This book explores a question central to philosophy--namely, what does it take for a belief to be justified or rational? According to a widespread view, whether one has justification for believing a proposition is determined by how probable that proposition is, given one's evidence. In this book this view is rejected and replaced with another: in order for one to have justification for believing a proposition, one's evidence must normically support it--roughly, one's evidence must make the falsity of that proposition (...) abnormal in the sense of calling for special, independent explanation. This conception of justification bears upon a range of topics in epistemology and beyond. Ultimately, this way of looking at justification guides us to a new, unfamiliar picture of how we should respond to our evidence and manage our own fallibility. This picture is developed here. (shrink)

This paper develops an information-sensitive theory of the semantics and probability of conditionals and statements involving epistemic modals. The theory validates a number of principles linking probability and modality, including the principle that the probability of a conditional If A, then C equals the probability of C, updated with A. The theory avoids so-called triviality results, which are standardly taken to show that principles of this sort cannot be validated. To achieve this, we deny that rational (...) agents update their credences via conditionalization. We offer a new rule of update, Hyperconditionalization, which agrees with Conditionalization whenever nonmodal statements are at stake but differs for modal and conditional sentences. (shrink)

The problem of probabilities of conditionals is one of the long-standing puzzles in philosophy of language. We defend and update Adams' solution to the puzzle: the probability of an epistemic conditional is not the probability of a proposition, but a probability under a supposition. Close inspection of how a triviality result unfolds in a concrete scenario does not provide counterexamples to the view that probabilities of conditionals are conditional probabilities: instead, it supports the conclusion that probabilities of (...) conditionals violate standard probability theory. This does not call into question probability theory per se; rather, it calls for a more careful understanding of its role: probability theory is a theory of probabilities of propositions; but as conditionals do not express propositions, their probabilities are not subject to the standard laws. We argue that both conditional probabilities and probabilities of conditionals are best understood in terms of the dynamics of supposing, modeled as a restriction operation on a probability space. This version of the suppositionalist view allows us to connect Adams' Thesis to the widely held restrictor view of the semantics of conditionals. We address two common objections to Adams' view: that the relevant probabilities are ‘probabilities only in name’, and that giving up conditional propositions puts us at a disadvantage when it comes to interpreting compounds. Finally, we argue that some putative counterexamples to Adams' Thesis can be diagnosed as fallacies of probabilistic reasoning: they arise from applying to conditionals laws of standard probability theory which are invalid for them. (shrink)

I argue that when we use ‘probability’ language in epistemic contexts—e.g., when we ask how probable some hypothesis is, given the evidence available to us—we are talking about degrees of support, rather than degrees of belief. The epistemic probability of A given B is the mind-independent degree to which B supports A, not the degree to which someone with B as their evidence believes A, or the degree to which someone would or should believe A if they had (...) B as their evidence. My central argument is that the degree-of-support interpretation lets us better model good reasoning in certain cases involving old evidence. Degree-of-belief interpretations make the wrong predictions not only about whether old evidence confirms new hypotheses, but about the values of the probabilities that enter into Bayes’ Theorem when we calculate the probability of hypotheses conditional on old evidence and new background information. (shrink)

In response to cases that involve tiny probabilities of huge payoffs, some argue that we ought to discount small probabilities down to zero. However, this paper shows that doing so violates Independence and Continuity, and as a result of these violations, those who discount small probabilities can be exploited by money pumps. Various possible ways of avoiding exploitation will be discussed. This paper concludes that the money pump for Independence undermines the plausibility of discounting small probabilities. Much of the discussion (...) on Independence generalizes to other views that also violate Independence. (shrink)

Within the τ-Logic program, this paper reconstructs scalar probability as a licensed projection/readout from a declared normalized τ phase regime. It uses the Layer-0 τ-identity reference and the zero/projection-nullity analysis as public continuity references, without treating either work as a hidden premise for the probability theorems. Its formal core is conservative: once τ-identity is represented in a declared native normalized compact phase domain, S¹ ≅ U(1) supplies normalized Haar phase measure, measurable τ-readout maps induce pushforward laws, and (...) class='Hi'>probability laws on standard Borel spaces are representable as such pushforwards. The paper preserves Kolmogorov probability as the ordinary measure-theoretic readout layer; it does not claim that named distributions are uniquely forced from bare τ. Instead, τ-Probability supplies structural provenance only under declared phase-completion, compact phase-domain, Haar-measure, measurable-readout, and local-licensing assumptions, so probability values remain late scalar readouts whose licensing structure includes a source space, σ-algebra, measure, readout/projection map, and local constraint regime. Universal representability is global; distributional necessity is regime-specific. Probability-zero is treated as measure-nullity under a declared regime, not as impossibility, event absence, or ontological nullity. The Borel–Kolmogorov boundary confirms this discipline: E = 0, C = 0, and E = C are phase-coordinate null-readout conditions, not intrinsic conditional events or carrier-nullities, and exact conditioning on them is licensed only by a declared conditioning/readout regime. (shrink)

This chapter will review selected aspects of the terrain of discussions about probabilities in statistical mechanics (with no pretensions to exhaustiveness, though the major issues will be touched upon), and will argue for a number of claims. None of the claims to be defended is entirely original, but all deserve emphasis. The first, and least controversial, is that probabilistic notions are needed to make sense of statistical mechanics. The reason for this is the same reason that convinced Maxwell, Gibbs, and (...) Boltzmann that probabilities would be needed, namely, that the second law of thermodynamics, which in its original formulation says that certain processes are impossible, must, on the kinetic theory, be replaced by a weaker formulation according to which what the original version deems impossible is merely improbable. Second is that we ought not take the standard measures invoked in equilibrium statistical mechanics as giving, in any sense, the correct probabilities about microstates of the system. We can settle for a much weaker claim: that the probabilities for outcomes of experiments yielded by the standard distributions are effectively the same as those yielded by any distribution that we should take as a representing probabilities over microstates. Lastly, (and most controversially): in asking about the status of probabilities in statistical mechanics, the familiar dichotomy between epistemic probabilities (credences, or degrees of belief) and ontic (physical) probabilities is insufficient; the concept of probability that is best suited to the needs of statistical mechanics is one that combines epistemic and physical considerations. (shrink)

Probabilism—the doctrine that ideally rational degrees of belief conform to the calculus of probabilities—is habitually defended by citing Dutch-book and accuracy-dominance arguments. Yet these arguments rely on premises that are not plausibly necessary truths. I argue that the premises in question can be understood as fixing a convention. Relaxing these premises leads to alternative calculi which—under permissive though not universal assumptions—are intertranslatable with the probability calculus. This applies to Dutch-book and accuracy-dominance arguments alike; a mathematical correspondence is shown to (...) hold between them. (shrink)

Many philosophers accept that, as well as having a right that others not harm us, we also have a right that others not subject us to a risk of harm. And yet, when we attempt to spell out precisely what this ‘right against risk imposition’ involves, we encounter a series of notorious puzzles. Existing attempts to deal with these puzzles have tended to focus on the nature of rights – but I propose an approach that focusses instead on the nature (...) of risk. The key move is to distinguish two different ways in which to conceptualise the risk that a given action presents – one of which is linked to the notion of probability and the other to the notion of normalcy. (shrink)

Once upon a time, some thought that indicative conditionals could be effectively analyzed as material conditionals. Later on, an alternative theoretical construct has prevailed and received wide acceptance, namely, the conditional probability of the consequent given the antecedent. Partly following critical remarks recently ap- peared in the literature, we suggest that evidential support—rather than conditional probability alone—is key to understand indicative conditionals. There have been motivated concerns that a theory of evidential conditionals (unlike their more tra- ditional counterparts) (...) cannot generate a sufficiently interesting logical system. Here, we will describe results dispelling these worries. Happily, and perhaps surprisingly, appropriate technical variations of Ernst Adams’s classical approach allow for the construction of a logic of evidential conditionals with distinctive fea- tures, which is also well-behaved and reasonably strong. (shrink)

It is well known that classical, aka ‘sharp’, Bayesian decision theory, which models belief states as single probability functions, faces a number of serious difficulties with respect to its handling of agnosticism. These difficulties have led to the increasing popularity of so-called ‘imprecise’ models of decision-making, which represent belief states as sets of probability functions. In a recent paper, however, Adam Elga has argued in favour of a putative normative principle of sequential choice that he claims to be (...) borne out by the sharp model but not by any promising incarnation of its imprecise counterpart. After first pointing out that Elga has fallen short of establishing that his principle is indeed uniquely borne out by the sharp model, I cast aspersions on its plausibility. I show that a slight weakening of the principle is satisfied by at least one, but interestingly not all, varieties of the imprecise model and point out that Elga has failed to motivate his stronger commitment. (shrink)

Conditional probability is one of the central concepts in probability theory. Some notion of conditional probability is part of every interpretation of probability. The basic mathematical fact about conditional probability is that p(A |B) = p(A ∧B)/p(B) where this is defined. However, while it has been typical to take this as a definition or analysis of conditional probability, some (perhaps most prominently Hájek, 2003) have argued that conditional probability should instead be taken as (...) the primitive notion, so that this formula is at best coextensive, and at worst sometimes gets it wrong. -/- Section 1.1 considers the concept of conditional probability in each of the major families of interpretation of probability. Section 1.2 considers a conceptual argument for the claim that conditional probability is prior to unconditional probability, while Section 1.3 considers a family of mathematical arguments for this claim, leading to consideration specifically of the question of how to understand probability conditional on events of probability 0. Section 1.4 discusses several mathematical principles that have been alleged to be important for understanding how probability 0 behaves, and raises a dilemma for probability conditional on events of probability 0. Section 2 and Section 3 take the two horns of this dilemma and describe the two main competing families of mathematical accounts of conditional probability for events of probability 0. Section 4 summarizes the results, and their significance for the two arguments that conditional probability is prior to unconditional probability. (shrink)

While the skeptical undercurrents of early modern thought have received sustained scholarly attention, such work has tended to be inattentive to hermeneutical or exegetical skepticism. This is a form of skepticism that threatened to stop hermeneutical theorizing in its tracks and absorbed several central hermeneutical concepts in its orbit. Hermeneutical probability was one of them. In this paper, I aim to examine whether the doctrine of hermeneutical probability as it was originally formulated by Christian Thomasius is a surrogate (...) of exegetical skepticism. I argue that not only is the examined view incompatible with skepticism, but that it can indeed articulate a promising response to it. Since general reflection on probability shaped the philosophical landscape of the Early German Enlightenment, the viability of the Thomasian view has much broader ramifications that extend beyond hermeneutics to historics, ethics, and physics. (shrink)

Many have argued that a rational agent's attitude towards a proposition may be better represented by a probability range than by a single number. I show that in such cases an agent will have unstable betting behaviour, and so will behave in an unpredictable way. I use this point to argue against a range of responses to the ‘two bets’ argument for sharp probabilities.

Subjective probability plays an increasingly important role in many fields concerned with human cognition and behavior. Yet there have been significant criticisms of the idea that probabilities could actually be represented in the mind. This paper presents and elaborates a view of subjective probability as a kind of sampling propensity associated with internally represented generative models. The resulting view answers to some of the most well known criticisms of subjective probability, and is also supported by empirical work (...) in neuroscience and behavioral psychology. The repercussions of the view for how we conceive of many ordinary instances of subjective probability, and how it relates to more traditional conceptions of subjective probability, are discussed in some detail. (shrink)

The Born probability measure describes the statistics of measurements in which observers self-locate themselves in some region of reality. In psi-ontic quantum theories, reality is directly represented by the wavefunction. We show that quantum probabilities may be identified using fractions of a universal multiple-time wavefunction containing both causal and retrocausal temporal parts. This wavefunction is defined in an appropriately generalized history space on the Keldysh time contour. Our deterministic formulation of quantum mechanics replaces the initial condition of standard Schrödinger (...) dynamics, with a network of ‘fixed points’ defining quantum histories on the contour. The Born measure is derived by summing up the wavefunction along these histories. We then apply the same technique to the derivation of the statistics of measurements with pre- and postselection. (shrink)

John Maynard Keynes’s A Treatise on Probability is the seminal text for the logical interpretation of probability. According to his analysis, probabilities are evidential relations between a hypothesis and some evidence, just like the relations of deductive logic. While some philosophers had suggested similar ideas prior to Keynes, it was not until his Treatise that the logical interpretation of probability was advocated in a clear, systematic and rigorous way. I trace Keynes’s influence in the philosophy of (...) class='Hi'>probability through a heterogeneous sample of thinkers who adopted his interpretation. This sample consists of Frederick C. Benenson, Roy Harrod, Donald C. Williams, Henry E. Kyburg and David Stove. The ideas of Keynes prove to be adaptable to their diverse theories of probability. My discussion indicates both the robustness of Keynes’s probability theory and the importance of its influence on the philosophers whom I describe. I also discuss the Problem of the Priors. I argue that none of those I discuss have obviously improved on Keynes’s theory with respect to this issue. (shrink)

Suppose that, yesterday at noon, Maria considered flipping a fair coin, but didn't. What probability do you assign to "If Maria had flipped the coin, the coin would have landed heads"? Now suppose that, contrary to fact, Maria did indeed flip the coin. In that counterfactual scenario, what is the probability of "The coin will land tails"? The two questions sound strikingly similar. I argue that they sound similar because they are equivalent. The chance of a counterfactual "If (...) A, would C" equals the chance of C, in the counterfactual scenario that A (and a similar principle holds for credence). This principle does better than similar principles that have been defended (like Skyrms' Thesis), avoids triviality, and gives us important clues for a semantics for counterfactuals. (shrink)

Probability can be used to measure degree of belief in two ways: objectively and subjectively. The objective measure is a measure of the rational degree of belief in a proposition given a set of evidential propositions. The subjective measure is the measure of a particular subject’s dispositions to decide between options. In both measures, certainty is a degree of belief 1. I will show, however, that there can be cases where one belief is stronger than another yet both beliefs (...) are plausibly measurable as objectively and subjectively certain. In ordinary language, we can say that while both beliefs are certain, one belief is more certain than the other. I will then propose second, non probabilistic dimension of measurement, which tracks this variation in certainty in such cases where the probability is 1. A general principle of rationality is that one’s subjective degree of belief should match the rational degree of belief given the evidence available. In this paper I hope to show that it is also a rational principle that the maximum stake size at which one should remain certain should match the rational weight of certainty given the evidence available. Neither objective nor subjective measures of certainty conform to the axioms of probability, but instead are measured in utility. This has the consequence that, although it is often rational to be certain to some degree, there is no such thing as absolute certainty. (shrink)

This paper argues that probability is structurally downstream of identity persistence, invariant structure, and bounded observability. The proposed ordering is: identity → invariant → observable → probability. Probability remains indispensable for bounded reasoning, but it cannot govern identity because it presupposes stable observables, and stable observables presuppose invariant structure sufficient for same/not-same attribution. -/- The paper applies this ordering to hallucination in generative AI. Hallucination is characterized as symbolic continuation that has drifted beyond lawful invariant anchoring while (...) preserving local probabilistic plausibility. This explains why hallucinated outputs can feel coherent while remaining source-disconnected, persistence-inconsistent, or structurally false. -/- The argument is structural and epistemic, not ontological. It does not claim that identity is the metaphysical ground of reality, nor that probabilistic methods are invalid. It claims only that in bounded epistemic systems, probability operates over observable domains whose identity conditions must already be settled. The paper distinguishes scalar governance as the form of invariant compression from scalar estimation as a downstream epistemic act, then draws implications for AI hallucination, verification infrastructure, replay-stable artifacts, auditability, and trust surfaces. (shrink)

