13.07.2026
The central goal of science is to gain knowledge. Scientists generally assume to be capable of achieving this goal within their respective disciplines. In philosophy, knowledge is often defined as a justified true belief. Based on this definition, scientific findings must not only be true but also justified in order to be considered knowledge.
To say that knowledge must be justified leaves open the question of what counts as good justification. In science, however, trust is not normally considered good justification. Instead, the idea of epistemic autonomy is held up as an ideal: those who think for themselves and critically question things are acting in an exemplary manner. Those who, on the other hand, blindly trust others are considered irrational. At the same time, however, modern science cannot function without teamwork, which seems to require trust among scientists.
The philosopher John Hardwig argues in his Paper The role of trust in science that trust in others is a necessary condition for knowledge in modern science.
About the Author: Ana Maria Abril won a silver medal in the 2023 Philosophy Olympiad finals. She is now studying philosophy and mathematics in Bern and volunteers for the Philosophy Olympiad.
Hardwig begins his argument with an analysis of the structure of modern science. This structure is fundamentally collegial. Scientists work based on the results of previous scientists. In addition, they work in large groups, some of which are internationally organized. One reason for this collaboration is that collecting and analyzing data, or conducting certain experiments, is sometimes so time-consuming or complex that it cannot be carried out by a single person.
According to Hardwig, this gives rise to an epistemological problem, that is, a philosophical problem concerning knowledge and its conditions. If scientific results are to be considered knowledge, then they must be justified. In large research groups, however, no single person possesses the complete justification for the collective results. Many scientific results therefore do not appear to be (fully) justified and can, strictly not be regarded as knowledge. Nevertheless, we wish to refer to the results of this research as knowledge.
Harwig sees the importance of testimony in this lack of justification. In epistemology, testimony generally refers to the transmission of beliefs from one person to another. Hardwig’s idea is that justification can be shared through testimony. Researchers in a research group can share justification, and thus knowledge, with one another through testimony. This gives every participating researcher, even if indirectly, access to the justification for the results produced by their joint research. Thus, only when researchers accept the testimony of other researchers can they call the results of their joint research “knowledge.”
According to Hardwig, this testimony is closely linked to trust: Person B’s testimony that p can only serve as justification for Person A to believe p if Person A trusts B. Testimony thus presupposes trust if it is to be justificatory.
An expert is giving a lecture. According to Hardwig, we will only accept his testimony as a justification for believing something if we trust him. (Photo: Philosophy Olympiad)
Hardwig further analyzes the concept of trust and provides an account of when a person is considered trustworthy (honesty, competence, conscientiousness, …). It is important to note that Hardwig’s analysis, even though he does not use the term himself, refers to trust in people (personal trust).
Annual theme “Look up!”: Who do we look up to? What do we see, when we look up — maybe using a telescope at the new Astronomy Olympiad? Just as a ship can use the stars to navigate, role models can help us find our path. In 2026 we meet inspiring people and those inspired by them. We present talent promotion practices worth looking up to. And we gaze at the sky. Subscribe to the newsletter or follow us on Instagram so you don’t miss anything!
On this basis, Hardwig concludes that in modern science, personal trust is a necessary condition for knowledge. His argument can be summarized as follows:
Premise 1: In modern science, testimony is a necessary condition for scientific knowledge.
Premise 2: Testimony presupposes trust.
Conclusion: In modern science, trust is a necessary condition for scientific knowledge
Hardwig’s analysis of the role of trust in science is limited exclusively to trust in people. One possible criticism of this is that it is insufficient to explain how scientific knowledge is possible in modern science. As Hardwig himself notes, scientists also work in international research groups. In such settings, scientists do not necessarily know one another personally. Thus, the kind of trusting relationship necessary to accept the testimony of other scientists cannot exist between them.
A philosophical text always invites its readers to reflect on their own views about its content and consider possible criticisms. (Photo: Philosophy Olympiad)
The following example illustrates that in situations where there is no personal relationship between the researchers involved, another form of trust, institutional trust, can take on the role of personal trust. This example is intended to make the thesis intuitively understandable. Institutional trust refers to trust in structures such as universities, private research institutions, peer reviews, and meta-analyses.
Leonie, a conscientious scientist, is reading a newly published scientific study that she could use as the basis for her own research. She has no personal relationship with the study’s authors. Before using the study’s results in her own research, she, conscientious as she is, first examines the study’s content and checks whether it is internally consistent. She also verifies how it aligns with the knowledge she already possesses. Leonie also takes into account who conducted the study and at which institution it was published. If she concludes that she can use the study’s results in her own research, this decision is not based on personal trust. She does not primarily trust the study’s authors, but rather the institutions and research programs within which the study was conducted.
This example does not yet provide a truly convincing argument for institutional trust. At this point, you are invited to think further. Consider these questions for example: What exactly does trust in institutions mean (as opposed to trust in persons)? When is an institution trustworthy? And how exactly does institutional trust justify the sharing of justification?
Does this make you think? If you're interested in questions like these, you can participate in the preliminary round of the Philosophy Olympiad from September 1 through October 4.
