Tuesday, September 23, 2008

In-Between Believing: The Implicit Racist

Many Caucasians in academia sincerely profess that all races are of equal intelligence. Juliet, let's suppose, is one such person, a Jewish-American philosophy professor. She has, perhaps, studied the matter more than most: She has criticially examined the literature on racial differences in intelligence, and she finds the case for racial equality compelling. She is prepared to argue coherently, authentically, and vehemently for equality of intelligence, and she has argued the point repeatedly in the past. Her egalitarianism in this matter coheres with her overarching "liberal" stance, according to which the sexes too possess equal intelligence, and racial and sexual discrimination are odious.

And yet -- I'm sure you see this coming -- Juliet is systematically racist in her spontaneous reactions, judgments, and unguarded behavior. When she gazes out on class the first day of each term, she can't help but think that some students look brighter than others -- and to her, the black students never look bright. When a black student makes an insightful comment or submits an excellent essay, she feels more surprise than she would were a white or Asian student to do so, even though her black students make insightful comments and submit excellent essays at the same rate as the others; and, worse, her bias affects her grading and the way she guides class discussion. When Juliet is on the hiring committee for a new office manager, it won't seem to her that the black applicants are the most intellectually capable, even if they are; or if she does become convinced, it will have taken more evidence than if the applicant had been white. When she converses with a janitor or cashier, she expects less wit if the person is black. Juliet may be perfectly aware of these facts about herself; she may aspire to reform; she may not be self-deceived in any way.

So here's the question: Does Juliet believe that the races are intellectually equal? Considering similar cases, Aaron Zimmerman and Tamar Gendler have said yes: Our beliefs are what in us is responsive to evidence, what is under rational control; and for Juliet, that's her egalitarian avowals. Her other responses are merely habitual or uncontrolled responses. But this seems to me to draw too sharp a line between the rational and the irrational or merely habitual. Our habits and gut responses are inextricably intertwined with our reason, perhaps often themselves a form of reasoning. Furthermore, imagine two black students in Juliet's class. One says to the other, in full knowledge of Juliet's sincere statements about equality, "and yet, she doesn't really believe that the black people are as smart as white people". Is that student so far wrong? Aren't our beliefs as much about how we live as about what we say?

Does Juliet have contradictory beliefs on the issue (as Brie Gertler seems to suggest about a similar case)? Does she believe both that the races are intellectually equal and that they're not? It's hard for me to know what to make of such an attribution. We might say that part of her believes one thing and part believes another, but there are serious problems with taking such a division literally. (Like: How do the different parts communicate? How much duplication is there in the attitudes held by the different parts and in the neural systems underlying those attitudes?)

Should we simply say that Juliet does not believe that the races are intellectually equal? That doesn't seem to do justice to her sincerity in arguing otherwise.

I recommend we treat this as an "in-between" case of believing. It's not quite right to say that Juliet believes that the races are intellectually equal; but neither is it quite right to deny her that belief. Her dispositions are splintered -- she has, we might say, a splintered mind -- and our attribution must be nuanced. Just as it's not quite right either to say that someone is or is not courageous simpliciter if he's courageous on Wednesdays and not on Tuesdays, it's not quite right to simply ascribe or deny belief when someone's actions and reactions are divided as Juliet's are.

Zimmerman tells me I'm whimsically throwing overboard classical two-valued logic -- the view, standard in logic, that all meaningful propositions are either true or false -- but I say if two-valued logic cannot handle vague cases (and I think it can't, not really), so much the worse for two-valued logic!


Unknown said...

This is weird, but maybe 'believe' is contextually sensitive like 'know'.

Let P be: She believes the races are equally intelligent.

In contexts where the standard for belief focuses on that which responds to evidence and is under rational control, P is true.

In contexts where belief is like Gendler's alief, P is not true.

Predicted problem: What about the context where the standard for belief equally takes into account the conditions of the first two cases I just described (high-falutent belief + alief)? It looks like we run into the same problem of whether we throw out 2-valued logic.

Still, it's a putative option.

Anonymous said...

I'll resist your warning and say that she doesn't really believe in racial equality. If she did, she would be perturbed by her dispositions and hence do everything (everything!) in her power to correct them. Don't we judge the sincerity of a person's beliefs by their behavior?

Genius said...

"Our beliefs are what in us is responsive to evidence, what is under rational control"

The problem is 'beliefs' have moments when they are like that and moments when they aren't.

