The aims of psychotherapy vary with who is being helped: a child, an adult, or a family. The aims vary with different problems: depression, post-traumatic stress, anxiety, compulsions, and so forth. Further variety comes from the different schools and theories behind the approach. Underneath all this is a common aim: release from distress by removing psychological obstacles to a good life.

This vague aim usually takes three more specific forms. One is helping people see and escape from cognitive and emotional traps. Another is helping people see how their relationships with each other may be working badly, and to find ways of untangling them. A third is trying to help people discover things about themselves and—if they so decide—to help them turn self-understanding into self-creation.

Release from Cognitive and Emotional Traps

Stressful situations, including such life events as divorce or losing a job, are powerful triggers of depression. Clinically depressed people often underrate the causal role of these triggers. Instead they may blame their own supposed inadequacies. In depression this is one way that interacting beliefs and emotions cause a downward spiral. Being depressed supports a pessimistic bias in thinking about the future. This also gives people a very poor view of themselves, making them feel hopeless. In this circular trap, depression distorts beliefs, which in turn maintain depression.1

Evidence supports therapy as one of the more effective treatments. A UK Department of Health report found that both cognitive behavioral therapy and interpersonal therapy are effective, and that there are indications of possible benefit in some other types, including psychodynamic therapy.² Paul Biegler has made a strong case for the view that cognitive and emotional traps found in depression give good reasons to expect this. Spinoza’s two perspectives on the same state are important here. Only medication directly attacks the chemical side of depression. Only therapy directly attacks the cognitive side. Seeing depression simply as a chemical problem obscures the role of stress. Therapy can help people escape the trap through seeing both the causal role of stress and the distorting effects of depression.

Relationships, Families, and Therapy

Family therapists could reasonably sigh at Tolstoy’s famously confident assertion that all happy families are alike, while an unhappy family is unhappy in its own way. Questions can be asked about both claims, but family therapists have a stake in challenging the second claim. Part of their expertise rests on experience of what has been useful in understanding other families’ problems. Lack of any common structure would make their experience useless.

Therapists look for particular patterns. Families may be trapped in vicious circles. They may communicate poorly. They may have unrealistic mutual expectations, or stereotype each other. There may be too much criticism and intolerance. They may blame problems too much on certain members. Some of them may have low self-esteem.

How real are the claimed patterns? Are they present in the families or only in the belief systems of therapists? They are obviously plausible causes of family stress and disruption. But what evidence is needed to justify overriding how families see their own problems? What are the causal theories family therapists draw on? How good is the evidence for them? And how is it decided which one fits a particular family? When these patterns are involved, how should therapists respond? Some answers are shaped by therapists’ theories, notably about families as systems.

Families Seen as “Systems”

The historical origins of the systems approach are important. It started in the mid-twentieth century with the general systems theory of Ludwig von Bertalanffy and with the “frames and paradoxes” of the anthropologist Gregory Bateson. Both of these influenced Mara Selvini Palazzoli and the still-influential Milan school of family therapy.3

Bertalanffy’s general systems theory was “a theory, not of systems of a more or less special kind, but of universal principles applying to systems in general.”⁴ A “system” can be anything with interacting components: the central heating, cells in a living organism, a family, an army, the international financial system. It is hard to see that a general theory of systems can have much specific content. Is it more useful than a “general theory of events,” meant to fit anything from the melting of an ice cap to the French Revolution or a child patting a dog?

Bertalanffy applied his approach to psychology: “Any system as an entity which can be investigated in its own right must have boundaries … Psychopathology shows the paradox that the ego boundary is at once too fluid and too rigid. Syncretic perception, animistic feeling, delusions and hallucinations, and so on, make for insecurity of the ego boundary; but within his self-created universe the schizophrenic lives ‘in a shell,’ much in the way animals live in the ‘soap bubbles’ of their organization-bound worlds.” This account of “the schizophrenic” does not suggest Bertalanffy has much familiarity with people with schizophrenia, but he concludes: “Psychiatric disturbances can be neatly defined in terms of system functions.”⁵

Gregory Bateson’s view of psychotherapy centered on frames and paradoxes. A psychological frame “is (or delimits) a class or set of messages (or meaningful actions).” Frames are “metacommunicative,” by which he meant that “the subject of the discourse is the relationship between the speakers.” He explained these abstractions with an illustration. When young monkeys nip each other in play, they signal that the aggression is not real. Humans do this too. We frame what we are saying as playing or joking.

