IV. Foundations for Moral Relativism
DOI: 10.11647/OBP.0029.04
I am not going to argue for moral relativism. The case for moral relativism is not an argument; it’s a pair of observations. The first observation is that people live and have lived by mutually incompatible moral norms. The second observation is that no one has ever succeeded in showing any one set of norms to be universally valid.
These observations do not prove that there is no universally valid morality, but they do lead us to wonder: If there weren’t a universally valid morality, would there be any valid morality at all? Could there be multiple moralities, each of merely local validity? To explain how there could be would be to lay foundations for moral relativism.
Formulating Relativism
According to moral relativism, saying that an action is wrong is like saying that someone is tall, a claim that is elliptical unless indexed to a reference class, since someone who is tall for an Mbuti may not be tall for a Kikuyu, and it makes no sense to ask whether he is tall simpliciter.1 Similarly, says relativism, it makes no sense to ask whether an action or practice is wrong simpliciter. Claims of wrongness must be about wrongness-for-members-of-x, where x ranges over different cultures or societies or, as I will call them, communities.2
The reason why it makes no sense to speak of tallness simpliciter is that there is no universal standard for who qualifies as tall. The standard applicable to the Kikuyu can’t be applied to the Mbuti, nor to the Inuit or Uighur, either. Similarly, says relativism, there is no universal standard of what qualifies as wrong; the only standards that exist are restricted in application to particular communities.
This claim implies that when the Kikuyu say that there isn’t anything wrong with female circumcision and the Mbuti say there is, both may be speaking the truth, because one group is speaking of what’s wrong-for-the-Kikuyu while the other is speaking of what’s wrong-for-the-Mbuti.3 Of course, the Kikuyu and the Mbuti have a practical disagreement: they disagree over how to treat young women. According to moral relativism, however, there is no proposition whose truth is at issue between them.4
Moral relativism cannot rest with this negative conclusion, however. It must go on to claim that being wrong-for-the-Mbuti is a way of being morally wrong, just as being tall-for-an-Mbuti is a way of being physically tall. In other words, moral relativism must not only deny the existence of universal morality; it must also assert the existence of local moralities. Otherwise, it won’t be relativism; it will just be nihilism.
The problem is that the relevant local institutions are mores, which seem to lack normative force.5 “Female circumcision is permissible among the Kikuyu but not among the Mbuti” is the sort of statement found in academic ethnographies, which are fastidiously non-judgmental. An ethnographer might add that members of a community adhere to its mores from a desire for solidarity or a fear of sanctions, social or divine; but these additions would just pile on more ethnographic facts. They might also suggest instrumental reasons for community members to obey the local mores, given a desire for solidarity or a fear of sanctions. But moral relativism is not the view that the universal norm of instrumental reasoning leads to different conclusions under different circumstances.
A moral relativist must claim that the mores of a community can be fundamental, underived norms. The problem for the relativist is that mores and morality are as different as facts and values. How can relativism bridge that difference?
Perspectival Normativity
The difference can be bridged by the connection between facts and their action-guiding modes-of-presentation. ‘A is wrong-for-members-of-x’ is not the complete expression of a fact until the value of x is supplied; but it cannot guide action if that value is supplied explicitly. The value of x is explicitly supplied by anthropology textbooks, which name the community in question, thereby stating a normatively neutral fact. What members of the community say, however, is simply that A is wrong, a statement that is normatively valenced. The latter statement should be interpreted as containing an implicit indexical, as in ‘wrong-for-us’, the reference of ‘us’ being supplied by the context of utterance, so that the statement expresses the fact that A is wrong for members of that community, the same fact expressed by the former statement. But the latter statement is normatively valenced because the reference of ‘us’ is left to be supplied by the context. “Female circumcision is wrong”, said by an Mbuti, is action-guiding; “Female circumcision is wrong for the Mbuti” is not.
The essential indexical
Here I rely on insights in John Perry’s paper “The Problem of the Essential Indexical”.6 Let me illustrate Perry’s thesis with a mundane example.
If I am walking down Fifth Avenue at noon on New Year’s Day, 2020, and I ask someone the way to Washington Square, he will say “It’s straight ahead”. The proposition he expresses, fully spelled out, will be that Washington Square lies straight ahead of David Velleman at noon on 1/1/2020. But the intended role of his utterance will be to say that Washington Square is straight ahead of me then, irrespective of who I am or what day and time it is. Even if I think that it is 1920 and that I am Edith Wharton, I will not be misled by this statement, although it is actually about David Velleman in the twenty-first century. Whether it expresses a proposition about David Velleman or Edith Wharton doesn’t matter for the purpose of guiding me to Washington Square.
What’s more, the statement would cease to serve that purpose if it were rephrased so as to specify the time and person concerned. “Washington Square is straight ahead of David Velleman at noon on 1/1/2020” would not tell me how to get to Washington Square — not, that is, unless I knew that I was David Velleman and that it was noon on 1/1/2020, so that I could infer that Washington Square was straight ahead of me then, irrespective of the time and person concerned. Thus, practical guidance is, in Perry’s phrase, essentially indexical, in the sense that its function depends not only on which proposition it expresses but also on how that proposition is determined by the context — specifically, on its being determined in the same way as the reference of indexical expressions such as ‘I’, ‘you’, ‘here’, and ‘now’. Spelling out the proposition so as to eliminate the role of context defeats the purpose of practical guidance.7
Yet spelling out the proposition is necessary to specifying the relevant fact. The fact relevant to my search for Washington Square is that Washington Square lies straight ahead of David Velleman at noon on 1/1/2020. The fact that Washington Square once lay straight ahead of Edith Wharton is irrelevant. My guide’s directions will state the former, relevant fact, but as we have seen, they must do so without explicitly distinguishing it from the irrelevant fact about Edith Wharton.
Normative guidance works the same way. The fact that female circumcision is permissible among the Kikuyu has, among the Kikuyu, an action-guiding mode-of-presentation — namely, that the practice is permissible, or “permissible-for-us”. Specifying the frame of reference, by saying that the practice is permissible for the Kikuyu, transforms it from the Kikuyu’s norm into the ethnographer’s report, hence from normative to factual. When the Kikuyu say that the practice is permissible and the Mbuti say that it is not, their statements are diametrically opposed, but they are diametrically opposed presentations of different facts, like “Straight ahead” and “Straight back” spoken to pedestrians headed in opposite directions.
