It is commonly assumed that preferences are determinate; that is, that an agent who has a preference knows that she has the preference in question and is disposed to act upon it. This paper argues the dubiousness of that assumption. An account of indeterminate preferences in terms of self-predicting subjective probabilities is given, and a decision rule for choices involving indeterminate preferences is proposed. Wolfgang Spohn’s and Isaac Levi ’s arguments against self-predicting probabilities are also considered, in light of Wlodek (...) Rabinowicz’s recent criticism. (shrink)

This paper presents an axiomatization of the principle of maximizing expected utility that does not rely on the independence axiom or sure-thing principle. Perhaps more importantly the new axiomatization is based on an ex ante approach, instead of the standard ex post approach. An ex post approach utilizes the decision maker's preferences among risky acts for generating a utility and a probability function, whereas in the ex ante approach a set of preferences among potential outcomes are on the input side (...) of the theory and the decision maker's preferences among risky acts on the output side. (shrink)

The main result of this paper is a formal argument for the principle of maximizing expected utility that does not rely on the law of large numbers. Unlike the well-known arguments by Savage and von Neumann & Morgenstern, this argument does not presuppose the sure-thing principle or the independence axiom. The principal idea is to use the concept of transformative decision rules for decomposing the principle of maximizing expected utility into a sequence of normatively reasonable subrules. It is shown that (...) this procedure provides a resolution of Allais's paradox that cannot be obtained by Savage-style or von Neumann & Morgenstern-style arguments. (shrink)

In the literature there are at least two models for probabilistic belief revision: Bayesian updating and imaging [Lewis, D. K. (1973), Counterfactuals, Blackwell, Oxford; Gärdenfors, P. (1988), Knowledge in flux: modeling the dynamics of epistemic states, MIT Press, Cambridge, MA]. In this paper we focus on imaging rules that can be described by the following procedure: (1) Identify every state with some real valued vector of characteristics, and accordingly identify every probabilistic belief with an expected vector of characteristics; (2) For (...) every initial belief and every piece of information, choose the revised belief which is compatible with this information and for which the expected vector of characteristics has minimal Euclidean distance to the expected vector of characteristics of the initial belief. This class of rules thus satisfies an intuitive notion of minimal belief revision. The main result in this paper is to provide an axiomatic characterization of this class of imaging rules. (shrink)

Recent criticisms of traditional understandings of risk, responsibility and the division of labour between science and politics build on the idea of the co-produced character of the natural and social orders, making a case for less ambitious and more inclusive policy processes, where questions of values and goals may be addressed together with questions of facts and means, causal liabilities and principled responsibilities. Within the neo-liberal political economy, however, the contingency of the world is depicted as a source of unprecedented (...) opportunities for human craftsmanship, rather than of possibly unmanageable surprises. Gene technologies offer a vantage point for reflecting on the consequences of the drift from discovery to invention as a master frame in the appraisal of human intermingling with the world. Biotech patenting regulations carve out a sovereign agency which, by crafting nature, also crafts the distinction between the manufactured and non-manufactured world. Difficulties in ascribing responsibility stem as a consequence. It is likely that politics and economy can be democratized and responsibilities rearranged not by ‘democratizing’ knowledge production, but rather the reverse. (shrink)

Zeckhauser’s paradox has puzzled and entertained many rationality enthusiasts for almost half a century. You are forced to play a Russian Roulette with a 6-chamber revolver containing either two bullets, or four bullets. Would you pay more to remove the two bullets in than you would to remove one in? Most would say yes, but rational considerations based on the classical utility theory suggest you should not. We discuss a possible solution within the classical framework, by explicitly stating and accounting (...) for more detailed preferences in terms of fewer bullets and smaller debt. To a large extent, the paradox arises due to a surreptitious trespassing of Savage’s Small-World utilities implied by a limited set of preferences to govern a larger world containing potentially conflicting preferences. To avoid logical issues associated with death in the roulette, we also describe a non-fatal game-show version, where you choose one box out of six that could be either empty or contain prize money. Here, the paradox arises when you pay from the prize money, but not when you pay from your own money. In summary, the paradox provides a useful lesson about the normative role of the utility function as a rational guide for our decisions and preferences. (shrink)

