Contextualism and Relativism

Let’s distinguish relativism and contextualism in the following way. Relativism introduces variability in the calculation of a truth value given a proposition, and contextualism introduces variability in the calculation of the proposition expressed.

I’ve taught contextualism regularly for the past several years, and a pattern has emerged. I introduce contextualism as providing a middle ground between Moorean optimism and skeptical pessimism. I’ve taught this both to graduate students and to upper division undergraduates. Each time I face a dilemma.

So, first I sometimes try the careful and precise method. I carefully distinguish between sentences, propositions, contexts, the theory of expression, and the theory of alethic evaluation. I tell them what a relativist says and what a contextualist says, and I insist on the distinction. I never, never, never, when doing it this way, say or allow to be said statements like “S knows that p in ordinary standards contexts,” or “S doesn’t know that p in high standards contexts”. These remarks mask the difference between relativism and contextualism, as well as violating use-mention strictures. So I insist. And the result is that no one finds contextualism very appealing.

The other way is to allow the remarks above, and talk this way myself in characterizing the view. I don’t talk initially about the difference between relativism and contextualism, but leave that until the very end. Students find the view very appealing when presented in this way, but upon questioning, it is clear that they hear it as a version of relativism. (The reason that the use/mention violation leads to a version of relativism is that a relativist, unlike a contextualist, can endorse a standard disquotation principle about truth for sentences. Well, that’s not quite right, but it’s close: if you’re a relativist, you’ll disquote with abandon, and if you’re a contextualist, you shouldn’t.) That is, they think of the truth value of the claim that S knows p as being relative to whether or not the context is a high standards context. And, so they think, relativism is cool!

I find this immensely frustrating, since I have a pretty strong aversion to relativistic views anywhere (though I try to put that to the side when presented with arguments for such views 🙂 ). So I wonder whether this experience is a common one. Have others tried to teach the view precisely, being meticulous about disallowing use-mention lapses? Or does the reception of the view really piggy-back on seeing it as a kind of relativism?


Contextualism and Relativism — 15 Comments

  1. Hi Jon,
    A small caveman point: the ‘relativism’ bit of relativism always seemed to me to be in the ‘variability’ bit, and no tied so closely to truth (I think noncognitivists can be relativists, for instance). So contextualism seems to me to be just ‘relativism about which propositions get expressed’ (or, misleadingly, ‘meaning relativism’ concerning knowledge claims). What you call relativism is just a kind of relativism (misleadingly, ‘sematic relativism’ concerning knowledge claims).
    Perhaps the caveman point offers some help: I think your scruples about relativism may be too strong; let students think contextualism is relativism and that it’s thereby cool. Surely not every relativist view gives you hives; you don’t think *your* etiquette is the right one, better than the rest. You can then argue students into your view (if you can), that relativism in epistemology is dorky and uncool.

  2. Nice points, Robert. I think I’ll resist just a little here, though of course not enough to be a realist about matters of etiquette! A first point is that contextual terms don’t vary in meaning from one context to another, and nobody should run around saying that relativism is true because we’ve got indexicals in our language, or gradeable adjectives in our language. So the theory of expression takes sentences with fixed meanings and assigns propositional contents depending on the context of utterance or inscription. So I think that relativism proper must be restricted to a certain kind of variability, to avoid saying that relativism is true because of indexicality and demonstratives. Relativism, so understood, will involve variability, but variability about truth.

    What to say about matters of etiquette, I don’t know, though…except that I don’t want to be a realist here. Coordination stories look most plausible here, don’t you think?

  3. Jon,
    Well, I think relativism in a certain area *is* true because we have indexicals. It’s just not interesting relativism, not a claim that raises the sorts of issues that we worry about under that heading. Holding that the propositions expressed by sentences with indexicals in them is relative to the context of their utterance (as presumably we all do) just shows that we are all relativists in a certain respect; it is uncontroversial. The reason the label ‘relativism’ makes people’s heart beat faster is that it is relativism about a certain area of concern…morals, mostly…not because people are mad dog global absolutists.
    You may think this is all just arguing over a label (that’s why i called it a small point). Bu the problem with tying relativism too closely to truth is that it makes it more difficult to see noncognitivism as a potentially relativist view. Surely the attitudes expressed by moral statements might vary from culture to culture. What say you about that? Morals are relative to culture, on that view, but not because the truth of moral statements vary. Or do you just think that noncognitism has an easy argument against relativism (no truth, therefore, no relativism)?
    Etiquette? Yes, you are a relativist about it.
    I think we need the broadest banner possible to wave over ‘relativism’. It doesn’t have to be contentious. Or that’s my current view.

