Side Note: The Documentary Fallacy Fallacy

In Who’s afraid of A. C. Bradley?

Henry Oliver

comes out in favor of “talk[ing] about Shakespeare’s characters as if they were people.” If “character criticism” is abandoned, you’ll miss most of what is good and important in the plays.

But who has said otherwise? Turns out, many. Henry summarizes L. C. Knights’ influential argument:

drama before the age of Shakespeare was largely that of the morality play, in which archetypes were used to make moral points. The most obvious descendant of that sort of drama are the characters in Ben Jonson who represent certain “humours”—these are not individuals, but types, comparable with the work of the Roman playwright Terence.

But Knights’ position—“that we cannot infer anything about a character that is not stated in the text—is flat wrong”:

As A.D. Nuttall pointed out in A New Mimesis, when a character sits up and yawns, we infer they were asleep. When they give a start, in certain situations, we assume they are guilty. (e.g., Macbeth.) Nuttall says that Shakespeare never mentions Hamlet’s legs. Are we not allowed to infer them?... It sounds silly, but those are the terms Knights set, and Nuttall makes clear that Knights was arguing against a straw-man.

Still, many were, and continue to be, on Knights’ side. Their reasons vary. Henry mentions arguments presupposing Marxist or postmodernist literary theory. Not all are so sophisticated. In 1966 A. J. A. Waldock (whom I generally admire) warned against “The Documentary Fallacy,” the mistake of thinking

that Hamlet [for example] is a document: that it is a literal transcript of fact: that it somehow records what, at that given time and place, an interlinked set of people said and did. [If this were right,] all the instruments of historical research, all the power of modern psychology, might justifiably be brought to bear ... on every issue raised by the play...A document, preserving by some miraculous means the record of what really took place, would open up endless possibilities of conjecture; not a trifle, not the obscurest detail, but might be the key to the ultimate truth. (Sophocles the Dramatist)

What is Waldock’s case against this “mistake” that, c’mon, no one is so confused as to actually make? Well, he tells us,

What renders [this] thesis...ridiculous is the simple fact that a play is a work of art: a piece of writing most carefully put together in order that it may achieve a most carefully calculated effect. ...[U]nlike life, [a play] has no depth. Literature operates on a thinnish crust, and there is nothing underneath this crust.

What renders this argument ridiculous is the simple fact that appreciating a play is an imaginative enterprise; and that what one is to imagine is, exactly, that one is witnessing facts about what an interlinked set of people said and did. Indeed it is not wild to suggest that, in this imaginative project, all the instruments of historical research, all the power of modern psychology, might justifiably be brought to bear. Yes, all too obviously, Hamlet is fiction, not a document; it may still be, for all that, that properly reading it requires treating it (in one’s imagination) like a document. That the documentary fallacy is a fallacy is itself a fallacy.

The Knights/Waldrock thesis does, however, hold within itself a kernel of truth, which emerges when Waldrock asks us to

Consider for a moment King Lear. We know a good deal about Lear the man; we feel his personality very intensely; in a sense he is more real to us than most of the people we know. But we do not know what he fancied for his dinner...we do not know what sort of a chairman he made at his council meetings; and it is not the slightest use our trying to find out these things.

Of course we also do not know what Shakespeare’s great-grandfather fancied for dinner. But here what did he fancy? does have an answer; we just lack any evidence that could make that answer known to us. It’s different (isn’t it?) with King Lear. There we lack knowledge because there is nothing to know. What did Lear fancy for dinner? has no answer. If Knights was wrong to think that there is nothing to a character beyond what is explicit in the text, it remains true that fictional characters are incomplete in ways real humans are not. But if Knights was wrong about where they become fuzzy, and where they fade out together, where then does the boundary lie?

---

See also: Against incomplete story worlds.