Abstract
I offer an account of ad hocness that explains why the adoption of an ad hoc auxiliary is accompanied by the disconfirmation of a hypothesis H. H must be conjoined with an auxiliary (or set of auxiliaries) a′, which is improbable antecedently given H, while ~H does not have this disability. This account renders it unnecessary to require, for identifying (bad) ad hocness, that either a′ or H have a posterior probability less than or equal to 0.5; there are also other reasons for abandoning that condition. I distinguish between formal ad hocness, which is bad in the probabilistic sense that it results in disconfirmation of H, and argumentative ad hocness, which actually involves bad reasoning on the part of a subject. The latter is what I call “not counting the cost.” This distinction allows us to see why the 0.5 condition appeared attractive in the first place. The concept of not counting the cost also has implications for other areas of research, including both a Bayesian concept of unfalsifiability and the classic epistemological question of the problem of the external world.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Strevens (2001, p. 535) holds that bad ad hocness occurs when the prior probability of the main hypothesis H is much lower than the prior probability of an auxiliary a, which is later replaced by a′. My own account here makes no reference to a previous auxiliary a, and I use the typographical convention of referring to an auxiliary which may be ad hoc as a′ only because that convention is so widespread in the literature. It seems to me that the common assumption that there is a replaced auxiliary a, which previously had an important role to play in conjunction with H, is frequently artificial and unhelpful. Sometimes (indeed, often) the alleged replaced auxiliary is merely the negation of an auxiliary a′, which is historically postulated later on in response to new evidence. See also footnote 8.
Neither Howson and Urbach nor Horwich explicitly states that, in paradigm cases of (bad) ad hocness, H decreases in probability, though that conclusion is implied by Horwich’s discussion.
An anonymous reviewer has suggested that the Hubbard case is funny because there were other and better explanations Hubbard should have tried instead such as memory loss due to stage fright. However, if the girl was really a “Clear” in Dianetics, as this demonstration was to show, she would not have been susceptible to mundane problems like stage fright, according to the theory of Dianetics.
In the first edition, Howson and Urbach (1989, p. 110) imply that their account is the same as Howson’s.
It is not necessary for there to be any particular connection among the disjuncts if a′ is a disjunction, nor between a′ and H. In fact, the more jumbled and disharmonious Ha′ is, when ~H can easily explain E without resorting to such a messy combination, the more obvious is the reason for the disconfirmation of H.
See Smart (1947, p. 302). Theories other than Airy’s non-Newtonian proposal included the existence of a previously unobserved satellite of Uranus, the impact of a comet on Uranus around 1781, and the existence of a resisting medium. All of these are as compatible with Newtonian theory generally as the hypothesis of a new planet.
It is possible that Airy insisted so vociferously that a correction of Newton had been a reasonable hypothesis partly because he was in a defensive frame of mind. His dilatoriness in following up on John Adams’s suggested position of a new planet had resulted in LeVerrier’s being credited with the discovery of Neptune, and Airy was sharply criticized by his fellow Englishmen for mishandling the whole sequence of events (Smart 1947, p. 307ff).
This criticism is also relevant to Strevens’s account (2001, p. 535), which ties the evaluation of a “desperate rescue” to the prior probability of H and its relation to the prior probability of an auxiliary a. (See footnote 1.) It seems possible that H might happen to have sufficient independent evidence in its favor to have as high a prior as an auxiliary a (not necessarily a sub-hypothesis of H), yet still suffer significant disconfirmation if P(E| ~ H) > > P(E|H).
Not counting the cost does not necessarily result in an inconsistent posterior probability distribution. The ad hoc reasoner may adjust all of his probabilities so that the new distribution is consistent, though it did not arise from conditionalizing on E using the probabilities in his prior distribution. It is not always rationally necessary to change one’s probabilities using conditionalization or even Jeffrey conditioning, but it is necessary to have an evidential reason for not conditionalizing, such as the sudden disappearance of foundational evidence (L. McGrew 2010, pp. 96, 103 n.7). Such a rational motivation for not conditionalizing is lacking in the case of argumentative ad hocness.
See, for example, Timothy McGrew’s argument for realism against Cartesian skepticism based on simplicity considerations. McGrew takes it for granted that the realist must answer an argument that treats deceiverism as empirically equivalent to realism (T. McGrew 1995, p. 135).
A similar question can be raised about the multiverse hypothesis. Apropos of (among other things) the multiverse, John Leslie (1989, p. 194) says that “strong evidence for something” is “whatever causes a puzzlement which the existence of that something would reduce or remove.” But that description could be satisfied by an ad hoc hypothesis. See L. McGrew (2005).
I am indebted to Timothy McGrew for comments on an earlier version of this paper and for help in my research on the perturbations of the orbit of Neptune.
References
Adams, J. C. (1896) The scientific papers of John Couch Adams, Vol. 1, W. G. Adams, ed. Reprint: Hong Kong: Forgotten Books, 2013.
Hempel, C. (1966). Philosophy of natural science. Englewood Cliffs, NJ: Prentice Hall.
Horwich, P. (1982). Probability and evidence. Cambridge: Cambridge University Press.
Howson, C., & Urbach, P. (1989). Scientific reasoning: The Bayesian approach (1st ed.). La Salle, IL: Open Court.
Howson, C., & Urbach, P. (1993). Scientific reasoning: The Bayesian approach (2nd ed.). La Salle, IL: Open Court.
Leslie, J. (1989). Universes. London: Routledge.
McGrew, T. (1995). The foundations of knowledge. Lanham, MD: Littlefield Adams.
McGrew, L. (2005). Likelihoods, multiple universes, and epistemic context. Philosophia Christi, 7, 475–481.
McGrew, L. (2010). Probability kinematics and probability dynamics. Journal of Philosophical Research, 35, 89–104.
McGrew, T., & McGrew, L. (2009). The argument from miracles: A cumulative case for the resurrection of Jesus of Nazareth. In W. L. Craig & J. P. Moreland (Eds.), The Blackwell companion to natural theology (pp. 593–662). Oxford: Wiley-Blackwell.
Salmon, W. (1990). Rationality and objectivity in science or Tom Kuhn meets Tom Bayes. In C. Wade Savage (Ed.), Scientific theories, volume 14, Minnesota Studies in the Philosophy of Science (pp. 175–204). Minneapolis, MN: University of Minnesota Press.
Scott, S. (2007) Invoking the Tooth Fairy twice, or how to identify cases of ad hoc hypothesis acceptance. PhD dissertation, University of North Carolina, Chapel Hill.
Smart, W. M. (1947). John Couch Adams and the discovery of Neptune. Popular Astronomy, 55, 301–311.
Strevens, M. (2001). The Bayesian treatment of auxiliary hypotheses. British Journal for the Philosophy of Science, 52, 515–537.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
McGrew, L. On Not Counting the Cost: Ad Hocness and Disconfirmation. Acta Anal 29, 491–505 (2014). https://doi.org/10.1007/s12136-014-0225-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12136-014-0225-9