Love and Identification with the Beloved

Abstract

1. Introduction

This is one of two essays in which I return to arguments I ran in the ‘self-preservation’ section of “Love: Self-Propagation, Self-Preservation, or Ekstasis?” (Whiting 2013, henceforth “Love,” in Whiting [2016]). Each tackles an obvious question “Love” failed to ask.¹

“Love” focused on two questions raised by Harry Frankfurt’s (2004) claim that the lover’s identification with her beloved is one of four ‘conceptually necessary’ features of any variety of love. First, what work does Frankfurt take the alleged identification to do? And second, can identification, whatever exactly it consists in, bear the weight Frankfurt places on it? Or is the weight in fact borne in any given case by whatever accounts in that case for the alleged identification?²

Here I focus on a prior question: why take identification to be conceptually necessary in the first place? This question should have been obvious given the case sketched at the end of “Love” for taking the ideal form of love to involve at least some degree of ekstasis, conceived as a process in which the lover is taken outside of (ek) her normal condition (or stasis) into the world of another. For it would seem, at least prima facie, that the more a lover identifies with her beloved, the less she travels outside of her normal condition into his world. Much of course depends on how identification is construed. But before I turn to that, let me introduce two philosophers whose accounts of love seem not to assume any such conceptual necessity.

The first, whose ideal plays a leading role in “Love,” is Aristotle. The second, who must have been lurking offstage, is Iris Murdoch. For “Love” tied the ekstatic ideal to the very passage in Plato’s Phaedrus that Murdoch (1970) cites in making what she calls ‘unselfing’ central to her ideal. Surely my countless readings of Murdoch had set me, if only subconsciously, on her path.

Murdoch’s work casts the present question in sharp relief. For, in making ‘unselfing’ central to her ideal, she seems not simply to deny the conceptual necessity of identification but to question the bona fides of any love that rests on it. Aristotle’s position is less clear. Irwin (1988, 311) reads into his talk of the friend as an ‘other self’ the idea that the activities of one’s friend are an ‘extension’ of one’s own. But the list of endoxa around which his discussion is organized includes no mention of identification. And Aristotle goes on immediately to single out mothers’ concern for their children as paradigms of the first (and manifestly most important) item: namely, active concern for another for the other’s sake.³

Murdoch too cites mothers—the “inarticulate, unselfish mothers of large families”—as paradigms of the goodness so rarely found in this world. So I propose to introduce the story of a mother who—like the father invoked by Frankfurt—takes the life of a beloved daughter: the story of Sethe in Toni Morrison’s Beloved (1987).

My question is: how plausible is it to view Sethe’s act as involving some form of identification with her daughter, either the form invoked in Frankfurt’s account of Agamemnon’s sacrifice of Iphigenia or some other form?

2. Frankfurt and Agamemnon

Let me begin with the first three of Frankfurt’s four features, also called ‘conceptually indispensable conditions’.⁴

[A] First, [love] consists most basically in a disinterested concern for the well-being or flourishing of the one who is loved. It is not driven by any ulterior purpose but seeks the good of the beloved as something that is desired for its own sake.

Second, love is unlike other modes of disinterested concern for people—such as charity—in that it is ineluctably personal . . . . The [beloved] is loved for himself or for herself as such, and not as an instance of a type.

Third, the lover identifies with the beloved: that is, he takes the interests of his beloved as his own. Consequently, he benefits or suffers depending on whether those interests are or are not adequately served.

Frankfurt sees a “particularly snug fit” between self-love and his four conditions (2004, 80). He argues separately for each that self-love satisfies it to a higher degree than does any other ‘mode’ of love and concludes that self-love is the ‘purest’ of all ‘modes’.⁵

Consider what he says about the first and the third conditions, on whose relationship I focus here.

[B] The unalloyed purity of self-love is almost never spoiled by the intrusion of any extrinsic or ulterior purpose. Only rarely do we seek our own well-being primarily because we expect it will lead to some other good. In the love we devote to ourselves, the flourishing of the beloved is sought—to a greater degree than in other sorts of love—not only for its own sake but for its own sake alone . . . . Indeed, self-love is nearly always entirely disinterested, in the clear and literal sense of being motivated by no interests other than those of the beloved. (Frankfurt’s italics, my bold, 2004 3.7)⁶

[C] It will surely be conceded without too much argument that when a person loves himself, the identification of the lover with the beloved is distinctively robust and uncurtailed. For someone who loves himself, needless to say, his own interests and those of his beloved are identical. (2004, 3.7)⁷

Frankfurt goes on in 3.8 to assimilate parental love to self-love. Each, he says in 2.3, is grounded in the pressures of natural selection. Here he takes the “extraordinary degree to which [parents identify] naturally and more or less irresistibly with [their offspring]” to render parental love close in degree of purity to self-love (2004, 3.8).