The normativity of reasons
I am going to assume that morality obligates its subjects by being rationally binding on them — more specifically, by generating complete and compelling reasons for them to act, or to hold practical attitudes such as desires or intentions. On this assumption, whether different communities can have different moralities will depend on whether they can have differently constituted reasons. Can the same set of facts constitute reasons for members of different communities to adopt different actions or attitudes?
This question must be formulated carefully. It must be about complete reasons: sets of facts that militate for or against an action or attitude without any additional assumptions. And it must ask whether community membership can modify the force of such reasons without entering into their content or into the descriptions of the actions or attitudes involved. As a relativist, that is, I cannot be content to treat a subject’s community membership as a circumstance that appears in the content of the reasons that apply to him. I cannot be content to say, for example, that needing shelter and being an Mbuti is a reason for building a lean-to, whereas needing shelter and being a Kikuyu is a reason for building a hut. Even in the context of Mbuti deliberations, being a homeless Kikuyu would qualify as a reason for building a hut, though of course that circumstance would never arise. Nor can I, as a relativist, be content to include the subject’s community membership in conditional act-descriptions, such as building a hut if one is Kikuyu or building a lean-to if one is Mbuti. Even in Kikuyu deliberations, needing shelter qualifies as a reason for building a lean-to if one is Mbuti, though a Kikuyu would satisfy this act-description by default, without having to build anything. I must rather treat community membership as determining what qualifies as a reason for what. I must say that one and the same set of facts gives the Mbuti a complete reason to build a lean-to and gives the Kikuyu a complete reason to build a hut, thus militating in different directions for reasoners from different communities.
A relativist about reasons cannot also be an anti-reductionist about them. Anti-reductionism about reasons is the view that there is no explaining or analyzing the relation between reasons, on the one hand, and the actions or attitudes that they are reasons for, on the other.8 According to anti-reductionism, we can say that reasons are considerations that count or weigh or militate in favor of such things, but we are then using the phrases ‘count …’ or ‘weigh …’ or ‘militate in favor’ in a sense that means no more than “be a reason for”.
The reason why a relativist cannot be anti-reductionist is that he needs to explain how one and the same set of facts can count or weigh or militate in favor of different things in different communities. Asked how reasons can do such a thing, an anti-reductionist would have to say that there is no explaining how: they just do. Not a satisfactory reply. If relativism is to be more than this bare and implausible assertion, it had better explain how the counting or weighing or militating relation can be modified by the subject’s community membership; and so it had better have something to say about the nature of that relation.
At the same time, the relativist had better not go so far as to say that different communities reason in accordance with different relations between reasons and what they are reasons for, as if communities use different methods of practical reasoning. Such methods of reasoning would be merely conventional — the deliberative mores of one’s community — and so the problem of explaining the normative force of mores would recur at the level of practical reasoning. Why would one be obligated to reason by local methods? To avoid this question, the relativist must characterize a single relation that reasons always bear to what they are reasons for. His relativism must then consist in the claim that one and the same relation is sensitive to differences among communities. The unavoidable question is: How?
In order to answer this question, I will draw on the image of reasons as having weight, though unlike the anti-reductionist, I will use that image as a tool for analyzing the relation between reasons and that which they favor. Comparing the force of reasons to weight can be fruitful because both phenomena give application to the notion of frames of reference. If the same set of facts can have different normative force in different communities, the explanation will be that different communities have different frames of reference — or, more colloquially, different perspectives or points of view. And of course there are different gravitational frames of reference, in which massive objects are pulled in different directions with different weight. So let me develop this analogy, beginning with some obvious facts about gravitation.
Rocks are heavy; that is, they tend to fall; that is, to accelerate downwards. But ‘downwards’ is an indexical, and so ‘Rocks are heavy’ is implicitly indexical as well. Whereas it expresses a true proposition when spoken by someone standing on Earth, it expresses a false proposition when spoken by an astronaut in outer space. The true proposition expressed on Earth by this sentence is that rocks tend to accelerate toward the Earth.
Of course, ‘Rocks tend to accelerate toward the Earth’ is true when spoken by anyone, including astronauts in space. But saying “Rocks tend to accelerate toward Earth” provides no practical guidance to anyone — for example, to someone who is wondering whether to let go of a rock. Saying “Rocks tend to accelerate toward Earth” gives him practical guidance only if he knows that he is standing on Earth, precisely so that he can derive an indexical mode-of-presentation, such as ‘Rocks are heavy’ or ‘Rocks tend to fall’. What would provide the most immediate practical guidance, however, is hefting the rock in his hand to feel its weight. Saying “Rocks tend to fall” is an action-guiding description of the force by which he would be guided directly in handling the rock.
In outer space, rocks are weightless, and there is no direction that counts as down. A rock can have weight only where there is a gravitational force that establishes some direction as down, which is the direction of falling, which is the direction in which the rock tends to accelerate. Which direction is down depends on the direction of gravitational force, which determines the direction in which rocks tend to fall.
What if reasons were like rocks? In that case, a consideration would have the weight of a reason only where there was a force that established a direction in which reasons militate. The force by which reasons militate in some direction is normativity. To the direction in which they militate, let us give the name ‘to be adopted’, since we can speak of both actions and attitudes as being adopted by a subject. Just as gravity determines what’s down, by causing material objects to accelerate in that direction, so normativity would determine what is to be adopted, by guiding subjects in the direction of adopting actions and attitudes. Like a rock, then, a reason would exert its weight within a frame of reference established by some weight-conferring force.
Corresponding to the statement that rocks are heavy on Earth would be a statement that some fact F weighs in favor of some action or attitude A from some perspective P. Like the statement that rocks are heavy on Earth, the statement that F weighs in favor of A from P would have no guiding force. By the same token, saying to someone who occupies P that F weighs in favor of A, like saying to someone on Earth that rocks are heavy, would offer guidance by describing a force exerted by the weighty item. Finally, considering weighty reason F while occupying P would be like hefting a heavy rock while standing on Earth: it would be the most immediate form of guidance.
