A discussion of epistemic reasons, theoretical rationality, and the relationship between them. Discusses the ontology of reasons and evidence, the relationship between reasons (motivating, normative, possessed, apparent, genuine, etc.) and rationality, the relationship between epistemic reasons and evidence, the relationship between rationality, justification, and knowledge, and many other related topics.

Philosophers have long distinguished between practical and theoretical rationality. The first section of this chapter begins by discussing the ways in which this distinction was drawn by Aristotle and Kant; then it sketches what seems to be the general consensus today about how, at least roughly, the distinction should be drawn. The rest of this chapter explores what practical and theoretical rationality have in common: in the second section, several parallels between practical and theoretical (...) class='Hi'>rationality are outlined, and it is argued that these parallels make it plausible that a unifying account of rationality can be given. Finally, in the third section, a number of such unifying accounts of practical and theoretical rationality are surveyed. These include accounts that are inspired in various ways by Hume and by the results of formal decision theory, as well as views that appeal to reasons and to the distinctive value of correct or appropriate attitudes. (shrink)

A common view in philosophy is that the way human beings reason is not only gradually better, but that our way of reasoning is fundamentally distinctive. Findings in the psychology of reasoning challenge the traditional view according to which human beings reason in accordance with the laws of logic and probability theory, but rather suggest that human reasoning consists in the application of domain specific rules of thumb similar to those that we ascribe to some intelligent non-human animals as well. (...) However, this view on human reasoning is unable to explain human accomplishments like technological innovations or scientific progress. David Papineau offers a theory of human theoretical rationality that is consistent with the psychological view on human reasoning but that can also explain how humans sometimes are able to transcend the limitations of their biologically quick and dirty modes of thought and thereby reach a high level of accuracy. Papineau claims that the abilities that constitute theoretical rationality are unique to the human species and thus, that human reasoning is fundamentally distinctive after all. In this paper I am going to discuss to what extent these abilities in fact are unique to our species and whether this theoretical rationality can be called an anthropological difference. (shrink)

An important advance in normativity research over the last decade is an increased understanding of the distinction, and difference, between normativity and rationality. Normativity concerns or picks out a broad set of concepts that have in common that they are, put loosely, guiding. For example, consider two commonly used normative concepts: that of a normative reason and that of ought. To have a normative reason to perform some action is for there to be something that counts in favour of (...) performing that action. Likewise with ought, when there is sufficient evidence for something, one ought to believe it (at least under normal circumstances). Not all guidance need be directed towards a specific state or a specific action. Subject to the requirements of normativity, too, are relations. It is commonly believed, for example, that we ought not to hold contradictory beliefs.1 At least some of the requirements that concern relations amongst an agent’s mental states are, or seem, distinctive. Agents who fail to satisfy these requirements are considered, at least to some degree, irrational. On many current views, being irrational is distinct in some way from not being how one ought to be; rationality is a concept distinct from normativity. Much of the literature on this topic over the last decade stems from attempts to capture the characteristic features of the requirements of rationality. Two influential views in particular did much to set the agenda. The first of these two was put forward John Broome.2 His view, the particulars of which I shall discuss in more detail below, is that the requirements of rationality could be expressed using a normative relation, which he calls a ‘normative requirement’. Normative requirements are conditionals governed by an all-thingsconsidered ought. In the case of rationality, the conditional is made up entirely of mental states.. (shrink)

The focus of this essay is Kant’s argument in the Groundwork for the Metaphysics of Morals (GMS) III that regarding oneself as rational implies regarding oneself as free. After setting out an interpretation of how the argument is meant to go (§§1-2), I argue that Kant fails to show that regarding oneself as free is incompatible with accepting universal causal determinism (§3). However, I argue that the argument succeeds in showing that regarding oneself as rational is inconsistent with accepting universal (...) causal determinism if one accepts a certain, plausible view of the explanation of events (§4). (shrink)

