The What and Why of Loves Reasons 13OCTOBER2015

[Pages:18]THE 'WHAT' AND 'WHY' OF LOVE'S REASONS

Michael Smith

1. The question

We can be said to love many things. These include people (people in general, but also people with whom we have a special relationship like parents, siblings, children, friends, romantic partners, spouses...), non-human animals (dogs, cats, horses...) objects (rings, shoes, paintings...), activities (eating, dancing, playing guitar, watching movies...), ideas (truth, justice, wisdom...), and so on. Though to say that we love some of these things amounts to little more than the claim that we like them, in many cases something more, and in other cases something quite different, is being claimed. The question is what that more or different thing amounts to, and what reasons for action we have in virtue of loving someone or something in that more or different way.

This question will be more or less urgent depending on what you think about the nature of reasons for action as such. In what follows I begin by describing the view of reasons for action I find most plausible, and I explain why that view makes the question very urgent. It turns out that there are two main ways in which loving someone could provide reasons for action. But one of these makes it difficult to see how we could love anything other than rational beings, and how our love, even among rational beings, could be selective in the way it is, giving us reasons to do things for those we love that we don't have to do for strangers--think of the reasons we have to do things for our friends and romantic partners, for example. And the other, though it makes it easy to see how our love, even among rational beings, could be selective, and how it could extend to things besides rational beings, makes it difficult to see how our loving someone or something could amount to anything other than merely liking them or it.

With the problem squarely in view, I go on to describe an account of the nature of love which, when combined with the theory of reasons for action I described at the beginning, makes it clear why we can love such a vast array of objects, why our love can be selective, and why loving something differs in an important way from merely liking it. I then explain what reasons for action we have in virtue of loving something in this way. The account I provide of love falls short of a set of necessary and sufficient conditions. Instead it is an account of what it is to love something in what I think of as the paradigm cases of loving something. In order to move from the account provided to an account of necessary and sufficient conditions we would need to add a condition that says something like '...and anything sufficiently similar to the paradigm cases is also a case of loving something'. Unsurprisingly, on my view love therefore admits of many borderline cases. This is as it should be, given that there are so many borderline cases of love.

To anticipate, since part of the task is to come up with an account of what it is to love someone or something, and since what you think such an account should look like will depend on many factors, including your views about the nature of reasons for action, the account I go on to provide will inevitably be controversial, presupposing as it does my own view about the nature of reasons for action. So as to be as up-front as possible about these controversies, in the course of the discussion I will make various comparisons with the views of others. As will become clear, the view of love I go on to provide owes a great deal to the work of Dean Cocking and Jeanette Kennett, and also to the work of Alexander Nehamas. However my own view differs in some important ways from theirs, largely because of the view of reasons for action that lies in the background, so I will make these differences clear as the discussion proceeds as well.

2

2. What makes it the case that we have reasons to act in certain ways in certain circumstances?

To have reasons for acting at all, in certain circumstances, we must have some array of options in those circumstances, options that have different outcomes associated with them. My completely unoriginal suggestion, as regards the nature of reasons for action, is that we have a reason to pursue one of these options, rather than another, just in case the outcome associated with our taking that option is better than that the outcome associated with our taking the other, the reason being provided by the feature that makes the option better. This account of what it is to have reasons for action is what Joseph Raz calls the 'classical account' (Raz 1999). The classical account is, however, just a schema that needs filling out with independent accounts of what makes the production of an outcome an option for an agent, and what makes one outcome more desirable than another.

Let's begin with what makes something an option. Though I do not have a fully worked out theory of what options are, for present purposes we can suppose that having an option to bring about an outcome requires two things: first, that an agent could form an overriding desire to produce that outcome, and second, that if she had such an overriding desire, and if she believed that the time had come to produce it, she would produce it. An agent's options are thus limited by her conceptual capacities, on the one hand, and by her physical abilities, on the other. Agents do not have the option of bringing about outcomes that they either cannot conceptualize or are unable to bring about, and they will therefore no reason to perform such actions either. Think of some examples. Wild animals can have no reason not to cause the harms that they cause for the first reason: they cannot conceptualize the outcomes of their actions as harms. And when we are in New York, we can have no reason to get from to Melbourne in nano-seconds for the second reason: we have no ability to travel faster than light.

