Why Metaphysical Abstinence Should Prevail in the Debate on Reductionism

Abstract

My main aim in this paper is to show that influential antireductionist arguments such as Fodor’s, Kitcher’s, and Dupré’s state stronger conclusions than they actually succeed in establishing. By putting to the fore the role of metaphysical presuppositions in these arguments, I argue that they are convincing only as ‘temporally qualified argument’, and not as ‘generally valid ones’. I also challenge the validity of the strategy consisting in drawing metaphysical lessons from the failure of reductionist programmes. What most of these antireductionist standpoints have in common is a pretension to methodological imports. I conclude by explaining why, in order to remain relevant for scientific practice, antireductionist arguments should stay clear of metaphysics, for the latter, I argue, does not mix well with a taste for methodological prescriptions.

Acknowledgements

I would like to thank two anonymous referees for very helpful comments and suggestions on an earlier draft of this paper.

Notes

[1] I take the terminology ‘temporally qualified’ from Nagel (1961). I will return later to Nagel’s use of the terminology.

[2] I leave aside ‘local’ antireductionist arguments, that is, arguments concerned with specific issues of reducibility in light of our current knowledge in a given field, for such arguments are clearly temporally qualified arguments. An example is Kincaid’s (1990) discussion of a number of recent central results from molecular biology, showing their irreducibility to biochemistry, and Robinson’s (1992) critique of it, based on a close look at what biochemistry actually offers today. In both cases, the arguments are grounded (or purport to be) in what happens to be our current state of knowledge in these disciplines.

[3] For a recent review of the differences between Nagel’s account of reduction and other accounts of reduction as an explanatory strategy (such as Putnam and Oppenheim’s), see Steel (2004). These differences do not matter here, since antireductionist arguments aim mainly at the minimal common core of reductionist claims just mentioned. Note also that the antireductionist standpoints discussed here do not respond to subsequent refinements of the classical concept of reduction, such as that proposed by Hooker (1981), or to the ‘structuralist’ approach of Nagelian reduction (see, for instance, Balzer, Moulines, and Sneed 1987, chap. 4, or Bickle 1998, chap. 3); nor do they deal specifically with the functionalist conception of reduction developed by Kim (1998, chap. 4).

[4] ‘Natural kinds’ simply refers here to objects or processes discriminated by a science. It need not be understood in an essentialist way or in the platonic sense of ‘carving nature at its joints’.

[5] In a nutshell, the macroexplanation goes like this: imagine in a population at sexual maturity a departure from a sex ratio of 1:1—females, for instance, outnumber males. With males having more chance to mate, they will have more offspring than females. Parents genetically predisposed to produce more males will thus be favoured by natural selection, hence a return to equilibrium (i.e. to a sex ratio of 1:1).

[6] And indeed, Kitcher writes the following: ‘Explanatory patterns that deploy the concepts of cytology will endure in our science because we would foreswear significant unification (or fail to employ relevant laws, or fail to identify causally relevant properties) by attempting to derive the conclusions to which they are applied using the vocabulary and reasoning patterns of molecular biology’ (Kitcher 1984, 371; my italics).

[7] See, for instance, the ongoing debate between Fodor (1974, 1997) and Kim (1992, 1993, 1998) on the issue of genuine levels of causally relevant properties, and Davies (1996) on how Dupré’s ontological thesis of ‘equal causal status’ (discussed later in this paper) is also undermined by Kim’s argument on causal powers.

[8] For distinct, more radical lines of attack against the multiple realizability argument, see Sober (1999).

[9] The thesis is for instance defended by Kitcher (2001, chap. 4), with special attention paid to biology. Hacking (1999) also provides a nice illustration of the thesis in his chapter ‘Kind‐making: the case of child abuse’.

[10] See Bickle (2003) for case studies of successful reductionist approaches in neuroscience, such as the explanation of memory consolidation in terms of molecular mechanisms of long‐term potentiation or sensory experiences induced by cortical microstimulation. Cases of successful reductions in physics are analysed, for instance, by Morrison (2000).

[11] ‘The activities of the philosophers are simply irrelevant to my scientific life. This view, or one very similar to it, is most certainly held by a vast majority of practising scientists in the English‐speaking world’ (Gale, 1984; italics in the original). This is how a paper published in 1984 in Nature, about the relationship between science and philosophy, begins. It is indeed usually admitted that the time when philosophical considerations were commonly invoked in scientific debate is gone. As interestingly noticed by Nickles (1987 536, n. 33), the drastic drop‐off in scientists’ interest in philosophy is, for instance, indicated by the change in the editorial board of Philosophy of Science. The original board of 1934 was composed almost entirely of leading scientists. Fifty years later, every member of the board was a card‐carrying philosopher of science.