A DEFENSE OF INTUITIONS
Forthcoming in Philosophical Studies. (Some minor revisions might still come.)
© S. MATTHEW LIAO
Philosophy, Oxford University, Littlegate House, 16/17 St. Ebbes St., Oxford
OX1 1PT, UK; e-mail:
27 June 2007
A Defense of Intuitions
Radical experimentalists argue that we should give up using intuitions as evidence in philosophy. In this paper, I first argue that the studies presented by the radical experimentalists in fact suggest that some intuitions are reliable. I next consider and reject a different way of handling the radical experimentalists’ challenge, what I call the Argument from Robust Intuitions. I then propose a way of understanding why some intuitions can be unreliable and how intuitions can conflict, and I argue that on this understanding, both moderate experimentalism and the standard philosophical practice of using intuitions as evidence can help resolve these conflicts.
A Defense of Intuitions*
I. The Radical Experimentalists’ Challenge
In recent years, a number of philosophers have been conducting empirical studies that surveyed people’s intuitions about various subject matters in philosophy. Other philosophers such as Gilbert Harman and Frank Jackson have also indicated support for this kind of empirical approach to philosophy even if they have not (yet) carried out such empirical surveys. Call this empirical approach to philosophy ‘experimentalism’ and those who are sympathetic to this approach ‘experimentalists.’
On the whole, experimentalism seems to be a good development for philosophy. When philosophers run out of arguments, they often appeal to intuitions. If something is intuitive, this tends to count in favor of a position, and if something is counterintuitive, this tends to count against the position. Moreover, if a philosopher discovers that others have different intuitions than she does regarding a particular case, such a revelation can be quite informative, because she must then investigate whether there is a genuine conflict, and if so, whether she or the others are mistaken. Call this the Intuition as Evidence approach to philosophy (IAE). To the extent that experimentalism has the potential to uncover unreliable intuitions, experimentalism seems to be something that philosophers should welcome.
However, among those who are sympathetic to experimentalism, one can distinguish the ‘moderate experimentalists,’ who believe that experimentalism can complement IAE, from the ‘radical experimentalists,’ who seem to hold the view that experimentalism should replace IAE. As examples of the latter, in “Normativity and Epistemic Intuitions,” Jonathan Weinberg, Shaun Nichols, and Stephen Stich (WNS) argue that their empirical studies show that intuitions about certain cases, which philosophers have taken for granted as being uniform, in fact vary according to factors such as cultural and educational background. Or, in “The Instability of Philosophical Intuitions,” Stacey Swain, Joshua Alexander and Jonathan Weinberg (SAW) argue that their research reveals that intuitions about cases that have been regarded as uniform can in fact vary according to what other cases have recently been considered. As it seems that intuitions should not vary according to these seemingly irrelevant factors, WNS believe that “the entire tradition of Epistemic Romanticism,” that is, philosophical approaches that take intuitions as inputs, “has been a very bad idea.” Or, SAW assert that “there is unlikely to be a fixed set of intuitions about a particular thought-experiment to which we can appeal.” Or, in another paper, Alexander and Weinberg claim that “empirical research into the nature of intuitions generated in response to thought-experiments, rather than supporting the use of intuitions as evidence, challenges the suitability of intuitions to function in any evidentiary role” (my italics).
There have been some attempts to address the radical experimentalists’ challenge to IAE. For example, Timothy Williamson argues that the radical experimentalists’ skepticism about intuitions is a special kind of skepticism about a contextually relevant judgment that targets our standards for applying ordinary concepts in experience. According to Williamson, the skeptic of intuitions insists that we have available as evidence only the fact that it intellectually appears to us that something is so, and therefore asks by what right we treat the fact that it intellectually appears to us that something is so as good evidence that something is true. Williamson argues that this kind of skepticism rests on bad epistemology, namely, a self-defeating operational standard for evidence that requires that one is always in the position to know what one’s evidence is. According to Williamson, once we give up this “hopelessly demanding” operational standard for evidence, it is unclear why our evidence could not include intuitions.
Or, Ernest Sosa also questions WNS’s research along several lines. For example, the hypotheses investigated by WNS are that culture and socioeconomic class affect the epistemic intuitions of members. But, as Sosa points out, epistemic intuitions may vary from group to group only in strength. If so, the variation found may still be compatible with total agreement across the cultures and socioeconomic groups. Sosa’s other concerns include whether it is clear what question the subjects disagree about given that people often import different assumptions that are not explicit in a text; whether the subjects are given adequate choices; and whether the disagreements among the subjects are merely verbal.
In addition, one might also complain about the sample sizes of the radical experimentalists’ studies. Since experimentalism is an a posterori enterprise, it seems odd for it to be making such a large generalization from very limited studies. Indeed, even if the radical experimentalists have found that some intuitions can vary across culture and socioeconomic background, it seems that they have to conduct a lot more empirical research to show that any intuition will be like this.
Furthermore, both the radical and the moderate experimentalists take an intuition to be something like “a spontaneous judgment about the truth or falsity of a proposition.” Some might argue though that an intuition is instead a non-inferentially-based belief in some proposition. If so, it might be difficult to tell whether the responses the experimentalists have gathered are in fact non-inferentially based beliefs.
