To most of us, Sigmund Freud, who was born 150 years ago next Saturday, is known chiefly as a provocative and highly controversial student of individual psychology. He is the man who theorized the unconscious and the Oedipus complex. What is less well known — and now perhaps more important — is that Freud devoted the final, and maybe most fruitful, phase of his career to reflections on culture and politics. In his later work, Freud brought forward striking ideas about the inner dynamics of political life in general and of tyranny in particular.

Adolf Hitler, who rolled into Freud's home city of Vienna on March 14, 1938, preceded by thousands of troops, was no surprise to Sigmund Freud. Nor would the many forms of tyrannical fundamentalism that have grown up in Hitler's wake and have extended into the 21st century have shocked him very much. In books like "Group Psychology and the Analysis of the Ego" and "Totem and Taboo," Freud predicted Hitler and his descendants almost perfectly. Now, in an age threatened by fundamentalisms of many sorts, Freud's thinking may be more usefully illuminating than ever before.

It is possible that Hitler and Freud actually encountered each other. Hitler spent some of the unhappiest years of his life in Vienna, just before the beginning of World War I. He had come to the great city with hope of becoming a major artist, but he was rejected from art school, not once but twice. In short order, he ran out of money and was reduced to sleeping in doorways and even to begging from time to time. If Hitler and Freud had passed each other on the streets of Vienna, after Freud's return from his highly successful 1909 trip to America, Freud would have seen a street rat, a rank denizen of the mob. (Freud was no populist.) Hitler would have seen a Viennese burgher (he despised the upper middle class) and probably would have identified Freud as a Jew as well. Hitler would perhaps have drawn back in shame at his threadbare overcoat and his broken shoes. Though he might, if things were bad enough, have extended his hand to beg. Whether Freud gave or not (he could well have; he was generally good-hearted) would have made no difference; the encounter would still have left young Adolf seething.

In March 1938, the street rat was back, in full dress uniform, riding through the center of Vienna in an open-topped Mercedes, holding onto the windshield with his left hand and with his right giving the salute to thousands of Viennese citizens, greeting the absorption of their small country into the Reich with howls of joy. ("Finis Austriae," Freud wrote in his diary that March.) Now Hitler was in robust middle age; Sigmund Freud was 81 and desperately ill with the cancer of the jaw that would, in London, a year and a half later, end his life.

Freud, sick as he was during the early spring of 1938, generally refused to take any medication stronger than aspirin. He wanted to think and to write, and for that he needed to keep his mind clear. He stayed away from morphine and from liquor. But staying away from intoxicants and keeping his mind clear meant more than that to Freud. It meant staying away from religion (Freud was a lifelong atheist); it meant staying away from romantic love (Freud called it "the overestimation of the erotic object") and it meant staying away from the kind of politics embodied by the onetime street rat now traversing Vienna in his Mercedes. Why were people so potently and ruinously drawn to Hitler and to all of the other agents of collective intoxication on offer in the world? Freud believed that he knew.

t the center of Freud's work lies a fundamental perception: human beings are not generally unified creatures. Our psyches are not whole, but divided into parts, and those parts are usually in conflict with one another. The id, or the "it," is an agent of pure desire: it wants and wants and does not readily take no for an answer. The superego, or over-I, is the internal agent of authority. It often looks harshly upon the id and its manifold wants. The superego, in fact, frequently punishes the self simply for wishing for forbidden things, even if the self does not act on those wishes at all. Then there is the ego, trying to broker between the it and the over-I, and doing so with the greatest of difficulty, in part because both agencies tend to operate outside the circle of the ego's awareness. The over-I and the it often function unconsciously. Add to this problem the fact that "the poor ego," as Freud often calls it, must navigate a frequently hostile outside world, and it is easy to see how, for Freud, life is best defined as ongoing conflict. In a passage in "The Ego and the Id," Freud observes that the ego is a "poor creature owing service to three masters and consequently menaced by three dangers: from the external world, from the libido of the id and from the severity of the superego. Three kinds of anxiety correspond to these three dangers, since anxiety is the expression of a retreat from danger."

About this conflict — about this painful anxiety — what is to be done? Humanity, Freud says, has come up with many different solutions to the problem of internal conflict and the pain it inevitably brings. Most of these solutions, Freud thinks, are best described as forms of intoxication. What the intoxicants in question generally do is to revise the superego to make it more bearable. We like to have one glass of wine, then two, Freud suggests, because for some reason — he's not quite sure what it is in scientific terms — alcohol relaxes the demands of the over-I. Falling in love, Freud (and a thousand or so years of Western poetry) attests, has a similar effect. Love — romantic love, the full-out passionate variety — allows the ego to be dominated by the wishes and judgment of the beloved, not by the wishes of the demanding over-I. The beloved supplants the over-I, at least for a while, and, if all is going well, sheds glorious approval on the beloved and so creates a feeling of almost magical well-being. Take a drink (or two), take a lover, and suddenly the internal conflict in the psyche calms down. A divided being becomes a whole, united and (temporarily) happier one.

Freud had no compunction in calling the relationship that crowds forge with an absolute leader an erotic one. (In this he was seconded by Hitler, who suggested that in his speeches he made love to the German masses.) What happens when members of the crowd are "hypnotized" (that is the word Freud uses) by a tyrant? The tyrant takes the place of the over-I, and for a variety of reasons, he stays there. What he offers to individuals is a new, psychological dispensation. Where the individual superego is inconsistent and often inaccessible because it is unconscious, the collective superego, the leader, is clear and absolute in his values. By promulgating one code — one fundamental way of being — he wipes away the differences between different people, with different codes and different values, which are a source of anxiety to the psyche. Now we all love the fatherland, believe in the folk, blame the Jews, have a grand imperial destiny. The tyrant is also, in his way, permissive. Where the original superego has prohibited violence and theft and destruction, the new superego, the leader, allows for it, albeit under prescribed circumstances. Freud's major insistence as a theorist of group behavior is on the centrality of the leader and the dynamics of his relation to the group. In this he sees himself as pressing beyond the thinking of predecessors like the French writer Gustave Le Bon, who, to Freud's way of thinking, overemphasized the determining power of the group mind. To Freud, crowds on their own can be dangerous, but they only constitute a long-term brutal threat when a certain sort of figure takes over the superego slot in ways that are both prohibitive and permissive.

As the Nazis arrived in Vienna, many gentile Viennese, who had apparently been tolerant and cosmopolitan people, turned on their Jewish neighbors. They broke into Jewish apartments and stole what they wanted to. They trashed Jewish shops. They made Jews scrub liberal political slogans off the sidewalk, first with brushes and later with their hands. And they did all of this with a sense of righteous conviction — they were operating in accord with the new cultural superego, epitomized by the former corporal and dispatch runner, Adolf Hitler.