Alternatively, saying that fact F favors action or attitude A from perspective P would be like saying that Washington Square is straight ahead from the perspective of heading south on Fifth Avenue: no practical guidance. Saying to someone in P that F favors A would be like saying to someone headed south on Fifth Avenue that Washington Square is straight ahead: an action-guiding description. Getting someone in P to consider F would be like pointing a tourist straight ahead down Fifth Avenue: direct practical guidance.
But what plays the role of gravity in the case of reasons? And what plays the role of Earth? Or, in other words, what is normativity and what generates it? According to the analogy, normativity must be a force that is present wherever considerations have the weight of reasons, as gravity is present wherever things have physical weight. Where present, the force must establish a “direction” to be adopted, as gravity establishes the direction down. And to be adopted must be a direction in which normativity guides subjects, as down is the direction in which things are weighed by gravity.
If we can identify such a force, we will see how the normativity of reasons might vary between communities without being simply a matter of deliberative mores. Reasons will turn out to have, not just a role in a local method of reasoning, but a constant nature, as considerations that have weight in virtue of coming under a particular force, which establishes a frame of reference. Their nature will nevertheless be compatible with their favoring different actions or attitudes, or the same ones to different degrees, in different frames. The remaining task for relativism will then be to explain how different communities generate different rational frames of reference, as Earth and Mars generate different gravitational frames.
Note the order of constitution suggested by this analogy. The force of gravity does not draw things in a direction that is antecedently constituted as down; rather, a direction is constituted as down by the force of gravity, which guides things toward massive objects such as Earth. If the analogy between normativity and gravity holds, then we should not expect normativity to draw us in the direction of what is antecedently constituted as to be adopted; rather, what is to be adopted will be constituted by the force of normativity, which draws us toward — well, toward whatever plays the role of Earth in the practical realm. What plays the role of Earth, thus generating normativity, remains to be seen.
Being Ordinary
As a moral relativist, I expect normative gravity to emanate from mores, which establish an agent’s social frame of reference, within which he finds reasons for acting and reacting. So what force do mores generate to guide actions and attitudes?
At this point, I must veer into the realm of speculative sociology, since the present question hangs on the nature of social mores. Before I take that turn, let me note that I have completed my outline of the form that a relativist metaethical theory should take. From here on, I will be proposing a specific content for one such metaethical theory. I find this theory compelling, but the reader may wish to regard it as no more than an illustration of how the above outline can be filled in.
A genealogy of mores
Human beings have a practical need and a psychological drive to live together with other people — not just in proximity to them but in personal interaction with them. Personal interactions require mutual interpretation: you cannot deal with others as persons without trying to understand their actions and attitudes, and to make yourself similarly understood. Your drive toward sociality therefore entails a drive toward mutual interpretation.
I speak of a drive toward sociality rather than a desire because this motive is inchoate and multiply specifiable. It can be described as a drive toward connection with other people, a drive to function as a person among other persons, indeed simply to be a person, insofar as sociality is essential to personhood or personhood is a social status. No matter how it is described, this drive requires you to engage in mutual interpretation. And your role in mutual interpretation requires not only interpreting but also being interpretable.
Whenever you interact with others, it’s as if you are on the computer’s end of a classic Turing Test, trying to gain and maintain recognition from the person on the other end of the line. In the Turing Test, the computer must avoid being relegated to the status of a machine; in real life, you must avoid being relegated to the status of mentally ill or deficient, or just too weird to bother with. No matter what in particular is at stake in a particular interaction, your eligibility for social interaction in general is also at stake: the interaction can always be broken off on the grounds that you are not a qualified interactant. In order for your qualifications to be acknowledged, you not only have to demonstrate an ability to interpret the other person; you also have to make yourself interpretable as a person.
Donald Davidson argued that in order to interpret other people, you have to narrow down the range of possible interpretations by assuming that they believe what is true and desire what is good by your lights.9 Davidson thought that this charitable assumption would be necessary in principle, no matter how much evidence or time or intelligence was available to you. For purposes of speculative sociology, however, it suffices to say that you have to make such an assumption in practice, because you must interpret people on the fly. You need, as it were, a library of sub-routines for real-time interpretation of other people, and they need sub-routines for real-time interpretation of you, if you and they are to interact. That’s why you and they need to exercise Davidsonian charity: the sub-routines most ready to hand are those drawn from the interpreter’s own beliefs and desires.
Because you need to be interpreted as well as to interpret, however, you need to exercise more than charity. Even as you extend charity to others by assuming that they believe and desire what you do, you must rise to their charity by satisfying their corresponding assumption, thus making yourself susceptible to their interpretation. They will try to understand you by assuming that you believe and desire as they do, and you must gratefully comply, so as to make yourself understood. They must do likewise, by gratefully satisfying your charitable assumption about them. The result is that you and they must converge on what to believe and desire. You needn’t converge perfectly, but eccentricities must form no more than a thin albeit salient layer atop a deep fund of shared attitudes. Even eccentric attitudes must come from a fairly limited set of alternatives.10
As the sociologist Harvey Sacks put it, people have to be ordinary — not completely ordinary, of course, but ordinary to a very large extent.11 Even if they want to be extraordinary, or out of the ordinary, there are more or less ordinary ways of doing so, beyond which they would strike others in their community as humanoid creatures of some unrecognizable kind. An idiosyncratic sense of humor still has to qualify as a sense of humor, and a disposition to laugh at manhole covers doesn’t qualify. A unique sartorial style still has to qualify as a style, not an inability to dress oneself. One can coin new slang expressions, invent new dances, but only within limits.
Before people can be ordinary, however, there has to be such a thing as ordinariness: there have to be ways that people ordinarily think, feel, and act. That’s where mores come in. People who need to interact with one another need to converge on ways of thinking, feeling, and acting that will suggest plausible first-pass interpretations of one another in their swiftly developing interactions. Their social mores are ways of thinking, feeling, and acting on which they converge.
