Though the conception of self I pursue draws on many sources, some quite old, one especially prominent source is contemporary: the influential writings on this topic by Professor Harry Frankfurt. His views are frequently in the background of what I say, even when there is no specific acknowledgment. The opportunity to engage with his views directly is, accordingly, particularly welcome. I was given such an opportunity in commenting on Frankfurt’s Tanner Lectures.1 Seen within the context of this book, this chapter considers three main themes. One is a continuation of a theme broached in the first chapter, concerning the constructive significance of ascribing responsibility and of the role played by different social actors in doing so. The second theme, further developed in the next chapter, is an analogy between the boundaries of self and those of the state. And the third, which is picked up mostly in Chapters 5 and 6, links the meaning-conception of self to an appeal to levels of abstraction as the medium or the source of social as well as universal moralities.
There are two main themes in Frankfurt’s Tanner Lectures, as there are in his work in general: one concerns autonomy and freedom of the will, and the second concerns the nature of normativity. They can be summarized as follows. First, we either identify with an attitude, or we don’t. This defines the shape of the will, the extent of our autonomy, and, as he puts it elsewhere, the boundaries of the self. Second, we either care for something or we don’t. This provides the ground of normativity. My general point is that Frankfurt’s focus on individual psychology has to be expanded to take account of the intersubjective or the social. Using the distinction I have introduced in Chapter 1, this suggestion can be best seen as a comment on the we in the italicized statements: Frankfurt uses the pronoun distributively, whereas in extending the theory in the direction I propose, the we would better serve if used collectively.
Frankfurt’s lectures address the shape of the will and the nature of autonomy, but they also treat of responsibility. My main interest is in the latter term. When are we responsible? What are we responsible for? The key to Frankfurt’s conception of autonomy, but also of responsibility, is the idea of identification. In his lectures, the connection between responsibility and identification is indicated most explicitly in the discussion of character. According to Frankfurt, responsibility for character
is not essentially a matter of producing that character but of taking responsibility for it. This happens when a person selectively identifies with certain of his own attitudes and dispositions, whether or not it was he that caused himself to have them. In identifying with them, he incorporates those attitudes and dispositions into himself and makes them his own.2
It’s a short step from this account of responsibility for the character traits themselves to a similar account of the responsibility the agent bears for actions that issue from those character traits and in which those traits are exhibited or expressed.
This extension of the theory of responsibility can be applied to a dramatic hypothetical that Frankfurt presents. Frankfurt imagines himself as a loving father who is beset by a desire to kill his son. “The desire,” he says “is wildly exogenous; it comes entirely out of the blue,” and “is ordinarily safely repressed.”3 But now consider the harrowing situation in which the repression is unsuccessful and in which the desire does prevail. This in fact happened in the case of Regina v. Charlson.4 The defendant’s ten-year-old son entered his father’s study. For apparently no reason, Charlson hit the child over the head with a heavy mallet and threw him out of the window. Fortunately, the child was not killed. At his trial, Charlson successfully pleaded involuntariness: he was suspected of suffering from a brain tumor, and he alleged that this explained his behavior. A claim of involuntariness amounts to a total denial of responsibility equivalent to the statement, “I didn’t really do it.” Now on the conventional understanding of involuntariness, this claim is read with the stress on the word do. The inquiry is: was an action involved here? And this we tend to interpret as raising a question of control: could Charlson have acted otherwise? Was compliance with the law an option for him at the time?
The difficulty with this conventional interpretation can be seen starkly if we compare Charlson with another case, State v. Snowden.5 Snowden was involved in what appeared to be a minor quarrel with a woman outside a bar. At some point, he claimed at his trial, she kicked him. In response Snowden took out a knife and stabbed her to death, inflicting more than ninety wounds over her entire body. Snowden’s explanation of this response was simple: when kicked by the victim, he flew into a rage and lost control; in his own words, “I blew my top.” On his version of the events, his action was no more up to him than Charlson’s was up to Charlson. But, I think not surprisingly, Snowden’s defense was unsuccessful; he was convicted of first-degree murder. Now when interpreted in terms of the idea of control, the difference in result between the two cases is puzzling. Can it be said beyond reasonable doubt—which, after all, is the standard of proof in a criminal trial—that Snowden could have contained his temper, reined in his fury, and subdued his murderous impulse? Indeed, on Frankfurt’s view this counterfactual inquiry would be misguided.6 Frankfurt’s approach suggests instead that we reorient the inquiry by reading the claim “I didn’t really do it,” implicitly made in these cases, with a different intonation, accenting the “I.” Applied to Charlson, the claim is, “It was not really ‘I’ who brought about the injury; it was the tumor.” In Frankfurt’s terms, Charlson refuses to identify with whatever prompted his murderous outburst and to take responsibility for it. He “banishes” these promptings by placing them outside the boundaries of his self, or to reverse the metaphor, he draws his boundary so as to leave these promptings outside. Either way, the control such promptings exercise over Charlson is “external” and “tyrannical.” This would explain why he was indeed acquitted. Frankfurt’s approach also explains why we don’t seem to be particularly perturbed by whether or not Snowden was able to contain his rage and subdue his outburst. On Frankfurt’s view, as far as responsibility is concerned, the fact that Snowden couldn’t help but act the way he did is neither here nor there. The difficulty, however, is that as it stands Frankfurt’s own account may exempt Snowden of responsibility as well. On this account, the maneuver attempted by Snowden closely resembles Charlson’s. By saying that he blew his top, Snowden can be understood to convey his refusal to identify with this irresistible rage; like Charlson, he too would rather draw the boundary of his self in a way that leaves the fury outside. If such dissociation were successful, it would, after all, keep him, as it did Charlson, out of jail. But at least as far as the jury in this case was concerned, the maneuver failed. What are we to make of Charlson’s success in defending himself against criminal charges and Snowden’s failure?