Radical experimentalists have responded to some of these concerns. For example, in SAW’s surveys, they offer five choices of varying strengths of agreement and disagreement in order to address the worry that intuitions may only vary in strengths. As we shall see shortly, they have also provided some cases to ensure that the subjects understood the questions properly. Moreover, the radical experimentalists have argued that the cases they have presented to their subjects are the same kinds of cases that most philosophers have regarded as being appropriate for eliciting intuitions. Hence, they believe that the spontaneous judgments they have collected do qualify as intuitions.
In the following, I shall grant the experimentalists that the judgments they have collected are intuitions. I shall also not try to offer a general theory of why intuitions are reliable, which has been done elsewhere. My specific aim in this paper is to present another argument, using the experimentalists’ own methodology, to persuade the radical experimentalists to become moderates. In particular, I shall first demonstrate that the studies presented by the radical experimentalists in fact suggest that some intuitions are reliable. This, I claim, is a serious problem for those radical experimentalists who believe that there is not a fixed set of intuitions about a particular thought-experiment to which we can appeal. Next I examine and reject a different way of handling the radical experimentalists’ challenge, what I call the Argument from Robust Intuitions, which says that since the experimentalists’ surveys do not engage in any serious dialogue with the test subjects, the surveys at best capture only surface intuitions but not robust intuitions, the latter of which are intuitions that competent speakers would have in ideal conditions. Finally, I offer an explanation of why intuitions can be unreliable and how intuitions can conflict, and I suggest that on this understanding, both moderate experimentalism and IAE can play a role in helping to resolve these conflicts.
To begin, let me provide more details of the empirical studies conducted by WNS and SAW.
II. The Radical Experimentalists’ Case against Intuitions
WNS present several cases to their subjects aimed to show that intuitive responses to particular cases that philosophers in epistemology have taken for granted in fact vary across culture and socio-economic backgrounds. One set of cases are the Truetemp cases inspired by Keith Lehrer. These are cases designed to explore externalist/internalist dimensions of the subjects’ intuitions – where internalism, with respect to some epistemically evaluative property, holds that only factors within an agent’s introspective grasp can be relevant to whether the agent’s beliefs have that property; and where other factors beyond the scope of introspection such as the reliability of the psychological mechanisms that actually produced the belief are epistemically external to the agent.
For example, in the Individualistic Truetemp Case, WNS ask,
One day Charles is suddenly knocked out by a falling rock, and his brain becomes re–wired so that he is always absolutely right whenever he estimates the temperature where he is. Charles is completely unaware that his brain has been altered in this way. A few weeks later, this brain re–wiring leads him to believe that it is 71 degrees in his room. Apart from his estimation, he has no other reasons to think that it is 71 degrees. In fact, it is at that time 71 degrees in his room. Does Charles really know that it was 71 degrees in the room, or does he only believe it?
Charles’ belief is produced by a reliable mechanism, but it is stipulated that he is completely unaware of this reliability, which means that his reliability is epistemically external. WNS find that East Asian (EA) subjects are much more likely to deny knowledge than their Western (W) classmates.
In another set of cases, the Gettier cases, in which a person has good – but, as it happens, false, or only accidentally true, or in some other way warrant-deprived – evidence for a belief which is true, WNS find that EAs are less inclined than Ws to withhold the attribution of knowledge. For example, they ask the following:
Bob has a friend,
Jill, who has driven a Buick for many years. Bob therefore thinks that Jill
drives an American car. He is not aware, however, that her Buick has recently
been stolen, and he is also not aware that Jill has replaced it with a
WNS find that a large majority of Ws give the standard answer in the philosophical literature, namely, “Only Believes,” but a majority of EAs say that Bob really knows. WNS find even more striking differences when they survey students from the Indian sub-continent (SCs) and compare them with W students.
In addition to cultural differences, WNS find that epistemic intuitions seem to be sensitive to socioeconomic status of the people offering the intuitions. WNS classify subjects as low SES if they report that they have never attended college, and subjects who report that they have one or more years of college are coded as high SES. When WNS present a Dretske-type case to these two different groups of people, again they find significant difference:
Pat is at the zoo with his son, and when they come to the zebra cage, Pat points to the animal and says, “that’s a zebra.” Pat is right –– it is a zebra. However, given the distance the spectators are from the cage, Pat would not be able to tell the difference between a real zebra and a mule that is cleverly disguised to look like a zebra. And if the animal had really been a cleverly disguised mule, Pat still would have thought that it was a zebra. Does Pat really know that the animal is a zebra, or does he only believe that it is?
Low SES subjects are more likely to attribute knowledge in such a case than high SES subjects.
As WNS acknowledge, they “are not sure how to explain these results.” However, WNS believe that the fact that Ws, EAs, SCs, low SESs, and high SESs have different epistemic intuitions is enough to show that philosophical approaches that take intuitions as inputs, that is, IAE, would “yield significantly different normative pronouncements as outputs.” And, this, WNS believe, is bad news for IAE.
