(A little more friendly to Williams this time, and a little bit of a tease at the end...I'm in the midst of the continuation right now...more to come...)
Conviction and reflection might seem to be awkward partners, their relationship constantly strained. Although Socrates claimed that the unexamined life is not worth living, we might suspect that because reflection tends to undermine our sense of certainty in our own beliefs—since in reflecting we question and critique those very beliefs—that reflection also tends to undermine conviction. This might seem to be especially true of moral beliefs, since a little exposure to the diversity of moral opinion and practice throughout the world may lead us to reflect upon the extent to which our own beliefs are influenced by our own particular upbringing and cultural and historical situation. Bernard Williams has suggested in such cases that “reflection can destroy knowledge,” which is to say that reflection can deprive us of confidence in our own ways of living and valuing, and thereby our convictions. At least, reflection might destroy our certainty that our way living and valuing is the right way.
Destruction is not, it should be kept in mind, always a bad thing. We want to destroy our illusions and false opinions, the biases that prevent us from recognizing truth. However, one might think that if knowledge is good and reflection can destroy it, then something has gone wrong in those cases when reflection does destroy knowledge. But how could reflection ever do such a thing? Shouldn’t what we call knowledge be more durable than that? A short—and somewhat unhelpful—answer is that it depends on what exactly we are willing to count as knowledge.
What is potentially unsettling about Williams’ point does not turn so much on how we choose to use the term knowledge but rather on the fact that reflection—as well as exposure to other ways of living and valuing—can undermine our confidence in our own moral beliefs. We can be led to question whether the certainty we have about our own convictions is justifiable when we consider that others have felt just as certain about the rectitude of various other systems and practices. What is equally unsettling is that simply refusing to reflect on such matters will seem dishonest (or dogmatic or lazy): if we refuse to reflect, then we may well fall (or have fallen) for just about anything. Unless we are foolish or arrogant enough to believe that we have already got everything right, then we will see that we cannot make any progress in our own moral or intellectual life without reflection.
The danger of reflection is not that it unsettles us; sometimes we need to be roused from cheap comfort. Rather, the danger is that reflection can lead to a cramped kind of skepticism that induces paralysis or despair. We can lose our grip and not know how to go on. In having lost our certainty, we may be led to the thought that we have lost everything—that without certainty, we cannot (or should not) allow ourselves to have convictions.
But this is a strange position.
Very good!
ReplyDeleteI think there could be an On Certainty-based response here, to the "strangeness" you mention at the end.
There are various papers by different authors that differ vastly in their approaches and argumentation, but whose shared message seems to be is roughly that On Certainty dissolves the disagreement between moral cognitivism and noncognitivism (or moral realism and antirealism) in much the way it dissolves the disagreement between Moore and the epistemic sceptic. Examples include Judith Lichtenberg's Moral Certainty, Michael Kober's On Epistemic and Moral Certainty: A Wittgensteinian Approach, or Nigel Pleasants's Wittgenstein and Basic Moral Certainty.
I think myself that the Wittgensteinian analysis of "knowledge" in Moore's sense could very well be extended to the analysis of the knowledge which Williams says can be destroyed by reflection, with potentially interesting results. What you call the "strange position" could be come to be seen as strange in the same way not just Moore's claims about his two hands but (importantly) also the sceptic's denials of Moore's strange claims are strange.
Peter Winch's Certainty and Authority, which is a similar attempt at marrying On Certainty to the problem in political philosophy of grounding political authority, would seem to be at least as relevant as well. It's a very neglected paper, but one of my biggest personal favourites from him.
Thanks, Tommi. I was looking at Lichtenberg's paper last night, which is one I think VERY highly of. I've read Pleasants, too (though maybe I should look back at that paper to see how it might help here). I don't know the Kober paper, so many thanks for that. (And if there's time--of which there's never enough--I'm always happy to look at Winch.)
ReplyDeleteWhere I'm actually going turns out to be close to what you had pointed out about Williams in the previous post: that convictions don't require certainty, even if there are some moral judgments--such as those with which Lichtenberg begins her paper--about which we are as certain as anything. Those judgments (e.g. that it's wrong to have sex with three-year-olds) aren't (at least not necessarily) much like the sorts of convictions we have in mind if we want to know what a person's moral convictions are. We are--or at least I would be--asking a very different question, which is related to what Cavell has in mind when he talks about wanting to understand what position a person is taking responsibility for and whether it is something that--even if I disagree, or just don't share that view--I can respect.
What is equally unsettling is that simply refusing to reflect on such matters will seem dishonest (or dogmatic or lazy): if we refuse to reflect, then we may well fall (or have fallen) for just about anything. [...] [T]he danger is that reflection can lead to a cramped kind of skepticism that induces paralysis or despair.
ReplyDeleteThis was strongly reminiscent of something I couldn't put my finger on. But I remembered it now - it's the closing passage of Renford Bambrough's The Scope of Reason: An Epistle to the Persians:
The sceptic thinks of himself as the enemy of credulity, but his scepticism is itself a form of what he takes himself to be fighting. The suggestion that something is uncertain needs to be supported by just as high a standard of evidence as any other view. Whether something is doubtful is not a question on which one side can be allowed to award itself a lighter burden of proof than another. Scepticism is a form of credulity, just as cynicism is a form of sentimentality. Cynicism is not usually seen in this light, mainly because we tend to think of a particular group of sentiments when we think of sentimentality. On reflection we can see that it is just as easy and just as sentimental to indulge or wallow in one's toughness or realism or brutality as in one's sensibility and compassion. Similarly, it is just as credulous to question the obvious as to decline to question the doubtful or the obviously false. To somebody who does not accept that Moore has two hands we might suitably say - and this is what I say to those who are sceptical or relativist about the scope of reason - "If you disbelieve that you'll believe anything".
Bambrough is an interesting figure - a pupil of John Wisdom, and a stereotypical old-school British university philosopher (at Cambridge), very Wittgensteinian, but also very rationalist and objectivist in a way commonly associated with neither "old" nor "new" Wittgensteinians. (His best-known paper is probably the 1970 "A Proof of the Objectivity of Morals".) I remember how crestfallen I was when I asked Cora Diamond for her opinion of Bambrough - he was editor of Philosophy while she published several classic papers there, including "Eating Meat and Eating People" and "Throwing Away the Ladder" - and she turned out to have had quite a low opinion of him. In hindsight, he's probably not at all for everyone (in violation of his own aspiration to objectivity), but he is for me.
Where I'm actually going turns out to be close to what you had pointed out about Williams in the previous post: that convictions don't require certainty, even if there are some moral judgments--such as those with which Lichtenberg begins her paper--about which we are as certain as anything.
André Maury, who supervised my master's thesis, has a short, very little-known paper (published only in a privately printed Festschrift) where he argues that On Certainty simply severs the conceptual link that has traditionally been assumed between certainty and knowledge - because there seem to be some empirical, a posteriori propositions that are in a sense too certain for "knowledge" to be a natural-sounding name for them. Convictions don't require certainty, but perhaps certainty doesn't require knowledge.