This paper defends strict invariantism against some philosophical and empirical data that have been taken to compromise it. The defence involves a combination of a priori philosophical arguments and empirically informed theorizing. The positive account of the data is an epistemic focal bias account that draws on cognitive psychology. It involves the assumption that, owing to limitations of the involved cognitive resources, intuitive judgments about knowledge ascriptions are generated by processing only a limited part of the available (...) information—the part that is in focus. According to the epistemic focal bias account, the intuitive judgments about knowledge ascriptions that constitute contrast effects amount to false positives, whereas the intuitive judgments that constitute salient alternatives effects amount to false negatives. I conclude by considering how the basic epistemic focal bias account may be developed further by reference to relevant alternatives theory in epistemology, pragmatics in the philosophy of language, and dual process theory in cognitive psychology. (shrink)

In this response to Michael Blome-Tillmann's comments on On Folk Epistemology, I defend the book's epistemic focal bias account of the salient alternatives effect that Blome-Tillmann takes to motivate epistemic contextualism. First, I defend the epistemic focal bias account against Blome-Tillmann's criticism that it is insufficiently general insofar as it fails to account for a range of cases. Second, I defend the epistemic focal bias account from Blome-Tillmann's charge that it (...) overgeneralizes insofar as it can ‘explain away the context-sensitivity of gradable adjectives.’ Along the way, I discuss the metaphilosophical principles that guide the epistemic focal bias account. (shrink)

I develop an epistemic focal bias account of certain patterns of judgments about knowledge ascriptions by integrating it with a general dual process framework of human cognition. According to the focal bias account, judgments about knowledge ascriptions are generally reliable but systematically fallible because the cognitive processes that generate them are affected by what is in focus. I begin by considering some puzzling patters of judgments about knowledge ascriptions and sketch how a basic focal (...) bias account seeks to account for them. In doing so, I argue that the basic focal bias account should be integrated in a more general framework of human cognition. Consequently, I present some central aspects of a prominent general dual process theory of human cognition and discuss how focal bias may figure at various levels of processing. On the basis of this discussion, I attempt to categorize the relevant judgments about knowledge ascriptions. Given this categorization, I argue that the basic epistemic focal bias account of certain contrast effects and salient alternatives effects can be plausibly integrated with the dual process framework. Likewise, I try to explain the absence of strong intuitions in cases of far-fetched salient alternatives. As a manner of conclusion, I consider some methodological issues concerning the relationship between cognitive psychology, experimental data and epistemological theorizing. -/- . (shrink)

Solomon has made the case, in Social Empicism (2001) for socially naturalized analysis of the dynamics of scientific inquiry that takes seriously two critical insights: that scientific rationality is contingent, disunified, and socially emergent; and that scientific progress is often fostered by factors traditionally regarded as compromising sources of bias. While elements of this framework are widely shared, Solomon intends it to be more resolutely social, more thoroughly naturalizing, and more ambitiously normative than other contextualizing epistemologies currently on offer. (...) Four focal issues are addressed in the commentaries that follow: Solomon's characterization of empirical success as a goal of science (Clough); her distinction between empirical and non-empirical decision vectors and the viability of the multivariate analysis she proposes for assessing epistemic fairness in their distribution (Clough; Richardson); the plausibility of her thesis that normatively appropriate consensus is a (rare) limiting case rather than an intrinsically desirable outcome of inquiry (Oreskes; Richardson); and her conviction that a socially naturalized analysis of science can ground norms of scientific rationality (Longino; Oreskes). (shrink)

It has been argued that implicit biases are operative in philosophy and lead to significant epistemic costs in the field. Philosophers working on this issue have focussed mainly on implicit gender and race biases. They have overlooked ideological bias, which targets political orientations. Psychologists have found ideological bias in their field and have argued that it has negative epistemic effects on scientific research. I relate this debate to the field of philosophy and argue that if, as (...) some studies suggest, the same bias also exists in philosophy then it will lead to hitherto unrecognised epistemic hazards in the field. Furthermore, the bias is epistemically different from the more familiar biases in respects that are important for epistemology, ethics, and metaphilosophy. (shrink)

