Abstract
The underdetermination of theory by data argument (UD) is traditionally construed as an argument that tells us that we ought to favour an anti-realist position over a realist position. I argue that when UD is constructed as an argument saying that theory choice is to proceed between theories that are empirically equivalent and adequate to the phenomena up until now, the argument will not favour constructive empiricism over realism. A constructive empiricist cannot account for why scientists are reasonable in expecting one theory to be empirically adequate rather than another, given the criteria he suggests for theory choice.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Ladyman et al. (1997) discuss a related issue, namely whether CE can admit a version of IBE. I will later show how my points differ from their discussion.
The strongest version of UD tells us that that there are cases of strong empirical equivalence. Two theories are strongly empirically equivalent if they are mutually incompatible but have the same empirical content all the same. Furthermore, no empirical facts past, present or future will give us evidence for choosing one over the other. In cases like this I think there are different strategies open to a realist in countering UD. If for example two theories are equally supported by the evidence but one theory accommodates the evidence post hoc and the other predicts it, realists would argue; that a theory has novel predictions would count in its favour. (There is a wide discussion of whether evidence post hoc. can ever count in a theory’s favour or not, see for example Psillos (1999), for a discussion of so-called use-novelty). In cases where two theories are equally supported by the evidence, and each theory is supported to an equal extent by old and new evidence in turn, the most common strategy for a realist will be to adopt an explanationist strategy and postulate the existence of super-empirical values. Even though this is arguably a question-begging procedure, it is my opinion that the debate between realists and anti-realist leaves it open who has the burden of proof in this case. Even though the realist will not convince any prospective anti-realists by doing so, she can at least do so while being internally consistent. There is a large body of literature on this, arguing in favour of realism; see for example Devitt (2002, 2005). For an argument that what shall count as evidence in the future is contingent (thereby denying the premise that all evidence is equal between the two theories) see Laudan and Leplin (1991). For an argument saying that the strongest versions of UD are produced by ‘semantic trickery’ see Hoefer and Rosenberg (1994). Finally, for an argument against the points made by Laudan and Leplin and Hoefer and Rosenberg see Kukla (1998).
Traditionally, voluntarism is the view that having a belief is within the control of a person. Our epistemic life is conceived of as involving certain subjective choices. Van Fraassen’s voluntarism does not however entail that someone can believe anything they want. There are certain constraints that van Fraassen calls ‘epistemic policies’. “If we choose an epistemic policy to govern under what conditions and how far, we will go beyond the evidence in our beliefs, we will be setting down certain boundaries” (van Fraassen 1985, p. 254). These policies are hot however derived from the empirical evidence that we have got, rather they are endorsed by virtue of us having certain commitments. This is more fully explored in van Fraassen (2002), where having such commitments is spelled out in terms of one taking a stance. If CE can be defended while combined with a different view in epistemology some of the points that I will go on to raise will not apply. However, it is not easy to work out what other kind of position in epistemology CE could be combined with. Van Fraassen has made the distinction between belief and acceptance central to CE and voluntarism is perfect for spelling out how we are to understand acceptance.
It should be noted at this point that acceptance is not to be thought of as being binary. Acceptance can be qualified — it is a matter of degree, where the degree to which one accepts a hypothesis is proportional to the evidence one has in favour of it.
According to van Fraassen, the context determines which scientific factors are explanatorily relevant. However as the context is always relative to the specific interests and goals of the person in question explanation cannot be epistemically relevant.
At this point it should be mentioned that according to van Fraassen, being rational is not necessarily a matter of following rules. Therefore it is open to him to object to a reference to intuition at this point, that that by itself does not introduce the need for rules. We may sometimes go beyond what is evident to us at a particular time without that involving any rule based principle, and yet be entirely rational. (I wish to thank an anonymous referee for this Journal for pointing this reply out to me). I think this is a reply that is entirely consistent with the constructive empiricist tradition, but in so far as we form expectations about the future, it is very hard to make sense of the idea that such expectations would be formed if we did not believe that certain regularities obtained. If it is indeed granted that such regularities obtain, it would be tempting to assume that our expectations are formed on the basis of inferences that we assume correspond to these regularities in one way or another (regardless of whether they actually do so). If one holds a view according to which it is reasonable to hold certain expectations about the future, without relying on any rule based reasoning, it seems to undermine the very idea of it ever being more reasonable to expect anything rather than something else.