[D] In self-love, there can be no discrepancy between the interests of the self-lover and those of the person to whom his self-love is devoted. The characteristic identification of the parent with child is generally rather more limited and less certain. Nevertheless it is as a rule distinctively extensive and compelling. After all, the child originates literally within the bodies of its parents; and parents normally continue long after a child’s birth to experience it as being still, in some less organic way, a part of them. (2004, 3.8)

This is one of two passages that led me to pair Frankfurt’s account of love with the account of erôs that Irwin sees in Plato’s Symposium, an account grounded in the lover’s desire for ‘self-propagation’.⁸ The other is [E], where Frankfurt seeks to ground obedience to the ‘commands of love’ in the lover’s desire for ‘self-preservation’ and claims in support of this that Agamemnon himself was destroyed by his sacrifice of Iphigenia. Each passage struck me as reflecting something like the foundational egoism embodied in Irwin’s account of Platonic erôs. Hence the structure of “Love,” which associates Irwin’s ‘self-propagation’ view with Frankfurt’s ‘self-preservation’ view and casts the ‘ekstatic’ view as preferable to each.

But I now think the ‘foundational egoism’ diagnosis of Frankfurt insufficiently precise. Irwin’s account of Platonic erôs is shaped by his conception of ancient eudaimonism as a form of ‘rational egoism’.⁹ Frankfurt’s account of love is shaped instead by his preoccupation with autonomy. He introduces identification to account for the way in which the ‘commands of love’ move the lover without threatening her autonomy. On his account, obedience to the commands of love is obedience to commands expressive of the will with which the lover identifies her very ‘self’.

[E] There is, I believe, a quite primitive human need to establish and maintain volitional unity. Any threat to this unity—that is, any threat to the cohesion of the self—tends to alarm a person and to mobilize him for an attempt at “self-preservation”. It seems to me that the authority that love has for us is closely related to this primitive and irreducible need to protect the unity of the self. Since the commands of love derive from the essential nature of a person’s will, a person who voluntarily disobeys those commands is thereby acting voluntarily against the requirements of his own will. He is opposing ends and interests that are essential to his nature as a person.¹⁰ In other words, he is betraying himself. We are naturally averse to inflicting upon ourselves such drastic psychic injuries. (Frankfurt 1994, 139, Frankfurt's italics; my bold)

The footnote moves beyond ‘drastic psychic injury’ to annihilation of ‘the self’ in question.

[F] Agamemnon at Aulis is destroyed by the conflict between two equally defining elements of his nature: his love for his daughter and his love for the army he commands. His ideals for himself include both being a devoted father and being devoted to the welfare of his men. When he is forced to sacrifice one of these, he is thereby forced to betray himself. Rarely, if ever do tragedies of this sort have sequels. Since the volitional nature of the tragic hero has been irreparably ruptured, there is a sense in which the person he had been no longer exists. Hence, there can be no continuation of his story. (Frankfurt’s italics; my bold)

Here, as elsewhere, Frankfurt speaks of self-preservation in what he calls the “unfamiliarly literal sense . . . of sustaining not the life of the organism but the persistence and vitality of the self” (2004, 2.8). Please note the ontological tone. Note also the way in which Frankfurt’s appeals to the pressures of natural selection suggest that he is at least sometimes concerned with self-preservation in the familiar sense, the sense in play when people sacrifice their ideals to save their skins, sometimes even ideals by which they had previously taken their ‘selves’ to be defined.

Of course, reading the appeal to identification as driven primarily by Frankfurt’s concern with the agent’s autonomy does not by itself remove the whiff of foundational egoism. It is after all the lover’s autonomy with which Frankfurt is primarily concerned. And there is still [D], where he seems to be concerned with self-propagation in the familiar sense.

[D] troubled me because I saw in it signs of a ‘colonizing ego’ account of love, one in which the beloved is—or is at least viewed by the lover as—an ‘extension’ or ‘part’ of lover (see Whiting 1991; in Whiting [2016]). I was also troubled by Frankfurt’s gender-neutral talk of ‘the parents’ experiencing the child as first organically and then in some less organic way a part of themselves. For even if gestating mothers experience the fetus as in some organic way a part of themselves, and even if some descendant of this experience continues for some time in those who go on to nurse their infants, it seems problematic for a theorist of love to take this sort of relationship as the proper paradigm for love of others.