And what would ones test be regarding this? Is it only where we understand our own train of thought and where we happen to follow that back to a set of premises that we find acceptable?

Brandon said...

If we take your suggestion about in-between believing, would it follow that there are never really any cases of weakness of will? One way Juliet's case could be glossed is that she believes the races are intellectually equal, but has habits that lead her to fail to act consistently with that belief; i.e., that she really does believe what she says, but that she is akratic in her spontaneous reactions. After all, if we take the view that her dispositions are splintered, there are many more dispositions than just beliefs.

Consider another case. In the CN Tower in Toronto there is a glass floor 1122 feet above the ground. I believe that it was engineered in such a way that it can hold my weight (and a great deal more). But walking out onto the floor is a great deal harder than that. And while little kids will often just run right out onto it, adults, who are capable of having a better grasp of its ability to hold their weight, nonetheless have to force themselves onto it, and some never succeed. So do they believe the floor will hold their weight? If we say no, then one can argue that we are simply playing games with the word 'belief', and treating every disposition to act (or not to act) as if it were a belief, or a sort-of belief. I have a disposition to jump when someone grabs me suddenly; this is not a belief of any sort, and I doubt many people would want to argue that it is. So it would seem we would have to be careful about what dispositions we allow to take the label 'belief'.

Unknown said...

Why can't we say Juliet believes in equality but not with the same strength as someone who answers the races are equal and acts as if they are equal. I don't quite understand why belief can't come in degrees of strength as that seems to handle the problem quite easily.

Eric Schwitzgebel said...

Thanks for all the interesting comments, folks!

Chad: I certainly agree that attributions of belief are contextually sensitive, in the sense that in some contexts it may be appropriate to attribute the belief that P and in other context not so. I also agree that only postpones and does not ultimately avoid the question of what to do about vague cases.

Anon 7:39: In my statement of the case I indicated that Juliet aspires to change her reactions. I'm not sure this is strictly necessary for the case to count as in-between, but I'm willing to go along with you that it helps. Is there any reason to think we can't amp up that aspect of the example?

Genius: I agree completely, especially with your implied reservations at the end. I suspect we have much less self-control of and self-knowledge about our trains of thought than may be implicit in views like Zimmerman's and Gendler's (and Burge's for that matter). To put it more extremely than I actually think: All that inner speech is just epiphenomenal froth, having little to do with the actual mechanisms of thinking.

Unknown said...

Weird, my comment didn't show up.

Anyway I'd pointed out the Peircean solution to this problem. He treats beliefs as habits of behavior in potential circumstances but allows for habit to come in degrees.

Thus a person who would answer to any intellectual question the equality of races but unconsciously acts in racist manners has the same belief as a person who doesn't act in subtly racist manners. The latter just believes with a higher degree of belief.

Eric Schwitzgebel said...

Brandon, you raise good points. (Are you aware that your glass floor example could be straight from Gendler?) Here's my take:

Yes, many cases of "weakness of will" can be analyzed as cases of in-between believing. But not all, and not the glass-floor example. The difference between Juliet's case and the glass floor example is that in Juliet's case there's a *broad range* of dispositions stereotypical of the belief that the races are not intellectually equal while in the glass floor case the range of dispositions is more limited. Especially pertinent are epistemic dispositions like the disposition to make certain inferences and reach certain related judgments (e.g., about a job applicant) that have little analog in the floor case.

Art: I agree that there are differences in degree of confidence. But Juliet, as I'm imagining her, is highly confident of the truth of the egalitarian view -- perhaps even more confident than someone without any pattern of racist responses. So I don't think appeal to degree of confidence resolves the issue. If you mean something else by "degree of belief" than degree of confidence, then maybe you could clarify a bit more? (I think the use of "degree of belief" to refer to patterns of betting behavior and degree of confidence in the decision theory literature masks over a more complex phenomenon.)

Clark Goble said...

Sorry Eric, someone had signed into Google with my name (which was why I didn't see my earlier post)

Anyway the issue is that confidence is not necessarily a good guide for what we believe. The obvious example of this is someone in denial about their beliefs. One could call this "in-betweenness" but I think degree of belief handles it much better. But of course this rests upon what we mean by "belief."

I fully admit I'm buying into the Peircean view of belief as a kind of habit that manifests itself in behavior (or potential behaviors). So I don't mean degree of belief or strength of belief in anything like Bayesian terms. Rather I mean that we can conceive of what behaviors a belief ideally ought entail. Then we can recognize that holders of that belief will not have all those behaviors. So we can talk about degree of belief.