SAW argue that their studies reveal that intuitions about certain Truetemp cases vary according to whether, and what, other thought experiments are considered first. To show this, they present eight different survey versions to 220 undergraduates. The survey versions differ only in the order in which four different thought experiments are presented. The central thought-experiment is the Individualistic Truetemp case, which WNS also use, involving Charles. The other thought-experiments include a clear case of non-knowledge (the Coinflip Case); a clear case of knowledge (the Chemist Case), and a more exploratory case (the Goldman-style Fakebarn Case). In the Coinflip Case, they ask,
Dave likes to play a game with flipping a coin. He sometimes gets a “special feeling” that the next flip will come out heads. When he gets this “special feeling”, he is right about half the time, and wrong about half the time. Just before the next flip, Dave gets that “special feeling”, and the feeling leads him to believe that the coin will land heads. He flips the coin, and it does land heads. Please indicate to what extent you agree or disagree with the following claim: “Dave knew that the coin was going to land heads.”
In the Chemist Case, SAW ask,
Karen is a distinguished professor of chemistry. This morning, she read an article in a leading scientific journal that mixing two common floor disinfectants, Cleano Plus and Washaway, will create a poisonous gas that is deadly to humans. In fact, the article is correct: mixing the two products does create a poisonous gas. At noon, Karen sees a janitor mixing Cleano Plus and Washaway and yells to him, “Get away! Mixing those two products creates a poisonous gas!” Please indicate to what extent you agree or disagree with the following claim: “Karen knows that mixing these two products creates a poisonous gas.”
And, in the Fakebarn Case, they ask,
Suzy looks out the window of her car and sees a barn near the road, and so she comes to believe that there’s a barn near the road. However, Suzy doesn’t realize that the countryside she is driving through is currently being used as the set of a film, and that the set designers have constructed many fake barn facades in this area that look as though they are real barns. In fact, Suzy is looking at the only real barn in the area. Please indicate to what extent you agree or disagree with the following claim: “Suzy knows there is a barn near the road.”
SAW find that the subjects’ willingness to attribute knowledge to Charles in the Truetemp case vary depending on whether, and what, other cases are presented before it. In particular, SAW find that the subjects’ willingness to attribute knowledge to Charles increase after being presented with a case of non-knowledge (the Coinflip Case); and the subjects’ willingness to attribute knowledge to Charles decrease after being presented with a clear case of knowledge (the Chemist Case). On this basis, SAW argue that there is not a fixed set of intuitions about a particular thought-experiment to which we can appeal.
III. Radical Experimentalism Also Relies on Intuitions
I shall now argue that in fact the studies by the radical experimentalists suggest that some intuitions are reliable.
As we have seen, in SAW’s and WNS’s studies, they included a case, what SAW called the Coinflip Case, and what WNS called the Special Feeling Case, involving Dave and coinflipping. In addition, in SAW’s studies, they included a “clear case of knowledge,” that is, the Chemist Case, involving Karen the chemistry professor.
As WNS and SAW explained, the Coinflip/Special Feeling Case is designed in part to ensure that their subjects understood the distinction between “Really Know” and “Only Believe,” and that their subjects were using a sense of “know” that was of philosophical interest rather than just a subjective certainty sense of the word such as “Drat! I just knew that X was going to win,” even if X did not win. I believe though that these two cases offer a different lesson, namely, they suggest that some intuitions are reliable across different cultural and educational backgrounds. As WNS and SAW have pointed out, in the Coinflip/Special Feeling Case, whether between different cultural groups or between the high and the low SES groups, almost none of the subjects judged that this was a case of knowledge. And as SAW have pointed out, irrespective of cultural or socioeconomic background, almost all the subjects judged that the Chemist Case was a case of knowledge. Now it is true that the mere fact that there is such wide agreement about these cases do not prove, especially to a real skeptic, that these intuitions are in fact reliable. But from the experimentalists’ own, empirical methodology, the fact that these cases exist do suggest that we can sometimes rely on our intuitions; or at the minimum, they suggest there is no empirical evidence to think that we could not rely on these intuitions.
Indeed, WNS seem to accept that some epistemic intuitions could be reliable in this way (SAW did not mention this as a concern at all). Hence, WNS say,
the fact that subjects from all the groups we studied agreed in not classifying beliefs based on “special feelings” as knowledge suggests that there may well be a universal core to “folk epistemology”. Whether this conjecture is true, and, if it is, how this common core is best characterized, are questions that will require a great deal more research. Obviously, these are not issues that can be settled from the philosopher’s armchair.
But they fail to recognize that this is a large concession, in fact, to the extent that it undermines their claim that there is not a fixed set of intuitions about a particular thought-experiment to which we can appeal. Moreover, once it is granted that some intuitions can serve as evidence, this opens up the possibility that other intuitions might also be part of this common core and might also be able to serve as evidence.
