Abstract
Among the most serious challenges to scientific realism are arguments for the underdetermination of theory by evidence. This paper defends a version of scientific realism against what is perhaps the most influential recent argument of this sort, viz. Kyle Stanford’s New Induction over the History of Science. An essential part of the defense consists in a probabilistic analysis of the slogan “absence of evidence is not evidence of absence”. On this basis it is argued that the likelihood of a theory being underdetermined depends crucially on social and historical factors, such as the structure of scientific communities and the time that has passed since the theory first became accepted. This is then shown to serve as the epistemological foundation for a version of scientific realism which avoids Stanford’s New Induction in a principled and non-question-begging way.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Since nothing in this paper turns on the distinction between true and approximately true theories, I shall henceforth use “true” to stand for “true or approximately true”.
This is not to say that algorithms of this sort would not be a problem for scientific realism as well. The point is rather that the problem would extend also to a host of other philosophical positions, including anti-realist positions such as Stanford’s own instrumentalism [which does not deny the possibility of inductive knowledge in general and even allows for the existence of some scientific knowledge concerning unobservable entities and processes (see e.g. Stanford 2006, pp 31–33)].
Stanford focuses on scientific claims about “fundamental constituents of the natural world” (Stanford 2006, p 32) rather than those about unobservable objects. In what follows, this difference will not be important.
In stating RTU I have deliberately been somewhat vague on what counts as a serious rival since the positions and arguments that follow do not turn on precisely how this notion is defined. Two points are worth emphasizing however: (i) For the purposes of this paper, a rival theory is simply a theory that is incompatible with the theory it rivals. Some would argue that a theory doesn’t count as a rival theory unless it at least addresses the question answered by the theory it rivals, so that (for example) the negation of a theory doesn’t normally count as a rival to the theory, despite being incompatible with it. (Leplin and Laudan 1993) While I agree (see [reference omitted]), I will operate with the more inclusive definition of “rival theory” in order to ensure that Stanford’s argument is not being construed uncharitably. (ii) I am following Stanford (2001, 2006) in characterizing a serious rival to a given theory as a theory that would be viewed by working scientists as a genuine non-skeptical alternative to the theory. This is admittedly somewhat vague since it is not entirely clear what it is for scientists to view a theory as a genuine non-skeptical alternative. Thus one could complain that RTU should be dismissed on the grounds that the alleged problem has not been clearly posed. However, since I do not want to decide the matter on a technicality of this sort, I will operate with Stanford’s admittedly somewhat vague notion in what follows.
See also Roush (2005), Magnus (2010), and Wray (2011). Of course, the name and general structure of the argument stems from Laudan’s (1981) hugely influential discussion of the Pessimistic Meta-Induction. Nevertheless, for the purposes of this paper it is important not to confuse the two arguments since the argument that follows is concerned specifically with the existence of underdetermination rivals to current scientific theories—an issue on which the Pessimistic Meta-Induction is silent.
Sober (2009) also points out that there is something to the slogan that “absence of evidence is not evidence of absence” in that absence of evidence is usually not very strong evidence for absence, given a “likelihoodist” measure of strength of confirmation. As Strevens (2009) points out, however, this is hardly a completely satisfactory interpretation of the original slogan, for it contradicts the slogan’s apparent meaning. I offer a different way of interpreting the slogan below— one that does not contradict the apparent meaning of the slogan.
Salmon (1975) famously distinguished between these two interpretations, referring to the first as the “relevance concept of confirmation” and the second as the “absolute concept of confirmation”.
This simplification obviously depends on \(F_X\) being the proposition that one has in fact found an X (e.g. as opposed to the proposition that one merely believes oneself to have found an X). It is of course possible for an agent to be mistaken about whether she has found something of a given kind, e.g. as Rene Blondlot mistakenly believed in 1903 that he had discovered a new form of radiation, “N-rays”. However, even an agent who believes herself to have found something that does not in fact exist should assign probability 1 to \( p(\lnot F_X | A_X \& B)\), since by her lights the conditional “If Xs did not exist, I would not have found any Xs” is still necessarily true. Such an agent, if rational, would not object to the conditional itself, but instead deny the consequent of the conditional and thus also its antecendent.
This frequency interpretation is appropriate in the context of of evaluating arguments for RTU, which holds that most empirically successful theories are underdetermined—i.e., that the frequency of underdetermined theories is high among empirically successful theories.
Here I am abusing notation by writing \(p(\mathcal {A} | \mathcal {B})\) instead of \(p(x \in \mathcal {A} | x \in \mathcal {B})\).