In this chapter, we explore whether agents have an epistemic duty to eradicate implicit bias. Recent research shows that implicit biases are widespread and they have a wide variety of epistemic effects on our doxastic attitudes. First, we offer some examples and features of implicit biases. Second, we clarify what it means to have an epistemic duty, and discuss the kind of epistemic duties we might have regarding implicit bias. Third, we argue that we (...) have an epistemic duty to eradicate implicit biases that have negative epistemic impact. Finally, we defend this view against the objection that we lack the relevant control over implicit bias that’s required for such a duty. We argue that we have a kind of reflective control over the implicit biases that we are duty-bound to eradicate. And since, as we show, we have this control over a wide variety of implicit biases, there are a lot of implicit biases that we have epistemic duties to eradicate. (shrink)

In this paper I explore the nature of confabulatory explanations of action guided by implicit bias. I claim that such explanations can have significant epistemic benefits in spite of their obvious epistemic costs, and that such benefits are not otherwise obtainable by the subject at the time at which the explanation is offered. I start by outlining the kinds of cases I have in mind, before characterising the phenomenon of confabulation by focusing on a few common features. (...) Then I introduce the notion of epistemic innocence to capture the epistemic status of those cognitions which have both obvious epistemic faults and some significant epistemic benefit. A cognition is epistemically innocent if it delivers some epistemic benefit to the subject which would not be attainable otherwise because alternative (less epistemically faulty) cognitions that could deliver the same benefit are unavailable to the subject at that time. I ask whether confabulatory explanations of actions guided by implicit bias have epistemic benefits and whether there are genuine alternatives to forming a confabulatory explanation in the circumstances in which subjects confabulate. On the basis of my analysis of confabulatory explanations of actions guided by implicit bias, I argue that such explanations have the potential for epistemic innocence. I conclude that epistemic evaluation of confabulatory explanations of action guided by implicit bias ought to tell a richer story, one which takes into account the context in which the explanation occurs. (shrink)

It is often argued that while biases routinely influence the generation of scientific theories, a subsequent rational evaluation of such theories will ensure that biases do not affect which theories are ultimately accepted. Against this line of thought, this paper shows that the existence of certain kinds of biases at the generation-stage implies the existence of biases at the evaluation-stage. The key argumentative move is to recognize that a scientist who comes up with a new theory about some phenomena has (...) thereby gained an unusual type of evidence, viz. information about the space of theories that could be true of the phenomena. It follows that if there is bias in the generation of scientific theories in a given domain, then the rational evaluation of theories with reference to the total evidence in that domain will also be biased. (shrink)

It has been argued that humans can face an ethical/epistemic dilemma over the automatic stereotyping involved in implicit bias: ethical demands require that we consistently treat people equally, as equally likely to possess certain traits, but if our aim is knowledge or understanding our responses should reflect social inequalities meaning that members of certain social groups are statistically more likely than others to possess particular features. I use psychological research to argue that often the best choice from the (...) epistemic perspective is the same as the best choice from the ethical perspective: to avoid automatic stereotyping even when this involves failing to reflect social realities in our judgements. This argument has an important implication: it shows that it is not possible to successfully defend an act of automatic stereotyping simply on the basis that the stereotype reflects an aspect of social reality. An act of automatic stereotyping can be poor from an epistemic perspective even if the stereotype that is activated reflects reality. (shrink)

We report and discuss the results of a series of experiments that address a contrast effect exhibited by folk judgments about knowledge ascriptions. The contrast effect, which was first reported by Schaffer and Knobe, is an important aspect of our folk epistemology. However, there are competing theoretical accounts of it. We shed light on the various accounts by providing novel empirical data and theoretical considerations. Our key findings are, firstly, that belief ascriptions exhibit a similar contrast effect and, secondly, that (...) the contrast effect is systematically sensitive to the content of what is in contrast. We argue that these data pose significant challenges to contrastivist accounts of the contrast effect. Furthermore, some of the data set provides, in conjunction with some non-empirical epistemological arguments, some limited evidence for what we call a focal bias account of the data. According to the focal bias account, the contrast effects arise at least in part because epistemically relevant facts are not always adequately processed when they are presented in certain ways. (shrink)

Motivating reforms to address discrimination and exclusion is important. But what epistemic practices characterize better or worse ways of doing this? Recently, the phenomena of implicit biases have played a large role in motivating reforms. We argue that this strategy risks perpetuating two kinds of epistemic oppression: the vindication dynamic and contributory injustice. We offer positive proposals for avoiding these forms of epistemic oppression when confronting racism.

