Abstract
This paper uses the idea of a proxy, which figures in discussions of bounded rationality, to construct an argument for a revisionary conclusion about ideal instrumental rationality. I consider how subjective responses can figure as proxies in heuristics and develop the following argument: (1) Proxies, even if relatively easy to recognize, can sometimes be messy, prompting incomplete or cyclic preferences. (2) From the point of view of ideal instrumental rationality, it is permissible for an agent to be concerned with a proxy rather than with what it is a proxy for. And so, (3) although it is standardly assumed, as part of the prevailing conception of ideal instrumental rationality, that rational preferences are complete and acyclic, neither incomplete preferences nor cyclic preferences can be dismissed as invariably irrational from the point of view of ideal instrumental rationality.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
I borrow the phrase “fast and frugal heuristics” from Gigerenzer, Todd, and the ABC Research Group (1999). Delving into debate regarding the use of heuristics is outside the scope of this paper; but for a sense of the range of approaches, see, in addition to the work of Gigerenzer et al., Simon (1955) and the classic collection Judgment under Uncertainty: Heuristics and Biases, edited by Kahneman, Slovic, and Tversky (1982).
Note that preferences are here understood as subjective comparative appraisals that can sometimes, but not invariably, explain choice (which is also affected by, for example, habits and plans). For some relevant discussion regarding this prominent notion of preference, see Hansson and Grüne-Yanoff (2018). Note that acting counterpreferentially is generally accepted as rational in at least some theoretically possible cases, such as in “EverBetter” cases. EverBetter cases involve a choice situation in which one is faced with an infinite sequence of increasingly better options, none of which will be enjoyed if one does not at some point select an option even though it is dispreferred. The famous EverBetter Wine case can be found in (Pollock 1983).
Note that all this is perfectly compatible with the idea that there can be a great deal of intersubjective agreement about tastiness and comfort since, for example, for someone with a normal palate, lightly salted chips are tastier than sawdust, and, similarly, for someone with a normal physical constitution, walking on carpet is more comfortable than walking on burning coals.
In line with a strong presumption of orderliness, one influential way of proceeding in discussions concerning cyclic preferences is to consider possible mistakes cyclic preferences “might” be based on and, even when the preferences “arise in an apparently reasonable way,” default to the assumption that the agent is making one or more of these mistakes. See, for example, Arntzenius and McCarthy (1997, Sect. 4). There is, however, the worry that this approach makes things “too easy on the theorist” seeking neatness (Quinn 1993).
This parenthetical remark is, of course, compatible with the idea that there is a close connection between an agent’s sensibility and the sorts of claims that she would, if prompted, sincerely avow or disavow; it is just meant to avoid suggesting an oversimplified picture of the connection.
A clear and concise discussion of the small-improvement argument, now famous in the literature on parity and incomparability, can be found in Chang (1997), at section III.7. See, also, and more recently, Chang (2017). For earlier variations on the argument, see, for example, Raz (1986), chapter 13, and de Sousa (1974). See, relatedly, Savage (1972), p. 17. Gustafsson and Espinoza (2010) distinguish between “preferential and axiological versions of the argument” and pose what might seem like a decisive argument against the preferential version; a rebuttal of Gustafsson’s and Espinoza’s argument can be found in Carlson (2011). Both versions of the small-improvement argument remain influential, with variations on each differing in terms of how modest or controversial they are.
See Temkin (1996).
See, for example, Arntzenius and McCarthy (1997, Sect. 4) for a compact and concrete instance of the following familiar line of thought:
-
(1)
An agent with cyclic preferences “could be turned into a money pump” if he follows his preferences.
-
(2)
So, even if his preferences “arise in an apparently reasonable way,” something is “wrong with them.”.
Arntzenius and McCarthy are here just adopting the going interpretation of the “money-pump argument.” According to the going interpretation, the argument speaks against the rationality of cyclic preferences by showing that agents with cyclic preferences can, in following their cyclic preferences, foreseeably be led astray. I say a little more about the money-pump argument and about a revisionary interpretation of the argument in note 11.
-
(1)
See, relatedly, note 2, regarding “EverBetter” cases, which are generally recognized as cases in which it is appropriate for the agent to settle on a dispreferred option even though the agent’s preferences are rational.
For some interesting related discussion, see Fishburn (1991, 117). It is there suggested that, although allowing for certain disorderly preferences, including cyclic preferences, “complicates matters,” since such preferences “can make life more difficult,” this is “not cause enough” to reject them; and it cannot just be assumed that “the relationship between preference and choice” (or between “preferences” and “good choices”) is simple.
