Agnes Callard
The University of Chicago
4/1/15
Self-Creation Without a Regress1
Adults are, to varying degrees, responsible for the kind of people they have become. They play
some role in shaping their own interests, passions, idiosyncrasies, and moral sensibilities.
Genetics, parenting, and many different kinds of environmental factors also contribute, more in
some cases than others. If you become a bad person under unusually harsh conditions, and I do
so while having had the benefit of every advantage, I deserve more blame than you.
These commonsense intuitions point both to the existence of, and to the existence of
qualifications on, someone’s responsibility for being whatever kind of person she is. We take
ourselves to be responsible not only for what we do (or think or feel), but also for being such as
to do (or think or feel) those things. The second kind of responsibility is responsibility for the
persisting cognitive, emotional, motivational and evaluative condition that sometimes goes by
the name of “character” or “self.”2 We presuppose responsibility for self when we decry wasted
talent and laziness: “you had it in you to succeed”; as well as when we take privilege into
account when considering otherwise similar applicants: “describe the challenges you have
faced…”
These intuitions surface sharply when we hear a story such as that of Ryan Loskarn, former chief
of Staff to Tennessee Senator Lamar Alexander. On Dec. 11, 2013 Loskarn was arrested on
charges of possessing and distributing child pornography. His arrest was met with shock,
dismay, disapproval and an effort on behalf of his former associates to distance themselves from
him as much and as quickly as possible. In an instant, Loskarn went from up-and-coming
political player to being branded with that special kind of evil mark that we reserve for crimes
we don’t even want to think about. A month later, Loskarn committed suicide, and left a note in
which he revealed that he had been sexually abused as a child, at ages 5 and 9. The note was met
with a wave of empathy, such as we find in expressed in Ruth Marcus’ Washington Post
editorial, “The Tragedy of Ryan Loskarn.” Marcus writes, “Some people do terrible things
because they are purely evil, others because they are terribly damaged.” Seeing Ryan Loskarn
as a victim of abuse re-framed for us what it meant that he was, as he described himself, an
abuser of young children. Details of his troubled past seem to mitigate his responsibility for
ending up as he did, since they suggest that, to some degree, something other than Loskarn
himself made him into what he became.
1
Thanks to Jason Bridges for comments on an early draft of this paper, and to the attendees of the 2015 Workshop,
Steve White in particular, for very helpful criticisms.
2
I prefer a phrase such as “responsibility for self” or “responsibility for character” over the more succinct
“autonomy,” because that word invites one to conflate responsibility for self with that for action. The concept of
character has come under attack by ‘situationalists’ such as John Doris and Gilbert Harman, to whom Rachana
Kamtekar responds with a defense. In accounting for responsibility for self or character, we don’t need to address
ourselves to these arguments, for reasons Kieran Setiya has given: situationalists only deny the existence of “unified
traits that range across much of what we do, their influence apparent in widely varying situations.” (Setiya, p.75) It
is open to the defender of character might deny the existence of such ‘traits’ and understand someone’s character as
composed of dispositions that are more fine-grained, and not necessarily widely shared.
1
When philosophers discuss examples of agents like Loskarn, they are usually interested in the
question of how diminished responsibility for character might detract from someone’s
responsibility for the actions he performs from that character3. That is not the direction of
reasoning suggested by Loskarn himself in the note he left behind4:
“I understand that some people – maybe most – will view this as a contrived story designed to
find some defense for defenseless behavior. That it’s an excuse….But I’m sharing this with you
because it is the truth, not an excuse. And I believe it played a role in my story.”
Loskarn ends his note with an apology “to the children in the images”: “I should have known
better. I perpetuated your abuse and that will be a burden on my soul for the rest of my life.”
Loskarn apologizes to the children because, as he says, he sees himself as responsible for having
perpetuated their abuse. He does not seek to be exculpated for responsibility for, as he puts it
elsewhere in the note, “my crime.” He offers the details of his abuse because he believes “it
played a role in my story.” Loskarn recounts his abuse in order to secure (some) exculpation for
having become the kind of person that he is. But he does not show any signs of seeking
exculpation for what he did (or thought or felt or saw). Marcus’ editorial seems to take its cue
from Loskarn’s self-representation, granting that Loskarn “did terrible things” but only because
he was “terribly damaged.” The line plays on the word “terrible,” first invoking a sentiment of
righteous indignation at the immorality of his (presumably voluntary) evil actions, and then shifts
to an empathetic sadness in response to Loskarn’s victimization.
It is a deep philosophical question whether (and if so, how) it is possible for someone to be
responsible for -ing, if she is not (completely) responsible for the character disposing her to 5.
But that is not a question I want to explore here. My interest is in responsibility for character per
se, setting aside questions about the implications for responsibility for action6. My aim is to
For a canonical example, see Gary Watson’s 1987 discussion of the vicious murderer/victim of brutal child abuse
Robert Alton Harris.
4
Parts of Loskarn’s note are cited in Marcus’ article, and it can be found in its entirety at
http://www.jesseryanloskarnslastmessage.com/ It does not appear that he intended to kill himself at the time he
wrote this note.
5
Galen Strawson argues that responsibility for action entails responsibility for character, and that the impossibility
of the latter entails the impossibility of the former. Strawson’s central contention (discussed below) is that
responsibility for character must be based on something unachievable, namely self-creation. Fischer’s response to
Strawson’s argument is to encourage us to “scale back our demands for the sort of autonomy and ‘‘sourcehood’’
required for moral responsibility.” (p.117) I think such measures are unnecessary, for two reasons. The first is that
self-creation is achievable. The second is that I think Strawson is wrong about the entailment between the two
forms of responsibility: responsibility for action does not rest on responsibility for self. I argue for the first point,
against Strawson, below. For an argument for the second, see Wolf, forthcoming.
