically, Dennett had to engage Harris at the level that you correctly summarize because Harris is extremely confused about free will, and mixes up entirely what he thinks most people think free will is with serious discussions in metaphysics and philosophy of mind. I hope this helps and shows that my analogy – though certainly debatable – isn’t quite so preposterous.


%d bloggers like this: