As of late, I've been doing a lot of reading on sentimentalist accounts of ethics. In particular, Adam Smith's Theory of Moral Sentiments has kept me up at nights. Here, I'm beginning to believe, is the beginnings of the right theory of morality. It is just the beginning, or the root, of the right moral theory.
There are many concerns. Not the least of which is this: Smith's account is intended to be primarily descriptive, and not normative. The theory tells us what Smith thought we do do when we make ethical judgments, and what, in fact, we want when we use ethics. It does not tell us what we ought to do.
I'm guessing that, in the not-too-distant future, I will post about the possibility of making Smith's account normative, and will post some significant differences between the account that I am pursuing for my dissertation, and both classical and neo-sentimentalist accounts. I will also describe the significant overlap, and why my dissertation can best be described as a sentimentalist account.
The title of my dissertation is "Me, Myself & Mine." I will give it a subtitle, which will make plain that my primary interest is not personal identity (or what makes me me), and neither is it political philosophy (as the "Mine" part might imply), but patterns of concern or caring. I want to explore what truly matters when it comes to what we care about, what we have good reason to care about, and, possibly, what we ought to care about. What I'm interested in is the patterns of concern that we have for the person that is me (Me), the self that is me (Myself), and the objects and things in the world that relate to me in particular ways (Mine).
Clearly, a definition or account of "caring" and "concern" are central to my dissertation. I'm beginning to believe that the best description of "caring" or "concern" is primarily, or most significantly, a sentimentalist account. In what follows, I want to begin exploring what it means to "care" about someone else.
A pure sentimentalist account of caring would be this. For P to care about someone (Q) or something (x) just is for P to feel a certain set of in principle specifiable sentiments when P judges that Q or x is either in a good or bad state. When P judges that Q, for instance, is worse off, P experiences negative affects, like grief, sadness, despair, and so on. Likewise, when P judges Q to be better off, P feels associated positive affects, like joy, happiness, and so on.
If we believe that feeling positive affects is part of what it means to be better off, and feeling negative affects is part of what it means to be worse off, then for P to judge that Q is better off is for P to be better off. This is so in the case of friends, lovers, our children, and our family in general (for the most part). When we judge that our children are better off, we feel positive affects and so are ourselves better off.
This means that improvements in Q's life are improvements in P's life, when P cares about Q. This is part of what it means to care about someone else. I suggest that a similar account can be given in the case of objects, but I won't offer it here.
Alternative accounts of caring could lean on desires or judgments. A desire account of caring would look like this. For P to care about Q is for P to desire that Q be in a good state, for Q's life to go better, and so on. A pure account would insist that this is all there is to caring--having these particular desires.
Desires and sentiments are, at least in principle, distinguishable. Consider this. Many times, friends will feel envy and jealousy at an improvement in the other's state of affairs. When Jones gets a good job, Smith, Jones' friend, may feel this way. On a desire account of caring, this would not undermine the status of Smith and Jones as friends, and we could still insist that Smith cares about Jones. We can insist on this quite apart from how Smith feels, so long as we can be certain that Smith does desire things to go better for Jones, and that Jones' getting this good job is consistent with those desires.
I strongly suspect that there can be a third, pure account of caring using judgment, but I have not yet thought of it. I'm working on it (help a brother out... can you think of an alternative account of caring that is principally about judgments and not desires/sentiments?)
If we want to defend the pure sentimentalist account of caring against the desire account, we can respond to the case of jealousy like this. We can say that Smith is not really a friend to Jones if he feels jealousy at her success. Without the appropriate sentiments, we cannot be said to truly care about someone. If Smith feels jealous, that's good reason to think that Smith does not really care about Jones, even if she thinks she does.
I do not think that this is a good strategy to take. It seems fairly clear to me that we can care about someone and yet still feel sentiments that would imply otherwise. In addition, this would be assuming the definition of caring, precisely the thing that is at issue.
An alternative strategy is to insist that jealousy and envy are not sentiments that constitute the subset of sentiments we use to judge caring. Since I do not yet have the first clue what sentiments are rightly affiliated with caring, we can use placeholders. Let the lower-case letters a through z stand for all possible sentiments. Let a through m be "positive" affect sentiments (joy, happy, glad, elation, and so on), and n through z be "negative" affect sentiments (sad, miserable, angry, jealous, and so on). (Yes, of course there are sentiments that cannot be neatly placed under "positive" or "negative" affect. Let's pretend these do not exist for our purposes. Or, if this is too much to ask, just add a third category of "neutral" or whatever affects).