On the day after Hitler arrived in Vienna, a gang of Nazis stormed into Freud's apartment, at 19 Berggasse. They ransacked the place and made off with a fairly large sum of money. ("I never got so much for a single session," Freud, never at a loss, observed.) They only left, it is said, when the old man, trembling and frail, appeared from out of his consulting room and fixed them in his long-practiced stare. The Nazis, the story continues, scrambled for the door.

In his last days, Freud became increasingly concerned about our longing for inner peace — our longing, in particular, to replace our old, inconsistent and often inscrutable over-I with something clearer, simpler and ultimately more permissive. We want a strong man with a simple doctrine that accounts for our sufferings, identifies our enemies, focuses our energies and gives us, more enduringly than wine or even love, a sense of being whole. This man, as Freud says in his great book on politics, "Group Psychology and the Analysis of the Ego," must appear completely masterful. He must seem to have perfect confidence, to need no one and to be entirely sufficient unto himself. Sometimes this man will evoke a god as his source of authority, sometimes not. But in whatever form he comes — whether he is called Hitler, Stalin, Mao — he will promise to deliver people from their confusion and to dispense unity and purpose where before there were only fracture and incessant anxiety. But, of course, the price is likely to be high, because the simplifications the great man offers will almost inevitably involve hatred and violence.

Freud's implicit morality is counterintuitive. Though Freud acknowledged the uses of mild intoxicants like love and art, he was nonetheless extremely suspicious of any doctrine or activity that promised to unify the psyche — or to unify the nation, the people — without remainder and to do so forever. Freud believed that the inner tensions that we experience are by and large necessary tensions, not because they are so enjoyable in themselves — they are not — but because the alternatives to them are so much worse. For Freud, a healthy psyche is not always a psyche that feels good. For Herbert Marcuse, author of a brilliant meditation on Freud, "Eros and Civilization," Freud's politics are potentially the politics of ecstasy. We can collectively undo our repressions and regress toward collective erotic bliss. For Philip Rieff, author of the equally perceptive and original "Freud: The Mind of the Moralist," Freud appears to be a deep political pessimist who thinks that the healthiest individuals will probably be those who turn completely away from politics. But another way to look at Freud is to see him as someone who suggests that a considerable measure of freedom and even relative happiness can come from following a self-aware middle way. If we are willing to live with some inner tension, political as well as personal, we need never be overwhelmed by tyranny or fall into the anarchy that giving into the unconscious completely can bring.

For Freud, we might infer, a healthy body politic is one that allows for a good deal of continuing tension. A healthy polis is one that it doesn't always feel good to be a part of. There's too much argument, controversy, difference. But in that difference, annoying and difficult as it may be, lies the community's well-being. When a relatively free nation is threatened by terrorists with totalitarian goals, as ours is now, there is, of course, an urge to come together and to fight back by any means necessary. But the danger is that in fighting back we will become just as fierce, monolithic and, in the worst sense, as unified as our foes. We will seek our own great man; we will be blind to his foibles; we will stop questioning, stop arguing. When that happens, a war of fundamentalisms has begun, and of that war there can be no victor.

Mark Edmundson teaches English at the University of Virginia. He is currently completing a book about the last two years of Freud's life.

It gave him satisfaction, though nothing else did, to tell us the harrowing stories of his life.

The early years: Russian roulette with his father's gun, his sister killed by a rampaging truck, a brother who took pleasure in obscure but unforgettable forms of abuse.

The wild middle years: gangs and deaths by deliberate violence. The despairing years: chaos, random fatherhood, slicing himself with knives just to see his blood run, breaking bones in drunken rages by punching walls.

The present: homelessness and suicide attempts, which had brought him to a psychiatric ward and then, reluctantly, to the clinic.

All through the first year here, he told terrible stories.

"My life's a punishment for the things that went wrong," he said.

The girlfriend who had died in his arms, the gang buddy he'd buried after an overdose; each memory seemingly darker than the last. He came in thin, grim and intoxicated, and he could not promise to return the next month. Yet he did. There was something else to tell.

His depression began to clear. He moved into a sober house, began a day program, gave clean urine samples each week and considered looking for a maintenance job. One day, he arrived with a new expression. He said he had reunited with an old girlfriend. It turned out they had a grown daughter together.

Now, though he had never known her before, he was helping to plan her wedding. Imagine: father of the bride.

Each meeting after that was filled with giddy details. She wanted orchids and a waterfront reception. He would wear a purple tux. There would be a white limo and a DVD. He was taking out a loan to finance the affair. He was also spending every weekend with his old girlfriend and new daughter, the sudden family. Just recently, they had told him he was going to be a grandfather. Imagine.

There were no more stories about badges of violent honor, no lists of reasons to die. "You can only lay in the slop so long," he said.

He put on weight, grinned from a double chin, and talked about another wedding — maybe the father of the bride would marry the mother of the bride. This loan was bleeding him already, for heaven's sake.

His stories were so full of light that a student therapist on the team finally thought to call the halfway house for corroboration. The staff listened with surprise. The patient had never mentioned any girlfriend or daughter to them. He stayed in his room at night and did not mingle with the other guys. He never got phone calls. He was always home on weekends, and utterly solitary. They encouraged him to socialize. But he never went anywhere, ever.

The student, properly astounded, presented this at the team meeting. Suddenly everything in the history began to waver. Maybe all the stories — harrowing as well as fantastic — had been untrue.

No one doubted his depression. But maybe he had trapped himself in a mine of his own construction, and now shame required him to keep digging. If we did not confront him on his dishonesty, weren't we colluding with him? Gently encouraged but respectfully insisted upon, the truth would free him to change his life — the real one.

It sounded right, the high therapeutic path. But it felt wrong. Why insist on truth, when the stories were growing happier and happier? They had begun full of horror, and were now full of dreams. He was living them earnestly in his mind, if not in his life. He was sober. He had a happy double chin. Wasn't he better for dreaming?

Our team has agreed to disagree. The student, who earnestly wants to help, has decided to begin a firmer line of inquiry. I have decided to listen unquestioningly and will not ask to see the wedding album.

