Augustine was the first of his circle of friends to disembark from the Manicheanism train. Thereafter, and especially once he’d become a Christian, he spared no effort in turning others away from what he always acknowledged was a seductive cosmology that offered a straightforward and simple (to his mind, simplistic) account of the origins of evil. Most of what Augustine wrote in service of this end is polemical, and Augustine’s willingness to slander his co-religionists makes these texts a funny but unenlightening read.
Perhaps because it takes the form of an epistle to one of his still-Manichean friends, De utilitate credendi is different. In this interpersonal rather than polemic scenario, Augustine cannot speak as though, since one must believe in something, the choice between Catholicism and Manicheanism is a given. He not only needs to show that Manicheanism deserves to be abandoned, but also to demonstrate that Catholicism, rather than a general skepticism, ought to take its place.
It would be going too far to say that Augustine proves this case, or even, on the whole, that he argues it well. Augustine’s own option for Catholicism over some other form of Christian belief (or even a “secular” neo-platonism) will seem arbitrary to modern readers, and De utilitate credendi suffers from a certain degree of bad conscience about this. The best argument that Augustine can muster in favor of Catholicism is that it draws a crowd. Schools of rhetoric, where one can indeed learn something, also draw a crowd. But so does a garbage fire.
Where Augustine’s arguments carry more weight is on an antecedent question, whether we should believe things on faith or only accept claims on the basis of rational proof. Since Catholicism insists on faith, whereas heretics (and Manicheans) promise rational proof for their doctrines, Augustine finds himself in the unusual position of having to argue, not only for the validity of belief, but indeed for its superiority in many cases over rational knowledge.
Augustine’s strategy for showing this looks to us like a radical extension of what counts as belief. We know either what we perceive with our senses (although see his Retractiones on this) or what we can rationally understand by way of facts and concepts. Everything else, we only believe – or, if we think we know it rather than believe it, “opinate.” This is a theory of knowledge with deep Platonic roots, as Miles Burnyeat has shown, but Augustine applies it in some rather original ways. Here, for instance, is my favorite of Augustine’s test-cases:
“But, if what is not known should not be believed, I ask how children are supposed to serve their parents and love them with mutual family feeling, when they don’t believe that these are their parents. For this cannot be known by reason in any way, but is believed about the father on the mother’s authority; about the mother, in many cases, not even on the mother’s own authority but on that of midwives, nursemaids, servants. Now, isn’t it possible that someone who can steal a child and put another in its place might just as well deceive by accident, being in error themselves? Nevertheless we believe, and we believe without any doubt, that which we agree we cannot know. Unless this were so, who doesn’t see that family feeling, the most holy bond of humankind, would be violated by the most overweening crimes? For even a madman would not think him blameworthy who does his duty to those he believes to be his parents, even if they aren’t. Who, on the other hand, wouldn’t judge him worthy of death who, just because he fears to love false ones, fails to love his real parents?” (DUC 12.26)
Kinship – all kinship – is a matter of belief, not knowledge. If we accept the very plausible theses about infant memory that Augustine puts forward in the Confessions, then his argument here does in fact follow: we “know” who our parents are in the same sense that one could be said to “know” who Jesus’ parents are, because one has heard a story. In short, we only believe that our parents are our parents.
This is an argument that puts paid to most ideas about the “naturalness” of consanguinal kinship by contrast with the “artificialness” of affinal or otherwise ascriptive kinship. Both, on Augustine’s reasoning, are equally matters of belief. Yet it is not an argument that, to my knowledge, skeptical practitioners of anthropology in the tradition of Sahlins have ever put forth. This may be because it derives from a recondite source, or because it employs a notion of belief with which anthropologists have come to feel uncomfortable. There are negative reasons for this – the attacks launched by analytic philosophers like John Searle against forms of relativism that frame all mental commitments as “beliefs” – as well as positive ones – for instance, Eduardo Viveiros de Castro’s polemic against anthropology as a logging of delusional “beliefs” and in favor of anthropology as the study, through alien cultures, of “possible worlds.”