Okay: For P to care about Q is for P to feel either a, c, d, or f, or all of them, or some of them, when P judges that Q is in a good state; and to feel n, o, r, and s, or all of them, or some of them, when P judges that Q is in a bad state. P feeling b, for instance, is not implicated in caring. That is, b is not relevant to determining whether P cares about Q. We might now claim that jealousy is not a caring-relevant sentiment. P feeling jealous tells us nothing about whether or not P cares about Q.
Surely, given the wide panoply of possible human sentiments, not all of them will be relevant for determining caring. This is why, in principle, we can make these distinctions between sentiments that are, and sentiments that are not, caring-relevant.
But suppose that what Smith feels is jealousy and nothing else. Her feelings are not mixed with at least one of the positive affects on our sentimental account that would imply caring. She does not feel jealous and a, c, d, or f. She merely feels jealous.
That having mixed feelings is at least a possibility seems right to me. I suspect we have complicated emotional states, where we can feel both glad and jealous at once, or alternately, or whatever. It is not silly to say "I'm jealous, but I'm happy for you," and to actually feel both emotions. I need to look into the psychological literature to make certain of this, but I don't think it is controversial.
This account may sidestep a number of possible objections. But probably not this one. It is very improbable to suggest that jealousy is not one of the sentiments associated with caring. I take it that this is fairly plain, so I won't go into detail.
A third possibility is to say that, right then, P does not care about Q, even if P cares about Q in general. For me to care about something is not necessarily for me to always and consistently care about it. It is to have, say, a certain disposition towards the object of my care most of the time and in most circumstances. This would imply the following qualifier to our sentimentalist account of caring: For P to care about Q is for P to feel some subset of positive (a, c, d, or f) or negative (n, o, r, or s) affect sentiments when P judges that Q is in a correspondingly good or bad state, most of the time, and in most circumstances
Isn't it obvious that when people say "I wasn't a friend to you then" that they mean something like "I did not care about you in a way that is consistent with friendship then"? It might not be obvious. This is because, when we say something like "I wasn't a friend to you then," we might mean "I did not act
in a way consistent with friendship," rather than, "I did not feel
in a way that is consistent with friendship." While this is clearly so, we can make the following claim by way of rejoinder. Actions are not right or wrong all on their own. What makes an action right or wrong is the intention, or the motive behind it. While playing soccer (football for my European buddies), I may accidentally kick you while trying to play the ball (I'm not that good at this sport, and so have plenty of occasions to apologize for stepping on toes, kicking shins, and otherwise causing harm). Clearly, hoofing people in the shins is not a right thing to do. But the explanation "I'm sorry, I didn't mean to kick you, I was going for the ball," is typically accepted as a good nullifier of moral wrongness. We might say that what we mean
(our motives) to do trumps what we do do (our actions).
(It is more complicated in the case of negligence, but this complication does not undermine our account. Negligence is being insufficiently "good" at acquiring the right information given the circumstances, when we believe that being sufficiently good at this information-gathering is a requirement of all of us. But then again, I suspect that what is really doing the moral work is the claim that we lacked the motivation to acquire sufficient information. And it is this lack of a motive that we think is a requirement for most of us in those circumstances, is what is being addressed, and not the lack of acquiring the information in the first place).
We might be inclined to distinguish motives and sentiments, and claim that the given response fails to address the worry. If we come to believe that motives and sentiments are necessarily linked, we may believe that to say "I meant to do so-and-so" just is a claim about how we feel, since some combination of feelings will be the right explanation for the motives. That sentiments and motives (or motivation for action) are so linked is argued below.
How we count the times and circumstances in the proffered qualifier will be difficult to determine. How many cases do we need? And in what proportion of those cases must we have the corresponding sentimental states? I hope this difficulty does not lead us to abandon the project, even if, at the moment, it appears a terribly difficult problem.
We might insist that there are some circumstances which are significant enough such that, were we to feel a sentiment inconsistent with caring at those times, it would undermine the claim that we really care about the thing in question. If we accept this, we would add to "most of the time, and in most circumstances" the further qualifier "and always when this or that," where "this or that" stands in for a full description of the circumstances or cases we judge to be significant enough.
One reason not to insist is this: Suppose P is on anti-depressants, or some other drug, that alters moods and sentiments. Or suppose that P has certain portions of his brain cut out, such that it is not possible for him to feel certain sentiments. Now suppose Q is in one of these most significant of circumstances, and P does not feel the appropriate corresponding sentiments. Does this mean that P does not care about Q? Perhaps not. We can chalk it up to the drugs, or the missing brain parts, and not to P's failure to care. P may still care, but merely be incapable of feeling the right sentiments.
