Four Asymmetries Between Moral and Epistemic Trustworthiness‬, Susann Wagenknecht

Author Information: Susann Wagenknecht, Aarhus University,

Wagenknecht, Susann. “Four Asymmetries Between Moral and Epistemic Trustworthiness.” Social Epistemology Review and Reply Collective 3, no. 6 (2014): 82-86.

The PDF of the article gives specific page numbers. Shortlink:

Please refer to:

‪Questions of how the epistemic and the moral, typically conceived of as non-epistemic, are intertwined in the creation and corroboration of scientific knowledge have spurred long-standing debates (see, e.g., the debate on epistemic and non-epistemic values of theory appraisal in Rudner 1953, Longino 1990 and Douglas 2000). To unravel the intricacies of epistemic and moral aspects of science, it seems, is a paradigmatic riddle in the Philosophy and Social Epistemology of Science. So, when philosophers discuss the character of trust and trustworthiness as a personal attribute in scientific practice, the moral-epistemic intricacies of trust are again fascinating the philosophical mind. 

‪I have offered an account of trust in research group collaboration in a recently published paper (Wagenknecht 2014). Therein, I argue that trust in scientific practice is “incomplete”, i.e. it is not sufficient to manage the epistemic dependency that arises between collaborating scientists. For this reason, collaborating scientists resort to strategies that mitigate the limitations of the trust between them. The article then discusses a number of such strategies on the basis of an empirical case study that I have conducted.

‪In her reply to my article, Kristina Rolin (2014) discusses the ‘incompleteness’ of moral trustworthiness. Following Hardwig (1991,700), Rolin distinguishes, as I have done, a moral and an epistemic dimension of trustworthiness. The moral dimension of trustworthiness is the disposition towards honest behavior, whereas the epistemic dimension of trustworthiness is the capacity to act skilfully. Both honesty and expertise are implicated in the scientific trustworthiness of a collaborating researcher. In my article, I have focused on the epistemic dimension of trustworthiness. Now, Rolin argues that trust “[…] is incomplete because it involves a moral dimension” (2014, 75), that precisely the moral dimension of trust is “necessarily incomplete” (76) and that, at the same time, “[…] it is the moral dimension that enables scientists to cope with the incompleteness of epistemic trust” (75).

Moral-Epistemic Asymmetries

‪When elaborating upon these claims, Rolin finds that there is an “asymmetry” between the epistemic and the moral dimension of trustworthiness. This is a highly intriguing thought, which I would like to pursue. What I thus offer in this reply to Rolin’s reply is to distinguish the different ways in which the epistemic and moral dimension of trustworthiness can be seen as ‘asymmetric’ as implied in Rolin’s argument, and to comment on those from my own perspective. While I agree with some of the asymmetries found by Rolin, I do not entirely agree with others. In the end, this will lead me to re-state my position that trust, neither in its epistemic nor in its moral dimension, can sufficiently address epistemic dependence in collaborative scientific practice. There is, as I have argues, no simple trust-automatism in scientific practice. Collaborating scientists cannot use epistemic trust, not even moral trust, as a ready “patch” on the lack of directly controlled, first-hand evidence when epistemic dependence arises in testimonial exchanges.

In the following, I distinguish four different ways in which the epistemic and the moral dimension of trustworthiness can be seen as asymmetric. I provide what I believe is a charitable reconstruction of Rolin’s asymmetry argument, and I will make clear where I agree with her line of thought and where I am reluctant to do so.

Asymmetry 1: Moral trustworthiness is the foundation of epistemic interaction

‪Moral trustworthiness is absolutely crucial in collaborative scientific practice where the collaboration between different individuals involves forms of epistemic dependence. Dishonesty, in turn, destroys everything. It undermines the very foundation of epistemic interaction. The testimony of a dishonest, or occasionally dishonest, speaker is epistemically useless. The foundational role of the moral dimension trustworthiness becomes even more palpable when compared to epistemic dimension of trustworthiness. In collaborative scientific practice, a scientist’s lack of expertise may lead to erroneous results, but errors are common in science and can be analyzed. Dishonest behavior, however, occludes the analysis of error. For this reason, dishonest behavior is considered scientific misconduct and is — rightly, I believe — punished more severely than mere incompetence, negligence or ignorance (Rolin 2014, 76; see also Andersen 2014).