Where to converge? In the case of what to think, the salient point of convergence is determined by the facts, but points of convergence are less constrained in the case of how to feel. Within broad constraints imposed by human nature (as I will discuss shortly), people have a fair amount of leeway in their responses, and they still need to converge on recognizable kinds of responses to recognizable kinds of things in recognizable kinds of circumstances.
For example, you need for people in your social vicinity to be able to tell whether you’re being serious, and just saying so won’t help unless they can tell that you’re being serious. So you need for there to be matters about which seriousness is the recognized default in your social vicinity — matters that are known by your interactants to be taken seriously by their interactants — so that the people with whom you interact can assume that you’re serious about those matters. And then you had better be serious about them. Conversely, you need for there to be matters about which the recognized default is joking or teasing. Now, people will assume that you’re serious about what you care about, and so it’s helpful if there are things that people in your social vicinity know that people in their vicinity generally care about, and if you too care about those things but not about the things that are generally known to be laughing matters.
You need for people to be able to tell whether you’re angry, and to tell without having to ask, just in case you are. So you need for there to be matters about which anger is the recognized default — matters that are generally known to be generally such as to make people angry. Conversely, you need for there to be matters about which the recognized default is non-acknowledgement.12 People will assume that you’re angry, for example, when you have been offended, and so it’s helpful if there are ways of being treated that are generally known to give offense, and if you too feel offended when treated in those ways but not in others.
I cannot emphasize enough that these social necessities allow for exceptions. You can afford to care about things that are generally known to be laughing matters or to overlook things that are generally known to give offense, but you cannot afford to do so in general. The most affordable exceptions are ones that do not require on-the-spot interpretation — hobbies, for example. Taking bottle-caps seriously is unproblematic, especially because you can pursue that hobby in the privacy of your home or in the company of other bottle-cap fanciers. By and large, however, the things you take seriously have to be matters that are generally taken seriously and generally known to be such.
Similarly, there are times and places where you can afford to be deceptive, secretive, or inscrutable by preventing others from interpreting you correctly. But again, those occasions have to be rare exceptions, lest you fail the social Turing Test and end up as a non-person.
The Normativity of Reasons
Now, there is a view — and it has to be the relativist’s view — that the only reasons to value something are features that it shares with other things that are valued, by oneself and by people in general. This was Mill’s view,13 and it is the view presupposed in our usual justification of attitudes. That is, we justify our attitudes by showing that they are ordinary, for ourselves and for those in our social vicinity.
Imagine (just imagine!) that we live in a community that admires people along lines of wealth. About a particular millionaire, we will say, “Now, that’s the sort of person we admire,” which is a way of saying that the person is admirable. Of course, one of us could point to rich people whom we don’t admire and poor people whom we do, thereby initiating a discussion of whether rich people really are admirable, given that we don’t ordinarily admire them. Alternatively, the dissenter may say, “That’s the sort of person others admire, but I don’t admire people like that.” We can then ask what sort of people he does admire, what makes them desirable to him as role models or mentors, what he finds remarkable or estimable in them, and so on. If he points to things that we don’t ordinarily desire in role models and mentors, or note and esteem in others, we can ask what he hopes to learn from such people or regrets lacking in comparison to them. And unless those questions, and the natural follow-up questions, eventually lead to attitudes that are somehow ordinary, his attempt at justification will fall flat.
Or perhaps one of us will say, “I think we all admire the wrong sort of person.” This dissenter can then be asked what makes such people unremarkable and uninteresting, undesirable as role models or mentors, and so on; which will lead to questions such as what we ordinarily desire in role models and mentors, what we ordinarily take note of or an interest in, and so on; which may raise the objection that we desire the wrong sort of role model, for example; which may then lead to the question what about them is undesirable. But unless those questions, and natural follow-up questions, eventually lead to attitudes that are somehow ordinary, the dissenter’s attempt at justification will fall flat.
Finally, a dissenter may say, “Hell, I just don’t admire her.” This dissenter can do without ordinariness, but at the cost of doing without justification.
Within our imaginary community, then, the fact that a person is rich constitutes a presumptive reason for admiring him, although the presumption in favor of its being a reason can be defeated in any of the ways that I have just surveyed, all of which appeal to presumptive reasons for desire or esteem or regret or some other attitude. Wealth is a presumptive reason for admiration in this community, I claim, because we have converged on admiring rich people so as to facilitate mutual interpretation with respect both to whom we admire and to how we regard the rich. The normative force of this reason is the force of the drive toward mutual interpretability, which arises out of the drive toward sociality.
One might think: The fact that some people are ordinarily admired is merely evidence that they are admirable; it isn’t what makes them admirable. I say: Then why do communities converge within themselves but diverge from one another with respect to whom they admire? Can we residents of Greenwich Village assert categorically that widely cited scholars are admirable but widely viewed televangelists are not? Or must we rather conclude that there is no reason to admire one person more than another, hence no reason to admire anyone? Relativism offers a more plausible account of the phenomena. According to relativism, we residents of Greenwich Village have reason to admire Hannah Arendt and Elizabeth Anscombe, namely, that they were widely cited scholars; and their being so is a reason to admire them because that’s the sort of people we admire.
One might think: The fact that we ordinarily admire scholars is no reason to admire them. I say: Right. Our admiring scholars isn’t a reason to admire them; it’s the frame of reference within which scholarship is a reason to admire them — a reason, specifically, for each of us. Compare: There cannot be directions to Washington Square except from some perspective, but the perspective doesn’t add to the directions. “Straight ahead” gives the directions; “Heading south on Fifth Avenue” gives the perspective. “Straight ahead for someone heading south on Fifth Avenue” doesn’t give more complete directions; it gives no directions at all. “Straight ahead” gives complete directions. So too, “She’s a widely cited scholar” gives a complete reason: information about our community membership would not add but would rather detract.