Enjoying great popularity in decision theory, epistemology, and philosophy of science, Bayesianism as understood here is fundamentally concerned with epistemically ideal rationality. It assumes a tight connection between evidential probability and ideally rational credence, and usually interprets evidential probability in terms of such credence. Timothy Williamson challenges Bayesianism by arguing that evidential probabilities cannot be adequately interpreted as the credences of an ideal agent. From this and his assumption that evidential probabilities cannot be interpreted as the actual credences (...) of human agents either, he concludes that no interpretation of evidential probabilities in terms of credence is adequate. I argue to the contrary. My overarching aim is to show on behalf of Bayesians how one can still interpret evidential probabilities in terms of ideally rational credence and how one can maintain a tight connection between evidential probabilities and ideally rational credence even if the former cannot be interpreted in terms of the latter. By achieving this aim I illuminate the limits and prospects of Bayesianism. (shrink)

Modern scientific cosmology pushes the boundaries of knowledge and the knowable. This is prompting questions on the nature of scientific knowledge. A central issue is what defines a 'good' model. When addressing global properties of the Universe or its initial state this becomes a particularly pressing issue. How to assess the probability of the Universe as a whole is empirically ambiguous, since we can examine only part of a single realisation of the system under investigation: at some point, data (...) will run out. We review the basics of applying Bayesian statistical explanation to the Universe as a whole. We argue that a conventional Bayesian approach to model inference generally fails in such circumstances, and cannot resolve, e.g., the so-called 'measure problem' in inflationary cosmology. Implicit and non-empirical valuations inevitably enter model assessment in these cases. This undermines the possibility to perform Bayesian model comparison. One must therefore either stay silent, or pursue a more general form of systematic and rational model assessment. We outline a generalised axiological Bayesian model inference framework, based on mathematical lattices. This extends inference based on empirical data (evidence) to additionally consider the properties of model structure (elegance) and model possibility space (beneficence). We propose this as a natural and theoretically well-motivated framework for introducing an explicit, rational approach to theoretical model prejudice and inference beyond data. (shrink)

In finite probability theory, events are subsets S⊆U of the outcome set. Subsets can be represented by 1-dimensional column vectors. By extending the representation of events to two dimensional matrices, we can introduce "superposition events." Probabilities are introduced for classical events, superposition events, and their mixtures by using density matrices. Then probabilities for experiments or `measurements' of all these events can be determined in a manner exactly like in quantum mechanics (QM) using density matrices. Moreover the transformation of the (...) density matrices induced by the experiments or `measurements' is the Lüders mixture operation as in QM. And finally by moving the machinery into the n-dimensional vector space over ℤ₂, different basis sets become different outcome sets. That `non-commutative' extension of finite probability theory yields the pedagogical model of quantum mechanics over ℤ₂ that can model many characteristic non-classical results of QM. (shrink)

The role of probability is one of the most contested issues in the interpretation of contemporary physics. In this paper, I’ll be reevaluating some widely held assumptions about where and how probabilities arise. Larry Sklar voices the conventional wisdom about probability in classical physics in a piece in the Stanford Online Encyclopedia of Philosophy, when he writes that “Statistical mechanics was the first foundational physical theory in which probabilistic concepts and probabilistic explanation played a fundamental role.” And the (...) conventional wisdom about quantum probabilities is that they are basic, not reducible to the types of probabilities we see in statistical mechanics. In the first section of this paper, I’ll argue that in fact classical physics was steeped in probability long before statistical mechanics came on the scene, specifically, that an objective measure over phase space is an indispensable component of any informative physical theory. In the next section, I’ll argue that this objective measure is the fundamental form of physical probability and that quantum probabilities can be defined in terms of it. In the last, I’ll raise some questions about the metaphysical status of the fundamental measure. (shrink)

The article is a plea for ethicists to regard probability as one of their most important concerns. It outlines a series of topics of central importance in ethical theory in which probability is implicated, often in a surprisingly deep way, and lists a number of open problems. Topics covered include: interpretations of probability in ethical contexts; the evaluative and normative significance of risk or uncertainty; uses and abuses of expected utility theory; veils of ignorance; Harsanyi’s aggregation theorem; (...) population size problems; equality; fairness; giving priority to the worse off; continuity; incommensurability; nonexpected utility theory; evaluative measurement; aggregation; causal and evidential decision theory; act consequentialism; rule consequentialism; and deontology. (shrink)

Stalnaker's Thesis about indicative conditionals is, roughly, that the probability one ought to assign to an indicative conditional equals the probability that one ought to assign to its consequent conditional on its antecedent. The thesis seems right. If you draw a card from a standard 52-card deck, how confident are you that the card is a diamond if it's a red card? To answer this, you calculate the proportion of red cards that are diamonds -- that is, you (...) calculate the probability of drawing a diamond conditional on drawing a red card. Skyrms' Thesis about counterfactual conditionals is, roughly, that the probability that one ought to assign to a counterfactual equals one's rational expectation of the chance, at a relevant past time, of its consequent conditional on its antecedent. This thesis also seems right. If you decide not to enter a 100-ticket lottery, how confident are you that you would have won had you bought a ticket? To answer this, you calculate the prior chance--that is, the chance just before your decision not to buy a ticket---of winning conditional on entering the lottery. The central project of this article is to develop a new uniform theory of conditionals that allows us to derive a version of Skyrms' Thesis from a version of Stalnaker's Thesis, together with a chance-deference norm relating rational credence to beliefs about objective chance. (shrink)

