It is well-acknowledged that a revival of interest in classically-based virtue theory has occurred in recent philosophy, beginning during or even before the 1970s with discussions among ethicists about the role of the ethical virtues in evaluating agents and their actions. The extension of this interest into epistemology dates to 1980, when the suggestion of an account of justified belief (justification) based upon epistemic or intellectual virtues was made by Ernest Sosa in his paper "The Raft and the Pyramid." As a first approximation, intellectual virtues in contemporary virtue epistemology are those "capacities," "powers," "dispositions" and "acquired habits" of persons that contribute to their success in achieving the cognitive aim or aims. That aim most epistemologists agree is truth, and so Sosa understands intellectual virtue as "a quality bound to help maximize one’s surplus of truth over error." This Preface will introduce the reader to the study of the intellectual virtues, and to the dialogue both among self-described virtue epistemologists, and between themselves and their critics. I will first provide a brief overview of the initial motivations behind the revival of virtue theory, and explicate some parallels between the approaches taken to philosophical problems in virtue ethics and in virtue epistemology. The second part of the Preface will survey the central epistemological problems that frame the research interests of virtue epistemology, and the major directions taken in contemporary research. The third part will acquaint the reader with each of the authors and exemplary papers that I have selected for this collection. Finally, the fourth part of the Preface briefly raises and answers some queries about the extended sense of the term "character" as it is used in the title of this collection. I distinguish various connotations of the term, and suggest several senses of "intellectual character" that I believe are appropriate to connect with the acknowledged central epistemic topics of knowledge and justified belief. This requires, however, articulating not only the commonalities, but also the differences between contemporary virtue theories and its classical predecessors.

It is widely agreed today that the role of the virtues had been under-emphasized in ethical theory in a period prior to this recent revival in virtue theory. Some would identify this period with the positivist era through mid-twentieth century in Anglo-American philosophy, while others would more broadly implicate the Modern era in philosophy as a whole. This difference is related to the oft-noted distinction between a "theory of the virtues" and a "virtue theory" (see Julia Driver, chapter 10). We can pick up this distinction in the third part of the Preface; we do not need it for the moment since the example of positivist metaethics is quite sufficient as a basis for indicating concerns that led to a revival of virtue theory. During the positivist era ethics had been dominated by concerns with the philosophical analysis of "thin" concepts like ethical "goodness"; the view of philosophic analysis that accompanied this approach sharply separated logic and philosophical analysis from matters of psychology, and thereby discouraged philosophers from doing substantive work in normative ethics. Even those ethicists who persisted in applying theory to practice and in evaluating the ethical rightness or wrongness of actions, typically did so by criteria that did not directly involve consideration of the character of moral agents themselves. With the revival of interest in virtue theory, the ethical virtues and vices came to be seen as having more important roles to play, roles that make them valuable both for understanding the motivations underlying an agent’s actions, and for evaluating agents as praiseworthy or blameworthy. In the period since the revival of virtue ethics, ethical evaluation has come to focus to a greater extent on situated agents themselves, and the habits and dispositions that constitute their moral character.

In his 1980 article, Sosa proposed a virtue-centered epistemology analogous to a form of virtue ethics. Soon thereafter, the intellectual virtues (and their pertinence to the epistemic evaluation of agents and their beliefs) was further developed in Sosa’s own work, and in that of several others, including Lorraine Code, Hilary Kornblith, and Alvin Goldman. Forms of contemporary virtue theory differ in substantial ways from their classical predecessors, as we shall see. But the recent inclusion of intellectual in addition to ethical virtues as a focus of study reintroduces the distinction and relationship between the kinds of virtue that Aristotle identified. This focus on the intellectual virtues and their role in epistemic evaluation also introduces a kind of parity into the approach to ethical and epistemic normativity, a parity that some see as present at least potentially in the classical outlook. That the fields of ethics and epistemology were sharply divided through much of the twentieth century, and that a great deal of the support positivism lent to that sharp division has been undermined by trends in the last quarter of the century, hardly seems debatable today. The plausibility and philosophical advantages of instilling greater parity through a strongly analogical approach to ethical and epistemic evaluation, however, certainly are a matter of debate, and part of the background that should concern us here.

