Whose Intuitions Matter?

Maybe not all philosophical intuitions are created equal. This rather straightforward idea has become increasingly attractive, in part because it seems to offer an equally straightforward response to experimental philosophy.

One particularly popular version of this response has become known as the expertise defense.¹ According to the expertise defense, our interest in philosophical intuitions should be understood to be an interest in philosophers’ intuitions. Kirk Ludwig (2007) provides a wonderfully straightforward articulation of the expertise defense in his discussion of the role played by philosophical intuitions in the philosophy of language:

We should not expect antecedently that untrained subjects should be in an especially good position to give judgments in response to scenarios involving difficult questions about the semantics of proper names, for this is a domain of considerable complexity where our ordinary vocabulary is not especially precise. We should instead expect that the relevant experts in the field of philosophical semantics will be better placed to give answers which focus on the right features of the cases and what they are supposed to be responding to... What is called for is the development of a discipline in which general expertise in the conduct of thought experiments is inculcated and in which expertise in different fields of conceptual inquiry is developed and refined. There is such a discipline. It is called philosophy.

Education and experience are also emphasized by Steven Hales (2006), who writes:

Intuitions are and should be sensitive to education and training in the relevant domain. For example, the physical intuitions of professional scientists are much more trustworthy than those of undergraduates or random persons in a bus station. Scientists have and rely on physical intuitions, intuitions that are trained, educated, and informed and yet are good indicators of truth for those very reasons. In the same way, the modal intuitions of professional philosophers are much more reliable than either those of inexperienced students or the “folk”. (p. 171)

The common thought seems to be this. We should be interested in expert philosophical intuitions rather than folk philosophical intuitions. After all, philosophers have better concepts and theories, or at least a better understanding of the relevant concepts and theories, have thought long and hard about these concepts and theories, and have been trained in how best to read and think about philosophical thought experiments that call upon us to apply these concepts and theories. Surely, this makes expert philosophical intuitions more theoretically valuable than folk philosophical intuitions.

While this is an attractive idea, it turns out to be quite difficult to determine who has expertise about what and when. It seems that only certain kinds of training help improve task performance and, even then, only for certain kinds of tasks, and there is reason to worry that philosophical training doesn't seem to be the right kind of training nor does philosophical thought-experimenting seem to be the right kind of task (see, e.g., Shanteau 1992, Ericsson et al.

One reason that we might have for thinking that expert philosophical intuitions are more theoretically valuable than folk philosophical intuitions is that philosophers have better concepts and theories, or at least a better understanding of the relevant concepts and theories. Peter Singer (1972) advances this view in his discussion of moral expertise, writing:

[I]t would seem that the moral philosopher does have some important advantages over the ordinary man... his specific experience in moral philosophy gives him an understanding of moral concepts and of the logic of moral argument. The possibility of serious confusion arising if one engages in moral argument without a clear understanding of the concepts employed has been sufficiently emphasized in recent moral philosophy and does not need to be demonstrated here. Clarity is not an end in itself, but it is an aid to sound argument, and the need for clarity is something which moral philosophers have recognised. (p. 117)

Hilary Kornblith (2007) also emphasizes the benefit of having appropriate theoretical understanding, writing:

Theory-informed judgments in science may be more telling than the judgments of the uninformed because accurate background theory leads to more accurate theory-informed judgment. The uninformed observer and the sophisticated scientist are each trying to capture an independently existing phenomenon, and accurate background theory aids in that task. Experts are better observers than the uninitiated. If the situation of philosophical theory construction is analogous, as I believe it is, then we should see philosophers as attempting to characterize, not their concepts, let alone the concepts of the folk, but certain extra-mental phenomena, such as knowledge, justification, the good, the right, and so on.

Let's begin with the idea that philosophical expertise consists in having better concepts or at least better conceptual understanding. There are several ways of cashing this idea out. One way is to suggest that philosophical discussions typically involve technical concepts, and that philosophers have privileged access to these concepts by virtue of their philosophical education. The problem with this suggestion is that it simply doesn't seem to be true to say that most philosophical discussions involve technical concepts.² There are certainly some philosophical discussions that do; philosophical discussions about the nature of validity or warranted assertability come to mind. But many philosophical discussions involve rather ordinary concepts, and for good reason. Concerns about these ordinary concepts are precisely what gave rise to these philosophical discussions in the first place. If these discussions were then couched in purely technical terms, they would lose traction with the ordinary concerns that gave rise to them. It's hard to see, for example, how a philosophical discussion about some purely technical concept of knowledge could ever hope to tell us much about the kinds of epistemic worries involving our ordinary concept that motivate much of contemporary epistemology.

