How should a machine intelligence perform unsupervised structure discovery over streams of sensory input? One approach to this problem is to cast it as an apperception task [1]. Here, the task is to construct an explicit interpretable theory that both explains the sensory sequence and also satisfies a set of unity conditions, designed to ensure that the constituents of the theory are connected in a relational structure. However, the original formulation of the apperception task had one fundamental limitation: it assumed (...) the raw sensory input had already been parsed using a set of discrete categories, so that all the system had to do was receive this already-digested symbolic input, and make sense of it. But what if we don't have access to pre-parsed input? What if our sensory sequence is raw unprocessed information? The central contribution of this paper is a neuro-symbolic framework for distilling interpretable theories out of streams of raw, unprocessed sensory experience. First, we extend the definition of the apperception task to include ambiguous (but still symbolic) input: sequences of sets of disjunctions. Next, we use a neural network to map raw sensory input to disjunctive input. Our binary neural network is encoded as a logic program, so the weights of the network and the rules of the theory can be solved jointly as a single SAT problem. This way, we are able to jointly learn how to perceive (mapping raw sensory information to concepts) and apperceive (combining concepts into declarative rules). (shrink)

How should one react when one has a belief, but knows that other people—who have roughly the same evidence as one has, and seem roughly as likely to react to it correctly—disagree? This paper argues that the disagreement of other competent inquirers often requires one to be much less confident in one’s opinions than one would otherwise be.

I insist that I was able to raise my hand, and I acknowledge that a law would have been broken had I done so, but I deny that I am therefore able to break a law. To uphold my instance of soft determinism, I need not claim any incredible powers. To uphold the compatibilism that I actually believe, I need not claim that such powers are even possible. My incompatibilist opponent is a creature of fiction, but he has his prototypes (...) in real life. He is modeled partly after Peter van Inwagen and partly on myself when I first worried about van Inwagen's argument against compatibilism. (shrink)

This paper offers both a theoretical and an experimental perspective on the relationship between connectionist and Classical (symbol-processing) models. Firstly, a serious flaw in Fodor and Pylyshyn’s argument against connectionism is pointed out: if, in fact, a part of their argument is valid, then it establishes a conclusion quite different from that which they intend, a conclusion which is demonstrably false. The source of this flaw is traced to an underestimation of the differences between localist and distributed representation. It has (...) been claimed that distributed representations cannot support systematic operations, or that if they can, then they will be mere implementations of traditional ideas. This paper presents experimental evidence against this conclusion: distributed representations can be used to support direct structure-sensitive operations, in a man- ner quite unlike the Classical approach. Finally, it is argued that even if Fodor and Pylyshyn’s argument that connectionist models of compositionality must be mere implementations were correct, then this would still not be a serious argument against connectionism as a theory of mind. (shrink)

A critical examination of the Reflection principle in Bayesian epistemology, and of the diachronic Dutch-book-style arguments that have been invoked to support Reflection and Conditionalization.

It has often been noticed that conciliatory views of disagreement are "self-undermining" in a certain way: advocates of such views cannot consistently maintain them when other philosophers disagree. This leads to apparent problems of instability and even inconsistency. Does self-undermining, then, show conciliationism untenable? If so, the untenablity would extend not only to almost all views of disagreement, but to a wide range of other views supporting what one might call epistemic modesty: roughly, the idea that getting evidence that one (...) has made an epistemic error in arriving at one’s opinion may require adjusting that opinion. This paper argues that the phenomenon of self-undermining does not disclose any defect in views mandating epistemic modesty. Instead, it highlights an uncomfortable but natural consequence of reflecting on one's own possible epistemic imperfections, a sort of reflection that tends to cause epistemic ideals to conflict. (shrink)