This counter-example is not reason to abandon the "significant circumstances" qualifier. It is, instead, a reason to claim that, then, P really did not care about Q. Being incapable of feeling certain care-relevant sentiments at a time, according to the sentimentalist account of caring, is to be incapable of caring at that time. (Interestingly, the sentimental account of caring would suggest that, if P had portions of his brain cut out such that he could not feel care-relevant sentiments, we could no longer claim that P cares about anything
This view may be false. We may believe that the truth may be that we still care, even if we do not feel the corresponding care-relevant sentiments now. For instance, we may claim that we always
care about our children, say, even if, at some times, and for some reason, we cannot experience the care-relevant sentiments. (It should of course be clear that I do not mean that we care about things only so long as we continuously feel certain care-relevant sentiments, and do not when we are not feeling those. Instead, we care when we feel these sentiments when we are appropriately confronted with the object of our caring.)
In fact, the above account of caring which includes a "most of the time and under most circumstances" qualifier accepts just this possibility. It claims that, to care about someone, is to have care-relevant corresponding sentiments not always, but in at least a sufficient amount of cases. This is to say that our judgment that we care about so-and-so or such-and-such can override temporary or occasional failures to feel certain sentiments.
This is very much like the following. We do not determine whether or not P is intelligent by looking at this
test result. We average, aggregate, or use better tests to determine this. We look at P's history. P, just now, may be on anti-depressants or other drugs, and so perform poorly on this test, even if it is not true to say that P is not intelligent. It may be true to say that P is not intelligent now (because he is on drugs), but our global evaluation of P's intelligence will not hinge on this one case. So it appears plausible to say "P is intelligent, but not now (because he is on drugs)."
To be able to say this requires us to capture the truth about P's "normal" state, and to be able to distinguish this from "abnormal" states. To determine P's "actual" intelligence is to know when P is in a "normal" state, and to judge based on his performance when he is in this state. One (defeasible) heuristic is aggregating or averaging. About a month or so ago, I had three shots on goal in soccer (football), and scored all three times. One goal was extraordinary. I (falsely) believed that I was improving, and was beginning to feel somewhat confident. Subsequently, I haven't done much of anything, even though I've only had about one or two chances. Having chances, however, is part of what it takes to be good at soccer. You need to have a good idea of where to go, and what to do, and get in a good position at the right time. I am useless at this. When I am honest with myself, I find that I am extraordinarily talentless when it comes to soccer (although I often try to focus on the good, and convince myself that I am pretty good, even though this is false). When I am being honest with myself about soccer, I always try to think to myself of the fact that what really matters is wallyball, volleyball, whiffleball, and squash (all of which I am very good at), and not soccer, which is, as we all know, for hooligans.
This heuristic is defeasible because it is difficult to believe that we are ever in a "normal" state. It is, I suspect, always an approximation. Even when it is an approximation, judging this will also require an assumption that we know what is "normal" in order to know that this case is an approximation. We can be thoroughly skeptical about this, but I suspect that we do not have good enough reason to be quite so skeptical.
Let us leave the "significant circumstances" qualifier discussion to one side. I do not yet know what to say about it, except that I'm inclined to disagree with having it.
One reason to prefer sentimentalist to desire accounts of caring is because of the apparently more obvious link between sentiments and action. To feel a certain way is to be motivated to do something in particular ways. To feel "disgusted" is to be motivated to make certain facial expressions, and to engage in avoidance behaviour. This is so when we feel fear and joy too, and so on. Our emotional states are "reasons" for action in the sense of always giving us a certain amount of "action-juice" (and not in the sense of "reasons for action" that practical reason theorists use).
Suppose it takes one litre of action-juice to result in actual action. The claim here is that sentiments provide at least some amount of action-juice always
, even if it is not enough action-juice to lead all the way to actual action. This contrasts with desires like this. To have a desire is not necessarily to be motivated to do anything at all. Only the combination of desire and sentiment motivates.
I have claimed, above, that desires and sentiments are, at least in principle, separable. Someone may complain that desires and sentiments are, in practice, deeply linked. Certain desires and certain corresponding sentiments are co-extensive. It is not possible to have certain desires without at the same time feeling certain sentiments. This co-extensive correspondence view strikes me as a touch implausible, even though I do believe that, in very, very many cases desires and sentiments come in packages. What is implausible is that this is so in every
case, and not merely in the vast majority of cases, or in very many cases, and so on.
