Pessimistic Inductions: Four Varieties

Abstract

The pessimistic induction (PI) plays an important role in the contemporary realism/anti-realism debate in philosophy of science. But there is some disagreement about the structure and aim of the argument. And a number of scholars have noted that there is more than one type of PI in the philosophical literature. I review four different versions of the PI. I aim to show that PIs have been appealed to by philosophers of science for a variety of reasons. Even some realists have appealed to a PI. My goal is to advance our understanding of (1) what the various PIs can teach us about science and (2) the threat posed by PIs to scientific realism.

Acknowledgements

I thank Lori Nash for critical feedback on earlier drafts. This paper was inspired in part by Juha Saatsi's paper on Laudan's ‘A Confutation of Convergent Realism’ presented at the British Society for the Philosophy of Science meeting in Dublin, Ireland, in 2010, and a Skype conversation with Larry Laudan on 16 February 2014 about the structure of his argument in ‘A Confutation of Convergent Realism’. I also thank the editor and two referees for International Studies in the Philosophy of Science for their thoughtful feedback on my paper. Their suggestions have improved the paper significantly.

Notes

1. Saatsi, for example, argues for the importance of distinguishing between ‘Laudan's argument and Putnam's rhetoric’ (Saatsi 2005, 1091). He also believes that Poincaré presents a PI (Saatsi 2005, 1088).

2. I thank the editor and one of the referees for International Studies in the Philosophy of Science for suggesting I add this general statement about the form of a PI.

3. This remark is a bit misleading, implying that there is a single best theory of the electron. Ian Hacking, though, notes that ‘even people in a team, working on different parts of the same large experiment, may hold different and mutually incompatible accounts of electrons’ (Hacking 1983, 264). Hence, there may not be a single best theory of the electron.

4. This is the version of the PI that Nola (2008) criticizes. In his attack on the PI, Nola appeals to the causal theory of reference, a theory that Putnam is partly responsible for developing. Nola also suggests that an optimistic induction is better supported than Putnam's PI from the history of science. Devitt (2011, 289), though, rightly suggests that it is problematic to attempt to address the PI by appealing to a theory of reference, given the disagreement amongst philosophers about theories of reference.

5. Putnam suggests that the no-miracles argument is the strongest argument in support of realism.

6. Fahrbach (2011) makes a conjecture of this sort, though he is not so specific as to claim that 30% is the appropriate number. Fahrbach also argues that the vast majority of theories have been developed in the recent history of science, since 1900.

7. I have challenged this realist argument elsewhere (Wray 2012). Roughly, I argue that developments in methodology may merely enable us to develop theories that are empirically more successful but not necessarily theories that are closer to the truth than the theories we currently accept.

8. Fresnel's wave theory replaced the particle theory developed earlier by Newton. Fresnel's theory was replaced by Maxwell's, which dispensed with the ether, the medium through which Fresnel believed light waves travelled. Maxwell regarded light as a periodic disturbance ‘in the “disembodied” electromagnetic field’ (Worrall 1989, 116). ‘Einstein reintroduced particles; and finally the “probability waves” of Quantum Mechanics came up’ (Fahrbach 2011, 141–142).

9. Forster and Sober (1994, 28), Papineau (1996, 14), Lipton (2004, 145), Magnus and Callender (2004), Chakravartty (2007), and Fahrbach (2011, 141) attribute the PI to Laudan, but they do not reconstruct it as a reductio ad absurdum. Frost-Arnold (2011) is more cautious, and merely notes that Laudan's ‘Confutation’ is responsible for the recent attention to the PI.

10. My interpretation of Laudan is not only supported by my own reading of Laudan's article and Lyons's reading of Laudan, but also by remarks made by Laudan during a Skype conversation on 16 February 2014.

11. Laudan's list has generated a vast body of critical literature (see, for example, Bishop 2003, section 2.2; Mizrahi 2013, 3219–3220). Mizrahi is especially concerned that Laudan's examples are not a random sample of the target population of successful theories.

12. Some may question whether Poincaré does in fact present a PI. I hope my argument below shows that he does in fact present one. Nonetheless, he is read by others as presenting a PI (see, for example, Lange 2002, 281; Saatsi 2005, 1088).

13. Kuhn (2000, 97–98) makes a similar remark about the increasing specialization that characterizes science. For an extended discussion of Kuhn's view on specialization, see Wray (2011, chapter 7).

14. Doppelt (2007) also recognizes that the PI helps to clarify what a plausible form of realism entails or commits one to. There are now a number of forms of modest realism. Entity realists, for example, are committed to belief in the existence of the theoretical entities scientists manipulate routinely in laboratory operations, but recognize that scientists may be mistaken about some of the properties that they ascribe to these entities (see Hacking 1983, chapter 16). Alternatively, Chakravartty (2008, 155) has suggested that the most viable form of realism is ‘a realism about well-confirmed properties’ rather than entities.

15. One might object to Mizrahi looking at scientific laws to either support or undermine the PI, for newly discovered laws sometimes do not require the overthrow of the accepted theory. Boyle's Law, for example, merely identified a previously unnoticed relationship between two variables, the pressure of a gas and its volume (see Kuhn [1962] 2012, 28). Kuhn discusses this example in a chapter of Structure devoted to normal science. This reminds us that we must be clear about what the unit of analysis is in the PI.