Intellectual Virtues and the Labyrinth of Understanding — Epoche C2
Linda Zagzebski's Virtues of the Mind (1996) defines knowledge as true belief arising out of acts of intellectual virtue, and the question this essay puts to that definition is whether intellectual virtues do their most characteristic work there at all. On her account a virtue is a deep and enduring acquired excellence of a person with two components: a characteristic motivation — in the intellectual case, some specification of the motivation for truth — together with reliable success in attaining the end that motivation sets. Open-mindedness is not merely a tendency to arrive at true beliefs when confronted with rival views; it is that tendency growing out of a settled desire to get things right, and a person who stumbles into truths without the desire has not exercised the virtue. The definition of knowledge then leans its whole weight on the phrase arising out of : the belief must be true because of the virtuous features of the act, and not merely alongside them. My claim is that this apparatus, whatever its merits as an analysis of propositional knowledge, is not where the virtues earn their keep. They earn it in understanding — in the condition of a person who can move about inside a body of information rather than merely hold its items. One correction is owed before the argument begins, because the earlier and shorter version of this essay got it wrong in a way that changes what has to be proved. That version argued contra Zagzebski, casting her as committed to virtues as constituents of knowing and therefore as an opponent of the thesis just stated. She is not. In 'Recovering Understanding' (2001) Zagzebski argues that understanding has been mislaid by an epistemology fixated on propositional knowledge; that the object of understanding is not a proposition but the structure of a body of information; that it admits of degrees in a way knowledge does not; and that it is transparent to its possessor, since a person who understands can tell on reflection that she understands, whereas a person who knows cannot always tell that she knows. That is this essay's own thesis, published by the author it was being aimed at, five years after the book. The real opponent lies elsewhere, and naming him correctly raises the cost of the argument rather than lowering it. The reduction, stated at its strongest The position that genuinely denies the thesis is reductive: understanding is a species of knowledge, so there is nothing distinctive to say about the virtues that produce it beyond what is already said about the virtues that produce knowledge. Stephen Grimm defends this in 'Is Understanding a Species of Knowledge?' (2006), and his answer to his own title is yes. To understand why an event occurred is to know its causes; more generally, to understand a subject matter is to know the relations of dependence that hold among its elements — which items would have been otherwise had which others been otherwise. On this view understanding is not a further state layered on top of knowledge but knowledge with a particular sort of content, and it inherits knowledge's factivity, its susceptibility to defeat, and its analysis. The reduction is not a straw position, and it answers the two arguments most often brought against it. The first is Jonathan Kvanvig's, in The Value of Knowledge and the Pursuit of Understanding (2003): a reader acquires an accurate account of Comanche history from a reliable book that happens to sit in a library otherwise stocked with fabrications, so that his true beliefs are environmentally lucky. Kvanvig's intuition is that the reader's knowledge is undermined while his understanding survives, which would show the two states come apart. Grimm's reply is that the case does not isolate understanding but only relocates the luck: shift the fabrications so that they bear on the dependence relations themselves rather than on isolated dates and names, and the intuition of surviving understanding disappears along with the knowledge. The second argument is from degrees — understanding comes in more and less, knowledge does not — and here the reply is that a body of knowledge also comes in more and less, since one may know more or fewer of the dependence facts, and gradability of the body does not establish a difference in kind. So the reduction cannot be defeated by pointing at gradability or at a single Gettier case. It can, however, be made to look much less consequential than it appears, and that is the strategy this essay takes: grant Grimm the content claim entirely, and show that the epistemic problem of acquiring that content is of a different character from the problem of accumulating truths. If it is, then the traits of character that solve it are not a subspecies of the traits that make individual beliefs reliable, and the interesting theory of the virtues sits on the understanding side whatever the taxonomy of states turns out to be. How much a list of truths leaves open Take the labyrinth in the title literally, because the metaphor is doing quantitative work. A labyrinth is a place where local facts are cheap and global structure is dear: every wall can be seen and touched, and no wall tells you the route. Suppose an inquirer has secured $n$ true propositions about some domain — all of them, no Gettier trouble, each held for good reasons. On the reductive account itself, understanding requires in addition that she know which of these depend on which. A dependence relation is directed and irreflexive: $p$ may depend on $q$ without $q$ depending on $p$, and nothing depends on itself. The candidate relations are therefore the ordered pairs of distinct propositions, of which there are $n(n-1)$ — not $n^2$, because the $n$ self-pairs are excluded, and not $\binom{n}{2}$, because direction matters and each unordered pair supplies two candidates, so that $2\binom{n}{2}$ recovers the same figure. Each candidate is independently in or out, so the number of dependence structures compatib