Compare again: Someone’s being ordinarily admired has no weight from our perspective, no more weight than the Earth has on Earth. What plays the role of Earth in our evaluative universe is personal interaction with co-members of our community, which is made possible by mutual interpretability, which is made possible by convergence on ordinary attitudes. The community’s evaluative frame of reference is established by the drive toward sociality plus the shared ways of thinking, feeling, and acting to which members of the community are thereby drawn. Other communities have their own evaluative frames of reference, established by the same force drawing them toward other ways of thinking, feeling, and acting, whichever are theirs. Hence reasons are relative to a community — specifically, to the community’s mores, or shared way of life. In one community, scholarship is admirable; in others, not.
One might think: The drive that constitutes the force of reasons should be the drive toward doing what ought to be done and feeling what ought to be felt, not a drive toward some arbitrary aim like mutual interpretability. I say: Mutual interpretability is not an arbitrary aim in relation to the force of reasons. Actions and reactions are interpreted in light of reasons for adopting them. Whatever force makes one responsive to reasons makes one responsive to the very considerations that figure in interpretation.
One might think: Okay, but considerations figure in interpretation because they are reasons, not vice versa. I say: Yes, vice versa; that’s the best account of the phenomena. On the one hand, we find genuine reasons for admiration; on the other, we find individually convergent and mutually divergent communities of admirers. The best explanation for these phenomena is that there is no such thing as what categorically ought to be admired; there are only reasons that acquire their weight from some perspective-establishing force, which cannot be the force of what categorically ought to be admired. That force is whatever force gives weight to reasons in general, everywhere. Our practices of justification, and their connection to interpretation, suggest that it is the drive to converge with our community on what to feel, which in turn is best explained by our drive toward mutual interpretability as a prerequisite of sociality.
One might think: This theory purports to reduce an ‘ought’ to an ‘is’, or a value to a fact, which we all know is impossible. The value is that scholars are admirable; the fact is that scholarship draws admiration from residents of Greenwich Village, who ordinarily admire scholars. The one cannot be reduced to the other. I say: Of course the value can’t be reduced to that fact; what it can be — and has to be, given the nature of normativity — is an indexical mode-of-presentation of the fact. That scholars are admirable (that is, to us) is a normative, hence indexical, expression of the non-normative fact that scholarship draws admiration from residents of Greenwich Village.
Reasons for acting
Thus far, I have spoken of reasons for attitudes, which I have treated as the basis of values such as admirability, desirability, and so on. Morality involves not only reasons for attitudes but reasons for actions. I can be brief in presenting relativism about reasons for actions, since much of the work has already been done.
Reasons for attitudes acquire their normative force, I have said, from the drive toward mutual interpretability for the sake of sociality. If reasons for acting are to exert normative force, they should acquire it from the same drive in a similar way. The question, then, is what would make for mutual interpretability of behavior.
Interpretation is holistic. That is, an interpreter tries to figure out all at once what a subject is feeling, believing, and doing, and he tries to figure it out by looking for the suite of attitude- and act-descriptions that best fits the subject’s present and past behavior overall. In order for the subject to make himself interpretable in real time, he must therefore behave in a way that clearly suggests some overall suite as the best fit. And what lies most within his control is of course his behavior, which he can fashion so as to fit his attitudes — or, preferably, those attitudes which he can most readily be interpreted as having.14 As a rule, then, the subject’s reasons for acting will consist in his attitudes, especially those attitudes which he and others ordinarily have.15
This rule has many exceptions, however, as is inevitable whenever holistic reasoning is at work. For example, co-members of a community may coordinate on behaving in a particular way under particular circumstances no matter what they think or feel. They will then be able to read the applicable act-descriptions directly off the circumstances, because they will know what “we” do in circumstances of that kind. If they want to interpret the behavior in terms of the attitudes behind it, they will interpret it, not as an expression of attitudes peculiar to the agent at the time, but as conformity to a social convention for the sake of sociality, and they will find departures from the convention intelligible only if they can understand why the agent would have strong motives for departing from it.
And then there will be cases in which a subject has strong motives for being uninterpretable to co-members of his community — that is, for lying or keeping secrets or simply being inscrutable. Despite his drive toward interpretability to co-members of the community, such motives will give the subject reason for being in some respect uninterpretable them, lest he become uninterpretable to the minimal community consisting of himself.16 The subject himself is, as it were, the core of his own normative Earth.
My account thus far has the following implications. The Kikuyu may have reasons for practicing female circumcision, and if they have such reasons, they have them because they live like Kikuyu. Westerners have reason to abominate the practice, and they have those reasons because they live like Westerners.
One might wonder: “Perspectives aside, whose reasons are the real reasons?” I say: That’s a nonsensical question, like asking “Perspectives aside, which are the correct directions to Washington Square?” or “Frames-of-reference aside, how heavy is this rock?” One might conclude: “Well, then, there are no real reasons, only reasons-from-a-perspective.” I say: That’s the wrong conclusion. From within a perspective, some facts really and truly have the normative force of reasons, just as from within a perspective, some utterances are really and truly the directions to Washington Square. Indeed — and this is the fundamental point — there is no other kind of normative force for reasons to have. Perspective-independent reasons are impossible, just like perspective-independent directions, because reasons and directions are action-guiding, and guidance is necessarily relative to a perspective.
Of course, the Kikuyu may actually have reasons within their own perspective to abolish the practice of female circumcision. The practice may be in violation of other Kikuyu mores, which generate reasons to abolish it. In that case, we can say, as a matter of anthropological fact, that the Kikuyu have reason to abolish the practice, and we can say to them, “You have reason to abolish that practice”, while directing their attention to the considerations that might guide them to abolish it. Even without knowing whether the Kikuyu way of life generates such reasons, we may feel optimistic that it must, and we can engage them in conversation with the hope of discovering that it does. Relativism doesn’t counsel despair over the possibility of moral coordination among communities.
What relativism does counsel, however, is humility. We cannot assume that the Kikuyu have reason to change their ways. We have to allow for the possibility that at the end of the conversation, common ground will still be out of reach.
By now, the reader may wonder whether I am really a relativist. And indeed there are several respects in which my view departs from the extreme and simplistic relativism that is usually associated with the name.
Universality vs. ubiquity
One reason for doubting whether I am a relativist is that I seem to have allowed my evaluative universe to fall under universal purposes and principles — first the aim of mutual interpretability and then the various rules that subserve it, such as Davidsonian charity and (as I call it) generosity. How can a relativist allow such norms to govern universally?