6
Character-responsibility includes cognitive, emotional and practical dispositions, whereas the action-responsibility
that I forbear from discussing covers all, and not merely agential, activations of those dispositions. Angela Smith
objects to a voluntaristic construal of what she calls “responsibility for attitudes.” She argues that we ought to
construe such responsibility in terms of the subject’s rational answerability for the thought or feeling; and that we
are responsible when those thoughts and feelings are traceable to our own evaluations. Her phrase “responsibility
for attitudes” obscures an important distinction between the disposition to, e.g., harbor racist beliefs or be jealous of
one’s sister and the activation of such a disposition in the form of thinking a racist thought or feeling a pang of
3
2
offer philosophical vindication for moral intuitions such as those described in my opening
paragraph, intuitions speaking directly to responsibility for character. Loskarn’s letter makes a
powerful case for the conclusion that he is not (fully) responsible for being a pedophile. But that
requires us to render the contrast case intelligible, one in which a person is responsible for the
kind of person she is. Meeting that requirement is the aim of this paper.
In virtue of what, then, is someone responsible for (some feature of) herself? One approach to
accounting for such responsibility is hierarchical and synchronic: we distinguish those features of
an agent for which she is responsible from those for which she is not by looking to her own selfevaluations. We are familiar with the phenomenon of agents such as akratics or addicts who
evaluatively reject some part or aspect of their motivational makeup. It might appear that a view
on which evaluative endorsement is necessary for responsibility is well placed to exculpate
Loskarn, given his expressions of self-revulsion. I argue, however, that if we attend to Loskarn’s
self-descriptions we will find that his unhappiness with himself is importantly dissimilar to the
self-rejection manifested by akratics or addicts. No part of Loskarn is able to withdraw to the
safe reflective position of an unimpeachable evaluator. His problems infect his values
themselves, and this fact does not escape his own notice. The case of Loskarn raises a standing
problem for any hierarchical view: how do we negatively evaluate our values themselves?
A diachronic account of responsibility for self will, by contrast, place emphasis on the adverse
conditions under which Loskarn’s desires and values were formed. On such an account,
someone is responsible for those features of herself which are the products of her own choices or
rational commitments. But voluntaristic self-creation is threatened by a paradox compellingly
articulated by Galen Strawson: if the feature I ‘create’ in myself follows rationally from values I
already have, then my ‘creation’ of it does not amount to a real change, but merely an
elaboration; but if the new feature is rationally unconnected to my earlier values, then its advent
in my life cannot be my own doing. Rational change is not self-creation, and irrational change is
not self-creation.
Both the problem of evaluating one’s own values and Strawson’s paradox are versions of a wellknown regress. In order to articulate the regress in its full generality, let us use the word ‘create’
to describe either the act of evaluatively sanctioning some feature of myself—for in this way I
am said to make it truly ‘mine’—or that of bringing the feature into being through antecedent
choice. The trouble for the theorist of self-creation begins when she is asked to identify the
creator’s creator. Unless she refuses to answer, ending the regress by arbitrary fiat, she will be
forced to posit as creator a yet higher or yet earlier self. The demand will then be repeated, and
she will be forced upwards or backwards through a series of ever less likely candidates to count
as a source of ultimate responsibility. Hence Nietzsche, in a passage cited approvingly by
Strawson and other skeptics of self-creation, scoffed at the philosophical impulse “to pull oneself
up into existence by the hair, out of the swamps of nothingness.” He describes self-creation as a
“rape and perversion of logic7.”
jealousy. Smith’s account may work as an account of responsibility for the activation of some evaluative disposition,
but I argue below that views such as hers leave open a question about responsibility for having the disposition.
7
Beyond Good and Evil, section 21. It is, however, far from clear that these lines represent Nietzsche’s final word
on self-creation. For in other places he waxes eloquent on the unique power of the strong to create themselves. See
3
The situation is not, I will argue, as dire as it appears. The regress depends on a prevailing, but
unnecessary, assumption about self-creation: that the creator self must be authoritative over the
created self. The alternative is to envision the created self as, instead, the agent’s authoritative
self. The key to dissolving the regress is to invert the traditional relation of conceptual and
epistemic priority between the two selves, allotting priority to the created over the creator. Selfcreation can coherently be construed as the aspirational process by which a human being works
her way towards responsibility for who she is. If I aspire, I do not see myself as fashioning,
controlling, sanctioning or shaping the self I create. Instead, I look up to her, try to understand
her, find my way to her. In aspiration, it is they created self who, through the creator’s imperfect
but gradually improving understanding of her, explains the path someone’s life takes.
I. Self-Endorsement vs. Self-Cultivation
Responsibility for self is naturally8, and I think correctly, understood self-referentially: I am
responsible for myself in virtue of some relation in which I stand to myself. Existing selfreferential accounts can be classified by whether they specify a synchronic or diachronic relation
between the selves. I will call the synchronic account, “the self-endorsement account,” and the
diachronic account “the self-cultivation account.”9
I begin with self-endorsement. One might hold that someone is responsible for those features of
herself of which she approves. Such approval can be spelled out in a variety of ways: as desiring
to desire what one fact desires (Frankfurt 1971), as wholehearted identification with a desire
(Frankfurt 1988), as reflective endorsement of one or another practical identity (Korsgaard, ch.
3), as being motivated in line with one’s evaluations (Watson 1975). The differences among
these views will not concern us here; the important point is that an agent, or some part of her,
steps back from, appraises, and attaches a positive or negative evaluation to the aspect of herself
she evaluates. Can we explain Loskarn’s diminished responsibility by pointing to the negative
outcome of such a reflective procedure? I think that a closer reading of his note suggests not:
“I found myself drawn to videos that matched my own childhood abuse. It’s painful and
humiliating to admit to myself, let alone the whole world, but I pictured myself as a child in the
image or video. The more an image mirrored some element of my memories and took me back,
the more I felt a connection.
the passages cited in Owen and Ridley, who are responding to the more straightforwardly fatalist reading offered in
Leiter.
8
An alternative view, one that I will not discuss, ties responsibility for self to meeting some substantive
requirement: psychological or physiological health (Buss), sanity (Wolf 1987), objectivity (Berofsky), resistance to
being ruled by others (Garnett). Though proponents of such accounts sometimes cite the fact that self-referential
accounts face a regress as motivation for their view (see Buss p.656, Wolf 1987 p.52), the positing of a substantive
requirement is not, in principle, incompatible with a self-referential account. It might, for instance, be that
fulfillment of the self-referentiality condition entails meeting one or more of these substantive conditions.