Next consider what makes an outcome associated with one of an agent's options more desirable than another. My view is that such facts are fixed in the way suggested by a dispositional theory of value (Smith 1994). According to the dispositional theory, the desirability of the outcome of an agent's acting in a certain way in certain circumstances is a function of the desires her ideal counterpart has concerning that outcome in those circumstances. The more desirable outcome is the one an agent's ideal counterpart desires more. Of course, agents who lack the conceptual resources to conceive of the outcomes of their actions in certain ways--think of a wild animal, or an infant, or someone who is brain-damaged--may well have ideal counterparts who do have those conceptual resources. This will depend on what it is for a counterpart to be ideal, about which I will have more to say presently. But, assuming that this is so, though the outcomes of the different actions that these agents could perform will still be more or less desirable, depending on their ideal counterparts' desires, these agents won't have reasons to produce those desirable outcomes, as they won't have the option of doing so in the technical sense we've defined.

Putting our account of what makes one option more desirable than another in terms of the dispositional theory together with our account of what it is for an agent to have the option of producing an outcome, it turns out that agents who have the conceptual resources to conceive of outcomes in the ways in which their ideal counterparts do, and who also have physical abilities to produce such outcomes, have the option to produce them. These agents therefore have reasons to produce such outcomes. To repeat, this is because what agents have reasons to do is determined by the overriding desires for outcomes that they have the capacity to form and

3

realize, on the one hand, and by which desires their ideal counterparts have concerning those outcomes, on the other.

One feature of the dispositional theory worth emphasizing is that it makes desirability evaluator-relative (Sen 1982). Suppose a certain outcome o would come about if an agent a were to in circumstances C, where -ing is one of a's options in C. In that case, o in C is intrinsically desirablea just in case a's ideal counterpart intrinsically desires that o obtains in C, and, if a's ideal counterpart has several different intrinsic desires, then how intrinsically desirablea o is in C is fixed by the relative strengths of the intrinsic desire a's ideal counterpart has concerning o in C and the intrinsic desires that a's ideal counterpart has concerning other outcomes.

Because a's ideal counterpart's desires fix the desirabilitya of the outcomes of a's actions, the intrinsic desirability-making features of the outcomes of a's actions may themselves be characterizable only in terms of their relation to a. The upshot is that intrinsic desirability claims of both the following forms may be true:

(x)(It is intrinsically desirablex that Fx)

and

(x)(y)(It is intrinsically desirablex that (Fy))

The difference between these is that the first requires that each agent's ideal counterpart intrinsically desires that he has feature F, whereas the second requires that each agent's ideal counterpart intrinsically desires that everyone has feature F. Though both entail that it is desirablea that Fa, the desirabilitya of a's having F is extrinsic, according to the second, but not according to the first. According to the first, it is it is desirablea that a has F because a is who she is. A's having F has relative value, not neutral value. According to the second, it is desirablea that a has F merely because a is someone. A's having F has neutral value, not relative value.

This will prove to be important in what follows, as it will help explain why the reasons for action we have in virtue of loving someone can be selective, notwithstanding the fact that reasons are fixed in the way suggested by the classical account of reasons, that is, by the desirability of the outcomes associated with the outcomes of the actions we have reasons to perform. Reasons can be selective because the most desirable outcomes of an agent's actions are always most-desirable-relative-to-that-agent, and this means that an outcome featuring someone or something to which that agent stands in a certain relation--the x-loves-y relation, whatever that turns out to be--can be more desirable relative to that agent than those in which someone figures to whom she does not stand in that relation. In these cases the outcome will have more relative value, even though it may not have more neutral value.

To give the obvious example, given the focus of this paper, if for x to love y were to turn out to be merely a matter of x's liking y, and if our liking someone leads us to prefer outcomes in which they fare well rather than badly, over those outcomes in which others, even their doppelgangers, fare well rather than badly, and if these preferences survive the process of idealization, then the dispositional theory would tell us that the outcomes featuring those we love faring well would, to that extent, be more desirableus. Outcomes featuring those we love faring well would have more relative value than those in which they fare badly, even though though they may not have more neutral value. In this way, the dispositional theory would entail that we

4

have reasons to make it the case that those we love fare well, rather than badly, reasons that we don't have concerning those we don't love.

Let's now focus on what it is for an agent's counterpart to be ideal. What we are after is an account of what it is for an agent to have reasons for action, so what it is for her counterpart to be ideal must be fixed by the kind of thing she has to be insofar as questions about her reasons for action arise in the first place. Since she has to be an agent, and since the function of an agent is to know the world in which she lives and realize her desires in it, something she might do well or badly, we can restate our account what is desirable in the following terms. The desirabilitysome particular agent of the outcomes of that agent's actions is fixed by the desires she has in in the nearest possible world in which she is an ideal agent: that is, the possible world in which she functions optimally as a knower and desire-realizer. Everything therefore turns on what it is for an agent, a being whose function is to know the world in which she lives and realize her desires in it, to function optimally, and the answer is that such a being would have to meet a modal condition. She would have to have and exercise maximal capacities to know what the world is like, no matter what it is like, and to realize her desires in that world, no matter what she happens to desire.