The radical experimentalists might reply that intuitions that are part of the common core are bound to be uninteresting ones. For example, they might claim that the intuitions elicited from the Coinflip/Special Feeling Case and the Chemist Case are uninteresting. However, whether an intuition is interesting or not surely depends on the context. In the context of the radical experimentalists’ arguments, arguably, not only are the intuitions elicited from the Coinflip/Special Feeling Case and the Chemist Case interesting, they are also indispensable. Indeed, as WNS and SAW have pointed out, one of the main rationale for including the Coinflip/Special Feeling Case is to ensure that the subjects were using a sense of “know” that was of philosophical interest. Without such a case, Sosa’s concerns that the subjects may have imported different assumptions that are not explicit in the survey, and that the subjects may just be having verbal disagreements, would surface and would seriously undermine the interpretability of WNS’s and SAW’s studies. The Coinflip/Special Feeling Case and the Chemist Case are even more important for SAW’s studies, because these cases were used to demonstrate that some intuitions are susceptible to the order by which the cases were presented. Without these cases, SAW’s studies would not have been possible.
In addition, even if one grants that these intuitions are uninteresting, given the limited number of studies the radical experimentalists have conducted, it seems a stretch to assert with certainty that any reliable intuition will be uninteresting.
At this point, some radical experimentalists might seek to alter their claim. Rather than denying there are intuitions to which we can sometime appeal, they might argue that their results demonstrate that only those intuitions that have been confirmed empirically can be reliable. As SAW note at one point, “even if one were to grant that, in principle, intuitions can be used as evidence, [their] results suggest that, at this time, we cannot tell which intuitions can safely be deployed.”
In response to this kind of skeptical argument, Sosa, Williamson, and others have argued that the fact that perception can sometimes fail us does not mean that we must confirm each time that our perception is reliable. According to them, as perceptions are relevantly similar to intuitions, the fact that intuitions can sometimes fail us also should not mean that we must prove each time that our intuitions are reliable.
Such a response has been challenged on the ground that perceptions might not be relevantly similar to intuitions. For example, it might be pointed out that while we have considerable knowledge regarding how and when our perception is likely to be unreliable, arguably, we have less knowledge regarding how and when our intuition is likely to be unreliable. Also, it might be said that our explanation of how perception works entails that, typically at least, perception tracks the truth, but it is less clear how intuitions are truth-tracking.
Sosa, Alvin Goldman, and others have offered accounts of how intuitions can be reliable and truth-tracking. For Sosa, to have an intuition is to be attracted to assenting to a proposition on the basis of a reliable cognitive ability, and in so far as an intuition is the result of such a reliable faculty, its outputs have probative force. For Goldman, it is part of the constitutive nature of concepts (in what Goldman calls the ‘personal psychological sense) that possessing a concept tends to give rise to beliefs and intuitions that accord with the contents of the concept. These accounts are obviously not uncontroversial. Sosa has further noted that some intuitions (e.g. our grasp of simple a priori truths) may in fact be more reliable than some perceptions. Moreover, as I shall shortly point out, we do in fact have some knowledge of how and when our intuitions can be unreliable. In any case, even if it were true that intuitions are on the whole not as reliable perceptions, this may still leave open the issue of whether intuitions need to be as reliable as perceptions before we can use intuitions without confirming them each time. After all, our memories are typically also not as reliable as our perceptions; but usually we are justified in using our memories without confirming them each time.
However, these matters need not be settled here. The radical experimentalists’ altered claim still demands too much, because the radical experimentalists rely on a number of intuitions that they have not confirmed. For example, they assume that an intuition regarding a particular case is confirmed if there is near universal agreement regarding the case. For one thing, an intuition can still be unreliable even if there were universal agreement. More pertinently, as far as I can tell, they have not empirically tested this assumption. Or, consider their claim that only those intuitions that have been empirically confirmed can be reliable. Again, as far as I can tell, they have not empirically tested this claim. The radical experimentalists might deny that their methodological assumptions appeal to intuitions. In particular, they might claim that they are merely employing the methods of experimental psychology, which has been shown to be successful in psychology, or that they are just appealing to some coherentist view of epistemology. However, Williamson has argued that the standard philosophical practice also does not depend on intuitions, understood as some sort of mental seeming, but rather, the facts in the world. Alexander and Weinberg have responded though that the radical experimentalists’ challenge applies whether the standard philosophical practice is framed in terms of intuitions or some other kind of evidence. But if this is so, then even if the radical experimentalists were not appealing to intuitions per se, they are still appealing to evidence that they have not yet confirmed. Indeed, as far as I know, they have not empirically confirmed the assumption that what works in psychology will also work in philosophy, or that a coherentist view is truth-tracking. Note that I am not arguing that these assumptions could not be true; I am only suggesting that the radical experimentalists have not empirically confirmed them. Sosa has observed that “It is hard to avoid appeal to direct intuition sooner or later.” Michael Lynch has concurred, “Indeed, more than hard, practically impossible.” If the radical experimentalists are right that only those evidence – glossed typically in terms of intuitions – that have been confirmed empirically can be reliable, they would be undermining their own position.
Here it is worth pointing out that my aim has not been to argue that philosophers need not worry about their use of intuitions. To the extent that intuitions about central cases in philosophy seem to vary according to irrelevant factors, philosophers should certainly reconsider these intuitions again, and in general, be even more vigilant about confirming the reliability of their intuitions. My aim here has rather been to argue that the radical experimentalists’ results do not entail a) that we need to abandon the use of intuitions altogether; or b) that only those intuitions that have been confirmed empirically can be reliable. On the latter, as Williamson has argued, an operational standard for evidence that requires that one is always in the position to know one’s evidence is “hopelessly demanding.”