Sosa’s original foray in the conclusion of his article did not get much beyond a proposal for a reorientation of epistemology. Since that time, he and others have developed intellectual virtue-centered epistemologies in substantial detail. His initial proposal and the reasoning behind it remains informative, though, especially when seen in the context of epistemological debate within which it is set. An agent’s actions are understood to be the result of her virtues or stable dispositions to act, and these dispositions are conceived in turn to embody rules. "Primary ethical justification," that is, the primary meaning of ascribing rightness or wrongness to actions, as Sosa understands it, attaches to (or derives from) virtues or stable dispositions to act, "through their greater contribution of value when compared with alternatives." An analogous strategy, he then suggests, might prove fruitful in epistemology: we might address the question of what confers justification upon an agent’s beliefs by a similar strategy. Following Sosa’s analogy, primary epistemic justification would attach "to intellectual virtues, to stable dispositions for belief acquisition, through their greater contribution towards getting us to the truth."

This places us in a position to identify a defining methodological feature setting virtue epistemology off from its alternatives. John Greco has described this key feature as a change in the direction of analysis. Justification, a property of belief, is explained in terms of the belief’s source in an intellectual virtue—that is, a property of persons. This parallels the way that virtue theories of ethics understand the normative properties of actions (rightness/wrongness) in terms of the normative properties of moral agents. Using the analogy between determining an act to be "right" and determining a belief to be "justified," one can understand this change in the direction of analysis by saying that virtue theories makes rightness (or justifiedness) follow from an action’s (or belief’s) source in a virtue, rather than the other way around. According to Greco, "Non-virtue theories try to analyze virtuous character in terms of justified belief, defining the former in terms of dispositions to achieve the latter. I am following Sosa’s suggestion that we do things the other way around, defining justified belief in terms of virtuous character. Virtuous character is then defined in terms of successful and stable dispositions to form belief." Another of our contributors, Christopher Hookway, also captures this reversal well when he writes, "Justified beliefs are those that issue from the responsible inquiries of virtuous inquirers. It is a mistake to put it the other way round: epistemic virtues are those habits and dispositions that lead us to have justified beliefs. The primary focus is on how we order activities directed at answering questions and assessing methods of answering questions; it is not upon the epistemic status of beliefs."

All forms of virtue epistemology embody this change in the direction of analysis that Greco and Hookway define. Virtue epistemologies typically define not only justified belief but also knowledge through essential reference to their source in intellectual virtues. But answers to the question of how, in turn, to define virtuous character have not enjoyed the same level of consensus. The claim Greco makes above (in step with Sosa’s proposal), that "virtuous character is then defined in terms of successful and stable dispositions to form belief" is a claim associated with a "reliabilist" virtue epistemology. Contrarily, Hookway’s final sentence, which suggests decentralizing questions of the epistemic status of beliefs in favor of questions of agency and inquiry, is a claim associated with a "responsibilist" virtue epistemology. Those authors oriented towards responsibilism typically do not agree with the "consequential" approach Sosa, Greco, Alvin Goldman and others take to defining the virtues—that is, by their reliable success in producing true belief.

The defining interests and concerns of responsibilists will come into sharper focus later in this Preface. Responsibilist themes are represented in this collection through contributions by Linda Zagzebski, Jonathan Dancy, James Montmarquet and Christopher Hookway and Richard Paul. In the next section, we will begin by discussing the epistemological problems that frame the research interests of the reliabilist strain of contemporary virtue epistemology that Ernest Sosa inaugurates. After this, following the general arrangement of papers in the collection, we will further elaborate on the epistemological problems that frame the research interests of responsibilist virtue epistemology.

Sosa made his initial suggestion for an intellectual virtue-centered account of knowledge and justification in support of externalist or reliabilist epistemology. Indeed he drew his own comparisons with the reliabilist conceptions of justification and the "causal theory of knowing" of Alvin Goldman and others circa 1980. A driving force for the development of externalist accounts of knowledge and justification, in turn, has been discussion of Gettier-type problems, which pose a very serious challenge to the traditional conception of knowledge as justified true belief. It merits a brief digression to explain the nature of Gettier-problems and the general character of reliabilist approaches to them.

