David Henderson

The overall work of this chapter is two-fold: First, it illustrates how philosophical reflection leading to transglobal reliabilism fully conforms to the model of low-grade _a priori_ reflection advanced in Chapter 2. Second, the chapter serves to provide a sustained defense of transglobal reliabilism. The defense involves the full range of data that is characteristic of low-grade _a priori_ philosophical reflection. Of course the data to be accommodated includes judgments provoked by scenarios—so this chapter discusses a diversity of such judgments (some featured in the previous to chapters, some newly introduced here). Various judgment tendencies that initially seem to point in directions incompatible with transglobal reliabilism are discussed. It is shown how these tendencies can be accounted for by transglobal reliabilists. However, the data to be accommodated also includes some sense for the points and purposes of the evaluative concept of epistemic justification. It is not uncommon in philosophical reflection to find oneself pulled in apparently opposing directions when thinking about particularly difficult cases—and the concept of epistemic justification has its difficult applications and “dual pulls.” Ultimately, reflection on the points and purposes of the concept of epistemic justification allows one to understand this dual pull within a fleshed out transglobal reliabilist account.

This chapter develops a revisionary account of a kind of _a priori_ reflection that is central to much philosophy—viz., conceptual analysis. It argues that this kind of _a priori_ inquiry has subtle empirical elements rather than being a wholly non-empirical enterprise. This break with more traditional conceptions of the _a priori_ is marked by writing of “low-grad_e a priori”_ justification. The chapter begins with the observation that philosophical reflection commonly commences with intuitive judgments having to do with concrete scenarios. This is as it should be, for one's conceptual competence typically is much better at applying concepts to concrete cases then it is at directly yielding highly abstract, highly general, conceptually grounded necessary truths. To arrive at the sorts of general conceptually grounded necessary truths to which philosophers typically aspire, one must go beyond the direct and reasonably reliable deliverances of one's conceptual competence: one must engage in what is essentially an abductive inference—an inference to the best explanatory account of the workings of the relevant concepts _and_ of one's conceptual competence. It is argued that the needed abductive inference has an ineliminable empirical component.

Justificatory cognitive processes must be tractable. That one ought to produce and sustain beliefs in certain ways entails that one can. Fitting epistemic standards for human epistemic agents must be sensitive to which potential belief-forming processes humans are capable of employing, at least with training. Such (low-grade _a priori_) points call for a naturalized epistemology. This chapter clarifies this demand by elaborating upon the kind of idealized normative standards one can expect from a naturalized epistemology, and on the range of disciplines that can contribute to normative naturalized epistemology. It also advances a way of thinking about naturalized epistemology as encompassing a broad spectrum of inquiry ranging from low-grade reflection about central evaluative concepts to richly empirical inquiry revealing the character of actual human cognitive processes and the possibilities for variation. The organizing principle that holds the spectrum together is the concern for a kind of engineering for truth seeking. Just as various kinds of inquiry find a place within common engineering disciplines, so likewise a wide range of inquiries can find a place within naturalized epistemology. The spectrum of inquiry falling within naturalized epistemology is the E-spectrum (the ‘E’ standing both for epistemology and for engineering).

The neoclassical reliabilism of chapter three is found to be flawed. Its inadequacy is strongly suggested by variants of the so-called “new evil demon problem.” Discussion of such scenarios, and the general association of epistemic safety with robustness of reliability, leads to a position that better captures the concern for epistemic safety associated with objective epistemic justification: _transglobal_ reliabilism. Transglobal reliability is reliability relative to the wide reference class made up of experientially relevant possible global environments. Further scenarios reinforce the idea that modulational control remains important. The resulting position is termed _transglobal reliabilism_. On this account, for an agent to be objectively justified in holding a belief, that belief must arise or be maintained by way of processes that are transglobally reliable under suitable modulational control. The main outlines of transglobal reliabilism are developed here and supported by reflection on a number of scenarios.

