Consider the regularity thesis that each possible event has non-zero probability. Hájek challenges this in two ways: there can be nonmeasurable events that have no probability at all and on a large enough sample space, some probabilities will have to be zero. But arguments for the existence of nonmeasurable events depend on the axiom of choice. We shall show that the existence of anything like regular probabilities is by itself enough to imply a weak version of (...) AC sufficient to prove the Banach–Tarski Paradox on the decomposition of a ball into two equally sized balls, and hence to show the existence of nonmeasurable events. This provides a powerful argument against unrestricted orthodox Bayesianism that works even without AC. A corollary of our formal result is that if every partial order extends to a total preorder while maintaining strict comparisons, then the Banach–Tarski Paradox holds. This yields an argument that incommensurability cannot be avoided in ambitious versions of decision theory. (shrink)

Nebel argues for the Repugnant Conclusion via the “Intrapersonal Repugnant Conclusion,” on which certainty of a mediocre life is better for individuals than a sufficiently small chance of an excellent life. In this article, I deny that acceptance of the Intrapersonal Repugnant Conclusion leads us to the Repugnant Conclusion. I point out that on many views which avoid the Repugnant Conclusion we should discount very small probabilities down to zero. If we do, then Nebel’s crucial premise of Ex Ante (...) Pareto fails, because discounting at the individual level can fail to match up with discounting at the population level. (shrink)

Cumulative Prospect Theory (CPT) does not explain the St. Petersburg Paradox. We show that the solutions related to probability weighting proposed to solve this paradox, (Blavatskyy, Management Science 51:677–678, 2005; Rieger and Wang, Economic Theory 28:665–679, 2006) have to cope with limitations. In that framework, CPT fails to accommodate both gambling and insurance behavior. We suggest replacing the weighting functions generally proposed in the literature by another specification which respects the following properties: (1) to solve the (...) class='Hi'>paradox, the slope at zero has to be finite. (2) to account for the fourfold pattern of risk attitudes, the probability weighting has to be strong enough. (shrink)

The topic of this thesis is how we should treat tiny probabilities of vast value. This thesis consists of six independent papers. Chapter 1 discusses the idea that utilities are bounded. It shows that bounded decision theories prescribe prospects that are better for no one and worse for some if combined with an additive axiology. Chapter 2, in turn, points out that standard axiomatizations of Expected Utility Theory violate dominance in cases that involve possible states of zero probability. (...) Chapters 3–6 discuss the idea that we should ignore tiny probabilities in practical decision-making. Chapter 3 argues that discounting small probabilities solves the ‘Intrapersonal Addition Paradox’ and thus helps avoid the Repugnant Conclusion. Chapter 4 explores what the most plausible version of this view might look like and what problems the different versions have. Chapter 5 focuses on one type of problem, namely, money pumps. The Independence Money Pump, in particular, presents a difficult challenge for those who wish to discount small probabilities. Finally, Chapter 6 discusses the implications of discounting small probabilities for the value of the far future. (shrink)

This article explores some open questions related to the problem of verification of theories in the context of empirical sciences by contrasting three epistemological frameworks. Each of these epistemological frameworks is based on a corresponding central metaphor, namely: (a) Neo-empiricism and the gambling metaphor; (b) Popperian falsificationism and the scientific tribunal metaphor; (c) Cognitive constructivism and the object as eigen-solution metaphor. Each of one of these epistemological frameworks has also historically co-evolved with a certain statistical theory and method for testing (...) scientific hypotheses, respectively: (a) Decision theoretic Bayesian statistics and Bayes factors; (b) Frequentist statistics and p-values; (c) Constructive Bayesian statistics and e-values. This article examines with special care the Zero Probability Paradox (ZPP), related to the verification of sharp or precise hypotheses. Finally, this article makes some remarks on Lakatos’ view of mathematics as a quasi-empirical science. (shrink)

Hardy’s non-locality paradox is a proof without inequalities showing that certain non-local correlations violate local realism. It is ‘possibilistic’ in the sense that one only distinguishes between possible outcomes (positive probability) and impossible outcomes (zero probability). Here we show that Hardy’s paradox is quite universal: in any (2,2,l) or (2,k,2) Bell scenario, the occurrence of Hardy’s paradox is a necessary and sufficient condition for possibilistic non-locality. In particular, it subsumes all ladder paradoxes. This universality (...) of Hardy’s paradox is not true more generally: we find a new ‘proof without inequalities’ in the (2,3,3) scenario that can witness non-locality even for correlations that do not display the Hardy paradox. We discuss the ramifications of our results for the computational complexity of recognising possibilistic non-locality. (shrink)

