Abstract
Orthodox Bayesianism is a highly idealized theory of how we ought to live our epistemic lives. One of the most widely discussed idealizations is that of logical omniscience: the assumption that an agent’s degrees of belief must be probabilistically coherent to be rational. It is widely agreed that this assumption is problematic if we want to reason about bounded rationality, logical learning, or other aspects of non-ideal epistemic agency. Yet, we still lack a satisfying way to avoid logical omniscience within a Bayesian framework. Some proposals merely replace logical omniscience with a different logical idealization; others sacrifice all traits of logical competence on the altar of logical non-omniscience. We think a better strategy is available: by enriching the Bayesian framework with tools that allow us to capture what agents can and cannot infer given their limited cognitive resources, we can avoid logical omniscience while retaining the idea that rational degrees of belief are in an important way constrained by the laws of probability. In this paper, we offer a formal implementation of this strategy, show how the resulting framework solves the problem of logical omniscience, and compare it to orthodox Bayesianism as we know it.
Similar content being viewed by others
Notes
Both principles follow from the Kolmogorov axioms; see Earman (1992) and Titelbaum (forthcoming) for relevant background.
In a broader perspective, Garber’s approach to logical omniscience forms part of an attempt to solve the so-called ‘problem of old evidence’ introduced by Glymour (1980). Related approaches to this problem can be found in Gaifman (2004), Good (1968) and Jeffrey (1983). See also Fitelson and Hartmann (2015) and Sprenger (2015) for more recent developments in this direction.
See, e.g., Nolan (1997).
Bjerring and Skipper (forthcoming). The step-based model was initially inspired by work in active logic; see Elgot-Drapkin and Perlis (1990) for background.
Slightly abusing notation, we will use ‘L’ both as the name of our object language and as a variable that ranges over all sentences of that language.
A notable exception is Hájek (2003).
Here is a sketch of a proof of the first part of n-conjunction: suppose \(w \in |A \wedge B|^n\), for any \(w \in W\). Assuming that R contains standard introduction and elimination rules for conjunction, \(|A \wedge B|^n \subseteq |A|^{n+1}\) and \(|A \wedge B|^n \subseteq |B|^{n+1}\). Hence, \(w \in |A|^{n+1} \cap |B|^{n+1}\). The other subset relations can be established in similar ways.
References
Baltag, A., & Renne, B. (2016). Dynamic epistemic logic. In E. N. Zalta (Ed.), The Stanford encyclopedia of philosophy. Stanford: Metaphysics Research Lab, Stanford University.
Berto, F., & Jago, M. (2018). Impossible worlds. In E. N. Zalta (Ed.), The Stanford encyclopedia of philosophy. Stanford: Metaphysics Research Lab, Stanford University.
Berto, F., & Jago, M. (2019). Impossible worlds. Oxford: Oxford University Press.
Bjerring, J. C. (2013). Impossible worlds and logical omniscience: An impossibility result. Synthese, 190(13), 2505–2524.
Bjerring, J. C., & Schwarz, W. (2017). Granularity problems. Philosophical Quarterly, 67(266), 22–37.
Bjerring, J. C., & Skipper, M. (forthcoming). A dynamic solution to the problem of logical omniscience. The Journal of Philosophical Logic.
Christensen, D. (2007). Does Murphy’s law apply in epistemology? Self-doubt and rational ideals. Oxford Studies in Epistemology, 2, 3–31.
Cresswell, M. (1973). Logics and languages. London: Methuen.
Earman, J. (1992). Bayes or bust?. Denver: Bradford.
Easwaran, K. (2011). Bayesianism II: Applications and criticisms. Philosophy Compass, 6(5), 321–332.
Eells, E. (1985). Problems of old evidence. Pacific Philosophical Quarterly, 66(3), 283.
Elga, A., & Rayo, A. (ms.). Logical omniscience and decision theory: A no-go result. Unpublished manuscript.
Elgot-Drapkin, J., & Perlis, D. (1990). Reasoning situated in time I: Basic concepts. Journal of Experimental and Theoretical Artificial Intelligence, 2, 75–98.
