Portal de Periódicos da CAPES

Some hope for intuitions: A reply to Weinberg

2010; Taylor & Francis; Volume: 23; Issue: 4 Linguagem: Inglês

10.1080/09515089.2010.505958

1465-394X

Thomas Grundmann,

Decision-Making and Behavioral Economics

Abstract In a recent paper Weinberg (2007 Weinberg, J. 2007. How to challenge intuitions empirically without risking skepticism. Midwest Studies in Philosophy, 31: 318–343. [Crossref], [Web of Science ®] , [Google Scholar]) claims that there is an essential mark of trustworthiness which typical sources of evidence as perception or memory have, but philosophical intuitions lack, namely that we are able to detect and correct errors produced by these "hopeful" sources. In my paper I will argue that being a hopeful source isn't necessary for providing us with evidence. I then will show that, given some plausible background assumptions, intuitions at least come close to being hopeful, if they are reliable. If this is true, Weinberg's new challenge comes down to the claim that philosophical intuitions are not reliable since they are significantly unstable. In the second part of my paper I will argue that and why the experimentally established instability of folk intuitions about philosophical cases does not show that philosopher's expert intuitions about these cases are instable. Keywords: EpistemologyExperimental PhilosophyIntuitions Acknowledgements The author is grateful for extremely helpful comments from Joshua Alexander, Magdalena Balcerak Jackson, Alvin Goldman, Frank Hofmann, Joachim Horvath, Jens Kipper, Max Seeger, Jonathan Weinberg, Anand Vaidya and an anonymous referee as well as for stimulating discussions with the participants of the workshop "Armchair in Flames – Experimental Philosophy and its Critics" in September, 2008 at the University of Cologne, Germany. Notes [1] For discussion see Grundmann (2007 Grundmann, T. 2007. The nature of rational intuitions and a fresh look at the explanationist objection. Grazer Philosophische Studien, 74: 69–87. [Crossref] , [Google Scholar]) and Pust (2000 Pust, J. 2000. Intuitions as evidence, New York: Garland Publishing. [Google Scholar], 2004 Pust, J. 2004. On explaining knowledge of necessity. Dialectica, 58: 71–87. [Crossref] , [Google Scholar]). [2] For a representative statement on this matter see the following quote from Jackson: "often we know that our own case is typical and so can generalize from it to others. It was surely not a surprise to Gettier that so many people agreed about his cases" (1998, p. 37). [3] To be precise: Weinberg claims that error detection and correction is only required for the hopefulness of practically fallible sources, i.e., sources that fail very rarely and only under unusual conditions. See Weinberg, 2007 Weinberg, J. 2007. How to challenge intuitions empirically without risking skepticism. Midwest Studies in Philosophy, 31: 318–343. [Crossref], [Web of Science ®] , [Google Scholar], p. 324. [4] To be more precise: the criterion of internal coherence depends on two independent factors. First, the deliverances of a token source or of different tokens of the same type of source have to be comparable. Later I will call this feature "redundancy" of a source. Only if sources have this feature, errors can be detected. Second, the deliverances of a token source or different tokens of the same type of source must fit together in an overall coherent way. Only then particular errors can be corrected on the basis of the local incoherence they would produce. Thanks to Max Seeger for helping me to get clearer about this point. [5] This is not a case of purely internal considerations, though. Certain background assumptions also seem to be in place, e.g., that other people roughly see the same things in the same way and that testimony is generally trustworthy. Thanks to Joachim Horvath who helped me to get clearer on the issue of internal coherence. [6] Why is nothing more required than at least one of these four resources of error-correction? The answer to this question is pretty obvious. For most of our basic sources we do not possess more than one such resource. If Weinberg were to ask for more, a very far-reaching scepticism would be the consequence. Hence, condition (C1) would be violated. For most basic sources we do not have any resource of external corroboration since basic sources are often exclusive sources with respect to their domain. How should we be able to corroborate colour vision, introspective awareness of special features of mental content (e.g., is this pain sharp or nagging?), or preservative memory by any external procedure? For similar reasons theoretical understanding cannot be necessary for a source being trustworthy. Since many basic sources are exclusive with respect to their domain, their proper functioning within their typical environment can only be explained by