Bateson thought that in delusions, people with schizophrenia take fantasies that are metaphors and wrongly frame them as literal truth.6 He drew on Russell’s paradox: “Is the class of classes that are not members of themselves, a member of itself ?” Either answer, yes or no, seems self-defeating. One way out, the theory of types, postulates a hierarchy of levels of language. The class of classes is on a higher level than the class of houses or the class of dogs. So the (level two) class of (level one) classes that are not members of themselves cannot be a member of itself. Finding this paradoxical is said to come from confusing different levels of discourse.

Bateson saw the monkey’s nip and the signal “only playing” as being on different levels. He thought that schizophrenic conflation of the metaphorical and the literal called for reframing. Therapy could help a person with schizophrenia to see that what appeared to be literal truth is really metaphor. Like the “different levels” approach to Russell’s paradox, it dissolves the problem by reframing it.⁷ There are limits to this view of schizophrenia. It is purely cognitive, focusing only on delusions, and only on the view that schizophrenics confuse metaphor and reality. (What is hearing hostile voices a metaphor for? How can we tell?)

Bateson used framing problems in developing his “double bind” theory. An example of a double bind is a child’s being given conflicting commands, one verbal and the other by gesture or tone of voice. This might cause schizophrenia through a failure to discriminate between frames. This interesting theory influenced R. D. Laing and others in the 1960s. But it is backed by little evidence and is ignored in current discussions.⁸

The theoretical approach of the Milan school started from Bertalanffy’s ideas about self-maintaining systems, and those of Bateson about paradox and the hidden rules governing mutual interpretation and interaction. Its classic exposition is Mara Selvini Palazzoli’s Paradox and Counter-Paradox, the stated aim of which is to test the hypothesis that “the family is a self-regulating mechanism which controls itself according to rules formed over a period of time through a process of trial and error.”⁹

The hypothesis is hardly controversial. In hundreds of ways families avoid chaos by regular patterns of behavior. On days when she goes to work, he does the shopping. At meals some families have regular places to sit; others sit randomly. They do not hang up on calls for each other, but take messages. For these patterns the family therapist John Byng-Hall has drawn on the apt theatrical metaphor of “family scripts.”¹⁰ Families were not born with all this, so the rules must have evolved over time. And errors are corrected: “There was no towel, so I dried my hands on her dress. Her response wasn’t good, so I haven’t done it again.”

The interest comes from applying to schizophrenia this emphasis on unconscious or barely conscious family rules. The central claim is: “Families in which one or more members present behaviors traditionally diagnosed as ‘pathological’ are held together by transactions and, therefore, by rules peculiar to the pathology” (italics mine).11

It is claimed that this points to a cure: “Since the symptomatic behavior is part of the transactional pattern peculiar to the system in which it occurs, the way to eliminate it is to change the rules … when we are able to discover and change one fundamental rule, pathological behavior quickly disappears.”¹² Given the evidence about genetic, early environmental, and other factors in schizophrenia,¹³ it is perhaps optimistic to hope that it can be cured by a family rule change.

The claim is that a pattern of pathological family rule systems causes schizophrenia. Each of the parental couple needs approval. Each bolsters his or her own standing by withholding approval from others. These two unpromising parents share a reluctance to admit either their mutual influence or the rules of their system. “Each member perceives open declaration as extremely dangerous. Therefore, everyone cooperates in keeping it hidden.”

The contrast is with other couples, in which each parent still wants their own view of the relationship to prevail but also accepts that it might not. The claim is that parents of schizophrenic children cannot accept this. So, although each wants to control the relationship, the only solution is not to define it. “Thus the great game begins and its secret rules are formed.” The parents “become expert in the use of paradox, taking advantage of that possibility, specific to man, to communicate simultaneously on the verbal and the non-verbal level, jumping from one logical class to a member of that class, as if they were the same thing, thus becoming acrobats in the world of the Russellian paradox.”¹⁴ “At the beginning of our work with these families, it often happened that we were taken in by the family’s game to the point that our resulting frustration and anger became transferred to the relationship between ourselves.”¹⁵

This account contains a large ratio of assertion to support. What is the evidence that parents of children with schizophrenia are those who, to an exceptional degree, both need and withhold approval? Or that they are unusually reluctant to admit their mutual influence or the existence of family rules? Why should we believe in the great game and its secret rules? Or that the parents are such good acrobats with Russellian paradox?

Some of the systems approach is overconfident speculation. But one thought about systems is important. Families are “systems” in the sense that their members interact and interpret each other. Changes in one person are likely to have reverberations on others, whose responses may in turn feed back. Therapists need to take account of these complications.

