Intellectual Virtue and Causation
Intellectual virtue and causation is one of my favourite topics, mostly because I think it's really important. Here's why: for starters, I think that virtue-based epistemology is the right way to go. (I won't defend this here, but just trust me :) If you're on board so far, which you should be, then it should be clear that an important task will be to clarify the relationship between, on the one hand, an agent's exhibiting intellectual virtue, and on the other, her coming to form a true belief. Literature on VE and epistemic luck shows us that within a VE account, there must not be a disconnect here.
How to block worries about a disconnect? Greco, Sosa and Zagzebski all invoke (in slightly different ways) the notion of 'causation.' The story goes something like this: S knows that p just in case S's true belief that p is because of S's exhibition of intellectual virtue. That's a general template.
Now, all of the 'big three' seem to understand the conditions under which the relevant causal claims would be true or false as something that is either (i) intuitive, or (ii) a function of what is salient in a causal explanation. This is so because they quite frequently illustrate cases in which exhibition of intellectual virtue does or does not 'cause' the agent's having a true belief by appealing to example cases, and the example cases are ones in which it is 'intuitive' that what is salient in a causal explanation for the agent's true belief is or is not intellectually virtuous agency (in accord with their examples). For example, Sosa's ballerina example, Zagzebski's example of the judge, and Greco's example of the gambler all are intended to make these sort of 'causal' points. I don't dispute that their examples constitute helpful ways of thinking about how intellectual virtue should be connected with an agent's true belief in cases of knowing. What is problematic, though, is that none of these authors gives us much more to go on over and above the intuitive appeal to example. (Greco is an exception here)
A general worry should be this: if, for any theory, the conditions under which one thing causes another are meant to function importantly within the theory, then the theory should be supplemented with a corresponding theory of causation. In particular, the theory should say just what the conditions are that must hold if one thing causes another.
It is here where our leading VE'rs are somewhat bankrupt.
I'm not sure what Zagzebski's current view is, but Sosa and Greco both rely importantly on the notion of "explanatory salience."
Their idea is, but crudely, that an agent's exhibiting IV causes her true belief in a way that is sufficient for her to know just in case her IV is 'salient' in a causal explanation for her coming to have a true belief.
Greco has a nuanced view about salience--one that cashes out salience as a function of the relevant interests and purposes that frame the context in which the explanation is given. Sosa's view is less nuanced.
Both, however, are taking--by relying on salience--a line that reduces causal relations to causal explanations. Helen Beebee, Donald Davidson and others take objection to this reduction. In particular, they reject that causal explanations entail causal relations. But this point aside, there is I think a more interesting worry: whether it be causal relations or causal explanations that are at issue in deciding whether a knower's IV is appropriately hooked up with her coming to have a true belief, we will be letting some strange causal ducks in the door once we let 'salience' be the adjudicator.
My thought here is this:
Salience clearly attaches to events, but also and perhaps just as frequently, to absences of events. For example, John's giving $5 to a beggar might be salient in explaining why the beggar was able to buy lunch, and similarly, Flora's failure to water her flowers might be salient in explaining why her flower's died (to cite an example from Beebee).
It stands to reason then that both the exhibition of intellectual virtue and the absence of the exhibition of intellectual vice might each be candidates for the role of 'salient' explainers.
Now, a venerable tradition led by Lewis and others holds that events are the sort of things that can stand in causal relations. Suppose that's right. If it is, we get a weird aporia:
1. Whether one knows depends on whether her exhibition of IV caused her to have a true belief. (Sosa and Greco)
2. Whether her exhibition of IV caused her to have a true belief is a matter of whether her IV is salient (Sosa and Greco)
3. Both events and absences of events can be salient in explaining her true belief.
4. Events, but not absences of events, can be causes (Event Theory of Causation).
One way to try to escape the aporia is to deny (4) and say that absences of events can be causes. But Beebee's "Causation and Nothingness" makes a very compelling case against taking this route.
The better routes would be to either reject (1) or (2). Because Greco and Sosa rely importantly on salience in explaining the role of intellectual virtue in cases of knowing, I think they should reject both (1) and (2). They should first deny (2) (for the reasons Beebee gives) and deny that whether an exhibition of IV causes one to have a true belief is a matter of whether her IV is salient. And next, they should reject (1); they should say that whether one knows depends on whether her exhibition of IV is salient in explaining her true belief, rather than on her exhibition of IV 'causing' her to have a true belief.
Rejecting (1) and (2) is needed if salience is to have an important role in explaining the role intellectual virtue has in cases of knowing, and a result of rejecting (1) and (2) is that the notion of 'causation' is made unnecessary.
This would seem to be a bad thing if explanations just are causal explanations. But they're not. There are both causal explanations and non-causal explanations. I'm increasingly inclined to think that the leading VE theorists--insofar as they are wedded to salience--have non-causal explanations in mind.
What is the difference?
Suppose you ask, "Why did the 8 ball drop into the corner pocket?" My striking the cue ball into the 8 ball explains why the 8 ball dropped into the corner pocket. This is a causal explanation.
Suppose you ask, after finding out that I am richer than my neighbor, "Why are you richer than your neighbor?" That I have a million dollars in the bank (I wish) would explain why I am richer than my neighbor, but it didn't cause me to be richer than my neighbor. My having a million dollars would be salient in explaining why I'm richer than my neighbor, but it is not a causal explanation.
Another example of a non-causal explanation: Suppose the Cardinals beat the Cubs by the score of 5-0 against the Cubs' best picther. Without knowing the score, but knowing they beat the Cubs, you ask "how did the Cardinals not get shut out?" I could explain to you why the Cardinals didn't get shut out by pointing out that Albert Pujols hit 5 home runs. But Pujols' hitting five home runs didn't 'cause' the Cardinals to not get shut out, even though it implies that they did. After all, whatever caused the Cardinals to not get shut out caused this to happen before Pujols hit his fifth home run.
What should be gathered, then, is this: insofar as the notion of 'salience' is meant to function importantly within a VE account for the purposes of articulating the conditions under which an agent's exhibiting IV and her coming to have a true belief are to be appropriately (in cases of knowing) connected, talk of causal relations, and perhaps even causal explanations should be dropped in favour of the notion of non-causal explanations.
A discalimer: Perhaps salience itself should be dropped from the picture. I'm open to that possibility. But if it's given an important place, then a consequence should be that the 'causal' language that is so common in these accounts should be excised. If, on the other hand salience is excised instead, then these accounts could continue to use the causal language that they do, so long as the supplemental account of causation bolstering the theory of knowledge isn't either a "reduce causation to salience" theory or a "derive your theory of causation from the intuitiveness of these examples" theory.