The Borel–Kolmogorov Paradox is typically taken to highlight a tension between our intuition that certain conditional probabilities with respect to probability zero conditioning events are well defined and the mathematical definition of conditional probability by Bayes’ formula, which loses its meaning when the conditioning event has probability zero. We argue in this paper that the theory of conditional expectations is the proper mathematical device to conditionalize and that this theory allows conditionalization with respect to (...) probability zero events. The conditional probabilities on probability zero events in the Borel–Kolmogorov Paradox also can be calculated using conditional expectations. The alleged clash arising from the fact that one obtains different values for the conditional probabilities on probability zero events depending on what conditional expectation one uses to calculate them is resolved by showing that the different conditional probabilities obtained using different conditional expectations cannot be interpreted as calculating in different parametrizations of the conditional probabilities of the same event with respect to the same conditioning conditions. We conclude that there is no clash between the correct intuition about what the conditional probabilities with respect to probability zero events are and the technically proper concept of conditionalization via conditional expectations—the Borel–Kolmogorov Paradox is just a pseudo-paradox. (shrink)

We provide a solution to the well-known “Shooting-Room” paradox, developed by John Leslie in connection with his Doomsday Argument. In the “Shooting-Room” paradox, the death of an individual is contingent upon an event that has a 1/36 chance of occurring, yet the relative frequency of death in the relevant population is 0.9. There are two intuitively plausible arguments, one concluding that the appropriate subjective probability of death is 1/36, the other that this probability is 0.9. How (...) are these two values to be reconciled? We show that only the first argument is valid for a standard, countably additive probability distribution. However, both lines of reasoning are legitimate if probabilities are non-standard. The subjective probability of death rises from 1/36 to 0.9 by conditionalizing on an event that is not measurable, or whose probability is zero. Thus we can sometimes meaningfully ascribe conditional probabilities even when the event conditionalized upon is not of positive finite (or even infinitesimal) measure. (shrink)

This article discusses the classical problem of zero probability of universal generalizations in Rudolf Carnap’s inductive logic. A correction rule for updating the inductive method on the basis of evidence will be presented. It will be shown that this rule has the effect that infinite streams of uniform evidence assume a non-zero limit probability. Since Carnap’s inductive logic is based on finite domains of individuals, the probability of the corresponding universal quantification changes accordingly. This implies (...) that universal generalizations can receive positive prior and posterior probabilities, even for (countably) infinite domains. (shrink)

Resolving Bertrand’s probability paradox.

In this note, it is shown that in a Bayesian model with unawareness of impossible, or vanishingly improbable, events, awareness can only change after such an improbable event has actually been observed.

Dorothy Edgington’s work has been at the centre of a range of ongoing debates in philosophical logic, philosophy of mind and language, metaphysics, and epistemology. This work has focused, although by no means exclusively, on the overlapping areas of conditionals, probability, and paradox. In what follows, I briefly sketch some themes from these three areas relevant to Dorothy’s work, highlighting how some of Dorothy’s work and some of the contributions of this volume fit in to these debates.

A festschrift for Dorothy Edgington, containing contributions from Cleo Condoravdi, Dorothy Edgington, Kit Fine, Alan Hájek, John Hawthorne, Sabine Iatridou, Nick Jones, Rosanna Keefe, Angelika Kratzer, David Over, Daniel Rothschild, Robert Stalnaker, Scott Sturgeon, and Timothy Williamson.

Popper functions allow one to take conditional probabilities as primitive instead of deriving them from unconditional probabilities via the ratio formula P=P/P. A major advantage of this approach is it allows one to condition on events of zero probability. I will show that under plausible symmetry conditions, Popper functions often fail to do what they were supposed to do. For instance, suppose we want to define the Popper function for an isometrically invariant case in two dimensions and hence (...) require the Popper function to be rotationally invariant and defined on pairs of sets from some algebra that contains at least all countable subsets. Then it turns out that the Popper function trivializes for all finite sets: P=1 for all A ) if B is finite. Likewise, Popper functions invariant under all sequence reflections can’t be defined in a way that models a bidirectionally infinite sequence of independent coin tosses. (shrink)

Classic work by one of the most brilliant figures in post-war analytic philosophy.

