Nowadays the argument for the existence of God based on the common consent of mankind is taken to be so bad that contemporary atheists do not even bother to mention it. And it seems very few theists think that the argument is worth defending. In this paper I shall argue to the contrary: not only is the argument better than usually thought, but widespread belief in God constitutes a prima facie defeater for every reasonable atheist.

While it is widely acknowledged that knowledge can be acquired via testimony, it has been argued that understanding cannot. While there is no consensus about what the epistemic relationship of understanding consists in, I argue here that regardless of how understanding is conceived there are kinds of understanding that can be acquired through testimony: easy understanding and easy-s understanding. I address a number of aspects of understanding that might stand in the way of being able to acquire understanding through testimony, (...) focusing on understanding ’s paradigmatic form and what it means to say that in order to understand something you need to “grasp” some information or the relationship between bits of information. I argue that in cases of both easy and easy-s understanding, no aspect of understanding stands in the way of it being able to acquire it through testimony. As a result, while not all understanding be acquired through testimony in all instances and for all subjects, this failure of acquisition is only a product of the complexity of the relevant information or one’s unfamiliarity with it, and not a product of the epistemic relationship of understanding. (shrink)

One sometimes believes a proposition without grasping it. For example, a complete achromat might believe that ripe tomatoes are red without grasping this proposition. My aim in this paper is to shed light on the difference between merely believing a proposition and grasping it. I focus on two possible theories of grasping: the inferential theory, which explains grasping in terms of inferential role, and the phenomenal theory, which explains grasping in terms of phenomenal consciousness. I argue that the phenomenal theory (...) is more plausible than the inferential theory. (shrink)

There is a virtual consensus in contemporary epistemology that knowledge must be reliably produced. Everyone, it seems, is a reliabilist about knowledge in that sense. I present and defend two arguments that unreliable knowledge is possible. My first argument proceeds from an observation about the nature of achievements, namely, that achievements can proceed from unreliable abilities. My second argument proceeds from an observation about the epistemic efficacy of explanatory inference, namely, that inference to the best explanation seems to produce knowledge, (...) even if it isn't reliable. I also propose a successor to standard versions of reliabilism, which I call ‘ecumenical reliabilism’. Ecumenical reliabilism is consistent with unreliably produced knowledge and helps explain why unreliably produced knowledge is possible. (shrink)

Defenders of pragmatic encroachment in epistemology (or what I call practicalism) need to address two main problems. First, the view seems to imply, absurdly, that knowledge can come and go quite easily—in particular, that it might come and go along with our variable practical interests. We can call this the stability problem. Second, there seems to be no fully satisfying way of explaining whose practical interests matter. We can call this the “whose stakes?” problem. I argue that both problems can (...) be addressed in roughly the same terms. More exactly, I argue that by first clarifying the whose stakes? problem an answer to the stability problem naturally falls out. (shrink)

Although hypointellectuals have long been a part of our cultural landscape, it is in post-conflict societies, such as those in Bosnia and Kosovo, that there has arisen a strong need for a different breed of intellectual, one who is more than simply a social critic, an educator, a person of action, and a compassionate individual. Enter the non-partisan intellectual—the hyperintellectual. It is the hyperintellectual, whose non-partisanship is manifested through a reciprocating critique and defense of both the nationalist enterprise and strong (...) interventionism of the International Community, who strives to create a climate of understanding and to enlarge the moral space so as to reduce the divisiveness between opposing parties. It is in this way that the hyperintellectual acts as a catalyst for the creation of a democratic culture within the civil societies of Bosnia and Kosovo. (shrink)