using these sources themselves. But then, the theoretical understanding itself could never be justified. The reason is that any justification by a source would already presuppose theoretical understanding of that very source. Concerning the detectability of margins, many basic sources simply do not represent the epistemic quality of their performances (e.g., introspection in cases where it is not practically infallible, or preservative memory). It seems clear that for most basic sources internal coherence is the only available procedure of error-correction. So, Weinberg could not ask for more here without challenging most of our ordinary sources of evidence. [7] One might object to this claim that even if intuitions were sensitive to irrelevant factors this would not necessarily diminish their truth-ratio. Think of perceptual states. They are surely sensitive to neural activities of the brain which are not identical to the truth-makers of the perceptual states. Hence, they are sensitive to irrelevant factors. But this is compatible with the view that perception is extremely reliable. The fact that perceptual states are caused by irrelevant factors does not show that they are not reliable (thanks to Joachim Horvath for the example). In response to this objection I would like to stress that everything depends on how we understand the phrase "irrelevant factors." When I use this phrase I want to convey the information that the factors are neither truth-makers of the intuitions nor are they correlated with the truth-makers. But then it seems obvious that intuitions are fallible if they are sensitive to irrelevant factors in this sense. Thanks to Joachim Horvath and Joshua Alexander for pressing me on this point. [8] In correspondence Weinberg has indicated to me that he hesitates to accept (P3). He wants to put the argument in terms of trustworthiness. So it would run like this: first, practically fallible sources are epistemically trustworthy only if they are hopeful; second, philosophical intutions are hopeless; therefore, philosophical intuitions are not epistemically trustworthy. Weinberg's main reason for preferring these terms is that he thinks that a source of evidence might still not be trustworthy as long as we are not in a position to distinguish what is really good evidence from what is non-evidence or as long as the quality of evidence delivered by the source is not good enough. But if one understands evidence as what justifies a belief, then a source of evidence provides all of its outputs with prima facie justification. So I do not see why we should not trust a source of evidence in this sense. I think all the reasons one can give here are not purely epistemic reasons. Pragmatic, social, institutional or discipline-specific reasons may play an additional role. [9] Notice that Weinberg restricts this statement to philosophical intuitions about bizarre cases. He thinks that there are other kinds of intuitions, like intuitions about logic or mathematics, that are embedded in practices that are hopeful (see Weinberg, 2007 Weinberg, J. 2007. How to challenge intuitions empirically without risking skepticism. Midwest Studies in Philosophy, 31: 318–343. [Crossref], [Web of Science ®] , [Google Scholar], p. 340). And even with respect to philosophical intuitions about bizarre cases our practices might be improved. For this reason Weinberg says that philosophical intuitions are not hopelessly hopeless (2007 Weinberg, J. 2007. How to challenge intuitions empirically without risking skepticism. Midwest Studies in Philosophy, 31: 318–343. [Crossref], [Web of Science ®] , [Google Scholar], p. 341). [10] I take this example from BonJour (1998 BonJour, L. 1998. In defense of pure reason, London: Cambridge University Press. [Google Scholar], p. 119). [11] Thanks to Anand Vaidya for drawing my attention to these findings. [12] Weinberg is well aware of the fact that a theoretical understanding cannot generally be required for basic sources of evidence (2007, p. 331) Humans used perception, memory or testimony as sources of evidence long before they acquired a rudimentary scientific understanding of these sources and children still do. Weinberg, however, does not realize that different requirements may hold for scientific sources of evidence. As Goldman (1988 Goldman, A. 1988. Strong and weak justification. Philosophical Perspectives, 2: 51–69. [Crossref] , [Google Scholar], p. 54), points out, a scientific method will only confer justification, if it is reliable and, in addition, is acquired by a reliable method or process. The latter condition can be satisfied very elegantly when we understand that and under what conditions the scientific method works reliably. [13] Again: this condition holds only for practically fallible sources. [14] Thanks to Jonathan Weinberg for pressing me on this point. [15] Nonetheless