Eric Schwitzgebel said...

Clark, I'm completely with you (and Peirce) on all that, I think. But then there's one more step: There's no sharp line between having a degree sufficient for the ascription of belief simpliciter and not having such a degree. It's vague, with borderline cases. So then you get the in-betweenness I stress. I think the Juliet case is vague in the relevant way, but if you emphasize her in-the-world behavior as opposed to her theoretical avowals, you might see it differently.

One infelicity in the "degree of belief" locution is that it suggests a simple spectrum from .5 to .6 to .7 etc., whereas in reality, I think, there are lots of ways to be between full belief and non-belief.

Clark Goble said...

Well, I agree there. And of course by vague you are meaning something more like Sorites cases. I'm still not sure how to handle that sense of vagueness. I'm very sympathetic to Williamson's view that there is a truth of the matter and we just can't figure it out. Although I'm not convinced by his arguments.

In that case the problem is less whether our 'racist' has racist beliefs or not than whether we can identify easily how we ought label them.

Clark Goble said...

To add, if we take Peirce's claim seriously this kind of vagueness will always be the case since the only way to determine a belief is to conduct all the verifications that would tell us if they have the belief. But of course we can't do that. There will be an astounding number of potential behaviors from a belief and we just can't test them all to see if (or to what degree) a person has the belief.

This isn't a problem to Peirce since he's a major fallibilist. But I think it highlights the intrinsic problem in any belief ascription.

Eric Schwitzgebel said...

I've got the opposite take on Williamson: I disagree with him about vagueness, but I find his arguments interesting! My guess is that the problem is in the background assumption that it makes sense to try to capture vagueness in formal logic.

On your further point about Peirce, it seems to me a question of induction: Of course we can't check the whole dispositional profile. But can we get enough evidence to be warranted in assuming "...and so on"?

kvond said...

I see that some have brought up Peirce. Aside from the notion of habit, belief is seen as a rule for action to resolve the discomfort of doubt. My two cents, as I understand belief to be both an ascription for both conscious and unconscious behavior. I don't find the answer to be a question of degrees, but of circumstances and examples. Occasions for rules of action to be applied. This is really a Wittgensteinian advisement, look to the variety and kinds of examples.

Clearly the professor believes and disbelieves that races are equal (if we can broadly assume such a statement is revelatory). In some contexts the ascription explains certain intentional behaviors, and in other contexts the alternate ascription serves.

But really assuming that there is one state of belief which would explain all behaviors in the context of race seems naive to me. It would be like saying, either Jim loves or does not love his wife. The answer to that kind of question is really contingent to shifting criteria.


Clark Goble said...

Eric, yes, for Peirce via abduction and induction we can to varying degrees have a good ideal of what degree of belief someone has.

The issue of vaguess in formal logic is interesting since it was one of the keys of Peirce's logic. So I think it can be done. (I suppose folks can disagree with Peirce here, but I think in terms of logic he has a fairly good answer - the real question is more whether there is a truth about the matter in some cases where we use the English term vagueness)

Kvond I agree doubt is pretty key for understanding belief for Peirce. I just didn't want to make things too complex here. Peirce (correctly I think) doesn't believe we have control over our doubts or beliefs. (One of the major points of his attacks on Descartes) So doubt leads us to inquiry or adjusts our habits. The nature of our beliefs move us in a dynamic way to engage beliefs.

Regarding whether it is degrees or not depends upon whether you buy into Peirce's adoption of the idea of counterfactual tests. If you do then I think degree makes a lot of sense. If you don't (and originally Peirce didn't) then yeah, context becomes crucial, but you end up having other problems (which is why Peirce modified his view)

Anonymous said...

I don't see how Juliet can believe P when her behavior confirms that her honest feelings show that she really believes not-P. I was in family therapy with my first wife about 37 years ago and I honestly believed (i.e., I honestly thought that I believed) that my then wife was a good woman and that I loved her and wanted to continue living with her. But when when the words "You whore!" -- directed at my wife -- exploded from my mouth, I knew that I really believed no such things. If you cannot act consistently with your putative beliefs, then those beliefs are self-deceptions. Juliet believes in the superiority of her ideology, perhaps, but not in the equality of intelligence between the sexes or the races. Maybe, however, I'm taking too common-sense a view of Juliet's state of mind. Does a symbolic logic analysis really lead to a different conclusion?

Eric Schwitzgebel said...

Thanks for the thoughtful and interesting comments, folks!