Given the success of the formal approach, within contemporary epistemology, to understanding degreed belief, some philosophers have recently considered its extension to the challenge of understanding intention. According to them, (1) intentions can also admit of degrees, as beliefs do, and (2) these degreed states are all governed by the norms of the probability calculus, such that the rational norms for belief and for intention exhibit certain structural similarity. This paper, however, raises some worries about (2). It considers two schemes (...) for representing degreed intention, and casts some doubt on applying probabilistic norms to degreed intention on each scheme. And it argues that the norms of Intention–Belief Consistency and Enkrasia cannot plausibly be turned into norms prescribing a simple correspondence relation between one’s degrees of belief and intention. The results suggest a potential structural discrepancy between the norms of theoretical and practical rationality—at least for degreed belief and intention. This potential discrepancy not only raises interesting questions about the intention–belief relation and rational norms for degreed intention. It also poses a new challenge for intention-based expressivism about normative judgment and cognitivism about practical rationality. (shrink)

We present an axiomatic approach for a class of finite, extensive form games of perfect information that makes use of notions like “rationality at a node” and “knowledge at a node.” We distinguish between the game theorist's and the players' own “theory of the game.” The latter is a theory that is sufficient for each player to infer a certain sequence of moves, whereas the former is intended as a justification of such a sequence of moves. While in general (...) the game theorist's theory of the game is not and need not be axiomatized, the players' theory must be an axiomatic one, since we model players as analogous to automatic theorem provers that play the game by inferring (or computing) a sequence of moves. We provide the players with an axiomatic theory sufficient to infer a solution for the game (in our case, the backwards induction equilibrium), and prove its consistency. We then inquire what happens when the theory of the game is augmented with information that a move outside the inferred solution has occurred. We show that a theory that is sufficient for the players to infer a solution and still remains consistent in the face of deviations must be modular. By this we mean that players have distributed knowledge of it. Finally, we show that whenever the theory of the game is group-knowledge (or common knowledge) among the players (i.e., it is the same at each node), a deviation from the solution gives rise to inconsistencies and therefore forces a revision of the theory at later nodes. On the contrary, whenever a theory of the game is modular, a deviation from equilibrium play does not induce a revision of the theory. (shrink)

Rationality postulates for preferences are developed from two basic decision theoretic principles, namely: (1) the logic of preference is determined by paradigmatic cases in which preferences are choice-guiding, and (2) excessive comparison costs should be avoided. It is shown how the logical requirements on preferences depend on the structure of comparison costs. The preference postulates necessary for choice guidance in a single decision problem are much weaker than completeness and transitivity. Stronger postulates, such as completeness and transitivity, can be (...) derived under the further assumption that the original preference relation should also be capable of guiding choice after any restriction of the original set of alternatives. (shrink)

The conceptual foundations, features, and scope of the notion of rationality are increasingly being affected by developments in embodied cognitive science. This article starts from the idea of embodied rationality, and aims to develop a frame in which a debate with the classical, possibly bounded, notion of rationality-as-consistency can take place. To this end, I develop a game theoretic description of a real time interaction setup in which participants' behaviors can be used to compare the enactive approach, (...) which underlies embodied rationality, with game theoretic approaches to human interaction. The Perceptual Crossing Paradigm is a minimal interaction interface where two participants each control an avatar on a shared virtual line, and are tasked with cooperatively finding each other among distractor objects. It is well known that the best performance on this task is obtained when both participants let their movements coordinate with the objects they encounter, which they do without any prior knowledge of efficient interaction strategies in the system. A game theoretic model of this paradigm shows that this task can be described as an Assurance game, which allows for comparing game theoretical approaches and the enactive approach on two main fronts. First, accounting for the ability of participants to interactively solve the Assurance game; second, accounting for the evolution of choice landscapes resulting from evolving normative realms in the task. Similarly to the series of paradoxes which have fueled debates in economics in the past century, this analysis aims to serve as an interpretation testbed which can fuel the current debate on rationality. (shrink)