In fact, I haven’t granted universality to any norms. As Sharon Street has pointed out, a norm needn’t hold universally in order to hold within every perspective, since it can hold independently within each one.17 I will express this point by saying that a norm can be ubiquitous but not universal. Ubiquitous norms govern only locally, but they govern locally everywhere, within every perspective. An example is the taboo against incest. Some people point to the incest taboo as a “human universal”, but they don’t usually mean “universal” in the metaethical sense, since they don’t think that there is a perspective-independent moral requirement that would be violated by a community that permitted incest. In the metaethical sense, then, they regard the taboo as ubiquitous but not universal.
A moral relativist had better deny the existence of any universal norms, moral or otherwise; for if he concedes the existence of universal norms, he will be hard-pressed to explain why moral norms are not among them.18 But a moral relativist must go further. Although there being no universal norms would entail that moral norms are at most ubiquitous, they might still be ubiquitous in a way that the moral relativist must also deny; for even if moral norms were merely ubiquitous, they might be necessarily so. The principles of charity and generosity, for example, are necessarily ubiquitous norms, in the sense that they are operative independently in every normative frame of reference. The fact that these principles are locally operative everywhere is no accident: each normative frame of reference must be established by the drive of its occupants toward sociality, which requires mutual interpretability, which calls for charity on their part as interpreters and generosity on their part as targets of interpretation.19 Where this force is absent, there are no reasons for acting or reacting, and no actions or attitudes are to be adopted: everything is normatively weightless. As a believer in the necessary ubiquity of these principles, I am not a relativist about Davidsonian charity or generosity. I am a moral relativist, however, because I deny that any moral norms have the same status as those principles. In sum, I deny that there are universal norms of any kind, and that there are necessarily ubiquitous norms of morality.
Plural moralities
Another reason for doubting my credentials as a relativist is my assumption that the mores of actual communities will not diverge so far from ours as to be utterly a-moral by our lights. I assume that the mores of actual communities always have enough in common with our morality to be recognizable to us as versions — often misguided versions, even horrifically or appallingly misguided versions, but still recognizably versions — of what we call morality.
Why so? If there is no single true morality but many moralities, then why aren’t there communities with no morality at all, because their mores are utterly a-moral?
The beginning of an answer is that members of a community cannot achieve mutual interpretability by converging on just any attitudes and actions. The eligible points of convergence are constrained by human nature. There are some attitudes on which we humans cannot help but converge. They include an aversion to pain, separation, and frustration; an inclination toward pleasure, connection, and the fluid exercise of skill; the inborn and automatic fight-or-flight response; an interest in the human face and form; an initial dislike of snakes, spiders, blood, and the dark; plus an array of physiological appetites. Human nature also gives every attitude a distinctive role in causing behavior. Admiration (to stick with my example) naturally disposes one to emulate the admired person, to defer to him, and to approve of his words and actions. These behaviors may be more or less readily interpretable in light of one’s other attitudes. One may have beliefs that harmonize or clash with the person’s opinions; ideals that he may or may not exemplify; interests that he may or may not share; likes and dislikes of other people whom he may resemble. Now consider a category of people who tend to have opinions we ordinarily reject, interests we ordinarily disdain, resemblances to people we ordinarily hate. Converging on admiration for such people will not serve the purpose of mutual interpretability. We will rather tend to converge on admiring people admiration for whom makes us more readily interpretable, because admiration for them harmonizes with other attitudes on which we converge. And we will tend to converge on types of action that are readily interpretable in light of such attitudes
There is reason to think that the resulting constellation of attitudes and actions will tend to be pro-social rather than anti-social, in the sense that they will favor mutual benefit over mutual harm. The reason is that our convergence must result from spontaneous, unmanaged coordination, which favors mutually beneficial arrangements.
Hume illustrates this point with the example of two people rowing a boat together.20 If these people were riding a tandem bicycle instead, then each would be tempted to ease off the pedals and let the other do most of the work. (Maybe that’s why tandem bicycles are used for leisurely sightseeing but not for travelling from point A to point B.) But if two people are travelling by rowboat and each is pulling one of the oars, then trying to shift the workload will be self-defeating, since the boat will go around in circles.21 In order to go anywhere, the rowers need to produce equal work: they need to coordinate on a level of effort. Fortunately, the rowers can coordinate spontaneously, without exchanging a word, provided that there is a uniquely salient level on which to converge, as there will be if some point of convergence is obviously preferred by both. Each rower prefers a level high enough to get them where they are going but not so high as to wear them out; if some such level is obvious, then they will spontaneously coordinate on it. The need to coordinate thus produces mutually beneficial joint effort.
Or consider two people who are trying to move a sofa by picking it up at either end. If they were hoisting the sofa with a block and tackle, each would be tempted to slack off and let the other do the pulling, but straight lifting requires them to lift their ends equally fast and equally high. The need to coordinate will lead them to converge on that degree of force, if there is one, that is uniquely salient as the one on which to converge; the most salient degree will be the one, if there is one, that is obviously preferred by both; and the obviously preferable degree will the one, if there is one, that will get the job done without straining anyone’s back. If there is such a degree of force, then the movers will converge on it, and the result, again, will be mutually beneficial joint effort.22
For the same reason, what becomes ordinary in a community — the constellation of feelings and actions on which its members converge — is likely to favor mutual benefit over harm. Different communities, already made alike by human nature, will also be shaped alike by the need for coordination, which favors their pro-social over their anti-social human tendencies. The variance among social mores will therefore resemble the variance among variations on a theme, where the theme is recognizably moral.
So although I believe that there is no necessarily ubiquitous morality, I also believe that having a recognizably moral way of life is indeed necessarily ubiquitous. The difference between our community and others is not that we have a morality and they have none; the difference is that their ways of life and ours embody common moral themes in incompatible ways. And the fact that all ways of life embody those themes is no accident. Shared ways of life arise from the need for mutual interpretability, which requires co-ordination, which favors mutually beneficial arrangements; and so ways of life, by their very nature, tend to be recognizably moral, however horrifically or appallingly so.