The epistemological problem of the external world asks: (1) “How is knowledge of the external world possible given certain obstacles which make it look impossible?” This is a “how-possible?” question: it asks how something is possible given certain obstacles which make it look impossible (cf. Cassam 2007; Nozick 1981; Stroud 1984). Now consider the following question, which asks: (2) “How is a philosophically satisfying answer to (1) possible?” Skepticism is the thesis that knowledge of the external world is impossible. It (...) therefore represents a negative answer to the first question. Meta-epistemological skepticism is the thesis that a satisfying philosophical explanation of how our knowledge of the external world is possible is itself impossible. It therefore represents a negative answer to the second question. In this thesis, I explore the prospects of meta-epistemological skepticism. In particular, I structure the thesis around two master arguments from for meta-epistemological skepticism. The first argument is that skepticism represents a kind of paradox which requires dispensing with a plausible principle or belief, or even 'platitude', which means that there's no philosophically satisfying response to the paradox (see, e.g., Stroud 1984), and the second argument is “Stroud’s dilemma” (cf. Cassam 2009), that either we explain how external world knowledge is possible by reference to what external world knowledge we have, thereby failing to answer the philosophical question adequately, or else we don't, but then we succumb to first-order skepticism. I argue that the first argument fails to provide adequate support for meta-epistemological skepticism. However, the second argument, Stroud’s dilemma, is plausible and withstands serious objections (e.g., from Sosa 1994, Williams 1996, and Cassam 2009). In short, while the worry emerging out of the 'skeptical paradox' view fails to adequately support meta-epistemological skepticism, Stroud’s dilemma succeeds. Fortunately, it doesn't show that we lack external world knowledge; we rather just can't explain, philosophically, how it's possible. We may have first-order knowledge without understanding how it's possible. This thesis therefore represents a partial defense of meta-epistemological skepticism. Meta-epistemological skepticism is a live option in epistemology. (shrink)

This paper offers an analysis of the nature of epistemic trust. With increased philosophical attention to social epistemology in general and testimony in particular, the role for an epistemic or intellectual version of trust has loomed large in recent debates. But, too often, epistemologists talk about trust without really providing a sustained examination of the concept. After some introductory comments, I begin by addressing various components key to trust simpliciter. In particular, I examine what we might think of when we (...) consider what it means to place trust in someone. Once we have examined trust in general, we can modify this discussion to derive an epistemic version of trust—placing trust in someone for an epistemic reason. I argue that ET includes four components: belief, communication, reliance, and confidence. The first two sets are distinctively epistemic and the second set of conditions form the core of any kind of trust. Put together,.. (shrink)

Orthodoxy has it that knowledge is a composite of belief and non-mental factors. However, Timothy Williamson suggests that orthodoxy implies that the concept of belief is acquired before the concept of knowledge, whereas developmental data suggest the reverse. More recently, Jennifer Nagel reviews the psychological evidence, building a psychological case that the concept of knowledge emerges prior to belief. I assess the psychological state of the art and find support for the opposite conclusion. Overall the empirical evidence supports the orthodox (...) view that the concept of belief is prior to the concept of knowledge. (shrink)

What is it that makes someone wise, or one person wiser than another? I argue that wisdom consists in knowledge of how to live well, and that this knowledge of how to live well is constituted by various further kinds of knowledge. One concern for this view is that knowledge is not needed for wisdom but rather some state short of knowledge, such as having rational or justified beliefs about various topics. Another concern is that the emphasis on knowing how (...) to live well fails to do justice to the ancient tradition of ‘theoretical wisdom’. I address both of these concerns in filling out the account. (shrink)

Leading virtue epistemologists defend the view that knowledge must proceed from intellectual virtue and they understand virtues either as refned character traits cultivated by the agent over time through deliberate effort, or as reliable cognitive abilities. Philosophical situationists argue that results from empirical psychology should make us doubt that we have either sort of epistemic virtue, thereby discrediting virtue epistemology’s empirical adequacy. I evaluate this situationist challenge and outline a successor to virtue epistemology: abilism . Abilism delivers all the main (...) benefts of virtue epistemology and is as empirically adequate as any theory in philosophy or the social sciences could hope to be. (shrink)