The crucial question here is how self-love is to be understood. For the propriety of taking self-love—or the approximation of it Frankfurt sees in parental love—as one’s paradigm for love of others depends on that. So let me turn to Aristotle, whose view is at least superficially similar to Frankfurt’s.

3. Frankfurt and Aristotle

Each assimilates love of others to love of self. But Frankfurt’s conception of value is fundamentally subjectivist and in that sense egocentric: the value the beloved has for the lover is a function of the lover’s actual attitudes towards the beloved, not of features that belong to the beloved independently of the lover’s attitudes towards her.

[G] It need not be a perception of value in what he loves that moves the lover to love it. The truly essential relationship between love and the value of the beloved goes in the opposite direction . . . . [W]hat we love necessarily acquires value for us because we love it. (2004, 2.3)

Even in the case of his own children he says that the particular value he ascribes to them is “not inherent in them but depends on [his] love for them.”

[H] The reason they are so precious to me is simply that I love them so much. As for why it is that human beings do tend generally to love their children, the explanation presumably lies in the evolutionary pressures of natural selection. In any case, it is plainly on account of my love for them that they have acquired in my eyes a value that otherwise they would certainly not possess. (2004, 2.3)

This has a tautologous air that goes with the subjectivist territory, an air his children might well find unsettling: their value to him depends on the fact that he loves them, which is something he might well fail to do. His children may have the pressures of natural selection on their side, but that is hardly reassuring. For though parents “tend generally to love their children,” some parents—arguably a significant number—fail to do so. And perhaps more importantly, children (like adults) surely want to be loved ‘for themselves’, by which I (following Aristotle) mean, for features that belong to them, insofar are they are the individuals they are.

Honoring this desire is surely part of the point of Frankfurt’s second ‘conceptually necessary’ feature. And the importance of honoring it is central to Aristotle’s ideal. In friendships on account of utility and pleasure, the friend is loved on account of features that happen either to serve the needs or to fuel the pleasure of the lover. So these relationships are fundamentally egocentric. But in friendship on account of virtue, the friend is loved on account of her character (êthos), which is what by her own lights makes her who she is. Aristotle’s ideal is thus ethocentric.¹¹

This brings me to another related difference: Frankfurt’s conception of love, starting with self-love, is ‘brute’ as distinct from ‘motivated’.

On ‘brute’ conceptions, self-love is simply given. We might seek to explain it—like the parental extension or analogue of which Frankfurt speaks—by appeal to the pressures of natural selection. But it is not viewed as requiring any warrant or justification. A theorist who assimilates philia to self-love so construed will see philia in much the same way, as a matter of brute fact to be explained rather than warranted or justified.

Some ‘brute’ theorists see this as a virtue of their accounts: they prize ‘unconditional’ love and so reject the idea of grounding love of one’s friend in her instantiation of features that warrant love. But to the extent that their objection to ‘motivated’ conceptions rests on the ideas, taken together, that we should love others as we love our selves and that we love ourselves no matter what we are like, the force of the objection may be softened by the thought that there is an important sense in which self-love both is and ought to be based on appreciation of one’s own love-worthy features (Whiting 1991, in 2016).

On ‘motivated’ conceptions, self-love not only admits of warrant but to some extent calls for it. The idea is that self-love, properly construed, is a disposition to value oneself on account of what one takes to be one’s own merits. These merits may include the ideals one aims to realize, in which case the warrant may rest less on their actual realization than on the agent’s steadfast pursuit of that.¹² A theorist who assimilates philia to self-love so construed will see philia in much the same way, as a disposition to value the beloved on account of his merits, including ideals he aims to realize. Here too the warrant may rest less on actual realization than on the beloved’s steadfast pursuit of that.

The motivated conception is clearly present in what Aristotle takes to be the ideal form of philia, now known as ‘character friendship’. The motivated conception is also present, in ways not always appreciated, in Aristotle’s conception of self-love properly construed Whiting (1996). This is clear from Nicomachean Ethics 9.4, where Aristotle ties the features by which friendships are defined to features of the decent agent’s relations toherself.

Aristotle begins not with a bold assertion of ‘conceptually necessary’ features but with a survey of the endoxa (i.e., the entrenched views of the many and/or those reputed to be wise). He says that people take a friend to be one who:

• (a) wishes and does good (or apparently good) things for the sake of [the friend] (heneka ekeinou);

• (b) wishes the friend to exist and to live for his sake (autou charin);

• (c) spends time with, chooses the same things as, and feels pains and joys together with the friend.