However, since the exercise of each of these capacities would seem to be in tension with the exercise of the other--just imagine the possible world in which someone desires that she believes certain things, whether or not they are true--it follows that an ideal agent would have to have the wherewithal to reconcile this tension in the exercise of her two capacities, if such a reconciliation is possible. Is it possible? For reasons I have outlined elsewhere, it seems to me that it is (Smith 2011, Smith 2012, Smith 2013). An ideal agent is able to reconcile this tension because she has a pair of dominant intrinsic desires that bear on her possession and exercise of these two capacities in the present, and consistency demands that she extends these desires to herself in the future and to other agents. To function optimally as knowers and desire-realizers, every agent's ideal counterpart must therefore have a dominant intrinsic desire that, no matter what circumstances they find themselves in, they do not interfere with their own or anyone else's current or future exercise of their capacities to know the world in which they live or realize their desires in it (on condition, of course, that the realization of those desires wouldn't itself constitute such an interference), and they must also have a dominant intrinsic desire that they do what they can to see to it that they themselves, and others as well, acquire and maintain this pair of capacities. In a phrase, every agent's ideal counterpart has dominant intrinsic desires that, in whatever circumstances she finds herself, she helps and does not interfere, and beyond this, that she does whatever her non-ideal self wants to do. There is therefore a limited convergence in the intrinsic desires of ideal agents. They converge on dominant intrinsic desires to help and not interfere.

Putting the pieces together, the picture that emerges is this. No matter what circumstances an agent finds herself in, certain intrinsic desirability claims of the first of the two forms described above--(x)(It is intrinsically desirablex that Fx)--must be true. No matter what circumstances an agent finds herself in, it is most intrinsically desirableher that she helps and does not interfere in those circumstances, and, if there is nothing she can do to help, or if there are various ways in which she can help that she must choose between, then the next most intrinsically desirableher thing she can do is whatever she next most intrinsically desires to do (on condition, of course, that her satisfying these further intrinsic desires wouldn't require her to interfere). Given the classical account of reasons for action it follows that, so long as agents

5

have the capacity to form and realize dominant intrinsic desires to help and not interfere, no matter what circumstances they find themselves in, they all have most reason to help and not interfere in those circumstances, and, on condition they have done that, they have next most reason to do whatever they want to do.

Kant tells us that having respect for persons is a matter of believing that those who can set ends for themselves are never to be treated merely as means, but also always as ends (Kant 1786). Treating such beings as ends is in turn, inter alia, a matter of believing that there are reasons to leave them free to lead lives of their own choosing--this looks a lot like believing that there are beings we have reasons not to interfere with, one of the reasons for action delivered up by the quite general account of reasons just outlined--and it is also, inter alia, a matter of believing that those who can set ends for themselves are owed the basic wherewithal to live lives of their own choosing--this looks a lot like believing that there are beings we have reasons to help, the second of the two reasons for action delivered up by our quite general account of reasons. Notwithstanding its classical roots, the account of reasons for action we have just outline thus seems to be in basic agreement with Kant: we have respect for beings when we believe that they are among the class of beings who we have dominant reasons to help and not interfere with, and we show such beings no disrespect when, so long as we act on those dominant reasons, we do whatever we want to do.

The equation isn't perfect, however, and it is worthwhile emphasizing why this is so. Kant's own idea of having respect for persons is an attitude that can only be taken towards beings with a rational nature that includes not just possession of minimal desiderative and epistemic capacities that qualify someone as an agent, albeit a very non-ideal one, but also the much more sophisticated capacity for reflective self-control: that is, the capacity to have reasons and respond to them. In our terms, such a being wouldn't just be one that we have dominant reasons to help and not interfere with, but would also have dominant reasons to help and not interfere with others herself--recall the earlier discussion of the conceptual sophistication required to be a being with such reasons. Wild animals, infants, and the severely disabled, though agents in our sense, lack this sophisticated capacity and so cannot be objects of respect of the kind Kant has mind.

Our account of respect, by contrast, suggests a rather different picture. All beings with desiderative and epistemic capacities are agents, and simply in virtue of being agents, they merit respect: we have reasons to help and not interfere with all of them, including those who lack the specific capacity for reflective self-control. Possession of the capacity for reflective self-control turns out to be significant, on our account, not because it is a condition of someone's being an object of respect, but rather because, on the one hand, it is a condition of someone's being able to mete out respect, and hence someone of whom we can have such expectations, and, on the other, because whether or not beings have the capacity for reflective self-control turns out to affect what it would be to help and not interfere with them. Possession of that capacity is thus of great significance on our account of respect, but it does not have the significance it has for Kant.