In the real world, there can be constraints on rational decision-making: there can be limitations on what I can know and on what you can know. There can also be constraints on my ability to deliberate or on your ability to deliberate. It is useful to know what the norms of rational deliberation should be in ideal contexts, for fully informed agents, in an ideal world. But it is also useful to know what the norms of rational deliberation should be (...) in the actual world, in non-ideal contexts, for imperfectly informed agents, especially for big, life-changing decisions. That is, we want to know how to deliberate as best we can, given the real-world limitations on what we can know, and given real-world limitations on how we are able to deliberate. In this paper, our concern is with the norms of rational deliberation in certain, important, non-ideal contexts, where the reasoning occurs from the agent’s first person, subjective point of view. The norms governing the process of deliberation for real people in the sorts of non-ideal contexts we’ll consider need to reflect the way that real agents, with an incomplete grasp on the facts and an imperfect ability to deliberate, can be expected to proceed. Our central contention is that framing and exploring the deliberative process from the first person perspective allows us to uncover and explore important, real-world constraints on boundedly rational agents deliberating from the subjective perspective. (shrink)

Each of us has a story which we can think of as a set of beliefs, hopefully consistent. We make our decisions in view of our beliefs which may be probabilistic, in the general case, but simple yes or no as in this paper. Our beliefs are our envelope just as the shell of a tortoise is its envelope. Decision theory—or single agent game theory tells us when to make the best choice in a game of us against nature. But (...) nature has no desire to further or frustrate our efforts. Nature is mysterious but not malign. Things change when there are other agents involved. Then the best thing for us to do will depend on what they do. And they will think the same. And we predict their actions in terms of what we think their beliefs are. But how do we coordinate with others whose beliefs are different? This paper addresses the issue of working together despite different beliefs. (shrink)

Popper’s ‘Situational Analysis’ (SA) constitutes his methodological proposal for the social sciences. We claim that the two hallmarks of SA are: (i) that scientists assume they possess a ‘wider’ view of the problem-situation than actors do, and (ii) use the model as an ideal ‘benchmark’ scenario to identify thedeviationof actors’ actual behaviour from the former. We argue that SA is not a generalization of the neoclassical theory of individual behaviour but captures instead the methodology adopted by modern behavioural economists. Last, (...) we argue that SA highlights a way of acquiring knowledge that has gone unnoticed in the literature. (shrink)

When preferences are such that there is no unique additive prior, the issue of which updating rule to use is of extreme importance. This paper presents an axiomatization of the rule which requires updating of all the priors by Bayes rule. The decision maker has conditional preferences over acts. It is assumed that preferences over acts conditional on event E happening, do not depend on lotteries received on Ec, obey axioms which lead to maxmin expected utility representation with multiple priors, (...) and have common induced preferences over lotteries. The paper shows that when all priors give positive probability to an event E, a certain coherence property between conditional and unconditional preferences is satisfied if and only if the set of subjective probability measures considered by the agent given E is obtained by updating all subjective prior probability measures using Bayes rule. (shrink)

This paper starts out from the proposition that case-based decision theory is an appropriate tool to explain human decision behavior in situations of structural ignorance. Although the developers of CBDT suggest its reality adequacy, CBDT has not yet been tested empirically very often, especially not in repetitive decision situations. Therefore, our main objective is to analyse the decision behavior of subjects in a repeated-choice experiment by comparing the explanation power of CBDT to reinforcement learning and to classical decision criteria under (...) uncertainty namely maximin, maximax, and pessimism-optimism. Our findings substantiate a predominant significantly higher validity of CBDT compared to the classical criteria and to reinforcement learning. For this reason, the experimental results confirm the suggested reality adequacy of CBDT in repetitive decision situations of structural ignorance. (shrink)

We describe a logic of preference in which modal connectives reflect reasons to desire that a sentence be true. Various conditions on models are introduced and analyzed.