It seems to me, though, that Augustine’s notion of belief is broad enough that it does in fact capture most of what we might mean by “mental commitments” – excepting only that strange class of commitments which Augustine calls “knowledge,” and which may be unique, conceptually, to cultures in the tradition of Ancient Greece. In any case, few anthropologists would disagree with Augustine that this kind of knowledge is rare – the exception rather than the rule – and surely no kind of basis for social life:
“I could set forth many examples by which it would be shown that nothing of human society would remain stable if we committed to believing nothing that we could not know by direct perception.” (DUC 12.26)
Where Augustine may seem to modern readers to have gone wrong is in asserting this broad notion of belief without, as we would do, drawing a further distinction within it. Some entities – tables, cars, parents – we can know to exist by direct perception; about these, we then believe predicates – “old,” “yours,” “mine” – which we cannot by Augustine’s standards know to be correctly applied. When it comes to Catholicism, though, we need to believe in God’s existence as well as that a particular set of predicates hold true about him. To accept that entities should be believed to exist without evidence, just on somebody else’s say-so, would be to commit ourselves to populating the universe with all sorts of gods, not to say entering into any given schizophrenic’s delusion.
There are plenty of arguments that Augustine would probably want to present on this point, but I don’t think that many of them would hold water for us. One argument that Augustine might not have cared to present, but which does carry a certain amount of weight, has to do with the theory of interpretation he presents in the DUC and elsewhere. There are, he says, three kinds of interpretive mistakes. You can take fiction for truth – which would be like someone reading Vergil and thinking that Aeneas really had gone to Carthage. Or you can fail to see that what an author takes for truth is really fiction – which Augustine says would be like believing Lucretius. Or, finally, you can derive a better lesson from a work than its author intended to put there – which would be finding the truth in fiction. The first two of these are evils; the last of them, though it is a form of error, may be not evil but good:
“What, then, if I should hear about a person I had some affection for that he, although he was already wearing a beard, said that in the presence of many that he so liked boyhood and infancy that he swore he wanted to live that way, and it should be so proved to me he had said this that I would be a fool to deny it? Would I really be deserving of rebuke if I supposed that he, in saying this, wanted to indicate that he liked innocence and a mind free of those desires in which the human race is ensnared, and for this reason I loved him more than I had before? Although perhaps he had really been hankering for a certain freedom to play and eat and a lazy leisure characteristic of boyhood? Now suppose he died after this was announced to me, and I was not able to ask anyone to explain his opinion. Would there be anyone so nasty as to blame me for praising the man’s purpose and intention through the words that I had heard? A just appraiser of things would perhaps not hesitate to praise my opinion and intention, since I both liked innocence and, being a man, thought well of another man in a dubious matter, when I might have thought ill of him?” (DUC 4.10)
When it comes to dubious matters, there’s no harm, says Augustine, in believing the “better” of two (or more) interpretations, and in fact “credulity” in this sense reflects well on the believer. Thinking back to the parallel example Augustine offers in the passage I quoted before of someone who acts with pietas towards people he takes to be his parents even though they are not really his parents, we can see the appeal of Augustine’s argument. If it could in addition be shown that believing in the God of the Catholics was really the “better” interpretation of the cosmos, then Augustine would have succeeded – not in showing that other people should share his belief in that god, but in showing that this belief was, in his own case, just fine. Not the usefulness of belief, then, but something like the non-usefulness of non-belief.
There’s a lesson in this for skeptics, which is that the will to truth can’t stand as its own justification: the mere fact that a belief is held on epistemologically unsatisfactory grounds is insufficient argument that the belief shouldn’t be held. What needs to be shown in addition is that the belief in question is bad, either consequentially (although this, given that beliefs tend to entail divergent visions of their own consequences, seems like an unproductive tack to take) or by itself. Only in a case like that would non-belief turn out to be useful.