Van Fraassen argued (1980) that we can explain the success of science by appeal to how theories come to be successful by natural selection. “Only the successful theories survive [...]” (van Fraassen 1980, p. 40). In this way it may be argued that van Fraassen has in fact coined a connection between pragmatic considerations and the aim of scientists. But as has been argued by Leplin (1997) (among others) the competition amongst theories does not satisfy some of the conditions necessary for a process of Darwinian selection to take place. Leplin sets out an analogy with tennis-playing abilities of Wimbledon finalists. The question we want to answer is: how come Wimbledon finalists are so good at tennis? We can interpret this question in two ways. If we want to know why it is in general that Wimbledon players are so good at tennis, we can cite the selection criteria for getting into the tournament. That answer though, does not tell us why individual finalists like Becker and Agassi are so good at tennis. In order to answer the latter question we have to cite the relevant properties of those particular players, which may involve training and genetics. The analogy with the Darwinian account of success of theories is the following. The Darwinian account may explain how we come to be in possession of successful theories, but it does not have anything to say about what it is about our particular successful theories, that make them successful.
Van Fraassen (1980) argued that IBE could perhaps be used at the observable level, but this is denied in Ladyman et al. (1997) that is a response to Psillos (1996) arguing that accepting IBE for observables but not for unobservable lead to inconsistency. Psillos (1997) argues that denying that IBE is reliable at even the observable level will force constructive empiricism into scepticism.
References
Devitt, M. (2002). Underdetermination and realism. In E. Sosa, & E. Villanueva (Eds.) Philosophical issues, 12, realism and relativism. Boston, MA & Oxford, UK: Blackwell Publishing.
Devitt, M. (2005). Scientific realism. In F. Jackson, & M. Smith (Eds.) The Oxford handbook of contemporary philosophy, Chapter 26 (pp. 767–791). UK: Oxford University Press.
Hoefer, C., & Rosenberg, A. (1994). Empirical equivalence, underdetermination and systems of the world. Philosophy of Science, 61, 592–602.
Kukla, A. (1998). Studies in scientific realism. New York, Oxford: Oxford University Press.
Ladyman, J., van Fraassen, B. C., Douven, I., & Horsten, L. (1997). A defence of van Fraassen’s critique of abductive reasoning: reply to Psillos. Philosophical Quarterly, 47, 305–321.
Laudan, L., & Leplin, J. (1991). Empirical equivalence and underdetermination. Journal of Philosophy, 88, 269–285.
Leplin, J. (1997). A novel defense of scientific realism. Oxford: Oxford University Press.
Psillos, S. (1996). On Van Fraassen’s critique of abductive reasoning. Philosophical Quarterly, 46(182), 31–47.
Psillos, S. (1997). How not to defend CE: a rejoinder. Philosophical Quarterly, 47(188), 369–372.
Psillos, S. (1999). Scientific realism, how science tracks the truth. London and New York: Routledge.
Van Fraassen, B. C. (1980). The scientific image. Oxford: Clarendon Press.
Van Fraassen, B. C. (1985). Empiricism in the philosophy of science. In P. M. Churchland, & C. A. Hooker (Eds.) Images of science (pp. 245–305). Chicago, IL: The University of Chicago Press.
Van Fraassen, B. C. (2002). The empirical stance. Connecticut: Yale University Press.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Busch, J. Underdetermination and Rational Choice of Theories. Philosophia 37, 55–65 (2009). https://doi.org/10.1007/s11406-008-9133-9
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11406-008-9133-9