Elementary, My Dear (Gary) Watson

Gary Watson’s work on virtue ethics contains some of my favorite work in the field. I like Watson — he’s a clear writer, and I find his work on taxonomy interesting. Reading a bit more of Zagzebski’s book (Divine Motivation Theory) I see that she discusses Watson’s call for a “pure virtue ethics” (which I’ve always wondered — is this just a call for a form of Michael Slote’s “agent basing”?). Reading her chapter started me thinking again about some issues in Watson (I’m thinking here of his “On the Primacy of Character” from his Identity, Character, and Morality, MIT Press, 1990). Specifically, I’m thinking of Watson’s attempt to simultaneously juggle different notions of the “good” when trying to understand how a virtue theory could be evaluatively primary and basic.

Here’s a central part of Watson (found in a footnote — I find that this piece has a lot going on in the footnotes!). It says:

Insofar as virtues must in the end by characterized by their contribution to the good for human beings, the notion of the good will be primary relative to virtue. But there would still be a point to thinking of the theory under consideration as an ethics of virtue, since virtue still remains basic relative to the concepts of right and a good state of affairs…So understood, on the theory I am trying to describe, virtue is prior to that notion, and so the priority claim is maintained. (pg. 70 in the Statman collection)

I gather that Watson is up to a few things here.

1. Only a pure virtue ethic is a real virtue ethics.

You know the complaint. I hear it all the time from my colleagues who study ethics. They say (things like) “but if virtue is what promotes (expresses) the good, then isn’t it just consequentialism (deontology)?” Here, it seems they are following Rawls; only consequentialism (C) and deontology (D) are basic, all other purported theories of ethics are derivative of these two.

In a way, they have a point. If “the good” (outcomes) or “the right” (duty) are basic and we then work backwards and then somewhere down the line construct a theory of virtue, then there really isn’t a “pure” theory of virtue that competes with C and D. So what we need, if we want to argue for virtue ethics as autonomous, is a theory that doesn’t have virtue as a secondary feature, but as the primary feature.

2. A pure virtue ethic has virtue as what is evaluatively primary.

I take Watson’s point here to be this: when we start off, at the ground floor level, suggesting what one should and should not do, these claims must emerge from virtuous states. Only in such a way is virtue primary. So, basically, what is good (as an outcome, say) must be what expresses virtue, not the other way around. Obviously here theories such as character consequentialism (such as Ivanhoe’s portrait of Confucianism) would not count as virtue ethics, but rather as indirect forms of outcome-based theories, where character virtues are instrumental means (similar, perhaps, to rule utilitarianism) towards advancing certain ends that are seen as good independently (Ivanhoe’s point is more subtle, and he argues for a constituent relationship, but this is incidental to my point in this post).

So, what is good as an outcome must be dictated by what the aim of expressed virtue happens to be (I wonder here if there is a relationship between this point and Foot’s point about morality as a system of hypothetical imperatives, but I am not sure).

3. Some Goods are Dictated By Virtue, Some Not, But That’s Fine

Yeah, so #3 is the tough spot. Virtues, he says, must be “characterized by their contribution to the human good.” So here it sounds, at least initially, as if the good is not what is expressed by virtue, but rather the other way around — that virtue tracks the good, a good that exists primary to virtuous states. Watson’s “good” here is “human nature” essentially. In another spot, Watson explains:

It may be useful to compare a theory of excellence for a non-human animal. The judgment that a lack of attention to her cubs is a sign of imperfection in a mother tiget (though not in the father) is based upon a notion of a good specimen of tiger. This idea in turn depends upon what is normal for or characteristic of tigers. None of the judgments is mediated by any notion of the value of a tiger’s living a life characteristic of its species. On an ethics of virtue, the same goes for people.

Watson’s point, I think, is this: this notion of good, or “good for” is different from the notion of good that he’s worried about, which is “good as an outcome” (Zagzebski mentions this as well). According to Watson, what is “good for” is not evaluative, just descriptive. Good knives cut well, but that doesn’t mean that I should want a knife that cuts well. On the other hand, “good outcome” is evaluative, such as “the baby is not in pain” or something of that sort.

So Watson’s claim is that a pure virtue ethic can be grounded in what is “good for” humans without sacrificing evaluative primary away from virtue. Basically, the virtuous person desires to express what a good human being would look like, but the normativity expressed by the aim of the desire is generated by and grounded in that virtuous state. That’s just what virtuous people do. At the same time, virtuous people who take on these aims will seek to promote (perhaps) certain outcomes as goods, but again these goods, even if they are seen as evaluative, will again be grounded in the virtue, and do not have any evaluative strength apart from those virtuous drives and desires.

A. As the post name implies, I wonder sometimes whether it is helpful to use two uses of “good” here, because it is inherently confusing. Let’s come up with a new word, eh? I prefer “authentic” for “good for” but I’m sure there are other terms out there.

I say “authentic” because Watson’s program here in some ways reminds me of Existentialism and the claims by some such authors that authenticity is something we can aim at, but not something that we ought or ought not to do. For almost all of the existentialists, they want to avoid normative claims that exist apart from the concerns of evaluation-making agents (God is dead, and all that, you know). So for Heidegger, whether we achieve authenticity or not is no strike against us; we aren’t good or bad people (in an evaluative sense), and it marks no detraction of our Being one way or the other. For Sartre, similarly, it’s not supposed to matter (really, at the end of the day in an evaluative sense) if we fall into “bad faith.”

In both cases, although especially for Heidegger, perhaps, we get the claim that the existential nature of our being has a certain structure, just as much as Watson might say that we have a certain human nature. Whether we align with that structure doesn’t make us good or bad (in that evaluative sense), but rather simply points out whether or not we have, in the end, “become what we already are” (to use a Nietzschean phrase).

The connection with Existentialism, however, brings me back to Watson, and my worry at times whether such claims about human nature are evaluative or not. Clearly his argument relies on the claims not being evaluative. But it’s hard to get away from the thought that they are not (I’m not sure if this intuition is well grounded — that’s part of my confusion here). Oddly enough, the Existentialists had a similar concern. Heidegger couldn’t help but to think that Sartre’s “bad faith” was dripping with moral evaluation; from his perspective, Sartre couldn’t help but roll his eyes at Heidegger’s notion of “authenticity,” which he himself saw as intrinsically evaluative.

Is it? Is “good for” different from “good outcome” in the sense that the latter is always evaluative, and the former not necessarily so? I find myself drawn to Watson’s analysis here, and I find his attempt to locate a pure virtue ethics very interesting. But I still cannot dislodge my intuitions that this use of “good” harbors evaluation (much as Sartre and Heidegger worried about one another).

B. I also have a side question: Is Watson’s call for a pure virtue ethics simply another name (ten years earlier) for what Michael Slote calls in 2001 an “agent-based” virtue ethic, where (a) what is value of the right is determined by whether it expresses virtue and (b) the value of virtue is on its own terms independent and primitive? Watson’s program appears to me to be similar to Slote’s.

As usual, a spattering of questions without any answers.