9
None of these theorists directs him or herself specifically at the project of account of responsibility for self as
opposed to responsibility for action. They do not tend to separate the two, and I leave open the possibility that they
might offer us a good account of the latter. My interest is only to explore why the strategy they deploy doesn’t
prove fruitful in accounting for responsibility for self.
4
This is my deepest, darkest secret.”
“In my mind I instigated and enjoyed the abuse – even as a five and nine year old – no matter the
age difference…. By my late teens I reached a sort of mental equilibrium on the matter. I
couldn’t stop the images from appearing altogether, but I generally controlled when they
appeared…As an adult I thought I was a tougher man because of the experience; that I was
mentally stronger and less emotional than most. I told myself that I was superior to other people
because I had dealt with this thing on my own. Those I worked with on the Hill would likely
describe me as a controlled, independent, and rational person who could analyze a situation with
little or no emotion. That’s how I viewed myself. In retrospect, the qualities that helped me
succeed on Capitol Hill were probably developed partly as a result of the abuse and how it
shaped me.”
In many ways, Loskarn identifies both with his abuse (“in my mind I instigated and enjoyed” it)
and with the person his abuse made him: “controlled, independent, and rational.” He writes,
speaking in the present tense, that “in my heart I still struggle to see my five-year-old self as a
victim.” He does not seek exculpation on the basis of his own rejection of the person that he is,
but rather on the basis of the fact that his abuse made him into the person that he is. For it
created a past which the images then “matched.”
Loskarn says he ‘felt a connection’ to child pornography, and he does not present himself as
safely alienated from this connection. He does not describe his desire as a compulsion which he
can evaluatively condemn as in any sense ‘external.’ He is unlike Frankfurt’s unwilling addict,
whose will retreats to the safety of second-order rectitude. Loskarn cannot seem make a clean
break from the negative features of himself—and this is because those negative aspects run very
deep.
Reflective self-appraisal has a blind spot: value. It is possible to endorse or not endorse virtually
any feature of oneself except one’s values. One can critically, detachedly assess one’s own
physical appearance, or one’s desires, skills, proclivities, etc. But I must assess whatever I assess
using values, and any values I have will be among the set I use for this purpose. Suppose I
approve of my thin physique from a desire to be attractive and a belief that thinness makes one
attractive, and suppose that I endorse this desire because I value the status and privileges
afforded to attractive women. How do I endorse this value? The theorist of self-endorsement
wants to put his foot down and end the regress at this point, by exempting values themselves
from the endorsement requirement10. He doesn’t see any need to make a space between a person
and her values; or perhaps he thinks, wrongly, that this can be done by invoking conflicts of
value.
Watson (1975, p.216) expresses our reflective attachment to our values too weakly when he says
that “the important feature of one’s evaluational system is that one cannot coherently dissociate
oneself from it in its entirety.” For surely it is also impossible to dissociate oneself from one’s
motivational system in its entirety. The point is: one cannot, via the procedure of rational
10
See Smith (p.251) and Watson (1975). Noggle complains (p.91) that Watson allows himself, but denies
Frankfurt, recourse to thus ending the regress by fiat. And yet Noggle himself makes the same move, offering a
historical account that simply stops at an ‘initial self’: “there is a bottom turtle that must stand on something that is
not itself a turtle.” (p.97)
5
reflection under discussion by Watson, dissociate oneself from one’s values at all, at least insofar
as they remain one’s values. Watson says “One can dissociate oneself from one set of ends and
principles only from the standpoint of another such set that one does not disclaim.” (p.216) But
when, for instance, two values interfere11 with one another’s realization, this doesn’t give the
agent grounds to withdraw her endorsement of either value. Wishing the world were such that
both values were realizable makes one regret that one must privilege the one over the other. But
such regret reflects the fact that one has not withdrawn endorsement from the other value. If
there is some “set of ends or principles” that I do not endorse, but is still in some sense, mine,
this can only be because it is represented in my “motivational system.” But such ‘ends and
principles’ are not values at all. They have devolved into the status of desires, impulses,
compulsions, or habit. Setting aside the diachronic question of whether I can change my
values—we will turn to this shortly—not a single one of my values is, insofar as it is my value at
all, a possible target for synchronic reflective endorsement or rejection.
I am not, I think, making exactly the same point about Loskarn that he is making about himself.
Loskarn seeks exculpation for the kind of person he has become on the basis of his past. He sees
himself as having been made into the person that he is. Loskarn is, effectively, likening himself
to Susan Wolf’s (1987) JoJo, who was indoctrinated into evil values at the hands of his sadistic
dictator father, Jo. JoJo doesn’t seem as responsible for having those values as his father, Jo,
who arrived at them without having been himself indoctrinated. Wolf seems right to observe that
this kind of case raises worries for the self-endorsement account12. But there is an additional,
deeper, worry for the self-endorsement account that is raised by the case of Loskarn—who is, so
to speak, a JoJo who knows himself to be a JoJo.
JoJo is perfectly happy with his corrupt evaluative system. Loskarn was so unhappy with his that
he was moved to take his own life. And yet Loskarn is no more capable than JoJo of clearly
assessing his own corruption from a reflexive distance. When he speaks of a continuing struggle
not to see himself as having invited the abuse, we know he cannot fully grasp the error in his
own way of thinking. And this is exactly why he hates himself. Loskarn expresses gratitude to
those who have offered compassion, but also notes that “compassion is harder to accept than
condemnation when you feel as disgusting and horrible as I do.” The theorist of reflective
endorsement is inclined to present us with the following options: either someone rejects the
problematic part of himself via reflective evaluation by the unproblematic part, or his problem
escapes his notice. But Loskarn shows us that these are not the only possibilities. The model of
reflective-self evaluation may provide a good framework for understanding akrasia, addiction, or
11
Here I imagine a contingent conflict between two values, such as the would-be resistance fighter from Sartre who
must choose between mother and motherland. There is a different, special kind of value conflict in which the values
conflict intrinsically. In this kind of case, appreciating the one value actually gets in the way of appreciating the
other. But this is precisely because one cannot get the value properly into view, not because one does so and yet still
withdraws endorsement. (I discuss this in “Internal and External Conflicts of Desire.”)