Families Across Generations

Some therapists add an extra dimension to this, seeing families as systems stretched across time. Patterns in families may have part of their origin in parents’ own past family problems, particularly with their own parents. The first line of Philip Larkin’s “This Be the Verse” is often remembered: “They fuck you up, your mum and dad.” Less remembered is the start of the second verse: “But they were fucked up in their turn / By fools in old-style hats and coats.” And no doubt those fools in their turn … the patterns may go back generations.

Sophocles, blaming the gods, noticed how disasters can cascade down the generations:

When high gods shake a house

That family is going to feel the blow

Generation after generation.

It starts like an undulation underwater,

A surge that hauls black sand up off the bottom,

Then turns itself into a tidal current

Lashing the shingle and shaking promontories.

I see the sorrows of this ancient house

Break on the inmates and keep breaking on them

Like foaming wave on wave across a strand.

They stagger to their feet and struggle on

But the gods do not relent, the living fall

Where the dead fell in their day

Generation after generation.16

Larkin and Sophocles meet here, but so do systems theorists and Freudians. It is not surprising that the tidal current across generations is strong if the father or grandfather is Oedipus.

How a family’s past is preserved in stories can be important. John Byng-Hall thought about how his own family’s stories affected him. In 1756 his ancestor, Admiral Byng, was sent to defend Minorca from a French attack. Finding his fleet heavily outnumbered, he fired token shots and avoided disaster by sailing away. He was found guilty of cowardice and was executed on the quarterdeck.

The story was passed down. Throughout his own youth, John Byng-Hall was preoccupied with worries about whether he would be a coward. In colonial Kenya there was danger from the Mau Mau and from wild animals. Would he run from a charging lion? Because he wanted to join the Navy, at the age of 12 he spent time on a naval cruiser. He remembers being terrified when standing at attention on the quarterdeck, trying to hide his fear from the ship’s second in command. He links these anxieties and fears to memories of the Admiral Byng story. He later spent time researching evidence suggesting that Admiral Byng may have been made a scapegoat for the failings of others.17

Not every family history has an Admiral Byng, so the way the remote past reached into John Byng-Hall’s own life may be very unusual. But as a therapist he reflected more generally on the effects of other families’ memories of their—usually closer—history. He suggests two false beliefs that may be harmful. One is the belief that family history has no relevance. The other is that the family members now have no choice but to repeat the patterns of the past. He suggests that these opposed beliefs are similar. Families who think they cannot avoid it, reenact the past consciously. And those who deny the relevance of the past do so unwittingly.¹⁸

What Should Be the Aims of Family Therapy?

In the literature on families as systems, the interesting points are independent of the abstract theories. Little work is done by thoughts about “systems” seen as homeostatic mechanisms. What comes over as valuable is the intuitive feel for particular people, how they see and react to each other, and how they do or do not get on. Where people are in danger of being trapped into reliving the family past, what matters is again independent of “systems.” It is the rejection of fatalism: knowing the past may be relevant but seeing the possibility of breaking free.

Families with different problems generate different aims: helping them cope with the needy and demanding grandparent now living with them; helping homophobic parents listen to their son who has come out as gay; helping the teenager to talk about previously hidden childhood traumas and helping the parents to listen.

Therapy may take the form of helping families live with conflicts or disagreements that cannot be eliminated. “Our religion tells us that homosexuality is sinful, but we do see you have to decide about your own life.” It might even help the teenager say, “I can’t give way on having to be myself, but I do see this hurts you and I wish it didn’t.” It may take the form of helping family members listen to each other and see that no single person’s narrative has the whole truth. It may involve helping people escape from bad cycles of mutual blame and see the possibility of better ways of getting on with each other.

Put so generally, these aims inevitably sound platitudinous. The controversial questions are about how to realize them. What strategies should family therapists adopt? This is discussed in some reflective papers by family therapists.19

One discussed blaming a person’s illness on an unhealthy allocation of responsibility: “Putting all the problems in some sort of externalized thing like dad’s depression or mum’s anxiety. It seems a very helpful thing for them.” Perhaps. But equally a family’s difficulties may really stem from one of them developing Alzheimer’s, schizophrenia, or depression. Obviously families will sometimes deceive themselves, using mistaken causal accounts to rationalize their problems. But there is a troubling hint of “therapist knows best” in these comments.

Katja Kurri and Jarl Wahlstrom used a couple’s therapy session to illustrate help in escaping cycles of mutual misunderstanding and blame. The account also brings out some linked problems.