Of course, this is not to say that one could not mistakenly believe oneself we have found an underdetermination rival to a given theory even when no such rivals exist. While the possibility of mistakenly believing oneself to have found an underdetermination rival does not speak against the fact that \(p(\overline{\mathcal {F}_U} | \mathcal {A}_U, \mathcal {X}) = 1\) – which is a necessary truth since it is impossible to find something of a kind that does not exist – it does highlight the fact that it is assumed in the debate as a whole that it is possible to reliably estimate whether a given theory that has already been discovered is indeed an underdetermination rival to another theory. Indeed, this assumption is essential to the New Induction’s argumentative strategy since if we could not reliably locate historical cases of underdetermination rivals then proponents of the New Induction could not argue that they are sufficiently common in the history of science to inductively warrant the conclusion that most current successful theories are underdetermined as well.
I have deliberately chosen terms for these two factors that are not widely used in the literature on scientific theory testing in order to prevent them from being confused with a number of factors discussed in that literature. For example, although sensitivity is superficially similar to “statistical power”, i.e. the probability that a false null hypothesis will be rejected in an empirical test, it would be misleading to refer to \(p(\mathcal {F}_U | \overline{\mathcal {A}_U}, \mathcal {X})\) as a given theory’s “statistical power” since that falsely suggests that the theory has this kind of statistical power with regard to a set of results from an empirical test. After all, note that sensitivity is concerned with the probability that a group of scientists discover underdetermination rivals to theories in a given set, which is a feature that does not depend directly on the features of any empirical test.
This argument is implicit in many realist responses to the New Induction, but Devitt (2011) arguably provides the clearest and most careful objection of this sort.
Thus, assuming that unobservables are, all other things being equal, harder to conceptualize, there is some truth to the anti-realist claim that theories that concern unobservables are, all other things being equal, more likely to be underdetermined. However, as I’ll emphasize below, other things are not always, or even usually, equal, and so theories about unobservables can very well be less likely to be underdetermined than theories about observables.
See Magnus (2010) for an argument along these lines.
Indeed, anti-dogmatism is really just compliance with the norm of “organized skepticism”, which the eminent sociologist of science Robert Merton argued was one of the four norms comprising the scientific ethos.
Note that the question is not whether it’s reasonable to believe now that Fresnel’s theory had such rivals. The question is whether it was reasonable at the time.
Of course, current scientists may be more confident in the photon theory than Fresnel’s contemporaries were in his ether theory, but we shouldn’t confuse confidence in theories with a dogmatic attitude towards alternative approaches. Besides, the photon theory itself underwent a significant period in the early 20th century where there was widespread doubt about the truth of the theory.
I am assuming here that Mendelian genetics is at least as successful empirically as each of the other theories, and thus that it can be assumed to have at least as high plausibility.
As a case in point, the genetics societies of Britain and the United States were not founded until 1919 and 1931 respectively. The official journals of each society, Heredity and Genetics, were established in 1947 and 1916 respectively.
Compare: Suppose I claim that most Swedish residents are protestants. It is no good objection to my claim to point out that Petersen, a Swedish resident, is catholic—especially not if the overwhelming majority of other Swedes that we have met have been protestants. That some of the Swedes we run into are non-protestants is exactly what we should expect if I were right that most (but not all) of them are protestants.
References
Anderson, M. S., Ronning, E. A., DeVries, R., & Martinson, B. C. (2010). Extending the mertonian norms: Scientists’ subscription to norms of research. Journal of Higher Education, 81, 366–393.
Blackburn, S. (2002). Realism: Deconstructing the debate. Ratio (new series), 15, 111–133.
Bockstaele, P. (2004). The mathematical and the exact sciences. In W. Ruegg (Ed.), A history of the University in Europe Vol. III: Universities in the nineteenth and early twentieth centuries (1800-1945) (pp. 493–518). Cambridge: Cambridge University Press.
Bowler, P. J. (1989). The Mendelian revolution: The emergence of hereditarian concepts in science and society. Baltimore: John Hopkins University Press.
Buchwald, J. Z. (1989). The rise of the wave theory of light: Optical theory and experiment in the early nineteenth century. Chicago, IL: University of Chicago Press.
Cartwright, N. (1983). How the laws of physics lie. Oxford: Oxford University Press.
Chakravartty, A. (2007). A metaphysics for scientific realism: Knowing the unobservable. Cambridge: Cambridge University Press.
Devitt, M. (2011). Are unconceived alternatives a problem for scientific realism. Journal for General Philosophy of Science, 42, 285–293.
Feyerabend, P. (1963). How to be a good empiricist: A plea of tolerance in matters epistemological. In B. Baumrin (Ed.), Philosophy of science: The Delaware seminar. New York, NY: Interscience Publishers.
Gillies, D. (1993). Philosophy of science in the twentieth century. Oxford: Blackwell Publishers.
Glymour, C. (1977). The epistemology of geometry. Noûs, 11, 227–251.
Hacking, I. (1983). Representing and intervening: Introductory topics in the philosophy of natural science. Cambridge: Cambridge University Press.
