THE PAST AND FUTURE OF EXPERIMENTAL PHILOSOPHY

Abstract

Experimental philosophy is the name for a recent movement whose participants use the methods of experimental psychology to probe the way people think about philosophical issues and then examine how the results of such studies bear on traditional philosophical debates. Given both the breadth of the research being carried out by experimental philosophers and the controversial nature of some of their central methodological assumptions, it is of no surprise that their work has recently come under attack. In this paper we respond to some criticisms of experimental philosophy that have recently been put forward by Antti Kauppinen. Unlike the critics of experimental philosophy, we do not think the fledgling movement either will or should fall before it has even had a chance to rise up to explain what it is, what it seeks to do (and not to do), and exactly how it plans to do it. Filling in some of the salient details is the main goal of the present paper.

Keywords:

• experimental philosophy
• conceptual analysis
• meta-philosophy
• folk intuitions

ACKNOWLEDGEMENTS

For helpful comments, we would like to thank Joshua Alexander, Adam Feltz, Antti Kauppinen, Joshua Knobe, Trevor Kvaran, Shaun Nichols, and Jonathan Weinberg.

Notes

1. Authorship is equal.

2. Action Theory: Adams and Steadman (2004a, 2004b), Knobe (2003a, 2003b, 2004b, 2005a, 2005b), Knobe and Burra (2006a, 2006b), Knobe and Mendlow (2004), Leslie, Knobe, and Cohen (2006), Machery (forthcoming), Malle (2001, 2006), Malle and Knobe (1997), McCann (2005), Meeks (2004), Nadelhoffer (2004a, 2004b, 2005, 2006a), Nichols and Ulatowski (forthcoming), Phelan and Sarkissian (forthcoming), and Young et al. (2006). Epistemology: Alexander and Weinberg (2006), Baron (1998), Bishop and Trout (2005a, 2005b), Feltz (n.d.), Nichols, Stich, and Weinberg (2003), Stich and Nisbett (1980), Stich and Weinberg (2001), Swain, Alexander, and Weinberg (forthcoming), and Weinberg, Nichols, and Stich (2001). Ethics: Cushman, Young, and Hauser (2006), Doris (2002), Greene (2002, 2003, forthcoming), Greene and Haidt (2002), Greene et al. (2004), Haidt (2001, 2003), Haidt, Bjorklund, and Murphy (2000), Hauser (2006), Hauser, Young, and Cushman (forthcoming), Knobe (2005a), Knobe and Leiter (forthcoming), Knobe and Roedder (forthcoming), Nichols (2002, 2004b, 2004c), Nichols and Mallon (2006), Pizarro and Bloom (2003), Pizarro, Uhlmann, and Bloom (2003), and Pizarro, Uhlmann, and Salovey (2003). Free Will & Moral Responsibility: Knobe and Doris (forthcoming), Nadelhoffer and Feltz (forthcoming), Nahmias (2006), Nahmias et al. (2005, 2006, forthcoming), Nichols (2004c, 2006a, 2006b), Nichols and Knobe (forthcoming), Turner and Nahmias (2006), Woolfolk and Doris (2002), and Woolfolk, Doris, and Darley (forthcoming). Philosophy of Language: Machery et al. (2004), and Mallon et al. (forthcoming). Philosophy of Law: Carlsmith, Darley, and Robinson (2002), Greene and Cohen (2004), Knobe (2005b), Malle and Nelson (2003), Nadelhoffer (2006a), and Sardjevéladzé and Machery (n.d.). Philosophy of Mind: Gray, Gray, and Wegner (2007), and Knobe and Prinz (forthcoming). Philosophy of Science: Stotz (forthcoming), Stotz and Griffiths (2004), and Stotz, Griffiths, and Knight (2004).

3. We will speak primarily of ordinary (or ‘folk’) intuitions, which we take roughly to be dispositions to make certain non-deductive, spontaneous judgments, for instance, about whether or not a particular concept applies in a particular situation (see Goldman and Pust 1998; Jackson 1998). Some experimental philosophy, however, may be better understood as investigating folk theories or folk concepts, and the complicated connections between intuitions, theories, and concepts adds a layer of complexity both to what experimental philosophers take themselves to be doing and what their critics take them to be doing.

4. In fact, we think the criticisms of experimental philosophy that have been put forward so far (e.g. Kauppinen 2007; Levy n.d.; Sosa 2005, 2007) are philosophically interesting enough that experimental philosophers deserve some credit for having inspired them! We also think it is essential for experimental philosophers themselves to examine these metaphilosophical questions as they present their research so that they are as clear as possible about how they think their own research bears on philosophical debates.

5. For the purposes of this paper, we need not try to resolve the tension that exists between the various projects within experimental philosophy. Nor do we need to take sides. Our present goal is just to identify and briefly explore three fairly distinct projects within experimental philosophy.

