Friday, July 9, 2010

Objectivism, Perspectivism, and Contextualism

For reasons too complicated to get into here, Alphonse is a jury of one. He reviewed the case carefully and judged that the accused, Dave, was guilty as charged. He says it:

(1a) He is guilty as charged.

Let’s suppose that he is confident:
(2a) I know he is guilty as charged.

Let’s suppose we’re confident in him, and so we say that he’s right both times.

For reasons too complicated to get into here, there is another jury of one. This one member jury has the same evidence for believing that the accused is guilty (well, change the names but otherwise it's the same evidence), but he is not guilty. So, while this second speaker, Bridget, is perfectly reasonable in believing that she speaks the truth, she speaks falsely when she asserts (1b) and (2b):

(1b) He is guilty as charged.
(2b) I know he is guilty as charged.

Now, consider:
(3a) I ought to convict.
(3b) I ought to convict.
(4a) I know I ought to convict.
(4b) I know I ought to convict.

Suppose (3a) is true. It does not follow that (4a) is true, but there is no principled reason to think it has to be false given everything we have said. So, suppose it is true, too. What should we say about (3b) and (4b)? This is where things get tricky. The natural thing for the objectivist to say is that these claims are both false. According to the objectivist, you ought to Φ if Φ-ing is best-objectively and since it seems that convicting is not the best thing for Bridget to do, (3b) is false. Owing to the factivity of ‘knows’, the objectivist should say (4b) is false, too. The natural thing for the prospectivist to say is that (3b) is true. According to the prospectivist, you ought to Φ if Φing is
best-prospectively (i.e., maximizes expected value). Since we are assuming that Alphonse ought to convict, the same should be true for Bridget. The same options are best-prospectively for them. If (3b) is true and we have no reason to think that Alphonse owes an epistemic advantage over Bridget, perhaps the prospectivist should say that (4b) is true as well. Or, more guardedly, (4b) is true if (4a) is.

In cases like this, I think the objectivist enjoys an advantage over the prospectivist. Suppose Bridget convicts Clarence and after serving twenty years, Clarence is cleared of the charges. They clear him of the charges because of DNA evidence, evidence that was not available at the time of the trial. Intuitively, Clarence was wronged and is owed compensation. The prospectivist, it seems, has to deny this but this is what seems counterintuitive. It would not just be beneficent to help Clarence after he is freed. He is owed something. But, if giving him what he is owed is righting some past wrong, there has to be a wrong in the past and the prospectivist cannot find one.

This seems like a plausible principle:

(PP) If the speaker knew at the time of the conviction that she ought to vote to convict, the convicted will not be owed compensation for being sentenced.

This is the sort of the principle that the objectivist and prospectivist might agree on, they just disagree about when you ought to vote to convict. The objectivist says that what you should do depends (in part, on occasion) upon the facts ‘beyond’ the evidence, but the prospectivist denies this. The objectivist and the prospectivist deny that ‘ought’ is ambiguous.

Enter the contextualist. On one version of contextualism, one designed to accommodate some intuitions about Regan's mineshaft example, if the speaker asserts 'A ought to Phi', that is true only if A's Phi-ing is best-prospectively given the contextually determined body of evidence. This might be A's evidence, but it needn't be. Suppose Clarence knows he's innocent and so knows that Bridget speaks falsely if she asserts (1b) and (2b). We might imagine that Clarence cannot be present for the trial and so he knows just this: Bridget is conscientious and will review the evidence very carefully. Clarence knows the evidence either strongly points in favor of guilt or innocence, but not which. His lawyer texts him to say that a verdict has been reached, but his lawyer likes to create suspense and so he doesn't say what the verdict was. Because Clarence knows what he knows about Bridget and the evidence, it seems that on the contextualist view, Clarence can say truthfully, 'Whatever Bridget says about (3b) and (4b), she speaks the truth'. His lawyer then tells him that Bridget said that he should be convicted and that she knew that she should convict him. So, it seems, Clarence can say, "I know she shouldn't have convicted me and so know that she didn't know she should convict me, but she did speak the truth." So, according to Clarence, "Bridget spoke truthfully when she said "I know I ought to convict Clarence" but I know she should not have convicted Clarence".

That seems bad and now I wonder about (PP). Do contextualists think he gets compensation when he's exonerated? If they can say that Clarence is owed compensation, it looks like they have to say that speakers say something true when they say ‘Clarence is owed compensation in spite of the fact that the speaker who said ‘I know I ought to convict Clarence’ spoke the truth’. If we could climb down from these quotation marks, it seems that the contextualist is coming out in favor of owing compensation to those who are not owed compensation. If we cannot climb down from these quotation markes, it seems that the contextualist can take no position on (PP). I guess that's why contextualists haven't joined the innocence project.

It seems to me that objectivists and prospectivists can have an honest debate about whether someone like Clarence is owed compensation for a past wrong, but I don't see how the contextualist can contribute to this debate.

4 comments:

CR said...

In both cases the evidence available to the jury was identical.

Now you said it doesn't follow from the proposition expressed in (3a) that I should convict, to the proposition that (4a) I know I ought to convict. I don't see how you are not expressing that there is some kind of logical gap between:

What a person S ought to do and what S knows that he ought to do.