The old evidence problem affects any probabilistic confirmation measure based on comparing pr(H/E) and pr(H). The article argues for the following points: (1) measures based on likelihood ratios also suffer old evidence difficulties; (2) the less-discussed synchronic old evidence problem is, in an important sense, the most acute; (3) prominent attempts to solve or dissolve the synchronic problem fail; (4) a little-discussed variant of the standard measure avoids the problem, in an appealing way; and (5) this measure nevertheless reveals a (...) different problem for probabilistic confirmation measures, a problem that is unlikely to lend itself to formal solution. (shrink)

Conciliationism holds that disagreement of apparent epistemic peers often substantially undermines rational confidence in our opinions. Uniqueness principles say that there is at most one maximally rational doxastic response to any given batch of total evidence. The two views are often thought to be tightly connected. This paper distinguishes two ways of motivating conciliationism, and two ways that conciliationism may be undermined by permissive accounts of rationality. It shows how conciliationism can flourish under certain strongly permissive accounts of rationality. This (...) occurs when the motivation for conciliationism does not come from taking disagreement as evidence of one's own rational failings. However, divorcing the motive for conciliating from worries about rationality does not remove a feature of conciliationism that some find troubling: that conciliationism can lead to cases of “rational toxicity,” in which the most rational response to one's evidence involves violating some rational ideal. (shrink)

Consequentialism, many philosophers have claimed, asks too much of us to be a plausible ethical theory. Indeed, the theory's severe demandingness is often claimed to be its chief flaw. My thesis is that as we come to better understand this objection, we see that, even if it signals or tracks the existence of a real problem for Consequentialism, it cannot itself be a fundamental problem with the view. The objection cannot itself provide good reason to break with Consequentialism, because it (...) must presuppose prior and independent breaks with the view. The way the objection measures the demandingness of an ethical theory reflects rather than justifies being in the grip of key anti-Consequentialist conclusions. We should reject Consequentialism independently of the Objection or not at all. Thus, we can reduce by one the list of worrisome fundamental complaints against Consequentialism. (shrink)

The principle of alternate possibilities (PAP), making the ability to do otherwise a necessary condition for moral responsibility, is supposed by Harry Frankfurt, John Fischer, and others to succumb to a peculiar kind of counterexample. The paper reviews the main problems with the counterexample that have surfaced over the years, and shows how most can be addressed within the terms of the current debate. But one problem seems ineliminable: because Frankfurt''s example relies on a counterfactual intervener to preclude alternatives to (...) the person''s action, it is not possible for it to preclude all alternatives (intervention that is contingent upon a trigger cannot bring it about that the trigger never occurred). This makes it possible for the determined PAPist to maintain that some pre-intervention deviation is always available to ground moral responsibility. In reply, the critic of PAP can examine all the candidate deviations and argue their irrelevance to moral responsibility (a daunting prospect); or the critic can dispense with counterfactual intervention altogether. The paper pursues the second of these strategies, developing three examples of noncounterfactual intervention in which (i) the agent has no alternatives (and a fortiori no morally relevant alternatives), yet (ii) there is just as much reason to think that the agent is morally responsible as there was in Frankfurt''s original example. The new counterexamples do suffer from one liability, but this is insufficient in the end to repair PAP''s conceptual connection between moral responsibility and alternate possibilities. (shrink)

This book introduces readers to the topic of explanation. The insights of Plato, Aristotle, J.S. Mill and Carl Hempel are examined, and are used to argue against the view that explanation is merely a problem for the philosophy of science. Having established its importance for understanding knowledge in general, the book concludes with a bold and original explanation of explanation.

A wide range of prominent epistemological theories include a principle of conservatism. Such principles take the fact that an agent currently holds a certain belief to constitute at least some measure of epistemic justification for her to maintain that belief. I examine the main arguments that have been made in conservatism's behalf, and find them unsound. Most interestingly, conservatism does not fall out of confirmational holism (the view that the justification of each of our beliefs is in part determined by (...) our other beliefs). I argue that conservatism is a mild form of dogmatism, and should be abandoned. (shrink)

According to the Principle of Sufficient Reason, every fact has an explanation. An important challenge to this principle is that it risks being a counterexample to itself. What explains why everything needs to be explained? My first goal is to distinguish two broad kinds of answers to this question, which I call “Humean Rationalism” and “Non-Humean Rationalism”. My second goal will be to defend the prospects of Humean Rationalism.