In the end, our divisions don't matter. The patient himself will decide what he wants to know and what he will imagine.

n the spring of 1993, when I was an intern at The New Haven Advocate, a local weekly, I met two girls named La'Tanya and Tichelle. La'Tanya was 13, Tichelle was 11 and, along with their two younger sisters, they had recently returned from a year in foster care to live with their mother, Jean. (I have used middle names to protect the family's privacy.) I was supposed to spend an hour or two with them and write an article for the paper about families that reunite. But I liked the girls, and I decided I needed to interview them again — the joy of being an intern, after all, is that no one really cares when you finish your article. The next time I showed up, about a week later, a worried-looking woman was talking to the girls. She was a prosecutor who was about to try Jean's boyfriend, Earl Osborn, for sexually abusing La'Tanya and Tichelle over several years and their 7-year-old sister, Charnelle, for a shorter period. The girls were her chief witnesses.

At the trial that May in a courtroom in New Haven, La'Tanya testified that Osborn started touching her when she was in kindergarten. She told her mother, and Jean put him out of the house. But somehow, though he wasn't violent, she couldn't make him stay out. She would later say that this was the greatest mistake of her life, but in court, she said as little as possible.

During the next five years, Jean warned Osborn to keep away from her daughters. He didn't. When La'Tanya and Tichelle were about 10 and 8, Jean testified, she put a lock on the door of their bedroom. Osborn broke the lock three times. Jean would find him lying on top of the girls, rubbing against them and putting his hands down their pajamas, or next to them masturbating. She gave her daughters a stick to sleep with. But she never banished him from their home.

In June 1991, after being tipped off that there might be a problem at home, a school social worker pulled La'Tanya out of her sixth-grade graduation party and asked if anyone was bothering her. La'Tanya shook her head no and then started to cry, and to talk. Jean lost custody of La'Tanya, Tichelle, Charnelle and their youngest sister, Chanté. That's when the girls lived in foster care, first with a family and then for several months with their grandmother. Osborn was arrested. In May 1993, on the strength of La'Tanya and Tichelle's testimony, a jury in New Haven convicted him of nine charges of sexual assault. Osborn was sentenced to 85 years in prison. The jury foreman was Jon Butler, a history professor, who is dean of the Yale Graduate School of Arts and Sciences. "The verdict hinged overwhelmingly on the credibility of the girls," he told me recently. "They were so good because they weren't so good. They weren't acting, this wasn't contrived, what happened had been deeply disturbing to them. They conveyed this with the kind of precision that made it completely believable."

During the summer after the trial, I spent time with the girls, taking them to the playground and the pool. I wrote about them for The Advocate. Then I left, as interns do. I did stay in touch with La'Tanya and Tichelle, because I was worried about them and because I admired them. They took care of each other, and they were resourceful. "They're very appealing kids, and I don't think anyone expected that, considering what they've been up against," Cecilia Wiederhold, the prosecutor, said in my Advocate article.

As the girls grew up, they kept exceeding my expectations. Study after study has shown that sexually-abused children — especially those who grow up in the sort of low-income, messy surroundings that the girls did — are more likely to develop a raft of emotional and health problems, including depression, post-traumatic stress disorder and suicidal thoughts. As adults, they are more likely to be unemployed, homeless, addicted to drugs or alcohol and alone. Now, at ages 26 and 24 respectively, La'Tanya and Tichelle are none of those things. La'Tanya works as a certified nursing assistant at St. Raphael's Hospital. She has her own apartment in a small town on the Connecticut shore. She is raising her two sons, who are 10 and 5. Tichelle is a computer operator for the city of Bridgeport. She lives with her 1-year-old son in the same apartment complex as La'Tanya. Both sisters graduated from high school and have their own cars.

By middle-class standards, these accomplishments seem modest. But financial independence and stability are rare and hard-won for anyone in the poor black New Haven neighborhood where the sisters grew up. Jean raised her children on welfare and never learned to drive. La'Tanya and Tichelle's fathers have been almost entirely absent. On the blocks of row houses and vacant lots where they were raised, teenage mothers far outnumber married ones.

Over the years, I've wondered what accounts for their relative success. Were La'Tanya and Tichelle different, and if so, why? Weren't their lives supposed to have fallen into chaos? How is it that some children show a certain resilience after experiencing a trauma and others do not?

he everyday meaning of the word "resilience" extends to anything that bounces back. Estée Lauder makes Resilience Lift Eye Crème and Hanes makes Resilience Pantyhose. But in psychology, resilience has a specific meaning. It's the word for springing back from serious adversity, like abuse, war or natural disasters. You exhibit resilience (as opposed to plain competence) if you cope with terrible misfortune and live a relatively successful life as defined by mental health, success in school or at work or solid relationships. In studies of the long-term effects of physical and sexual child abuse, 20 to 40 percent of victims show few signs of behavioral or mental-health problems. And many of them don't appear damaged later in life. As Ann Masten, a resilience researcher, has written, resilient children have the benefit of "ordinary magic." When it comes to abuse victims, though, this finding is rarely trumpeted, for fear that saying abuse isn't always inevitably harmful is tantamount to saying it's not always bad.

Over the last several decades, a small group of researchers has tried to understand how a minority of maltreated children exceed expectations. The grandfather of resilience theory is Norman Garmezy, who by the 1960's had begun asking why some children of schizophrenics fared better than others. In the 1970's, Ann Masten joined Garmezy at the University of Minnesota, and the two, along with others, started a project spanning more than two decades. They looked at a child's personality, among other things, imagining resilience as a function of temperament, will or intelligence. While children of average intelligence or above were more likely to exhibit resilience, the researchers noted that good relationships with adults can exert an effect that is as powerful, if not more, in mitigating the effects of adversity.

In recent years, biological science has proposed a new paradigm. The latest research shows that resilience can best be understood as an interplay between particular genes and environment — GxE, in the lingo of the field. Researchers are discovering that a particular variation of a gene can help promote resilience in the people who have it, acting as a buffer against the ruinous effects of adversity. In the absence of an adverse environment, however, the gene doesn't express itself in this way. It drops out of the psychological picture. "We now have well-replicated findings showing that genes play a major role in influencing people's responses to adverse environments," says Sir Michael Rutter, a leading British psychiatrist and longtime resilience expert. "But the genes don't do anything much on their own."

Rutter opened a GxE research center because he was frustrated that most psychiatric studies tracked the effects either of genes or of environment rather than looking at them in tandem. Despite a few initial successes, like the discovery of the gene that causes Down syndrome, most searches for genes that fully explain psychiatric outcomes — "the alcoholism gene," "the schizophrenic gene" — have failed. Meanwhile, in the field of medicine, it's increasingly common to consider external factors when studying the effects of genes. "With heart disease and cancer, genetic researchers have always known to include factors like smoking and exercise," says Terrie Moffitt, who is on the faculty of Rutter's research center at the Institute of Psychiatry in London and of the University of Wisconsin. "We wanted to do the same thing for the study of behavior."