But of course this too is riddled with assumptions. We might wonder whether that assumption above is true. I think quotidian conversational contexts permit one to place ‘knows’ in front of any assertible: if it is permissible for me to assert that snow is white is some context, I think that whatever permits that assertion also permits that assertion plus ‘...and I know it’. And this appears on the surface at least to cohere with Gricean maxims (perhaps they are not identical though). If their evidence which speakers recognize supports an assertion that p, I don't see how that same evidence would fail to permit to the addition of ‘and I know it’. This does not mean that I do know it. Depending on ones views about human knowledge, whether justified or permissible assertions such as ‘It is raining’ or even ‘I know it is raining’ are significantly related to knowledge is unclear.

And if from its being true that I ought to convict someone, that that does not entail I know that I ought to convict someone, seem to imply that the intension of ‘ought’ is disconnected from human knowledge. To put another, perhaps more recognizable way, that means that since the extension of ‘what I ought to do’ and what ‘I know I ought to do’ can diverge, you have assumed (in a non-psychological but logical sense) that some kind of moral realism is true. If certain varieties of moral anti-realism were true, its being true that you ought to convict so-and-so would just entail that you knew it: facts about what one ought to do would just be epistemic.

Second, your used of ‘objectively’ is worth discussing. Would a proposition be less of an ‘objective’ fact if that proposition were true just because I believed it? If moral subjectivism is true, then if I believe that murder is wrong, then murder is wrong. So, by disquotation (going in), ‘murder is wrong’ is true. But would that not be an ‘objective’ fact? Of course we tend to contrast ‘subjective’ with ‘objective’, and this is not entirely unjustified, but that alone would not explain how the proposition expressed by ‘murder is wrong’ failed to be true in some robust objective sense. Some of the minor points I touched on here are not relevant to the overall aim of your discussion but I thought they were worth noting.

Chris said...

In both cases the evidence available to the jury was identical.

Now you said it doesn't follow from the proposition expressed in (3a) that I should convict, to the proposition that (4a) I know I ought to convict. I don't see how you are not expressing that there is some kind of logical gap between:

What a person S ought to do and what S knows that he ought to do.

But of course this too is riddled with assumptions. We might wonder whether that assumption above is true. I think quotidian conversational contexts permit one to place ‘knows’ in front of any assertible: if it is permissible for me to assert that snow is white is some context, I think that whatever permits that assertion also permits that assertion plus ‘...and I know it’. And this appears on the surface at least to cohere with Gricean maxims (perhaps they are not identical though). If their evidence which speakers recognize supports an assertion that p, I don't see how that same evidence would fail to permit to the addition of ‘and I know it’. This does not mean that I do know it. Depending on ones views about human knowledge, whether justified or permissible assertions such as ‘It is raining’ or even ‘I know it is raining’ are significantly related to knowledge is unclear.

And if from its being true that I ought to convict someone, that that does not entail I know that I ought to convict someone, seem to imply that the intension of ‘ought’ is disconnected from human knowledge. To put another, perhaps more recognizable way, that means that since the extension of ‘what I ought to do’ and what ‘I know I ought to do’ can diverge, you have assumed (in a non-psychological but logical sense) that some kind of moral realism is true. If certain varieties of moral anti-realism were true, its being true that you ought to convict so-and-so would just entail that you knew it: facts about what one ought to do would just be epistemic.

Second, your used of ‘objectively’ is worth discussing. Would a proposition be less of an ‘objective’ fact if that proposition were true just because I believed it? If moral subjectivism is true, then if I believe that murder is wrong, then murder is wrong. So, by disquotation (going in), ‘murder is wrong’ is true. But would that not be an ‘objective’ fact? Of course we tend to contrast ‘subjective’ with ‘objective’, and this is not entirely unjustified, but that alone would not explain how the proposition expressed by ‘murder is wrong’ failed to be true in some robust objective sense. Some of the minor points I touched on here are not relevant to the overall aim of your discussion but I thought they were worth noting.

Clayton said...

CR,
I apologize for not getting back earlier, I've been swamped the past few days. I should start with clarifying (3a) and (4a). I didn't want to work from the assumption that (3a) entailed (4a), but that's because I wasn't taking a stand on the connection between them. So I don't disagree with the remarks you're making about asserting p and asserting you know p.

"o put another, perhaps more recognizable way, that means that since the extension of ‘what I ought to do’ and what ‘I know I ought to do’ can diverge, you have assumed (in a non-psychological but logical sense) that some kind of moral realism is true."

That's probably right, but it's also common ground for the objectivist and prospectivist views I have in mind. They are just competing versions of moral realism. They disagree about whether ignorance subverts obligation.

On the "objective" point, I think that's right. I didn't define objectivism and prospectivism here just because the post was taking too much time and I'm lazy. But the views I had in mind were from Michael Zimmerman's _Living with Uncertainty_. Basically, on objectivism, you ought to A only if there is no available alternative to A, B, such that you can do A, you can do B, and B > A in terms of deontic value. (What makes for deontic value might be subjective states.) On prospectivism, you ought to do what is best-prospectively, meaning (roughly) you ought to maximize expected value rather than actual value (except when this comes to the same thing).

Mandeville said...

Eye witness (subjective) testimony shouldn't be allowed, but since it is, if someone convicted is later found innocent by objective facts, the original jury should to liable for retribution. This will change the way jurors think and, to some degree, renders this discussion moot.