Results in the previous chapter call for an _iceberg epistemology_ with two complementary foci. One is the traditional focus on conscious and accessible psychological factors; the other focus is the full set of epistemologically relevant psychological factors, many of which may be only partially, piecemeal-fashion, accessible. This allows one to revisit and re-conceive certain traditional epistemological doctrines such as foundationalism and coherentism. One then finds that central themes associated with each position can be salvaged, but that doing so requires one to transform each position somewhat, limiting these doctrines in important ways. The holistic themes from coherentism can be honored, with qualifications, but only when abandoning the exclusive focus on occurrent doxastic states that has been characteristic of coherentism. These holistic chores are managed at the level of the full set of epistemically relevant psychological factors. The foundationalist insistence on terminating chains of justificatory bases does seem to characterize the justificatory processes at the level of accessible processes. But there will be some complications in just where justificatory bases leave off—and these will not be fully understandable at the level of accessible processes. Properly limited and conceived of as treating of processes understood in terms of the different foci just distinguished, coherentism and foundationalism are compatible and can each be honored—along with the position sometimes called contextualism.

Emerging results in cognitive science have far-reaching consequences concerning the cognitive processes that make for doxastic justification. Epistemologists commonly suppose that epistemological tasks can be, and are, managed by cognitive processes in which all the information bearing on a stretch of belief fixation is occurrently represented in the course of these processes. Recent work in cognitive science associated with “the frame problem” indicates that this common assumption is misguided (as are several tempting, weakened versions of it). It is argued that an epistemologically crucial aspect of effective belief-formation involves cognitive processing that accommodates a wide range of information in an implicit way—without the information being explicitly represented in the course of processing, either consciously or unconsciously. What emerges is a view of objectively justificatory human belief-fixing processes according to which much of the epistemic action, much of what contributes to one's systematic successes, lies below the readily accessible “surface” of one's cognition. This calls for an epistemology with two legitimate, and complementary, foci. One is the traditional focus on psychological factors that are conscious during belief formation, and/or are readily psychologically accessible after the fact. And the other focus is the full set of epistemologically relevant psychological factors, many of which may be only partially, piecemeal-fashion, accessible after the fact of belief-formation. Thus, one arrives at what is termed _iceberg epistemology_.

This chapter focuses on scenarios that point in the direction of a reliabilist account of _being objectively justified in believing._ A concern for some form of epistemic safety is found to be central to the concept of justification. However, the chapter does not purport to give a final account of the relevant sort of safety. What does emerge is a plausible, refined, version of what is now classical reliabilist thinking. The refinements turn on two ideas. The first is a distinction between global reliability and (merely) local reliability. _Local_ reliability is reliability relative to the specific, spatiotemporally local, environment that the cognitive agent happens to occupy when the belief is formed. _Global_ reliability is reliability relative to the class of actual and potential local environments within the cognitive agent's global environment. The second idea is that of “suitable modulational control.” Cognitive agents like humans deploy various belief-forming processes in ways that are _holistically integrated_ within the agent's overall cognitive architecture. Information afforded my some processes may then condition, enhance, trigger, restrain, or suppress the working of other cognitive processes. Such modulational control can enhance the reliability of one's cognitive processes. On the refined reliabilist account developed here, termed _neoclassical_ reliabilism, in order to be objectively justified in holding a belief, one must have formed that belief by way of processes that are globally reliable under what we call suitable “modulational control.” While neoclassical reliabilism will be criticized in subsequent chapters, it provides a charitable version of much reliabilist epistemology.