Much recent discussion in the philosophy of mathematics has concerned the indispensability argument—an argument which aims to establish the existence of abstract mathematical objects through appealing to the role that mathematics plays in empirical science. The indispensability argument is standardly attributed to W. V. Quine and Hilary Putnam. In this paper, I show that this attribution is mistaken. Quine's argument for the existence of abstract mathematical objects differs from the argument which many philosophers of mathematics ascribe to him. Contrary to (...) appearances, Putnam did not argue for the existence of abstract mathematical objects at all. I close by suggesting that attention to Quine and Putnam's writings reveals some neglected arguments for platonism which may be superior to the indispensability argument. (shrink)

The most immediately appealing model for formal constraints on degrees of belief is provided by probability theory, which tells us, for instance, that the probability of P can never be greater than that of (P v Q). But while this model has much intuitive appeal, many have been concerned to provide arguments showing that ideally rational degrees of belief would conform to the calculus of probabilities. The arguments most frequently used to make this claim plausible are the so-called "Dutch Book" (...) arguments. (shrink)

Current controversies about knowledge integration reflect conflicting ideas of what it means to “take Indigenous knowledge seriously”. While there is increased interest in integrating Indigenous and Western scientific knowledge in various disciplines such as anthropology and ethnobiology, integration projects are often accused of recognizing Indigenous knowledge only insofar as it is useful for Western scientists. The aim of this article is to use tools from philosophy of science to develop a model of both successful integration and integration failures. On the (...) one hand, I argue that cross-cultural recognition of property clusters leads to an ontological overlap that makes knowledge integration often epistemically productive and socially useful. On the other hand, I argue that knowledge integration is limited by ontological divergence. Adequate models of Indigenous knowledge will therefore have to take integration failures seriously and I argue that integration efforts need to be complemented by a political notion of ontological self-determination. (shrink)

This essay focuses on three recent books on morality and virtue, Michael Slote's 'Morals from Motives', Rosalind Hursthouse's 'On Virtue Ethics', and Philippa Foot's 'Natural Goodness'. Slote proposes an "agent-based" ethical theory according to which the ethical status of acts is derivative from assessments of virtue. Following Foot's lead, Hursthouse aims to vindicate an ethical naturalism that explains human goodness on the basis of views about human nature. Both Hursthouse and Slote take virtue to be morally basic in a way (...) that Foot, to her credit, does not. We argue that all three views face a range of serious difficulties. (shrink)

The aim of this article is to show that externalist accounts of cognition such as Clark and Chalmers' (1998) “active externalism” lead to an explosion of knowledge that is caused by online resources such as Wikipedia and Google. I argue that externalist accounts of cognition imply that subjects who integrate mobile Internet access in their cognitive routines have millions of standing beliefs on unexpected issues such as the birth dates of Moroccan politicians or the geographical coordinates of villages in southern (...) Indonesia. Although many externalists propose criteria for the bounds of cognition that are designed to avoid this explosion of knowledge, I argue that these criteria are flawed and that active externalism has to accept that information resources such as Wikipedia and Google constitute extended cognitive processes. (shrink)

The situationist literature in psychology claims that conduct is not determined by character and reflects the operation of the agent's situation or environment. For instance, due to situational factors, compassionate behavior is much less common than we might have expected from people we believe to be compassionate. This article focuses on whether situationism should revise our beliefs about moral responsibility. It assesses the implications of situationism against the backdrop of a conception of responsibility that is grounded in norms about the (...) fair opportunity to avoid wrongdoing that require that agents to be normatively competent and possess situational control. Despite the low incidence of compassionate behavior revealed in situationist studies, situationism threatens neither situational control nor normative competence. Nonetheless situationism may force revision of our views about responsibility in particular contexts, such as wartime wrongdoing. Whereas a good case can be made that the heat of battle can create situational pressures that significantly impair normative competence and thus sometimes provide a full or partial excuse, there is reason to be skeptical of attempts to generalize this excuse to other contexts of wartime wrongdoing. If so, moral responsibility can take situationism on board without capsizing the boat. (shrink)