E.L. Gettier (1963) rocked the traditional conception of knowledge through his framing of specific cases of justified true beliefs that appear not to constitute knowledge. This suggests the need for either a new account of justification, or else a search for an additional condition (of warrant or positive epistemic status) that together with justified true belief, constitutes the proper set of necessary and sufficient conditions for knowledge. Gettier-problems have been widely discussed since the mid-1960s without any single clear line of resolution. But one of the striking features of contemporary epistemology since Gettier’s influential paper was published has been a broadening consensus on the need to abandon many aspects of the "traditional" conception of knowledge. This is primarily because accounting for the problems raised by his cases appears to demand something besides the internalist construal of justified belief that the traditional conception holds: it appears to necessitate restricting genuine knowing only to instances where the truth of the agent’s belief is "linked" with the causal process that produced it in some reliable way.

Can an agent’s belief constitute knowledge if its truth was in fact a matter of luck—if its truth was in fact improbable or merely fortunate, given the process that produced it? In Gettier cases, the agent’s justification for her beliefs has come ‘unglued’ from the truth of her belief. In these cases their truth is really a matter of luck, given the process by which the belief was produced. In order to accord with the intuition that the true beliefs in Gettier cases do not amount to knowledge, epistemologists began looking at what went wrong, and at how to strengthen the conditions on knowledge so that truth and justification could not come apart. Where the traditional conception fails, many conclude, is in not demanding of knowledge an objective or external connection between the truth of the agent’s belief and the belief-producing cognitive process (BCP) that gave rise to it. Demanding such a truth-connection would be to place an "external" in contrast to an "internal" condition on justification, and theories that place any external conditions on justification are classed as externalist theories. An externalist theory that asserts that the production of a belief by a reliable cognitive process is (at least) a necessary condition for it to be justified, is called a "process reliabilist" theory of justification.

We return now to the point that Sosa proposed his intellectual virtue-centered epistemology as a means of improving the prospects of a process reliabilist account of epistemic justification. Why did it need improving? While the shift in epistemology toward reliabilism has been widespread, there are also serious objections to reliabilism. Noting the most serious of these, Sosa acknowledges that reliabilism "comes in a great variety of types most of which are clearly unacceptable." We have already seen the first and most incisive revision that Sosa proposed: No previous form of reliabilism explicitly entails the change in the direction of analysis described above. Other related revisions or modifications that ‘virtue-basing’ a reliabilist account would suggest are initially indicated by the problem (often seen as a crucial aspect of Gettier-problems) of the ‘accidental reliability’ of a belief-producing cognitive process (hereafter BCP). Can just any reliable BCP provide the kind of justification a true belief needs to in order to constitute knowledge? Can "strange" or "fleeting" processes, for instance, give rise to knowledge? Because virtues, whether ethical or intellectual, are understood as stable dispositions, defining epistemic justification in terms of intellectual virtues will respond directly to a serious objection to process reliabilism: that it must, by its very logic, attribute knowledge and justified belief whenever reliability is present (and even if the belief has its source in a process that is itself fleeting). To the extent that it can demonstrate a principled way to rule out fleeting, random and ad hoc processes, the conception of justification-making or justification-conferring BCP’s is greatly strengthened

Performing an action due to a demonstrable skill requires a different explanation than success due to mere luck; success due to luck is not reliable or likely repeatable under normal circumstances. To attribute an action or belief to the workings of an intellectual virtue is to identify its ground with an attribute of the agent that is still more stable and less fleeting than is a skill: Sosa and Goldman each require that justified beliefs be generated by a genuine power or capacity or competence of the agent to arrive at truth, and it is such powers of persons that are identified with intellectual virtue. Sosa identifies five basic powers, and the normal parameters and limits of reliability for each. These five are the faculties of perception, memory, reason, intuition and introspection.

While we have used the problem of fleeting processes to indicate an initial difference between virtue epistemologies and "generic" forms of reliabilism, the reader will encounter a number of additional problems in the early chapters of this collection. These help us to understand more fully the revisions that characterize reliabilist virtue epistemology. Goldman’s contribution in chapter 1, for instance, discusses problem cases of "strange processes," such as that of the ‘reliable clairvoyant’ (closely associated with what is called the "meta-incoherence problem"). An agent posited to have such a faculty might be unaware of it, and unable to connect those beliefs it causes her to have with any evidential considerations she may possess that would bear upon the rational acceptability of those beliefs. Is Goldman committed to viewing this reliable faculty as a virtue, and therefore as justification-conferring? How would the virtue epistemologist try to square our conceptual understanding of intellectual virtues with our intuitions regarding such cases?