Cases in which adherence to one’s preferences is foreseeably self-defeating include, for example, “EverBetter” cases (described in note 2), cases in which an agent allows herself to be “money pumped” (wherein she foreseeably ends up with the same option she started with but with less money), and “self-torturer” type cases (wherein the agent foreseeably ends up with an option he finds terrible even though options he would welcome as good were available). Money-pump cases can be found in Davidson, McKinsey, and Suppes (1955), with some more recent discussion in, for example, Rabinowicz (2000) and Dougherty (2014); and self-torturer type cases can be found in Quinn (1993), with some more recent discussion in, for example, Tenenbaum (2020) and Andreou (2022, 2023). Although the money-pump argument is generally interpreted as supporting the view that cyclic preferences are irrational, it is a controversial argument and, for some who think that cyclic preferences cannot be so easily dismissed, the money-pump argument is more plausibly interpreted as supporting the view that a rational agent does not just naïvely follow her preferences when they are disorderly. See Andreou (2023, especially chapter 1). Relatedly, see Tenenbaum (2020, especially chapter 3).
See especially Andreou (2023, chapter 3 and chapter 6). I there also address the complicated matter of what to say about the presumed acyclicity of “the better” than relation if rational preferences can be cyclic.
References
Andreou, C. (2022). Commitment and resoluteness in rational choice. In Elements in decision theory and philosophy, series ed. Martin Peterson. Cambridge University Press.
Andreou, C. (2023). Choosing well: The good, the bad, and the trivial. Oxford University Press.
Arntzenius, F., & McCarthy, D. (1997). Self torture and group beneficence. Erkenntnis, 47, 129–144.
Carlson, E. (2011). The small-improvement argument rescued. Philosophical Quarterly, 61, 171–174.
Chang, R. (1997). Introduction. In Chang (Ed.), Incommensurability, incomparability, and practical reason (pp. 1–34). Harvard University Press.
Chang, R. (2002). The possibility of parity. Ethics, 112, 659–688.
Chang, R. (2017). Hard choices. Journal of the American Philosophical Association, 3, 1–21.
Davidson, D., McKinsey, J., & Suppes, P. (1955). Outlines of a formal theory of value. Philosophy of Science, 22, 140–160.
de Sousa, R. (1974). The good and the true. Mind, 84, 534–551.
Dougherty, T. (2014). A deluxe money pump. Thought, 3, 21–29.
Fishburn, P. C. (1991). Nontransitive preferences in decision theory. Journal of Risk and Uncertainty, 4, 113–134.
Gigerenzer, G., Todd, P. M., & the ABC Research Group. (1999). Simple heuristics that make us smart. Oxford University Press.
Gustafsson, J., & Espinoza, N. (2010). Conflicting reasons in the small-improvement argument. Philosophical Quarterly, 60, 754–763.
Hansson, S. O., & Grüne-Yanoff, T. (2018). Preferences. In E. N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/archives/sum2018/entries/preferences/
Kahneman, D., Slovic, P., & Tversky, A. (1982). Judgment under uncertainty: Heuristics and biases. Cambridge University Press.
Pollock, J. L. (1983). How do you maximize expectation value? Noûs, 17, 409–421.
Quinn, W. (1993). The puzzle of the self-torturer. In Morality and action. Cambridge University Press.
Rabinowicz, W. (2000). Money pump with foresight. In M. J. Almeida (Ed.), Imperceptible harms and benefits (Library of ethics and applied philosophy (8) (pp. 123–154). Kluwer.
Raz, J. (1986). The morality of freedom. Clarendon Press.
Savage, L. (1972). The foundations of statistics. Dover.
Simon, H. (1955). A behavioral model of rational choice. The Quarterly Journal of Economics, 69, 99–118.
Temkin, L. (1996). A continuum argument for intransitivity. Philosophy & Public Affairs, 25, 175–210.
Tenenbaum, S. (2020). Rational powers in action. Oxford University Press.
Acknowledgements
My thanks to Elijah Millgram, Jonah Schupbach, Mariam Thalos, Frederik Van De Putte, Mike White, the referees, and the work-in-progress group participants in the Department of Philosophy at the University of Utah for their helpful comments on earlier drafts of this paper. I am also grateful to Kenny Easwaran, whose thought-provoking challenges [in conversation] to some claims in my earlier work prompted me to develop the argument in this paper.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The author is a faculty member of the University of Utah and was employed by the University of Utah during the preparation of this article. The author has no relevant financial or non-financial interests to disclose.
Human or animal participants
The article did not involve human or animal participants and so no informed consent information is provided.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Andreou, C. Preferences, Proxies, and Rationality. Erkenn (2024). https://doi.org/10.1007/s10670-023-00766-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10670-023-00766-4