According to Russell and Norvig’s (2009) classification, Artificial Intelligence (AI) is the field that aims at building systems which either think rationally, act rationally, think like humans, or...

BackgroundThere is a growing concern about artificial intelligence (AI) applications in healthcare that can disadvantage already under-represented and marginalised groups (eg, based on gender or race).ObjectivesOur objectives are to canvas the range of strategies stakeholders endorse in attempting to mitigate algorithmic bias, and to consider the ethical question of responsibility for algorithmic bias.MethodologyThe study involves in-depth, semistructured interviews with healthcare workers, screening programme managers, consumer health representatives, regulators, data scientists and developers.ResultsFindings reveal considerable divergent views on three key (...) issues. First, views on whether bias is a problem in healthcare AI varied, with most participants agreeing bias is a problem (which we call the bias-critical view), a small number believing the opposite (the bias-denial view), and some arguing that the benefits of AI outweigh any harms or wrongs arising from the bias problem (the bias-apologist view). Second, there was a disagreement on the strategies to mitigate bias, and who is responsible for such strategies. Finally, there were divergent views on whether to include or exclude sociocultural identifiers (eg, race, ethnicity or gender-diverse identities) in the development of AI as a way to mitigate bias.Conclusion/significanceBased on the views of participants, we set out responses that stakeholders might pursue, including greater interdisciplinary collaboration, tailored stakeholder engagement activities, empirical studies to understand algorithmic bias and strategies to modify dominant approaches in AI development such as the use of participatory methods, and increased diversity and inclusion in research teams and research participant recruitment and selection. (shrink)

Some recent work in philosophy of religion addresses what can be called the “axiological question,” i.e., regardless of whether God exists, would it be good or bad if God exists? Would the existence of God make the world a better or a worse place? Call the view that the existence of God would make the world a better place “Pro-Theism.” We argue that Pro-Theism is not implausible, and moreover, many Theists, at least, (often implicitly) think that it is true. That (...) is, many Theists think that various good outcomes would arise if Theism is true. We then discuss work in cognitive science concerning human cognitive bias, before discussing two noteworthy attempts to show that at least some religious beliefs arise because of cognitive bias: Hume’s, and Draper’s and Nichols’s. We then argue that, as a result of certain cognitive biases that result when good outcomes might be at stake, Pro-Theism causes many Theists to inflate the epistemic probability that God exists, and as a result, Theists should lower the probability they assign to God’s existence. Finally, based our arguments, we develop a novel objection to Pascal’s wager. (shrink)

A topic of special importance when it comes to responsibility and implicit bias is responsibility for knowledge. Are there strategies for becoming more responsible and respectful knowers? How might we work together, not just as individuals but members of collectives, to reduce the negative effects of bias on what we see and believe, as well as the wrongs associated with epistemic injustice? To explore these questions, Chapter 9 introduces the concept of epistemic responsibility, a set of (...) practices developed through the cultivation of basic epistemic virtues, such as open-mindedness, epistemic humility, and diligence that help knowers seek information about themselves, others, and the world. (shrink)

Philosophers of science have argued that epistemic diversity is an asset for the production of scientific knowledge, guarding against the effects of biases, among other advantages. The growing privatization of scientific research, on the contrary, has raised important concerns for philosophers of science, especially with respect to the growing sources of biases in research that it seems to promote. Recently, Holman and Bruner ( 2017 ) have shown, using a modified version of Zollman ( 2010 ) social network model, (...) that an industrial selection bias can emerge in a scientific community, without corrupting any individual scientist, if the community is epistemically diverse. In this paper, we examine the strength of industrial selection using a reinforcement learning model, which simulates the process of industrial decision-making when allocating funding to scientific projects. Contrary to Holman and Bruner’s model, in which the probability of success of the agents when performing an action is given a priori, in our model the industry learns about the success rate of individual scientists and updates the probability of success on each round. The results of our simulations show that even without previous knowledge of the probability of success of an individual scientist, the industry is still able to disrupt scientific consensus. In fact, the more epistemically diverse the scientific community, the easier it is for the industry to move scientific consensus to the opposite conclusion. Interestingly, our model also shows that having a random funding agent seems to effectively counteract industrial selection bias. Accordingly, we consider the random allocation of funding for research projects as a strategy to counteract industrial selection bias, avoiding commercial exploitation of epistemically diverse communities. (shrink)

