The Exchange of Words

Much of our knowledge of the world comes through testimony. When philosophers have attempted to explain this fact, it has been natural to apply existing epistemologies. Thus, testimony can be thought of as just another sign in the world, indicating its truth in the same way that smoke indicates fire. Understood in this way, testimonial knowledge might be explained in reliabilist terms, since both the receipt of a speaker's words and the observation of smoke can be regarded as reliable indicators. Alternatively, testimony might be thought of as secondhand perception. On this conception, confrontation with testimony functions as an epistemic substitute for confrontation with the perceptual state of affairs that the testimony represents.On these approaches to testimony, philosophers take the perspective of the recipient of testimony. The speaker figures merely as the source of testimony, which might be reliable or not. This philosophical landscape was revolutionized by Moran's paper “Getting Told and Being Believed” (Moran 2005). What is missing from above accounts, Moran argued, is the testimonial relationship. A speaker is not merely a source of testimony but someone who stands in a relationship of address to an audience. The epistemology of testimony should not be thought of in terms of consumption but in the bilateral terms established by this relation of address—that is, in second-personal terms.The Exchange of Words takes these central insights and builds on them. (“Getting Told and Being Believed” is reprinted, slightly improved, as chapter 2.) The result is a more developed and philosophically satisfying account of what was already an engaging and, in my opinion, largely persuasive view of testimony. Additionally, Moran outlines a more general account of what he calls, following Thomas Reid, ‘social acts of the mind’. This is philosophically rich and complex stuff and I cannot do justice to either its depth or breadth in this brief review. Instead, I will focus on Moran's account of tellings and on how The Exchange of Words fleshes out that account. This is a central theme of the book.If we leave the testimonial relationship aside and focus on the audience as a ‘consumer’ of testimony, the question is whether any given bit of testimony is ‘good’ or ‘bad’; whether it is true, well supported, an expression of knowledge, and so on, or not. In most cases, the key determiners will be the speaker's sincerity and competence. The problem is that from this theoretical perspective sincerity is valuable only insofar as it marks the absence of insincerity—insofar as it lets the testimony pass as a genuine indication of belief (thereby leading to the question of competence). It would be better if the audience did not have to rely on testimony but had some means of accessing the speaker's beliefs directly, by employing a brain scan or some such technology. This indicates that something has gone wrong with epistemic theory, because the act of being told something can be an epistemic positive. Compare apology. By using a brain scan an audience might know a speaker is sorry. But being sorry might not be enough; the speaker might need to say sorry because doing so is an act of taking responsibility. Similarly with telling. The act of giving testimony—of telling someone something—is one of taking responsibility, of standing behind the truth uttered, or at least presenting yourself as doing so. Here Moran borrows from Paul Grice's distinction between acts of telling and letting know. Both involve the intention to induce belief; what then distinguishes acts of telling is that speakers intend to accomplish this by way of their audiences’ recognizing that they have this intention. To this Moran adds: for the audience to see this intention to induce belief as a reason for belief requires them to see this intention as an assumption of responsibility. To take the speaker's telling as such would be then to believe the speaker. Because acts of telling have this aim of believing the speaker, they should be regarded as offering speakers’ assurance that what is said is true.So far this does not go beyond “Getting Told and Being Believed.” A common response to this position is to say that assurance offers a practical, not epistemic, reason for belief. The Exchange of Words offers two rejoinders.First, it is important to distinguish between the nature of epistemic reasons themselves and how these reasons are given in communication. Epistemic reasons are not second-personal but ‘monadic’, existing independently of the testimonial exchange. What the testimonial exchange does is simply make the speaker's epistemic reasons available to an audience, through presenting the speaker's knowledge or justified belief. What is second-personal is the act by means of which these epistemic reasons are made available. The speaker has the authority to determine that her utterance makes an epistemic reason available because, and insofar as, she has illocutionary authority over what she says. This authority is second-personal since it sees the speaker making herself accountable, or taking responsibility. The second rejoinder found in The Exchange of Words is then the claim that giving an account of this illocutionary act—by means of which we give each other epistemic reasons for belief—is actually a precondition of asking the kind of epistemological questions that characterize theorizing on testimony, questions about sincerity, competence, reliability, and so on. This is because the starting point of the epistemology of testimony is understanding, not merely as a semantic matter but as understanding what a speaker is doing in uttering. Suppose the hissing and crackling of a fire makes sounds that we hear as ‘the rain has stopped’. We don't take this to be testimony to that effect. Nor would we, if someone made these sounds when asleep, or uttered these words as a pronunciation exercise, or if we understood the utterance to be a joke. Thus, the epistemological starting point is that we understand an