Epistemic dependence
Also called: epistemic reliance, dependence on experts, intellectual dependence, epistemic deference
The condition of holding a belief on the strength of someone else’s knowledge or evidence, which one does not have and often could not obtain oneself.
You believe that the earth is about four and a half billion years old. You have not dated any rocks, and you could not assess the methods if they were shown to you. You hold the belief because geologists hold it. John Hardwig’s 1985 paper gave this ordinary situation its name and drew an uncomfortable conclusion from it.
Hardwig’s argument
The traditional ideal is intellectual self-reliance: believe only what you have good reasons for. Hardwig observed that for most of what an educated person believes, the reasons are possessed by other people. A layperson has good reason to think the expert has good reasons, and that is all. He argued that it is nonetheless rational, often more rational than thinking for oneself, to defer. The alternative is to believe far less, or to hold worse-supported beliefs out of pride.
The point applies inside science as much as outside it. A modern experiment may have dozens or hundreds of authors, none of whom could vouch for every part. Hardwig concluded that knowledge in such cases rests on trust between people, and in a second paper that trust is in that sense prior to evidence.
What it changes
If this is right, the unit that knows is sometimes a community and not an individual, and the character of informants, their honesty and competence, becomes part of epistemology. The view sits within the wider study of testimony and runs against any theory on which a knower must hold the justification herself.
Dependence on machines
Hardwig’s dependence is on people who know. Dependence on an AI system has the same outward shape, since one takes an answer one cannot check, but I argue that the structure underneath differs. When I rely on a geologist, her knowledge stands behind her word. When I rely on a language model, there is, on current evidence, no knower in that position, and whatever warrant I have must come from my own grounds for thinking the system reliable on questions of this kind (instrument or informant?). The dependence is real in both cases. What one depends on is not the same.
Related terms
Questions that turn on it
Referred to by
Further reading
- Hardwig, J. (1985). Epistemic dependence. The Journal of Philosophy 82(7): 335–349.doi.org/10.2307/2026523
- Hardwig, J. (1991). The role of trust in knowledge. The Journal of Philosophy 88(12): 693–708.doi.org/10.2307/2027007