Two further problems that both Sosa and Goldman address are the "Generality," and "New Evil Genie" problems, which we might describe here. In briefest terms, the Generality problem presents reliabilists with the challenge of describing those cognitive processes that are understood as justification-conferring, in a way that is neither too broad nor too narrow. Moving beyond the problems of strange and fleeting processes, the virtue epistemologists will also need to ask, ‘How and why do we count certain processes and not others among the intellectual virtues?’ It must not be the case that an account is so specific that every true belief is credited to a virtue or reliable faculty; but neither must it be the case that the account is so broad that a single faculty is taken to generate beliefs with widely varying epistemic statuses. The New Evil Genie (or Demon) problem revolves around cases of systematic deception or wholesale falsehood. It has been among the most difficult obstacles for reliabilist accounts of justification, since it highlights strong intuitions about the importance of subjective justification, intuitions that may be thought to support an internalist perspective on justification. How are reliabilists to square their method of attributing intellectual virtues to agents with the consideration that the degree of effort to attain the truth, (and the phenomenological experience of a victim in the Genie’s world), is by hypothesis indistinguishable from that of the victim’s twin, a reliable agent in the actual world? This problem is posed forcefully by Lawrence BonJour in our collection, and papers by Goldman, Sosa, and Greco each suggest different ways of responding to it.

Thus far we have focussed only upon those versions of virtue epistemology that initially developed, as we earlier said, out of concern to improve the prospects of reliabilism. But an alternative "responsibilist" strain of virtue epistemology began to emerge soon after Sosa’s initial formulations of his position, as was perhaps first evident in Lorraine Code’s Epistemic Responsibility (1987). The kinds of virtues responsibilists focus on are akin to those that Richard Paul, Director of the Foundation for Critical Thinking, defines and discusses in chapter 13: the virtues of intellectual humility, courage, empathy, integrity, perseverance, fair-mindedness, and faith in reason. Some of the differences between the reliabilist and responsibilist approaches often seem little more than matters of divergent focus or emphasis. However, in certain respects the differences between "virtue reliabilism" and "virtue responsibilism" are far more substantial and interesting than this. They surface, first and foremost, in the conceptual understanding and identification of the intellectual virtues themselves.

While most forms of virtue epistemology today place a success condition on the ascription of a virtue to an agent, not all share Greco’s claim that "the essential aspect of an intellectual virtue is its success component, or in Sosa’s and Goldman’s terms, its reliability." Responsibilists like Lorraine Code (1987) concede that reliability "maintains a closer connection with truth and warrantability than responsibility can establish." But they do not see reliability as the most important or defining feature of a virtue. Code’s book announces a primary interest in the virtues as they relate to active agency, and she defines the intellectual virtues more restrictively, as acquired intellectual habits and dispositions; only these, she points out, are the proper object of attributions of praise and blame, not the genetically-endowed "faculties" that the reliabilists would include among the intellectual virtues. Justification, responsibilists insist, depends greatly not just on the quantity of evidence an agent has for their beliefs, but on the quality of that evidence, and this is most often a function of how well one investigates. Related to this, the reliability of agents may depend crucially on acts they perform or choices they make in the context of inquiry.

What is important to notice about the responsibilists’ list of intellectual virtues is that many of them are analogues of recognized ethical virtues. They are conceived to have their source in acquired habits of character, yet to pertain (not to action but) to thought or belief. To accord with such considerations, responsibilists often seek an account of the intellectual virtues that builds into them the kind of motivational constraint that is integral to the Aristotelian conception of the moral virtues. For example, Linda Zagzebski (chapter 9 & 1996) restricts the intellectual virtues to acquired dispositions, and builds into her conception of them an especially strong motivational requirement, as may be required to defend her unique move (chapter 9) "to a virtue epistemology based on virtue in the ethical sense."