This paper concerns the rational reconstruction of physical theories initially advanced by F. P. Ramsey and later elaborated by Rudolf Carnap. The Carnap–Ramsey reconstruction of theoretical knowledge is a natural development of classical empiricist ideas, one that is informed by Russell's philosophical logic and his theories of propositional understanding and knowledge of matter ; as such, it is not merely a schematic representation of the notion of an empirical theory, but the backbone of a general account of our knowledge (...) of the physical world. Carnap–Ramsey is an illuminating approach to epistemological problems that remain with us, one whose difficulties are shared by accounts that have sought to replace it. 1 Introduction 2 Russell's theory of propositional understanding 3 Ramsey's primary and secondary systems 4 Carnap's reconstruction of the language of science and an observation of Newman 5 Extension of the foregoing to constructive empiricism 6 Putnam's model-theoretic argument and the semantic view of theories 7 The problem clarified and resolved. (shrink)

Rationality and reasonableness can be illustrated as Janus-faced concepts, not only in a descriptive diagnosis but also in a normative construction of adjudication, and in the analysis of its practical and rhetorical effects. Considering such an illustration, the present reflection returns to the discussion on the relevance of rationality and reasonableness in legal interpretation, aiming at distinguishing and/or connecting principles and criteria, beyond formally logical and/or procedurally argumentative decision-making, and, thus, within a normatively practical adjudication. Such an approach (...) will be presented in two main steps. In a first step, the rationality/reasonableness dichotomy will be discussed within a constitutive connection between principles and criteria in adjudication, and understanding legal principles as normatively substantial foundations to legal criteria—the former as axiological references filtered into the juridical system and the latter as their concretizations, and whose practical adequacy is rhetorically assigned by the interpretive communities involved in/by adjudication, and normatively referred both to their practical accordance to the axiological normativity of the principles they shall presuppose and to their suitability as practical operators to the problematic relevance of the sub judice cases. In a second step, and consequently, encompassing the projection of legal interpretation, in semiotic and in practical-normative perspectives, in order to discuss their constituting practical relevance in adjudication, the rationality/reasonableness dichotomy will be experimented within formal-logical procedures and within practical-dialectical connections between juridical problem and juridical system—therefore stating a discussion on the foundations, the intentionality and the consequences of adjudication, amongst deduction and analogy. (shrink)

The impossibility results in judgement aggregation show a clash between fair aggregation procedures and rational collective outcomes. In this paper, we are interested in analysing the notion of rational outcome by proposing a proof-theoretical understanding of collective rationality. In particular, we use the analysis of proofs and inferences provided by linear logic in order to define a fine-grained notion of group reasoning that allows for studying collective rationality with respect to a number of logics. We analyse the (...) well-known paradoxes in judgement aggregation and we pinpoint the reasoning steps that trigger the inconsistencies. Moreover, we extend the map of possibility and impossibility results in judgement aggregation by discussing the case of substructural logics. In particular, we show that there exist fragments of linear logic for which general possibility results can be obtained. (shrink)

Certain realist critics of U.S. foreign policy put forth an alternative model of “offshore balancing” as a definitively rational alternative to what they regard as the current, and utterly disastrous, policy of “liberal hegemony.” They predict that the public will eventually recognize the hollowness of liberalism and demand a foreign policy rooted in hardnosed realism. They also promise that this rational outline will also be a positive good, maximizing national interests and moral values with no tradeoffs between them. I argue (...) that offshore balancing packages realist arguments in an idealist framework, whereby the good is both self-evident and bound to triumph. This inhibits the actual realist task of revealing the harsh facts of politics, which inevitably interfere with preferences. Offshore balancing traps itself between its claim of inevitability, which undercuts the need for advocacy, and desirability, which pushes against its inevitability. The only way they can resolve this contradiction is with an increasingly caustic account of liberalism as so utterly wrong and immoral that its opponents earn the mark of reason and goodness by sheer virtue of their opposition. Offshore balancing subordinates its account of the world as it is to a demand of how the world must not be. (shrink)

Whereas Brian Skyrms in his chapter views rationality and evolution as alternative ways to derive decision behaviour, indirect evolution allows us to combine the two approaches. By focussing on Skyrms’ examples it will be illustrated how optimal decisions for given rules of interaction can influence the future rules of interaction. Here evolution does not determine behaviour directly, but only indirectly via the rules. We, furthermore, report on an experiment, related to Skyrms’ examples, revealing effects of deliberation in the sense (...) of forward-looking considerations and of path dependence in the sense of adapting to past success. (shrink)

Philosophy and Phenomenological Research, EarlyView.