The breakthrough moment for GxE came in 2003, when Moffitt and her husband and co-investigator, Avshalom Caspi, published a paper in Science that discussed the relationship between the gene, 5-HTT, and childhood maltreatment in causing depression. Scientists have determined that 5-HTT is critical for the regulation of serotonin to the brain. Proper regulation of serotonin helps promote well-being and protects against depression in response to trauma or stress. In humans, each 5-HTT gene has two alleles, and each allele occurs in either a short or a long version. Scientists are still figuring out how the short allele affects serotonin delivery, but it seems that people with at least one short 5-HTT allele are more prone to depression. And since depression is associated with unemployment, struggling relationships, poor health and substance abuse, the short allele could contribute to a life going awry.

About one-third of the white population have two copies of the protective long allele. About one-half have one long allele and one short one. And about 17 percent have two short alleles. (African-Americans are less likely to have a short allele; Asians are more likely.) In their 2003 study, Caspi and Moffitt looked at 847 New Zealand adults and found a link between having at least one short 5-HTT allele and elevated rates of depression for people who had been mistreated as children or experienced several "life stresses" — defined as major setbacks with jobs, housing, relationships, health and money. Having two short alleles made it highly likely that people who had been mistreated or exposed to unhinging stress would suffer depression. One short allele posed a moderate risk of depression in these circumstances. Two long alleles, on the other hand, gave their carriers a good chance of bouncing back under negative circumstances. In other words, as a group, children with two risky alleles lost out badly when their environments failed them, children with one risky allele were at some risk and children with good resilience alleles often carried a shield. The risky variation of the gene doesn't confer vulnerability, though, if an individual who carries it never experiences abuse or serious stress — in other words, it's not a "depression gene" in any general sense. It seems that only under dire circumstances — abuse, the strife of war, chronic stress — is the gene triggered. Eventually scientists hope to understand more about other genes that most likely play a role like 5-HTT's.

esearchers who study humans cannot, of course, run controlled experiments by randomly assigning some children to abusive homes. But primatologists can. At a laboratory in rural Maryland, run by the National Institutes of Health, Stephen Suomi studies 500 rhesus monkeys. Each year, Suomi divides newborn monkeys into several groups. One group live with their mothers, much as they would in the wild (except for the indoor pens and the daily rations). Another group, created to mimic the experience of a neglected or abused child, never see their mothers, spending two weeks in an incubator and then moving into a small group of peers.

Rhesus monkeys share with humans about 96 percent of their genes — including the long and short variations of 5-HTT. Using DNA samples, Suomi is able to track which of his monkeys have which allele. In an ongoing study, Suomi has found that motherless, peer-raised monkeys who have a copy of the short 5-HTT allele are more likely to experience fear, panic and aggression (accompanied by low levels of serotonin acid in spinal fluid) when a strange monkey in a cage is placed next to them. Motherless, peer-raised monkeys with two long alleles, on the other hand, are more likely to take the presence of the stranger in stride, as mother-raised monkeys do. (Only a tiny number of monkeys have two copies of the short allele, so they're not studied.) "How you grow up affects your hormonal output and the structure and function of the brain," Suomi says. "And these effects are tempered by the kind of gene the monkeys carry. So it's a true interaction."

In Suomi's lab, there's a room filled with large cages of 2- and 3-year-olds — adolescents approaching adulthood, in monkey years. "Go in quickly and quietly," Suomi tells me and then follows me through the door. Some of the monkeys stay in the middle of the cage, eyeing us without seeming preoccupied. Another group races to the back and huddles together in the farthest corner, their small fingers wrapping around one another's fur. They twitter and turn their faces away in distress.

The middle-of-the-cage monkeys were raised by their mothers. The freaked-out ones at the back raised one another. After a few minutes, some of the peer-raised monkeys begin to dart forward. After a few more minutes, they settle in with the mother-raised group. But others never move from the back of the cage. According to Suomi, you could approach the cage a hundred times and each time see the same result. And each time, the peer-raised monkeys would race to the back, and then a few would mirror human resilience by coming forward. And they would generally be the monkeys with two long 5-HTT alleles. The good version of the gene.

Suomi's peer-raised monkeys are deprived of their mothers and other adult monkeys. Abused children, by contrast, don't just live with other children. They may have in their lives grandparents, aunts, teachers, maybe an adult they know from church or a volunteer from a Big Brother or Big Sister program. And they are much more likely than other children to name one of these adults as the person on whom they most rely.

Caspi and Moffitt's research was important in showing a link between genes and an abusive environment, but they didn't explore the effect of mitigating factors in the abused children's lives. Joan Kaufman, a Yale psychiatry professor, has taken the next step by doing so. In a paper this month in the journal Biological Psychiatry, Kaufman reported on 196 children between the ages of 5 and 15, 109 of them removed from their homes in Connecticut because of reports of physical or sexual abuse or neglect. This group was compared with a second nonabused group with the same racial composition — about 28 percent white, 24 percent Hispanic, 28 percent African-American and 20 percent biracial — and the same income of $25,000 or less. (Physical and sexual abuse are more prevalent among poor families, though abuse happens at all economic levels. Studies like Caspi and Moffitt's, which include families of varied income, show that the resilience findings apply to middle- and upper-class kids as well as poor ones.)

Kaufman gave all 196 children a questionnaire about their moods, which measure mental health. She also used DNA tests to check their 5-HTT alleles. Kaufman's abused children with two short 5-HTT alleles had a higher mean score for depression than the abused children with two long alleles and the nonabused children, no matter what their alleles. (In Kaufman's study, which was smaller than Caspi and Moffitt's, the moderate risk of depression posed by one short allele didn't show up, though that finding has been replicated by other researchers.)

In her recent paper and in earlier research, Kaufman also built on the work of psychologists who have measured the quality of abused children's relationships to adults, asking the children to name the person they most often "talk to about personal things, count on to buy the things they need, share good news with, get together with to have fun and go to if they need advice." The mean depression score for abused children with two short alleles who rarely saw the adults they named was off the charts. If the children with two short alleles saw the adults they counted on daily or almost daily, their depression scores were very close to the scores of the children with two long protective alleles — and within reach of the children who had not been abused. (The children with the protective version of the gene were far less affected by a lack of contact with their primary adult.) "Good support ameliorates the effect of abuse and of the high-risk genotype," Kaufman says. While he notes that Kaufman's research is preliminary, Dennis Charney, a psychiatry and neuroscience professor at Mount Sinai School of Medicine, says that the study used "solid methodology and yielded very interesting findings."