Aristotle then argues that each of (a)–(c) is exhibited in some privileged or extreme way in a decent person’s relationship to herself—and also to some extent in anyone who takes herself to be decent.

Frankfurt introduces no such restriction: he does not even require that self-lovers take themselves to be decent. Here lies a profound difference. Aristotle recognizes the independent and objective value of features instantiated in lovers as well as their beloveds, and he thinks it proper for people to love themselves as well as others on account of such features; Frankfurt does not.

Frankfurt allows that some lovers in fact take themselves to love others on account of such features and take their love to be warranted by the relevant features. But even here he takes the value the beloved has for the lover to be a function of the lover’s attitudes, not of features that belong to the beloved independently of any relationship in which she stands to the lover.

We can best appreciate Aristotle’s view by noting the way 9.4 stresses (a) the distinctness of the beloved from the lover, and (b) the lover’s wish for the existence of the beloved for the beloved’s sake. Aristotle says that (b) is especially characteristic of mothers and of friends who have quarreled. His idea seems to be that such lovers wish for the existence (and presumably also the well-being) of the beloved for the beloved’s sake, even when they do not wish for what comes next: namely (c). For friends who have quarreled may in fact avoid spending time with one another and avoid choosing the same things as one another.

The closest we come to something like identification is in (c), where the friend is said to “spend time with, choose the same things as, and feel pains and joys together with the friend.” But even if this amounts to the sort of identification demanded by Frankfurt, there is no sign that Aristotle views it as conceptually necessary: he is simply referring to phenomena commonly associated, in the views of the many and the wise, with philia. In fact, he explicitly acknowledges that some lovers do not wish to spend time with or to share the pleasures and pains of their beloveds: fathers, for example, chose their offspring to exist but prefer to live together with others (Eudemian Ethics 1240a27–30).

Aristotle’s claim that (c) is especially characteristic of mothers deserves comment. Elsewhere he cites, as a sign that philia consists more in loving than in being loved, the fact that some mothers give up their children to be raised by others, loving them without seeking to be loved in return when they cannot have both, it being enough to see their children doing well (Nicomachean Ethics, 1159a25–34). He stresses the pleasure a mother takes in seeing her children flourish even when, because the children do not know her, she receives from them none of the things (including love) due to a mother. He makes a similar point at 1239a35–b2, where he refers to Antiphon’s Andromache arranging a secret adoption for her child. This practice, though less common than in past decades, is familiar to us.

On parental love, the contrast between Aristotle and Frankfurt is striking. Frankfurt seeks to assimilate the biological or genetic father’s relationship to the gestating fetus and the father’s experience of that to the gestational mother’s relationship to the fetus and her experience of that, thus suggesting that the parents have a kind of gender-neutral experience of their offspring as, at least for some time, organic ‘parts’ of themselves. But Aristotle cites mothers’ love for their offspring, where the ‘part’ talk might seem most apt, as paradigms of love for another for the other’s sake.

I cannot help but wonder whether the alleged phenomena are rooted in traditional conceptions of the father as the “true parent of the child” (Aeschylus, Eumenides 658–63). But I cannot now dig into this. What matters here is the contrast between Aristotle’s list of endoxa and Frankfurt’s list of ‘conceptually necessary’ conditions, especially the presence on Frankfurt’s list but not on Aristotle’s of the lover’s identification with his beloved, however exactly that is to be understood. So let us turn to that.

4. Identification: Psychological and Ontological

Let me begin by drawing two distinctions. The first is between (a) identifying with the beloved herself and (b) identifying with one’s love of the beloved. The second is between two general ways of understanding (a) and (b), one psychological, the other ontological.

Psychological identification is a matter of the lover’s attitudes, either towards the beloved herself or towards the lover’s love of his beloved. These attitudes can be understood in different ways. They may be either ‘brute’ or ‘motivated’. And if motivated, they may be motivated in different ways; some may rest on the lover’s beliefs about the relationships in which he stands to the beloved, some not.

Ontological identification is a matter of the lover’s and beloved’s respective boundaries. We can ask, with respect to (b), whether the lover’s love of his beloved is part of what makes him the particular person he is and so something without which he cannot survive as that person. And we can ask, with respect to (a), whether either party is in some sense an ‘extension’ (or ‘part’) of the other; or whether (as in ‘union’ accounts of love) each is in some sense ‘part’ of some larger whole, perhaps a ‘joint agent’.¹³ Here again there are different ways to construe the phenomena. There may be some fact of the matter, independent of either party’s attitudes, that determines their respective boundaries; or these boundaries may be determined at least partly by the parties’ attitudes.