12
Though Wolf (1987) draws a different positive conclusions from the example than I do, contending that JoJo is
not responsible because his values are insane. But sanity does not seem like a good criterion for distinguishing
between Jo and JoJo, since they have the same values. And she does suggest that JoJo’s exculpation is tied to his
upbringing: “it is unclear whether anyone with a childhood such as his could have developed into anything but the
twisted and perverse sort of person that he has become.” (p.54) She could, perhaps, deny that Jo and JoJo’s values
are equally insane, but only by presupposing a systematic connection between the process by which one arrives at a
normative outlook and the manner in which one holds the outlook thus arrived at. And this is just what the
diachronic account seek to establish.
6
psychological compulsion, but that list captures only a relatively shallow subset of the ways in
which someone can be disillusioned from, rejecting of, or unhappy with himself.
I want, now, to turn to a diachronic account. It is obvious why such an account should help us
emphasize the moral relevance of an upbringing like JoJo’s, or a history of abuse like Loskarn’s,
to someone’s responsibility for his character. It is less obvious, but also true, that a diachronic
account will help us articulate the distinctive form of dissatisfaction with himself that Loskarn
evinces in his note. What I call the ‘self-cultivation’ view has been adopted by many
philosophers who have sought a more historically sensitive self-referential13 account. Such
theorists are attentive to the question of whether someone’s valuations are themselves the
product of rational cultivation. On the self-cultivation view, the responsibility I currently have
for myself is derivative of the agency of my earlier incarnation. I am responsible for being who I
am because my earlier self chose the self I have now, and I make my future self responsible for
who she is through the choices I make now. Responsibility, on such a picture, is not a matter of
endorsing one’s current self but of voluntaristically shaping one’s future self.
But is that possible? It is undoubtedly possible to cultivate a certain physical appearance or a
friendships, a desires, a hobby. But the difficult case is, once again, the cultivation of the state of
valuation from which the agent will selects this physical appearance, that friend as a friend,
rejecting this desire in favor of that one. Can someone create her own values? Galen Strawson
has argued that, because we cannot create our own valuational condition, we cannot be
responsible for anything we do from it14. I want, once again, to set aside the question about
responsibility for action, and look only at the portion of his argument attacking value creation.
The basic dilemma is as follows: the new values, creation of which constitutes my act of selfcreation, must either be continuous or discontinuous with the ones I already have. If they are
continuous, my new values are entailed by my old ones. In this case, I don’t really change. If
they are discontinuous, the new values contradict or come at a tangent to my old values. In this
case, they arise accidentally or through external influence, rather than through my agency. I
change, but I do not change myself. Let us see how this dilemma plays out in a schematic case.
13
Many accounts of responsibility for self contain a historical element in only a negative form, e.g. Dworkin (p.2526), Elster (p.227-8). On these views, a desire or preference is autonomous (with respect to its origins) if it is not
formed in the wrong way, as a result of coercion, manipulation, implantation, and other such forms of interference.
Valdman raises a serious objection (pp.419-420) to such views: it seems that the normal course of human
development presents us with forms of interference (parental, genetic, cultural) that are not in principle different
from the coercion, implantation, and manipulation taken by such theorists to be defeaters of autonomy. Valdman
points us to the need to offer a positive historical self-relation as grounding responsibility for oneself, and offers up
the self-cultivation account: “What is involved in engaging with a desire in a way that makes it one’s own? Very
roughly, this involves evaluating it, deciding whether it is worth maintaining and worth acting on, and making this
decision effective.” “On my view, D is autonomous only if its bearer decided to grant D its special status – only if
she made D her own. And this can happen during D’s formation or long after.” (p.423) Philosophers are quite
generally inclined to present the self as developing by cultivating itself in this way (see fn. 16), even if they do not,
as Valdman does, always cite such self-cultivation as ground for responsibility for self.
14
Since he does not always clearly separate the arguments for these two claims, what I offer here is not a direct
paraphrase of any passage of his paper. Nonetheless, both style and the substance of this argument is heavily
indebted to Strawson’s dilemmatic way of arguing against moral responsibility.
7
Suppose that I value X, and I come to value Y because I see that Y is the means to X (or more
generally, that valuing X rationally requires me to value Y). Is this ‘self-creation’? No:
instrumental reasoning is not self-creation; it is a matter of working out the consequences of the
value condition I already have, rather than acquiring a new one. But we might suppose that,
prior to realizing the relationship between X and Y, I had actively disvalued Y. In deciding to
conquer my hatred of Y, I am, let us suppose, deciding to change in a substantive way. But in
that case, the mere discovery that valuing X rationally required me to value Y would be
insufficient to explain the change. For why didn’t I abandon X instead of embracing Y? Perhaps
I have other values that dictate this decision, or perhaps I simply valued X more than I disvalued
Y. In either case, we are once again back on the first horn: this isn’t a change in value but
merely an extension or making-consistent of the values I already have. If I neither valued X more
than disvaluing Y nor had other values dictating the overall importance of X, then we have not
yet explained my rationale for adopting Y. I may have done so randomly, or on a whim, but
such arbitrariness is not a proper basis for responsibility. Strawson makes this observation when
considering the libertarian response that character is the result of causally undetermined efforts
of will. He asks, “How can the occurrence of partly random or indeterministic events contribute
in any way to one's being truly morally responsible either for one's actions or for one's
character?” (p.18) This point applies equally to cases in which one’s efforts of self-change are
made for no reason, or arbitrarily. As Strawson points out, someone who succeeds as a result of
such an event is “merely lucky.”
At each point in the story of my choice regarding X and Y, it seems that we can characterize the
moves I make either as continuous with my previous values (no change), or as haphazard (no
responsibility). This is the dilemma that Strawson poses for the very idea of self-cultivation.