Hesse, M. (1976). Truth and the Growth of Scientific Knowledge (pp. 261–280). In PSA: Proceedings of the Biennial Meeting of the Philosophy of Science Association.
Kitcher, P. (1993). The advancement of science: Science without legend, objectivity without illusions. New York: Oxford University Press.
Kitcher, P. (2001). Real realism: The Galilean strategy. The Philosophical Review, 110, 151–197.
Kuhn, T. S. (1962/1996). The structure of scientific revolutions (3rd ed.). Chicago, IL: University of Chicago Press.
Kuhn, T. S. (1977). Objectivity, value judgments, and theory choice. In T. Kuhn (Ed.), The essential tension. Chicago, IL: University of Chicago Press.
Kukla, A. (1993). Laudan, Leplin, empirical equivalence and underdetermination. Analysis, 53, 1–7.
Larsen, P. O., & von Ins, M. (2010). The rate of growth in scientific publication and the decline in coverage provided by science citation index. Scientometrics, 84, 575–603.
Laudan, L. (1981). A confutation of convergent realism. Philosophy of Science, 48, 19–49.
Laudan, L. (1990). Demystifying underdetermination. In C. W. Savage (Ed.), Scientific theories, Minnesota studies in the philosophy of science (Vol. 14). Minneapolis, MT: University of Minnesota Press.
Leplin, J., & Laudan, L. (1993). Underdetermination underdeterred: Reply to Kukla. Analysis, 53, 8–16.
Magnus, P. (2010). Inductions, red herrings, and the best explanation for the mixed records of science. British Journal for the Philosophy of Science, 61, 803–819.
Magnus, P., & Callender, C. (2004). Realist Ennui and the Base Rate Fallacy. Philosophy of Science, 71, 320–338.
Mayr, E. (1982). The growth of biological thought: Diversity, evolution, and inheritance. Cambridge, MA: Harvard University Press.
Merton, R. K. (1973). The normative structure of science. In N. W. Storer (Ed.), The sociology of science (pp. 267–278). Chicago, IL: University of Chicago Press.
Poincaré, H. (1952). Science and hypothesis. New York: Dover. Republication of the first English translation, published by Walter Scott Publishing, London, 1905.
Price, D. J., & d, S. (1961). Science since Babylon. New Haven: Yale University Press.
Price, D. J., & d, S. (1963). Little science, big science. New York: Columbia University Press.
Psillos, S. (1999). Scientific realism: How science tracks truth. London: Routledge.
Putnam, H. (1978). Meaning and the moral sciences. Boston: Routledge and Kegan Paul.
Roush, S. (2005). Tracking truth: Knowledge, evidence, and science. Oxford: Clarendon Press.
Salmon, W. C. (1975). Confirmation and relevance. In G. Maxwell & R. Anderson (Eds.), Minnesota studies in the philosophy of science, induction, probability, and confirmation (Vol. 6, pp. 3–36). Minneapolis, MT: University of Minnesota.
Sklar, L. (1981). Do unborn hypotheses have rights? Pacific Philosophical Quarterly, 62, 17–29.
Sober, E. (2009). Absence of evidence and evidence of absence: Evidential transitivity in connection with fossils, fishing, fine-tuning, and firing Squads. Philosophical Studies, 143, 63–90.
Stanford, P. K. (2001). Refusing the devil’s bargain: What kind of underdetermination should we take seriously? Philosophy of Science (Proceedings Supplement), 68, S1–S12.
Stanford, P. K. (2006). Exceeding our grasp: Science, history, and the problem of unconceived alternatives. Oxford: Oxford University Press.
Stanford, P. K. (2009). Scientific realism, the atomic theory, and the catch-all hypothesis: Can we test fundamental theories against all serious alternatives. British Journal for the Philosophy of Science, 60, 253–269.
Stanford, P.K. (2013). Underdetermination of scientific theory. In E.N. Zalta (Ed.), Stanford encyclopedia of philosophy (Winter 2013 Edition).
Strevens, M. (2009). Objective evidence and absence: Comment on Sober. Philosophical Studies, 143, 91–100.
Worrall, J. (1989). Structural realism: The best of both worlds. Dialectica, 43, 99–124.
Worrall, J. (1994). How to remain (reasonably) optimistic: Scientific realism and the “Luminiferous Ether”. In R. M. Burian, D. Hull. M. Forbes (Eds.) PSA 1994: Proceedings of the 1994 Biennial Meeting of the Philosophy of Science Association, volume 1, pp 334–342.
Wray, K. B. (2011). Epistemic privilege and the success of science. Nous, 46, 375–385.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Dellsén, F. Realism and the absence of rivals. Synthese 194, 2427–2446 (2017). https://doi.org/10.1007/s11229-016-1059-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11229-016-1059-3