Non-epistemic values and scientific assessment: an adequacy-for-purpose view

Lusk, Greg; Elliott, Kevin C.

doi:10.1007/s13194-022-00458-w

Cited by 15 publications

(9 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…3 See Gilpin (2018) for a review of various ML interpretability methods. 4 See Lusk and Elliott (2022) for an adequacy-for-purpose theory of values in science. 5 In Sullivan [2022b], I call this the external problem of model opacity.…”

Section: Discussionmentioning

confidence: 99%

How Values Shape the Machine Learning Opacity Problem

Sullivan¹

2022

Scientific Understanding and Representation

View full text Add to dashboard Cite

One of the main worries with machine learning model opacity is that we cannot know enough about how the model works to fully understand the decisions they make. But how much is model opacity really a problem? This chapter argues that the problem of machine learning model opacity is entangled with non-epistemic values. The chapter considers three different stages of the machine learning modeling process that corresponds to understanding phenomena: (i) model acceptance and linking the model to the phenomenon, (ii) explanation, and (iii) attributions of understanding. At each of these stages, non-epistemic values can, in part, determine how much machine learning model opacity poses a problem.

show abstract

Section: Discussionmentioning

confidence: 99%

How Values Shape the Machine Learning Opacity Problem

Sullivan¹

2022

Scientific Understanding and Representation

View full text Add to dashboard Cite

show abstract

“…Doing away with the history and context of research objects, focusing on the challenges of sharing them rather than the challenges of interpreting them, is an attractive proposition in a world of increasingly distributed expertise, where personal links among research communities are frequently mediatedeven substitutedby communication technologies and digital platforms. And yet, the trustworthiness and reliability of those technologies and platforms ultimately depends on the collective willingness to keep scrutinizing their adequacy for purpose and the extent to which they embrace and support epistemic diversity and justice within an everchanging scientific and social landscape (Lusk and Elliott 2022). This, in turn, means having to invest some efforts into opening the black box of digital infrastructures (Bowker et al 2010, Nowotny 2021, identifying not only the choices and assumptions of relevance to one's own investigation but also the epistemic communities from which such choices and assumptions have emerged.…”

Section: Openness As Judicious Connection and The Process-oriented Vi...mentioning

confidence: 99%

Philosophy of Open Science

Leonelli

2023

View full text Add to dashboard Cite

The Open Science [OS] movement aims to foster the wide dissemination, scrutiny and re-use of research components for the good of science and society. This Element examines the role played by OS principles and practices within contemporary research and how this relates to the epistemology of science. After reviewing some of the concerns that have prompted calls for more openness, it highlights how the interpretation of openness as the sharing of resources, so often encountered in OS initiatives and policies, may have the unwanted effect of constraining epistemic diversity and worsening epistemic injustice, resulting in unreliable and unethical scientific knowledge. By contrast, this Element proposes to frame openness as the effort to establish judicious connections among systems of practice, predicated on a process-oriented view of research as a tool for effective and responsible agency. This title is also available as Open Access on Cambridge Core.

show abstract

“…One of the main reasons for adopting our normative approach is that it accords well with the insight that the characteristics of good scientific research can vary depending on the context. Because these norms rest on the aims of science, the relative weight given to different norms can vary somewhat depending on the aims associated with specific research contexts (Lusk and Elliott 2022 ). When researchers are doing applied work for regulatory purposes, for example, norms like social responsibility and engagement take on special significance.…”

Section: Toward a Solution To The New Demarcation Problemmentioning

confidence: 99%

Science, Values, and the New Demarcation Problem

Resnik

Elliott

2023

J Gen Philos Sci

Self Cite

View full text Add to dashboard Cite

In recent years, many philosophers of science have rejected the “value-free ideal” for science, arguing that non-epistemic values have a legitimate role to play in scientific inquiry. However, this philosophical position raises the question of how to distinguish between legitimate and illegitimate influences of values in science. In this paper, we argue that those seeking to address this “new” demarcation problem can benefit by drawing lessons from the “old” demarcation problem, in which philosophers tried to find a way of distinguishing between science and non-science. Many of those who worked on this problem ultimately found that efforts to provide necessary and sufficient conditions for defining science failed, and most concluded that the best solution to the problem was to characterize scientific hypotheses, theories, and research programs in terms of some common norms. We suggest that those seeking to distinguish between legitimate and illegitimate value influences on science would do well to adopt a similar approach. Rather than attempting to establish necessary and sufficient conditions for identifying appropriate value influences, it will be more fruitful to evaluate scientific activities based on their adherence to a set of epistemic and ethical norms that can be implemented in scientific practice by means of rules, conventions, policies, and procedures.

show abstract

Non-epistemic values and scientific assessment: an adequacy-for-purpose view

Cited by 15 publications

References 34 publications

How Values Shape the Machine Learning Opacity Problem

How Values Shape the Machine Learning Opacity Problem

Philosophy of Open Science

Science, Values, and the New Demarcation Problem

Contact Info

Product

Resources

About