The basic insight that greatly contributes to our understanding of these cases seems to me the connection indicated by Frankfurt between responsibility and the boundaries of self. But when it comes to the ascription of responsibility, Frankfurt’s approach must be supplemented in order to account for the difference between the two cases. The crucial point is that the self’s boundaries are not drawn unilaterally, not only from within. The shape of the self is at least in part the product of what we may call constructive practices, including those of law and morality. Central among these practices are those of ascribing or withholding responsibility. As the cases suggest, the drawing of the self’s boundaries may involve a process of negotiation, in which the agent participates, but over which she has no unilateral control. Through the jury, society plays an active role in drawing the defendant’s boundary. On this view, the verdict in Snowden amounts to a determination that, unlike a tumor, rage is internal to the self, a regrettable yet legitimate component of one’s character and personality, and so something for which one bears responsibility.
Although questions of responsibility arise on innumerable other occasions as well, the criminal trial provides a particularly visible and stylized setting for the kind of negotiation involved. The normative stakes in drawing the self’s boundary are also particularly high in this context.
Defendants are typically eager to draw their boundary narrowly so as to escape the nasty ramifications of legal responsibility. This need not be just strategic posturing on their part: the phenomenology of withdrawal or flight from responsibility is altogether familiar and real. The prosecutor, eager to pin down responsibility to advance law enforcement and carry out justice, can be prompted by equally genuine indignation and resentment to advocate drawing the self’s boundary widely. These momentary pressures and concerns of the trial should not, however, be allowed to eclipse the long-term and more general normative incidents of the self’s boundary. The latter are obviously more complex than the immediate, momentary ones, but the political context from which the boundaries metaphor is drawn provides a useful, if simplified, analogy that affords a glimpse of the main considerations. A state’s boundary settles at once the scope of both its sovereignty and responsibilities. Replacing sovereignty with autonomy, the more apt label for an individual’s self-rule, we get a picture, alluded to in Chapter 1, in which autonomy and responsibility are coextensive, both defined by the boundaries of the self.7 To abdicate responsibility by contracting the self’s boundary is accordingly also to forfeit part of one’s autonomy, since by evacuating potential responsibility bases we also give up regions of autonomy and self-rule. Moreover, responsibility is itself a two-sided concept. The moral and especially the legal context focus for the most part on forbidden behavior and thus bring to mind responsibility’s negative side, as a source of blame and a basis for sanctions. But questions of responsibility also arise concerning credit due for positive actions and events. By defining the scope of one’s responsibility, the boundaries of the self thus determine not only the extent of one’s vulnerability to blame and punishment, but also the sources of satisfaction and gratification, of praise and reward. Drawing the self’s boundary is a delicate balancing act, in which both the momentary and the longterm perspectives play a part, and in which conflicting considerations and difficult tradeoffs apply. I will not expand any further these cursory remarks on the nature of this process and will instead briefly comment on the connection between the ascription of responsibility, which Frankfurt does not explicitly consider, and the assumption of responsibility, which he does. I do so by relating this connection to another, I think particularly moving, point in the lectures.
Frankfurt speaks of harmony within the self, a congruity between higher-order attitudes and lower-order ones, and links this state to Spinoza’s ideal of “acquiescentia in se ipso” (“acquiescence to oneself”).8 If the self’s boundaries are drawn, as I suggest, through social practices in the public domain, and in a process that involves something like a negotiation between the agent and others, another kind of harmony or dissonance comes into view. The negotiation may end in agreement, as it apparently did in the Charlson case, where society, represented by the jury, came to accept the defendant’s dissociative maneuver and drew the boundary accordingly. There is, however, the possibility of a breakdown in negotiations, with each party insisting on a different version as to where the borderline is drawn. Snowden may be such a case. I say may be, because more than one scenario may unfold. One possibility is for Snowden to persist in the face of the conviction in denying his responsibility. Either in proud defiance or in embittered self-pity, he’ll consider himself the victim of two external forces that ruined his life: his rage is one, a cruel and uncomprehending jury the other. There is another possibility, though: Snowden may come to accept the verdict. This means that he will now align the boundary of his self as he conceives of it with society’s. Contrition and remorse are mechanisms through which such harmony between Snowden and society can be restored. A single version of his self, rather than two incompatible and competing versions, will emerge.