This chapter provides an overview of the chapters to follow, their central arguments, and how the various pieces of the view developed in the book fit together. Two of the subsequent chapters are identified as pivotal methodological chapters: Chapter Two—on the _a priori_ and on philosophical reflection in particular—and Chapter Six—on the spectrum of inquiry to be pursued within a fittingly naturalized epistemology. The chapters that follow each of these methodological chapters implement and illustrate their lessons. The discussion of various versions of reliabilism in the third, fourth, and fifth chapters provides an extended instance of the “low-grade priori” reflection characterized in Chapter Two. The discussion and criticism (in Chapter Seven) of the psychological “proto-theories” that have mis-informed much epistemology reflects the far-reaching significance of certain empirical parts of the epistemological spectrum, while Chapter Eight articulates some consequences of an empirically better informed understanding of epistemically relevant processes. Thus, the revisionary treatments (in Chapter Eight) of foundationalism, coherentism, and contextualism constitute important lessons to be gained from a fittingly naturalized epistemology.

The chapter rehearses the main outlines of _gatekeeping contextualism_—the view that it is central to the concept of _knowledge_ that attributions of knowledge function in a kind of epistemic gatekeeping for contextually salient communities. The case for gatekeeping contextualism is clarified within an extended discussion of the character of philosophical reflection. The chapter argues that normatively valenced, evaluative concepts constitute a broad class of concepts for which a sociolinguistic point or purpose may be readily sensed—and for which the intimate connection between point/purpose and the concept itself is readily understandable. Commonly at the inception of philosophical reflection one’s sense for the point or purpose of a specific concept is seldom highly articulate. The chapter argues that this does not prevent one’s grasp of point/purpose from doing real work in philosophical reflection.

The papers in this issue all concern the normative standards by which we do or should regulate our joint epistemic lives in communities. Plausibly, reflection on how we should regulate ourselves—what one should insist on in one's own practice and that of one's epistemic partners—takes some cues from reflection on what we do insist on. The reverse is plausibly also the case. These papers also, more or less explicitly, suggest that our epistemic sensibilities themselves reflect the demands of epistemic practice within a community of interdependent agents.

This paper is a beginning—an initial attempt to think of the function and character of epistemic norms as a kind of social norm. We draw on social scientific thinking about social norms and the social games to which they respond. Assume that people individually follow epistemic norms for the sake of acquiring a stock of true beliefs. When they live in groups and share information with each other, they will in turn produce a shared store of true beliefs, an epistemic public good. True beliefs, produced individually or in groups, constitute an epistemic good—one commonly stockpiled and distributed within a community. Epistemic norms can then be understood as a kind of socially developed and transmitted normative sensibility having to do with the production of this individual and public good. Epistemic norms should serve to regulate this practice—coordinating the practice of individuals so as to afford the benefits of life in an interdependent epistemic community—and to manage the risks of being epistemically dependent on others within such a community. Here, we provide some attempts to characterize the central aspects of "the epistemic game"—the epistemic choice situation confronted by communities of epistemic agents.

In "Epistemic Norms and the 'Epistemic Game' They Regulate", we advance a general case for the idea that epistemic norms regulating the production of beliefs might usefully be understood as social norms. There, we drew on the influential account of social norms developed by Cristina Bicchieri, and we managed to give a crude recognizable picture of important elements of what are recognizable as central epistemic norms. Here, we consider much needed elaboration, suggesting models that help one think about epistemic communities and, ultimately, the temptations confronted in one's epistemic life. We suggest that once "the epistemic game" is embedded into a wider set of gains and losses, one can understand the epistemic choice situation as a form of mixed-motive game. This allows one to understand epistemic games more straightforwardly using Bicchieri's framework for thinking about social norms. That is, once the choice situation denominated only in terms of classical epistemic gains and losses is embedded in a wider accounting of goods that may compete with the production of true beliefs, one can see how an agent's individual and community epistemic project can be furthered by social norms regulating epistemic practices. We recommend thinking of epistemic norms as analogous to social norms for hygiene.