3. Love and respect

Where do the reasons we have in virtue of loving someone or something fit into this picture? One obvious place is among the conditional reasons we have to do whatever we want to do. As already noted, if loving someone or something were simply a matter of liking them or it, then given that liking certain people or things gives rise to preferences concerning outcomes that

6

involve them, we would in turn have conditional reasons to satisfy those preferences. This would explain both why we can love so many different kinds of things, and why loving someone or something is selective in the way that it is. I will have more to say about this idea presently. For now, however, I want to focus on the other possibility, the possibility that loves fits in via its connection with respect.

We saw at the very beginning that one of the things we can be said to love is people in general. The idea that we should love people in general has a long and venerable history, a history that goes back at least to the Greek conception of agape, and, via the Greeks, to the Christian view that we should love everyone because Jesus loves us, and because we owe it to him to follow his example. But it is also an idea that surfaced in a more secular mode in the 1960s among those who rejected the prevailing norms in Western societies, norms that seemed to them to legitimize the domination of women by men, blacks by whites, homosexuals by heterosexuals, the indigenous by colonizers, socialists by capitalists, and so on. The Beatles performance of "All You Need is Love" on BBC's Our World in 1967, a performance that was watched by over 400 million people in twenty-five countries as part of the world's first global television link, was perhaps the most salient expression of that more secular idea of love as an attitude we should have towards people in general.

Love in this more secular sense would seem to amount to no more or less than the idea that rational beings, no matter what their differences, merit respect. We love others as we love ourselves, on this view, to the extent that we take the condition of the rational permissibility of the satisfaction of our own desires to be that that doesn't come at the cost of deceiving or manipulating others, or failing to do what's required to enable them to have the wherewithal to lead their lives in the way that they wish. But while it seems very plausible to suppose that this is one of the forms that love can take, it seems quite wrong to suppose that it exhausts the forms that love can take. Perhaps we can and should love people in general, but if we do, then there must also be a more selective sense in which we can and are permitted to love more selectively in the way we love our family and friends. The aspiration to find a single account of love that captures all of the paradigm cases would in that case be dashed. There would have to be at least two distinct accounts of love.

David Velleman in effect suggests that this isn't so because love in the more selective sense is a response to the very same thing to which respect is a response.

The Kantian view is that respect is a mode of valuation that the very capacity for valuation must pay to instances of itself. My view is that love is a mode of valuation that this capacity may also pay to instances of itself. I regard respect and love as the required minimum and optional maximum responses to one and the same value. Respect for others is required, in Kant's view, because the capacity for valuation cannot take seriously the values that it attributes to things unless it first takes itself seriously; and it cannot first take itself seriously if it treats instances of itself as nothing more than means to things that it already values. That's why the capacity for valuation, when facing instances of itself, must respond in the manner constitutive of respect, by restraining its self-interested tendency to treat them as means. In my view, love for others is possible when we find in them a capacity for valuation like ours, which can be constrained by respect for ours, and which therefore makes our emotional defenses against them feel unnecessary. That's why our capacity for valuation, when facing instances of itself, feels able to respond in the manner constitutive of love, by suspending our emotional defenses. Love, like respect, is

7

the heart's response to the realization that it is not alone. (Velleman 1999, p.366).

This idea needs some unpacking.

To begin, it is important to remember that respect already rationalizes a suspension of our emotional defences. If we respect others in the way we should, then when they have been wronged, we will not be indifferent to how they have been treated. We will be indignant on their behalf, and may even join with them in a righteous demand for justice. We show no-one disrespect in feeling these emotions, and would in fact show disrespect if we failed to express solidarity in some form with those who have been wronged. These emotions are thus permissible. The upshot is that even love of the non-selective kind, the kind that is no different from respect, has at least this much by way of permissible emotional vulnerability already packed into it. So if Velleman is right that love of the more selective kind is a matter of permissible emotional vulnerability to those we respect, then there must be more to that permissible vulnerability than love as respect already requires. Here is a suggestion as to what that more might be.