Asymmetry 2: The default assumption of moral trustworthiness

‪Rolin holds that “[…] [w]hile scientists may expect to have some empirical warrant to support their trust in the epistemic character of their collaborator, the moral character of the collaborator is to a large extent taken for granted” (Rolin 2014, 76). Put differently, honesty typically is and should be assumed by default, i.e., moral trustworthiness should be granted generously unless “… there is evidence of intentional attempts to distort [the] research process” (76). The acknowledgement of expertise, in contrast, is at least partially expected to be underwritten by first-hand or testimonially-conveyed forms of experiential evidence.

‪According to Rolin, moral trustworthiness is assumed by default because a satisfactory evidential basis for the assumption of honest character is not to be had. Here, however, a subtle approach is needed and it is important to point out that an initial default assumption of moral trustworthiness will in the course of continued interaction be qualified and fine-tuned to gathered experience. When collaborating, scientists observe one another. These observations help to specify expectations of trustworthiness in so far as they help to tell when vigilant attention and additional measures of indirect and direct control are worthwhile to consider (Sperber et al. 2010). This is all the more important in collaborations that cannot be easily terminated even though the moral trustworthiness of a collaborator is sincerely damaged (think of, e.g., student-teacher relationships or intricate set-ups with multiple mutually-dependent collaborators).

Asymmetry 3: The evidence of moral character is unavoidably incomplete

‪Rolin argues that the evidence of moral character is unavoidably incomplete, i.e., it is permanently under-determined by evidence because moral character is a disposition to behave in a certain way across a range of social situations at present and in the future. Therefore, Rolin proceeds, “[…] trust in the moral character of other scientists is at least partly based on a principle of charity” (Rolin 2014, 76). Now, what about the epistemic dimension of scientific trustworthiness, i.e., what about the possession of scientific expertise? Rolin suggests that evidence for moral character is necessarily incomplete, while evidence for expertise is not.

‪This is, in fact, a very useful distinction that helps to refine the position that I developed in my article. For my purposes, I suggest to reconstruct this facet of the asymmetry between the moral and epistemic dimension of trustworthiness as follows: Epistemic trustworthiness is rooted in expertise, which is had, not had or partially had. It is a resource and the degree to which this resource is present in a collaborator to be trusted can, albeit with considerable effort, be reliably enough determined. Put differently, whether or not expertise enables to deal with future situations is usually taken to be a questioned whose answer can be sufficiently determined. To follow this line of thought is equivalent to saying that the, as I have called it in my article, inherent incompleteness of trust as future-oriented is essentially rooted in the moral dimension of trustworthiness. I am inclined to adopt this line of thought because it is very difficult to show that expertise is as ephemeral, as pervasive and as hard to predict as moral character (although one may try to do so for the case of embodied, tacit and/or highly context-sensitive expertise).

Asymmetry 4: To assume moral trustworthiness is a moral imperative

‪Furthermore, Rolin argues that it is right to assume moral trustworthiness by default because the decision to do so is a moral decision, not an empirical judgment: “The moral reason is the belief that it is morally wrong to doubt another group member’s honesty when one does not have a reason to do so” (Rolin 2014, 76, my emphasis). Remember that it would not be possible to make moral trustworthiness the matter of a fully satisfactory empirical judgment because any assumption of moral character is under-determined by empirical evidence.

‪Still, I am not entirely convinced of the claim that the ascription of moral trustworthiness is solely a moral decision. Rather, I believe this decision to be a judgment that may change character over time and become increasingly consolidated by experience. To my mind, Rolin herself indicates this possibility when she defines what it means to assume moral trustworthiness by default: It means to assume moral trustworthiness until further notice, i.e. “[…] unless one has reason to doubt it” (ibid.). Hence, the decision not to ascribe moral trustworthiness is not a moral, but an empirical judgment — and in the case of a default assumption as described here, considerations whether not to ascribe moral trustworthiness are an inherent element in the continuous decision to ascribe this attribute.

