When we learn from others, do we learn from their beliefs, or from their words?
Jennifer Lackey insists that we learn from words rather than beliefs, as the title of her book indicates:
Her "creationist teacher" thought experiment is used in making the case. She asks us to imagine an elementary school science teacher who is convinced that human evolution did not happen, but who is required to nevertheless teach her students the evolutionary account of human origins. Out of fear of losing her job, she does so. This would seem to be a paradigm case of an untrustworthy speaker: she presents as true things she believes to be false. Assuming that the evolutionary account is true after all, Lackey asks: do the students, believing their teacher, come to know how humans originated? Your intuitive response is supposed to be that yes, they do. If that is so, then, she argues, the students learned from what she said, not what she believed.
(Lackey calls her own view in the epistemology of testimony the "statement view of testimony": it is reliable statements, rather than trustworthy speakers, that make testimonial knowledge available to hearers. This is a "process reliabilist" epistemology of testimony. But this is not what I am interested in in this post--so on to other matters.)
I think that the conclusion that the students have learned from words instead of beliefs is not as well supported by this thought experiment, since it fails to rule out a plausible alternative explanation of the students' knowledge. Granting all the particulars in the case, we can note that the students would be less inclined to believe adults giving similar utterances in different contexts. Consider, for example, the same students gathered in the library for story time. The librarian loves to regale the students with fantastic tales that he presents with vigor and conviction, and he never gives any indication than that he fully believes that what is is telling the students is the truth. Nevertheless, the students do not believe that things he tells them are true--they do not believe, for example, that Arthur really did pull a magical sword from a stone, or that Lucy really did step through a wardrobe into the magical land of Narnia.
What explains the difference is the background knowledge that the students have regarding the context of utterance, and even of the particular roles of the speakers within the context. They do not expect the librarian telling them stories to be telling them things he actually believes, but they do expect this of the teacher.
But then aren't we back to the problem of how students could acquire knowledge from simply trusting a speaker who is untrustworthy--even if that speaker happens to say something true? I think that this is resolvable without appealing to an epistemology of testimony that makes speaker-beliefs dispensable. Here's how:
First, we distinguish two different ways a teacher could be, or fail to be, trustworthy: a teacher can fail to be trustworthy through incompetence or through insincerity (or both). It certainly seems like the speaker in Lackey's case is insincere: ex hypothesi, she is telling the students things that she doesn't believe. But she may still be competent, and in the case, the competence may be all that is required for knowledge transmission via belief communication.
Given her social role as a teacher, the things that she tells the students in the context of a science classroom are, and the students understand them to be, communications regarding not (only) her personal beliefs, but the beliefs of the community--that is, the propositions that are held to be true by the relevant part of the community. The children, however tacitly, assume not just that the teacher knows these things, but that the teacher is a teacher, and that teachers communicate what we believe about the subject matter at hand. The teacher competently fulfills her role as teacher when she teaches the children the things that are believed widely in their community--even when she herself does not believe those same things.
So the teacher doesn't "just happen" to tell the students something true in Lackey's thought experiment. She reliability communicates, in her role as teacher, what the community believes is true (or what the relevant experts within the community believe to be true). The librarian is neither communicating what she believes to be true, nor what the community believes to be true, and the children understand this (however tacitly).
But then aren't we appealing to the reliability of the teacher's statement to explain how the students can acquire knowledge, thus affirming Lackey's view after all, and undercutting the view that we learn from others' beliefs?
No, because it is the students' attunement to beliefs that explains why they come to believe as they do--whether to the beliefs of the individual speaking, or to the communal beliefs expressed by the one speaking. This is not to deny the importance of reliability in the transmission of knowledge, but it is to deny that when we learn via testimony, we are not learning from the beliefs of others.
I realize that this explanation needs more work, and that there is much more that could be said. For example, I believe that there is good empirical evidence that children are in fact naturally inclined to adapt their beliefs to what their community believes, rather than to what is reliably communicated. But hey, this is just a blog post.