We can illustrate the force of the paradox by seeing how hard it is to construct a case of selfcultivation that avoids it. Suppose that there exists a pill which will produce unshakeable faith in
the existence of God. An atheist might see fit to take the pill if, for instance, he learns that
believers live longer than atheists. Taking the pill out of a desire to prolong his life, he arrives as
a result of his own rational choice at a fundamentally different set of values. Does such selfmanipulation constitute a solution to the paradox? No. The self-manipulator has broken down
self-creation into two steps: a first step in which the agent decides she wants to be a theist, and
the second step in which she becomes one. The first step succumbs to the ‘no change’ horn of
the dilemma, and the second to the ‘external influence’ horn. The result is that we do not have
‘self-creation’ in the sense we are looking for it: to the extent that he’s responsible he’s not
changing, and to the extent that he’s changing he’s not responsible.
It is, perhaps worth focusing some extra attention on the first horn, since proponents of selfcultivation are more likely to grab it. Why, it might be wondered, does someone’s embrace of a
value fail to count as “change” if the value was derivable from her old values? The derivation
might have required time and effort, so that there is at least some sense in which the agent who
has performed the derivation differs from the person she was before15. We ought to allow, for
instance, that coming to value something that is instrumentally conducive to ends one had all
along might occasions observably significant changes in acting, thinking and feeling. What
15
This will be the line pushed by those, such as Ekstrom, who understand self-creation as a drive towards
coherence.
8
stands in the way of classifying those changes as “real” changes? The answer is: the regress.
The problem with allowing that such an agent develops her new self by unraveling the
implications of the materials already present in her old one is that this simply pushes the question
of self-creation back one step. For we then want to know how she acquired those materials. Her
valuation of the means is a new expression of a value that she acquired whenever she came to
value that end. The time when the value was “new” must have been some time prior, and we
have yet to tell the story of that genesis.
Self-cultivation seems structurally designed to push the problem of self-creation backwards,
locating an agent’s responsibility for who she is at a previous time of self-making. The
endorsement view had a similar problem, stumbling on the question of how the endorser gets
endorsed.
S1
Endorser Self
Cultivating Self
—
—
S2
Endorsed Self
Cultivated Self
In each case, we were inquiring after S2’s responsibility for himself, and the theorist directed us
to S1. But now we want to ask the same question again. If S1 is responsible for the way S2 is,
who is responsible for the way S1 is? Reflection on the case of the theism pill reveals that
deriving S2’s responsibility from S1 may be the source of the problem.
It seems that the pre-pill agent (call him P1) is responsible for the post-pill agent’s (P2) condition
of belief in God: P1 deemed that condition instrumentally valuable, and took measures to bring it
about. Nonetheless P2 does not seem responsible for his own condition. If we thought it was a
mistake for the atheist to come to believe in God in this way, we might get angry at him, before
he takes the pill, for what he is about to do. Our anger pre-emptively attributes the bad results of
pill-taking taking to P1, now. Once he takes the pill, however, our anger dissolves. P2 will
claim that his belief in God reflects the evident truth of God’s existence, and is only incidentally
connected to the pill. Insofar as P2 honestly believes this, and we honestly believe that his belief
is a product of the pill, we cannot hold him (now) responsible for being as he is.
The dissolution of our anger should not be confused with the frustration of our desire to stop him
taking the pill. For if the pill, once ingested, takes a few hours to take effect, we will still be
angry during those hours. We will be angry at P1 for what he is doing to himself, even though
neither we nor he can change it. Moreover, if it emerges, after he has taken the pill, that there is
an antidote, we will not be angry at P2 for not taking it. We may try to get him to take the
antidote, perhaps even resorting to trickery, but we will not think that it is ‘his fault’ that he is the
way he is. It is P1’s, and not P2’s, fault that P2 is the way he is now.
It seems to be a general problem with self-referential views that they hold S2 responsible for S2
only in a derivative way, via S1. For the pill case shows that it does not follow from the fact that
S1 is responsible for S2’s being the way S2 is that S2 is responsible for being the way S2 is.
Such shunting of responsibility is not, however, a necessary feature of any self-referential view.
We needn’t assume that S2 is responsible for himself only because S1 is in the first place
responsible for S2. S1 might cause or make or bring about S2 and yet it might be true that S2 is,
9
in the first instance, responsible for himself. I am suggesting a reversal of the standard picture
of the conceptual priority relation between S1 and S2.
The standard (and, I grant, natural) way of thinking about this relation is to understand S1 as
conceptually prior to S2. We find a particularly evocative expression of this standard picture in
Joseph Raz’s use of promising as an analogy for self-creation. Raz describes S1, the creator self,
as having done something akin to making a promise, and S2, the created self, as therefore having
a new reason for action that is akin to one’s reason to keep a promise. Raz says that in “a
creation of new values and reasons…our past forms the reasons which apply to us at present… it
is like the change occasioned by promising: a creation, in that case, of a duty one did not have
before.” (Morality of freedom ch. 14 p.387) In Razian self-creation, the earlier self holds the
reins, creating values for the later self.
Likewise, Robert Noggle (fn. 26) presupposes a “basic picture of the self,” one “quite common
among philosophers,”16 on which it “evolves according to its own internal logic—its own
contents determine whether an how it is to change in response to new information, internal
conflicts and changing condition….when psychological changes happen in this way, it seems
correct to say that the new configuration is an authentic continuation of the previous
configuration. On the other hand, a psychological change—especially a change in the core
attitudes—that does not occur in this way produces a new configuration that is not an authentic
continuation of the previous one…If the changes are radical enough, it might be proper to speak
of the destruction of one self and its replacement by a new one.” (pp.100-101)
As examples of changes in the second category, Noggle offers “sudden organic trauma” and
“nefarious brain surgery.” He contrasts these with changes that are “internally motivated in such
a way that they seem to be intelligible reflections of the contents of the core attitudes.” (p.100,
emphasis original) Noggle’s division recapitulates Strawson’s dilemma: he sees changes in a self
as either rational extrapolations of previous conditions (“according to its own inner logic”) or
traumatic intrusions. But we do not have to choose between a self that unravels its inner
resources and a self that is changed from the outside.