  4. Robert, I think the differences here are almost all terminological. I like distinguishing relativism from contextualism to keep clear on whether alethic evaluation is involved or just the theory of expression itself. On the non-cognitivism point, I think there are couple options. My favorite is that a noncognitivist ought to try to save the view by adopting a minimal approach to truth, and then noncognitivists will be relativists (in my usage). If they can’t do that, then I’d say they are eliminativists of a certain kind–they are eliminativists about moral facts and hence about moral truth.

    That still leaves your point unaddressed, though, since there is *something* relative about the attitudes.

    In either your terminology or mine, however, contextualists do not want the variability to enter in at the point of alethic evaluation, but rather at the point of determining what is expressed. And so making acceptance of the view depend on misdescribing the view in terms of variability of the first sort should still be troublesome.

  5. In ethics, most relativists say that the content of a moral judgement shifts with features of the context, such as who is speaking, who is listening, whose action the judgement is about, as a way of explaining why the same sentence can get different truth values in different contexts. So I take it that people like Harman and Dreier (my paraidgm case ethical relativists) postulate variability in “what is expressed” to explain variability in truth values.
    (What I say here needs to be complicated a bit to handle two judgements about the very same action involving different words, and also to not count the same form of words about different actions as generating versions of relativism, but the above simple version gets the main point across.) So I’m with Robert here.


  6. Mark, welcome to the blog! It’s great to smoke out lurking ethicists!

    I agree that one can use the term ‘relativism’ in different ways, and that there’s a sense in which any divergence from a Fregean logically perfect language can introduce relativity. But if the relativity is contained in the theory of expression only, we need an additional story to make it anything for anti-relativists (those for whom ‘relativism’ is pejorative) to worry about. There isn’t such an explanation for ordinary contextualism, since no defender of orthodoxy about truth would be interested in denying that the thought expressed by “I’m tired” varies by context. In the case of Harman’s relativism, I can see how the additional explanation might go, perhaps in terms of opposition by Kantian absolutists. If we wish, we can call all these views relativistic, and distinguish between two types: the type that finds variability in the theory of expression, and the kind that finds variability in the theory of truth (after we’ve settled the question of proposition expressed). I don’t think much hinges on which way we go here.

    There is an interesting question, though, as to why a position like Harman’s is worrisome to non-relativists about truth when contextualism about indexicals and demonstratives is not worrisome at all. Perhaps the difference is in terms of the properties taken to be expressed by the supposedly contextual moral terms. But mere contextualism in ethics shouldn’t worry non-relativists. A utilitarian could adopt an account of the proposition expressed so that the utilities and their bases in desires were constituents of the proposition in question, and then we get utilitarian contextualism, a moral theory that wouldn’t trouble those who oppose relativism at all (I mean, again, those for whom the use of the term ‘relativism’ is pejorative). So if Harman’s contextualism is troubling to such anti-relativists, what’s the difference between Harman’s view and this wacky version of utilitarianism? I mean: what’s the difference that could explain the differential response to the two by the anti-relativist? I’m not sure…

  7. Jon,
    I may not have been clear in what I said, so let me clarify something and then give some answer to the question you end with.

    I took it that what you mean by ‘theory of expression’ includes what content is expressed in a judgement. Thus I was saying that the moral relativists of which I am most aware explain the shift in truth values of the same utterance in different mouths, by saying that the contents of those judgements also shift. That way they are not committed to accepting a contradiction because one content may be true while the other is false since they are different.

    Jamie Dreier’s version has the content of “such and such is right” mean something like, such and such is what would be allowed by the most consistent and coherent working out of a contextually salient moral system. What makes a moral system salient is a complicated matter, but among other things the shared commitments of speaker, agent and listener may play a role in determining the salient moral system. But since different people may play those roles in different contexts and this may change which system is salient, the truth value of the claim made by “such and such is right”may shift beccause the content expressed by the form of words may shift.

    In Gil Harman’s “What is Moral Relativism” in the old Kim and Goldman volume Gil argues for three claims one one of which is labeled ‘normative moral relativism’ and the other of which is labeled ‘moral judgement relativism,’ if my memory is correct. The former just says that seemingly conflicting moral judgements can both be true, and the second that the contents of moral judgements contain some sort of relativity to standards or commitments of the agent. I take it that it is because of this relativity that Gil thinks that the seemingly conflicting judgements can shift truth values.