Much contemporary epistemology is informed by a kind of confirmational holism, and a consequent rejection of the assumption that all confirmation rests on experiential certainties. Another prominent theme is that belief comes in degrees, and that rationality requires apportioning one's degrees of belief reasonably. Bayesian confirmation models based on Jeffrey Conditionalization attempt to bring together these two appealing strands. I argue, however, that these models cannot account for a certain aspect of confirmation that would be accounted for in any adequate (...) holistic confirmation theory. I then survey the prospects for constructing a formal epistemology that better accommodates holistic insights. (shrink)

This paper defends the view that perceptual ascriptions such as “Jones sees a cat” are sometimes intensional. I offer a range of examples of intensional perceptual ascriptions, respond to objections to intensional readings of perceptual ascriptions, and show how widely accepted semantic accounts of intensionality can explain the key features of intensional perceptual ascriptions.

A number of philosophers, from Thomas Reid1 through C. A. J. Coady2, have argued that one is justified in relying on the testimony of others, and furthermore, that this should be taken as a basic epistemic presumption. If such a general presumption were not ultimately dependent on evidence for the reliability of other people, the ground for this presumption would be a priori. Such a presumption would then have a status like that which Roderick Chisholm claims for the epistemic principle (...) that we are justified in believing what our senses tell us. (shrink)

Platonism in the philosophy of mathematics is the doctrine that there are mathematical objects such as numbers. John Burgess and Gideon Rosen have argued that that there is no good epistemological argument against platonism. They propose a dilemma, claiming that epistemological arguments against platonism either rely on a dubious epistemology, or resemble a dubious sceptical argument concerning perceptual knowledge. Against Burgess and Rosen, I show that an epistemological anti- platonist argument proposed by Hartry Field avoids both horns of their dilemma.

Can we adequately account for our reasons of mere taste without holding that our desires ground such reasons? Recently, Scanlon and Parfit have argued that we can, pointing to pleasure and pain as the grounds of such reasons. In this paper I take issue with each of their accounts. I conclude that we do not yet have a plausible rival to a desire-based understanding of the grounds of such reasons.

Truthmaker theory promises to do some useful philosophical work: equipping us to argue against phenomenalism and Rylean behaviourism, for instance, and helping us decide what exists (Lewis 1999, 207; Armstrong 1997, 113-119). But it has proved hard to formulate a truthmaker theory that is both useful and believable. I want to suggest that a neglected approach to truthmakers – that of Ian McFetridge – can surmount some of the problems that make other theories of truthmaking unattractive. To begin with, I’ll (...) outline some of the most prominent accounts of truthmaking in the current literature. Then the second part of the paper will explain McFetridge’s theory and argue for its superiority over these accounts. (shrink)

This article develops an account of local epistemic practices on the basis of case studies from ethnobiology. I argue that current debates about objectivity often stand in the way of a more adequate understanding of local knowledge and ethnobiological practices in general. While local knowledge about the biological world often meets criteria for objectivity in philosophy of science, general debates about the objectivity of local knowledge can also obscure their unique epistemic features. In modification of Ian Hacking’s suggestion to discuss (...) “ground level questions” instead of objectivity, I propose an account that focuses on both epistemic virtues and vices of local epistemic practices. (shrink)

Both Representation Theorem Arguments and Dutch Book Arguments support taking probabilistic coherence as an epistemic norm. Both depend on connecting beliefs to preferences, which are not clearly within the epistemic domain. Moreover, these connections are standardly grounded in questionable definitional/metaphysical claims. The paper argues that these definitional/metaphysical claims are insupportable. It offers a way of reconceiving Representation Theorem arguments which avoids the untenable premises. It then develops a parallel approach to Dutch Book Arguments, and compares the results. In each case (...) preferencedefects serve as a diagnostic tool, indicating purely epistemic defects. (shrink)

The authors argue against direction of fit accounts of the distinction between belief and desire.