On Frankfurt’s account, it seems that Agamemnon identifies, as a matter of psychological fact, both with being a devoted father and with being a devoted leader of his men, and that this is supposed to have ontological consequences. One might wonder whether the psychological attitudes, described in terms of Agamemnon’s sense of his own identity, entail actual love of his daughter and/or his army, but let’s spot him love for now. The idea is that psychological identification with his love of his daughter serves together with psychological identification with his love of his army to constitute a ‘volitional nature’ by which he is defined in a way such that he, speaking now ontologically, cannot survive the loss of either attitude. But why think psychological identification in this ontologically robust sense conceptuallyrequired for love?

Here Frankfurt faces a dilemma. Either he adopts an ontologically robust conception of identification, in which case it is implausible to say that identification is conceptuallynecessary. Or he adopts the sort of deflationary conception invited by the ‘that is’ clause in [A]’s presentation of the third condition, in which case his view is vulnerable to the Nagel-inspired argument I ran in “Love.”¹⁴

On the deflationary conception, identification is simply a matter of the lover being disposed to count the beloved’s interests among his own. There may or may not be a backstory here. And there may be different backstories in the case of different sorts of lovers.

A rational egoist may feel the need to tell himself a backstory. He may find himself unable to count the interests of his beloved among his own unless he can view her as in some sense an ‘extension’ or ‘part’ of himself. Or he may need to tell himself that she is part of some larger whole of which he himself is a part, perhaps a whole with which he ‘identifies’ in the sense that he takes his own existence to depend on his being and continuing to be a part of that whole. In other words, the lover’s disposition may as a matter of psychological fact depend on his assumption (whether correct or not) that some form of ontological relation obtains.¹⁵ But the disposition is itself a matter of his psychology.

Other lovers may feel no need for a backstory. A lover may simply find himself, as a matter of brute psychological fact, caring for his beloved in ways such that he counts her interests among his own, ways not so different from those in which he counts among his own the interests of his dog, his favorite movie star, or the local team for which he roots.¹⁶ In these cases, there is no assumption of ontological identification and there may be little if any of the common parlance form of psychological identification. Whatever exactly explains his attachment, it is the attachment itself that gives rise to his disposition to count the beloved’s interests among his own.

There is no need here for any ontological identification with his beloved. But [D] suggests that Frankfurt views lovers as at least sometimes assuming—whether correctly or not—some such identification. For while the last sentence of [D] is easily read as invoking psychological identification, the preceding sentence seems to invoke an ontological relation: the child’s being, at least early on, in some organic way a ‘part’ of the parents. So the final sentence makes it sound as if the degree to which parents identify psychologically with their offspring tends to correspond to the degree to which they see their offspring as ontologically connected with them, as somehow ‘parts’ (or perhaps ‘extensions’) of themselves.

Even so, the final sentence does not entail that either parent correctly sees the child as somehow a ‘part’ of itself. The idea could be simply (a) that biological parents are “naturally and more or less irresistibly” disposed to see their offspring, at least early on, as ‘parts’ of themselves, and (b) that this psychological tendency is what leads parents to count the interests of the offspring among their own.

This ‘deflationary’ reading has two clear advantages. First, it does not assume the truth of any ontological claim to the effect that an infant is at some point (even if only in utero) a ‘part’ of either parent. Second, it provides a ready explanation for the failure of some parents to count among their own interests the interests of offspring they take to be their own, parents who neglect their children or are downright abusive: perhaps they fail to view their ostensible offspring as ‘parts’ or ‘extensions’ of themselves.

But here we might wonder: is the love in which parents do count the interests of their offspring among their own best explained by taking these parents to view their offspring (whether correctly or not) as ‘parts’ or ‘extensions’ of themselves? What should we say about adoptive parents and step-parents, or stepsiblings and even teachers, who love and put first children whom they do not take to stand in any biological or genetic relation to themselves?

Let me approach these questions by turning to Iris Murdoch, whose ideal not only fails to assume identification but seems in fact to deny it. Can Frankfurt plausibly claim that her ideal rests on conceptual confusion?

5. Murdoch on Frankfurt’s First Two Features

The salient features of Murdoch’s account correspond to Frankfurt’s first two features. She treats love as a kind of knowledge of an individual as such, a kind of knowledge that involves ekstasis in the sense that it requires its subject to attend in a detached way to what lies outside itself. And she views such knowledge as being, in spite of its detached nature, practical in the sense that it “occasions right conduct.”

Murdoch likens the relevant sort of detachment, which she calls ‘impersonal’, to the detachment involved both in the production and in the proper appreciation of great art.