    So why do some people find this sort of view frightening while epistemic contextualism of the sort that started the thread seems not to be? Well, one thing that people are often worried about with respect to ethics is partly that the rightness or wrongness of an action is supposed to be tightly connected to actions. If such and such is wrong, you are not supposed to do it. If the kind of relativity that moral relativism builds in is relativity to the “Motivational set” of the agent in question, then two agents could be such that such and such is right for one and wrong for the other. But that would mean that one agent could do the one and the other could not and they would both be acting fully correctly. Depending on what we fill in for ‘such and such’ that could seem very troubling. Relativism of this sort seems to make the judgements escapable by changing one’s motivational commitments.

    Relatedly, I don’t think that the epistemic varieties of contextualism lead to the same worries because they are not tied to action as tightly. A person could accept all of the different judgements expressed by the varying contextually determined standards and not be in internal conflict. On the other hand, if a person believes that he or she should do whatever is right or wrong she will be in conflict over what to do if she accepts more than one of the judgements in her own case.

    It is relevant here that it is generally the moral commitments and motives of the person or persons who are relevant in the context which shifts the contents of the judgements, whereas that does not seem to be what is going on with the epistemic contextualist positions of which I am aware.

    Am I making any sense?


  8. Mark, Your explanation of what is troubling here may work, but I’m not sure. If knowledge is the norm of assertion, then we get the same connection to action in the epistemic case that you cite in the moral case. Moreover, some versions of contextualism rely heavily, almost exclusively, on salience of the risk of error, and those theories seem to make shifts in context quite subjective in much the same way as you describe for the moral theories in question.

    My suspicion here is this. The difference between, in my terminology, contextualism and relativism isn’t a significant difference in moral theory (in roughly the sense that anyone incensed by one will be equally incensed by the other). The differences between the two are significant in other areas, however, so it’s not surprising to find moral theorists wanting to classify the two kinds of variability in the same category, when those outside of moral theory think of them as fundamentally different. Does that seem right?

  9. Jon,

    If what you mean by “knowledge is the norm of assertion,” is that people who know something are thereby entitled to assert it, it is the case that there is a connection between the the truth of claims to know and the action of asserting. But what I don’t think is the same is that the actions which go with the truth of the claim exclude the actions which go with the truth of the seemingly contradictory claim. I take it that a contextualist could claim to understand what someone means by “I know it but I don’t *know* it,” if their theory says that the change in inflection changes the standards in mid-sentence. So a speaker need not choose between asserting the two seemingly but not actually contradictory things. But I don’t see the same thing happening so easily with moral judgements insofar as the action that goes with thinking something wrong tends to exclude the actions that are permitted when we deny that it is wrong. If the only actions that went with moral judgements were assertions of their contents the ethical and epistemic cases would be more similar. (I hope I am understanding what epistemic contextualism is so that this much makes sense.)

    You started the thread with, “Relativism introduces variability in the calculation of a truth value given a proposition, and contextualism introduces variability in the calculation of the proposition expressed.” In ethics, most of the people labeled “relativist” are contextualists in your sense. Very few are also relativists in the sense defined. I would have expected the same thing to be true in other domains as well. Would this fit with your suspicion?

    Robert raised the issue of how to characterize the sense in which expressivism is/can also be a form of relativism. Clearly it seems there are versions on which two seemingly conflicting moral judgements are fully correct. (I myself think these are the more natural versions to hold if you like non-cognitivvism, but some disagree.) I want to call this a kind of relativism, though of a non-cognitivist sort. I suppose that simple noncognitivist views don’t need to shift the contents of the judgments with the shift in agents because the judgements have no propositional contents to begin with. So maybe it is right to say that the relativist noncognitivists need not be contextualists, but that there is a strong push for cognitivist relativists to also be contextualists, on pain of being commtted to asserting a contradiction.

  10. Mark, sorry for the cryptic reference to the norm of assertion point. The idea is that you shouldn’t say what you don’t know to be true. I’m not a fan of this claim, but many are.

    On the argument of the first paragraph, I’m not sure I grasp what your after. Contextualists can countenance true instances of “p&~p” as long as the context shifts at the conjunction. But then what is assertible before the conjunction will track with true sentences of the form ‘S knows p’, and the same after the conjunction. So what the speaker can assert before the conjunction is different from what can be asserted after it.