This is not to say that most responsibilists do not also acknowledge the importance of a success condition on the attribution of a virtue to an agent. It is indeed very Aristotelian to include both success and motivation conditions (again though, if we base this on how Aristotle understood moral virtue, and not on how he himself understood intellectual virtue). "An act is an act of virtue A" according to Zagzebski, "just in case it arises from the motivational component of A, is something a person with virtue A would characteristically do in the circumstances, and is successful in bringing about the end of virtue A because of these features of the act." The centrality Zagzebski gives to the motivational component of acts of virtue leads us to the important recognition that there is a spectrum of positions from stronger to weaker on the understanding of what constitutes a virtue epistemology. Responsibilists share concerns over issues like proper motivation, rationality, and doxastic responsibility. They appear especially interested in how agency works, and in how responsible agents can practically apply the virtues of an ideal critical thinker to the context of their inquiry. These issues can be seen as aspects of an account of active epistemic agency, and no dubious voluntarism about belief is suggested by such concerns in and of themselves. But the concerns of responsibilists are enveloped within conceptions of virtue theory that are generally "stronger" than those current among virtue reliabilists.

To a substantial degree, the range or spectrum of views I am alluding to can be understood by referring back to debates within ethics over the past two decades. One rather contentious issue in ethics has been the dividing line between "a theory of virtue" and "a virtue theory." It is very difficult to be clear about the dividing line between the positions these terms represent, and language often confounds. There are those who have a theory of virtue, but would not describe themselves as "virtue theorists," which term they reserve for a thesis concerning the primacy of the virtues that they consider "radical"; and as we are now seeing, even among those who describe themselves as virtue theorists, there is significant divergence over the implications of such an approach. What we can say is that the strongest forms of virtue ethics are generally characterized by an additional claim about the conceptual primacy of the virtues: Not only are virtues conceptually prior to specifications of rules, they are also conceptually prior to specification of ends; not only the "right," but also the "end" of the ethical life (the Good or the Good life), is derivative from or at least partly constituted by the virtues.

For its proponents such a view is not radical, but simply reflects a return out of modernism to some of the assumptions that were basic to classical thought. For Aristotle it seems, the life according to virtue is not merely a means, but in a complete life is also partly constitutive of eudaimonia, or human flourishing. Staying with virtue ethics for the moment will also help to clarify how these concerns express themselves in contemporary philosophical debate. Those who would implicate the entire modernist tradition in ethics—who would characterize it as "act" rather than "agent" centered, or as concerned with "doing" in contrast to "being"—also tend to reject modernist typologies of normative ethical theories. While certainly conceding that consequentialists and deontologists, etc., may incorporate within themselves a theory of virtue, they themselves would likely identify virtue theory as indicating a horse apart—an agent-centered perspective in ethics that takes the concept of a virtue as prior to or more basic than other central ethical concepts, of which the "Good" and the "Right" are of course also top candidates. Reflecting this formal difference in their way of ordering the relationships of priority between ends, rules, and virtues, those described as "strong" virtue theorists typically understand virtue theory to resist subsumption under labels such as consequentialism or deontology.

Certainly not all self-described virtue ethicists hold this strong thesis. But what this shows is that the term "virtue theory" admits of a range of denotations, and is often used to indicate a weaker or stronger thesis from one author to the next. This is a well-recognized aspect of debates within ethics, but I raise it here because it is also helpful for understanding certain key differences both among virtue epistemologists, and between them and their critics. It is worth noting in this regard that contemporary textbooks often classify epistemological theories as deontological or consequentialist, (or deontological/non-deontological), on analogy with theories in ethics (Steup 1996). It should not be surprising, then, that in reading the works of virtue epistemologists we also encounter a range of positions reflecting weaker and stronger claims about what constitutes virtue theory, and a debate strongly analogous in certain respects to one ongoing in ethics.

