Augustine on Testimony

Philosophical work on testimony has flourished in recent years. Testimony roughly involves a source affirming or stating something in an attempt to transfer information to one or more persons. It is often said that the topic of testimony has been neglected throughout most of the history of philosophy, aside from contributions by David Hume (1711-1776) and Thomas Reid (1710-1796).¹ True as this may be, Hume and Reid aren't the only ones who deserve a tip of the hat for recognizing the importance of testimony: Augustine of Hippo (354-430) affirms the place of testimony in human cognition, at least in his later writings.

In what follows, we consider three questions raised by Augustine's thinking about testimony: the analytical question of what sources count as testimony (Section I); the epistemological question about the status of testimony-based belief (Section II); and the doxastic question about the circumstances in which it is appropriate to believe on the basis of testimony (Section III). We outline Augustine's view of testimony by examining his answers to these three questions. Finally, we'll briefly situate Augustine within the tradition of thinking about testimony (Section IV), by way of conclusion.

A few preliminaries. Augustine occasionally uses 'testimonia' to talk about one thing being a sign for another thing. For example, Augustine remarks that a Stoic philosopher's pallor during heavy seas testified to his fear (civ. 9.4).² This isn't the philosophically interesting sort of testimony; any sign would 'testify' to its significate in this sense. Yet Augustine doesn't always indulge his tendency to use 'testimony' to talk about any sign whatsoever. Often he explicitly treats testimony as a source that affirms something in an attempt to transfer information, a specific type of intentional activity. Brief passages are found in his De libero arbitrio 2.2.5.14-15, Confessiones 6, De Trinitate 15, De civitate Dei 11, with more sustained discussions in the course of his De utilitate credendi, De fide rerum inuisibilium, and Epistula 147. These texts will serve to explain Augustine's answers to the three questions mentioned above and to clarify the development of his views on testimony.

I The Analytical Question

What sources count as testimony? Augustine maintains that testimonial sources include both spoken and written words: 'We are,' he says, 'informed by spoken or written words, or some other means' (ep. 147.3.8). These 'other means' include gestures, such as hand signals, nods, and the like.³ Now for some terminology: call the testimonial source the testifier, the testifier's affirmation the testimony or testimonial report, and the recipient of the testimony the hearer. In his several discussions, Augustine appears to assume that a testifier doesn't need the firsthand authority of an eyewitness or source - that is, a particular testifier may be many stages removed from the primary testifier (ep. 147.4.10). Since Augustine does not discuss his assumption explicitly, however, we will mainly consider cases involving primary testifiers. Note that eyewitness testimony should not be confused with expert testimony. Testimony from experts - be they philosophers, physicians, scientists, or whatever - needn't concern something witnessed firsthand. Instead, expert testimony may concern the implications of some theory or method. Augustine doesn't discuss expert testimony as such, though he does consider Biblical authors experts in their particular domain. Our attention will be directed to the general case of testimony, not to expert testimony.

Now Augustine's many examples of testimonial reports concern places, people, events, and states of affairs. These examples involve testimonial reports about (i) foreign places,⁴ (ii) histories of people and nations,⁵ (iii) current events,⁶ and (iv) a hearer's own biographical information.⁷ Augustine draws a further distinction between testimonial reports. Since a hearer may be 'unable to demonstrate the fact because the event is already in the past' (f. inuis. 2.4),⁸ she cannot always check or verify a report's authenticity. We'll say that a report is distant only if the hearer cannot, even in principle, check it using non-testimonial evidence. However, the...