Regulative virtue epistemology argues that intellectual virtues can adjust and guide one’s epistemic actions as well as improve on the quality of the epistemic actions. For regulative virtue epistemologists, intellectual virtues can be cultivated to a higher degree; when the quality of intellectual virtue is better, the resulting quality of epistemic action is better. The intellectual virtues that regulative epistemologists talk about are character virtues (such as intellectual courage and open-mindedness) rather than faculty virtues (such as sight and hearing), since (...) they don’t think that faculty virtues could be cultivated. This article refers to Xunzi’s philosophy, explaining how a regulative faculty-based virtue epistemology is possible. If this explanation works, on the one hand, a new branch of contemporary virtue epistemology is shown, and, on the other hand, a clear theoretical framework of Xunzi’s epistemology is constructed. (shrink)

This paper offers a new account of reflective knowledge’s value, building on recent work on the epistemic norms of speech acts. Reflective knowledge is valuable because it licenses us to make guarantees and promises.

I review several theoretical and empirical developments relevant to assessing contemporary virtue epistemology’s theory of knowledge. What emerges is a leaner theory of knowledge that is more empirically adequate, better captures the ordinary conception of knowledge, and is ripe for cross-fertilization with cognitive science. I call this view abilism. Along the way I identify several topics for future research.

Jennifer Lackey challenges the sufficiency version of the knowledge-action principle, viz., that knowledge that p is sufficient to rationally act on p, by proposing a set of alleged counterexamples. Her aim is not only to attack the knowledge-action principle, but also to undermine an argument for subject-sensitive invariantism. Lackey holds that her examples are counterexamples to the sufficiency version of the knowledge-action principle because (a) S knows the proposition in question, but (b) it is not rational for S to act (...) on it. In this paper, first, I argue against (a) on intuitive and on theoretical grounds. Second, I point out that (b), even if combined with (a), is not sufficient to make for counterexamples to the knowledge-action principle of the relevant kind. Third, I offer two alternative explanations of the intuition Lackey relies on. If either one of them is right, (b) may not be satisfied in her examples. (shrink)

We distinguish between two categories of belief—thin belief and thick belief—and provide evidence that they approximate genuinely distinct categories within folk psychology. We use the distinction to make informative predictions about how laypeople view the relationship between knowledge and belief. More specifically, we show that if the distinction is genuine, then we can make sense of otherwise extremely puzzling recent experimental findings on the entailment thesis (i.e. the widely held philosophical thesis that knowledge entails belief). We also suggest that the (...) distinction can be applied to debates in the philosophy of mind and metaethics. (shrink)

Recently, Timothy Williamson has argued that considerations about margins of errors can generate a new class of cases where agents have justified true beliefs without knowledge. I think this is a great argument, and it has a number of interesting philosophical conclusions. In this note I’m going to go over the assumptions of Williamson’s argument. I’m going to argue that the assumptions which generate the justification without knowledge are true. I’m then going to go over some of the recent arguments (...) in epistemology that are refuted by Williamson’s work. And I’m going to end with an admittedly inconclusive discussion of what we can know when using an imperfect measuring device. (shrink)

An overview of contemporary debates and topics in epistemology.

In my remarks, I discuss Sosa's attempt to deal with the sceptical threat posed by dreaming. Sosa explores two replies to the problem of dreaming scepticism. First, he argues that, on the imagination model of dreaming, dreaming does not threaten the safety of our beliefs. Second, he argues that knowledge does not require safety, but a weaker condition which is not threatened by dreaming skepticism. I raise questions about both elements of his reply.

This paper provides a principled and elegant solution to the Gettier problem. The key move is to draw a general metaphysical distinction and conscript it for epistemological purposes. Section 1 introduces the Gettier problem. Sections 2–5 discuss instructively wrong or incomplete previous proposals. Section 6 presents my solution and explains its virtues. Section 7 answers the most common objection.