utterance to be a case of a speaker being intentionally informative, telling us something. (Not that we can't learn something from a snoring speaker or a joke, but we don't take this learning to be testimonial.) Only when this understanding is in place can questions of believing or not arise. But this means that epistemology must first turn to the prior question of the nature of the illocutionary act.Let's put lies aside for a moment. As speakers, what we think we are doing when we tell someone something is sharing knowledge, or what we justifiably believe to be so. Moreover, we think that we are sharing knowledge not by laying down clues that enable our audience to work out the truth of what we know, but by the very act of telling our audience what it is that we know or justifiably believe. Thus, tellings have the mark of the illocutionary. For instance, in normal circumstances, it is enough for a speaker to promise something for her to have done that very thing. This is marked by the legitimacy of including ‘hereby’ before the performative verb, as in ‘I hereby promise to...’. Tellings similarly allow such announcements: ‘I hereby tell you that...’. This is the dimension of second-personal authority: the authority a speaker has to determine that their utterance is an illocutionary act, and, in this case, the authority we have as speakers to determine that our utterance is a telling and gives the addressed audience an epistemic reason for belief by making our knowledge available.A speaker enjoys this authority in normal circumstances but not all circumstances. This is because second-personal authority requires recognition: it requires that the audience understand the speech act to be one of telling, or to share this understanding with the speaker. In fact, Moran proposes three conditions here: sincerity, understanding, and manifest conditions. In telling an audience something, a speaker must present herself as sincere. Thus, in telling lies, a speaker must also purport to be sincere. Lies are necessarily intentionally deceptive. Perceived insincerity will block audience belief but will not stop the audience from thinking of the lie as a telling; all that is needed for the audience to think this is that the speaker present herself as sincere. Lies are still told. The act of telling is undermined in the case of Moorean utterances (“It's raining out but I don't believe it”). Such utterances annul their own telling in declaring their own insincerity. Audience uptake, or a shared understanding, is equally necessary for the accomplishment of an illocutionary act. For instance, if an audience hears a promise as an exercise in pronunciation it will be annulled as a promise. Similarly, if an audience merely takes a speaker's telling to be the laying down of clues, and so as just a piece of evidence, this understanding of the telling undermines it as a telling. It ceases to be the intended telling, and so cannot be known to be such by the speaker. (Thus, the ‘disharmony’ described in “Getting Told and Being Believed” cuts deep in leading to a failure of self-knowledge.)However, in normal circumstances speaker and audience share knowledge of the illocutionary act performed when the speaker tells the audience something. Moran acknowledges that capturing this ‘mutual knowledge’ condition has been philosophically problematic, and offers an interesting explanation as to why: higher order states can always be achieved by each party acting as a ‘Holmesian detective’. He then develops the Strawsonian idea that the telling must be essentially avowable. The idea, I think, is this. The speaker's knowing what she is doing when she tells an audience something requires her to share a perspective on the illocutionary act with the audience. This shared perspective is necessary since without it neither party knows what their role in the conversation amounts to, what each is or should be doing. Thus, what becomes philosophically central is the idea of a conversation, or shared practice. And in this respect tellings—along with other illocutionary acts like promises, apologies, commands, and so on—are fundamentally social acts, or, as Reid says, social acts of the mind.It was with this last bit that I struggled: how to square the illocutionary authority a speaker has over her utterance with her dependence on her audience for this utterance being the illocutionary act she intended. I still think that the assurance account needs to include reference to a background of trust as an essential part of the explanation as to how speaker and audience come to a shared understanding of acts of telling (Faulkner 2011, 2020). The worry here is that speakers lie, and in lying a speaker presents herself as sincere. Since we know this, we don't always take tellings at face value. This raises the following question: when we do take them at face value—when we take a speaker to be really assuming responsibility and so seeing their intention that we believe as a reason to do so—what is our reason for doing this? A natural answer is to appeal to what we believe about the exchange: that we believe this speaker to be trustworthy, for instance. But if this explanation is tendered, the next question is, what warrant have we for this belief? We seem pushed back to evidence or reliability; the epistemological options seem exhausted by reductive and nonreductive theories. The way out of this, I've argued, is to appeal to an attitude of trust, which is itself invited and expected. This is the ingredient that allows the audience to take things at face value, and so achieve the shared understanding that can be exemplified in a conversational exchange.As important as trust might be, this is a worry about a detail. In the round, The Exchange of Words is a brilliant and philosophical transformative read

Epistemology · Perception · Receipt · Relation (database) · Sign (mathematics) · Sociology · Testimonial · Epistemology, Ethics, and Metaphysics · Hermeneutics and Narrative Identity · Philosophy · Psychology · Rhetoric and Communication Studies