One therapist invites the husband to consider whether his thoughts about his wife’s “emotionality” play a role in his agitated shouting, and asks him to consider why she reacts as she does. This implies that the husband might accept some responsibility for the situation. But later the therapist is said to emphasize the good intentions of both husband and wife: “The therapist’s statement emphasizes the reformulation of the transactional pattern as expressing good intentions on both sides. The therapist appears here to deconstruct blame … If the intentions are good but the outcome undesired, then it appears as if something like bad luck or fate is in operation. In that case none of the participants could be held responsible for the course of events. The responsibility is given to the transactional pattern, so to speak.”

As with the Milan school’s view of schizophrenia, it invites skepticism to ascribe responsibility for the relationship’s going wrong not to human agency but to fate, bad luck, or “the transactional pattern.” Yet there are cases of mutual misunderstanding despite good intentions on both sides. In such cases, exposing the good intentions may help the couple escape from mutual recrimination. But intentions are not always good. What should happen when only one of the couple is open to reasonable criticism for being selfish or insensitive?

Kurri and Wahlstrom rightly see a problem for the moral neutrality of therapists. There may be no neutral way of formulating a critical comment about only one of a couple. They ask whether therapists can be neutral between two “discourses.” The “discourse of autonomy” emphasizes a person’s right to make choices and to pursue his or her own interests. The “discourse of relationality” emphasizes emotional responsiveness to others. If, in a couple, he is strong on autonomy and she feels he is emotionally insensitive to her, what should the therapist do? Kerri and Wahlstrom say, approvingly, that the conflict was resolved by the “principle of relational autonomy,” which “indicates that identity and autonomy are generated from a matrix of relations.” The therapists’ interventions “produced a situation where the partners’ autonomy was necessarily understood as relational.”

The principle of relational autonomy seems too platitudinous and too vague to be much of a success in changing behavior. (“Yes, I agree my identity evolved in the context of relationships, but I still don’t see why in this case I should put my autonomy second to what I see as my wife’s unreasonable feelings.”) And is it really true that therapists have to decide between “reinforcing” one or other of the values or else both? Is it really for the therapist to decide on priorities for the couple, guided by some principle that somehow combines the two values? What would be wrong with asking questions designed to make the conflict of values explicit and asking whether the couple are willing to explore compromises?

The Therapist as Freud or as Socrates?

Searching through and talking about my emotions was no more than the psychiatrists wanted to do, of course, but I couldn’t, and didn’t want to, tell them. At that time, they would have started telling me my own story back; they would have made me into this character in their own version of events.

—Grace Bowman, A Shape of My Own: A Memoir of Anorexia and Recovery

Why not ask questions to make the conflict of values between husband and wife explicit? Here there is a clash between two models of therapy. With some caricature, these can be seen as the therapist as Freud and the therapist as Socrates. The therapist as Freud says, “I have special psychological knowledge—in this case knowledge of how identity is generated from a matrix of relations—and my interventions will be designed to help you understand yourselves as I do.” The therapist as Socrates says, “I don’t know the answers, but I will ask questions to help make your values explicit. If you turn out to have different values, I will ask questions designed to find out whether you can still get on with each other. Everything will depend on the answers you decide to give.”

An obvious limitation of the model of Socrates is that his questions explored only beliefs and not emotions. For therapists, Freud’s approach was a clear advance. But another of Freud’s ideas is more debatable: belief in special psychological knowledge, possessed only by initiates, giving the therapist a privileged interpretation. The Milan school’s belief that systems theory gives greater insight than the family has into “the family’s game” is in this sense thoroughly “Freudian.” This belief in the therapist’s privileged interpretation is still influential, even if the esoteric knowledge less often comes from psychoanalysis.

A deep ethical issue concerns the role of truth. Here the Freudian and Socratic models may pull in different directions. Does it matter in family therapy how far each of the different narratives is true? The therapist as Socrates may think this is worth exploring. If one of the family has paranoid fantasies about the rest, the others may want to “put the record straight,” to whatever extent that is possible. The therapist as Socrates may be willing to discuss the evidence. The therapist as Freud, who perhaps “knows” that the real problems are about systems and boundaries, might try to exclude the issue as unhelpful. Paradoxically, the Socratic approach may sometimes be more helpful, because of the importance people attach to their true history being recognized.

There are parallels to political peacemaking, as between Israelis and Palestinians. Is it better to forget the disputed past and start afresh? Or do both sides need acknowledgment of the truths in their respective narratives? It might not be enough to say, “We accept that your version is true for you, just as theirs is true for them.” This may elicit the response that this account is not just “true for us”: “We want some recognition that these events actually happened.”