La'Tanya and Tichelle had relatively good support. When the state removed the girls and their sisters from their home, their grandmother took them in. "She did everything for us," La'Tanya remembers. Later, a half-sister who is 10 years older than La'Tanya began picking her up on weekends. "I spent a lot of time with her," La'Tanya says. "We'd rent movies, go places, do a lot of things together."

Having "good support" isn't just a question of good luck. Researchers have found that children who are resilient are skillful at creating beneficial relationships with adults, and those relationships in turn contribute to the children's resilience. La'Tanya and Tichelle were both good at forging these bonds. When I left New Haven in 1994, they wrote me. I moved back a few years later, and Tichelle called regularly, came to my office to meet me for lunch, asked me to stop by her house on the weekends. La'Tanya soon started calling, too. Sometimes the sisters were behind on their bills and, always with embarrassment, asked me for money. But more often they called, and still call, to check in, to ask after my kids or tell me about theirs. They let me know that I matter to them, and that has made them matter more to me.

In the last year or so, I've become more aware that La'Tanya and Tichelle are quite different and that the darkness of childhood seems to have left a more indelible mark on the older sister. La'Tanya and her sister were both molested. And they lived with their mother, who was remote — I don't think I ever saw Jean hug her daughters in the months surrounding the trial. But La'Tanya shouldered more of a burden than Tichelle did. As the oldest, La'Tanya often had to look after her sisters — make them dinner, put them to sleep. "La'Tanya raised us," Tichelle says. "She's more like a mom than a sister." La'Tanya wrote in her diary last year, "Ever since I can remember I've taken on more than any one person should have to."

In March 1994, when they were 14 and 12, the girls sent me letters. Tichelle wrote: "My grades are excellent. I got all A's and two B+'s. In school I am a cheerleader we cheer every Tuesday and Thursday. I am still in double dutch it's very fun." La'Tanya also reported her good grades. But mostly she described the heaviness of her world. "A lot of people have been getting killed," she wrote. "A house almost got burned down and nobody ever goes outside."A year later, La'Tanya ran away from home. She moved in with her 17-year-old boyfriend and decided to have a baby. The thread that runs through these decisions is her anger and disappointment with Jean. "I had a baby to be loved by someone," La'Tanya says. "When all the mess fell out, my mother didn't do anything about it. That's what made me think she didn't love us."

Tichelle, by contrast, forgave her mother and relies on her. "That's my mom, and I'm not going to let anyone take her away." On the day Tichelle went into labor last spring, the father of her baby was arrested for selling drugs. She called her mother, and after the birth Jean slept at the hospital with her daughter and grandson. Jean takes care of the baby every day while Tichelle is at work, as she has done for her other grandchildren.

Before her baby was born, Tichelle landed a permanent position in her office after impressing her boss as a temp. "I'm 23, and I know what I want for myself," she told me last year. La'Tanya, meanwhile, was struggling with crying spells and panic attacks. After a recurrent nightmare about Osborn, she would wake up and compulsively check and recheck the locks in her apartment. She couldn't stop thinking about Osborn's gun — "a black shotgun with a light brown barrel."

There are a lot of reasons that La'Tanya has had a harder time than her sister, not least of them that she was abused for a longer period of time. But reading all the GxE research made me wonder whether she was also more genetically vulnerable. I asked the girls if they'd be willing to be tested, and they agreed — they said they were curious. Last month, La'Tanya, Tichelle and Charnelle (who had been abused by Osborn for a shorter period when she was 3) sent cheek swabs with their DNA to a lab run by a Colorado-based company called NeuroMark, which tests for the 5-HTT alleles.

The study of resilience is nearly 50 years old. Yet its contribution to our understanding of the effects of child abuse has gained little traction beyond a small subset of academics. Historically, the study of resilience inadvertently collided with the movement to treat child abuse as a national cause for alarm. In the 1950's, experts like Alfred Kinsey minimized the damage of sexual abuse. The fright described by children who'd had sexual contact with adults was "nearer the level that children will show when they see insects [or] spiders," Kinsey wrote, as Joseph E. Davis, a sociologist, recounts in his recent book, "Accounts of Innocence." Until the mid 1970's, standard psychology textbooks also played down the effects of abuse and put the incidence of incest at one in a million.

By the beginning of the next decade, the textbooks were being rewritten. Led by Judith Lewis Herman, a professor of clinical psychiatry at Harvard, feminists shredded the myth that sexual abuse is rare and does little significant harm. They argued that even a brief, single incident of abuse could and often did scar victims for life. In 1980, Herman helped win an official psychiatric diagnosis for post-traumatic stress disorder, as a response to trauma that causes people to "dissociate," or fragment, by alternately feeling numb and reliving the event. According to its clinical definition, PTSD can strike a victim of rape or child abuse as easily as a combat veteran. The orthodoxy that abuse necessarily causes trauma grew and still remains entrenched. It has extensive institutional support — $29 million a year in government financing goes to a national "traumatic stress" center and 44 hospitals and community-based programs around the nation. "The problem in our country isn't that we overidentify trauma," said Ellen Gerrity, associate director for the national center. "It's that we underidentify it."

All along, a few experts have raised doubts about equating childhood sexual experience with trauma and about the assumption that abuse destroys victims' lives. But they got little attention — until an academic brush fire over terminology exploded into a public war. In 1998, Bruce Rind, Robert Bauserman and Philip Tromovitch published an article in Psychological Bulletin, a journal of the American Psychological Association, analyzing 59 studies of the long-term effects of sexual abuse and adult-child sexual contact on college students. "At the time, the starting hypothesis in the field was that child sexual abuse, broadly defined, was extremely harmful in all cases," Rind says. "Our idea was to take this very strong statement and to be statistically and methodologically rigorous about testing it."

The Rind paper found only a marginal difference between the psychological well-being of college students who'd been "sexually abused" and those who hadn't. But there was a catch: Some of the studies being analyzed defined sexual abuse broadly to include exhibitionism and consensual contact between teenagers and adults. When abuse was limited to lack of consent, force or incest, its deleterious effects were more pronounced. So Rind and his co-authors recommended narrowing the definition of abuse.

Conservatives condemned the Rind paper, and Congress denounced it. The uproar virtually derailed the hope of opening child sex studies to rigorous inquiry. "There had been an underestimation of the extent to which children can recover from sexual abuse," says David Finkelhor, a sociologist who directs the Crimes Against Children Research Center at the University of New Hampshire and who has found that pre-existing depression may make children vulnerable to sexual abuse and may help account for the problems some suffer afterward. "But that article started a trend in the opposite direction," he says — by discouraging investigations of the differences between harmful sexual abuse and other sexual contact.

