Abstract
Most of the literature on belief change presupposes a theory of rationality according to which rational agents are optimizers: they chose the best among the feasible epistemic options available to them. But when epistemic values are indeterminate this view of rationality is not applicable. In this situation it is still rational to maximize, i.e., to deem an option as choosable when it is not known to be worse that any other Sen (1997). In this article we present some basic results about rationalizability in terms of maximization rather than optimization.
In recent work Arl)o-Costa (2006) I propose to maximize the categorical preference that arises when one takes the set of shared pairs from all the permissible value orderings. This categorical preference need not be complete but it can be maximized. One possible shortcoming of this procedure is that in many cases the set of shared agreements might be empty. In this case one might use a different approach in cases of indeterminacy. This approach (first proposed in a different form by Isaac Levi in Levi (1986)) has two parts, or tiers. The idea is to first take as admissible any option that is deemed as maximal by some permissible ordering. This is the first tier of the decision rule. Of course, this can generate an admissible set that is not a singleton. To untie ties, one then uses an auxiliary notion of security (qualitatively represented by an additional security ordering). This is the second tier of the decision rule.
The problems of fully characterizing the two-tier decision rule procedure via conditions on choice functions is open. We survey here some attempts to solve this problem and consider applications in the theory of belief change under indeterminacy. If one limits ones attention to the first-tier decision rule there are results by M. Aizerman and A. Malishevksi characterizing a choice function that corresponds to the decision rule Aizerman and Malishevski (1981). This result is applied to axiomatize a belief revision operator where indeterminacy is resolved via the first-tier decision rule with an empty security order. The techniques used to correlate conditions on choice functions with belief revision axioms extend the usual methods first investigated by Hans Rott in Rott (2001).
We conclude with a discussion of applications in philosophy of science. In partic-ular we consider an approach proposed by Sidney Morgenbesser and Arnold Koslow in an unpublished paper Morgenbesser and Koslow (2008). This paper contains various important insights related to the problem of indeterminacy in the process of theory change and theory choice. Most of these insights are related to applications in the philosophy of science. We compare our proposal to the one offered in Morgenbesser and Koslow (2008) and consider some of the questions that are left open in this manuscript.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
The usual procedure in the belief revision literature is to define a selection function γ as a function that returns a non-empty output when applied to a non-empty remainder set. This makes implicit requirements on the marking-off relation ≤. For example, the relation cannot be irreflexive.
- 2.
- 3.
- 4.
Much of the presentation here follow the one introduced in Arló-Costa and Pedersen (2009b). In both cases the section introduces the reader to material that is well known in the belief revision literature.
- 5.
The “g” in \((\ast 7g)\) is for “Gärdenfors” (Rott 2001, p. 110).
- 6.
For a belief set K and a sentence ϕ, a remainder set \(K\bot\varphi\) is the set of maximal consistent subsets of K that do not imply ϕ. Members of \(K\bot\varphi\) are called remainders. Thus, in the AGM framework, a belief set K is fixed, and for every sentence ϕ such that \(\varphi\notin{\textup{Cn}({\emptyset})}\), \(\gamma(K\bot\varphi)\) selects a set of remainders of \(K\bot\varphi\). The situation in which \(\varphi\in{\textup{Cn}({\emptyset})}\) can be handled as a limiting case at the level of the selection function Alchourrón et al. (1985) or at the level of the revision operator.
- 7.
In Alchourrón et al. (1985, pp. 517–518), a relation ≥ is defined over remainder sets for a fixed belief set K, and (Eq≥) is called the marking off identity:
$$\gamma(K\bot\varphi)=\{B\in K\bot\varphi : B\geq B^{\prime}\textup{for all }B^{\prime}\in K\bot\varphi\}.$$ - 8.
Now let \((X,\mathcal{S})\) be a choice space. We call \(\,\mathbb{K}\,\) finitely additive if it is closed under finite unions; we call \(\,\mathbb{K}\,\) subtractive if it is closed under relative complements; and we say that \(\,\mathbb{K}\,\) is compact if for every \(S\in\mathcal{S}\) and \(I\subseteq\mathcal{S}\), if \(S\subseteq\bigcup_{T\in I}T\), then there is some finite \(I_{0}\subseteq I\) such that \(S\subseteq\bigcup_{T\in I_{0}}T\). If γ is a selection function on \((X,\mathcal{S})\), we call γ finitely additive (subtractive, compact) if \(\,\mathbb{K}\,\) is finitely additive (subtractive, compact).