In the case of the mutually blaming couple, when the therapist suggests that they both had good intentions, is this said because it is true or because it is helpful? Of course, if true, it is likely to be helpful. But supposing it is not true? Should it still be suggested, because believing it may help the couple escape the bad cycle? The therapist as Freud might say yes, claiming to understand their problem better than they do. The therapist as Socrates, not claiming a privileged view, lacks this reason for deception.

Moving the family toward a shared view, allowing escape from mutual blame, is often assumed to require the therapist’s neutrality between the rival versions. Empathy with one can be seen as siding against others. John Stancombe and Sue White report that this leads to therapists changing the subject, or changing the blame-laden account into a nonblaming version. As they point out, either response may leave the person who expressed blame feeling that the core of his or her version has gone unheard and so saying it again. Neutrality may reinforce blame.

This problem arises only because of some disingenuousness in the therapist as Freud. She has heard the expression of blame. Fearing the consequences of indicating this, she gives the impression of not having heard. Hence the evasive or off-the-point responses that give the family an irritating sense of being manipulated. The therapist as Socrates may feel that all points are worth taking seriously, and doing so might not send a message about taking sides. (“Sarah, I understand that you are blaming Henry for causing the problem by getting so angry. Henry, I know you have a different view, and we will come to that in a moment. But first let’s explore Sarah’s view.”) This may lead to a shared account that does some justice to both views. That more complex account, created by acknowledging rather than evading the partisan points, could even be both helpful and true.

According to Stancombe and White, some such account, based on different bits of the partisan versions, is what therapists construct for themselves. They are said to do this “in the backstage” as part of creating a neutral version to be “performed frontstage without compromising the therapist’s ‘neutrality.’ ”

The therapist as Freud is clearly visible in this contrast between the backstage account and the frontstage one offered to the family. Why all this make-believe? Why not go for the Socratic version, using sympathetic questioning to coax the family into themselves constructing the complex account? Perhaps this is what, at their best, therapists already do. Modest therapists understand how small is their knowledge of a family compared to what the family themselves know. They will expect the family’s own constructed complex account to be richer than anything concocted “backstage.” Good family therapists, instead of being in the grip of a theory, have the human sensitivity that goes with being a good listener. Then, with Socratic openness to what they hear, they will help the family dig themselves out of whatever trap they are in. Using gentle questioning, not to move families but to help them move themselves, is a wonderful thing for a therapist to do.

I argued this in the Journal of Family Therapy, commenting on the articles cited above. The editor, Dr. Ivan Eisler, made good criticisms in an editorial. He expressed sympathy for the Socratic over the Freudian approach. But “Glover’s discussion of the Freudian and Socratic positions did not always fit my own preconceptions of what they might mean and kept switching like the corners of the Necker cube from inside to outside.” He supported using gentle questions to help families move themselves. But was this a liking for the Socratic position per se? “Or is it based on our conceptualization of the process of change (clearly a Freudian position in Glover’s terms)?” Therapists who ask questions aimed at eliciting new narratives conducive to change are clearly Socratic, “but the intention is, in Glover’s terms, that of therapist as Freud.” This rightly complicates the account I gave. Freudian intention can motivate Socratic method. (I am happy for Ivan Eisler’s acute point to be the last word here. He is a therapist and I am a philosophy professor. In each role, as we both know, leaving things open can be more fruitful than seeking a decisive verdict.)

Self-Discovery and Self-Creation

And so it was, in term one of my third and last year at Oxford, that I met with Louise. She calls herself a humanistic therapist and talks to you without looking down at you … Louise taught me to take care of myself. She allowed me to see that the best part of growing up is acknowledging a part of us that always remains a child, and to cherish that child always. I discovered feelings I had not admitted, not even recognised, in the past. Fear. I was so afraid. Afraid of uncertainty. Afraid of not being good enough. Afraid of disappointing my parents, other people. Afraid that I was not, could not be, perfect.

—Emily Halban, Perfect: Anorexia and Me

Emily Halban brings out some help a psychotherapist can give. Talking to the person seeking help without looking down is essential for creating the atmosphere for a shared enterprise of assisted self-exploration and self-interpretation. Most of us now know that people are sometimes partly unaware of their own motives and emotions. Any therapist must have the hope, and with luck the satisfaction, of helping the growth of self-understanding.