Kvond: I agree that it some contexts it will be appropriate to say that Juliet believes in the intellectual equality of the races (say, in a debate on the subject) and in other context it will be appropriate to say she does not (say, if two unfairly treated black students are talking to each other). However, I don't think that this implies, or that it is very useful to say, that she both believes and fails to believe. Why not say she is "in-between"? Would we say of someone courageous on Tuesdays and not on Wednesdays that he is both courageous and not courageous? I don't think that's very helpful. Better to refuse to attribute either courage or its lack simpliciter.

Clark: I must confess I don't know Peirce's logic of vagueness. Something to look up! I'm not deeply committed to the impossibility of a logic of vagueness. It just seems to me that no attempt I've seen to logicize it is even close to adequate -- with one of the main problems being allowing for metavagueness (it's being vague whether something is vague). And: Isn't context crucial to attribution in such cases, with or without degrees?

Hui-zhe: I don't think symbolic logic decides the matter one way or another. You may be right to privilege one's overall pattern of action and reaction over one's "sincere" avowals. When push comes to shove, I think the former the more important. Yet I do continue to think that if one's avowals are consistent enough and deeply enough rooted, they should not be excluded from the dispositional base relevant to deciding whether it is appropriate to attribute belief.

Clark Goble said...

Isn't context crucial to attribution in such cases, with or without degrees?

Context is crucial to meaning I think. But if it is a matter of counterfactuals rather than what is actual then that should handle it since the meaning is the ideal view across all possibilities.

Put an other way you can see things in terms of what is possible being privileged or what is actual being privileged. I think you can see movements in both directions. I think that historically analytic philosophy has tended to privilege the actual over the possible. Which is why, I suspect, Peirce's view isn't popular. What many want isn't an analysis in terms of possibility but in terms of actuality or what is before us right now.

with one of the main problems being allowing for metavagueness (it's being vague whether something is vague).

I think the way Peirce frames the issue this is always a possibility. It arises naturally in his thought since there's nothing we encounter unmediated. Now one can attack him in terms of his adopting what ends up being an epistemic infinitism. But I'm not sure how to attack him on his own grounds.

Put an other way it really depends upon what you demand for grounds of reason in a judgment. If ones intuitions make Peirce's conception of continuity seem ludicrous then a lot of the rest of his thought won't work. If you're fine with continuity and it's applications in logic and epistemology then this won't be a problem in the least.

Clark Goble said...

I don't see how Juliet can believe P when her behavior confirms that her honest feelings show that she really believes not-P.

Chen, I think the case Eric provides is so interesting precisely because this doesn't appear. For most of the "tests" of racism Juliet appears like she isn't a racist. For some of the tests it appears like she is.

If the situation were clear cut then yeah, what you say would be the case. But even if you don't think the Juliet thought experiment works one could easily adjust it to find a middle ground to fit the problem Eric points out.

Eric Schwitzgebel said...

Thanks for all your comments, Clark! I'd already been planning to spend a bit more time with Peirce, but you've inspired me even more.

kvond said...

Eric: "Kvond: I agree that it some contexts it will be appropriate to say that Juliet believes in the intellectual equality of the races (say, in a debate on the subject) and in other context it will be appropriate to say she does not (say, if two unfairly treated black students are talking to each other). However, I don't think that this implies, or that it is very useful to say, that she both believes and fails to believe. Why not say she is "in-between"? Would we say of someone courageous on Tuesdays and not on Wednesdays that he is both courageous and not courageous? I don't think that's very helpful. Better to refuse to attribute either courage or its lack simpliciter."

Hmmmm. Yours seems to be an urge to essentialize someone, to be able to make a statement that somehow describes EVERY state she would potentially be in, and in absence of that, to pose some kind of essential indeterminate state. This does not make much sense for me, if we grant that our belief ascriptions are meant to bring out the intentions and likely results of those intentions. Yes, one could say "She is inbetween believing x and not x" if one wants to somehow generalize about an unpredictability in her actions. But if we pay very close attention, we might very well be able to predict exactly how she will act. In cases of an intellectual debate, for instance, saying that she is "in-between" would simply be obfuscating. In those cases she is not in-between at all. Rather, she is quite committed and predictable.

I think the safe thing to be said is that none of these descriptions alone are sufficient, that the structure of her beliefs is simply too complex to be reduced to a simple sentence. I think that this if often the case with most belief ascriptions. What they reduce is simply a shorthand for much more complex relations, and the reduction is at best only a work-station for interpretation and prediction.