Nonmonotonic reasoning is a discipline of computer science, epistemology, and cognition: It models inferences where classical logic is inadequate in symbolic AI, defines normative models for reasoning with defeasible information in epistemology, and models human reasoning under information change in cognition. Its building blocks are defeasible rules formalised as DeFinetti conditionals. In this thesis, Christian Eichhorn examines qualitative and semi-quantitative inference relations on top said conditionals, using the conditional structure of the knowledge base and Spohn’s Ordinal Conditional Functions, using established (...) properties. Converting network approaches from probabilistics, he shows how to approach the relations with regard to implementation. (shrink)

Two principles are central to Rational Causation. Causalism: Believing and acting for a reason are causal phenomena in the sense that there is in both domains a causal connection between ground and grounded. Equivalence: There is a necessary connection between something's being the reason why I believe or act and my taking it to favour the belief or action. Kieran Setiya argues that Causalism is false in the theoretical case and that Equivalence is false in the practical case. I (...) reply to these and other arguments. (shrink)

Philosophy and Phenomenological Research, EarlyView.

Much of cognitive science seeks to provide principled descriptions of various kinds and aspects of rational behaviour, especially in beings like us or AI simulacra of beings like us. For the most part, these investigators presuppose an unarticulated common sense appreciation of the rationality that such behaviour consists in. On those occasions when they undertake to bring the relevant norms to the surface and to give an account of that to which they owe their legitimacy, these investigators tend to (...) favour one or other of three approaches to the normativity question. They are the analyticity or truth-in-a-model approach; the pluralism approach; and the reffective equilibrium approach.All three of these approaches to the normativity question are seriously flawed, never mind that the first two have some substantial provenance among logicians and the third has enjoyed a flourishing philosophical career.Against these views, we propose a strong version of what might be called normatively immanent descriptivism. We attempt to elucidate its virtues and to deal with what appears to be its most central vulnerability, embodied in the plain fact that actual human behaviour is sometimes irrational. (shrink)