Fagin, R., et al. (1995). Reasoning about knowledge. Cambridge: MIT Press.
Gaifman, H. (2004). Reasoning with limited resources and assigning probabilities to arithmetical statements. Synthese, 140, 97–119.
Garber, D. (1983). Old evidence and logical omniscience in Bayesian confirmation theory. In J. Earman (Ed.), Testing scientific theories. Minneapolis: University of Minnesota Press.
Glymour, C. (1980). Theory and evidence. Princeton: Princeton University Press.
Good, I. (1968). Corroboration, explanation, evolving probability, simplicity and a sharpened razor. British Journal for the Philosophy of Science, 19(2), 123–143.
Hacking, I. (1967). Slightly more realistic personal probability. Philosophy of Science, 34(4), 311–325.
Hájek, A. (2003). What conditional probability could not be. Synthese, 137(3), 273–323.
Harman, G. (1986). Change in view. Cambridge: MIT Press.
Hartmann, S., & Fitelson, B. (2015). A new Garber-style solution to the problem of old evidence. Philosophy of Science, 82(4), 712–717.
Hintikka, J. (1962). Knowledge and belief. Ithaca, NY: Cornell University Press.
Hintikka, J. (1975). Impossible possible worlds vindicated. Journal of Philosophical Logic, 4, 475–484.
Jago, M. (2013). The problem of rational knowledge. Erkenntnis, 6, 1–18.
Jeffrey, R. (1983). Bayesianism with a human face. In J. Earman (Ed.), Testing scientific theories (pp. 133–156). Minneapolis: University of Minnesota Press.
Lakemeyer, G. (1987). Tractable meta-reasoning in propositional logics of belief. In Tenth international joint conference on artificial intelligence (pp. 402–408).
Levesque, H. (1984). A logic of implicit and explicit belief. In National conference on artificial intelligence (pp. 198–202).
Nolan, D. (1997). Impossible worlds: A modest approach. Notre Dame Journal of Formal Logic, 38(4), 535–572.
Rantala, V. (1982). Impossible worlds semantics and logical omniscience. Acta Philosophica Fennica, 35, 106–15.
Rasmussen, M. S. (2015). Dynamic epistemic logic and logical omniscience. Logic and Logical Philosophy, 24, 377–399.
Smithies, D. (2015). Ideal rationality and logical omniscience. Synthese, 192(9), 2769–2793.
Sprenger, J. (2015). A novel solution to the problem of old evidence. Philosophy of Science, 82(3), 383–401.
Talbott, W. (2016). Bayesian epistemology. In E. Zalta (Ed.), The Stanford encyclopedia of philosophy. Stanford: Metaphysics Research Lab, Stanford University.
Titelbaum, M. (forthcoming). Fundamentals of bayesian epistemology. Oxford: Oxford University Press.
Titelbaum, M. (2015). Rationality’s fixed point (Or. in defense of right reasons). Oxford Studies in Epistemology, 5, 253–94.
van Ditmarsch, H., van der Hoek, W., & Kooi, B. (2008). Dynamic epistemic logic. Berlin: Springer.
von Wright, G. (1951). An essay in modal logic. Amsterdam: NorthHolland Pub. Co.
Acknowledgements
An earlier version of this paper was presented at the “Normative Notions Formalized” Workshop in Munich. Many thanks to the audience on that occasion. Thanks also to two anonymous referees from Erkenntnis for very detailed and helpful comments.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
This appendix contains proofs of three main results of the paper. All definitions can be found in §3. The results are repeated here for convenience.
Theorem 1
(n-preservation) If \(A \vdash _R^n B\), then \(Cr(A) = x \models \langle n \rangle (Cr(B) \ge x)\).