The increasing success of the evidence-based policy movement is raising the demand of empirically informed decision making. As arguably any policy decision happens under conditions of uncertainty, following our best available evidence to reduce the uncertainty seems a requirement of good decision making. However, not all the uncertainty faced by decision makers can be resolved by evidence. In this paper, we build on a philosophical analysis of uncertainty to identify the boundaries of scientific advice in policy decision making. We start (...) by introducing a distinction between empirical and non-empirical types of uncertainty, and we explore the role of two non-empirical uncertainties in the context of policy making. We argue that the authority of scientific advisors is limited to empirical uncertainty and cannot extend beyond it. While the appeal of evidence-based policy rests on a view of scientific advice as limited to empirical uncertainty, in practice there is a risk of over reliance on experts beyond the legitimate scope of their authority. We conclude by applying our framework to a real-world case of evidence-based policy, where experts have overstepped their boundaries by ignoring non-empirical types of uncertainty. (shrink)

An evolutionary basis for Bayesian rationality is suggested, by considering how natural selection would operate on an organism’s ‘policy’ for choosing an action depending on an environmental signal. It is shown that the evolutionarily optimal policy, as judged by the criterion of maximal expected reproductive output, is the policy that, for each signal, picks an action that maximizes conditional expected output given that signal. This suggests a possible route by which Bayes-rational creatures might have evolved.

The problem of how to make optimal choices in the face of risk arises in both economics/decision theory and also evolutionary biology; in the former, ‘optimal’ means utility maximizing, while in the latter it means fitness maximizing. This article explores the links, thematic and formal, between the economic and evolutionary theories of optimal choice in risky situations, with particular reference to the relationship between utility and fitness. It is argued that the link is strongest between evolution and ‘nonexpected’ utility theory, (...) rather than traditional expected utility theory. (shrink)

:This paper explores the contrast between mentalistic and behaviouristic interpretations of decision theory. The former regards credences and utilities as psychologically real, while the latter regards them as mere representations of an agent's preferences. Philosophers typically adopt the former interpretation, economists the latter. It is argued that the mentalistic interpretation is preferable if our aim is to use decision theory for descriptive purposes, but if our aim is normative then the behaviouristic interpretation cannot be dispensed with.

It is customary to distinguish experimental from purely observational sciences. The former include physics and molecular biology, the latter astronomy and palaeontology. Experiments involve actively intervening in the course of nature, as opposed to observing events that would have happened anyway. When a molecular biologist inserts viral DNA into a bacterium in his laboratory, this is an experiment; but when an astronomer points his telescope at the heavens, this is an observation. Without the biologist’s handiwork the bacterium would never have (...) contained foreign DNA; but the planets would have continued orbiting the sun whether or not the astronomer had directed his telescope skyward. The observational/experimental distinction would probably be difficult to make precise 1, as the notion of an ‘intervention’ is not easily defined, but it is intuitively fairly clear, and is frequently invoked by scientists and historians of science. Experimentation, or ‘putting questions to nature’, is often cited as a hallmark of the modern scientific method, something that permitted the enormous advances of the last 350 years. And it is sometimes said that the social sciences lag behind the natural because controlled experiments cannot be done so readily in the former. Moreover in certain sciences, e.g. epidemiology, students are explicitly taught that experimental data is preferable to observational data, particularly for doing causal inference. So the distinction between observational and experimental science has quite wide currency, and is often regarded as methodologically significant. Surprisingly, mainstream philosophy of science has had rather little to say about the observational/experimental distinction. 2 For example, discussions of confirmation usually invoke a notion of ‘evidence’, to be contrasted with ‘theory’ or ‘hypothesis’; the aim is to understand how the evidence bears on the hypothesis. But whether this ‘evidence’ comes from observation or experiment generally plays no role in the discussion; this is true …. (shrink)