[J] . . . great art teaches us how real things can be looked at and loved without being seized and used, without being appropriated into the greedy organism of the self . . . . Unsentimental contemplation of nature exhibits the same quality of detachment: selfish concerns vanish, nothing exists excepts the things which are seen . . . .¹⁷

It is also clear that in moral situations a similar exactness is called for. I would suggest that the authority of the Good seems to us something necessary because the realism (ability to perceive reality) required for goodness is a kind of intellectual ability to perceive what is true, which is automatically at the same time a suppression of self.¹⁸ (1970, 353–54)

It is plausible to associate the sort of “unsentimental, detached, unselfish, objective attention” of which Murdoch speaks with disinterested concern in what Frankfurt calls the “clear and literal sense of being motivated by no interests other than those of the beloved.”

Consider the following example, which highlights the problems with taking parental love, on account of its proximity to self-love, as a general model for love of others: the parental model does not even work for one’s own maturing offspring. Imagine the father of someone in her late teens who is struggling to become her own person. However sentimental the father may be when he recalls his ‘little darling’, he needs to start checking his sentiments at his deliberative door. However concerned he himself is with financial security—perhaps he is a stock broker and she wants to become a professional philosopher—he should seek ways to detach from what he himself wants for her so as to gain a clear view of what she wants for herself. And however much he would stand to benefit from her staying at home to help care for her cognitively challenged brother, he should at least consider the potential costs to her of foregoing some apparently tremendous opportunity in order to do so. And he should seek in considering such things a kind of objectivity that allows (and indeed requires) him to assess how tremendous the opportunity really is along with the probable costs to her, given the nature of her own attachment to her brother, of pursuing this opportunity at this point in time. In other words, the likelihood (if real) of her future regret is among the phenomena he may—indeed must—seek to view in an objective way. There are no easy answers here, either for him or for her. But I think we all have an intuitive sense of the sort of attention Murdoch takes to be required of a loving father, and of the way in which the father’s paying such attention may lead to improvement in the ways in which he interacts with his daughter.

Murdoch associates such attention with what she calls ‘unselfing’ and sees in the conception of erôs as ekstasis found in Plato’s Phaedrus. It is important to Murdoch that this is a form of experience with which we are all to some extent familiar through encounters with beauty, especially perhaps natural beauty. The subject, struck by a beautiful sight, sound, or whatever, is transported out of his ordinary condition: he is ‘detached’ from his own everyday concerns and focused exclusively on the object of his attention.

[K] Our minds are continually active, fabricating an anxious, usually self-preoccupied, often falsifying veil which partially conceals the world . . . . Following a hint in Plato (Phaedrus 250) I shall start by speaking of what is perhaps the most obvious thing in our surroundings which is an occasion for ‘unselfing’, and that is what is popularly called beauty . . . . I am looking out of my window in an anxious and resentful state of mind, oblivious of my surroundings, brooding perhaps on some damage done to my prestige. Then suddenly I observe a hovering kestrel. In a moment everything is altered. The brooding self with its hurt vanity has disappeared. There is nothing now but kestrel. And when I return to thinking of the other matter it seems less important. (1970, 84–85)

Like Plato, Murdoch associates the sort of experience she has in mind with the highest form of love. And qua Platonist, she surely thinks this an ideal to which any form deserving the name should aspire.

That lovers are so moved, and that it is characteristic of the best sort of love to be so moved, seems to me the point of Socrates’s second speech in the Phaedrus. That’s why he connects love not (as Diotima had urged him in the Symposium) with self-propagation, but rather with ekstasis—i.e., with the lover’s being taken outside of herself. This happens when the lover is so struck by and so appreciative of the beloved that she forgets herself, and the preservation or propagation of herself, and allows herself to be moved directly by the interests of her beloved. Frankfurt can say if he likes that her very identity is altered, that she becomes a different person. But her in-love identity is a consequence of her appreciation of the beloved and not what explains it. So this identity plays no significant explanatory (or justificatory) role.¹⁹

In sum, it is difficult to see unselfing as a model for the sort of disinterested concern in which Frankfurt takes love of another ‘most basically’ to consist. For the third of his ‘conceptually necessary’ features seems to get in the way: if love consists ‘most basically’ in disinterested concern for the beloved, then why require the lover to ‘identify with’ the beloved in the first place?

This is the challenge posed by Murdoch’s view. Suppose we encounter an agent who habitually assigns priority to what are ordinarily regarded as the interests of another person. Should we assume that an adequate explanation requires us to view the agent as taking the other (whether correctly or not) to stand in some sort of privileged (perhaps ontological) relation to her? Why in the absence of a theoretical commitment to psychological egoism should we assume that?