We begin by showing that every theory of the value of uncertain prospects must have one of three unpalatable properties. _Reckless_ theories recommend giving up a sure thing, no matter how good, for an arbitrarily tiny chance of enormous gain; _timid_ theories permit passing up an arbitrarily large potential gain to prevent a tiny increase in risk; _non-transitive_ theories deny the principle that, if A is better than B and B is better than C, then A must be better than (...) C. Having set up this trilemma, we study its horns. Non-transitivity has been much discussed; we focus on drawing out the costs and benefits of recklessness and timidity when it comes to axiology, decision theory, and normative uncertainty. (shrink)

The so-called Preface Paradox seems to show that one can rationally believe two logically incompatible propositions. We address this puzzle, relying on the notions of truthlikeness and approximate truth as studied within the post-Popperian research programme on verisimilitude. In particular, we show that adequately combining probability, approximate truth, and truthlikeness leads to an explanation of how rational belief is possible in the face of the Preface Paradox. We argue that our account is superior to other solutions of (...) the paradox, including a recent one advanced by Hannes Leitgeb (Analysis 74.1). (shrink)

This paper aims to investigate the so-called paradox of deterministic probabilities: in a deterministic world, all probabilities should be subjective; however, they also seem to play important explanatory and predictive roles which suggest they are objective. The problem is then to understand what these deterministic probabilities are. Recent proposed solutions of this paradox are the Mentaculus vision, the range account of probability, and a version of frequentism based on typicality. All these approaches aim at defining deterministic objective (...) probabilities as to make them compatible with determinism. In this paper I argue that one can think of the equivalent of subjective and objective deterministic probabilities in terms of typicality. Also, I show that only what I identify with objective probabilities play the necessary explanatory and predictive roles, while the subjective components essentially guide beliefs. In this way, the paradox is solved. (shrink)

We describe two anomalies in Bohm’s quantum theory that shed light on the notion of probability zero in quantum mechanics. In one anomaly the preferred reference frame may be discovered.

Exhumation and study of the 1945 paradox of confirmation brings out the defect of its formulation. In the context of quantifier conditional-probability logic it is shown that a repair can be accomplished if the truth-functional conditional used in the statement of the paradox is replaced with a connective that is appropriate to the probabilistic context. Description of the quantifier probability logic involved in the resolution of the paradox is presented in stages. Careful distinction is maintained (...) between a formal logic language and its semantics, as the same language may be outfitted with different semantics. An acquaintance with sections 1? 5 of Hailperin (2006) covering the sentential aspects of probability logic is assumed as background information for quantifier probability logic. (shrink)

We call something a paradox if it strikes us as peculiar in a certain way, if it strikes us as something that is not simply nonsense, and yet it poses some difficulty in seeing how it could make sense. When we examine paradoxes more closely, we find that for some the peculiarity is relieved and for others it intensifies. Some are peculiar because they jar with how we expect things to go, but the jarring is to do with imprecision (...) and misunderstandings in our thought, failures to appreciate the breadth of possibility consistent with our beliefs. Other paradoxes, however, pose deep problems. Closer examination does not explain them away. Instead, they challenge the coherence of certain conceptual resources and hence challenge the significance of beliefs which deploy those resources. I shall call the former kind weak paradoxes and the latter, strong paradoxes. Whether a particular paradox is weak or strong is sometimes a matter of controversy—sometimes it has been realised that what was thought strong is in fact weak, and vice versa,— but the distinction between the two kinds is generally thought to be worth drawing. In this Cchapter, I shall cover both weak and strong probabilistic paradoxes. (shrink)

Classic work by one of the most brilliant figures in post-war analytic philosophy.

For a finite universe of discourse, if Φ → and ~(Ψ → Φ) , then P(Ψ) > P(Φ), i.e., there is always a loss of information, there is an increase in probability, in a non reversible implication. But consider the two propositions, "All ravens are black", (i.e., "(x)(Rx ⊃ Bx)"), and "Some ravens are black" (i.e., "(∃x)(Rx & Bx)"). In a world of one individual, called "a", these two propositions are equivalent to "~Ra ∨ Ba" and "Ra & Ba" (...) respectively. However, (Ra & Ba) → (~Ra ∨ Ba) and ~[(~Ra ∨ Ba) → (Ra & Ba)]. Consequently, in a world of one individual it is more probable that all ravens are black than that some ravens are black! (shrink)