IV. Argument from Robust Intuitions
At this point, some people might feel that there is an easier way to rebut the radical experimentalists’ claim. In particular, some might think that one should distinguish between surface intuitions, which are ‘first-off’ intuitions that may be little better than mere guesses; and robust intuitions, which are intuitions that a competent speaker might have under sufficiently ideal conditions such as when they are not biased. In other words, when philosophers assert that ‘Everyone would agree that …’ or ‘Intuitively, we would all find it obvious that …’ or ‘It is clear to us that . . .’, the ‘we’ and ‘us’ should be interpreted as applying only to competent speakers in certain non-distorting conditions. According to this line of thought, because the surveys do not engage in any serious discussion and reflection with the test subjects, e.g., to find out if they are biased or not, the surveys at best capture only surface intuitions but not robust intuitions. Call this the Argument from Robust Intuitions (ARI).
Whatever the merit of the distinction between surface and robust intuitions might be, certainly we should be interested in something more than surface intuitions where the speakers have a competent grasp of the concepts at issue. After all, we would not typically ask a child what she thought of a particular thought experiment (though we might if she was particularly clever). Also, notice that if ARI were successful, it would not just pose problems for the radical experimentalists, it would also pose problems for the moderate experimentalists, since they both employ the same methodology.
However, arguably, the radical experimentalists’ surveys do capture more than surface intuitions. As we have seen, both WNS and SAW have cases to test whether the subjects were using a sense of “know” that was of philosophical interest. Given this, they could argue – and in fact, they have – that the intuitions they have elicited are at least minimally reflective ones.
Of course, a proponent of ARI will argue that minimally reflective intuitions are not robust intuitions. To obtain the latter, an individual must have engaged in certain serious reflection and dialogue. But why think that an intuition can count as credible evidence only if one has engaged in serious reflection and dialogue? That is, why believe that an intuition that has been subjected to philosophical examination will necessarily be better than one that has not been subjected to such examination? In fact, the converse might even be true, namely, a philosopher’s intuitions might be so colored by her philosophical commitments that her intuitions are no longer as reliable as minimally reflective ones. Note that I am not claiming that serious reflection and dialogue can never help us improve the reliability of our intuitions. In fact, I shall shortly point out that our intuitions are sometimes unreliable owing to partiality, clouded emotions, mistake heuristics, and so on; and serious reflection and dialogue no doubt can often help us remove these biases. The limited claim I am making on both the moderate and the radical experimentalists’ behalf is that there is no reason to believe that serious reflection and dialogue will necessarily improve the reliability of our intuitions. If this is right, then it is not clear that the experimentalists’ surveys must capture robust intuitions.
Proponents of ARI will no doubt have further rebuttals to this point, but for our purpose, I do not immediately see how ARI is a more effective argument against the radical experimentalists than the argument I have presented earlier.
V. The Unreliability of Some Intuitions and Adjudicating Conflicts of Intuitions
Although the radical experimentalists have not shown that intuitions cannot function in any evidentiary role, their studies do raise some interesting questions, namely, why are our intuitions sometimes unreliable and what should we do when intuitions conflict. I provide the following explanations.
Our intuitions are sometimes unreliable because of partiality, clouded emotions, and mistaken heuristics. For example, Bob may have the intuition that Chinese cooking is better than British cooking, because Bob is ethnically Chinese and has some partiality towards Chinese food. Or, someone who is afraid of flying may have the intuition that plane crashes have a higher chance of occurrence than common causes of death. Heuristics are simple decision procedures such as “the more expensive the better.” Hence, one may think that the more expensive the wine, the better it should taste.
When intuitions are the results of some form of bias, we should clearly not take these intuitions as being definitive without further consideration. For example, given that Bob has partiality towards Chinese food, we should take his intuition that Chinese cooking is better than British cooking with a grain of salt. Or, we may wish to reflect on whether it is in fact the case that a more expensive wine always tastes better.
That partiality, clouded emotions, and mistaken heuristics can cause our intuitions to be unreliable in fact provides another explanation as to why philosophers often assert that ‘Everyone would agree that …’ or ‘Intuitively, we would all find it obvious that …’ or ‘It is clear to us that . . . .’ Both radical and moderate experimentalists take these assertions to mean that claims of intuitions are empirical claims about what everyone believes. Consequently, some of them hold the view that evidence of disagreements implies that these claims of intuitions are false. Proponents of ARI interpret these assertions as not claims about what everyone believes, but only what competent speakers in ideal conditions believe. A third explanation is that philosophers make these assertions to indicate that they are not biased. That is, agreement is used as a useful heuristic to rule out biased intuitions. In other words, if other people also have the same intuition as one does, this suggests, although it does not prove, that one’s intuition is likely not to be based on partiality, clouded emotions, and mistaken heuristics. Unlike the other explanations, on this explanation, evidence of disagreement may suggest that one has biased intuitions (a point we shall come to shortly), but agreement is neither necessary nor sufficient for an intuition to be correct.