Some beliefs seem more significant than others. This paper suggests an approach to explaining this apparent fact. As there are multiple senses in which one belief may be more significant than another, multiple possible sources of such significance, and, moreover, no prima facie reason to expect a single, unified account under which all these senses and sources can be subsumed, I propose the modest approach of articulating just one feature in virtue of which a belief may fairly be called significant: (...) that of bearing a certain relation to human flourishing, a relation that more trivial truths do not bear. From such modest projects can a complete solution to the problem (or, more accurately, problems) of significance emerge. (shrink)

In a recent paper, Erasmus et al. (2021) defend the idea that the ambiguity of the term “explanation” in explainable AI (XAI) can be solved by adopting any of four different extant accounts of explanation in the philosophy of science: the Deductive Nomological, Inductive Statistical, Causal Mechanical, and New Mechanist models. In this chapter, I show that the authors’ claim that these accounts can be applied to deep neural networks as they would to any natural phenomenon is mistaken. I also (...) provide a more general argument as to why the notion of explainability as it is currently used in the XAI literature bears little resemblance to the traditional concept of scientific explanation. It would be more fruitful to use the label “understandable AI” to avoid the confusion that surrounds the goal and purposes of XAI. In the second half of the chapter, I argue for a pragmatic conception of understanding that is better suited to play the central role attributed to explanation in XAI. Following De Regt (2017) and Kuorikoski and Ylikoski (2015), the conditions of satisfaction for understanding an ML system are fleshed out in terms of an agent’s success in using the system. (shrink)

In Charles Mills’ essay, “White Ignorance,” and his trail-blazing monograph, The Racial Contract, he developed a view of how Whiteness or anti-Black-Indigenous-and-Latinx racism causes individuals to hold false beliefs or lack beliefs about racial injustice in particular and the world in general. I will defend a novel exegetical claim that Mills’ view is part of a more general view regarding how racial injustice can affect a subject’s epistemic standing such as whether they are justified in a belief and whether their (...) degree of confidence in the belief is rational given their evidence. Then, in light of this novel exegetical claim, I show how this interpretation of Mills’ view that racial injustice causes ignorance relates to how epistemically justified philosophers and social scientists are in the views that dominate their respective scholarly literature. (shrink)

What is the nature and structure of phronesis or practical wisdom? According to the view widely held by philosophers and psychologists, a person S is wise if and only if S knows how to live well. Given this view of practical wisdom, the guiding question is this: What exactly is “knowing how to live well”? It seems that no one has a clear idea of how to answer this simple but fundamental question. This paper explores knowing how to live well (...) (or “life know-how”) by showing how its nature and structure can be understood through contemporary epistemology of knowledge-how. I will achieve this by doing the following. In Section I, I highlight the two as-yet unanswered “integration questions” about life know-how. In Section II, I explain why the epistemology of knowledge-how has good potential to address the integration questions. In Sections III and IV, I construct two positions—intellectualism and anti-intellectualism—for the epistemology of life know-how and show how they address the two integration questions. In Section V, I show how the epistemology of life know-how established in the previous sections can be used in the philosophy of wisdom and the psychology of wisdom. (shrink)

This book offers a defence of Wrightean epistemic entitlement, one of the most prominent approaches to hinge epistemology. It also systematically explores the connections between virtue epistemology and hinge epistemology. -/- According to hinge epistemology, any human belief set is built within and upon a framework of pre-evidential propositions – hinges – that cannot be justified. Epistemic entitlement argues that we are entitled to trust our hinges. But there remains a problem. Entitlement is inherently unconstrained and arbitrary: We can be (...) entitled to any hinge proposition under the right circumstances. In this book, the author argues that we need a non-arbitrariness clause that protects entitlement from defeat. This clause, he argues, is to require epistemic virtue. Virtuous cognitive dispositions provide the non-arbitrariness clause that protects entitlement from defeat. The epistemic character of the agent who holds a particular set of hinges tells us something about the hinges’ epistemic status. Conversely, epistemic virtues are cognitive dispositions and capacities that rely on hinge propositions – without trusting in some hinges, we would be unable to exercise our virtues. -/- Trust Responsibly will appeal to scholars and advanced students working on epistemology, Wittgenstein, and virtues. (shrink)