Armchair reflection commonly provides a low-grade a priori form of justification for certain claims which, if true at all, are necessarily true and are rendered necessarily true solely by virtue of their constituent concepts, independently of any contingent facts about our actual world. The justification afforded is aptly understood as a priori, yet of a low-grade sort that has an ineliminable empirical dimension. The conceptual competence that folk possess for the non-deferential application of concepts features prominently, as does the general capacity for abductive inference. Armchair methodology in linguistics provides a useful model for armchair philosophical reasoning. We argue for two somewhat surprising hypotheses: first, metaphysically necessary truths that are semantically non-analytic and epistemologically a posteriori are underwritten by yet-more-fundamental necessary truths that are analytic; and second, it is a conceptually grounded necessary truth that some statements expressing epistemic possibilities do not express metaphysical possibilities.

This introductory chapter proceeds in three parts. The first section characterizes the general approach to epistemology around which the volume revolves—purposeful epistemology—and examines the general motivation for that approach. The guiding idea is that considerations about the point and purpose of epistemic evaluation might fruitfully constrain epistemological theory and yield insights for epistemological reflection. The second section explores the approach by characterizing some important versions of it. Several themes and issues that we see running through the volume are here articulated and highlighted. The third section provides a brief summary of each contribution to the collection.

Epistemic Evaluation aims to explore and apply a particular methodology in epistemology. The methodology is to consider the point or purpose of our epistemic evaluations, and to pursue epistemological theory in light of such matters. Call this purposeful epistemology. The idea is that considerations about the point and purpose of epistemic evaluation might fruitfully constrain epistemological theory and yield insights for epistemological reflection. Several contributions to this volume explicitly address this general methodology, or some version of it. Others focus on advancing some application of the methodology rather than on theorizing about it. The papers go on to explore the idea that purposes allow one to understand the conceptual demands on knowing, examine how purposeful epistemology might shed light on the debate between internalist and externalist epistemologies, and further develop the idea of purposeful epistemology.

In What Philosophers Know, Gary Gutting provides an epistemology of philosophical reflection. This paper focuses on the roles that various intuitive inputs are said to play in philosophical thought. Gutting argues that philosophers are defeasibly entitled to believe some of these, prior to the outcome of the philosophical reflection, and that they then rightly serve as significant (again defeasible) anchors on reflection. This paper develops a view of epistemic entitlement and applies it to argue that many prephilosophical convictions of the kind Gutting discusses would be just the sort of belief for which entitlement would plausibly be defeated from the start. They then could not properly play the role in philosophical reflection that Gutting envisions for them

Accounts of what it is for an agent to be justified in holding a belief commonly carry commitments concerning what cognitive processes can and should be like. A concern for the plausibility of such commitments leads to a multi-faceted epistemology in which elements of traditionally conflicting epistemologies are vindicated within a single epistemological account. The accessible and articulable states that have been the exclusive focus of much epistemology must constitute only a proper subset of epistemologically relevant processing. The interaction of such states looks rather contextualist. It might also be called quasi-foundationalist. However, in attending to our epistemological tasks we must rely on processing that is sensitive to information that we could not articulate, that is not accessible in the standard internalist sense. When focusing on the full range of epistemologically important processes, the structure of what makes for justification is rather more like that envisioned by some coherentists.

Disagreement affords humans as members of epistemic communities important opportunities for refining or improving their epistemic situations with respect to many of their beliefs. To get such epistemic gains, one needs to explore and gauge one’s own epistemic situation and the epistemic situations of others. Accordingly, a fitting response to disagreement regarding some matter, p, typically will turn on the resolution of two strongly interrelated questions: (1) whether p, and (2) why one’s interlocutor disagrees with oneself about p. When one has high intellectual respect for one’s interlocutor, answering question (2) involves arriving at a sympathetic explanatory understanding of the interlocutor’s own epistemic attitude toward p. Sorting out (2) is an abductive matter. Further, so far as the abductive explanation conditions one’s epistemic take regarding (1), there will be an abductive character to one’s epistemic position with respect to whether p—even where one’s initial purchase on whether p was not an abductive matter. We explain here how this can be managed naturally and tractably.