Can a lover not insist that what she appreciates in her beloved is value that resides in the beloved independently of her own appreciation of it, indeed independently of any relationships (such as parenthood) in which she stands to the beloved? And in cases where a lover reports her experience as being like this, shouldn’t the theorist of love give some weight to what she says? Should he not seek to accommodate her experience within his theory? And shouldn’t he do that even if he himself, perhaps in a theory-laden way, experiences the value of his beloveds as in the eye of their beholder?

There are large questions here about the kinds of experience to which ethical theories should appeal and the proper use of experiential phenomena in defending such theories. I cannot dig into them here. What matters is that Murdoch seeks to support her ideal by citing relatively familiar experiences in which features of a beautiful object capture the attention of a subject in ways that move the subject to want to get to know, to honor, and to promote the interests of the object as what it itself is, apart from the subject’s attitudes towards it.

Murdoch’s account embodies a kind of commonsense realism that many philosophers (for example projectivists) would reject. We cannot resolve these issues here. The point is simply that her account, like Frankfurt’s, is embedded in a broader Weltanschauung. Murdoch is attacking the sort of post-Kantian philosophy Frankfurt embraces, a philosophy dominated by the “notion of the will as the creator of value,” one in which “the idea of the good remains indefinable and empty so that human choice may fill it” and “the sovereign moral concept is freedom . . .” (Murdoch 1970, 80–81). She seems moreover to be working in opposite direction from Frankfurt. His account is shaped largely by his persistent concern with autonomy together with his subjectivist conception of value, while Murdoch starts from a kind of commonsense realism concerning the nature of love and explicitly proposes to interpret talk of ‘the will’ and ‘freedom’ in light of that. This is clear from [J], which appears in a context where Murdoch is arguing that

[K] consideration of what the effort to face reality is like, and what its techniques are, may serve both to illuminate the necessity or certainty which seems to attach to ‘the Good’; and also to lead on to a reinterpretation of ‘will’ and ‘freedom’ in relation to the conception of love. (1970, 64)

6. Beloved

I want to conclude by comparing what Frankfurt says about Agamemnon with the story of Sethe, the mother in Toni Morrison’s Beloved. Sethe’s story was inspired by a newspaper clipping about an escaped slave, Margaret Garner, who in 1856 took the life of one of her four children and was preparing to take the lives of the others rather than let them be captured by fugitive slave-hunters who had come to return them all to the Kentucky plantation from which they had escaped.²⁰ Her story, at least as Morrison tells it, exposes the tension between the first and third of Frankfurt’s ‘conceptually necessary’ features. And it does so in a way that supports dropping the third.

Sethe’s act comes 28 days after a harrowing escape in which she gives birth while seeking to carry herself—and, more importantly in her mind, her milk—to the ‘still-nursing’ daughter she had sent ahead with her other children. The story, as Morrison tells it, raises questions about the ‘purity’ (as in [B]) of Sethe’s self-love. As she describes her escape, she says “all I knew was I had to get my milk to my baby girl” (p. 19), milk that was stolen from her by white boys on the plantation in an act of torture that was (unbeknownst to Sethe) observed by Halle, the father of her four children.

Sethe later describes her escape to Denver, the daughter she birthed along the way.

[L] Concerned as she was for the life of her children’s mother, Sethe told Denver, she remembered thinking: Well, at least I don’t have to take another step. An dying thought if ever there was one, and she wanted the little antelope [viz.Denver in utero] to protest . . . “I believe this baby’s ma’am is gonna die in the wild onions on the bloody side of the Ohio River.” That’s what was on her mind and what she told Denver. Her exact words. And it didn’t seem like such a bad idea, all in all, in view of the step she would not have to take, but the thought of herself stretched out dead while the little antelope lived on—an hour? a day? a day and a night?—in her lifeless body grieved her so, she made the groan that made the person walking on a path yards away halt and stand still . . . . The voice, saying “Who’s in there?”

Sethe is at this point rescued by a white girl, Amy, who coaxes her on and cares for her through the birthing before abandoning her on “the bloody side of the Ohio” to be rescued and carried across to find her children.