The history of statistics is filled with many controversies, in which the prime focus has been the difference in the “interpretation of probability” between Fre- quentist and Bayesian theories. Many philosophical arguments have been elabo- rated to examine the problems of both theories based on this dichotomized view of statistics, including the well-known stopping-rule problem and the catch-all hy- pothesis problem. However, there are also several “hybrid” approaches in theory, practice, and philosophical analysis. This poses many fundamental questions. This paper (...) reviews three cases and argues that the interpretation problem of probabil- ity is insufficient to begin a philosophical analysis of the current issues in the field of statistics. A novel viewpoint is proposed to examine the relationship between the stopping-rule problem and the catch-all hypothesis problem. (shrink)

A complete calculus of inductive inference captures the totality of facts about inductive support within some domain of propositions as relations or theorems within the calculus. It is demonstrated that there can be no complete, non-trivial calculus of inductive inference.

It is argued that models of H. Jeffreys' axioms of probability (Jeffreys [1939] 1967) are not monotone even with I. J. Good's proposed modification (Good 1950). Hence the additivity axiom seems essential to a theory of probability as it is with Kolmogorov's system (Kolmogorov 1950).

How much would you pay to play a lottery with an “infinite expected payoff?” In the case of the century old, St. Petersburg Paradox, the answer is that the vast majority of people would only pay a small amount. The authors seek to understand this paradox by providing an explanation consistent with a broad, process‐level model of human decision‐making under risk.

I discuss the trilemma that consists of the following three principles being inconsistent: 1. The Common Principle: if one distribution, A, necessarily brings a higher total sum of personal value that is distributed in a more egalitarian way than another distribution, B, A is more valuable than B. 2. (Weak) ex-ante Pareto: if one uncertain distribution, A, is more valuable than another uncertain distribution, B, for each patient, A is more valuable than B. 3. Pluralism about attitudes to risk (Pluralism): (...) the personal value of a prospect is a weighted sum of the values of the prospect’s outcomes, but the weight each outcome gets might be different from the probability the prospect assigns to the outcome. (shrink)

: In some cases in which rational and moral agents experience moral uncertainty, they are unable to assign exact degrees of moral value—in a non-arbitrary way—to some of the different acts available to them, and so are unable to choose with certainty the best act. This article presents a new justification for the use of lotteries in this kind of situation. It is argued that sometimes the only rational thing for a morally motivated agent to do here is to use (...) a lottery. (shrink)

A triviality result for what Lewis called “the Desire by Necessity Thesis” and Broome : 265–267, 1991) called “the Desire as Expectation Thesis” is presented. The result shows that this thesis and three other reasonable conditions can be jointly satisfied only in trivial cases. Some meta-ethical implications of the result are discussed. The discussion also highlights several issues regarding Lewis ’ original triviality result for “the Desire as Belief Thesis” that have not been properly understood in the literature.

I show that de Finetti’s coherence theorem is equivalent to the Hahn-Banach theorem and discuss some consequences of this result. First, the result unites two aspects of de Finetti’s thought in a nice way: a corollary of the result is that the coherence theorem implies the existence of a fair countable lottery, which de Finetti appealed to in his arguments against countable additivity. Another corollary of the result is the existence of sets that are not Lebesgue measurable. I offer a (...) subjectivist interpretation of this corollary that is concordant with de Finetti’s views. I conclude by pointing out that my result shows that there is a sense in which de Finetti’s theory of subjective probability is necessarily nonconstructive. This raises questions about whether the coherence theorem can underwrite a legitimate theory of rational belief. (shrink)

A standard way to challenge convergence-based accounts of inductive success is to claim that they are too weak to constrain inductive inferences in the short run. We respond to such a challenge by answering some questions raised by Juhl (1994). When it comes to predicting limiting relative frequencies in the framework of Reichenbach, we show that speed-optimal convergence—a long-run success condition—induces dynamic coherence in the short run.