The argument I present here is an example of the manner in which naturalizing epistemology can help address fairly traditional epistemological issues. I develop one argument against coherentist epistemologies of empirical knowledge. In doing so, I draw on BonJour (1985), for that account seems to me to indicate the direction in which any plausible coherentist account would need to be developed, at least insofar as such accounts are to conceive of justification in terms of an agent (minimally) possessing articul able reasons and arguments, as is standard. I end by indicating important elements of coherentist epistemology that can be salvaged in the face of my argument, provided we are willing to drop the traditional commitment to characterizing justification in terms of the structure of articulable argument.

Naturalized epistemology is not a recent invention, nor is it a philosophical invention. Rather, it is a cognitive phenomena that is pervasive and desirable in the way of human epistemic engagement with their world. It is a matter of the way that one’s cognitive processes can be modulated by information gotten from those same or wider cognitive processes. Such modulational control enhances the reliability of one’s cognitive processes in many ways ‐ and judgments about objective epistemic justification consistently evince a reasonable demand for it. However, with suitable modulational control in place within an agent or a community of agents, the fitting cognitive processes take time to generate information that then engenders changes in processes and norms. Further, as there are significant historical and biographical contingencies involving trajectories through one’s environment, there are contingencies in the information and modifications that will be engendered by suitable modulational control. As a result, what makes for objectively justified belief at a time will vary ‐ as the fruits of suitable modulational control accrue over time. This is a moderate form of historicism about epistemic justification

This paper explores a position that combines contextualism regarding knowledge with the idea that the central point or purpose of the concept of knowledge is to feature in attributions that keep epistemic gate for contextually salient communities. After highlighting the main outlines and virtues of the suggested gate-keeping contextualism, two issues are pursued. First, the motivation for the view is clarified in a discussion of the relation between evaluative concepts and the purposes they serve. This clarifies why one's sense for the point of an evaluative concept ought to constrain and inform one's understanding of the concept. Second, the paper explores ways of avoiding a problem in the author's earlier development of gate-keeping contextualism. The initial development of the view opened the door to a form of skepticism that would hobble an important facet of our social-epistemic lives

The chapter rehearses the main outlines of gatekeeping contextualism—the view that it is central to the concept of knowledge that attributions of knowledge function in a kind of epistemic gatekeeping for contextually salient communities. The case for gatekeeping contextualism is clarified within an extended discussion of the character of philosophical reflection. The chapter argues that normatively valenced, evaluative concepts constitute a broad class of concepts for which a sociolinguistic point or purpose may be readily sensed—and for which the intimate connection between point/purpose and the concept itself is readily understandable. Commonly at the inception of philosophical reflection one’s sense for the point or purpose of a specific concept is seldom highly articulate. The chapter argues that this does not prevent one’s grasp of point/purpose from doing real work in philosophical reflection.

We respond to the central concerns raised by our commentators to our book, The Epistemological Spectrum. Casullo believes that our account of what we term “low-grade a priori” justification provides important clarification of a kind of philosophical reflection. However he objects to calling such reflection a priori. We explain what we think is at stake. Along the way, we comment on his idea of that there may be an epistemic payoff to making a distinction between assumptions and presumptions. In the book, we argued that an epistemically important form of nonaccidental reliability can be understood as a matter of processes being “transglobally reliable under modulational control.” Graham recommends another form of nonaccidental reliability, one rooted in evolutionary etiology. We explain why we think that the reliability of perceptual processes is best understood as turning of the kinds of modulational control that we highlight. We clarify how this approach represents a kind of reasonable epistemic patience—modulational control takes time, as it must turn on agents generating information about their own capacities and foibles. Lyons raises interesting questions regarding how (what we term) morphological content possessed by the agent can do the work that we set for it. We argue that it is necessary in order for agents to accommodate the background information that is relevant to many central problems of belief formation. We clarify how it can be expected to work