Still, in trying to understand why some children are not scarred for life, resilience research brightens a picture that is often painted in black. And the promise of GxE brings with it new excitement — and grant money. Caspi and Moffitt, whose article about 5-HTT received widespread praise, didn't bother to try to get government money when they began collecting DNA data in 1998. Now they receive $500,000 a year in N.I.M.H. and U.K. government grants. Other researchers have similar support for GxE work.

If GxE pans out as the enthusiasts hope, it could change not just our understanding of the effects of abuse but also our treatment of it. Neurobiological research on mice and rats has begun to look at the effect that the 5-HTT gene has on the brain at the molecular level. Eventually, a designer drug might succeed in mimicking precisely what the long-allele variation of 5-HTT does to foster resilience. "A magic drug down the line — yes, that's the whole point of understanding the neurological mechanisms," Joan Kaufman says.

Other experts, however, are skeptical. Whatever an abused child's genes, they argue, she still needs the ingredients that promote resilience — adults she can trust, the reinforcements that make her believe in herself. "It's nice to know what's going on in the body," says Suniya Luthar, a psychology professor at Columbia University's Teachers College. "But what's the real promise here? We already know what people need to be resilient. From the standpoint of intervention, I'd rather see money go toward things that are more likely to make real change," like developing effective interventions. Luthar also worries about genetic profiling. "Are we going to think about genetic engineering" to weed out the high-risk variation of a gene? she asks.

Kaufman, too, warns that finding out which variation of 5-HTT you carry is not like getting a diagnosis. The short allele increases vulnerability across a large group rather than exposing any one person who has it. Other genes, as well as relationships, contribute. "Think about it as one factor on a scale," Kaufman says. "It can tip the scale toward depression or away from it. But other factors can tip the scale, more powerfully, in the opposite direction."

Still, the test results for La'Tanya, Tichelle and their younger sister Charnelle are intriguing. As it turns out, Tichelle carries only the protective version of the gene (two long alleles). So does Charnelle, who at 20 is thriving, with a steady job at a nursing home and an apartment she shares with a boyfriend whom her family likes.

La'Tanya, though, carries one copy of the short 5-HTT allele, putting her in the group of abused children who are at moderate risk of depression. Perhaps her genes help account for the times she has gone to the hospital because she's so anxious she can't breathe and the days she can't stop crying or get out of bed.

Yet, as Kaufman says, La'Tanya's genes don't doom her to unhappiness. She has good days too. Last year she signed up with a home-health-aide agency and was frustrated — and broke — because she was getting only scattered hours of work. She wanted a hospital job. But that required a state recertification. I gave her $700 for the four-week course she needed.

At the graduation ceremony last spring, two dozen students ate pizza next to the hospital beds and dummy patient they'd practiced on. "I got a 93 on the final," La'Tanya told me twice. She was wearing a pink turtleneck sweater, pressed jeans and high-heeled black boots. She'd taken out her lip piercing because the teacher said it might put off potential employers. "Now I just need to take the licensing test," she said to a friend. "When we leave from here I'm going over to sign up." She did. She passed. She got the hospital job she wanted. Keeping it hasn't been easy, especially because her younger son has been acting up in school. But she is doing it.

Next month, NeuroMark will begin selling the 5-HTT test to people whose doctors request it. The results won't solve the riddle of which survivors of abuse fare better than others. But they may provide a clue. "I think for me it helps explain things," La'Tanya said. "I feel a little better that there is a reason, another reason, for my life being hard. And I understand that what I'm able to do for myself and my kids, even with this, is good. It's good."

Emily Bazelon is a senior editor at Slate and a recent Soros Media Justice Fellow.

Leesburg, Va.

I RECENTLY did some research for a satirical novel set at a university. The idea was to have a bunch of gags about how colleges prostitute themselves to improve their U.S. News & World Report rankings and keep up a healthy supply of tuition-paying students, while wrapping their craven commercialism in high-minded-sounding academic blather.

I would keep coming up with what I thought were pretty outrageous burlesques of this stuff and then run them by one of my professor friends and he'd say, Oh, yeah, we're doing that.

One of my best bits, or so I thought, was about how the fictional university in my novel had hired a branding consultant to come up with a new name with the hip, possibility-rich freshness needed to appeal to today's students. Two weeks later, a friend called to say it was on the front page of The Times: "To Woo Students, Colleges Choose Names That Sell." Exhibit A was Beaver College, which had changed its name to Arcadia University. Applications doubled.

I also had created a character, a former breakfast-cereal executive who returns to his alma mater as vice president for finance (to give something back) and tries to get everyone to call the students customers. It turns out Yale was already doing that.

I knew that Tom Lehrer, the great satirical songwriter of the 60's, had said he had to give up satire when it kept being overtaken by reality. The final straw, he said, was Henry Kissinger winning the Nobel Peace Prize.

My final straw came when a friend at Case Western Reserve University (now referred to as Case, after their consultant concluded that all great universities have single-word names) sent me a packet of information on the university's new showcase undergraduate seminar program. Called SAGES (this supposedly stands for Seminar Approach to General Education and Scholarship), the program offers as an essential component of its core intellectual experience an upscale cafe that serves Peet's Coffee and is "staffed by baristas whose expertise in preparing espresso is matched only by their authoritative knowledge of all things SAGES."

As the program's Web site explains (complete with footnotes, bibliography and quotes from the urban theorist Jane Jacobs): "In the bustling personal-but-impersonal rhythms of campus activity, as in the streets of a big city, proprietors of public establishments occupy a special position... The SAGES cafe staff are patently not interested in providing grades or passing judgment." And, not only that, but "there are no compromises that would undermine the quality of our drinks.... Our chai latte is made not from a bottled concentrate, but from a fresh-brewed base made from scratch every day on site."

As a model of pandering to students in the guise of lofty academic purpose, I thought that was pretty hard to top. Then I started reading the 92-page guide Case has created for teachers of these seminars.

If students fidget, talk or walk out of class, the guide advises seminar leaders not to "manage" such behaviors, but to explore their underlying causes. Instructors must remember that to such characteristically American cultural beliefs as the importance of morality, rationality and personal responsibility, there are equally valid alternatives that must be respected.

Instructors must be wary of spurious objectivity, such as a 0-100 grading scale; much better is a 0-5 scale, or, best of all, a check, check-plus, check-minus scale. And finally, if students do not contribute to discussions at all, seminar leaders should "make space for silence."

It's enough to drive a satirist to something stronger than chai latte.

Stephen Budiansky is the author, most recently, of "Her Majesty's Spymaster."

A mother, a father and a teacher sit down for a conference. A question soon arises: Are they talking about the same child?