This collection contains papers expressing a wide and exemplary range of virtue-theoretical perspectives on epistemological topics. Some are previously published papers, and others appear in print here for the first time. Our readings divide into four Parts. The selections of Part I begin with contributions by Sosa and Goldman wherein each author gives an account of knowledge and justified belief in terms of their source in intellectual virtues. Each tries to show how he can satisfactorily handle some of the familiar problems that process reliabilism faces within a virtue-centered version of the more general reliabilist approach. In chapter 1 Alvin Goldman provides a sketch for such a new understanding of epistemology, one which significantly modifies his own previous views. Goldman’s "Epistemic Folkways and Scientific Epistemology" is set in the context of a broader effort that he, Sosa and Hilary Kornblith have shared, of developing a naturalized epistemology. The author provides a guideline for the fuller development of his non-eliminative naturalistic account of knowledge and justification, by distinguishing two separate but inter-related missions of epistemology. One is the descriptive task of detailing our commonsense or "folk" epistemic concepts, and the other is the evaluative task of critically evaluating and if necessary revising these concepts. Goldman explains how both missions of epistemology integrally depend on empirical studies of the human reasoner in the social and cognitive sciences, and why this leads him to characterize his naturalized account as "scientific epistemology."

In chapters 2 "Reliabilism and Intellectual Virtue," and 3, "Three Forms of Virtue Epistemology," we have papers by Ernest Sosa that provide a sound introduction to his developed accounts of knowledge and justification. Here we see Sosa grappling with those major problems for reliabilist accounts that we have already outlined in section two of the Preface, above. But together with this we see Sosa develop the view he calls "virtue perspectivism," which leads away from a strict causal account in the direction of what can be called a "mixed" or mixed externalist account of justification. Distinguishing between animal and reflective knowledge, and between aptness and justification, Sosa develops both subjective and objective conditions for the justified true beliefs that constitute human reflective knowledge. In chapter 3, Sosa explains the similarities and differences between the latest virtue-epistemic version of Goldman’s causal theory of knowing, the "proper functionalism" of Alvin Plantinga, and his own virtue perspectivism. In the conclusion he makes the conciliatory suggestion that the disagreements between the three "seem relatively small, when compared with the large areas of agreement…It seems appropriate to view the three approaches as varieties of a single more fundamental option in epistemology, one which puts the explicative emphasis on truth-conducive intellectual virtues or faculties, and is properly termed ‘virtue epistemology’."

In chapter 4, "Ever Since Descartes," Hilary Kornblith argues for a form of reliabilism with the flexibility to accommodate the important intuitions that have long informed internalist conceptions of justification. Kornblith’s work has exuded a strong influence in the development of "mixed" accounts of justification, and this characteristic is in turn a reason why virtue epistemologists think that their approach contributes substantially to resolving the dispute between internalism and externalism in epistemology. The author here develops a framework that makes it both historically and theoretically understandable how the reliability constraint on justification is compatible with an "internal" (or inner coherence) constraint, and with an "action-theoretic" (or responsibility/motivational) constraint. The latter, the important focus of responsibilist virtue epistemology, concerns whether the chosen acts by which an agent’s belief-forming cognitive processes are influenced were responsibly motivated by a desire for true beliefs. In connecting the action-theoretic constraint with the issue of subjective justification, Kornblith’s paper also serves to introduce the reader to a range of issues concerning active epistemic agency that receive further direct attention in Parts III and IV of the collection.

Part II contains papers that further develop the naturalistic orientation of reliabilist virtue epistemology; these contributions provide a closer focus on a virtue-based account of knowledge and the resources it provides for responding to the challenge of skepticism. In chapter 5, a specially edited selection from Putting Skeptics in their Place (Cambridge, 2000), John Greco develops "agent reliabilism" as an account of knowledge as arising from the cognitive abilities or powers of epistemic agents. Greco argues that agent reliabilism has unique resources for addressing an important kind of skeptical argument, one that trades on the assumption that for agents to have knowledge they must be able to discriminate the truth of their beliefs from alternative possibilities. In the course of explicating a distinction between relevant and irrelevant alternative possibilities and using it to answer the skeptic’s charge, Greco develops a sophisticated possible-worlds analysis for abilities, including cognitive abilities.

In chapters 6 and 7, Jonathan Dancy and Lawrence BonJour discuss Sosa’s epistemology as presented in the collected essays of Knowledge in Perspective (1991). Both authors are concerned with aspects and implications of Sosa’s naturalism, especially with his understanding of supervenience. Dancy attempts to drive a wedge between Sosa’s "formal foundationalism" and the doctrine of supervenience which he sees supporting it. He also challenges Sosa’s conception of what makes a feature a virtue. Dancy here argues for an alternative non-consequentialist conception of virtue, thereby directing our attention to a major issue of debate among virtue epistemologists which will come under closer scrutiny in Part III of the collection. BonJour develops a skeptical dilemma for Sosa with respect to the latter’s iterative condition on reflective knowledge, the requirement that an agent have "epistemic perspective" on the reliability of his/her own belief-producing cognitive faculties. In his responses to both authors in chapter 8, Sosa details his understanding of the supervenience of evaluative epistemic properties on non-evaluative properties. He also engages BonJour in debate over the problem of skepticism, and Dancy in debate over the background assumptions informing his approach to the definition and identification of the intellectual virtues.