The subject of my dissertation is "rationality". In this book I undertake a comprehensive, systematic and independent treatment of the problem of rationality. This furthers progress toward a general theory of rationality, one that represents and defends a uniform conception of reason. The structure and general outline are as follows: Part I: General Definition of the Concept; Part II: Rationality in the Theoretical Realm; Part III: Rationality in the Practical Realm (parts II and III (...) are divided respectively into A. Relative Reasons and B. Absolute Reasons); Part IV: Grounds for the Principle of Reason. My aim in this work is to prove that a conception of rationality as justification of our usage of language is more significant than has been classically recognized, without, however, reducing the concept of rationality. After a definition of the term is provided in the first chapter by means of language analysis, the following three questions are dealt with in the subsequent chapters: a) Exhaustion of rationality in relative reasons—i.e., is reasoning exhausted by reference to existing views and goals, or does rationality also exist in a strong sense? The latter would be the case if the views and goals referred to the relative conception must themselves be justified de novo. I argue that such an absolute justification of views and goals cannot be provided. To this end, in the theoretical domain, I discuss a question prominent in the debate over the rationality of worldviews, namely the question of whether the standards of theoretical rationality are merely culturally relative (chapter V.). Here I arrive at the position that standards of theoretical rationality are only transcultural insofar as there are goals which are not themselves culturally-dependent. With regard to practically fundamental goals, the core standards of modern science can be justified. This justification strategy therefore requires that such goals be established. In practical terms, first I address (in Chapter VII) the so-called "final justification" of norms problem. Here I discuss above all the transcendental-pragmatic proposal and the contractualist conception. According to my view, neither succeed in providing a final rational grounding for norms. That grounding pertains instead to the rationale for objectives, and Chapter VIII demonstrates that even a final justification of objectives is impossible. This does not mean, however, that a person's goals cannot be criticized as irrational. But the warrant in this case depends on the rules that govern the genesis of desires being accepted by the subject. b) If substantive rationality does not exist because no absolute grounding for opinions and goals can obtain, then the question of how strong or weak the concept of rationality is acquires further weight. On that basis, relative theoretical and practical rationality is examined in detail. I aim with this analysis of (relative) rationality to apprehend the structural parallelism between rationality in both the practical and the theoretical domain, and therefore to defend a uniform conception of reason or rationality. This is accomplished by identifying a concept that exhibits the same structure in both theoretical and practical domains: the term "rational" is applied to actions and opinions in order to claim, first of all, that they are "well-founded." For example, one must inquire into the precise rational justification for opinions or actions. To determine the individually necessary and jointly sufficient conditions required for an opinion or action to be rational, one must first provide a formal characterization of the rationality of those opinions or actions. What is to count as a rational ground can only be explained by formal procedures and results achievable via those procedures. To be more specific, I claim that the process of opinion formation and practical reflection is subject to rational justification based on internalist rules. The rules governing theoretical rationality can have their content transformed in a further move, if one takes the result of the process of rational opinion-formation as the product of an epistemological determination, that is, as a decision about which opinion to hold that is made on the basis of epistemological or prudential goals. Implicit in this position is the controversial idea that one can actually decide to believe something. The rules of practical reason are the rules that guide the rational selection of objectives and actions. The structural parallelism of theoretical and practical rationality consists, therefore, in the fact that both refer to mental processes that are controlled by rules which one freely accepts and follows. In both cases, these rules are to be understood as governing the rational choice of a mental or physical action. c) Therefore, the question of whether rationality—being rational—can in turn be grounded by reasons can be answered in the following way: The rules of rationality can only be justified with reference to a person’s goals, as the best strategy for achieving those goals. If someone does not desire the optimal realization of his goals, there is no other way of arguing directly for rationality. One can only attempt to demonstrate—if the person is open to argument—that it is typically in the person's own best interest to adhere to the rules of rationality. (shrink)

Philosophy and Phenomenological Research, EarlyView.

David Gauthier thinks agents facing a prisoner's dilemma ('pd') should find it rational to dispose themselves to co-operate with those inclined to reciprocate (i.e., to acquire a constrained maximizer--'cm'--disposition), and to co-operate with other 'cmers'. Richmond Campbell argues that since dominance reasoning shows it remains to the agent's advantage to defect, his co-operation is only rational if cm "determines" him to co-operate, forcing him not to cheat. I argue that if cm "forces" the agent to co-operate, he is not acting (...) at all, never mind rationally. Thus, neither author has shown that co-operation is rational action in a pd. (shrink)

This chapter addresses an ambiguity in some of the literature on rational peer disagreement about the use of the term 'rational'. In the literature 'rational' is used to describe a variety of normative statuses related to reasons, justification, and reasoning. This chapter focuses most closely on the upshot of peer disagreement for what is rationally required of parties to a peer disagreement. This follows recent work in theoretical reason which treats rationality as a system of requirements among an (...) agent's mental states. It is argued that peer disagreement has either no, or a very limited, affect on what rationality requires of an agent in a given circumstance. This is in part because of difficulties generated by a novel example of evidence of evidence of p being evidence against p. This example calls into question the mechanisms whereby peer disagreement might affect what is rationally required of an agent. The chapter also reevaluates the importance of actual peer disagreement against the backdrop of prior expectations about whether disagreement is believed to be likely, arguing that peer disagreement is most likely to change what is rationally required of an agent when it is believed to be unlikely. (shrink)

A method of linguistically-oriented reasoning comprehension is proposed. It is based on semiological principles of text comprehension. Both content and form are essential for comprehending argumentative texts. A text recipient is viewed as a rational judge trying to detect all the components of the argument he/she considers and thus to see if the argument is consistent. Elementary and higher level argumentative units of the text are discovered by applying a modified S. Toulmin's model of argumentative functions. Validity and correctness of (...) arguments are established by means of a new – linguistic – interpretation of traditional syllogistic. (shrink)

