Wlodek Rabinowicz

The Puzzle of the Hats is a betting arrangement which seems to show that a Dutch book can be made against a group of rational players with common priors who act in the common interest and have full trust in the other players’ rationality. But we show that appearances are misleading—no such Dutch book can be made. There are four morals. First, what can be learned from the puzzle is that there is a class of situations in which credences and betting rates diverge. Second, there is an analogy between ways of dealing with situations of this kind and different policies for sequential choice. Third, there is an analogy with strategic voting, showing that the common interest is not always served by expressing how things seem to you in social decision-making. And fourth, our analysis of the Puzzle of the Hats casts light on a recent controversy about the Dutch book argument for the Sleeping Beauty.

The authors of this paper earlier argued that concrete objects, such as things or persons, may have final value (value for their own sake), which is not reducible to the value of states of affairs that concern the object in question.Our arguments have been challenged. This paper is an attempt to respond to some of these challenges, viz. those that concern the reducibility issue. The discussion presupposes a Brentano‐inspired account of value in terms of fitting responses to value bearers. Attention is given to a yet another type of reduction proposal, according to which the ultimate bearers of final value are abstract particulars (so‐called tropes) rather than abstract states or facts. While the proposal is attractive ((fone entertains the existence of tropes), it confronts serious difficulties. To recognise tropes as potential bearers of final value, along with other objects, is one thing; but to reduce the final value of concrete objects to the final value of tropes is another matter.

Suppose one sets up a sequence of less and less valuable objects such that each object in the sequence is only marginally worse than its immediate predecessor. Could one in this way arrive at something that is dramatically inferior to the point of departure? It has been claimed that if there is a radical value difference between the objects at each end of the sequence, then at some point there must be a corresponding radical difference between the adjacent elements. The underlying picture seems to be that a radical gap cannot be scaled by a series of steps, if none of the steps itself is radical. We show that this picture is incorrect on a stronger interpretation of value superiority, but correct on a weaker one. Thus, the conclusion we reach is that, in some sense at least, abrupt breaks in such decreasing sequences cannot be avoided, but that such unavoidable breaks are less drastic than has been suggested. In an appendix written by John Broome and Wlodek Rabinowicz, the distinction between two kinds of value superiority is extended from objects to their attributes.

In Rabinowicz 2008, I considered how value relations can best be analyzed in terms of fitting pro-­‐attitudes. In the formal model presented in that paper fitting pro-­‐attitudes are represented by the class of permissible preference orderings on a domain of items that are being compared. As it turns out, this approach opens up for a multiplicity of different types of value relationships, along with the standard relations of "better", "worse", "equally as good as" and "incomparable in value". Unfortunately, though, the approach is vulnerable to a number of objections. I believe these objections can be avoided if one re-interprets the underlying notion of preference: Instead of treating preference as a 'dyadic' attitude directed towards a pair of items, we can think of it as a difference of degree between 'monadic' attitudes of favouring. Each such monadic attitude has just one item as its object. Given this re-interpretation, permissible preferences can be modelled by the class of permissible assignments of degrees of favouring to items in the domain. From this construction, we can then recover the old modelling in terms of the class of permissible preference orderings, but the previous objections to that model no longer apply. Thus, what we get is the old wine in new and hopefully tighter barrels.

My focus is on pragmatic arguments for various ‘rationality constraints’ on a decision maker’s state of mind: on his beliefs or preferences. An argument of this kind purports to show that a violator of a given constraint can be exposed to a decision problem in which he will act to his guaranteed disadvantage. Dramatically put, he can be exploited by a clever bookie who doesn’t know more than the agent himself. Examples of pragmatic arguments of this kind are synchronic Dutch Books, for the standard probability axioms, diachronic Dutch Books, for the more controversial principles of reflection and conditionalization, and Money Pumps, for the transitivity requirement on preferences. The proposed exploitation set-ups share a common feature. If the violator of a given constraint is logically and mathematically competent, he can be exploited only if he is disunified in his decision-making. Exploitation is possible only if the agent makes decisions on various issues he confronts one by one, rather than on all of them together. Unity in decision making may be quite costly and is often inconvenient, especially when it concerns opportunity packages that are spread over time. On my view, therefore, pragmatic arguments should be seen as delivering conditional conclusions: “If you want to afford being disunified as a decision maker, then you’d better satisfy these constraints.” The arguments of this kind fail to establish the inherent rationality of the constraints under consideration. Levi’s view of the status of pragmatic arguments is opposed to mine. According to him, only synchronic pragmatic arguments are valid. The diachronic ones, he argues, lack any validity at all. This line of reasoning is questioned in my paper.