Biased research occurs frequently in the sciences. In this paper, I will focus on one particular kind of biased research: research that is subject to sponsorship bias. I will address the following epistemological question: what precisely is epistemically wrong with biased research of this kind? I will defend the evidential account of epistemic wrongness: that is, research affected by sponsorship bias is epistemically wrong if and only if the researchers in question make false claims about the evidential (...) support of some hypothesis H by data E. I will argue that the evidential account captures the epistemic wrongness of three paradigmatic types of sponsorship bias. (shrink)

Implicit bias results from living in a society structured by race. Tamar Gendler has drawn attention to several epistemic costs of implicit bias and concludes that paying some costs is unavoidable. In this paper, we reconstruct Gendler’s argument and argue that the epistemic costs she highlights can be avoided. Though epistemic agents encode discriminatory information from the environment, not all encoded information is activated. Agents can construct local epistemic environments that do not activate biasing (...) representations, effectively avoiding the consequences of activation. We conclude that changing our local environments provides a way to avoid paying implicit bias’s epistemic costs. (shrink)

Can implicit biases be properly thought of as epistemic vices? I start by sketching the contours of implicit biases (1), and then turn to the recent claim, from Cassam, that implicit biases are epistemic vices (2). However, I argue that concerns about the stability of implicit biases and their role in producing behavior make for difficulties in establishing that implicit biases of individuals are epistemic vices (3). I then consider a recently developed model which prompts us to (...) consider implicit biases as properties of groups (4). This raises the question of whether implicit biases might constitute collective epistemic vice. I suggest that there is a way to make sense of this claim, but it requires rethinking how we conceptualise collective epistemic vice (5). These re-conceptualisations can be independently motivated. I close by marshalling some considerations in favour of using the terminology of vice to capture these defects of collective epistemic practice (6). (shrink)

Although the impact of so-called “sponsorship bias” has been the subject of increased attention in the philosophy of science, what exactly constitutes its epistemic wrongness is still debated. In this paper, I will argue that neither evidential accounts nor social–epistemological accounts can fully account for the epistemic wrongness of sponsorship bias, but there are good reasons to prefer social–epistemological to evidential accounts. I will defend this claim by examining how both accounts deal with a paradigm case (...) from medical epistemology, recently discussed in a paper by Bennett Holman. I will argue that evidential accounts cannot adequately capture cases of sponsorship bias that involve the manufacturing of certainty because of their neutrality with respect to the role of non-epistemic values in scientific practice. If my argument holds, it further highlights the importance of integrating social and ethical concerns into epistemological analysis, especially in applied contexts. One can only properly grasp sponsorship bias as an epistemological problem if one resists the methodological tendency to analyze social, ethical, and epistemological issues in isolation from each other. (shrink)

It has recently been argued that beliefs formed on the basis of implicit biases pose a challenge for accessibilism, since implicit biases are consciously inaccessible, yet they seem to be relevant to epistemic justification. Recent empirical evidence suggests, however, that while we may typically lack conscious access to the source of implicit attitudes and their impact on our beliefs and behaviour, we do have access to their content. In this paper, I discuss the notion of accessibility required for this (...) argument to work vis-à-vis these empirical results and offer two ways in which the accessibilist could meet the challenge posed by implicit biases. Ultimately both strategies fail, but the way in which they do, I conclude, reveals something general and important about our epistemic obligations and about the intuitions that inform the role of implicit biases in accessibilist justification. (shrink)

In this paper I discuss and develop the risk account of scientific objectivity, which I have recently introduced, contrasting it to some alternatives. I then use the account in order to analyse a practice that is relatively common in anthropology, in the history of science, and in the sociology of scientific knowledge: withholding epistemic judgement. I argue that withholding epistemic judgement on the beliefs one is studying can be a relatively efficient strategy against collective bias in these (...) fields. However, taking into account the criticisms presented against the strategy, I also argue that it is a usable strategy only when the distance between the researchers and their ideas, and the people and ideas being studied, is significant enough. (shrink)