Meyer and Mortensen’s Alien Intruder Theorem includes the extraor- dinary observation that the rationals can be extended to a model of the relevant arithmetic R♯, thereby serving as integers themselves. Al- though the mysteriousness of this observation is acknowledged, little is done to explain why such rationals-as-integers exist or how they operate. In this paper, we show that Meyer and Mortensen’s models can be identified with a class of ultraproducts of finite models of R♯, providing insights into some of the (...) more mysterious phenomena of the rational models. (shrink)

This article brings interconnects three debates to show what this might imply for the ‘redemocratisation’ of UK society and for pedagogical reform. One debate concerns deliberative types of democratic reform, arguing in favour of a ‘creative agnosticism’ towards the two philosophical frameworks which dominate this literature. This leads into a discussion of education and critical rationality, arguing for an aptitude-based account of moral agency, one which relates to the sociocultural resources we inherit from the past. The final debate therefore (...) concerns social memory. I do not offer a theory of social memory but critique various commentators in arguing for what I call a ‘multipolar analysis’. This both describes the analytical method adopted in this article and the skills and capacities which a system based upon ‘education for discourse’ would need to facilitate. The article therefore proposes that democracy, education and memory—deliberativeness, critical rationality and social remembering—be brought closer together for the mutual benefit of each debate. (shrink)

The question of the rationality of conspiratorial belief ¬divides philosophers into mainly two camps. The particularists believe that each conspiracy theory ought to be examined on its own merits. The generalist, by contrast, argues that there is something inherently suspect about conspiracy theories that makes belief in them irrational. Recent empirical findings indicate that conspiratorial thinking is commonplace among ordinary people, which has naturally shifted attention to the particularists. Yet, even the particularist must agree that not all conspiracy belief (...) is rational, in which case she must explain what separates rational from non-rational conspiratorial thinking. In this paper, I contrast three strategies to this end: 1) the probabilistic objectivist, who assesses the objective probability of conspiracies; 2) the subjectivist, who rather focuses on the perspective of the believer, and typically views the decision to believe in a conspiracy as a problem of decision making under risk. Approaches 1) and 2) rely on assessments of the probability of conspiracy which, I argue, limits their applicability. Instead, I explore 3) viewing the problem facing the potential believer as a decision problem under uncertainty about probabilities. I argue, furthermore, that focusing solely on epistemic utilities fails to do justice to the particular character of conspiracy beliefs, which are not exclusively epistemically motivated, and I investigate the rationality of such beliefs under a number of standard decision rules. (shrink)

NOTE: this is a substantial revision of a previously uploaded draft. Intuitions are often thought of as inputs to theoretical reasoning. For example, you might form a belief by taking an intuition at face value, or you might take your intuitions as starting points in the method of reflective equilibrium. The aim of this paper is to argue that in addition to these roles intuitions also play action-guiding roles. The argument proceeds by reflection on the transmission of justification through (...) inference. According to inferential internalists, in order to gain justification for believing the conclusion of an argument by inferring it from the premises in that argument one must “see” that the premises support the conclusion. I motivate this view and endorse the idea that one’s “seeing” such a support relation consists of one’s having an intuition. In a number of recent papers, Paul Boghossian has pressed a regress argument against inferential internalism inspired by Lewis Carroll’s dialogue “What the Tortoise Said to Achilles.” I develop a response to Bogossian’s argument according to which intuitions work like mental imperatives and inferences are mental actions performed by obeying them. After developing this response to Boghossian’s argument, I take up the question of what it is in virtue of which intuitions play a guidance role, when they do so. (shrink)

Decision theory elucidates, in more ways than one, the, concept of rational behavior under imperfect knowledge of the consequences. On the other hand, the generally accepted concept of rationality refers to the end-means relation. This relation is not translatable into the language of decision theory. Consequently, the latter's claim to have defined in a general way rationality of behavior appears not to be valid.