A promisor must look backwards to find her reasons. To avoid regress, we want the self looking
in the other direction: instead of imagining my future self as beholden to its past self, I suggest
we imagine my past self as looking forward, trying to live up the person she hopes to become.
S1 doesn’t make a promise, she sees (to take up another facet of the concept of a promise) a
16
See fn. 16 and accompanying text. I trust the view is familiar enough that I do not need to reproduce his list, but I
will add to it by singling out Michael Bratman’s especially sensitive recent exposition of the standard view. Bratman
goes so far as to gesture at the limits of such a view by highlighting what he calls the problem of the
underdetermination of the self by its values. But he does not consider that such an underdetermined condition might
indicate a self-in-transition, since he takes for granted that even the undetermined agent relates to herself via some
kind of a “reflexive self-governing policy.” (p.175) Elijah Millgram tries to present an alternative picture in his
Practical Induction. Millgram’s conception of ‘learning from experience’ purports to pave the way for the self to
respond to more than what Bernard Williams called ‘internal reasons.’ But does ‘practical induction’ really represent
an alternative to Williams’ “instrumentalism” (as Millgram calls it)? The worry that “practical” induction can be
reduced to theoretical induction plus instrumentalism is, I believe, substantiated by the two-stage character of
Millgram’s account: an agent may form judgments via practical induction but must act on them via standard (noninductive) practical reasoning. If practical induction were a genuinely practical mode of reasoning, it would
describe a form of agency and not only a form of thought.
10
promise of a better self. When we speak of some prospect as a promising one, we do not use the
word ‘promise’ literally, since, among other reasons, what doesn’t exist yet cannot make
promises. But the locution is suggestive, and what it suggests a reversal of the Razian order.
Promising presupposes a certain stability and predictability in one’s self and one’s
circumstances; it is only S2, and not S1, who would be in a position to take on such
commitments. It is not S1’s place to embrace goals on behalf of, form reasons for, or create
duties that will bind S2. S1 is, by her own reckoning, not as she should be, she is in transition, a
work in progress.
Theorists of self-endorsement identify an agent’s true self—the self for which she is
responsible—as some subset of what we would ordinarily call her whole self. The idea that
someone’s self might be in some respect untrue is a deep insight that I want to preserve from that
view. I also want to grant that one of the ways in which such untruth manifests itself is the
existence of desires, habits or compulsions from which the true self withdraws its approval. But
untruth in one’s self can run deeper than that. If you are trying to change yourself, then you must
be unhappy with the way you are. This phenomenon cannot always be resolved cleanly into a
right-thinking “evaluator self” and those errant motivations from which it would be freed. The
untrue self can be the self-in-progress. Thus it is Loskarn who unhappily experiences his whole
present self as in some way untrue.
When S1 is the self-in-progress, there is no full accounting for her as she presently stands. She is
best understood derivatively, as something on her way to S2. If we force ourselves to produce a
determinate description of S1, we will have the analogue of a freeze frame of her at a particular
moment of time. What is left out of such a picture is the crucial fact that she is in motion.
Visually, we might better represent S1 as a blurry image. S2 is what she really looks like—her
true self—when things eventually come into focus. S1 is not the author or master of S2, she is
the prototype of or rough sketch for S2.
I have been developing, in very abstract terms, the formula for avoiding regress: we need to
understand the self that is to be responsible for itself as conceptually and explanatorily prior—
though temporally and evaluatively posterior—to the self generating or endorsing it. The throwweight of the process of self-creation is not the self at the beginning, but the one at the end. The
created self is not responsible for itself because of how it is created. Rather, it was created in
that way because it is the true self, the one that can be responsible for itself, the one in a position
to make promises. Now I want to try to flesh out this formula into a recognizable human
phenomenon.
II. Aspiration
It sometimes happens that one is suddenly—unprompted by a pill—confronted with a value
one’s prior valuations did not lead one to anticipate. One might experience such a
transformation at a musical performance, or in a church, or among a group of people talking in
an enticing way about something they value. But such experiences are ‘transformative’ only in
the inchoative sense. They will not bear the weight that the theorist of self-cultivation wants to
put on them, if she tries to see them as containing the resources from which the endpoint can be
rationally extrapolated. The transformed agent will look back, years later, and say of that initial
11
event ‘little did I know, at that time, what was really valuable about music,’ (or religion, or
whatever). That value is one she knows now as a result of the work she has done to come to
know it, work whose target she could not exactly foresee as she was working towards it. I want
to offer an account of self-creation in which that work figures centrally.
Let us consider the genesis of a foodie. No one starts off life with an interest in or expertise
about the technical details of cooking methodology, kinds of cookware, specialty ingredients, the
local restaurant scene, food journalism, etc. So how does someone become a foodie? Does she
derive the necessity of valuing food expertise from some commitments she already has? Or does
she just find herself, one day, with this new interest? These are the two choices Strawson’s
paradox offers us, and neither offers a recognizable story. Given the expertise and work
involved, it is implausible that anything but the earliest stages of such a transformation can be
explained by accident. For instance: the fact that someone found himself, for incidental reasons,
in the exceptional gastronomic environment of Osaka, Japan might be the beginning of the story.
Those experiences could ignite a spark of interest, but then something more would be needed to
drive someone’s systematic development of that initial spark into a full-fledged passion. The
‘something more’ in question is unlikely to be a value to which she was antecedently committed,
from which a passionate interest in culinary excellence could be derived. I suppose a committed
hedonist might turn to food with the aim of branching out into a new arena of sensible pleasure,
but this is not how most of us end up with subscriptions to Cook’s Illustrated.
