Many AI systems that make important decisions are black boxes: how they function is opaque even to their developers. This is due to their high complexity and to the fact that they are trained rather ...than programmed. Efforts to alleviate the opacity of black box systems are typically discussed in terms of transparency, interpretability, and explainability. However, there is little agreement about what these key concepts mean, which makes it difficult to adjudicate the success or promise of opacity alleviation methods. I argue for a unified account of these key concepts that treats the concept of understanding as fundamental. This allows resources from the philosophy of science and the epistemology of understanding to help guide opacity alleviation efforts. A first significant benefit of this understanding account is that it defuses one of the primary, in-principle objections to post hoc explainable AI (XAI) methods. This “rationalization objection” argues that XAI methods provide mere rationalizations rather than genuine explanations. This is because XAI methods involve using a separate “explanation” system to approximate the original black box system. These explanation systems function in a completely different way than the original system, yet XAI methods make inferences about the original system based on the behavior of the explanation system. I argue that, if we conceive of XAI methods as idealized scientific models, this rationalization worry is dissolved. Idealized scientific models misrepresent their target phenomena, yet are capable of providing significant and genuine understanding of their targets.
Particularists maintain that conspiracy theories are to be assessed individually, while generalists hold that conspiracy theories may be assessed as a class. This paper seeks to clarify the nature ...and importance of the debate between particularism and generalism, while offering an argument for a version of generalism. I begin by considering three approaches to the definition of conspiracy theory, and offer reason to prefer an approach that defines conspiracy theories in opposition to the claims of epistemic authorities. I argue that particularists rely on an untenably broad definition of conspiracy theory. Then, I argue that particularism and its counterpart are best understood as constellations of theses, rather than a pair of incompatible theses. While some particularist theses are highly plausible, I argue that one important particularist thesis is false. The argument for this conclusion draws on the history of false conspiracy theories. I then defend this conclusion against a pair of potential objections.
Given that conspiracies exist, can we be justified in dismissing conspiracy theories without concerning ourselves with specific details? I answer this question by focusing on
contrarian
conspiracy
...theories
, theories about conspiracies that conflict with testimony from reliable sources of information. For example, theories that say the CIA masterminded the assassination of John F. Kennedy, 9/11 was an inside job, or the Freemasons are secretly running the world are contrarian conspiracy theories. When someone argues for a contrarian conspiracy theory, their options are to appeal to eyewitness testimony, incriminating documents, photographs, audio recordings, video recordings, allegedly scientific evidence, anomalies, or considerations of
cui bono
(who benefits). Most laypeople don’t know how to adequately respond to arguments for contrarian conspiracy theories, but this doesn’t mean that it’s dogmatic to reject them. I argue that many laypeople are defeasibly justified in rejecting a variety of contrarian conspiracy theories based on testimony and this justification is rarely defeated by arguments for contrarian conspiracy theories even if we don’t know how to respond to these arguments.
Complex environmental problems require well-researched policies that integrate knowledge from both the natural and social sciences. Epistemic differences can impede interdisciplinary collaboration, ...as shown by debates between conservation biologists and anthropologists who are working to preserve biological diversity and support economic development in central Africa. Disciplinary differences with regard to 1) facts, 2) rigor, 3) causal explanation, and 4) research goals reinforce each other, such that early decisions about how to define concepts or which methods to adopt may tilt research design and data interpretation toward one discipline's epistemological framework. If one of the contributing fields imposes a solution to an epistemic problem, this sets the stage for what I call disciplinary capture. Avoiding disciplinary capture requires clear communication between collaborators, but beyond this it also requires that collaborators craft research questions and innovate research designs which are different from the inherited epistemological frameworks of contributing disciplines.
•Interdisciplinary collaboration can fail due to incompatible philosophical beliefs.•Incompatible philosophies are more resistant to change than communication problems.•Facts, evidentiary standards, causal inferences, and research goals interrelate.•Disciplinary capture involves a systematic default to one philosophical framework.•Avoiding disciplinary capture requires explicit attention to philosophical decisions.