Here is yet a third respect in which my version of relativism may seem un-relativistic. According to my version, the fact that reasons are always relative to a perspective does not entail that perspectives themselves are on a par. Even if people do have reason for practicing female circumcision, I say, the possibility remains that those reasons depend on perspectives that are backwards, and not just from a particular perspective.
My fellow relativists will be shocked by the suggestion that one community can be less advanced than another, and not just from someone’s perspective. Nothing could be further from the spirit of relativism. But as I said at the outset, a relativist has to characterize a single relation that reasons bear to actions or practical attitudes, lest he end up with deliberative mores whose normativity needs to be explained. The guiding force mediated by that relation will be a single normative force, the same force in every perspective, perspective-dependent only as to its direction. Such a force will unavoidably provide a necessarily ubiquitous parameter in relation to which ways of life can be more or less advanced.
I say that the necessarily ubiquitous parameter is mutual interpretability, which is a prerequisite for social life. The standard of comparison for practical perspectives is thus the degree to which they facilitate mutual interpretability. How well have members of a community managed to converge on reasons for acting and reacting? How well do those reasons help them to understand themselves as the kind of creatures they are, endowed with a somewhat fixed nature as human beings? How well, in other words, have the members of a community managed to develop a shared way of human life?
The idea is that there is something that ways of life characteristically do.23 Members of a community, any community, develop a way of life for the sake of its doing that thing. Some ways of life do it better than others. Those ways of life are more advanced with respect to an aim shared by all communities in developing their ways of life. Those ways of life are more advanced, in other words, with respect to a necessarily ubiquitous social aim.
At this point, my fellow relativists may want to banish me from their midst. My so-called relativism, they will say, is no relativism at all, because it allows for evaluative distinctions among ways of life. What is left of relativism in my view?
For one thing, the evaluative distinctions that remain are not moral. Communities do not qualify as more or less advanced by falling closer or further from some universal or ubiquitous morality. There is no universal or even ubiquitous morality, and there are no universal norms of any kind. What there are, however, are ubiquitous norms of interpretation and interpretability, which are the fundamental prerequisites of sociality, and it is in relation to these norms that communities can be more or less advanced. They can be more or less advanced, in other words, in terms of the prerequisites of sociality.
Secondly, my view says that reasons for actions and attitudes are relative to the way of life that actually prevails in an agent’s actual community. We Westerners are therefore in no position to say that a Kikuyu mother has reason not to circumcise her daughter — unless, that is, we can locate such reasons within the Kikuyu way of life. Even if our Western way of life is more advanced, it cannot provide reasons to the members of communities who follow different ways.
Thirdly, even if the Kikuyu community as a whole can have reasons for revising its way of life, those reasons will be relative to the way of life it already has, and there are no grounds for assuming that they will lead it to converge with other communities. If a whole community is to have reasons to change, those reasons must consist in circumstances in light of which social change would be interpretable, at least to members of that community, and what’s interpretable by way of change in a community depends on what the community is already like. Reason-guided change is path-dependent: where it ends up depends on where it began. So different communities may have reason to change in ways that still lead to different ways of life, even if those ways of life are equally advanced by necessarily ubiquitous standards.24
There may some day be world-wide convergence, if there is a world-wide community — the proverbial global village — but even then, relativism would hold. If as a result of advances in transportation and communication, everyone has to be prepared to interact with just about anyone, then a global way of life may develop, and cultural diversity will vanish. But which way of life became global would still be path-dependent, and what people had reason for feeling and doing would still be relative to the way of life in which mankind happened to end up, given where it began.
Moral Debate
Finally, a warning to philosophers. We cannot eyeball various communities and see how well their ways of life facilitate mutual intelligibility. Differences in success between ways of life are usually too subtle to discern from an academic perspective, least of all from the philosopher’s study. We just have to inhabit a particular way of life and do the daily work of interpreting, being interpretable, and helping to develop a common ground that facilitates mutual interpretation. Progress comes from a collective experiment in living, and there is no substitute for participating in the experiment.
So there is no point in appealing to an explicit standard of progress when engaging in philosophical debate or in face-to-face disagreement with members of other communities. The rational way to disagree with those who live differently is to articulate our own self-understanding, listen as they articulate theirs, and then go back to our respective experiments to see whether we have learned something by which to understand ourselves better by living differently. We can thereby make progress of a sort that cannot be detected or directed from without.
The reason for talking with those who live differently is that we and they share at least some common ground, since all of us are trying to figure out how to make better sense of and to ourselves as human beings. We even have reason to think that conversation will lead to progress. Indeed, we have reason to think that it will lead to progress that is recognizably moral, because our need for mutual intelligibility has its source in our sociality.
Footnotes
1Yes, there may be a standard for human beings, tall for a human, which applies to all of us. But that standard is still relative to a reference class, namely, human beings. What’s tall for a human is not tall for a giraffe. What’s tall for a giraffe is not tall for a tree. The Milky Way is said to be 2,000 light years tall.
2I will use the word ‘community’ to emphasize that I am speaking of people who regularly interact, usually because they live together. I will use the word ‘social’ as the corresponding adjective. I will speak as if communities are well defined and as if every individual belongs to one and only one community. Both of these assumptions are false but helpful as idealizations.
3The proper term for this practice is a matter of controversy. I chose ‘female circumcision’ because it is widely used and somewhat value-neutral, though it is far from ideal. (Its evaluative force may depend on whether male circumcision becomes widely viewed as immoral.)
4Some claim that there is a version of moral relativism according which the Kikuyu and Mbuti are disagreeing about a single proposition but faultlessly so, since both are right. I think that faultless disagreement is impossible, and so I ignore this version of relativism.
5This objection is equivalent to one that is raised by Paul Boghossian. See his “What is Relativism?”, in Truth and Realism, ed. Patrick Greenough and Michael P. Lynch (Oxford: Oxford University Press, 2006), 13–37, and “Three Kinds of Relativism”, in A Companion to Relativism, ed. Steven D. Hales (Oxford: Wiley-Blackwell, 2011), 53–69.