I don’t want to deny that there might be cases in which an accidental and transformative initial
experience or an anterior value commitment suffices to explain someone’s transition to
becoming a foodie. But most of the foodies you know will not identify with either of those
narratives. Most of us embark on culinary (and other) adventures without thinking that we
know, in advance, exactly what we are to get out of them. But that is not to say that we take them
up on a whim or for no reason or by accident. There seems to be an intermediate possibility that
Strawson’s dilemma has directed us away from recognizing, the possibility that someone has an
inkling of a value he does not fully grasp. He doesn’t have a fully worked out sense of how this
value fits into the rest of his values, because he doesn’t have a fully worked out sense of what
this value is. How could he, if the value corresponds to the intrinsic pleasures of the fine
discriminations he is not yet capable of making? The point is a quite general one. Most of the
profoundly important activities, relationships, and forms of knowledge that human beings pursue
are such that one can appreciate them fully only from the inside, by being well-acquainted with
them. And our question concerns (the rationality of) the process of becoming acquainted. If you
had to acquire values either by accident or by working through the entailments of your prior
commitments, there would be no such process. You would either already be at your valuedestination or have no way to get started. But those don’t seem to be our real choices.
The way in which people stand towards many of the values that they do not fully appreciate is
that they partly appreciate them. And with respect to some of these partly-appreciated values,
they also have the inclination to appreciate them more; they have a sense that their inchoate
appreciation is incomplete. They are, as it were, non-reflectively aware of something being
wrong with their values. Someone who seeks a better grasp of a value is an aspirant. Is
aspiration a rational activity? It is, but the reasons in question are of a special kind.
12
I am building here on work I do in another paper (“Proleptic Reasons”), where I introduce a
distinctive kind of reason for action that I call proleptic. It will be useful here to summarize the
central contention of that paper, since the aspirational account of self-creation I am offering must
invoke proleptic reasons. The agent familiar to us in the philosophical literature in moral
psychology is one who initiates her action only after having settled what value is to be achieved
in so acting. She knows exactly why she is doing what she’s doing. We might think that, to the
extent that someone doesn’t know why she is -ing, her -ing is to that extent irrational. I argue
that -ing without a complete grasp of the value to be achieved by -ing is not necessarily a
manifestation of irrationality, because it is sometimes a manifestation of learning. Some rational
agents have only inchoate, anticipatory and indirect grasp of a good they are trying to know
better.
If you are trying to get better acquainted with some value, then you take your antecedent
conception of that value to be inadequate. You act in order to grasp the value better, but your
reason for wanting to grasp that value must be the very value you don’t yet fully appreciate.
Someone who takes a music appreciation class in order to secure a future spouse or an easy A is
not really there to learn. If he has an ulterior motive for taking the class, he does not act on a
proleptic reason. The proleptic reasoner is there for the music he doesn’t yet know how to enjoy.
Most big life decisions are like this. They find the agent contemplating some obscure value from
a great distance. We can’t comprehend the value of child-raising for us, let alone the value of the
life of the child we will raise, before starting a family. Likewise, we go to college for the
education college will teach us to appreciate; we leave our familiar hometown for the
unfathomable big city; we date in order to love, and get married in order to love in a new way;
we choose a career because of the as yet unfamiliar joys of expertly doing the work in question.
In pursuing these values, our attitude is not merely that of hope that we will one day come to
appreciate them. We work to appreciate them, and this work is rationalized by the (proleptic)
appreciation of the value we are coming to know.
In non-proleptic (standard) practical reasoning, the agent looks to the world for a match for her
antecedently fixed goal. For instance, she goes to the store in order to get milk for her cereal. If
the store is closed, she will be disappointed—and she knows this in advance. Which is to say,
she knows in advance what will constitute the success of her action, and what failure. The
proleptic reasoner, by contrast, does not demand that the end result match a schema she currently
claims in her possession. For she wouldn’t know what to ask for: she hopes, eventually, to get
more out of what she is doing than she can yet conceive of. Proleptic rationality illustrates the
possibility that we might reason in order to improve the very conception of the good driving our
reasoning. This conception of valuational growth is elusive on both of the views of
responsibility for self I have considered.
It is on this point that we should contrast aspiration with ambition. Both the aspirant and the
ambitious person seek to become better than they currently are, but the ambitious person takes
herself to have a fully adequate grasp on what ‘better’ consists in. She does not seek to improve
her conception of the good, only to satisfy it. Hence ambition is a common attitude towards
those goods—wealth, power, fame—that can be, or can be taken to be, fully appreciated even by
those who do not have them. The characteristic danger of ambition is that of becoming trapped
13
in the impoverished appreciation of a value available to someone who doesn’t have (much)
access to it. Turning ambition into aspiration is one of the job descriptions of any teacher, but it
belongs especially to those of us who, as philosophers, carry the Socratic torch. Socrates may
not have had great success with, e.g., Alcibiades, Thrasymachus, and Callicles, but there is no
question that he was trying to transform their ambitious political strivings into an aspiration to
virtue. Aspirants make good learners because, quite simply, they take themselves to have
something to learn.
It has become a commonplace17 to describe the developing self using the image of Neurath’s
boat, which must be rebuilt as it sails. The fact that the boat never docks is, I grant, an elegant
way of expressing the unsheddability of character. One cannot ‘operate’ on one’s self from the
outside, as a doctor might operate on his own anaesthetized leg. But it does not follow that one’s
self must stay true to its initial form, or else be traumatically violated from without. Neurath’s
boat isn’t Theseus’ ship, whose job is to stay the same through all its many changes. On
Theseus’ ship, a rotting plank must be replaced with a plank of a similar shape and size. On
Neurath’s boat, the only requirement is that it must be replaced by sailors on board the boat.
Neurath’s rowboat might be gradually rebuilt into a trireme, or, for that matter, an airplane,
without the builder ever setting foot onto dry land.
We can do more than maintain ourselves: we can build ourselves up into something better.
It may be worth spelling out some of the ways in which this activity of self-building manifests.
We become spectators, take classes, do exercises. We find mentors to emulate or fellow
travelers with whom to commiserate—and compete. We do the same thing over and over and
over again until we get it right, without knowing in advance what ‘right’ is. We do work we
don’t always enjoy, and we pretend—even to ourselves—that we enjoy it. We leave ourselves
open to certain kinds of experiences and closed to others, knowingly risking disappointment and
disillusionment down the line. We alert ourselves to and steel ourselves against temptations to
abandon course in favor of a more readily available and more immediately intelligible form of
value. Candy, television, alcohol, a nap, video games, internet surfing—pick your poison, it’s
waiting in the wings. We struggle against implicit or explicit messages, from individuals or
groups of individuals, to the effect that this kind of value is “not for you.” Often these struggles
are heightened by the fact that we have internalized the judgments in question. The work that we
are engaged in is the work of bringing something into view. But because what we are bringing
into view is something practical—a value—the work is a matter of acting, and feeling, as well as
thinking.