Man is a rational animal, or so says Aristotle.1 But what is rationality? It is the use of reason to reach a certain level of reasonableness or unreasonableness. To be rational is to be extremely reasonable as to be irrational is to be extremely unreasonable. The varieties of rationality and distinctions about rationality are many. First of all, there is a distinction between practical and theoretical rationality. Secondly, there is a distinction between synchronic or static (...) rationality and diachronic or dynamic rationality. Thirdly, there is a distinction between personal rationality and interpersonal rationality. Combinations of these kinds of rationality yield eight kinds of rationality. (shrink)

The article discusses some of the fundamental conceptual and theoretical aspects of rational choice and moral order. A distinction is drawn between constitutional interests and compliance interests, and it is argued that a viable moral order requires that the two interests somehow be brought into congruence. It is shown that with regard to the prospects for a spontaneous emergence of such congruence, a distinction between two kinds of moral rules which we call trust-rules and solidarity-rules is of crucial importance.

In this paper, I explore the relationship between epistemic rationality and instrumental rationality, and I attempt to delineate their respective roles in typical instances of theoretical reasoning. My primary concern is with the instrumentalist conception of epistemic rationality: the view that epistemic rationality is simply a species of instrumental rationality, viz. instrumental rationality in the service of one's cognitive or epistemic goals. After sketching the relevance of the instrumentalist conception to debates over naturalism (...) and 'the ethics of belief', I argue that, despite enjoying considerable popularity among both epistemologists and philosophers of science, it is ultimately indefensible. Having thus argued for the distinctness of epistemic and instrumental rationality, I attempt to clarify the role played by each in typical instances of theoretical reasoning. I suggest that being theoretically rational--that is, being proficient with respect to theoretical reasoning--is best construed as a hybrid virtue, inasmuch as it involves manifesting sensitivity to two very different kinds of reasons. (shrink)

Predictable polarization is everywhere: we can often predict how people’s opinions, including our own, will shift over time. Extant theories either neglect the fact that we can predict our own polarization, or explain it through irrational mechanisms. They needn’t. Empirical studies suggest that polarization is predictable when evidence is ambiguous, that is, when the rational response is not obvious. I show how Bayesians should model such ambiguity and then prove that—assuming rational updates are those which obey the value of evidence—ambiguity (...) is necessary and sufficient for the rationality of predictable polarization. The main theoretical result is that there can be a series of such updates, each of which is individually expected to make you more accurate, but which together will predictably polarize you. Polarization results from asymmetric increases in accuracy. This mechanism is not only theoretically possible, but empirically plausible. I argue that cognitive search—searching a cognitively accessible space for a particular item—often yields asymmetrically ambiguous evidence, I present an experiment supporting its polarizing effects, and I use simulations to show how it can explain two of the core causes of polarization: confirmation bias and the group polarization effect. (shrink)

Suppose you think that whether you believe some proposition A at some future time t might have a causal influence on whether A is true. For instance, maybe you think a woman can read your mind, and either (1) you think she will snap her fingers shortly after t if and only if you believe at t that she will, or (2) you think she will snap her fingers shortly after t if and only if you don't believe at t (...) that she will. Let A be the proposition that she snaps her fingers shortly after t. In case (1), theoretical rationality seems to leave it open whether you should believe A or not. Perhaps, for all it has to say, you could just directly choose whether to believe A. David Velleman seems to be committed to something close to that, but his view has been unpopular. In case (2), you seem to be in a theoretical dilemma, a situation where any attitude you adopt toward A will be self-undermining in a way that makes you irrational. Such theoretical dilemmas ought to be impossible, just as genuine moral dilemmas ought to be impossible, but it is surprisingly hard to show that they are (perhaps because they aren't). I study cases analogous to (1) and (2) in a probabilistic framework where degrees of belief rather than all-or-nothing beliefs are taken as basic. My principal conclusions are that Velleman's view is closer to the truth than it is generally thought to be, that case (2) type theoretical dilemmas only arise for hyperidealised agents unlike ourselves, and that there are related cases that can arise for agents like us that are very disturbing but might not quite amount to theoretical dilemmas. (shrink)