Reductionism and anti-reductionism in the epistemology of testimony

Earlier, I suggested that reductionism in the epistemology of testimony is a form of epistemic individualism. Reductionists allow that the individual can get testimonial knowledge from oth­ers, but the kind of dependence it allows here is merely a pseudo-dependence: Testimonial knowledge requires, on this view, that the speaker be adequately vetted.That is, using the non­testimonial resources that the hearer has available to her, she may extend her knowledge by means of validated testimony.

One way to understand anti-reductionism in the epistemology of testimony is that it insists on real epistemic dependence on other persons, as opposed to mere pseudo-dependence. Put differently, anti-reductionist views required a dependence on the speaker that is not exhausted by what can be evaluated and validated by the hearer.That is why anti-reductionist views often talk in terms of trusting the speaker, where this is explicitly opposed to believing the speaker on good evidence.¹⁸

I end by offering an example of an anti-reductionist view along such lines. The anti-reduc­tionist position I put forward also shows how epistemic dependence on others can add epistemic value, and how individual intellectual autonomy can be advanced by means of cooperation with others.¹9

In order to present the view that I have in mind, it will be helpful to introduce the notion of an “epistemic community,” understood as a group of persons who share some set of informa­tion-dependent tasks, and who share a set of norms for acquiring and distributing task-relevant information. That is, an epistemic community will have a need for quality information, and therefore will have norms and standards in place that are relevant for achieving that quality.

The next point to note is that the norms or standards that are appropriate for the acquisition of information will be somewhat different from the norms or standards that are appropriate for the distribution of information.

On the view of testimonial knowledge that I want to propose, the characteristic function of testimony is to distribute quality information within an epistemic community.That is, the primary function is to take information that has already passed the quality control test at the gatekeeping stage, and to distribute that information to those who need it. Put differently, the characteristic function of testimony is to move knowledge around within an epistemic community.

That central idea has to be qualified, however, because this characteristic function of testi­mony is not its sole function. In particular, testimony can be used to acquire information for an epistemic community as well as to distribute it. This happens when the speaker is not herself a member of the hearer's epistemic community, and therefore is not in the same relationship of cooperation with the hearer as are members of the same community. For example, suppose that a personnel director is interviewing a job candidate for her company. The personnel director and the job candidate are in a different relationship, governed by different norms for giving and receiving testimony, than the personnel director is in with her co-workers or with her boss.

Finally, let's make a distinction between the generation or production of knowledge for an epistemic community, and the transmission of knowledge within that community. The present suggestion is that the transmission of knowledge is governed by different norms and standards from the generation of knowledge. Insofar as the characteristic purpose of testimony is to trans­mit knowledge within an epistemic community, testimonial exchanges are (characteristically) governed by relevant norms and standards.

The current view is anti-reductionist in the sense explained above. That is, at least some testimonial knowledge is subject to its own standards and norms, and therefore not reducible to knowledge produced by inductive reasoning, or knowledge of some other kind.The view is also anti-individualist, in that the transmission of knowledge depends on both speaker and hearer playing their appropriate role in a cooperative testimonial exchange. Accordingly, the epistemic standing of the hearer—that she knows the thing she has been told—at least sometimes depends on the epistemic standing of the speaker—that she had knowledge to transmit, and did so suc­cessfully It follows from all this that the view is also externalist in the sense explained above.That

is, whether a speaker knows, and whether the speaker and hearer play their appropriate roles in a cooperative testimonial exchange, in not something that the hearer can know “by reflection alone,” and is certainly not simply a matter of the hearer's own mental states.

In fact, the present approach is able to accommodate more diverse social factors and the broader social environment in its account of the transmission of knowledge. For the norms that govern relevant testimonial exchanges are often tied up with interpersonal relations, social norms and roles, institutional rules, and even positive law. In particular, all of these may con­tribute to successful knowledge transmission in healthy epistemic communities, and undermine transmission in unhealthy ones. And so, again, we have various ways in which the epistemic standing of individuals very much depends on features of the broader social environment.

The phenomenon I have in mind here involves the notion of what I will call “transmission channels.” Much as the contours of a physical environment can promote or inhibit the move­ment of physical objects within it, the contours of an epistemic community can promote or inhibit the movement of knowledge. For example, imagine a social environment where there is a deficit of trust and good will, and various disincentives for cooperation. To that extent, conditions for epistemic cooperation will be limited as well, including opportunities for reli­ably sharing information and transmitting knowledge. Likewise, social environments that enable cooperation in general will also enable epistemic cooperation. In various cases, norms that are primarily practical or moral underwrite good epistemic practices in this regard. I will end with a few examples.

First, consider a normal, healthy relationship between parent and child. Care and love in the relationship will guide the behaviors of both parent and child in epistemically beneficial ways. For example, parents are motivated to provide their children with relevant information and to make sure that they are in a position to receive it effectively. Likewise, young children are moti­vated to trust their caregivers for needed information. Similar things can be said about relations among friends, which are governed by norms of care, loyalty, and mutual benefit.Various profes­sional-client relationships also give rise to epistemically effective transmission channels, which are created and maintained not only by norms governing the professional relationship, but also by broader social factors such as professional ethics, licensing requirements, economic incentives, and even positive law. In all of these ways, the transfer of knowledge is enabled by social factors that outrun the intellectual abilities and control of the hearer.

When we consider how ubiquitous these various relationships are, the extent of real epis­temic dependence on others becomes clear.What also becomes clear is that both practical and intellectual autonomy is enabled and enhanced by such dependence, rather than undermined by

it. Which is to say that human agency, both practical and intellectual, is largely a social agency.