This paper contributes to a recent research program that extends arguments supporting elementary conditionalization to arguments supporting conditionalization with general, measure-theoretic conditional probabilities. I begin by suggesting an amendment to the framework that Rescorla (2018) has used to characterize regular conditional probabilities in terms of avoiding Dutch book. If we wish to model learning scenarios in which an agent gains complete membership knowledge about some subcollection of the events of interest to her, then we should focus on updating policies that (...) are what I shall call proper. I go on to characterize regular conditional probabilities in proper learning scenarios using what van Fraassen (1999) calls The General Reflection Principle. (shrink)

Must probabilities be countably additive? On the one hand, arguably, requiring countable additivity is too restrictive. As de Finetti pointed out, there are situations in which it is reasonable to use merely finitely additive probabilities. On the other hand, countable additivity is fruitful. It can be used to prove deep mathematical theorems that do not follow from finite additivity alone. One of the most philosophically important examples of such a result is the Bayesian convergence to the truth theorem, which says (...) that conditional probabilities converge to 1 for true hypotheses and to 0 for false hypotheses. In view of the long-standing debate about countable additivity, it is natural to ask in what circumstances finitely additive theories deliver the same results as the countably additive theory. This paper addresses that question and initiates a systematic study of convergence to the truth in a finitely additive setting. There is also some discussion of how the formal results can be applied to ongoing debates in epistemology and the philosophy of science. (shrink)

Bayesians since Savage (1972) have appealed to asymptotic results to counter charges of excessive subjectivity. Their claim is that objectionable differences in prior probability judgments will vanish as agents learn from evidence, and individual agents will converge to the truth. Glymour (1980), Earman (1992) and others have voiced the complaint that the theorems used to support these claims tell us, not how probabilities updated on evidence will actually}behave in the limit, but merely how Bayesian agents believe they will behave, suggesting (...) that the theorems are too weak to underwrite notions of scientific objectivity and intersubjective agreement. I investigate, in a very general framework, the conditions under which updated probabilities actually converge to a settled opinion and the conditions under which the updated probabilities of two agents actually converge to the same settled opinion. I call this mode of convergence deterministic, and derive results that extend those found in Huttegger (2015b). The results here lead to a simple characterization of deterministic convergence for Bayesian learners and give rise to an interesting argument for what I call strong regularity, the view that probabilities of non-empty events should be bounded away from zero. (shrink)

Much work in the philosophy of action in the last few decades has focused on the elucidation and justification of a series of purported norms of practical rationality that concern the presence or absence of intention in light of belief, and that demand a kind of structural coherence in the psychology of an agent. Examples of such norms include: Intention Detachment, which proscribes intending to do something in case some condition obtains, believing that such condition obtains, and not intending to (...) do that thing; Intention-Belief Consistency, which proscribes intending to do what you believe you will not do; Intention Consistency, which proscribes intending each of two ends you believe to be inconsistent; and Means-End Coherence, which proscribes intending an end and not intending the means you believe to be implied by your end. In this paper, I present a series of examples that show that these requirements are not genuine requirements of rationality. The reason for this is simple: these requirements concern the presence or absence of intention in light of all-out belief. Rational agents like us, however, do not, and in fact should not, always form or revise their intentions in light of what they all-out believe. When such agents do not form or revise their intentions in light of what they all-out believe, they need not be irrational if they do not conform to these requirements. (shrink)

This chapter contains sections titled: Highlights of Past Literature Current Work Future Work.