A plausible view in the literature on the epistemic condition of blameworthiness is the Reasonable Expectation View (RE). According to RE, whether ignorance excuses an agent from deserving blame is a matter of whether the agent could have reasonably been expected to have avoided or corrected the ignorance. This paper does not take up the task of defending this view, but instead examines what it implies for an interesting type of ignorance: moral or political ignorance rooted in myside (...) class='Hi'>bias. With the prevalence of increasing political polarization, it is worth examining what a plausible view like RE implies about when, if ever, myside bias-based ignorance excuses the agent from blame—from deserving resentment or indignation. To assess this issue, the paper examines the empirical literature on what it takes to mitigate myside bias and then takes the United States as a case study to examine what RE implies for the actual non-ideal circumstances many agents are in. The paper argues that RE has two revisionary implications for our practice of blame. First, the paper argues that RE implies that political or moral ignorance rooted in myside bias is currently an excuse in a surprising number of cases. Second, our epistemic position to know whether a given instance of political or moral ignorance is excusing is often inadequate. (shrink)

Recent empirical research has substantiated the finding that very many of us harbour implicit biases: fast, automatic, and difficult to control processes that encode stereotypes and evaluative content, and influence how we think and behave. Since it is difficult to be aware of these processes - they have sometimes been referred to as operating 'unconsciously' - we may not know that we harbour them, nor be alert to their influence on our cognition and action. And since they are difficult to (...) control, considerable work is required to prevent their influence. We here focus on the implications of these findings for epistemology. We first look at ways in which implicit biases thwart our knowledge seeking practices (sections 1 & 2). Then we set out putative epistemic benefits of implicit bias, before considering ways in which epistemic practices might be improved (section 3). Finally, we consider the distinctive challenges that the findings about implicit bias pose to us as philosophers, in the context of feminist philosophy in particular (section 4). (shrink)

In Epistemic injustice, Miranda Fricker makes a tremendous contribution to theorizing the intersection of social epistemology with theories of justice. Theories of justice often take as their object of assessment either interpersonal transactions (specific exchanges between persons) or particular institutions. They may also take a more comprehensive perspective in assessing systems of institutions. This systemic perspective may enable control of the cumulative effects of millions of individual transactions that cannot be controlled at the individual or institutional levels. This is (...) true not only with respect to the overall distribution of such goods as income and wealth, but also with respect to the goods of testimonial and hermeneutical justice. Cognitive biases that may be difficult for even epistemically virtuous individuals to correct on their own may be more susceptible to correction if we focus on the principles that should govern our systems of testimonial gathering and assessment. Hence, while Fricker’s focus on individual epistemic virtue is important, we also need to consider what epistemic justice as a virtue of social systems would require. My paper will indicate some directions forward on this front, focusing on the need for integration of diverse institutions and persons engaged in inquiry. (shrink)

Lisa Bortolotti argues that some irrational beliefs are epistemically innocent and deliver significant epistemic benefits that could not be easily attained otherwise. While the benefits of the irrational belief may not outweigh the costs, epistemic innocence helps to clarify the epistemic and psychological effects of irrational beliefs on agency.

It is well known that AI-based language technology—large language models, machine translation systems, multilingual dictionaries, and corpora—is currently limited to three percent of the world’s most widely spoken, financially and politically backed languages. In response, recent efforts have sought to address the “digital language divide” by extending the reach of large language models to “underserved languages.” We show how some of these efforts tend to produce flawed solutions that adhere to a hard-wired representational preference for certain languages, which we call (...) language modeling bias. Language modeling bias is a specific and under-studied form of linguistic bias were language technology by design favors certain languages, dialects, or sociolects with respect to others. We show that language modeling bias can result in systems that, while being precise regarding languages and cultures of dominant powers, are limited in the expression of socio-culturally relevant notions of other communities. We further argue that at the root of this problem lies a systematic tendency of technology developer communities to apply a simplistic understanding of diversity which does not do justice to the more profound differences that languages, and ultimately the communities that speak them, embody. Drawing on the concept of epistemic injustice, we point to the broader ethico-political implications and show how it can lead not only to a disregard for valuable aspects of diversity but also to an under-representation of the needs of marginalized language communities. Finally, we present an alternative socio-technical approach that is designed to tackle some of the analyzed problems. (shrink)

Recent philosophy has paid considerable attention to the way our biases are liable to encroach upon our cognitive lives, diminishing our capacity to know and unjustly denigrating the knowledge of others. The extent of the bias, and the range of domains to which it applies, has struck some as so great as to license talk of a new form of skepticism. I argue that these depressing consequences are real and, in some ways, even more intractable than has previously been (...) recognized. For the difficulties we face in this domain are fueled not only by implicit biases but by various other sorts of entrenched cognitive attitudes we bear toward others, whether or not we judge them to be our peers. Inasmuch as the epistemic standing of this broader set of attitudes is itself quite dubious, the problem of epistemic injustice turns out to be just one special case—albeit of a particularly nasty kind—from a broader domain of cases where the collaborative character of knowledge clashes with tendencies that make collaboration difficult. This makes the threat of skepticism all the greater, and at the same time makes it harder to see what path of escape there might be. (shrink)