Abilitism is an approach to the metaphysics of concepts according to which each concept consists of a managing cognitive ability coordinating other abilities (cognitive and non-cognitive) and a set of subordinate abilities associated with this managing ability. As I argue here, if we accept the abilitist approach, we can efficiently solve such puzzles in the metaphysics of concepts as the partial possession problem, the concept pluralism problem, etc. However, there are some possible objections to abilitism, concerning the abilitist explanation of (...) compositional properties of concepts, knowledge-that, an extension/intension of concepts, and the idea that concepts are constituents of thought. As I demonstrate here, these objections can be answered. (shrink)

This paper suggests that some climate denialism is bullshit. Those who spread it do not display a proper concern for the truth. This paper also shows that this bullshit is harmful in some significant ways. It undermines the epistemic demands imposed on us by what we care about, by the social roles we occupy, and by morality. It is also harmful because it corrodes epistemic trust.

Wir Menschen streben danach, die Wirklichkeit zu verstehen. Eine Welt, die wir gut verstehen, ist eine, die wir "im Griff" haben, mit der wir gut umgehen können. Aber was heißt es genau, ein Phänomen der Wirklichkeit zu verstehen? Wie sieht unser Weltbild aus, wenn wir ein Phänomen verstanden haben? Welche Bedingungen müssen erfüllt sein, damit Verstehen gelingt? Die Kernthese des Buches ist, dass wir Phänomene der Wirklichkeit durch noetische Integration verstehen. Wir verstehen Phänomene, indem wir den entsprechenden Informationseinheiten eine sinnvolle (...) und angemessene Position in unserem Weltbild zuschreiben und insofern unser Weltbild in gewissem Maße der Wirklichkeit entspricht. (shrink)

Some, if not most, philosophy program admissions committee members assume that they can determine that one applicant will likely manifest a higher degree of philosophical skill than another applicant on the basis of differences between their materials. I challenge this assumption by explaining how applicants’ materials in significant measure reflect the racially unjust environment in which they manifest their philosophical skill. I explain how applicants’ racial-group membership in similar measure determines what these materials consist in.

The initial successes in recent years in harnessing machine learning technologies to improve medical practice and benefit patients have attracted attention in a wide range of healthcare fields. Particularly, it should be achieved by providing automated decision recommendations to the treating clinician. Some hopes placed in such ML-based systems for healthcare, however, seem to be unwarranted, at least partially because of their inherent lack of transparency, although their results seem convincing in accuracy and reliability. Skepticism arises when the physician as (...) the agent responsible for the implementation of diagnosis, therapy, and care is unable to access the generation of findings and recommendations. There is widespread agreement that, generally, a complete traceability is preferable to opaque recommendations; however, there are differences about addressing ML-based systems whose functioning seems to remain opaque to some degree—even if so-called explicable or interpretable systems gain increasing amounts of interest. This essay approaches the epistemic foundations of ML-generated information specifically and medical knowledge generally to advocate differentiations of decision-making situations in clinical contexts regarding their necessary depth of insight into the process of information generation. Empirically accurate or reliable outcomes are sufficient for some decision situations in healthcare, whereas other clinical decisions require extensive insight into ML-generated outcomes because of their inherently normative implications. (shrink)

In this paper, I argue that hinge propositions are ways of acting that constitute abilities or skills. My starting point is Moyal-Sharrock's account of hinge propositions. However, Moyal-Sharrock's account leaves gaps to be filled, as it does not offer a unified explanation of the origin of our ungrounded grounds. Her account also lacks resources to respond to the issue of demarcation, since it does not provide a criterion for distinguishing ways of acting that can legitimately fulfill the role of ungrounded (...) grounds from those that cannot. Without an answer to this issue, the relativistic threat is serious. I then propose that by narrowing the ways of acting to those that are constitutive of abilities, we can deal with the relativistic threat. I provide an ecological approach to abilities through which I explain why abilities are reality-soaked and therefore why the ways of acting that constitute them are legitimate ungrounded grounds. Based on that approach, I provide an answer to the issue of demarcation that defuses the relativistic threat. (shrink)

Can understanding be transmitted by testimony, in the same sense that propositional knowledge can be transmitted by testimony? Some contemporary philosophers – call them testimonial understanding pessimists – say No, and others – call them testimonial understanding optimists – say Yes. In this chapter I will articulate testimonial understanding pessimism (§1) and consider some arguments for it (§2).