I suspect that this is a crazily controversial claim that I have just made. To convince anyone of this view would require plenty of evidence and argument. One very good reason for accepting this account is this. I recollect reading some psychological research about the relation between emotional states and actions. That people who failed to have certain emotional states failed also to behave in certain ways (I think I read something about people who failed to feel fear, and that this was disastrous for them because they also failed to behave in certain ways). Was there not also some research on people who lacked very many emotional states and, while they desired certain things, they never did anything about it? I need to find this research to get anywhere with this. So consider this another plea for help.
Why would the link between motivation and sentiment benefit a sentimentalist account of caring? One reason to think this is because of the common sense notion of caring. A common sense account seems to include something like "to care is to be motivated to do something for the object of the caring." By "do something" is meant doing good things for things we care about. If we were not so motivated, it would be hard to claim that we really care about the thing in question.
I want to suggest that we ought to stick with the common sense account of caring as much as possible. Our ordinary meaning of caring includes being able to say things like, "if you cared about me, you would do so-and-so." A more philosophically technical account would read this as "if you cared about me, you would be motivated to do so-and-so." If desire and sentiment are in principle separable, claims like the previous would only make sense using a sentimentalist conception of caring (provided the claim that desires, on their own, do not motivate action). Even if, in practice, desires and sentiments come in packages, we might wonder which, the desire or the sentiment, is doing the motivational work. The claim is that sentiments are motivational, and desires all on their own are not.
This distinction, and the claim that desires do not motivate, may be nothing more than philosophical fictions. Why wouldn't desires motivate? Many economists, and many of us in general, speak of doing so-and-so because we desired to. We often treat desires as primitive explanations of behaviour, primitive in the sense of requiring no further explanation. On this account, we would be forced to go deeper. To claim that we did it not because "we wanted
to," but because "we felt
like it." That the two are often used interchangably, as though they mean the same thing, is reason not to accept hard-and-fast distinctions.
Perhaps we shouldn't. To insist that desires do not motivate is to present a radical revision to our ordinary, as well as technical, understandings of the role of desires in action. It seems clear to most of us that explanations of action in terms of desires are not merely ways of speaking, but the truth about why we do certain things. "Feeling" like doing something is a good explanation (if not yet a justification), while "wanting" to do something is also a good explanation (if not yet a justification) of something we have done. Until I take a longer look at the psychological literature, we can go on the assumption that both desires and sentiments are able to motivate action.
A pure sentimentalist account of caring may be false for other reasons. If we can avoid controversial (and revisionist) accounts, we should do so. Consider that, very many of us, and non-philosophers for sure, do not believe that certain animals, that are capable of having analogous sentiments to ours, care about anything. If all that mattered were feeling certain things in response to other things, we could gauge whether the animals do feel a particular way, and conclude, contrary to our beliefs, that they care about this or that.
Even the sentimentalist account, however, is not bereft of cognitive requirements that at least many animals are incapable of. The sentimentalist account requires that we be able to judge that so-and-so or such-and-such is either in a good or bad state. Making this judgment is a cognitive requirement possibly too steep for at least very many animals (I doubt very, very much that dolphins and apes and monkeys are incapable of caring, for instance. But I doubt even more that chickens are capable of caring, even if they may have some limited range of sentiments).
Even if this is not a good reason to enrich our pure sentimentalist account, there are other reasons. The best one, in my mind, is that our ordinary notion of "caring" is deeply rich. It appears to include not merely certain judgments and corresponding sentiments, but also desires. It is terribly natural to say that caring means, amongst other things, wanting (desiring) that things go better for the object of our concern. Including desires in our complete account of caring seems right given the sort of thing "caring" is.
My thought is that "caring" or "concern" incorporates the sentiments in a descriptive sense. When you care about Q, you have the corresponding care-relevant sentimental responses to judgments concerning Q's being in a better or worse state. This is not the whole of the story. Desires are included in the following way. We may desire that things go better for Q, as part of the meaning of caring about Q. We may also desire to have certain sentiments about Q, given that we care (or want to care) about Q. If we feel envy at Q's successes, we may desire that this not be so, that, instead, we feel pride or gladness, or something like that.
My hope is that normativity can enter the account in the following sort of way. We may insist that it is right, fitting, appropriate, or obligatory for P to care about Q when Q stands in particular relations to P. For instance, Q may be the daughter of P, and, in general and with exceptions, we probably think it obligatory for P to care about Q. By this we mean more than merely the claim that P should behave in all the ways that imply care or concern, but that P, instead, actually care, and actually have concern. This may be partly due to our belief that sentiments add motivation (if we do not believe that sentiments are the sole motivators of action), and we ought to be motivated to take certain actions with respect to certain Qs.
I am busy sorting this out. It will take a while. Boy, this is a long post... perhaps I should stop here.