6. The boundary is blurry (i) because most experimental philosophers are happy to obtain the data they take to be relevant to the philosophical issues they are exploring by drawing on already existing scientific research, and (ii) because some scientists are interested in the same kind of intuitions that interest philosophers and they, too, discuss the philosophical implications of their research (e.g. Baron 1998; Damasio 1994; Greene 2002, 2003, forthcoming; Haidt 2001, 2003; Hauser 2006; Hauser et al. forthcoming; Malle 2001, 2006; Nisbett 2003; Nisbett and Ross 1980; Wegner 2002). Because ‘experimental philosophy’ is perhaps best viewed as a family resemblance term, trying to explicate the movement in terms of necessary and jointly sufficient conditions would be wrong-headed even if well-intentioned. But the following two questions are nevertheless helpful. First, do you run controlled and systematic studies and use the resultant data to shed light on philosophical problems? Second, do you sometimes address the tension that exists between what philosophers say about intuition and human cognition, on the one hand, and what researchers are discovering about these things, on the other hand? Experimental philosophy can be seen as a movement at the cross-roads of psychology and philosophy that is being driven primarily by people who answer these two questions in the affirmative.

7. See, e.g. Jackson (1998), Jackson and Pettit (1995), Lewis (1972), and Smith (1994).

8. This shared distrust in armchair reflection and informal dialogue as generally reliable guides to folk intuitions is roughly what Kauppinen describes as ‘EXPERIMENTALISM − ’ (2007, 000).

9. See, e.g. Mynatt et al. (1977), Nickerson (1998), and Wason (1960).

10. See, e.g. Ditto and Lopez (1992), and Edwards and Smith (1996).

11. See, e.g. Fields and Schuman (1976), and Ross, Greene, and House (1977).

12. We are not claiming that philosophers are in fact more susceptible to these well-known biases—that, of course, is an empirical question that calls for controlled and systematic investigation. Rather, we are merely suggesting that given the central role that many philosophers give to folk intuitions, there is reason to worry about what Alvin Goldman and Joel Pust call ‘theory contamination’ (1998, 183). Minimally, we believe that the gathering data on cognitive biases place the burden clearly at the feet of philosophers who claim their own intuitions are a reliable guide to what people generally find intuitive.

13. The three projects we describe are not meant to label individual philosophers but rather individual research projects or articles. Some philosophers, such as Shaun Nichols, have worked on all three of the projects we will discuss.

14. What we are calling EA is roughly equivalent to what Kauppinen calls ‘optimistic experimentalism’ (2007, 99), and what Alexander and Weinberg call the ‘proper foundations view’ (2006). Some examples of EA include Knobe (2003a, 2003b, 2004a, 2004b), Nadelhoffer (2005, 2006a, 2006b), Nahmias et al. (2005, 2006, forthcoming), Nichols (2004a, 2004b), and Nichols and Ulatowski (forthcoming).

15. As we will see, not only is there no consensus among philosophers in general concerning the proper role (if any) that folk intuitions have to play in philosophy, but experimental philosophers are themselves divided on this issue.

16. Robert Kane exemplifies this view with his claim that ‘most ordinary persons start out as natural incompatibilists. They believe there is some kind of conflict between freedom and determinism … Ordinary persons have to be talked out of this natural incompatibilism by the clever arguments of philosophers’ (1999, 217).

17. See, e.g. Adams and Steadman (2004a, 2004b), Cushman, Young, and Hauser (2006), Machery (forthcoming), Malle (2006), McCann (2005), Meeks (2004), Nadelhoffer (2004b, 2005), Nichols and Ulatowski (forthcoming), and Phelan and Sarkissian (forthcoming).

18. Some examples of ED work in experimental philosophy include Greene (2002, 2003, forthcoming), Greene and Haidt (2002), Greene et al. (2004), Knobe and Doris (forthcoming), Nadelhoffer (2006a), and Nichols and Knobe (forthcoming). Knobe's (2007) response to Kauppinen (2007) offers a more extended defense of the philosophical relevance of at least this branch of experimental philosophy, arguing that it advances our understanding of human nature in the ways traditional philosophers have attempted to do since the Greeks.

19. Greene is an interesting case as he is appointed in a psychology department despite having done his graduate work in philosophy. He is also one of only a handful of psychologists who self-identifies with the experimental philosophy movement.

20. What we are calling ER maps roughly onto what Kauppinen calls ‘pessimistic experimentalists’ (2007, 98). For some examples of ER projects, see Alexander and Weinberg (2006), Machery et al. (2004), Mallon et al. (forthcoming), Nichols, Stich, and Weinberg (2003), and Weinberg, Nichols, and Stich (2001).

21. ER is heavily rooted in the pioneering work begun by Stich in the early 1980s. Indeed, not only did Stich pave the way for ER, he also participated in some of the first studies in experimental philosophy. See, e.g., Nichols, Stich, and Weinberg (2003), Stich (1983, 1984, 1985, 1990, 1994, 1996, 1998), Stich and Nisbett (1980), Stich and Weinberg (2001), and Weinberg, Nichols, and Stich (2001).