The Puzzle of the Hats is a betting arrangement which seems to show that a Dutch book can be made against a group of rational players with common priors who act in the common interest and have full trust in the other players’ rationality. But we show that appearances are misleading—no such Dutch book can be made. There are four morals. First, what can be learned from the puzzle is that there is a class of situations in which credences and betting rates diverge. Second, there is an analogy between ways of dealing with situations of this kind and different policies for sequential choice. Third, there is an analogy with strategic voting, showing that the common interest is not always served by expressing how things seem to you in social decision-making. And fourth, our analysis of the Puzzle of the Hats casts light on a recent controversy about the Dutch book argument for the Sleeping Beauty.

in Undetermined On an influential interpretation, the agent's degrees of belief asr identified with her betting rates. However, being placed in a betting situation can itself change one’s degree of belief in the proposition in question. The problem as such isn’t new. Ramsey, for example, was right on to this idea when he wrote: "… the proposal of a bet may inevitably alter [one’s] state of opinion; just as we could not always measure electric intensity by actually introducing a charge and seeing what force it was subject to, because the introduction of the charge would change the distribution to be measured." This ‘interference problem’ for the betting interpretation can best be put in the following way: The bets we are disposed to accept do not manifest our current unconditional degrees of belief in various propositions. Whether a bet on A would be accepted or not does not depend on the agent’s degree of belief in A but rather on the degree of belief she would have if she were confronted with this bet proposal, or – more generally – the degree of belief she would have if she were in a position to bet on A. Assuming the conditionalization model for belief change, this means that whether the agent would be willing to bet depends on her current conditional beliefs concerning A on the hypothetical supposition that she has an opportunity to make this bet. Furthermore, her disposition to bet also depends on the expected effects the act of betting would have on the truth of the proposition to be betted on. Both these phenomena imply that finding oneself in a betting situation might alter one’s expectations in important ways. Consequently, the identification of degrees of belief with betting rates is a mistake. We show that this problem comes up in several different versions, or – equivalently – that different kinds of arguments against the betting interpretation stem from the same source.

The paper presents and discusses the so-called Wrong Kind of Reasons Problem (WKR problem) that arises for the fitting-attitudes analysis of value. This format of analysis is exemplified for example by Scanlon's buck-passing account, on which an object's value consists in the existence of reasons to favour the object- to respond to it in a positive way. The WKR problem can be put as follows: It appears that in some situations we might well have reasons to have pro-attitudes toward objects that are not valuable. Or vice versa: we might have reasons not to have pro-attitudes toward some valuable objects. The paper goes through several attempts to solve (or dissolve) the WKR problem and argues that none of them is fully satisfactory.

Backward induction has been the standard method of solving finite extensive-form games with perfect information, notwithstanding the fact that this procedure leads to counter-intuitive results in various games. However, beginning in the late eighties, the method of backward induction became an object of criticism. It is claimed that the assumptions needed for its defence are quite implausible, if not incoherent. It is therefore natural to ask for the justification of backward induction: Can one show that rational players who know the structure of the game, have trust in each other's practical rationality and reason correctly, will act in accordance in backward induction? Several researchers have tried a justification of this kind, but the argument presented in Robert Aumann's paper from 1995 is perhaps the most well-known and influential attempt to provide such a justification. Clausing provides a sustained discussion of the justification problem for backward induction. It is an excellent work and the criticism I will present below does not detract from this evaluation: the issues discussed by the author are complex and it is difficult to get everything right. Furthermore, I hope that the criticism to be presented may be instructive; Even though it has not been Clausing's intention, his logical reconstruction of Aumann's defence of backward induction allows us to see very clearly what is wrong with that argument. It also provides us with * This paper was presented at workshops in Lund and in Uppsala, in the Fall of 1999. I am indebted to the participants for their useful comments. The work on the paper was supported by a generous research grant from The Bank of Sweden Tercentenary Foundation.