Unlike other classical arguments for the existence of God, Pascal’s Wager provides a pragmatic rationale for theistic belief. Its most popular version says that it is rationally mandatory to choose a way of life that seeks to cultivate belief in God because this is the option of maximum expected utility. Despite its initial attractiveness, this long-standing argument has been subject to various criticisms by many philosophers. What is less discussed, however, is the rationality of this choice in situations where the (...) decision-makers are confronted with greater uncertainty. In this paper, I examine the imprecise version of Pascal’s Wager: those scenarios where an agent’s credence that God exists is imprecise or vague rather than precise. After introducing some technical background on imprecise probabilities, I apply five different principles for decision-making to two cases of state uncertainty. In the final part of the paper, I argue that it is not rationally permitted to include zero as the lower probability of God’s existence. Although the conditions for what makes an act uniquely optimal vary significantly across those principles, I also show how the option of wagering for God can defeat any mixed strategy under two distinct interpretations of salvation. (shrink)

Von Mises defined a "physical" probability as a strict limit of the relative frequency of occurrence of an event in repeated trials. As a result of a number of criticisms of von Mises's approach, the more favored approach became the "propensity" interpretation. It is argued here that this interpretation is not compelling and that the only problem in von Mises's approach is the assumption that the relative frequency converges in a strict sense. This problem is then remedied by deducing the (...) axioms of probability theory from the assumption that the relative frequency converges only in the sense of the weak law of large numbers. (shrink)

This article assesses two major conceptual arguments against theories of choice.The first argument concerns the circularity of belief-desire psychology, on which decision theory is based. The second argument concerns the normativity arising from the concept of rationality. Each argument is evaluated against experimental practice in economics and psychology, and it is concluded that both arguments fail to establish their skeptical conclusion that there can be no science of intentional human actions.

We reexamine some of the classic problems connected with the use of cardinal utility functions in decision theory, and discuss Patrick Suppes's contributions to this field in light of a reinterpretation we propose for these problems. We analytically decompose the doctrine of ordinalism, which only accepts ordinal utility functions, and distinguish between several doctrines of cardinalism, depending on what components of ordinalism they specifically reject. We identify Suppes's doctrine with the major deviation from ordinalism that conceives of utility functions as (...) representing preference differences, while being non- etheless empirically related to choices. We highlight the originality, promises and limits of this choice-based cardinalism. (shrink)

This paper aims to open a dialogue between philosophers working in decision theory and operations researchers and engineers working on decision-making under deep uncertainty. Specifically, we assess the recommendation to follow a norm of robust satisficing when making decisions under deep uncertainty in the context of decision analyses that rely on the tools of Robust Decision-Making developed by Robert Lempert and colleagues at RAND. We discuss two challenges for robust satisficing: whether the norm might derive its plausibility from an implicit (...) appeal to probabilistic representations of uncertainty of the kind that deep uncertainty is supposed to preclude; and whether there is adequate justification for adopting a satisficing norm, as opposed to an optimizing norm that is sensitive to considerations of robustness. We discuss decision-theoretic and voting-theoretic motivations for robust satisficing, and use these motivations to select among candidate formulations of the robust satisficing norm. (shrink)

I argue that many of the priority rankings that have been proposed by effective altruists seem to be in tension with apparently reasonable assumptions about the rational pursuit of our aims in the face of uncertainty. The particular issue on which I focus arises from recognition of the overwhelming importance and inscrutability of the indirect effects of our actions, conjoined with the plausibility of a permissive decision principle governing cases of deep uncertainty, known as the maximality rule. I conclude that (...) we lack a compelling decision theory that is consistent with a longtermist perspective and does not downplay the depth of our uncertainty, while also supporting orthodox effective altruist conclusions about cause prioritization. (shrink)

In this paper we explore the absentminded driver problem using two different scenarios. In the first scenario we assume that the driver is capable of reasoning about his degree of absentmindedness before he hits the highway. This leads to a Savage-style model where the states are mutually exclusive and the act-state independence is in place. In the second we employ centred possibilities, by modelling the states (i.e. the events about which the driver is uncertain) as the possible final destinations indexed (...) by a time period. The optimal probability we find for continuing at an exit is different from almost all papers in the literature. In this scenario, act-state independence is still violated, but states are mutually exclusive and the driver arrives at his optimal choice probability via Bayesian updating. We show that our solution is the only one guaranteeing immunity from sure loss via a Dutch strategy, and that – despite initial appearances – it is time consistent. This raises important implications for the rationality of commitment in such scenarios. (shrink)