Privilege aggravates blame for failure because it facilitates value-acquisition. The conditions in
which one is raised determine how hard one has to work to see the value of, say, acquiring an
education, love, or civic engagement. When those who have been disenfranchised, whether de
jure or de facto, find their way to politics, we’re rightly impressed. When a victim takes on the
role of speaking out for other victims, we note how much it took to stand up, assert himself, and
take thought for others. And when a victim does not manage to transcend the strong forces
getting in the way of striking out on his own evaluative path, we have sympathy. Ryan
Loskarn’s enjoyment of the sexual exploitation of children is not a form of valuing he could
easily have gotten himself out of. Perhaps JoJo could have escaped the clutches of Jo’s ideology,
17
See Noggle fn. 26, McDowell pp.36-7.
14
but it would have been a long and hard road to walk. Whereas Jo—supposing, for the moment,
that he was raised in an ordinary, sane, loving home—had available to him the seeds of a
respectable moral sensibility. His failure to develop18 one was his own doing.
The story of Loskarn is sadder than the story of JoJo, and this is not only because Loskarn is real.
It is because Loskarn wasn’t trapped in his valuational error. When Loskarn’s offense initially
became known, people were outraged and angry. We reacted to Loskarn as we might react to Jo.
The suicide and the subsequent publication of the note diminished our moral outrage, softening it
into pity. We came to see Loskarn as a victim, more like JoJo than Jo. But when we read
Loskarn’s note with attention not only to the events described in it, but to the perspective from
which it is written, we glimpse the person he could have become. And we glimpse this because
Loskarn also glimpsed it. He sees his childhood self as having invited abuse, but he also, as it
were, sees another, future himself who doesn’t see things that way. He speaks of wanting to
“begin the process of trying to sort this out and fix myself.” I can only guess, with some
sympathy, that this aspirational task struck him, at a low moment, as simply unmanageable.
Works Cited
Berofsky, B. 1995: Liberation from self: A Theory of Personal Autonomy, Cambridge University
Press,
Bratman, M. 2003: “Autonomy and Heirarchy,” Social Philosophy and Policy.
Buss, S. 2012: “Self Determination in the Passive Mode,” Ethics.
Callard, A. “Proleptic Reasons,” under review.
_______ “Intrinsic and Extrinsic Conflicts of Desire,” under review.
Doris, J. 2002: Lack of Character, Cambridge University Press.
Dworkin, G. 1976: “Autonomy and behavior control,” The Hastings Center Report.
Ekstrom, L. 1993: “A Coherence Theory of Autonomy,” Philosophy and Phenomenological
Research.
Elster, J. 1982: “Sour Grapes: Utilitarianism and the Genesis of Wants,” in Amartya Sen &
Bernard Williams (eds.), Utilitarianism and Beyond. Cambridge University Press.
18
This phrase alludes to a point which I cannot develop here: many aspirational failures are not commissive but
omissive. A failure to be moral is typically not a result of an aspiration to be immoral, but a culpable failure to
aspire to be moral. Failure to aspire is very different from the aspiration to rid oneself of a value. There is
culpability-by-omission, but no creditability-by-omission. This suggests a deep asymmetry between success cases
and failure cases, one that I hope to explore in future work.
15
Fischer, J. 2006: “The Cards that Are Dealt You,” Ethics.
Frankfurt, H. 1971: “Freedom of the Will and the Concept of a Person,” Journal of Philosophy.
_________ 1988: “Identification and Wholeheartedness,” in his The Importance of What We
Care About (Cambridge University Press).
Garnett, M. 2013: “The Autonomous Life: A Pure Social View,” Australasian Journal of
Philosophy.
Harman, G. 2003: “No Character or Personality,” Business Ethics Quarterly.
Kamtekar, R. 2004: “Situationism and Virtue Ethics on the Content of Our Character,” Ethics.
Korsgaard, C. 1996: The Sources of Normativity, Cambridge University Press.
Leiter, B. 2011: “The Paradox of Fatalism and Self-Creation in Nietzsche,” in John Richardson
and Brian Leiter, eds., Nietzsche, Oxford University Press.
Marcus, R. 2014: “The Tragedy of Ryan Loskarn,” The Washington Post, Feb. 16th.
McDowell, 1998: “Some Issues in Aristotle’s Moral Psychology,” in his Mind Value and Reality
Harvard University Press.
Millgram, E. 1997: Practical Induction, Harvard University Press.
Nietzsche, F. 1966: Beyond Good and Evil, trans. Walter Kaufmann, Random House.
Noggle, R. 2005: “Autonomy and the Paradox of Self-Creation” in James Taylor, ed., Personal
Autonomy, Cambridge University Press.
Owen, D. and Ridley, A. 2003: “On Fate,” International Studies in Philosophy.
Raz, J. 1988: The Morality of Freedom, Oxford University Press.
Setiya, K. 2007: Reasons without Rationalism, Princeton University Press.
Smith, A. 2005: “Responsibility for Attitudes: Activity and Passivity in Mental Life,”
Ethics.
Valdman, M. 2011: “Autonomy History and the Origins of our Desires,” Journal of Moral
Philosophy.
Watson, G. 1975: “Free Agency,” The Journal of Philosophy.
16
_______1987: “Responsibility and the Limits of Evil: Variations on a Strawsonian theme” in
John Martin Fischer, ed., Perspectives on Moral Responsibility.
Williams, B. 1981: “Internal and External Reasons”, in his Moral Luck (Cambridge University
Press).
Wolf, S. 1987: “Sanity and the Metaphysics of Responsibility,” in Ferdinand Schoeman, ed.,
Responsibility, Character, and the Emotions, Cambridge University Press.
_______: “Character and Responsibility,” forthcoming in The Journal of Philosophy.
17