In Part III, our focus shifts to a debate between those forms of virtue epistemology with a predominantly reliabilist orientation, to those with a predominantly responsibilist orientation. This distinction between these two strains of virtue epistemology should not be used to exaggerate differences, since we have already pointed out many shared tenets and common interests among those who describe themselves as working within virtue epistemology. (Indeed I think that certain of the tensions that exist among virtue epistemologists are unnecessary, arising primarily from divergent interests in the complex topic of epistemic agency.) However that may ultimately be seen, these differences are interesting in their own right and at times quite sharply expressed; in Part III of the collection we see them laid out and debated in turn. Here we find a broadening range of interests in the intellectual virtues and an account directed specifically towards active, as contrasted with passive epistemic agency. Corresponding to these interests, we also find that many new and difficult questions arise. To identify but a few: Do all virtues carry a reliability constraint? Do some or all also carry a motivational constraint, such as that which one finds in an Aristotelian moral virtue? What is the relationship between a "motive" and an "end-in-view? Do responsibilists conform to the naturalistic orientation that characterizes the reliabilist approach? Is a "unifying" account of ethical and epistemic virtue plausible? What theoretical advantages might such an account hold?

Among our contributors, perhaps few differences will appear more marked to the reader than that between John Greco, who argues that any view making justification or knowledge a function of agent reliability constitutes as a version of virtue epistemology, and Linda Zagzebski, whose rejection of that view is apparent in her provocative title "From Reliabilism to Virtue Epistemology." Zagzebski (chapter 9) begins Part III with a discussion of process reliabilism and the difficulties it faces is adequately explaining what makes "the good of knowledge" greater than "the good of true belief." This problem is crucial, as the author sees it, and it "pushes us first in the direction of three offspring of process reliabilism—faculty reliabilism, proper functionalism, and agent reliabilism, and finally to a virtue epistemology based on virtue in the ethical sense."

Julia Driver (chapter 10) is known to us primarily for her outstanding past work in ethical theory, and her contribution here illustrates the fruitfulness of cross-fertilization between philosophical subdisciplines. The analysis she offers in "Ethical and Epistemic Virtue" illuminates disanalogies as well as analogies between ethical and epistemic evaluation. Her approach to this topic comes out of a consequentialist background aligned with reliabilist views in epistemology. From this perspective she offers a critique of Zagzebski’s "pure virtue theory," which she compares with Michael Slote’s agent-based form of virtue ethics.

Driver’s critique of Zagzebski underlines another issue that is a matter of pointed debate: the plausibility of a unified account of ethical and epistemic virtue. James Montmarquet’s paper in chapter 11, "An ‘Internalist’ Conception of Epistemic Virtue," provides an account of the relationship between the ethical and epistemic virtues that is critical, on the one hand, of Zagzebski’s strong "assimilationist position," and on the other, of the strong "externalist anti-assimilationist position." (which Driver may be thought to exemplify in arguing for "the disunity of virtue"). The alternative model Montmarquet develops for the relationship between the virtues is quite unique in that it does not arrange them by content or domain, but rather by the susceptibility of each disposition to direct control. This innovative approach issues in a model reflecting "a three-fold distinction cutting across the moral/epistemic divide."

In chapter 12, "Regulating Inquiry," Christopher Hookway begins with the topic of reasons for belief. The intellectual virtues, on Hookway’s view, play at least two important roles in respect to believing for reasons, for as he argues, "virtues…enable us to respond to reasons: they provide a sensitivity to rational requirements in particular cases, and they are motivating." To understand the first, normative aspect of reasons, Hookway leads us into issues of rationality, while to explore the second, motivational aspect, he asks us to consider cases of the intellectually akratic individual. If I accept that I have a sound reason to doubt or believe some proposition, must I be motivated to doubt or believe it? To resolve the problematic nature of views that would affirm this claim unqualifiedly, Hookway presents a pragmatist-inspired account on which motivation may be present, in certain instances (and to the advantage of the agent), not by conscious reflection but by force of unconscious habit.