it is at least possible that even the deliverances of an infallible and hence incorrigible source can be challenged by misleading defeaters from a corrupt source. But in these cases we may defeat these misleading defeaters by using undermining defeaters. We need not be able to defend the deliverances of the original source with further confirming evidence. Thanks to Joachim Horvath for reminding me of this point. [16] Of course, it is dialectically problematic to use intuitions about weird hypothetical cases against someone like Weinberg who challenges philosophers' appeal to intuitions in general. I agree that this part of my argument has only limited force. But one could still argue that the cases that I use here are not too far away from ordinary real-world cases. Unfortunately, I haven't found good real-world cases to illustrate my point. Thanks to Jonathan Weinberg and Jens Kipper for raising worries along these lines. [17] Thanks to Magdalena Balcerak Jackson, Joachim Horvath, Jens Kipper and Max Seeger for their helpful comments on this paragraph. [18] It is therefore not surprising that hopeful sources satisfy Popper's falsificationist criterion of science and the scientific requirement that experiments ought to be repeatable. [19] Thanks to Jonathan Weinberg and Magdalena Balcerak Jackson for raising this objection. [20] Joachim Horvath has reminded me of the fact that some qualifications are required here. Not all sciences claim to have a unified and systematic explanation of their domain. [21] For Goldman's view on this matter see note 15. [22] Weinberg himself does not claim that being a hopeful source is sufficient for being a trustworthy source. [23] Thanks to Jonathan Weinberg for helping me to make my case stronger here. [24] Here is a different way to defend the view that no further substantial methods are required to provide us with evidence if we have a fairly reliable source at our disposal. Call a source 'lucky' if it does not encounter any defeaters. A fairly reliable source is either lucky or unlucky. First consider the case in which it is lucky. In that case the source has to be exclusive about a certain domain and it also has to be non-redundant. Then there is simply no possibility to face any defeaters. But if there are no special challenges to the deliverances of the source and if it is fairly reliable, then it seems plausible to assume that it delivers evidence. Now consider the second case in which the source is unlucky. Then there will occur specific defeaters and challenges to its outputs. The defeaters may be either external or internal. If they are external, then we already have a source that is embedded in a hopeful practice, because the external defeaters must result from an independent source of evidence that we may use to corroborate the source in question. If they are internal and the conflicts are not resolvable, then we have always the option to suspend judgment in order to detect and avoid (though not correct) errors. Nevertheless, the source will provide us with evidence in those cases in which there is no defeater. If the source is basically reliable this cannot be an empty class of cases. So we can conclude that no matter whether the reliable source is lucky or unlucky it will provide us with evidence without requiring any further non-trivial methods. Thanks to Joshua Alexander for reminding me of the difference between lucky and unlucky sources. [25] To be honest, Weinberg (in correspondence) dismisses this argument as a sensible restrictionist strategy. His main reason is that even significant instability of a source is compatible with its basic reliability. So the restrictionist would have to show that intuition is radically unstable all over the place to challenge its reliability. As Weinberg puts it, the data we have so far certainly do not suggest that anything like that is true. Notice, however, that Weinberg himself sometimes uses the language of reliability to characterize what is wrong with intuitions. For example, he speaks about "the kinds of unreliability that seem to afflict the folk intuitions studied by experimental philosophers." (Weinberg, et al., 2010 Weinberg, J, Gonnerman, C, Buckner, C and & Alexander, J. 2010. Are philosophers expert intuiters?. Philosophical Psychology, 23.3: 331–355. [Taylor & Francis Online] , [Google Scholar], p. 333). In general, I agree with Weinberg that so far the reliability argument is not convincing. But in contrast to Weinberg, I do not believe that the restrictionist has any better argument to defend his position. [26] In fact, I will argue below that philosopher's intuitions are not influenced by these kinds of factors. [27] One could, of course, reply to this objection that it is not true that whenever we legitimately generalize from samples we have to prove in advance that the samples are