Why think that conscious experience of reality is any more epistemically valuable than testimony about it? I argue that conscious experience of reality is epistemically valuable because it provides cognitive contact with reality. Cognitive contact with reality is a goal of experiential inquiry which does not reduce to the goal of getting true beliefs or propositional knowledge. Such inquiry has awareness of the truth-makers of one’s true beliefs as its proper goal. As such, one reason why conscious experience of reality (...) is more epistemically valuable than testimony about reality is that it gives us more epistemic goods than only true belief or propositional knowledge. I defend this view from two rival accounts. First, that while conscious experience of reality has greater value than testimony, its value is only eudaimonic. Second, that while it has greater epistemic value than testimony, this value is not distinctive: for it only promotes truth better than testimony. (shrink)

Some philosophers (e.g., Pritchard, Grimm, and Hills) recently have objected that veritism cannot explain the epistemic value of understanding-why. And they have proposed two anti-veritist accounts. In this paper, I first introduce their objection and argue that it fails. Next, I consider a strengthened version of their objection and argue that it also fails. After that, I suggest a new veritist account: Understanding-why entails believing the truth that what is grasped is accurate (or accurate enough), and it is this true (...) belief, along with many other true beliefs understanding-why entails, that makes understanding-why finally epistemically valuable. Then, I explain why the two anti-veritist accounts are both false. Finally, I briefly discuss the idea that understanding involves a kind of know-how and show how veritism can explain the epistemic value of know-how in general. (shrink)

We are often urged to figure things out for ourselves rather than to rely on other people’s say-so, and thus be ‘epistemically autonomous’ in one sense of the term. But why? For almost any important question, there will be someone around you who is at least as well placed to answer it correctly. So why bother making up your own mind at all? I consider, and then reject, two ‘egoistic’ answers to this question according to which thinking for oneself is (...) beneficial for the autonomous agent herself. I go on to suggest that the reason we should (sometimes) think for ourselves is that doing so (sometimes) increases the collective reliability of the epistemic community to which we belong. In many cases, this will do nothing at all to increase our own chances of forming correct beliefs. So, at least in this respect, the rationale for being epistemically autonomous is entirely ‘altruistic’. (shrink)

In this thesis, I elaborate and defend Donald Davidson's account of knowing one's own mental states that exhibit first-person authority. To that end, I place Davidson's account among others in the philosophical landscape concerning self-knowledge. Next, I examine his response to philosophical challenges that arise from mental content externalism and self-deception. Finally, I draw some insights froms Davidson's account to the broader aims of epsitemology.

In this paper, I argue that hinge propositions are ways of acting that constitute abilities or skills. My starting point is Moyal-Sharrock's account of hinge propositions. However, Moyal-Sharrock's account leaves gaps to be filled, as it does not offer a unified explanation of the origin of our ungrounded grounds. Her account also lacks resources to respond to the issue of demarcation, since it does not provide a criterion for distinguishing ways of acting that can legitimately fulfill the role of ungrounded (...) grounds from those that cannot. Without an answer to this issue, the relativistic threat is serious. I then propose that by narrowing the ways of acting to those that are constitutive of abilities, we can deal with the relativistic threat. I provide an ecological approach to abilities through which I explain why abilities are reality-soaked and therefore why the ways of acting that constitute them are legitimate ungrounded grounds. Based on that approach, I provide an answer to the issue of demarcation that defuses the relativistic threat. (shrink)

We discuss the social-epistemic aspects of Catherine Elgin’s theory of reflective equilibrium and understanding and argue that it yields an argument for the view that a crucial social-epistemic function of epistemic authorities is to foster understanding in their communities. We explore the competences that enable epistemic authorities to fulfil this role and argue that among them is an epistemic virtue we call “epistemic empathy”.