To take up your suggestion of "in-between" as a most useful reduction, I would have to ask you: "What person of all the persons in the world would not be 'in-between' on issues of race prejudice?" Were there not white Africaaners who could find Mandela racist? If everyone could be found to be race-biased under a variety of definitions and imaginable contexts, then "in-between" simply again collapses into a variety of circumstances to be examined.

What troubles me by your story is that it carries an ideological component, one in which for those that work as advocates of race equality have to "qualify" for their place as spokespersons. There is a hidden sense in which the teacher describe is somehow disqualified as an advocate due to her unconscious behavior in the classroom. She is a hypocrite. To say that she is "in-between" implies that she somehow is "undecided", and this really is not the case. In terms of decisions, she determinatively has decided. In terms of predictions of behavior, some of which falls beneath the threshold of strict decision, there is variability. This is an inexact, in fact inaccurate description I believe.

The other reading of "in-between" is that somehow she has not gone through the kinds of personal processes that makes one free of racism, she is halfway of the hill, so to speak. This perhaps is more useful since it does not have the same difficulty of misleading implications, but it has the drawback of wholly ideological, putting a functional framework upon behavior.

What I suggest is that as much as we would like to distill possible states of mind into essential statements of over-arching belief. I really do think this boils down to statements such as "I believe in God" or "I believe in America", questions of categorical allegiance. Any critical response to these questions would be I think, "Yes, I believe in God in these ways, under these terms" or, "Yes, I believe in America under these circumstances." If one wants to say that such a belief is "in-between" then perhaps yes. But EVERYONE is in-between in belief. Belief is about what you would do, think and feel in conditions which are ever and necessarily contingent.

As far as courage goes, I'm also unsure that saying that someone is in-between being courageous is helpful in any real circumstances. When weighing the courage of others we want to know how they will likely act. Can we count on them. We all are in-between being courageous. We are sometimes courageous (genuinely) and sometimes we are not. If talking about a 911 firefighter who rushed into the burning building, but earlier in the morning did not have the courage to tell his wife that he was cheating on her, does it make sense to say that he is "in-between" courageous, or to say that he is courageous and he is not? I think that latter.

Anonymous said...

You might be interested in reading Nicholas D. Kristof's October 4th Op-Ed column "Racism Without Racists" in the New York Times. (http://snurl.com/43u5n)

Kristof cites research by Phillip Goff, a social psychologist at U.C.L.A., and John Dovidio, a psychologist at Yale. They both have evidence of the unconscious racism of which many people don't believe they are guilty. Dovidio calls it "aversive racism". Here's a brief quote -- it's about the current presidential election, but it seems germane:

“In the U.S., there’s a small percentage of people who in nationwide surveys say they won’t vote for a qualified black presidential candidate,” Professor Dovidio said. “But a bigger factor is the aversive racists, those who don’t think that they’re racist.”

Faced with a complex decision, he said, aversive racists feel doubts about a black person that they don’t feel about an identical white. “These doubts tend to be attributed not to the person’s race — because that would be racism — but deflected to other areas that can be talked about, such as lack of experience,” he added.

While I haven't read either scholar's work, I hypothesize that they both provide strong evidence that the world is filled with people who are profoundly convinced of the truth and sincerity of what are in truth their false beliefs.

Anonymous said...

Why not think that there isn't just one thing that 'belief' gets at? This kind of example is one reason that Brandom opts for the term 'doxastic commitment'. On the one hand, there's the kind of state that gets expressed by explicit avowals, and on the other there's the kind of state that shows up when we interpret someone as doing some practical reasoning, the state that plays the functional role of belief in explaining action.

(Interestingly (to me anyway), this is a motivation for eliminativism that doesn't rely on the Churchlandian argument from the likeliness that folk psychology is a false theory. And what it motivates is a kind of fractioning of the concept, analogous to the fractioning of 'memory' in psychology.)

Eric Schwitzgebel said...

Adam, I agree completely with your comment. Here's how I would develop that idea: Given that there's more than one thing we're getting at, or could potentially mean, by "belief", *which* of those things should we choose as the default interpretation of the concept in philosophy? That's a pragmatic question, turning on which approaches to belief best model the empirical facts about cases we care about, generating the most useful pattern of attributions.

Therefore, my argument for a dispositional approach to belief that allows for in-between cases of believing like this is a pragmatic one: It's *advice* about how to use the word "belief" given our values and the empirical facts.