Regarding what we should do when intuitions conflict, it is helpful first to distinguish a) conflicts within oneself, what might be called internal conflicts, from conflicts with others, what might be called external conflicts; and 2) conflicts that are about the same cases from conflicts that are about different (albeit similar) cases. An example of an internal conflict about the same case may be an individual’s having both the intuition that abortion is completely permissible (because she is sympathetic to the idea that a woman has the right to decide what happens to and insider her body), and the intuition that abortion is not completely permissible (because she is also sympathetic to the idea that the fetus has some moral status). An example of an external conflict about the same case may be when one individual has the intuition that abortion is completely permissible, while another has the intuition that abortion is not completely permissible. An example of an internal conflict of intuitions about different (albeit similar) cases may be an individual’s having both the intuition that embryonic stem cell research is completely permissible, and the intuition that abortion is not completely permissible. An example of an external conflict about different (albeit similar) cases may be that one individual has the intuition that embryonic stem cell research is completely permissible, while another has the intuition that abortion is not completely permissible. In this paper, the studies that the radical experimentalists have presented are all external conflicts about the same cases.
Secondly, it is useful to distinguish among Apparent Conflicts, Conflicts out of Biases, and Genuine Conflicts. In Apparent Conflicts, the parties involved just have verbal disagreements. For example, it has been suggested that the Knobe effect may be revealing only a verbal disagreement. In Conflicts out of Biases, the parties involved may have intuitions that are the result of partiality, clouded emotions, and mistaken heuristics, which we discussed earlier. For the purpose of resolving conflicts of intuitions, Apparent Conflicts and Conflicts out of Biases should not really concern us, because once verbal disagreements and biases have been revealed and confirmed, the conflicts tend to resolve themselves.
In Genuine Conflicts, on the other hand, the conflicts seem to remain after possible verbal disagreements and considerations of partiality, clouded emotions, mistaken heuristics, and so on, have been taken into account. Indeed, in these cases, two individuals seem to possess the same facts about a matter, but seem to interpret the facts differently. For example, consider the Rescue Case:
Suppose there are two islands, one with one person, A, and the other with two people, B&C. There is a tsunami and both islands will soon be immersed in water, killing whoever is on the island. You only have time to go to one of the islands to rescue the people on it. Other things being equal, e.g., assume that there is no morally relevant difference (e.g. special relationship or prior agreements) between these individuals. To which island should you go?
Some people have the intuition that you should save the greater number, while others have the intuition that it does not matter whom you save, as long as you save someone.
In such a case, the difference in intuitions does not seem to have arisen out of verbal disagreements, partiality, clouded emotions, mistaken heuristics, and so on. Indeed, the conflicting intuitions seem to be hinting at deeper disagreements.
To resolve Genuine Conflicts such as this one, there appears to be no additional empirical facts of the matter to be taken into account, since other things have already been presumed to be equal. Given this, one must be able to provide instead further philosophical arguments as a way to adjudicate the conflict. In particular, it will be helpful for adjudicating the conflict if one can provide a positive theory as to why certain intuitions should be included, and a negative, error theory as to why certain other, rival intuitions should be excluded. A positive theory might, for example, involve an argument by analogy (e.g. Mill’s defense of Utilitarianism); and an error theory might take the form of a reductio ad absurdum such as “If the rival intuitions were true, it would have a particular implication. The particular implication is absurd. Therefore, the rival intuitions cannot be true.” If neither side can produce positive arguments in favor of their intuitions and negative arguments against the rival intuitions, then the conflict is at a stalemate. It is worth noting that on this approach, a stalemate does not mean that each side has to give up his or her intuition. However, the existence of a stalemate should mean that one should hold onto one’s intuition with a dose of skepticism until there is further evidence either in favor or against one’s intuition.
To illustrate these points, consider again the Rescue Case. If we have the intuition that we should save the greater number, we should be able to explain why this intuition is correct and why the alternative intuition is not correct. Conversely, if we have the intuition that we are permitted to save either group as long as someone is saved, we should also be able to explain why this intuition is to be preferred over the alternative intuition. If we are unable to provide an explanation as to why our intuition is correct and why the rival intuition is wrong; and if our opponent is also unable to provide an adequate counter explanation, then there is a stalemate. In such a case, we are permitted to hold on to our intuition provided that we recognize that we may still be mistaken.
On this view of philosophizing, it should be clear that moderate experimentalism and IAE are not only compatible, but actually complement each other. Through quantitative and qualitative research, moderate experimentalism can help us identify areas in which we have conflicts of intuition, and may be able to help us distinguish among Apparent Conflicts, Conflicts out of Biases and Genuine Conflicts. When Genuine Conflicts have been uncovered, IAE may then be able to help us resolve these conflicts by providing positive theories for why certain intuitions should be included, and error theories for why certain rival intuitions should be excluded.