In this paper, we illustrate some serious difficulties involved in conveying information about uncertain risks and securing informed consent for risky interventions in a clinical setting. We argue that in order to secure informed consent for a medical intervention, physicians often need to do more than report a bare, numerical probability value. When probabilities are given, securing informed consent generally requires communicating how probability expressions are to be interpreted and communicating something about the quality and quantity of the (...) evidence for the probabilities reported. Patients may also require guidance on how probability claims may or may not be relevant to their decisions, and physicians should be ready to help patients understand these issues. (shrink)

A probability distribution is regular if no possible event is assigned probability zero. While some hold that probabilities should always be regular, three counter-arguments have been posed based on examples where, if regularity holds, then perfectly similar events must have different probabilities. Howson (2017) and Benci et al. (2016) have raised technical objections to these symmetry arguments, but we see here that their objections fail. Howson says that Williamson’s (2007) “isomorphic” events are not in fact isomorphic, but Howson (...) is speaking of set-theoretic representations of events in a probability model. While those sets are not isomorphic, Williamson’s physical events are, in the relevant sense. Benci et al. claim that all three arguments rest on a conflation of different models, but they do not. They are founded on the premise that similar events should have the same probability in the same model, or in one case, on the assumption that a single rotation-invariant distribution is possible. Having failed to refute the symmetry arguments on such technical grounds, one could deny their implicit premises, which is a heavy cost, or adopt varying degrees of instrumentalism or pluralism about regularity, but that would not serve the project of accurately modelling chances. (shrink)

This essay proposes fundamental ontological clarification within contemporary de-bates on quantum physics: the rigorous distinction between probability as a regime of prediction and the real as material dynamics. Against the widespread view ac-cording to which quantum indeterminacy would imply a vague, unstable, or “merely statistical” reality, the text argues that probability belongs to the epistemic plane of knowledge and anticipation, not to the ontological plane of what exists. Quantum physics does not describe an ontologically fluctuating world, but a (...) world whose symbolic access is structurally limited. On this basis, the essay examines the role of Born’s statistical interpretation, decoherence theory, Bell’s theorems, and the Pu-sey–Barrett–Rudolph result, showing how these impose formal costs on any ontol-ogy that conflates formalism and reality. In critical dialogue with QBist, relational, and agential-realist approaches, the text develops a conception of concretion grounded in the notion of mark: the concrete emerges when a material interaction is irreversibly inscribed in a physical support. Stability, far from meaning fixity, is un-derstood as emergent robustness—movement in regime—sufficient to sustain prac-tices, techniques, and historical continuity. The essay concludes by arguing that quantum physics, rather than dissolving the real, demands a more exacting ontology capable of distinguishing precisely between the real, the concrete, and theory. (shrink)

-/- Why do such different systems — quantum measurements, thermal ensembles, Bayesian updates and large language models — keep producing probability mappings with strikingly similar structure? We introduce relational probability fields: distributions over discrete possibilities shaped by context-dependent relational potentials and resolved by collapse, sampling, or flow. We show that the Born rule, Boltzmann statistics and softmax sampling can be expressed within a single template: potentials induce an outcome distribution, and resolution instantiates particular outcomes. -/- The central contribution (...) of this paper is a set of interface style descriptors — concentration, context sensitivity and memory — that characterise how resolution behaves across domains. These descriptors do not replace domain-specific formalisms; they form a comparative layer that makes cross-domain differences and similarities tractable. Two systems may share the same abstract template while exhibiting very different interface styles: one may generate sharp, low-sensitivity distributions with little memory; another diffuse, high-sensitivity distributions with strong history dependence. Together, these descriptors operationalise what we call a system’s relational mode: its characteristic way of turning contextual relations into realised outcomes. -/- We provide explicit definitions and an alignment map that shows how the template and descriptors apply in quantum measurement, statistical mechanics, Bayesian updating and next-token prediction. Built on this structural foundation, we sketch a speculative coherence-based interpretation: probabilities as measures of relational fit between configurations and their context, and collapse as a form of coherence optimisation. The paper is primarily conceptual: it does not propose a new microphysical theory, nor does it derive the Born rule or Einstein’s equations from a single principle. Instead, it offers a compact cross-domain language designed to generate tractable questions at the interface of quantum theory, statistical mechanics and machine learning — with possible extensions discussed in the outlook. (shrink)

Classical probability theory is built upon rigid structures: events, distributions, and measures defined on a fixed number line. Shinayaka Probability Theory, inspired by Shinayaka Geometry, introduces elas- ticity into probabilistic models, allowing distributions to flex according to contextual conditions. This paper defines elastic probability dis- tributions, derives their mathematical properties, and explores their significance in adaptive statistics and information theory.