What Moral Trustworthiness Cannot Do

‪It is clear that both the epistemic and moral dimension of trustworthiness deserve our attention. Both matter crucially in collaborative scientific practice, and Rolin has rightly pointed out that there are important asymmetries between those two dimensions. Yet the fact that moral trustworthiness in contrast to epistemic trustworthiness is granted by default, and is granted also for moral reasons, should not lead us to believe that moral trustworthiness would be less problematic, easier to handle or is a simple given of collaborative scientific practice.

‪Rolin indicates that “[…] it is the moral dimension that enables scientists to cope with the incompleteness of epistemic trust” (75). In fact, what moral trustworthiness can do is to trigger collaborative activities. Moral trustworthiness foregrounds epistemic trustworthiness. Without moral trustworthiness there is no epistemic trust to begin with. Also, and this seems to be what Rolin has in mind, the moral dimension of trust might sometimes be able to splint brittle epistemic trustworthiness. It may be the case that strong moral trustworthiness can occasionally compensate for disappointed expectations of epistemic trustworthiness. An unintended error, if discovered and not covered up, can be ridden out. Eventually, however, honesty cannot compensate a lack in expertise.

‪Moral trustworthiness is not a handy patch for doubts concerning the epistemic trustworthiness of a collaborator. The reality of collaborative scientific practice is piecemeal, fragile, fragmented and uncertain. What helps to handle the epistemic risks of collaboration, i.e. the risk of disappointed trust expectations, is the deliberate will to take this risk and to resort to a number of measures that can mitigate it: The reference to institutional trust, i.e. trust in community-borne gate-keeping functions, can partly compensate for a lack of familiarity with potential collaborators. In addition, epistemic vigilance as expressed in probing dialoguing practices helps to get to know collaborators, assess their skills and determine in which areas of expertise collaborators are epistemically trustworthy and in which they are not. And most importantly, perhaps, scientists fine-tune expectations towards one another in first tentative, then subsequently more consolidated iterative cycles of ongoing collaboration. In analogy to Chang’s “epistemic iteration” in experimental science (Chang, 2004), moments of socio-epistemic ‘iteration’ could be the key to understanding and managing the uncertainties of collaborative scientific practice.


‪Andersen, Hanne. “Epidemic Dependence in Contemporary Science: Practices and Malpractices.” In Science After the Practice Turn in the Philosophy, History and Social Studies of Science, edited by Léna Soler, Sjoert Zward, Michael Lynch, Vincent Israel-Jost. 161-173. New York: Routledge, 2014

Chang, Hasok. Inventing Temperature: Measurement and Scientific Progress. Oxford Studies in Philosophy of Science. Oxford, New York: Oxford University Press, 2014.

Douglas, Heather. “Inductive Risk and Values in Science.” Philosophy of Science 67, no. 4 (2000): 559-579.

‪Hardwig, John. “The Role of Trust in Knowledge.” Journal of Philosophy 88, no. 12 (1991): 693-708.

‪Longino, Helen E. Science as Social Knowledge: Values and Objectivity in Scientific Inquiry. Princeton University Press, 1990.

Rolin, Kristina. “‘Facing the Incompleteness of Epistemic Trust’ — A Critical Reply.” Social Epistemology Review and Reply Collective 3, no. 5 (2014): 74-78.

‪Rudner, Richard. “The Scientist Qua Scientist Makes Value Judgments.” Philosophy of Science (1953): 1-6.

‪Sperber, Dan, Fabrice Clément, Christophe Heintz, Olivier Mascaro, Hugo Mercier, Gloria Origgi, and Deirdre Wilson. “Epistemic Vigilance.” Mind & Language 25, no. 4 (2010): 359-393.

‪Wagenknecht, Susann. “Facing the Incompleteness of Epistemic Trust: Managing Dependence in Scientific Practice.” Social Epistemology (12 February 2014): doi: 10.1080/02691728.2013.794872

Categories: Critical Replies

Tags: , , , , ,

Leave a Reply