But what does it mean to speak about two competing versions of one and the same self? It may at first appear that there’s got to be a fact of the matter as to where a thing’s boundary lies. In a case of disagreement, one party—either Snowden or the jury—must have gotten it wrong. But this appearance is dispelled by the constructive conception of the process by which the boundary is drawn. Antecedent to the negotiations, there is no fact of the matter; the process fixes the segment of the boundary that is under dispute. It may be felt, however, that once the process is over and the boundary fixed one way or another, there can be only one self.
Refusing to acknowledge it at this point amounts to ignoring the facts. It is an advantage of the metaphor we’re using that it does not force this conclusion on us either. In the international arena from which the metaphor derives, indeterminacy of borders is all too familiar. There need not exist a single authority whose judgment is accepted by all. Hence different and incompatible versions may persist, though often with more or less disastrous consequences. These further implications of applying to the self the idiom of boundaries seem to me altogether apt. The self’s boundary is the product or reflection of normative determinations regarding a person’s responsibility, autonomy, and the like. Such putative determinations can be made by the person herself as well as by others, acting in various pertinent capacities. These determinations often converge. However, when they do not, no recognized supreme authority need exist to settle the dispute, and so, as in the case of the state, different versions of one and the same self may persist.
Let me now indicate some implications of this attempt to socialize Frankfurt’s theory for the second theme of his lectures, the issue of moral authority. Here again my interest is in the kinds of practices that the cases I’ve mentioned illustrate: not just ascribing responsibility, but blaming and punishing. Can Frankfurt’s approach account for these as well? According to Frankfurt, the authority of morality, and more broadly of all judgments of importance, is grounded at bottom in what we care about; in “the attitudes and dispositions of the individual.” “If what we should care about depends upon what we do care about, any answer to the normative question must be derived from considerations that are manifestly subjec- tive.”9 This view, I take it, parallels the one that was held by Professor Bernard Williams, that only internal reasons exist. On Williams’s version of this view, we cannot charge another person with a failure of rationality by reference to some objective values or standards, as long as these values and standards are not included in that person’s “subjective motivational set.”10 Both views imply a sense in which blaming others is pointless, unless they too care for what one cares about—unless their will is aligned in the relevant respect with one’s own. But Frankfurt’s view has the further and more striking implication that one can’t really blame others even when their will is aligned with one’s own. Surely my will has authority only over me. If the authority of morality derives from my will when my will endorses or accepts some moral precepts, morality too has authority only over me. By what right can I invoke its imperatives to blame others?
Put in other words, Frankfurt offers an attractive and metaphysically lean construal of the Kantian view that each person is a law unto himself. However, a question of jurisdiction now arises: even if the laws of two states have the same content, each state can prosecute only the violation of its own laws, not the other’s, because each legal system has authority only domestically. According to Frankfurt, if I violate my deep values and convictions, I betray myself; and by the same token, if you violate your values and convictions, even if they resemble mine, you betray yourself. What business is this of mine? By what authority can I condemn you? Of course, I can be mad at you for harming me or the things I love, or disparage you for your hypocrisy or for the weakness of your will. But neither anger nor disparagement is the same as blame. For me to be in a position to blame you for the violation of a moral norm, we must be both under its jurisdiction. One and the same norm must have authority over both of us. Where would such authority come from?
My comments on Frankfurt’s first theme indicate the general direction in which an answer may be sought. On Frankfurt’s view, the authority of what’s important comes from its importance for us. As I noted at the outset, Frankfurt uses the “us” distributively, whereas I propose to use it collectively. Support for this suggestion can be found in another insightful observation in Frankfurt’s lecture: “The fact that there are things that we do care about [or, to use Frankfurt’s other expression, that things are important to us] is plainly more basic to us—more constitutive of our essential nature—than what those things are.”11 What I take to be essential to human nature on this view is that some things appear to us under a certain description or designation, namely as “important.” The point as I understand it is that not only do some things appear as “important to me,” but also, and crucially, that certain things appear to me as “important.” Important, however, is a word, specifically an adjective, and what it takes for something to be important is that the adjective apply to it. But to say this is to withdraw exclusive authority from the individual over this bit of content or meaning. My suggestion is that “taking ourselves seriously,” as Frankfurt urges, requires that we take seriously the semantics of words, or concepts, such as important, that according to Frankfurt’s own view play in human beings an essential, constitutive role.
Once again, this is an appeal to an intersubjective context and to our mutual intelligibility. Though we often disagree vehemently about what is important, this is a disagreement in our understanding and interpretation of a single term or concept. All of us are willful subjects of one and the same authority, the authority of important and kindred basic normative terms such as perhaps right or appropriate, whose meanings are embedded in a shared conceptual framework that secures our common human intelligibility. When we blame each other, we invoke this authority, under which we all live. To use again the legal analogy, we are more like lawyers who disagree about the proper interpretation of one and the same statute whose authority they all concede than like the inhabitants of two different jurisdictions whose statutes happen to resemble one another.