- 9.
The parenthetical notation used here indicates that a is preferred to c and c to b.
References
Aizerman M, Malishevski A (1981) General theory of best variants choice: Some aspects. In: IEEE Transactions of Automatic Control, vol 26, pp 1030–1040
Alchourrón CE, Gärdenfors P, Makinson D (1985) On the logic of theory change: partial meet contraction and revision functions. Journal of Symbolic Logic 50:510–530
Arló-Costa H (2006) Rationality and value: The epistemological role of interdeterminate and agent-dependent values. Philosophical Studies 128(1):7–48
Arló-Costa H, Pedersen A (2009a) Bounded rationality: Models for some fast and frugal heuristics. In: Proceedings of the Second Indian Conference on Logic and its Relationship with Other Disciplines
Arló-Costa H, Pedersen A (2009b) Social norms, rational choice and belief change. In: Olsson E (ed) Science in Flux, Springer, Berlin
Arrow K (1951) Social choice and individual value. Cowles Foundations and Wiley, New York
Gärdenfors P (1988) Knowledge in flux: Modeling the dynamics of epistemic states. MIT Press, Cambridge
Gärdenfors P (1992) Belief revision: An introduction. In: Gäardenfor P (ed) Belief revision, Cambridge University Press, Cambridge pp 1–28
Gärdenfors P, Rott H (1995) Belief revision. In: Gabbay DM, Hogger CJ, Robinson JA (eds) Handbook of logic in artificial intelligence and logic programming, vol 4, Oxford University Press, Oxford pp 35–132
Hansson SO (1999) A textbook of belief dynamics, applied logic series, vol 11. Kluwer Academic Publishers, Dordrecht
Helzner J (2008) Indeterminacy in additive models of choice. In: Proceedings of Foundations of the Formal Sciences VI: Reasoning about Probabilities and Probabilistic Reasoning
Levi I (1986) Hard choices. Cambridge University Press, Cambridge
Levi I (1991) The fixation of belief and its undoing: Changing beliefs through inquiry. Cambridge University Press, Cambridge
Levi I (2004) Mild contractions: Evaluating loss of information due to loss of belief. Oxford University Press, Oxford
Morgenbesser S, Koslow A (2008) Theories and their worth. Tech. Rep., CUNY-Graduate Center, part I-II
Moulin H (1985) Choice functions over a finite set: A summary. Social Choice and Welfare 2:147–160
Pedersen A (2009) Pseudo-rationalizability over infinite choice spaces. Tech. Rep., Carnegie Mellon University
Poproski R (2008) The rationalizability of two-step choices. Master’s thesis, Carnegie Mellon University [Published on-line in the Journal of Philosophical Logic, 20 August 2010]
Rott H (2001) Change, choice and inference: A study of belief revision and non-monotonic reasoning. Oxford Science Publications, Oxford
Sen A (1971) Choice functions and revealed preference. The Review of Economic Studies 38(3):307–317
Sen A (1977) Social choice theory: A re-examination. Econometrica 45(1):53–89
Sen A (1997) Maximization and the act of choice. Econometrica 65:745–779
Suzumura K (1983) Rational choice, collective decisions, and social welfare. Cambridge University Press, Cambridge
Acknowledgments
I would like to thank specially Arnold Koslow for sharing with me an unpublished version of a paper he wrote in collaboration with Sidney Morgenbesser (discussed in last section of this chapter). I understand that this paper is currently under submission. So, perhaps the published version will be slightly different from the manuscript I read. In any case, I think that most of the discussed aspects of the original version of the paper are quite important and deserve a philosophical discussion.
I would like to thank also Arnie Koslow for an enlightening philosophical discussion of the contents of the unpublished manuscript. Finally I would like to thank Paul Pedersen who read the entire manuscript and made useful comments as well as an anonymous referee.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2011 Springer Science+Business Media B.V.
About this chapter
Cite this chapter
Arló-Costa, H. (2011). Indeterminacy and Belief Change. In: Girard, P., Roy, O., Marion, M. (eds) Dynamic Formal Epistemology. Synthese Library, vol 351. Springer, Dordrecht. https://doi.org/10.1007/978-94-007-0074-1_10
Download citation
DOI: https://doi.org/10.1007/978-94-007-0074-1_10
Published:
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-007-0073-4
Online ISBN: 978-94-007-0074-1
eBook Packages: Humanities, Social Sciences and LawPhilosophy and Religion (R0)