representative. That would create Hume's problem of induction and thereby lead to global skepticism. But I think that I need not rely on such a general principle in order to get my argument off the ground. Professional philosophers have so much more training and experience with the intuitive evaluation of hypothetical cases than laypersons that it is at least a real-life possibility that subjects from both camps will behave unlike with respect to their intuitions about these cases. In such a situation it seems clearly objectionable to generalize without possessing further indicators of laypersons' intuitions being representative of philosophers' intuitions. Here is an analogy. Suppose you travel to a country which is completely unknown to you. Suppose further that on arrival you first encounter a landscape that is completely flat. Is this sufficient evidence to conclude that the country is flat all over? I would say 'no'. The reason is that although some parts of the country are flat other parts may well be hilly. Being flat in one part and hilly in other parts seems to be a real-life possibility. Many existing countries satisfy this possibility. You are so far not in a position to exclude this possibility on the basis of your evidence. I claim that we are in a similar position with respect to philosophers' intuitions. We do not need prima facie evidence that suggests that professional intuitions are more stable than laypersons' intuitions, if we want to challenge Weinberg's generalization. All we need is the real-life possibility that philosophers' intuitions are superior expert intuitions. This is indeed a real-life possibility, since adequate training and experience often yield expertise. Thanks to Joachim Horvath for helping me to make the reply to my objection as strong as possible. [28] There are indeed a few people who claim that it is plausible to ascribe knowledge to the epistemic agent in the fake barn case. But on closer inspection, they do not claim that it is intuitive to ascribe knowledge in that case. For example, Hetherington (2006 Hetherington, S. 2006. "How to know (that knowledge-that is knowledge-how)". In Epistemology futures, Edited by: Hetherington, S. 71–94. Oxford: Clarendon Press. [Google Scholar], p. 88), argues that this case is analogous to a case in which we acquire knowledge on the basis of waking experience. Waking experience is indistinguishable from dream experience. Nevertheless, nobody would claim that this fact deprives us of knowledge by perception. Therefore, the fact that we might be deceived by our sense experience in the face of a fake barn cannot deprive us of knowledge when we actually see a real barn. Actually, I am not persuaded that this analogy really holds. (Forming a belief on the basis of dream experience seems to be a psychological process different from forming a belief on the basis of waking experience. Whereas the same type of perceptual process produces the belief in front of a real barn and a fake barn.) What really matters here is that Hetherington does not directly rely on a deviant intuition about the fake barn case. He rather argues that we should assess this case differently because it is analogue to a case which we intuitively assess differently. [29] One might object that a handful of cases with uniform intuitions cannot prove the claim that as a rule philosophers' intuitions are stable. That may be true. But then the findings of experimental philosophy also do not show that intuitions in general are unstable. In correspondence Weinberg clarifies that he takes the findings only as evidence for unexpected patches of error that call for being embedded in a hopeful practice of error-correction. But again, I don't think that Weinberg is successful in defending the claim that only hopeful sources are trustworthy. [30] Jens Kipper has suggested to me that it is dialectically inappropriate to counter the appeal to intuition just by claiming that one does not share the respective intuition. According to Kipper this would not be a very effective move in the debate and usually leads to stalemate. If this were true, the absence of such dissenting intuitions from the debate would not show that there are no deviant intuitions. But I think that Kipper's dialectical explanation does not fit the data. For example, in his discussion of Stanley's subject-sensitive invariantism, Blome-Tillmann (2007 Blome-Tillmann, M. (2007). Scepticism and contextualism. Unpublished thesis, Oxford [Google Scholar], pp. 180f) claims that he does not share Stanley's intuitions about knowledge-attributions to certain cases. One probably can find more examples of this sort in the literature. [31] Sometimes it is objected that there are (at least) a few dissenting philosophers on the Gettier cases. For example, Sartwell (1992 Sartwell, C. 1992. Why knowledge is merely true belief. The