Piccione and Rubinstein (1997) present and analyse the sequential decision problem of an “absentminded driver”. The driver's absentmindedness (imperfect recall) leads him to time-inconsistent strategy evaluations. His original evaluation gets replaced by a new one under impact of the information that the circumstances have changed, notwithstanding the fact that this change in circumstances has been expected by him all along. The time inconsistency in strategy evaluation suggests that such an agent might have reason to renege on his adopted strategy. As we shall see, however, this danger is only apparent. There is no serious problem of dynamic inconsistency in this case. My diagnosis of the case under consideration is in many respects similar to the one provided by Aumann, Hart and Perry (1997), but the analysis leading to this diagnosis is not quite the same.

In the standard money pump, an agent with cyclical preferences can avoid exploitation if he shows foresight and solves his sequential decision problem using backward induction (BI). This way out is foreclosed in a modified money pump, which has been presented in Rabinowicz (2000). There, BI will lead the agent to behave in a self-defeating way. The present paper describes another sequential decision problem of this kind, the Centipede for an Intransitive Preferrer, which in some respects is even more striking than the modified pump. In the new problem, the BI reasoning that implies self-defeating behavior does not rest on the controversial robustness assumption concerning beliefs in one's future rationality. This strengthens the claim that foresight cannot save the intransitive preferrer from a self-defeating course of action.

I begin, in section 1, with a presentation of the Interpersonal Addition Theorem. The theorem, due to John Broome (1991), is a re-formulation of the classical result by Harsanyi (1955). It implies that, given some seemingly mild assumptions, the overall utility of an uncertain prospect can be seen as the sum of its individual utilities. In sections 1 and 2, I discuss the theorem's connection with utilitarianism and in particular consider its implications for the Priority View, according to which benefits to the worse off count for more, in terms of overall utility, than comparable benefits to the better off (cf. Parfit 1995 [1991]). Broome (1991) and Klint Jensen (1996) have argued that, in view of the Interpersonal Addition Theorem, the Priority View should be rejected for measurement-theoretical reasons. Therefore, it cannot be seen as a plausible competitor to utilitarianism (cf. section 1). I will suggest, however, that a proponent of the Priority View would be well-advised, on independent grounds, to reject one of the basic assumptions on which the Addition Theorem is based. I have in mind the so-called Principle of Personal Good for uncertain prospects (cf. sections 4 and 5). If the theorem is disarmed in this way, then, as a side benefit, the Priority View will avoid the afore-mentioned problems with measurement.

In earlier papers (Lindström & Rabinowicz, 1989. 1990), we proposed a generalization of the AGM approach to belief revision. Our proposal was to view belief revision as a relation rather thanas a function on theories (or belief sets). The idea was to allow for there being several equally reasonable revisions of a theory with a given proposition. In the present paper, we show that the relational approach is the natural result of generalizing in a certain way an approach to belief revision due to Adam Grove. In his (1988) paper, Grove presents two closely related modelings of functional belief revision, one in terms of a family of "spheres" around the agent's theory G and the other in terms of an epistemic entrenchment ordering of propositions. The "sphere"-terminology is natural when one looks upon theories and propositions as being represented by sets of possible worlds. Grove's spheres may be thought of as possible "fallback" theories relative to the agent's original theory: theories that he may reach by deleting propositions that are not "sufficiently" entrenched (according to standards of sufficient entrenchment of varying stringency). To put it differently, fallbacks are theories that are closed upwards under entrenchment The entrenchment ordering can be recovered from the family of fallbacks by the definition: A is at least as entrenched as B iff A belongs to every fallback to which B belongs. To revise a theory T with a proposition A, we go to the smallest sphere that contain A-worlds and intersect it with A. The relational notion of belief revision that we are interested in, results from weakening epistemic entrenchment by not assuming it to be connected. I.e., we want to allow that some propositions may be incomparable with respect to epistemic entrenchment. As a result, the family of fallbacks around a given theory will no longer have to be nested. This change opens up the possibility for several different ways of revising a theory with a given proposition.