Smart Environments Ryan, Shane; Palermos, S. Orestis; Farina, Mirko
Social epistemology,
07/2024, Volume:
38, Issue:
4
Journal Article
Peer reviewed
This paper proposes epistemic environmentalism as a novel framework for accounting for the contribution of the environment - broadly construed - to epistemic standings and which can be used to ...improve or protect epistemic environments. The contribution of the environment to epistemic standings is explained through recent developments in epistemology and cognitive science, including embodied cognition, embedded cognition, extended cognition and distributed cognition. The paper examines how these developments support epistemic environmentalism, as well as contributes theoretical resources to make epistemic assessments of dynamic environments. The epistemic environmentalist procedure from the assessment of an individual environment to changes made to that environment based on promoting the attainment of epistemic goods is also discussed.
In this article we argue that philosophy can facilitate improvement in cross-disciplinary science. In particular, we discuss in detail the Toolbox Project, an effort in applied epistemology that ...deploys philosophical analysis for the purpose of enhancing collaborative, cross-disciplinary scientific research through improvements in cross-disciplinary communication. We begin by sketching the scientific context within which the Toolbox Project operates, a context that features a growing interest in and commitment to cross-disciplinary research (CDR). We then develop an argument for the leading idea behind this effort, namely, that philosophical dialogue can improve cross-disciplinary science by effecting epistemic changes that lead to better group communication. On the heels of this argument, we describe our approach and its output; in particular, we emphasize the Toolbox instrument that generates philosophical dialogue and the Toolbox workshop in which that dialogue takes place. Together, these constitute a philosophical intervention into the life of CDR teams. We conclude by considering the philosophical implications of this intervention.
This paper introduces and discusses the core themes explored in the special issue on the social dimension of the ethics of knowledge at the intersection between virtue and vice epistemology.
Is it generally rational to defer to official stories? On the affirmative view exemplified by Neil Levy, grounds for scepticism cannot outweigh the epistemic authority of the experts presumed to ...generate them. Yet sociological studies of how expertise is mediated into official communications reveal the epistemic potential of citizens' collaboratives. These may include, or advocate hearing, dissident experts. Such groups' epistemic position is arguably analogous to that of the 'other institutions of civil society' that Levy sees as underwriting the authority of official stories. An added advantage is that they are able to adhere to deliberative methods of inquiry. By contrast, an official story presupposes that deliberations have concluded, and what remains is to promote the agreed story. This is achieved by strategic communication, which can involve extensive coordination by organisations committed primarily to persuasion rather than truth-seeking. Sometimes it includes blocking dissidents' deliberative challenges. A case study highlights how even an organisation of indisputable epistemic authority can be politically influenced to block the testimony of its own scientific experts, while uncritical support for its official story is maintained by other institutions of civil society. Hence, an appropriate attitude towards official stories is critical receptiveness rather than automatic deference.
This paper uses metaphor theory and analysis to explore competing and sometimes contested claims about the nature and utility of social learning. Seven metaphor clusters — performance metaphors, ...action metaphors, communication metaphors, governance mechanism metaphors, social learning as a paradigm, social learning as a form of cognition and social learning as a wheelbarrow full of frogs, were identified from the sustainability and natural resource management (NRM) literature. Rather than seeking to define social learning rigidly, and thus limiting its potential utility to open up spaces for innovation in NR governance, social learning can be positioned in future discourse so that it holds a cluster of revealing and concealing features. This position shifts responsibility for clarity and rigour away from the concept, useful because of its fluidity, to the user of the concept who must then articulate the way(s) in which they choose to use it. This shift is consistent with reflexive, systems practice for systemic and adaptive governance and invites practitioner responsibility rather than conceptual reification.
► Uptake of social learning by policy makers has not paralleled research interest. ► Social learning meanings differ depending on user and their epistemological commitments. ► Metaphor diversity provides an opportunity rather than a constraint. ► Our analysis moves responsibility from the concept to the user of a concept. ► Certain critiques of social learning dissolve through metaphor analysis.