In this paper, we present three necessary conditions for morally responsible animal research that we believe people on both sides of this debate can accept. Specifically, we argue that, even if human beings have higher moral status than nonhuman animals, animal research is morally permissible only if it satisfies (a) an expectation of sufficient net benefit, (b) a worthwhile-life condition, and (c) a no unnecessary-harm/qualified-basic-needs condition. We then claim that, whether or not these necessary conditions are jointly sufficient conditions of (...) justified animal research, they are relatively demanding with the consequence that many animal experiments may fail to satisfy them. (shrink)

Presentists, who believe that only present objects exist, should accept a thisness ontology, since it can do considerable work in defence of presentism. In this paper, I propose a version of presentism that involves thisnesses of past and present entities and I argue this view solves important problems facing standard versions of presentism.

This essay examines the philosophical significance of $\Omega$-logic in Zermelo-Fraenkel set theory with choice (ZFC). The categorical duality between coalgebra and algebra permits Boolean-valued algebraic models of ZFC to be interpreted as coalgebras. The hyperintensional profile of $\Omega$-logical validity can then be countenanced within a coalgebraic logic. I argue that the philosophical significance of the foregoing is two-fold. First, because the epistemic and modal and hyperintensional profiles of $\Omega$-logical validity correspond to those of second-order logical consequence, $\Omega$-logical validity is genuinely (...) logical. Second, the foregoing provides a hyperintensional account of the interpretation of mathematical and metamathematical vocabulary. (shrink)

It is obvious that we would not want to demand that an agent' s beliefs at different times exhibit the same sort of consistency that we demand from an agent' s simultaneous beliefs; there' s nothing irrational about believing P at one time and not-P at another. Nevertheless, many have thought that some sort of coherence or stability of beliefs over time is an important component of epistemic rationality.

Glymour's "bootstrap" account of confirmation is designed to provide an analysis of evidential relevance, which has been a serious problem for hypothetico-deductivism. As set out in Theory and Evidence, however, the "bootstrap" condition allows confirmation in clear cases of evidential irrelevance. The difficulties with Glymour's account seem to be due to a basic feature which it shares with hypothetico-deductive accounts, and which may explain why neither can give a satisfactory analysis of evidential relevance.

A given causal system can be represented in a variety of ways. How do agents determine which variables to include in their causal representations, and at what level of granularity? Using techniques from Bayesian networks, information theory, and decision theory, we develop a formal theory according to which causal representations reflect a trade-off between compression and informativeness, where the optimal trade-off depends on the decision-theoretic value of information for a given agent in a given context. This theory predicts that, all (...) else being equal, agents prefer causal models that are as compressed as possible. When compression is associated with information loss, however, all else is not equal, and our theory predicts that agents will favor compressed models only when the information they sacrifice is not informative with respect to the agent’s anticipated decisions. We then show, across six studies reported here (N=2,364) and one study reported in the supplemental materials (N=182), that participants’ preferences over causal models are in keeping with the predictions of our theory. Our theory offers a unification of different dimensions of causal evaluation identified within the philosophy of science (proportionality and stability), and contributes to a more general picture of human cognition according to which the capacity to create compressed (causal) representations plays a central role. (shrink)

I discuss Gheaus's (2013) argument against the claim that the requirements of justice are not constrained by feasibility concerns. I show that the general strategy exemplified by this argument is not only dialectically puzzling, but also imposes a heavy cost on theories of justice -- puzzling because it simply sidesteps a presupposition of any plausible formulation of the so-called "feasibility requirement"; costly because it it deprives justice of its normative implications for action. I also show that Gheaus's attempt to recover (...) this normative force presupposes an epistemic dimension to the feasibility requirement that most proponents of that requirement would reject. (shrink)

For a stable visual world, the colours of objects should appear the same under different lights. This property of colour constancy has been assumed to be fundamental to vision, and many experimental attempts have been made to quantify it. I contend here, however, that the usual methods of measurement are either too coarse or concentrate not on colour constancy itself, but on other, complementary aspects of scene perception. Whether colour constancy exists other than in nominal terms remains unclear.