When interests and preferences of researchers or their sponsors cause bias in experimental design, data interpretation or dissemination of research results, we normally think of it as an epistemic shortcoming. But as a result of the debate on science and values, the idea that all extra-scientific influences on research could be singled out and separated from pure science is now widely believed to be an illusion. I argue that nonetheless, there are cases in which research is rightfully regarded (...) as epistemologically deficient due to the influence of preferences on its outcomes. I present examples from biomedical research and offer an analysis in terms of social epistemology. (shrink)

This paper argues that underlying social biases are able to affect the processes underlying linguistic interpretation. The result is a series of harms systematically inflicted on marginalised speakers. It is also argued that the role of biases and stereotypes in interpretation complicates Miranda Fricker's proposed solution to epistemic injustice.

I argue that evidentialism should be rejected because it cannot be reconciled with empirical work on bias in cognitive and social psychology.

Research on bias in peer review examines scholarly communication and funding processes to assess the epistemic and social legitimacy of the mechanisms by which knowledge communities vet and self-regulate their work. Despite vocal concerns, a closer look at the empirical and methodological limitations of research on bias raises questions about the existence and extent of many hypothesized forms of bias. In addition, the notion of bias is predicated on an implicit ideal that, once articulated, raises (...) questions about the normative implications of research on bias in peer review. This review provides a brief description of the function, history, and scope of peer review; articulates and critiques the conception of bias unifying research on bias in peer review; characterizes and examines the empirical, methodological, and normative claims of bias in peer review research; and assesses possible alternatives to the status quo. We close by identifying ways to expand conceptions and studies of bias to countenance the complexity of social interactions among actors involved directly and indirectly in peer review. (shrink)

Epistemologists spend a great deal of time thinking about how we should respond to our evidence. They spend far less time thinking about the ways that evidence can be acquired in the first place. This is an oversight. Some ways of acquiring evidence are better than others. Many normative epistemologies struggle to accommodate this fact. In this article I develop one that can and does. I identify a phenomenon – epistemic feedback loops – in which evidence acquisition has gone (...) awry, with the result that even beliefs based on the evidence are irrational. Examples include evidence acquired under the influence of confirmation bias and evidence acquired under the influence of cognitively penetrated experiences caused by implicit bias. I then develop a theoretical framework which enables us to understand why beliefs that are the outputs of epistemic feedback loops are irrational. Finally, I argue that many popular approaches to epistemic normativity may need to be abandoned on the grounds that they cannot comfortably explain feedback loops. The scope of this last claim is broad: it includes almost all contemporary theories of justified/rational belief and of the epistemology of cognitive penetration. (shrink)

In recent work, Mark Alfano and Jennifer Saul have put forward a similar kind of provocative sceptical challenge. Both appeal to recent literature in empirical psychology to show that our judgments across a wide range of cases are riddled with unreliable cognitive heuristics and biases. Likewise, they both conclude that we know a lot less than we have hitherto supposed, at least on standard conceptions of what knowledge involves. It is argued that even if one grants the empirical claims that (...) Saul and Alfano make, the sceptical conclusion that they canvass might not be as dramatic as it first appears. It is further argued, however, that one can reinstate a more dramatic sceptical conclusion by targeting their argument not at knowledge but rather at the distinct epistemic standing of understanding. (shrink)

This article presents Epistemic Spatialization as a new framework for investigating the interconnected patterns of biases when identifying objects with convolutional neural networks. It draws upon Foucault’s notion of spatialized knowledge to guide its method of enquiry. We argue that decisions involved in the creation of algorithms, alongside the labeling, ordering, presentation, and commercial prioritization of objects, together create a distorted “nomination of the visible”: they harden the visibility of some objects, make other objects excessively visible, and consign yet (...) others to permanent or haphazard invisibility. Our approach differs from those who focus on high-stakes misidentifications, such as errors tied to structural racism. Examining the far more dominant series of low-stakes mistakes shows the scope of errors, destabilizing the goal of image content identification with considerable societal impact. We explore these issues by closely examining the demonstration video of a popular convnet. This examination reveals an interlocking series of biases undermining the content identification process. The picture we paint is crucial for a better understanding of the errors that result as these convnets become further embedded in everyday products. The framework is valuable for critical work on computer vision, AI studies, and large-scale visual analysis. (shrink)