Think of the difference between how you feel when a stranger is wronged, and how you feel when a loved one is wronged. Though you may be equally indignant on behalf of the stranger and the loved one, you suffer along with the one you love in a way that you don't suffer along with a stranger. Our sense of our own life's going well or badly is bound up with how the life of those we love is going, not so with how strangers' lives are going. As Barry Manilow tells his beloved in the chorus of I Can't Smile Without You

I feel sad when you're sad I feel glad when you're glad

My hypothesis, accordingly, is that love of the more selective kind Velleman has in mind is a matter of our lowering our emotional defenses towards others with a capacity for valuation like ours in the quite specific sense of allowing our weal and woe to be bound up with theirs. We show no one disrespect in not lowering our emotional defenses towards them in this way, or almost no one--I will discuss exceptions in the final section--so it is permissible for us to allow ourselves to become emotionally vulnerable in this way to some but not to others. This, I take it, is why Velleman thinks that love is the optional maximum response to the same value to which respect is the required minimum response.

How plausible is this suggestion? One question to ask is whether emotional vulnerability of the kind described is necessary for love in the more selective sense. Another is whether it is sufficient. The idea that it is sufficient is very plausible indeed, so plausible as to be almost platitudinous. Not only does it have resonance in popular songs and the like, but it is an idea that is endorsed by a wide range of philosophers who write about love. For example, Harry Frankfurt tells us that,

...in the very nature of the case, a lover identifies himself with what he loves. In virtue of this identification, protecting the interests of his beloved is necessarily among the lover's own interests. The interests of his beloved are not actually other than his at all. They are his interests too. Far from being austerely detached from the fortunes of what he loves, he is personally affected by them. The fact that he cares about his beloved as he does means that his life is enhanced when its interests prevail and that he is harmed when those interests are defeated. The lover is invested in his beloved: he profits by its successes, and

8

its failures cause him to suffer. (Frankfurt 2004, p.61)

But even if Velleman is right that emotional vulnerability of the kind identified is sufficient for love, that doesn't imply that love simply consists in emotional vulnerability of that kind. For all that the sufficiency claim tells us, something deeper might explain that kind of emotional vulnerability, and love might consist in the thing that provides that deeper explanation.

This is in fact Frankfurt's view. As he sees things, our weal and woe is bound up with the weal and woe of those we love because we care about them, and loving someone should therefore be identified with caring about them. Loving thus differs from liking not because the objects of love and the objects of respect are one and the same, but rather because of the identification involved in caring as distinct from liking, an identification that Frankfurt spells out in terms of a second-order volition that constrains the will (Frankfurt 1971, 1987, 1992). We will have more to say about this idea presently.

Now consider whether allowing our weal and woe to be bound up with the weal and woe of those with a capacity for valuation like ours is necessary for love. Velleman thinks that it is necessary because possession of this capacity by those we love is required for "our emotional defenses against them...[to]...feel unnecessary." It is, however, quite obviously false that we never allow our weal and woe to be bound up with those who lack a capacity for valuation like ours. Many of us allow our weal and woe to be bound up with that of infants, the severely disabled, those with advanced dementia or Alzheimer's disease, and non-human animals, none of whom have a capacity for valuation like ours. There are risks involved in our doing so, of course, but we regularly act as if the gains make the risk worthwhile. Doing so thus isn't impossible.

Nor does it seem to be irrational to allow our weal and woe to be bound up with those who lack a capacity for evaluation like ours. Remember again what we have a reason to do, given our quite general account of reasons for action. We have a reason to do whatever we want to do, so long as our doing so doesn't prevent us from helping and not interfering. Having an emotional attachment to an infant, someone who is severely disabled, someone with advanced dementia or Alzheimer's disease, and a non-human animal is something that many of us want. Since our having such an attachment often won't prevent us from helping and not interfering, it is therefore often the case that there is a conditional reason for many of us to have such emotional attachments. Velleman's restriction of permissible emotional vulnerability of the kind he thinks constitutes love to those with a capacity for valuation like ours, those we must respect, would thus seem to be unmotivated.

Indeed, if love consists of allowing our weal and woe to be bound up with the weal and woe of others, then it seems that love needn't be an attitude that we have towards people at all. Remember again all of the things we can be said to love. To take just one example, some of us love paintings. Our sense of our life's going well or badly is bound up with the preservation of paintings, with opportunities to look at them, with affectively responding to them, with discussing them with others and making them more widely understood, and so on. In an admittedly somewhat loose sense, how things are going for us is therefore bound up with how things are going for the paintings, and we are glad that this is so. Given that we have a conditional reason to do whatever we want to do, it therefore seems that many of us therefore have a conditional reason to have such an attachment to paintings as well.

We began this section with the observation that love in one sense seems to be no different from respect. But since love in this sense is non-selective, an attitude we can and should have

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download