Recent years have witnessed a proliferation of attempts to apply the mathematical theory of probability to the semantics of natural language probability talk. These sorts of “probabilistic” semantics are often motivated by their ability to explain intuitions about inferences involving “likely” and “probably”—intuitions that Angelika Kratzer’s canonical semantics fails to accommodate through a semantics based solely on an ordering of worlds and a qualitative ranking of propositions. However, recent work by Wesley Holliday and Thomas Icard has been widely (...) thought to undercut this motivation: they present a world-ordering semantics that yields essentially the same logic as probabilistic semantics. In this paper, I argue that the challenge remains: defenders of world-ordering semantics have yet to offer a plausible semantics that captures the logic of comparative likelihood. Holliday & Icard’s semantics yields an adequate logic only if models are restricted to Noetherian pre-orders. But I argue that the Noetherian restriction faces problems in cases involving infinitely large domains of epistemic possibilities. As a result, probabilistic semantics remains the better explanation of the data. (shrink)

In standard probability theory, events of probability zero may still occur. That is, a real number $x$ can be sampled uniformly at random from $[0,1]$ despite having $\mathbb{P}(X=x)=0$. This goes against the intuitive principle that probability $0$ means impossible. In this paper, we make this tension explicit and argue that the usual resolution is backwards. We begin by analyzing two common claims: 1. (Zero) Probability $0$ means impossible for any possible outcome. 2. (Uniform) It is possible (...) to sample a real number: there exists an experiment whose possible outcomes are all $x\in[0,1]$, distributed according to the uniform (Lebesgue) measure, and which returns a completed real on each run. These two claims are incompatible. Standard probability theory keeps (Uniform) and abandons (Zero), allowing events of probability zero to occur. We argue for the opposite choice. We retain (Zero) and deny that there is any literal experiment that samples a completed real from $[0,1]$ with the uniform law. The main tool is definability. We propose that to count as a number at all, an object must be definable by a finite description in a fixed language. Undefinable ``real numbers'' then do not exist as numbers; they are set theoretic artifacts rather than genuine numerical quantities. Under any continuous distribution on $[0,1]$, the set of definable reals has probability zero, so an idealized ``random real'' is almost surely an undefinable pseudo object. On our view this is a sign that the usual continuum picture is being applied outside its proper domain. We develop this finite information perspective across several domains: definable reals and complex numbers, definable infinitesimals and infinite quantities in hyperreal fields, definable distributions such as the Dirac delta, and definable infinities in extended number systems. We stress the abundance of definable numerical objects: giving up undefinable reals does not impoverish the numerical universe but focuses attention on a rich and extensible world of definable numbers. We also revisit Cantor's diagonal argument. We accept its conclusion as a theorem of set theory about the uncountability of the set theoretic reals, but argue that its anti diagonal construction presupposes the existence of undefinable real codes. Once we restrict attention to definable numbers, Cantor's theorem is about a larger background universe than the domain of genuine numerical quantities. Finally, we articulate a process based view of sampling on the continuum. No physical or computational procedure ever produces an infinite real; what exists in practice is a sampling process that yields arbitrarily fine but always finite information. In this operational setting, events with probability zero that refer to completed infinite outcomes never occur, and the principle that probability $0$ means impossible holds for all physically meaningful events. (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 work examines in depth the methodological relationships that probability and statistics have maintained with the social sciences from their emergence. It covers both the history of thought and currrent methods. First it examines in detail the history of the different paradigms and axioms for probability, from their emergence in the seventeenth century up to the most recent developments of three main concepts: objective, subjective and logical probability. It shows the statistical inference they present, different applications to (...) social sciences and the main problems they encounter. On the other side, from social sciences-particularly population sciences-to probability, it shows the different uses they made of probabilistic concepts during their history, from the seventeenth century, according to their paradigms: cross-sectional, longitudinal, biographical, hierarchical contextual and multilevel approaches. While the ties may have seemed loose at times, they have more often been very close: some advances in probability were driven by the search of anwers to questions raised by the social sciences, conversly the latter have made progress thanks to advances in probability. This dual approach sheds new light on the historical development of the social sciences and probability, and the enduring relevance of their links. It permits also to solve a number of methodological problems encountered all along their history. (shrink)

It is tempting to think that a process of choosing a point at random from the surface of a sphere can be probabilistically symmetric, in the sense that any two regions of the sphere which differ by a rotation are equally likely to include the chosen point. Isaacs, Hájek, and Hawthorne (2022) argue from such symmetry principles and the mathematical paradoxes of measure to the existence of imprecise chances and the rationality of imprecise credences. Williamson (2007) has argued from a (...) related symmetry principle to the failure of probabilistic regularity. We contend that these arguments fail, because they rely on auxiliary assumptions about probability which are inconsistent with symmetry to begin with. We argue, moreover, that symmetry should be rejected in light of this inconsistency, and because it has implausible decision- theoretic implications. -/- The weaker principle of probabilistic invariance says that the probabilistic comparison of any two regions is unchanged by rotations of the sphere. This principle supports a more compelling argument for imprecise probability. We show, however, that invariance is incompatible with mundane judgments about what is probable. Ultimately, we find reason to be suspicious of the application of principles like symmetry and invariance to nonmeasurable regions. (shrink)

This paper argues that the technical notion of conditional probability, as given by the ratio analysis, is unsuitable for dealing with our pretheoretical and intuitive understanding of both conditionality and probability. This is an ontological account of conditionals that include an irreducible dispositional connection between the antecedent and consequent conditions and where the conditional has to be treated as an indivisible whole rather than compositional. The relevant type of conditionality is found in some well-defined group of conditional statements. (...) As an alternative, therefore, we briefly offer grounds for what we would call an ontological reading: for both conditionality and conditional probability in general. It is not offered as a fully developed theory of conditionality but can be used, we claim, to explain why calculations according to the RATIO scheme does not coincide with our intuitive notion of conditional probability. What it shows us is that for an understanding of the whole range of conditionals we will need what John Heil (2003), in response to Quine (1953), calls an ontological point of view. (shrink)