The philosophical literature on well-being and the good life contains very little explicit discussion of what makes for a better or worse death. The purpose of this essay is to highlight some commonly held views about the good death and investigate whether these views are recognized by the leading theories of well-being. While the most widely discussed theories do have implications about what constitutes a good death, they seem unable to fully accommodate these popular good death views. I offer two (...) partial explanations for why these views have been neglected in discussions of well-being and make two corresponding recommendations for future work in the philosophy of well-being. (shrink)

A critical review of “Gettier” cases and theoretical attempts to solve “the” "Gettier" "problem".

Virtue epistemologists define knowledge as true belief produced by intellectual virtue. In this paper, I review how this definition fails in three important ways. First, it fails as an account of the ordinary knowledge concept, because neither belief nor reliability is essential to knowledge ordinarily understood. Second, it fails as an account of the knowledge relation itself, insofar as that relation is operationalized in the scientific study of cognition. Third, it serves no prescriptive purpose identified up till now. An alternative (...) theory, abilism, provides a superior account of knowledge as it is ordinarily and scientifically understood. According to abilism, knowledge is an accurate representation produced by cognitive ability. (shrink)

In this paper I argue that the search for explainable models and interpretable decisions in AI must be reformulated in terms of the broader project of offering a pragmatic and naturalistic account of understanding in AI. Intuitively, the purpose of providing an explanation of a model or a decision is to make it understandable to its stakeholders. But without a previous grasp of what it means to say that an agent understands a model or a decision, the explanatory strategies will (...) lack a well-defined goal. Aside from providing a clearer objective for XAI, focusing on understanding also allows us to relax the factivity condition on explanation, which is impossible to fulfill in many machine learning models, and to focus instead on the pragmatic conditions that determine the best fit between a model and the methods and devices deployed to understand it. After an examination of the different types of understanding discussed in the philosophical and psychological literature, I conclude that interpretative or approximation models not only provide the best way to achieve the objectual understanding of a machine learning model, but are also a necessary condition to achieve post hoc interpretability. This conclusion is partly based on the shortcomings of the purely functionalist approach to post hoc interpretability that seems to be predominant in most recent literature. (shrink)

It is widely accepted in epistemology that knowledge is factive, meaning that only truths can be known. We argue that this theory creates a skeptical challenge: because many of our beliefs are only approximately true, and therefore false, they do not count as knowledge. We consider several responses to this challenge and propose a new one. We propose easing the truth requirement on knowledge to allow approximately true, practically adequate representations to count as knowledge. In addition to addressing the skeptical (...) challenge, this view also coheres with several previous theoretical proposals in epistemology. (shrink)

While social epistemologists have recently begun addressing questions about whether groups can possess beliefs or knowledge, little has yet been said about whether groups can properly be said to possess understanding. Here I want to make some progress on this question by considering two possible accounts of group understanding, modeled on accounts of group belief and knowledge: a deflationary account, according to which a group understands just in case most or all of its members understand, and an inflationary account, according (...) to which a group’s understanding does not depend solely on whether its members understand. I argue that both accounts face problems. The deflationary account has two such problems: aggregation problems that are familiar from discussions of group belief, and the problem of different bases, wherein members possess understanding for different but consistent reasons. The inflationary account faces what I call the problem of distributed grasping: while it is widely accepted that understanding requires a kind of “grasping”, it is hard to make sense of how this requirement could be met at the group level while not necessarily being met by any individual member. Despite its problems, I make a case for the inflationary account. This will require addressing the problem of distributed grasping: to do this, I propose a different way of thinking about the grasping relation at the group level, such that it is constituted by a dependency relationship between members. (shrink)

I present a puzzle – the “puzzle of aesthetic testimony” – along with a solution to it that appeals to the impossibility of testimonial understanding. I'll criticize this solution by defending the possibility of testimonial understanding, including testimonial aesthetic understanding.