Journal of Philosophy, 89: 167–180. [Crossref], [Web of Science ®] , [Google Scholar]) claims that knowledge is merely true belief. This view implies that a Gettier victim would have knowledge. Hetherington (1999 Hetherington, S. 1999. Knowing failably. The Journal of Philosophy, 96: 565–587. [Crossref], [Web of Science ®] , [Google Scholar]) explicitly maintains that there is knowledge in the Gettier case. But in both cases the authors do not claim to have the intuition that there is knowledge in the Gettier case. Sartwell defends his view by an abstract line of argument according to which 'knowledge' is a goal-concept and truth is our only epistemic goal. Hetherington supports his assessment of the Gettier cases by a general analysis of knowledge. See, e.g., "all things considered… if you are in a Gettier case… you do have an ability which may be regarded as constituting knowledge. Perhaps this knowledge would lack strong justificatory pedigree. But if the ability analysis is correct, that kind of pedigree is not always required by knowledge" (Hetherington, 2006 Hetherington, S. 2006. "How to know (that knowledge-that is knowledge-how)". In Epistemology futures, Edited by: Hetherington, S. 71–94. Oxford: Clarendon Press. [Google Scholar], p. 87). [32] For a more comprehensive criticism of the view that we have privileged first person access to conceptual intuitions compare Horvath (this issue). [33] This view is defended by Lewis (1994 Lewis, D. 1994. "Reduction of mind". In Papers in metaphysics and epistemology, Edited by: Lewis, D. 291–324. London: Cambridge University Press. [Google Scholar]) and Weatherson (2003 Weatherson, B. 2003. What good are counterexamples?. Philosophical Studies, 115: 1–31. [Crossref], [Web of Science ®] , [Google Scholar]). [34] I rather should say that this is what Weinberg would have to claim, if he were to use the reliability argument. As I mentioned before, he does not go this far. He need not make such a strong claim in order to show that intuitions are hopeless. But, as I said before, I don't think that intuitions have to be hopeful in order to be trustworthy. [35] Admittedly, there are further options to explain the stability of philosophers' intuition by irrelevant factors. Weinberg suggests that it is the uniform cultural background of professional philosophers that does this job in the Gettier cases. But this claim is not empirically established yet. So far, we have no experimental findings about the Gettier intuitions of non-Western professional philosophers. So Weinberg's suggestion is just a conjecture. And this conjecture does not fit well to the observation that other biases (as theory contamination) are absent from the philosopher's intuitions about Gettier cases. On my view the conjecture looks pretty much ad hoc. Max Seeger has suggested to me that there may be cognitive biases in play that are shared by professional philosophers only. Of course, it is true that stability neither guarantees reliability nor trustworthiness. [36] I do not want to dispute the fact that in some cases counterfactual judgments are based on the observation of factual regularities. Thanks to Jonathan Weinberg and Anand Vaidya for reminding me of this point. [37] Compare Cummins (1998 Cummins, R. 1998. "Reflection and reflective". In Rethinking intuition: The psychology of intuition and its role in philosophical inquiry, Edited by: DePaul, M and Ramsey, W. 113–127. Lanham, MD: Rowman & Littlefield. [Google Scholar], p. 118). [38] Clear cases are such that the subject forms a (highly) confident intuitive judgment about them. Wright (unpublished manuscript) provides empirical evidence that unstable judgments (as, for example, folk judgments about the Gettier cases) are accompanied by a low degree of subjective confidence. [39] This may also explain the findings on ethnic and gender variations of folk intuitions. [40] What about the following suggestion made by an anonymous refree: couldn't one overcome this difficulty by carefully characterizing the cases to the folk? One could also check their understanding by asking them about the relevant features. I do not want to deny this option. It may well be that laypersons can be put into a more ideal epistemic position if we confront them with more explicit descriptions of the respective hypothetical cases and if we teach them the background that is required to understand the relevance of some of the details of the cases. But it seems clear to me that the current methodology of experimental philosophers does not ensure that the test subjects are in anything like an optimal epistemic position when it comes to the evaluation of the tested hypothetical cases. Thanks to Joachim Horvath for helping me to clarify this point. [41] Thomas Nadelhoffer may be an exception here. He seems to have started a survey on ethical intuitions of philosophers.