22. Another project within the ER camp is to show that seemingly irrelevant features significantly influence the intuitions people express. For instance, the order in which cases are presented seems to influence people's epistemic intuitions (see Swain, Alexander, and Weinberg forthcoming).

23. See Nadelhoffer and Nahmias (n.d.).

24. For the purposes of this paper, we neither pursue nor examine various other responses experimental philosophers have already offered to critics. Knobe (forthcoming), for instance, suggests that even if Kauppinen were correct that experimental philosophers don't (and can't) get at the intuitions philosophers need for doing conceptual analysis, experimental philosophy is nevertheless useful for philosophers who are not interested in doing conceptual analysis but who are interested instead in the naturalistic project of understanding what the mind is and how it works. Another important issue that we will leave unaddressed is that many of the criticisms that have been put forward so far by Kauppinen, Sosa, and others rely too heavily on assumptions that are driven by unsettled issues in linguistics and the philosophy of language. As Machery (forthcoming) has correctly pointed out, until we have a better empirical understanding of the nature of concepts and how they are used, we simply cannot resolve several of the ongoing debates both about and within experimental philosophy concerning intuitions and conceptual analysis. We think Machery may be largely correct about this. But even so, as he goes on to suggest, experimental philosophers can nevertheless continue to fruitfully investigate the nature and limitations of human cognition. We believe that when the arguments put forward in the present paper are viewed in conjunction with the arguments developed by Knobe, Machery, Alexander and Weinberg, and others, we find a multi-pronged defense of experimental philosophy that successfully staves off all would-be assassins.

25. See Bealer (1996, 1998, 2000), Levy (n.d.), and Williamson (2004).

26. A similar point has been developed by Ernest Sosa—see, e.g. Sosa (2005, 2007). A response to Sosa can be found in Alexander and Weinberg (2006).

27. Some graduate students at Florida State University recently ran a pilot study that involved presenting participants with various cases about intentional action and allowing them to discuss and debate the cases among themselves. At the end of the study they took further surveys to examine how the students' views changed (or did not change). And while the results were inconclusive—owing primarily to some problems with the design of the studies—their strategy is certainly one that could prove useful in the future.

28. This is clearly not to suggest that philosophers should quit having informal conversations with their children, families, friends, students, and colleague about philosophical issues! Nor is to suggest that all of the information that we glean from these conversations is useless or uninteresting. We are simply cautioning against the dogmatic assumption that there is data that we can best (or only) obtain in informal one-on-one conversations.

29. Even in cases where people express consistent intuitions across cases, ERs argue that if people in different cultures or classes express consistently different intuitions, then picking out which people get to count as competent will often beg the question at issue. Hence, robust intuitions within a group may not satisfy Kauppinen's criterion of competence for robust intuitions.

30. A statistically significant result is one for which the probability that the result occurred due to chance factors was less than 5 percent—a p value of < 0.05 indicates this probability.

31. Contrast this with the way most in-class ‘surveys’ of students' intuitions are conducted: the teacher describes the thought experiment or scenario verbally, usually once quickly, and then asks students to raise their hands on the spot to indicate their intuitions about it. See Nadelhoffer and Nahmias (forthcoming) for an explanation of the usefulness of the controlled survey method for teaching philosophy in the classroom.

32. See, e.g. German and Nichols (2003), and Harris, German, and Mills (1996).

33. Some argue that all intuitions are suffused with background theoretical beliefs. As we mentioned earlier, the connections between intuitions and theories (and concepts) are complex, and we suspect experimental philosophy can help provoke philosophical debates about these connections, and presumably it can gather empirical data relevant to such debates.

34. A similar objection has been put forward by Ernest Sosa (see, e.g. 2005, 2007).

35. In a 2003 interview with Tamler Sommers for The Believer (http://www.naturalism.org/strawson_interview.htm).

36. Another explanation for divergent results is that some concepts may not have clear necessary and sufficient conditions (or at least, not the folk's concepts). In such cases, people may have shared intuitions about the paradigmatic applications of the concepts, but differing intuitions about the boundaries, as demonstrated by their different judgments about certain, often non-paradigmatic, cases.

37. We have worries about the way Nichols and Knobe describe determinism that make us skeptical that they have shown most ordinary people have incompatibilist intuitions (see Nahmias 2006; Turner and Nahmias 2006), but we think the differences they find across scenarios are fascinating and important.

38. For instance, an initial look at the data suggests that many people answer differently to questions about (i) free will compared to moral responsibility and (ii) praise compared to blame, while (iii) people answer almost identically to questions about free will compared to questions about whether decisions are ‘up to’ the agent.

39. See Vohs and Schooler (n.d.) who show that inducing people to doubt the existence of free will increases their willingness to cheat on a test.