Radical experimentalists believe that we should give up the standard philosophical practice of using intuitions as evidence (IAE), because their empirical studies have demonstrated that intuitions vary according to factors such as cultural and educational background, and what other cases have recently been considered. However, the studies presented by the radical experimentalists in fact suggested that some intuitions are reliable. Proponents of the Argument from Robust Intuitions might think that because the experimentalists’ surveys do not engage in any serious dialogue with the test subjects, the surveys at best capture only surface intuitions but not robust intuitions. I argued however that the experimentalists’ surveys do capture more than surface intuitions, namely, minimally reflective intuitions; and that it not clear why robust intuitions are always to be preferred over minimally reflective ones. I then explained that intuitions can be unreliable because of partiality, clouded emotions, and mistaken heuristics, and I proposed that there can be Apparent Conflicts, Conflicts out of Biases, as well as Genuine Conflicts of intuitions. Finally, I argued that moderate experimentalism and IAE can help resolve these different kinds of conflicts. The upshot is that moderate experimentalism is a healthy trend in philosophy, which we should embrace. But we should reject radical experimentalism.
* I would like to thank Joshua Alexander, Jonathan Weinberg, Roger Crisp, Steve Clarke, Robert Wachbroit, David Wasserman, Guy Kahane, Wibke Gruetjen, Walter Sinnott-Armstrong, Michael Blome-Tillmann, an anonymous reviewer at Philosophical Studies, and the audiences at the 2007 American Philosophical Association Pacific Division Meeting in San Francisco and at the Oxford University James Martin Advance Research Seminar for their very helpful comments on an earlier version of this paper.
 See, e.g., Weinberg, J., S. Nichols,
and S. Stich. "Normativity and Epistemic Intuitions." Philosophical
Topics 29, no. 1&2 (2001): 429-60; Machery, E., R. Mallon, Shaun
Nichols, and Stephen Stich.
"Semantics, Cross-Cultural Style." Cognition 92 (2004): B1-B12;
Knobe, J. "Intentional Action and Side Effects in Ordinary Language."
Analysis 63 (2003): 190-93; Nadelhoffer, T. "The
 Harman, G. "Doubts About
Conceptual Analysis." In Philosophy in Mind. The Place of Philosophy in
the Study of Mind, edited by M. Michael and J. O’Leary-Hawthorne, 43-48.
 See their blog at http://experimentalphilosophy.typepad.com/
See, e.g., Goldman, A. Epistemology and Cognition.
Weinberg, Nichols and Stich call this “Intuition-Driven Romanticism”
(“Normativity and Epistemic Intuitions,” pp. 430-436). See also Pust, J. Intuitions as Evidence.
 Weinberg, Nichols, and Stich,
“Normativity and Epistemic Intuitions,” op. cit. See also Nichols, S., S. Stich, and J.
Weinberg. "Meta-Skepticism: Meditations on Ethno-Epistemology." In
The Skeptics, edited by S. Luper, 227-47.
 Swain, S., J. Alexander, and J. Weinberg. "The Instability of Philosophical Intuitions: Running Hot & Cold on Truetemp," Philosophy and Phenomenological Research, forthcoming. For other research to which radical experimentalists have appealed to support their thesis, see, e.g., Machery, Mallon, Nichols, and Stich. "Semantics, Cross-Cultural Style," op. cit.; Nichols, S., and J. Knobe. "Moral Responsibility and Determinism: The Cognitive Science of Folk Intuitions." Nous (forthcoming).
 Weinberg, Nichols, and Stich, “Normativity and Epistemic Intuitions,” p. 434.
 Swain, Alexander, Weinberg. "The Instability of Philosophical Intuitions,” p. 19. In an earlier version of this paper, SAW’s claim is even more explicit. They say that their research entails that “philosophers should consider putting their intuitions in escrow.”
 Alexander, J., and J. Weinberg. "Analytic Epistemology and Experimental Philosophy." Philosophy Compass 2, no. 1 (2007): 56-80, p. 63. In this paper, Alexander and Weinberg have, as far as I am aware, independently introduced a distinction between the ‘proper foundation view’ and the ‘restrictionist view,’ which seems to correspond to my distinction between moderate and radical experimentalism. Their distinction is not quite my distinction though because they attribute the ‘proper foundation view’ the position that ‘Only the results of such research can deliver the intuitions that can serve as evidential basis for or against philosophical claims’ (p. 61). On my distinction, this position would fall under radical experimentalism. Also, the ‘restrictionist’ in the ‘restrictionist view’ may not be discriminating enough because even philosophers of traditional approaches would accept that there should be some restrictions to the use of intuitions. For example, intuitions elicited under biases, mistaken heuristics and so on, should not count.
 Williamson, T. "Philosophical 'Intuitions' and Skepticism About Judgment." Dialectica 58 (2004): 109-53, p. 114.
 Williamson, "Philosophical 'Intuitions' and Skepticism About Judgment,” p. 118.
 Williamson, "Philosophical 'Intuitions' and Skepticism About Judgment,” pp. 118-122.
 Williamson, "Philosophical 'Intuitions' and Skepticism About Judgment,” pp. 121.
 Sosa, E. "A Defense of the Use of Intuitions in Philosophy." In Stich and His Critics, edited by M. Bishop and D. Murphy: Blackwell, 2005.
 Nichols, Stich, and Weinberg, "Meta-Skepticism,” footnote 2; Weinberg, Nichols, Stich, “Normativity and Epistemic Intuitions,” op. cit.