The discussions of the relationship between ethical and epistemic virtues in Part III invite us to explore further, in Part IV, certain topics that have been of special interest for virtue ethicists and epistemologists alike. Chapter 13 is "Critical Thinking, Moral Integrity, and Citizenship: Teaching for the Intellectual Virtues," by Richard Paul. Paul, a developer and foremost advocate for a critical thinking curriculum at all levels of public education. Paul critiques pedagogical approaches that presume a sharp separation of the affective and moral from the cognitive dimensions of learning. In response to a model of "compartmentalized domains," Paul explores what he sees as the intimate connections between critical thinking, moral integrity, and responsible citizenship: "The problems of education for fairminded independence of thought, for genuine moral integrity, and for responsible citizenship are not three separate issues but one complex task." The intellectual virtues themselves are interdependent, and one does not develop students’ thinking skills "without in some sense simultaneously developing their autonomy, their rationality, and their character." Critical thinking in its richest and most valuable sense is not simply a matter of cognitive skills, any more than moral integrity and responsibility are merely matters of good intention. On Paul’s view, "Skills, values, insights, and intellectual traits are mutually and dynamically interrelated. It is the whole person who thinks, not some fragment of the person."

In my own contribution to the collection in chapter 14, "Virtue Theory and the Fact/Value Problem," I attempt both to analyze and to mediate the dispute between "virtue reliabilist" and "virtue responsibilist" epistemologies. The analysis casts the debate over the definition and identification of the virtues in terms of background understandings of the Greek tradition of virtue theory. The debate can be usefully understood in terms of competing interpretations of Aristotle’s own approach, or more generally if one prefers, in terms of contrasts between Platonic and Aristotelian conceptions of virtue. The mediating suggestions I make concerning a plausible form of a unified account of ethical and intellectual virtue reflect the unique resources I find in virtue theory and in American pragmatism.

In chapter 15, "Epistemic Vice," Casey Swank turns our attention from an account of epistemic virtue to that of epistemic vice, and challenges what he takes to be the received view of the latter. In sorting out the differences between understanding an epistemic vice as something "bad and epistemic" versus something "epistemically bad," Swank focuses criticism on truth-centered accounts of epistemic virtue and vice, and indicates the direction an alternative, non-truth-linked account would need to take.

In our final selection, Linda Zagzebski’s "Phronesis and Religious Belief," the author breaks significant new ground by outlining a theory of rationality with integral ties to virtue theory. She discusses why she thinks it is crucial, in debating the rationality of religious belief, to distinguish a weaker principle of rational permissibility for beliefs from a stronger principle of rational praiseworthiness. For the author, the phronimos or person of practical intelligence "is the paradigm in relation to which a host of concepts of epistemic evaluation can be defined." Her weaker and stronger principles are therefore articulated in terms of the differences between what the phronimos "might" believe, and what she/he "would" and "would not" believe, under suitable background and evidential circumstances. In her conclusion, she applies her principles to the appraisal of claims made by Alvin Plantinga on the part of Calvinistic "reformed epistemology." The upshot here is a quite original argument casting doubt on the rational defensibility of Plantinga’s model of "properly basic" Christian belief.

Those of responsibilist orientation perceive synergistic connections between the human reasoner and the human valuer. Therefore they emphasize not just the distinction between intellectual and ethical virtues, but also the intimate relationships among the kinds of traits that make for excellence in an integrated, properly attuned individual. They may also perceive the kind of parity that they find desirable in an approach to ethical and epistemic evaluation already present in classical virtue theory, at least potentially. We should distinguish inspiration and substantial support, however, and be clear that it is their own undertaking, and not that of the ancients, should they choose to develop virtue theory in the direction of a unified theory of value across epistemology and ethics. I hope that this adequately explains for the reader of our collection why—the noted problems not withstanding—I place the term "character" in the title, evoking as it must both the commonalities and differences among philosophers working today in the area of virtue epistemology.