It may not seem so. Several studies have found that evaluations of students by parents and teachers overlap on less than a third of the measures, a "pretty low" rate of agreement, said Timothy R. Konold, coordinator of research, statistics and evaluation at the University of Virginia's Curry School of Education.

Educators have generally assumed that the teacher is right, with some justification, Dr. Konold said. "Teachers have a whole classroom of kids to use as a standard" for assessing behavior, he said, "and can compare them with others."

But a new study by Dr. Konold concluded that parents and teachers focus on different aspects of children, with teachers more attuned to external behavior and parents more sensitive to emotional states.

Dr. Konold's study, which was presented earlier this month at the annual meeting of the American Educational Research Association in San Francisco, drew on evaluations of 562 first graders. He grouped traits into three he considered "internalizing" — whether a child was withdrawn, had physical complaints or seemed anxious or depressed — and two that were "externalizing" — delinquent and aggressive behavior.

Analyzed in that way, the disagreements fell into a pattern. The scores the mothers assigned for the emotional traits were roughly twice those assigned by teachers — an average score of 47, compared with 24 from teachers. On the external behaviors, the gap was even larger, with teachers assigning an average score of 77, and mothers 35.

Dr. Konold said the study did not answer "whether the kids are actually acting differently in home and at school." But it does argue that parents and teachers should try to meld their perspectives.

"Parents' ratings are as important and in some cases more important, because they might see something that explains the behavior" a teacher sees, he added.

Valerone Young, a teacher at the Ashley Elementary School in Winston-Salem, N.C., said that when parents doubt her reports of troubling behavior, she relies on documentation of specific incidents or even a videotape of their child acting badly in class. "The proof is in the pudding," Ms. Young said.

The issue, isn't winning the argument, she said, it is how to get enough information to get the fullest picture of the child. "I think that a lot of times their emotions dictate their behavior in class, and because I only know their behavior, I don't always know the why," Ms. Young said of her students. "If I talk to the parents, we can figure out what's changed."

Dr. Konold's study also found that fathers were the least likely to flag problems of any kind with a child. On external behaviors, their average rating was a mere 25.

Ms. Young said that finding meshed with her experience. "Fathers are more lenient when it comes to behaviors," she said. "Mothers are like, you have to do this!"

In only 10 minutes, my patient had come to hate me, and who could blame her? I had frowned at her blood pressure, rolled my eyes at her weight, clucked at her blood sugar readings, asked some pointed questions about her drinking habits and pointed out the possible relationship between her chronic stomachaches and her tourniquet-tight jeans.

First she just looked miserable and guilty, but by the time I got around to asking exactly what she had eaten for dinner the night before, she had become downright hostile. "Cheesecake," she spat out, with venom.

Under the circumstances, there was only one thing to say, and I didn't hesitate for a second. "By the way," I said, "I've been meaning to tell you, I love what you've done with your hair."

We were taught to call them lubricating comments: little morsels of oleaginous verbiage tucked into the usual miserable catechism to ease it along a little. Quite early on in medical school, we were handed a list to memorize. Most of us shuddered. It seemed then, in that nice, peaceful classroom, that the list's contents were just inane. "Tell me more about that." "That must have been very difficult for you." "I hear what you are saying." "Your story moves me." Surely, with all the other wisdom spilling from our lips, we would not be resorting to those viscous clichés.

But with experience came the knowledge that, without those clichés and innumerable others like them, the game is lost before it even begins. When almost every other word out of your mouth is destined to be an unwelcome reminder, dire prediction or insulting remark of some sort, it is really essential to be able to dilute it all with a little harmless verbal goo.

Presumably out there somewhere there are doctors who operate grimly goo-free. There are probably also those who loyally adhere to those medical school lists, in all their earnest, humorless orthodoxy. But most of us, over the years, slowly work out our own individual habits of senseless patter to smoothe over the hard parts.

I am among those who have grown to respect greatly the power of the minor compliment. And so, when someone snaps "cheesecake" at me in a certain tone of voice, I know to segue admiringly into hair.

I never lie. Sometimes I am guilty of exaggeration. Sometimes I overdo and start to sound like some slavish velvet-clad courtier making nice to Old King Cole. Some days I'm too tired and busy to bother with any of it. Occasionally I fish around for a nice thing to say and come up absolutely cold. But the experts tell us it is possible to find something to like about every patient, and it is also generally possible to construct a mild, inoffensive compliment for almost anyone.

The interesting part is that you never really know what you will get in return.

From most patients there are nothing but smiles. They love knowing that the doctor realizes they are more than their symptoms and prescriptions, that they have another dimension, even if it is just their taste in earrings. Should the compliment involve something more substantial — an intellectual achievement, a personal milestone, valor in the face of adversity — they almost begin to purr.

Others have been known to react differently.

The whole question of sexual harrassment aside — and if I were a man, I would have reconsidered many of my comments long ago — unnecessary personal remarks can get just a little too icky-sticky for some people's taste. They prefer their health care formal, businesslike and brief. They will do their socializing elsewhere.

Some people just aren't in the mood: they are feeling sick and belong in bed.

Some polite souls feel they must respond in kind, and then it all gets complicated. I like their earrings. They like my shoes. We wind up in an odd re-creation of a cocktail party without the alcohol. When does it all stop? I like your tie. You like my sense of humor. I appreciate your punctuality. You love my pen. Look, I'm sorry but I have run out of things to like about you, and I hate your cholesterol. Can we get back to work?

And then there are the clear-eyed few who see right through the ruse, and disdain me for the base manipulator that I am.

"You know you're my favorite patient," I once cooed to a particularly tough nut. He just snorted, annoyed. "You say that how many times a day?" All right, he had me there, the first one ever to call me on that particular remark, and I told him so.

And then, by God, he cracked a big fat smile. A happier man you've never seen. It had worked again.

AS April draws to a close, parents of college-bound children can finally relax. The verdicts are in. No more lying awake worrying that the A.P. biology teacher can't write a decent reference letter, or carrying around a pocket calculator to recheck G.P.A., or clogging supermarket aisles to compare rankings in U.S. News & World Report.

Having been caught in the insanity with my first child (and wanting to avoid it with my second) I decided to stand back and ask: Why do so many intelligent, rational people turn into basket cases when their children start the application process? Parents have always wanted the best for their children, but in the last decade this desire has taken an unusually obsessive form.

For the next crop of parents, before they succumb, here is my own sociological analysis of what's going on.