The paper's focus is on pragmatic arguments for various ‘rationality constraints’ on a decision maker’s state of mind: on his beliefs or preferences. An argument of this kind purports to show that a violator of a given constraint can be exposed to a decision problem in which he will act to his guaranteed disadvantage. Dramatically put, he can be exploited by a clever bookie who doesn’t know more than the agent himself. Examples of pragmatic arguments of this kind are synchronic Dutch Books, for the standard probability axioms, diachronic Dutch Books, for the more controversial principles of reflection and conditionalization, and Money Pumps, for the transitivity requirement on preferences. The proposed exploitation set-ups share a common feature. If the violator of a given constraint is logically and mathematically competent, he can be exploited only if he is disunified in his decision-making. Exploitation is possible only if the agent makes decisions on various issues he confronts one by one, rather than on all of them together. Unity in decision making may be quite costly and is often inconvenient, especially when it concerns opportunity packages that are spread over time. Therefore, pragmatic arguments should be seen as delivering conditional conclusions: “To afford being disunified as a decision maker, you’d better satisfy these constraints.” Arguments of this kind fail to establish the inherent rationality of the constraints under consideration. Levi’s view of the status of pragmatic arguments (cf. Levi 2002) is diametrally opposed. According to him, only synchronic pragmatic arguments are valid (indeed, categorically valid). The diachronic ones, he argues, lack any validity at all. This line of reasoning is questioned in the paper.

The ‘buck-passing’ account equates the value of an object with the existence of reasons to favour it. As we argued in an earlier paper, this analysis faces the ‘wrong kind of reasons’ problem: there may be reasons for pro-attitudes towards worthless objects, in particular if it is the pro-attitudes, rather than their objects, that are valuable. Jonas Olson has recently suggested how to resolve this difficulty: a reason to favour an object is of the right kind only if its formulation does not involve any reference to the attitudes for which it provides a reason. We argue that despite its merits, Olson's solution is unsatisfactory. We go on to suggest that the buck-passing account might be acceptable even if the problem in question turns out to be insoluble

This is a short version of Lindström & Rabinowicz 1991.In earlier papers, we proposed a generalization of the AGM approach to belief revision. The proposal was to view belief revision as a relation rather than as a function on theories (or belief sets). Going relational means that one allows for several equally reasonable revisions of a theory with a given proposition. In the present paper, we show that the relational approach is the natural result of generalizing in a certain way an approach to belief revision due to Adam Grove. In his (1988) paper, Grove presented two closely related modelings of functional belief revision, one in terms of a family of "spheres" around the agent's theory G and the other in terms of an epistemic entrenchment ordering of propositions. The "sphere"-terminology is natural when one looks upon theories and propositions as sets of possible worlds. Grove's spheres may be thought of as possible "fallback" theories relative to the agent's original theory: theories that he may reach by deleting propositions that are not "sufficiently" entrenched (according to standards of sufficient entrenchment of varying stringency). The entrenchment ordering can be recovered from the family of fallbacks by the definition: A is at least as entrenched as B iff A belongs to every fallback to which B belongs. To revise a theory T with a proposition A, we go to the smallest sphere that contain A-worlds and intersect it with A. The relational notion of belief revision that we are interested in, results if one allows that some propositions may be incomparable with respect to epistemic entrenchment. As a result, the family of fallbacks around a given theory will no longer have to be nested. This change opens up the possibility for several different ways of revising a theory with a given proposition.

Abstract:  The paper provides a general account of value relations. It takes its departure in a special type of value relation, parity, which according to Ruth Chang is a form of evaluative comparability that differs from the three standard forms of comparability: betterness, worseness and equal goodness. Recently, Joshua Gert has suggested that the notion of parity can be accounted for if value comparisons are interpreted as normative assessments of preference. While Gert's basic idea is attractive, the way he develops it is flawed: His modeling of values by intervals of permissible preference strengths is inadequate. Instead, I provide an alternative modeling in terms of intersections of rationally permissible preference orderings. This yields a general taxonomy of all binary value relations. The paper concludes with some implications of this approach for rational choice.

It is a popular view thatpractical deliberation excludes foreknowledge of one's choice. Wolfgang Spohn and Isaac Levi have argued that not even a purely probabilistic self-predictionis available to thedeliberator, if one takes subjective probabilities to be conceptually linked to betting rates. It makes no sense to have a betting rate for an option, for one's willingness to bet on the option depends on the net gain from the bet, in combination with the option's antecedent utility, rather than on the offered odds. And even apart from this consideration, assigning probabilities to the options among which one is choosing is futile since such probabilities could be of no possible use in choice. The paper subjects these arguments to critical examination and suggests that, appearances notwithstanding, practical deliberation need not crowd outself-prediction

An agent whose preferences violate the Independence Axiom or for some other reason are not representable by an expected utility function, can avoid 'dynamic inconsistency' either by foresight ('sophisticated choice') or by subsequent adjustment of preferences to the chosen plan of action ('resolute choice'). Contrary to McClennen and Machina, among others, it is argued these two seemingly conflicting approaches to 'dynamic rationality' need not be incompatible. 'Wise choice' reconciles foresight with a possibility of preference adjustment by rejecting the two assumptions that create the conflict: Separability of Preferences in the case of sophisticated choice and Reduction to Normal form in the case of resolute choice..