At the University of Sheffield during 2011 and 2012, a leading group of philosophers, psychologists, and others gathered to explore the nature and significance of implicit bias. The two volumes of Implicit Bias and Philosophy emerge from these workshops. Each volume philosophically examines core areas of psychological research on implicit bias as well as the ramifications of implicit bias for core areas of philosophy. Volume I: Metaphysics and Epistemology is comprised of two parts: “The Nature of (...) Implicit Attitudes, Implicit Bias, and Stereotype Threat,” and “Skepticism, Social Knowledge, and Rationality.” The first part contains chapters examining the relationship between implicit attitudes and “dual process” models of the mind; the role of affect in the formation and change of implicit associations; the unity (or disunity) of implicit attitudes; whether implicit biases are mental states at all; and whether performances on stereotype-relevant tasks are automatic and unconscious or intentional and strategic. The second part contains chapters examining implicit bias and skepticism; the effects of implicit bias on scientific research; the accessibility of social stereotypes in epistemic environments; the effects of implicit bias on the self-perception of members of stigmatized social groups as rational agents; the role of gender stereotypes in philosophy; and the role of heuristics in biased reasoning. (shrink)

Empirical psychology documents widespread evidence of bias in the ways that people select, interpret, and selectively interpret evidence in forming and revising their beliefs. These biases can function to create and perpetuate epistemologies of ignorance. I argue that virtue epistemology can help us explain what goes epistemically wrong in these cases, and can offer positive advice, orienting us toward ways to right it. In particular, I defend the virtue approach from epistemic situationist worries about the empirical plausibility of (...) individual agents' ability to develop corrective virtues. I argue that situational virtues are genuine responsibilist virtues, and are one of the tools at our disposal for overcoming epistemologies of ignorance. (shrink)

I argue that Kotzee’s model of meta- debate succeeds in identifying illegitimate or fallacious charges of bias but has the unintended consequence of classifying some legitimate and non-fallacious charges as fallacious. This makes the model, in some important cases, counter-productive. In particular, cases where the call for a meta- debate is prompted by the participant with epistemic privilege and a charge of bias is denied by the participant with social advantage, the impasse will put the epistemically advantaged (...) at far greater risk. Therefore, I propose treating epistemic privilege as a variety of expert opinion specifically in cases where meta- debate participants come to an impasse in deliberation. My proposal exposes the problem of interpreting debate contexts as both adversarial and free from social power differentials. (shrink)

Artificial intelligence (AI) systems are increasingly adopted to make decisions in domains such as business, education, health care, and criminal justice. However, such algorithmic decision systems can have prevalent biases against marginalized social groups and undermine social justice. Explainable artificial intelligence (XAI) is a recent development aiming to make an AI system’s decision processes less opaque and to expose its problematic biases. This paper argues against technical XAI, according to which the detection and interpretation of algorithmic bias can be (...) handled more or less independently by technical experts who specialize in XAI methods. Drawing on resources from feminist epistemology, we show why technical XAI is mistaken. Specifically, we demonstrate that the proper detection of algorithmic bias requires relevant interpretive resources, which can only be made available, in practice, by actively involving a diverse group of stakeholders. Finally, we suggest how feminist theories can help shape integrated XAI: an inclusive social-epistemic process that facilitates the amelioration of algorithmic bias. (shrink)

Sandra Harding's feminist standpoint epistemology makes two claims. The thesis of epistemic privilege claims that unprivileged social positions are likely to generate perspectives that are “less partial and less distorted” than perspectives generated by other social positions. The situated knowledge thesis claims that all scientific knowledge is socially situated. The bias paradox is the tension between these two claims. Whereas the thesis of epistemic privilege relies on the assumption that a standard of impartiality enables one to judge (...) some perspectives as better than others, the situated knowledge thesis seems to undermine this assumption by suggesting that all knowledge is partial. I argue that a contextualist theory of epistemic justification provides a solution to the bias paradox. Moreover, contextualism enables me to give empirical content to the thesis of epistemic privilege, thereby making it into a testable hypothesis. (shrink)