This is a collection of twenty-seven essays written by its author between 1962 and 1989 on topics in the history of science, the philosophy of science, and "the relevance of scientific practice to other parts of philosophy and culture". Twenty-one have been previously published, the remainder hitherto aired only as public presentations. The papers are gathered under six section-headings, including "Explanation," "Hume's Problem," "Logic and Causality," "Machines and Practices," "Scientific Knowledge--Its Scope and Limits," and "Science and Subjectivity"; yet their actual (...) range is both more diverse and less in keeping with prevailing currents in the philosophy of science than these tidily familiar rubrics suggest. "Explanation," for example, flirts with mainstream subject-matter in a much deserved encomium to Hempel's contributions in philosophy of science that reaches beyond the titular subject to the whole of Hempel's work. The section also embraces a discussion of General Systems Theory that distinguishes a doctrinaire, a prioristic approach to the study of science from a pragmatic-pluralistic version, urging the ills of the former and the virtues of the latter. It culminates in two presentations of the view that the theory of evolution rests on "circumstantial evidence" only as it involves the "logical fallacy" of "the inference of causes from effects" : the one is a historical piece on the little-known nineteenth-century naturalist, Phillip Henry Gosse, highlighting his creationistic "Omphalos" theory of the origins of biological species and the fossil record; the other is a short treatment of some conceptual issues involved in the creationist controversy. If the remaining five sections are not so heterogeneous in the scatter of subjects on which their articles touch, they are for the most part more idiosyncratic in the issues they address, the questions they raise, and the answers they suggest. Typical are "On Being in the Same Place at the Same Time," an argument for the ultimacy of subjectivity on the basis of relativity theory, and "Truth and Presence," explicating the attempt of Gaston Bachelard to blend science and poetry as complementary aspects of the human imagination. Sometimes they are also eyebrow-raising. "The Paradox of Induction and the Inductive Wager," would "show the impossibility of... logical proof [of the principle of induction]... by locating it among the paradoxes". Caw's case for paradox is simply that if inductive inference is legitimate then our past inability to justify such inference warrants the inductive conclusion that we shall never be able to justify it. But even if nothing else were suspect in this argument, its force is presumably at most probabilistic, making it hard to see how it could establish the "impossibility" of anything. Another paper, "Three Logics, or the Possibility of the Improbable," concerns "the logic of incredible events", touting the merits of so construing the natural world that events with zero probability occasionally occur and initiate spontaneous causal chains. Yet it is thoroughly unclear what interpretation of probability theory could both license identifying uncaused with impossible events and yet allow that such events may occur. Still another, "Rethinking Intentionality," includes the puzzling contention that "there can't be a science of intention because science presupposes intention". Unlike these last examples, most of the essays in the collection are provocative rather than upsetting, bearing the earmark of the independent, iconoclastic, speculative mind from which they spring, a former physicist not aligned with any of the current schools in philosophy of science, and more influenced by the classic giants of the Western philosophical tradition and by Hegel, Husserl, Heidegger, and Sartre than by empiricist philosophers of science or the recent analytic tradition. Thus, it is no surprise that the issues Caws addresses are nonstandard, and that his manner of addressing them embodies a looser, more broad-stroked, more humanistic, and more metaphysically self-conscious approach than is common in recent mainstream philosophy of science. This orientation doubtless accounts, too, for the somewhat subjective-idealistic stance, the view that one cannot definitively progress beyond one's own sensations and ideas, argued in "Yorick's World or the Universe in a Single Skull," and manifest in the tone, substance, and central thrust of a number of the essays in the collection. It is thus appropriate that Caws should have borrowed from this title to name the anthology itself.-Robert B. Barrett, Washington University in St. Louis. (shrink)

We provide a 'verisimilitudinarian' analysis of the well-known Linda paradox or conjunction fallacy, i.e., the fact that most people judge the probability of the conjunctive statement "Linda is a bank teller and is active in the feminist movement" (B & F) as more probable than the isolated statement "Linda is a bank teller" (B), contrary to an uncontroversial principle of probability theory. The basic idea is that experimental participants may judge B & F a better hypothesis about (...) Linda as compared to B because they evaluate B & F as more verisimilar than B. In fact, the hypothesis "feminist bank teller", while less likely to be true than "bank teller", may well be a better approximation to the truth about Linda. (shrink)

Many epistemologists have responded to the lottery paradox by proposing formal rules according to which high probability defeasibly warrants acceptance. Douven and Williamson present an ingenious argument purporting to show that such rules invariably trivialise, in that they reduce to the claim that a probability of 1 warrants acceptance. Douven and Williamson’s argument does, however, rest upon significant assumptions – amongst them a relatively strong structural assumption to the effect that the underlying probability space is both (...) finite and uniform. In this paper, I will show that something very like Douven and Williamson’s argument can in fact survive with much weaker structural assumptions – and, in particular, can apply to infinite probability spaces. (shrink)