My focus is on aggregation of individual value rankings of alternatives to a collective value ranking. This is compared with aggregation o individual prefrences to a collective preference. While in an individual preference ranking the alternatives are ordered in accordance with one’s preferences, the order in a value ranking expresses one’s comparative evaluation of the alternatives, from the best to the worst. I suggest that, despite their formal similarity as rankings, this difference in the nature of individual inputs in two aggregation scenarios has important implications: The kind of procedure that looks fine for aggregation of judgments is inappropriate for aggregation of preferences. The procedure I have in mind consists in similarity maximization, or – more precisely – in minimization of the average distance from individual inputs. When applied to judgment aggregation, this procedure can also be approached from the epistemic perspective: the questions are posed concerning its advantages as a truth-tracker. From that perspective, what matters is not only the probability of the outcome of the procedure being true, but also the expected verisimilitude of the outcome: its expected distance from truth.

This paper revisits Richard Hare's classical and much discussed argument for preference utilitarianism, which relies on the conception of moral deliberation as a process of thought experimentation, with concomitant preference change. The paper focuses on an apparent gap in Hare's reasoning, the so-called No-Conflict Problem. A solution to this difficulty which was proposed in is re-examined and shown to lead to a number of difficulties. The paper therefore also considers an alternative idea, due to Daniel Elstein. This new proposal may well turn out to be the best way of filling the gap in Hare's argument. The paper also examines whether the gap is there to begin with: The problem should perhaps be dissolved rather than solved. This suggestion goes back to an idea of Zeno Vendler. Unfortunately, it turns out that Vendler's move does not save Hare from criticism: It does dissolve the No-Conflict Problem, but at the same time it gives rise to another, potentially more serious difficulty.

The Interpersonal Addition Theorem, due to John Broome, states that, given certain seemingly innocuous assumptions, the overall utility of an uncertain prospect can be represented as the sum of its individual utilities. Given ‘Bernoulli's hypothesis’ according to which individual utility coincides with individual welfare, this result appears to be incompatible with the Priority View. On that view, due to Derek Parfit, the benefits to the worse off should count for more, in the overall evaluation, than the comparable benefits to the better off. Pace Broome, the paper argues that prioritarians should meet this challenge not by denying Bernoulli's hypothesis, but by rejecting one of the basic assumptions behind the addition theorem: that a prospect is better overall if it is better for everyone. This conclusion follows if one interprets the priority weights that are imposed by prioritarians as relevant only to moral, but not to prudential, evaluations of prospects

We discuss various possibilities for developing a dynamic doxastic logic (DDL) for introspective agents: agents who have the ability to form higher-order beliefs. Such agents can reflect upon and change their minds about their own beliefs. The project of constructing such a logic, full DDL or DDL unlimited, is ridden with difficulties due to the fact that the agent's own doxastic state now becomes a part of the reality he is trying to explore. When an introspective agent learns more about the world (and himself), the reality he holds beliefs about undergoes a change in its doxastic part, which means that his introspective beliefs have to be adjusted accordingly.

According to the fitting-attitude analysis of value , to be valuable is to be a fitting object of a pro-attitude. In earlier publications, setting off from this format of analysis, I proposed a modelling of value relations which makes room for incommensurability in value. In this paper, I first recapitulate the value modelling and then move on to suggest adopting a structurally similar analysis of probability. Indeed, many probability theorists from Poisson onwards did adopt an analysis of this kind. This move allows to formally model probability and probability relations in essentially the same way as value and value relations. One of the advantages of the model is that we get a new account of Keynesian incommensurable probabilities, which goes beyond Keynes in distinguishing between different types of incommensurability. It also becomes possible to draw a clear distinction between incommensurability and vagueness in probability comparisons