This self-discovery is one kind of (possibly unintended) self-creation. At the minimum, the person changes from being less aware to being someone more aware. Sometimes the self-knowledge may be enough. But the second, conscious layer of self-creation can be more important. Reclaiming feelings not admitted or recognized may help a person deal with problems through conscious change. It helps to control, or perhaps even eliminate, the emotions and motives from the dark that were once so influential. As in much psychiatry, a core aim must be the move from passivity to actively taking control. The rejection of fatalism is again essential: It is essential to see that it is possible to take some control of your own life.

Part of this therapy is helping people to think through their own deepest values. Alasdair MacIntyre has denied that therapists do this. A therapist “treats ends as given, as outside his scope; his concern also is with technique, with effectiveness,” so therapists cannot engage in moral debate.20 If MacIntyre’s view of therapists is correct, it should not remain so. One aim of therapy is to promote self-understanding, which can feed into self-creation. And self-understanding includes the articulation of a person’s values.

MacIntyre chides therapists for not engaging in moral debate. A lot depends on what that is taken to be. A woman seeking an abortion while in therapy is not going to be helped by a pro-life therapist attacking her moral outlook. That would be Therapist as Freud with a vengeance. But Therapist as Socrates will probe values. Asking questions to elicit hidden values is central to helping self-understanding. “Treating ends as given, as outside his scope,” would be crippling. Part of the Socratic aim is to help people create themselves by seeing what they most deeply want to be.

Horse and Rider

One thing is needful—to “give style” to one’s character—a great and rare art! It is practiced by those who survey all the strengths and weaknesses of their nature and then fit them into an artistic plan until every one of them appears as art and reason and even weaknesses delight the eye. Here a large mass of second nature has been added; there a piece of original nature has been removed—both times through long practice and daily work at it. Here the ugly that could not be removed is concealed; there it has been reinterpreted and made sublime … In the end, when the work is finished, it becomes evident how the constraints of a single taste governed everything large and small. Whether this taste was good or bad is less important than one might suppose—if only it was a single taste!

—Friedrich Nietzsche, The Gay Science

Few of us spend our lives, in the spirit of Nietzsche, as full-time landscape gardeners of the self. But many of us have some rough idea of the sort of person we want to be. And we often have a very clear idea of what we do not want to be. This sense of moral identity is part of the story. As well as caring about being one kind of person rather than another, people also value how they get there. They want to have some control. We cannot just choose our own characteristics. But it matters that, at least to some extent, we can shape the kind of person we are.

Not all self-creation is conscious. Through major decisions about our life, we influence unintentionally the sort of person we become. We are molded by choosing the person we marry or live with, by the friends we choose, the job we take and the place where we live. And there is Aristotle’s path: from actions to habits to character.

Then there is the minimally conscious kind of self-creation described by Sigmund Freud. The I (das Ich, often translated as “the Ego”) tries to bring the blind and conflicting unconscious impulses of the It (das Es, often translated as “the Id”) under some kind of coherent conscious control. But Freud insists that the unconscious desires can be strong enough to limit this: “Often a rider, if he is not to be parted from his horse, is obliged to guide it where it wants to go.”21 Marino Marini’s rather Freudian sculpture in Figure 28.1 captures this.

In Freud’s view, psychoanalysis helps us control the unconscious desires through understanding them. He thought of this as a work of reclamation like the draining of the Zuyder Zee.

“How do I feel about his success?” Answering such a question may be only partly discovery. If, on asking the question, what floats to the surface feels a bit like jealousy, this initial feeling can be questioned. “Am I really so mean-spirited?” Reflection can, sometimes, lead to rejecting the attitude. The mixture of discovery and decision can be microlevel self-creation.22

There are limits to self-creation. To make a sculpture out of wood requires respecting the natural constraints of shape and grain. Self-creation is bound up with self-discovery: finding our own shape and grain. We cannot just choose what will prompt the thought about the real me. This is a reminder of another limit to Aristotle’s thought about being responsible for our own character. What we are like depends partly on physical and chemical states of the brain, on experiences in childhood and even in the womb, and on the culture we live in. At most we only partly create ourselves. Some starting positions might make a particular self-creative project unattainable.

There are two different ways in which our self-creation is not fully ours. One is that how we are makes some transformations too difficult: we can only guide the horse where it is willing to go. The other is that the desires and values guiding our self-creation are not simply chosen by us. They too depend partly on things outside our control. We are horse and rider. Genetic inheritance, parents, early experiences, and chance encounters influence both the kind of horse and the kind of rider we are.