If philosophical expertise doesn't consist in having better concepts, maybe it consists in having a better understanding of ordinary concepts.³ Maybe philosophers are able to make more precise conceptual distinctions, for example.

This version of the expertise objection argument brings up a number of fascinating issues, but we don't see how it even begins to serve as an objection to the practice of experimental philosophy. On the contrary, we would love to know more about the ways in which philosophers differ from ordinary folks, and it seems to us that the best way to find out would be to run some experiments... Furthermore, even if we discover important differences between the philosophers and the folk, it would hardly follow that data from the folk are irrelevant. Rather, the whole pattern of data might tell us something important about the ultimate source of the philosophical problems. (p. 9)

In short, the suggestion that philosophers have a better understanding of ordinary concepts invites more experimental work on our philosophical intuitions, not less.

It also invites a question. Evidence that philosophers have a different understanding of ordinary concepts isn't evidence that they have a better understanding of those concepts, so why might we think that they do? The answer typically has something to do with philosophical education - philosophical education somehow improves our conceptual understanding. But, it is not clear exactly how this is supposed to happen. Jonathan Weinberg, Chad Gonnerman, Cameron Buckner, and Joshua Alexander (2010) suggest that most philosophers think it happens via some process of trial and error, where philosophy students train their conceptual competencies by checking their conceptual judgments against some received standard.

[V]arious of the psychological mechanisms that can plague human cognition in general - overgeneralization, overconfidence, cognitive dissonance, attribution error, belief bias, belief perseverance, and so on - might very well lead philosophers to believe that we have attuned intuitions, even if in reality in many places we have simply been systematically reaffirming early impressions and incorrectly attributing our professional successes (e.g., in debate and publication) to their validity. (p. 341)

Not only do we have reason to worry that philosophical education isn't particularly helpful, we have reason to worry that we aren't particularly well adapted to realize this.

Perhaps philosophical expertise consists in having mastered some set of philosophical theories and principles. Hilary Kornblith (2007) has something like this in mind when he writes,

Intuitions uninformed by any theory - or only minimally informed by theories common to the folk - would be no more useful [in philosophy] than observations performed by investigators wholly ignorant of relevant background theory in science. We do not go out of our way, in the sciences, to have observations made by individuals so ignorant of relevant theory that their corpus of beliefs contain no theories at all which might threaten to affect those observations... The suggestion that we should attempt to capture pre-theoretical intuition... seems to privilege the intuitions of the ignorant and the naive over those of responsible and well-informed investigators. I cannot see why this would be a better idea in philosophy than it is in science. (p. 34)

The idea seems to be that philosophical theories can help shape our intuitions, perhaps by helping to make certain features of a given hypothetical case salient or by guiding our interpretation of those features.⁴ While this is an appealing suggestion, it faces several difficulties. First, judgments filtered through philosophical theory might not even count as philosophical intuitions; at the very least, the filtering process would have to be unconscious or otherwise introspectively opaque in order for the subsequent judgments to count as genuine philosophical intuitions on some accounts (e.g., Lynch 2006). Second, theoretical commitments are just as likely to contaminate as they are to clarify. The fact that expert philosophical intuitions are theoretically informed doesn't ensure that they are more theoretically valuable than folk philosophical intuitions; in fact, they might be less theoretically valuable for that very reason. Kornblith recognizes this possibility, but argues that this simply means that we need to do whatever we can to make sure that the theories that influence our philosophical intuitions are accurate. It is not clear, though, what this would mean for philosophy’s intuition deploying practices. If our theoretical commitments shape our philosophical intuitions, it is hard to see how our philosophical intuitions can help us independently assess the accuracy of those theories. This isn’t a problem for someone like Kornblith, who thinks that our intuition deploying practices have quite limited philosophical value. But, it does present a significant challenge to anyone who thinks that we advance philosophical theories on the basis of their ability to explain our philosophical intuitions, defend their truth on the basis of their overall agreement with our philosophical intuitions, and justify our philosophical beliefs on the basis of their accordance with our philosophical intuitions.

Let’s switch gears a bit. Maybe philosophical expertise simply consists in the fact that philosophers spend more time thinking carefully about philosophical issues. Antti Kauppinen (2007) adopts something like this position when he argues that robust philosophical intuitions are more theoretically valuable than surface philosophical intuitions. One mark of robust philosophical intuitions is that they are formed in sufficiently ideal conditions, conditions in which we have the time to carefully examine and evaluate not only our judgments about hypothetical cases, but also the cases themselves and what influence our philosophical commitments might have on what details we find relevant in those cases.⁵ The underlying thought is that reflective judgments have a better chance of being right (see, e.g., Sosa 1991) and so, if philosophers spend more time engaged in this kind of reflective practice, then it might seem natural to think that their philosophical intuitions are more theoretically valuable than folk philosophical intuitions.