In the Bayesian approach to quantum mechanics, probabilities—and thus quantum states—represent an agent’s degrees of belief, rather than corresponding to objective properties of physical systems. In this paper we investigate the concept of certainty in quantum mechanics. Particularly, we show how the probability-1 predictions derived from pure quantum states highlight a fundamental difference between our Bayesian approach, on the one hand, and Copenhagen and similar interpretations on the other. We first review the main arguments for the general claim that (...) probabilities always represent degrees of belief. We then argue that a quantum state prepared by some physical device always depends on an agent’s prior beliefs, implying that the probability-1 predictions derived from that state also depend on the agent’s prior beliefs. Quantum certainty is therefore always some agent’s certainty. Conversely, if facts about an experimental setup could imply agent-independent certainty for a measurement outcome, as in many Copenhagen-like interpretations, that outcome would effectively correspond to a preexisting system property. The idea that measurement outcomes occurring with certainty correspond to preexisting system properties is, however, in conflict with locality. We emphasize this by giving a version of an argument of Stairs [(1983). Quantum logic, realism, and value-definiteness. Philosophy of Science, 50, 578], which applies the Kochen–Specker theorem to an entangled bipartite system. (shrink)

The epistemic probability of A given B is the degree to which B evidentially supports A, or makes A plausible. This paper is a first step in answering the question of what determines the values of epistemic probabilities. I break this question into two parts: the structural question and the substantive question. Just as an object’s weight is determined by its mass and gravitational acceleration, some probabilities are determined by other, more basic ones. The structural question asks what probabilities (...) are not determined in this way—these are the basic probabilities which determine values for all other probabilities. The substantive question asks how the values of these basic probabilities are determined. I defend an answer to the structural question on which basic probabilities are the probabilities of atomic propositions conditional on potential direct explanations. I defend this against the view, implicit in orthodox mathematical treatments of probability, that basic probabilities are the unconditional probabilities of complete worlds. I then apply my answer to the structural question to clear up common confusions in expositions of Bayesianism and shed light on the “problem of the priors.”. (shrink)

This paper argues that debates about probability often begin at the wrong level. Before asking whether probability is objective or subjective, ontic or epistemic, realist or instrumental, one should first ask what explanatory work probability is being asked to do. Once that prior question is made explicit, a recurrent pattern comes into view. Probability is repeatedly granted authority it has not earned: elevated into ontology, made to substitute for generation, and then allowed to move ambiguously between (...) primitive structure, generative mechanism, and descriptive appearance. Against this pattern, the paper defends a three-layer distinction between admissible basis, structural attainment, and readout. On that basis, it argues, first, that probability cannot securely do ontological work, because primitive probabilistic structure too easily imports into ontology the differentiated outcome-space that should remain under explanatory pressure. Second, probability cannot replace generation, because a distribution over possibilities, however refined, is not yet an account of attained standing. Distribution is not attainment, and repeated statistical order is not single-case generation. The paper then advances a positive reclassification. Probability is best understood not as an ontological primitive and not as a generative mechanism, but as a readout residue: the stable statistical appearance assumed by attained structure when it becomes legible only under selective and truncating conditions of access. Temporal windowing, bounded responsiveness, and threshold-conditioned triggerability are introduced as principal modes of truncating readout. The claim is not that truncation as such yields probability, but that where truncation is structured, stable, and repeatable, it can preserve an invariant profile of appearance that takes probabilistic form. The paper further argues that this residue thesis carries substantive commitments. If probability genuinely belongs to readout, the relevant readout conditions must be specifiable in principle, probabilistic appearance must be sensitive in principle to systematic variation in those conditions, and the account must expose itself to principled failure wherever those conditions prove explanatorially idle. Probability is thus retained as real, structured, and indispensable at the level of appearance, but denied the authority to define what reality fundamentally is or how structural attainment is fundamentally borne. (shrink)

This paper presents a unified theory of the truth conditions and probability of indicative conditionals and their compounds in a trivalent framework. The semantics validates a Reduction Theorem: any compound of conditionals is semantically equivalent to a simple conditional. This allows us to validate Stalnaker's Thesis in full generality and to use Adams's notion of $p$-validity as a criterion for valid inference. Finally, this gives us an elegant account of Bayesian update with indicative conditionals, establishing that despite differences in (...) meaning, it is tantamount to learning a material conditional. (shrink)

We will flip a fair coin infinitely many times. Al calls the first flip, claiming it will land heads. Betty calls every odd numbered flip, claiming they will all land heads. Carl calls every flip bar none, claiming they will all land heads. Pre-theoretically, it seems that Al's claim is infinitely more likely than Betty's, and that Betty's claim is infinitely more likely than Carl's. But standard, real-valued probability theory says that, while Al's claim is infinitely more likely than (...) Betty's, Betty's claim is exactly as likely as Carl's. I use John Conway's surreal numbers to capture these pre-theoretic judgements about the relative probabilities of Al's, Betty's, and Carl's claims. Surreal-valued probabilities also allow us to calculate precise numerical values for those claims. For instance, we will see that Betty's claim has a probability equal to the reciprocal of the product of the square root of cardinality of the continuum and the fourth root of 2. The article provides a 'user's guide' to surreal-valued probabilities and raises questions for further research. (shrink)