Proof
Suppose that \(A \vdash _R^n B\) and consider any pointed model, (M, w), such that \(M,w \models Cr(A) = x\), where \(M = (W^P,W^I,f,V)\) and \(w \in W^P\). We must show that \(M,w \models \langle n \rangle (Cr(B) \ge x)\). We proceed by defining a suitable n-accessible pointed model from (M, w). Let \(M' = (W'^P,W'^I,f',V')\) be a model such that \(W'^P = W^P\), \(W'^I = W^I\), and \(V' = V\). Since \(A \vdash _R^n B\), we can let \(f'\) be an n-variation of (M, w) for which it holds that \(f'(w) = (c,Pr_c)\), where \(M',v \models B\), for all \(v \in \{v' \in c: M',v' \models A\}\). By the definition of n-accessibility, then, \((M,w) {\mathop {\sim }\limits ^{n}} (M',w)\). Since \(M,w \models Cr(A) = x\), (P4) tells us that \(\sum _Q Pr(Q) = x\), where \(Q = \{v \in S: M,v \models A\}\). Hence, \(\sum _{Q'} Pr_c(Q') \ge x\), where \(Q' = \{v \in c: M',v \models B\}\). By another application of (P4), \(M',w \models Cr(B) \ge x\). So, by (P5), it follows that \(M,w \models \langle n \rangle (Cr(B) \ge x)\). \(\square \)
Theorem 2
(n-certainty) If \(\vdash _R^n A\), then \(\models \langle n \rangle (Cr(A) = 1)\).
Proof
Suppose that \(\vdash _R^n A\) and let (M, w) be any pointed model such that \(M = (W^P,W^I,f,V)\) and \(w \in W^P\). We must show that \(M,w \models \langle n \rangle (Cr(A) =1)\). We proceed by defining a suitable n-accessible pointed model from (M, w). Let \(M' = (W'^P,W'^I,f',V')\) be a model such that \(W'^P = W^P\), \(W'^I = W^I\), and \(V' = V\). Since \(\vdash _R^n A\), we can let \(f'\) be an n-variation of (M, w) such that \(f'(w) = (c,Pr_c)\), where \(M',v \models A\), for all \(v \in c\). Hence, \(\sum _{Q'} Pr_c(Q') = 1\), where \(Q' = \{v \in c: M',v \models A\}\). By (P4), \(M',w' \models Cr(A) = 1\). By the definition of n-accessibility, \((M,w) {\mathop {\sim }\limits ^{n}} (M',w)\). So, by (P5), it follows that \(M,w \models \langle n \rangle (Cr(A) = 1)\). \(\square \)
Theorem 3
(n-conditionality) If \(A \vdash _R^n B\), then \(\models \langle n \rangle (Cr(B \, | \, A) = 1)\).
Proof
Suppose that \(A \vdash _R^n B\) and let (M, w) be any pointed model such that \(M = (W^P,W^I,f,V)\) and \(w \in W^P\). We must show that \(M,w \models \langle n \rangle (Cr(B \,|\, A) = 1)\). Let \(M' = (W'^P,W'^I,f',V')\) be a model such that \(W'^P = W^P\), \(W'^I = W^I\), and \(V' = V\). Since \(A \vdash _R^n B\), we can let \(f'\) be an n-variation of (M, w) such that \(f'(w) = (c,Pr_c)\), where \(M',v \models B\), for all \(v \in \{v' \in c: M',v' \models A\}\). Hence, \(\sum _{Q} Pr_c(Q) = \sum _{Q'} Pr_c(Q')\), where \(Q = \{v \in c: M',v \models A\,\text {and}\,M',v \models B\}\) and \(Q' = \{v' \in c: M',v' \models q\}\). By (P8), \(M',w \models Cr(B \,|\, A) = 1\). By the definition of n-accessibility, \((M,w) {\mathop {\sim }\limits ^{n}} (M',w)\). So, by (P5), it follows that \(M,w \models \langle n \rangle (Cr(B \,|\, A) = 1)\). \(\square \)
Rights and permissions
About this article
Cite this article
Skipper, M., Bjerring, J.C. Bayesianism for Non-ideal Agents. Erkenn 87, 93–115 (2022). https://doi.org/10.1007/s10670-019-00186-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10670-019-00186-3