R. The Good in the Right. Princeton:
 Swain, Alexander, Weinberg, "The Instability of Philosophical Intuitions,” op. cit.
 Weinberg, Nichols, Stich, “Normativity and Epistemic Intuitions,” op. cit.
See, e.g., Sosa, E. "Intuitions and Truth." In
Truth and Realism, edited by P. Greenough and M. P. Lynch, 208-26.
 Initially, I was attracted to this approach. See also Kauppinen, A. "The Rise and Fall of Experimental Philosophy," Philosophical Explorations (forthcoming), for a good exposition of this position. The terms ‘surface intuitions’ and ‘robust intuitions’ are from Kauppinen. WNS call robust intuitions ‘Austinian intuitions,’ following a suggestion by Philip Kitcher.
 Lehrer, K. Theory of Knowledge.
 See Gettier, E. "Is Justified True Belief Knowledge?" Analysis 23 (1963): 121-23, which inspired these cases.
 See, e.g., Dretske, F. "Epistemic Operators." Journal of Philosophy 67, no. 24 (1970): 1007-23.
 Weinberg, Nichols, and Stich, “Normativity and Epistemic Intuitions,” p. 445.
 Weinberg, Nichols, and Stich, “Normativity and Epistemic Intuitions,” p. 445.
 Weinberg, Nichols, and Stich, “Normativity and Epistemic Intuitions,” p. 450.
 Alexander and Weinberg. "Analytic Epistemology and Experimental Philosophy," p. 63.
Swain, Alexander, Weinberg, "The Instability of Philosophical Intuitions,”
op. cit.; See also Sinnott-Armstrong,
W. "Moral Intuitionism Meets Empirical Psychology." In Metaethics
 Sosa, "Intuitions and Truth," op. cit.; Williamson, "Philosophical 'Intuitions' and Skepticism About Judgment,” op. cit.; Lynch, "Trusting Intuitions," op. cit.
 Sosa, E. "Experimental Philosophy and Philosophical Intuition." Philosophical Studies 132 (2007): 99-107, p. 105.
 See, e.g., Swain, Alexander, Weinberg, "The Instability of Philosophical Intuitions,” op. cit.
 I thank an anonymous reviewer for pressing this point.
 Sosa, "Intuitions and Truth," op. cit.
 Goldman, A. "Philosophical Intuitions: Their Target, Their Source, and Their Epistemic Status." Grazer Philosophische Studien 74 (2007): 1-25, p. 15.
 Sosa, "Intuitions and Truth," p. 225.
 Williamson, "Philosophical 'Intuitions' and Skepticism About Judgment,” op. cit.
 Alexander and Weinberg. "Analytic Epistemology and Experimental Philosophy," p. 72.
 Sosa, "Intuitions and Truth," op. cit.
 Lynch, "Trusting Intuitions," p. 236.
 Williamson, "Philosophical 'Intuitions' and Skepticism About Judgment,” pp. 121.
 See, e.g., Kauppinen, "The Rise and Fall of Experimental Philosophy," op. cit.
 Weinberg, Nichols, and Stich, “Normativity and Epistemic Intuitions,” op. cit.
 For a good discussion on some of these biases, see Sinnott-Armstrong, "Moral Intuitionism Meets Empirical Psychology," op. cit.
 There may be other ways by which intuitions can conflict, e.g. in their strengths, but I shall not consider them here. See, e.g., Sosa, “A Defense of the Use of Intuitions in Philosophy,” op. cit.
 Some may notice that I am using
moral examples here. Those who are
sceptical that intuitions are truth-tracking might be even more sceptical about
‘moral’ intuitions, since it could very well be the case that there is no truth
to be tracked in the moral case.
Personally, I do think that there are moral truths and that our
intuitions can track them. See also Crisp,
R. Reasons and the Good.
 The Knobe effect could be regarded as another example of an internal conflict of intuitions about different (albeit) similar cases. See, e.g., Knobe, J. "Intentional Action in Folk Psychology: An Experimental Investigation." Philosophical Psychology, no. 16 (2003): 309-24.
 Sosa, "Experimental Philosophy and Philosophical Intuition," op. cit.
 Nichols, S., and J. Ulatowski. "Intuitions and Individual Differences: The Knobe Effect Revisited." Mind and Language (forthcoming).
 For the latter view, see, e.g., Taurek, J. "Should the Numbers Count?" Philosophy and Public Affairs 6 (1977): 293-316.
 Kelly, T. "The Epistemic Significance of Disagreement." In Oxford Studies in Epistemology Vol 1, edited by J. Hawthorne and T. Szabo-Gendler, 2005; Pettit, P. "When to Defer to Majority Testimony - and When Not." Analysis 66, no. 3 (2006): 179-87.
 See, e.g., Williamson, "Philosophical 'Intuitions' and Skepticism About Judgment,” op. cit. For a view that we should adjust our beliefs and intuitions when there is a real disagreement with an “epistemic peer,” see, e.g., Christensen, D. "Epistemology of Disagreement: The Good News." Philosophical Review 116 (2007): 187-217.