One way in which philosophers have often sought to distinguish moral judgments from non-moral judgments is by using the “moral-conventional” distinction. I seek to raise serious questions about the significance of the moral-conventional distinction, at least for philosophers interested in moral judgment. I survey recent developments in the fields of philosophy, psychology, and cognitive science that have led many to the conclusion that moral judgment is not a distinctive kind of judgment or the result of a specific, identifiable cognitive process. (...) I argue that if this conclusion is largely correct, the moral-conventional distinction loses significance. If moral judgment does not correspond to a distinctive cognitive process, it is unclear how distinguishing between types of norms tracks anything of significance to human judgment formation. I then discuss the implications of abandoning the distinction for research in the field of moral psychology and tentatively propose a more modest way of conceiving of norm significance. (shrink)

Which premisses should we use to start our inquiries? Which transitions during inquiry should we take next? When should we switch lines of inquiry? In this paper, I address these open questions about inquiry, formulating novel norms for such decisions during deductive reasoning. I use the first-order predicate calculus, in combination with Carnap’s state description framework, to state such norms. Using that framework, I first demonstrate some properties of sets of sentences used in deduction. I then state some norms for (...) decisions made during deductive reasoning, establishing initial benchmarks for efficient deduction by ideal reasoners. When deciding which transition to make next, reasoners should choose the most informative transition, the one that maximally reduces uncertainty in the sense of ruling out the largest number of state descriptions relevant to their inquiry. Finally, inspired by optimal foraging theory, I show that, under certain assumptions of ignorance, reasoners should change premiss sets when their information intake drops below the global average information intake across premiss sets. (shrink)

Compare your pain when immersing your hand in freezing water and your pleasure when you taste your favourite wine. The relationship seems obvious. Your pain experience is unpleasant, aversive, negative, and bad. Your experience of the wine is pleasant, attractive, positive, and good. Pain and pleasure are straightforwardly opposites. Or that, at any rate, can seem beyond doubt, and to leave little more to be said. But, in fact, it is not beyond doubt. And, true or false, it leaves a (...) good deal more to be said: about the nature of sensory affect; its relations to perception, motivation, and rationality; its value; and the mechanisms underlying it. Much is said about these matters in the contributions that follow. Here, in this introductory essay, we map the dialectical landscape and locate our contributors’ papers within it. (shrink)

If the reliability of a source of testimony is open to question, it seems epistemically illegitimate to verify the source’s reliability by appealing to that source’s own testimony. Is this because it is illegitimate to trust a questionable source’s testimony on any matter whatsoever? Or is there a distinctive problem with appealing to the source’s testimony on the matter of that source’s own reliability? After distinguishing between two kinds of epistemically illegitimate circularity—bootstrapping and self-verification—I argue for a qualified version of (...) the claim that there is nothing especially illegitimate about using a questionable source to evaluate its own reliability. Instead, it is illegitimate to appeal to a questionable source’s testimony on any matter whatsoever, with the matter of the source’s own reliability serving only as a special case. (shrink)

According to a popular idea, freedom is grounded in orthonomy – the ability to be responsive to normative demands. But how exactly must an agent’s action relate to their reasons in order for this orthonomous relationship to hold? In this paper, I propose a novel explanationist answer to this question. I argue that extant answers – causalism and modalism about orthonomy – fail because they fail to account for the fact that intuitions about freedom and orthonomy track facts about explanation. (...) And these facts cannot be reduced to either causal or modal facts. Instead, I propose that we should understand freedom as grounded in special accidentality-dispersing explanations in terms of normative reasons. The resultant view is to be understood as following a Frankfurtian intuition: only factors that explain action can be relevant to whether the action is free. (shrink)