    On the expressivist question, I’m still perplexed. For one thing, I don’t see how to pull apart the notion of truth from the notion of correctness. If expressivist remarks lack propositional content, then I can see calling the remarks appropriate or inappropriate, but not correct or incorrect. And it’s obvious that what’s appropriate or inappropriate is relative to circumstance of utterance. So I understand that. And if expressivists grant propositional content, but go minimalist about truth, then the language of correctness doesn’t trouble me, and the claim of relativity is understandable as well.

  11. Jon, Thanks for the clarification. I think we are in agreement on the argument of the first paragraph. My point was that there was no conflict between the different contextualist judgements, nor really between courses of action since as I understand it, the way in which it is asserted is part of what changes the context. So while a person may not literally be able to say both seemingly conflicting things at once, they can say them within a few seconds of one another. It need really not be a choice between doing one or doing the other.

    I don’t think this is so for the typical moral case for two reasons. (1) typically the judgement that x is right and the judgement that x is not right call for conflicting courses of action. Even if the content ‘right’ contributes to that judgement shifts, so long as x does not shift it’s contribution there will be a practical conflict if x is right permits doing x, and x is not right forbids it.

    (2) Insofar as the typical feature of the context that relativists pay attention to is who is making the judgement (speaker relativism) or who is it about (agent relativism) (often it is both on the assumption that the two share a basic moral outlook), it is going to harder to generate the sorts of quick shifts of context necessary to get one person in a position to truly say ‘x is right and x is not right’ even if there is some time between.

    This last point also is relevant to the comment about whether expressivist theories count as relativist. Not every shift in content due to context makes a theory relativist in the sense that people usually intend when talking about ethics. If the situation changes the judgement is about a different act or act-type and then there is really no conflict in the relevant judgement. Relativists are generally taken to be talking about the same action or act type while shifting the person or group that the judgement is made relative to. So if I judge an act of a type right and you judge an action of the very same type wrong, and the only difference is either that I’m making the one judgement and you are making the other (for speaker relativism) or that you are doing the one action and I am doing the other (for agent relativism) then the view will count as relativist. But if the shift between me and you changes the act from one in which one of us harms a stranger and the other harms a close friend, that won’t be counted as relativism by most.

    So anyway, I do think that the most natural thing for expressivists to say (qua metaethical theorists) is that two people can express conflicting attitudes towards the same actions or act-types and be such that neither is making any mistake, except perhaps in the sense that relative to some particular ethical perspective the judgement counts as a mistake. But since there are no non-perspective relative ways of going wrong, this allows that each is symetrically situated with respect to the other’s perspective and that there are no neutral perspectives from which one can be judged more correct than the other. Of course the expressivist who says this might accept a certain perspective and go on to make the judgement that one of the two gets it right relative to this perspective and the other does not.

  12. Mark, I’ve been distracted by other things and didn’t get back to this discussion…sorry for the delay. One minor thought, though, on this part of your last comment:

    My point was that there was no conflict between the different contextualist judgements, nor really between courses of action since as I understand it, the way in which it is asserted is part of what changes the context. So while a person may not literally be able to say both seemingly conflicting things at once, they can say them within a few seconds of one another.

    Whether the way a knowledge claim is asserted changes the context depends on the version of contextualism one is referring to. For Lewis, for example, just using the concept of knowledge makes the stakes impossibly high. But it is not part of contextualism per se to endorse this point.

  13. I just taught about contextualism to my class today. I don’t know which strategy I followed. I don’t believe I said things like, “S knows that p in ordinary standards contexts”; instead, I think I converted to the meta-language, saying things like, “In ordinary contexts, ‘S knows p’ is true”. (But I might’ve slipped.) At the end, I discussed whether contextualism implies that “if I enterain more alternatives, I will know less,” and I compared this to the (false) statement, “If I fly to New York, New York will be here.” As far as I can tell, they seemed to understand this point. I haven’t asked any questions designed to test whether they’re confusing contextualism w/ relativism, though.

    About what ‘relativism’ is: I suppose one should define it so that it’s close to what most people who self-apply the term believe. If the students think they’re “relativists,” then we should figure out roughly what they believe that they’re trying to describe with this. (I suspect it is nothing coherent. I suspect it’s something like, “Whatever someone believes is true, but not really true but only ‘true for them,’ which means ‘believed by them,’ but there’s no such thing as truth, and also, p and ~p can be equally true.”

Leave a Reply

Your email address will not be published. Required fields are marked *