As Hilary Kornblith (2002, 2010) and Jonathan Weinberg and Joshua Alexander (forthcoming) point out, however, the relationship between reflection and reliability is not this straightforward. True, there are times when reflection helps improve our judgments. Reflection can sometimes help us to realize that we had been misinformed or uninformed about some relevant detail of the particular case, that we had lost track of some of the relevant details, or that our initial judgments about what details are relevant were contaminated by our theoretical commitments. But, reflection can just as easily serve as an echo chamber, simply ratifying whatever initial judgments we might have made. It can help increase our confidence in those judgments without increasing their reliability. One problem is that the cognitive processes involved in the formation of our conceptual judgments aren't introspectively accessible (Nisbett & Wilson 1977, Wilson 2002).⁶ Another is that we tend to be overconfident in our own reliability; we overestimate the degree to which our beliefs are based on the relevant details and underestimate the degree to which our decisions about what details are relevant are contaminated by our theoretical commitments. Perhaps the biggest problem, however, is that when we actually attempt to engage in the kind of reflection that Kauppinen has in mind, a host of cognitive biases are likely to get in the way (see, e.g., Wason 1960, Nickerson 1998, and Baron 2000). In fact, even being aware of this won't help, since we tend to do a particularly bad job of compensating for even known cognitive biases (Pronin et al. 2002). In short, it is far from obvious that reflection helps enough to support the claim that expert philosophical intuitions are more theoretically valuable than folk philosophical intuitions.⁷

There is another possibility; maybe philosophical expertise consists in some kind of procedural knowledge - some special “know-how” developed over the course of our philosophical education. If philosophers are better trained in how best to read and think about philosophical thought experiments, this might naturally make their philosophical intuitions seem more theoretically valuable than folk philosophical intuitions. So, what might this procedural expertise look like? Here are two possibilities. Ernest Sosa (2009) has recently observed that the vignettes typically used in philosophical thought experiments require the reader to import a certain amount of information not explicitly contained in the passage itself. In this way, reading philosophical thought experiments is similar to reading fiction:⁸

When we read fiction we import a great deal that is not explicit in the text. We import a lot that is normally presupposed about the physical and social structure construction as we follow the author's lead in our own imaginative construction. (p. 107)

If this is right, then we might construe philosophical procedural expertise in terms of our ability to properly get at the relevant details of a particular vignette, our ability to appropriately fill in details not explicitly contained in the vignette, and our ability to entertain these details in our imaginations. Timothy Williamson (2007, 2011) suggests a different kind of procedural expertise. On Williamson's account, thought experiments involve deductively valid arguments with counterfactual premises, and we are supposed to evaluate them using a mixture of imaginative simulation, background information, and logic. Williamson (2011) attempts to decompose the task of thought experimenting into discernible sub-tasks: we must read and digest the description of the scenario, judge what would be the case in the scenario described, judge whether the scenario is possible, and determine whether the premises entail the conclusion. If this reconstruction of the process of thinking about thought experiments is right, then we might construe philosophical procedural expertise in terms of our ability to pick out the relevant details of a particular vignette, engage in counterfactual reasoning, and make certain logical inferences. Both of these seem like fairly interesting pictures of what philosophical expertise might look like, but it is important to see that questions about comparative procedural expertise, like questions of comparative conceptual competence, are precisely the kinds of questions that experimental philosophy should be well suited to help us investigate. So, it is hard to transform the possibility that philosophers have greater procedural know-how into a reason to diminish the philosophical significance of experimental philosophy.

We have examined a number of reasons why we might think that expert philosophical intuitions are more theoretically valuable than folk philosophical intuitions: philosophers have better concepts and theories, or at least a better understanding of the relevant concepts and theories, have thought long and hard about these concepts and theories, and have been trained in how best to read and think about philosophical thought experiments that call upon us to apply these concepts and theories. None are particularly persuasive, at least at this point. At the very least, more work needs to be done to explain why we should think that expert philosophical intuitions are more theoretically valuable than folk philosophical intuitions. Much of this work is likely going to have to be empirical, and is going to involve carefully studying folk philosophical intuitions, expert philosophical intuitions, and the cognitive processes involved in thinking about philosophical thought experiments and generating philosophical intuitions.⁹

3.