We propose a new account of indicative conditionals, giving acceptability and logical closure conditions for them. We start from Adams’ Thesis: the claim that the acceptability of a simple indicative equals the corresponding conditional probability. The Thesis is widely endorsed, but arguably false and refuted by empirical research. To fix it, we submit, we need a relevance constraint: we accept a simple conditional 'If φ, then ψ' to the extent that (i) the conditional probability p(ψ|φ) is high, provided (...) that (ii) φ is relevant for ψ. How (i) should work is well-understood. It is (ii) that holds the key to improve our understanding of conditionals. Our account has (i) a probabilistic component, using Popper functions; (ii) a relevance component, given via an algebraic structure of topics or subject matters. We present a probabilistic logic for simple indicatives, and argue that its (in)validities are both theoretically desirable and in line with empirical results on how people reason with conditionals. (shrink)

Early work on the frequency theory of probability made extensive use of the notion of randomness, conceived of as a property possessed by disorderly collections of outcomes. Growing out of this work, a rich mathematical literature on algorithmic randomness and Kolmogorov complexity developed through the twentieth century, but largely lost contact with the philosophical literature on physical probability. The present chapter begins with a clarification of the notions of randomness and probability, conceiving of the former as a (...) property of a sequence of outcomes, and the latter as a property of the process generating those outcomes. A discussion follows of the nature and limits of the relationship between the two notions, with largely negative verdicts on the prospects for any reduction of one to the other, although the existence of an apparently random sequence of outcomes is good evidence for the involvement of a genuinely chancy process. (shrink)

Credences are beliefs about evidential probabilities. We give the view an assessment-sensitive formulation, show how it evades the standard objections, and give several arguments in support.

I conjecture that the probability amplitudes of a quantum state are fractions of the Planck frequency, stemming from the rich dynamics at the Planck scale. This offers a means to indirectly measure the fundamental properties of quantum spacetime and potentially resolves the measurement problem.

A longstanding issue in attempts to understand the Everett (Many-Worlds) approach to quantum mechanics is the origin of the Born rule: why is the probability given by the square of the amplitude? Following Vaidman, we note that observers are in a position of self-locating uncertainty during the period between the branches of the wave function splitting via decoherence and the observer registering the outcome of the measurement. In this period it is tempting to regard each branch as equiprobable, but (...) we argue that the temptation should be resisted. Applying lessons from this analysis, we demonstrate (using methods similar to those of Zurek's envariance-based derivation) that the Born rule is the uniquely rational way of apportioning credence in Everettian quantum mechanics. In doing so, we rely on a single key principle: changes purely to the environment do not affect the probabilities one ought to assign to measurement outcomes in a local subsystem. We arrive at a method for assigning probabilities in cases that involve both classical and quantum self-locating uncertainty. This method provides unique answers to quantum Sleeping Beauty problems, as well as a well-defined procedure for calculating probabilities in quantum cosmological multiverses with multiple similar observers. (shrink)

Epistemic closure under known implication is the principle that knowledge of "p" and knowledge of "p implies q", together, imply knowledge of "q". This principle is intuitive, yet several putative counterexamples have been formulated against it. This paper addresses the question, why is epistemic closure both intuitive and prone to counterexamples? In particular, the paper examines whether probability theory can offer an answer to this question based on four strategies. The first probability-based strategy rests on the accumulation of (...) risks. The problem with this strategy is that risk accumulation cannot accommodate certain counterexamples to epistemic closure. The second strategy is based on the idea of evidential support, that is, a piece of evidence supports a proposition whenever it increases the probability of the proposition. This strategy makes progress and can accommodate certain putative counterexamples to closure. However, this strategy also gives rise to a number of counterintuitive results. Finally, there are two broadly probabilistic strategies, one based on the idea of resilient probability and the other on the idea of assumptions that are taken for granted. These strategies are promising but are prone to some of the shortcomings of the second strategy. All in all, I conclude that each strategy fails. Probability theory, then, is unlikely to offer the account we need. (shrink)

There is a trade-off between specificity and accuracy in existing models of belief. Descriptions of agents in the tripartite model, which recognizes only three doxastic attitudes—belief, disbelief, and suspension of judgment—are typically accurate, but not sufficiently specific. The orthodox Bayesian model, which requires real-valued credences, is perfectly specific, but often inaccurate: we often lack precise credences. I argue, first, that a popular attempt to fix the Bayesian model by using sets of functions is also inaccurate, since it requires us to (...) have interval-valued credences with perfectly precise endpoints. We can see this problem as analogous to the problem of higher order vagueness. Ultimately, I argue, the only way to avoid these problems is to endorse Insurmountable Unclassifiability. This principle has some surprising and radical consequences. For example, it entails that the trade-off between accuracy and specificity is in-principle unavoidable: sometimes it is simply impossible to characterize an agent’s doxastic state in a way that is both fully accurate and maximally specific. What we can do, however, is improve on both the tripartite and existing Bayesian models. I construct a new model of belief—the minimal model—that allows us to characterize agents with much greater specificity than the tripartite model, and yet which remains, unlike existing Bayesian models, perfectly accurate. (shrink)