6John Perry, “The Problem of the Essential Indexical,” Noûs 13, no. 1 (1979): 3–21; reprinted in The Problem of the Essential Indexical and Other Essays, Expanded Edition (Stanford, CA: CSLI Publications, 2000). See also Perry’s “Self-Notions”, Logos (1990): 17–31, and “Myself and I”, Philosophie in Synthetischer Absicht, ed. Marcelo Stamm (Stuttgart: Klett-Cotta, 1998), 83–103, also reprinted in The Problem of the Essential Indexical. James Dreier based a version of speaker relativism on Perry’s “essential indexical” in “Internalism and Speaker Relativism”, Ethics 101, no. 1 (1990): 6–26. His goal is to explain how statements applying normative terms such as ‘good’ can express the speaker’s motives, so that speakers who agree on the facts can disagree about values. My goal is to explain how the facts in virtue of which reasons are action-guiding can fail to be action-guiding.
7All of the above applies, by the way, to the predicate ‘tall’ when it is used to guide action. If you ask whether someone is tall because you want to know whether to put on flats or heels, it won’t help to be told that the person is tall in comparison to someone who is five-foot-eight — unless, of course, you know that you are five-foot-eight, so that you can derive the action-guiding conclusion that the person is tall.
8See, e.g., T.M. Scanlon, What We Owe to Each Other (Cambridge, MA: Harvard University Press, 1999), Chapter 1: “Reasons”.
9See, e.g., ‘Radical Interpretation’, Dialectica 27 (1973): 314–328, reprinted in Inquiries into Truth and Interpretation (Oxford: Clarendon Press, 2001).
10When it comes to real life and real-time interpretation, the relevant interpreters are the people with whom you need to interact and by whom you therefore need to be recognized as an interactant. These people are your community, as I am using the term.
11“On Doing ‘Being Ordinary’”, in Structures of Social Action: Studies in Conversation Analysis, ed. J. Maxwell Atkinson and John Heritage (Cambridge: Cambridge University Press, 1984), 413–439.
12See Thomas Nagel, “Concealment and Exposure”, Philosophy & Public Affairs 27, no. 1 (1998): 3–30.
13In Chapter IV of Utilitarianism, in On Liberty and Utilitarianism (New York: Bantam Books, 1993). Often dismissed as a fallacious argument, it is rather a substantive claim.
14These may not be attitudes that the subject actually has. I discuss this issue in “From Self-Psychology to Moral Philosophy”, Philosophical Perspectives 14: Action and Freedom (2000): 349–377; reprinted in Self to Self: Selected Essays (New York: Cambridge University Press, 2005), 224–252.
15Some will object that reasons for acting are not attitudes but their contents or the satisfiers thereof. They will say that reasons for adding sage to a stew, for example, include the desirability of improving the taste and the fact that sage will improve it, not the desire to improve the taste or the belief that sage will do so. I am not sure whether Donald Davidson, the author of this example, disagrees. He thinks that the desire and belief are the reasons for which the agent adds sage to the stew, but he also thinks that ‘It is desirable to improve the taste’ and ‘It will improve the taste’ are the premises in the agent’s practical reasoning, and he doesn’t say whether the premises in practical reasoning are the relevant reasons for acting, as opposed to the reasons for which the agent acts. (“Intending”, in Essays on Actions and Events [Oxford: Clarendon Press, 1980], 86.) I believe that reasons for doing something have to be such as could become the reasons for which someone does it, hence that there can be no distinction. (See Bernard Williams, “Internal and External Reasons”, in Moral Luck [Cambridge: Cambridge University Press, 1982], 101–113.) I also believe that reasons in both cases are attitudes, not their contents or satisfiers, but I cannot defend that view here. See my Practical Reflection (Princeton, NJ: Princeton University Press, 1989) and How We Get Along (New York: Cambridge University Press, 2009).
16This claim and the associated theory of agency are defended in my Practical Reflection (Princeton, NJ: Princeton University Press, 1989), some essays in The Possibility of Practical Reason (Ann Arbor, MI: MPublishing, 2009), some essays in Self to Self (New York: Cambridge University Press, 2005), and How We Get Along (New York: Cambridge University Press, 2009).
17See Street’s “Objectivity and Truth: You’d Better Rethink It” (forthcoming) and “How to Be a Relativist About Normativity” (in progress).
18A similar point is made by Paul Boghossian in “Three Kinds of Relativism” and by Street in “How to Be a Relativist About Normativity”. Note that I am speaking here of practical norms; epistemic norms are a different matter.
19Norms can be necessarily ubiquitous for other reasons. See, again, Street’s “How to Be a Relativist About Normativity”.
20A Treatise of Human Nature, ed. L.A. Selby-Bigge, with text revised and notes by P.H. Nidditch (Oxford: Clarendon Press, 1976), III.ii.2, 490.
21David Lewis assumes that the rowers must synchronize their strokes (Convention: A Philosophical Study [Cambridge, MA: Harvard University Press, 1969], 44 et passim). In fact, the rowers need not synchronize their strokes, so long as their oars do equal amounts of work.
22Another example, further afield: What is the rationale for social sanctions against nonconformists? The rationale is not that nonconformity itself is harmful to society, nor that most people just happen to behave pro-socially, so that the sanctions happen to fall on the anti-social. The rationale is that the sanctions themselves militate in favor of pro-social behavior, by enforcing coordination. Given common knowledge that everyone will be penalized for being an exception, people will tend to converge on what they prefer to be the rule, and they prefer a rule of pro-social behavior to the alternatives. Thus, the point of sanctioning nonconformists is not that the nonconformists harm the group; the point is that the sanctions themselves benefit the group, by introducing an incentive for coordination, which favors mutually beneficial behavior.
23There is much in common between this functionalist view and the “pluralistic relativism” of David Wong in Natural Moralities: A Defense of Pluralistic Relativism (Oxford: Oxford University Press, 2006), esp. Chapter 2.
24Revolutionary change is another matter. I discuss this issue in “Motivation by Ideal”, Philosophical Explorations 5, no. 2 (2002): 89–103; reprinted in Self to Self.