The italicized phrases call attention to the way Sethe is moved by thoughts centered on her children, thoughts in which she characterizes herself as ‘her children’s mother’, ‘this baby’s ma’am’. It is thoughts of Denver (in utero) and the still-nursing child in Ohio in need of her milk that keep Sethe going when she is tempted to lie down and die. She is clearly driven by her love for her children, even when she seeks to take their lives. The still-nursing child whose life she takes is called ‘Beloved’ after the name engraved on its headstone, an engraving for which Sethe has to pay with sex. Sethe wonders later whether, for another 10 minutes or so, she could have bought ‘Dearly Beloved’, the only words she managed through her grief to hear the preacher say at the burial.

Sethe’s acts—first in taking the life of her daughter and then in paying for the engraving—are surely acts of love. Frankfurt might seek to explain them by saying that Sethe identifies, if not with Beloved with her love of Beloved, and so sees her acts as expressions of who she herself is. But that is not how Morrison tells the story. Morrison draws our attention to Sethe’s focus on the needs of the children as such: the still-nursing child’s need for "it’s ma’am’s" milk and the need of ‘the little antelope’ for "its ma’am’s" survival.

It is difficult to adopt an account of Sethe’s ‘sacrifice’ of Beloved like the one Frankfurt gives of Agamemnon’s sacrifice of Iphigenia. In Agamemnon’s case, there is clearly something that he himself stands to gain by taking the life of his daughter: viz. the prospect of military victory and with it the sort of eternal glory attendant on that. But in Sethe’s case, it is not clear that there is anything comparable she herself stands to gain. In other words, Sethe’s act seems to be motivated primarily—and perhaps even exclusively—by concern for Beloved.

One could of course say that Sethe acted to spare herself the pain of having to see Beloved—or simply to think of Beloved—growing up in the brutal conditions on the plantation from which they had just escaped. But the hypothesized pain assumes that Sethe has the sort of immediate and powerful concern for the welfare of Beloved that the hypothesis is designed to avoid. We might as well say straightaway that she acts primarily (and perhaps even exclusively) for the sake of her daughter. In other words, we might as well say that her act was disinterested in what [B] calls “clear and literal sense of being motivated by no interests other than those of the beloved.”

Consider Sethe’s act of taking Beloved’s life. However unlikely it would have been for Sethe to be thinking of this at the moment of action, she must at some level have been aware that the penalty for her act could well be her own death. So it is difficult to view her act as motivated by a concern for self-preservation, either in the familiar sense or in what Frankfurt calls the "unfamiliarly literal sense.” On Frankfurt’s account, Sethe was not only sacrificing her volitional self but risking literal death. Yet the moment she saw the slave-hunters, she ran to spare her children the horrors of the life from which they had escaped: whether she herself would be returned to the planation (as the historical Margaret Garner was) or hung (as proposed in the fictional trial of Sethe) was presumably the furthest thing from her mind. There was nothing then but the children.

What we see here is not an act of self-preservation but rather an act stemming from what Murdoch calls ‘unselfing’. As I read Beloved, Sethe’s life consisted largely in that—at least until Paul D showed up, 18 years after her escape, and started loving her into loving herself. Paul D had been fellow slave and friend of Halle at Sweet Home (the plantation taken over shortly before Sethe’s escape) by a man called ‘schoolteacher’. Paul D had long loved Sethe and had from the start honored her by awaiting and then respecting her choice of Halle as her partner. But Paul D now begins to express his love in ways that launch a slow and fitful process of selfing on Sethe’s part, a process interrupted when he learns, eventually, of her act and walks out on her, but not before judging her harshly. He charges that her love is "too thick", claiming that it did not work because her boys were gone (she knows not where), one girl is dead and the other afraid to leave the yard, to which Sethe replies, “They ain’t at Sweet Home. Schoolteacher ain’t got ‘em.” Paul D then says “what you did was wrong, Sethe . . . you got two feet, Sethe, not four.”

Still, Paul D loved Sethe a way such that he was willing to reflect on things from her point of view and he ended up, with minimal encouragement from Denver, returning to pick up where he left off in coaxing Sethe towards a kind of selfhood whose formation had been blocked throughout her life by her constant service to others, first to her masters and then to her children. This coaxing is epitomized in the final scene, where Paul D finds her crying and saying “She left me . . . she was my best thing.” It is not entirely clear to whom Sethe refers, most likely the ghost of Beloved that had been haunting her. But Paul D, leaning over her and holding her fingers, says “You your best thing, Sethe. You are.” And she replies “Me? Me?”

The moral here is this: even if human beings tend for the most part to need unselfing, there are some who need selfing—especially perhaps those who have been consigned by the free-ranging wills of others to social positions that tend to thwart the development and expression of the sort of volitional self that looms so large in Frankfurt’s Weltanschauung.

REFERENCES

Footnotes