THE NEED TO DO BETTER

Only a few generations ago, going to college was a significant achievement. But once higher education became widely available, it ceased to carry the same weight. More seemed called for: successful parents who received excellent educations at city and state universities now want their children to attend private universities, and those who thrived at good if not stellar private universities want their children to go to more elite ones. As for parents who attended Ivy League schools, they want their children to go there, too, but to work harder, make better connections and double major in economics and Chinese.

The drive to improve, to do things better and bigger, is inherently American. Yet there comes a time when there's less room for improvement, and we must find artificial goals — a wider TV screen, a bigger S.U.V. — to produce an illusion of progress. This seems to be the case with college. It used to be that getting an education was the goal; now, it is getting a brand-name education. This is a shift from content to form, from the thing itself to the way it is packaged.

THE NARCISSISM OF MINOR DIFFERENCES

Freud coined this phrase to refer to the phenomenon that similar countries or peoples, living in proximity to one another, develop rivalries and antagonisms. This makes the individual communities more cohesive, and allows each to assert its prominence. The same holds for colleges. Elite institutions of a certain size and locale tend to resemble one another and yet to fight fiercely to maintain a sense of superiority, which is to say, a higher U.S. News ranking.

Universities use rankings as a lure, making it seem that admission to a college ranked higher than another is crucial, only to turn around and blithely use the club of selectivity to beat away the hordes that, as a result, want desperately to gain entry.

A student is likely to experience great disappointment when rejected from, say, Amherst, despite being admitted to, say, Colby. But since these colleges are seemingly very similar, enormous time and energy are required to distinguish one from the other. Monet painted many pictures of Rouen Cathedral, which art historians have spent their careers studying. Likewise, parents pore over the catalogs of Dartmouth, Williams, Middlebury, Bowdoin, et al. As with all connoisseurship in which extremely small, elusive qualities are subject to analysis, the results are bound to be uncertain and contradictory, leading parents to an even deeper analysis of small New England liberal arts colleges.

The admissions process is not unlike the kabbalah or "Finnegans Wake": one works hard at coming to an understanding under the assumption that one never will. Urban (or rather suburban) legends abound: how the student with perfect scores — a third-generation legacy whose grandfather has a building named after him — was deferred from Yale. Parents are told "it's a crapshoot" so many times they may be tempted to join Gamblers Anonymous.

CONSPICUOUS ACHIEVEMENT

As the price of tuition began to rise in the 1980's, it looked as if the middle class would make a beeline for state universities. But the predicted backlash never happened. Instead, paying a high price for college became an example of what the sociologist Thorstein Veblen, in 1899, called conspicuous consumption — money spent as a sign of social status.

But attending certain colleges these days is not just conspicuous consumption; it is also conspicuous achievement — accomplishments displayed as a sign of social status. And children are the favored vehicles. If Johnny is proficient in American sign language, contract bridge and ice dancing, Janie must play the harp, collect prom dresses for the needy and be a member of the Olympic equestrian team. To fail at landing a seat in an elite college is to fall behind in the race that began with potty training.

ARTISTIC ANGST

The child-centered culture that has emerged the last several decades can be understood as an outgrowth of the creative self-expression of the 1960's. Parents came to see themselves as artists, and their children as canvases or lumps of clay. The problem, of course, is that children are not made of canvas or clay and are likely to resist being filled in or pummeled into shape.

In light of this resistance, the obsessive focus on college can be understood as a form of artistic desperation, a last-ditch effort to validate parental creativity: "Well, maybe Jenny isn't exactly what I had in mind, but at least she got into Duke." A great college can also be enlisted to finish the work begun by parents. The credit here might read: "Jenny Jones by Jeff and Amanda Jones, with assistance from Duke University."

APOCALYPTIC ANGST

Artistic angst reflects the fear of parents that they may not have shaped their progeny into the best or most finished form. Apocalyptic angst is the larger fear that the world will hamper their children in unforeseen ways.

This anxiety has always been around, but it greatly increased during the Vietnam era when it became clear that money and heavy artillery were not enough to protect our children. The uncertainty and vulnerability that are that legacy have only increased with the threat of terrorism. Thus, the college decal on the car is just a talisman against harm.

The parents who are most anxious over admissions tend to be highly educated. This is not surprising. They know a good education can pave the way for success. But what constitutes a good education? No doubt an elite college will provide one, but will it provide a better one? Isn't the notion of a good education as varied as the temperaments and interests of our children?

And what of success? For some, success means working in a Manhattan law firm with all its financial and social perks; for others it's teaching in a public school with summers free to travel and explore hobbies. Perhaps our children can "do better" when they send their own children to college — with a balanced, philosophical and individualized perspective on the meaning of success.

Paula Marantz Cohen teaches English at Drexel University. Her latest novel is "Jane Austen in Scarsdale or Love, Death and the SAT's" (St. Martin's).

Most prisons are notorious for the quality of their cuisine (pretty poor) and the behavior of their residents (pretty violent). They are therefore ideal locations to test a novel hypothesis: that violent aggression is largely a product of poor nutrition. Toward that end, researchers are studying whether inmates become less violent when put on a diet rich in vitamins and in the fatty acids found in seafood.

Could a salmon steak and a side of spinach really help curb violence, not just in prison but everywhere? In 2001, Dr. Joseph Hibbeln, a senior clinical investigator at the National Institutes of Health, published a study, provocatively titled "Seafood Consumption and Homicide Mortality," that found a correlation between a higher intake of omega-3 fatty acids (most often obtained from fish) and lower murder rates.

Of course, seeing a correlation between fatty acids and nonviolence doesn't necessarily prove that fatty acids inhibit violence. Bernard Gesch, a senior research scientist at Oxford University, set out to show that better nutrition does, in fact, decrease violence. He enrolled 231 volunteers at a British prison in his study; one-half received a placebo, while the other half received fatty acids and other supplements. Over time, the antisocial behavior (as measured by assaults and other violations) of the inmates who had been given the supplements dropped by more than a third relative to their previous records. The control group showed little change. Gesch published his results in 2002 and plans to start a larger study later this year. Similar trials are already under way in Holland and Norway.

What would it mean if we found a clear link between diet and violent behavior? To start with, it might challenge the notion that violence is a product of free will. "But how do you exercise that free will without using your brain?" Gesch asks. "And how, exactly, is the brain going to work properly without an adequate nutrient supply?" The belief that people choose to be violent may be irrelevant if the brain isn't firing on all cylinders. This may especially be the case for impulsive acts of violence, which are less a choice than a failure to rein in one's worst instincts.

Consider, for example, a study conducted by researchers in Finland. They tested prisoners convicted of violent crimes and found that they had lower levels of omega-3 fatty acids than ordinary, healthy subjects. Why? Omega-3's foster the growth of neurons in the brain's frontal cortex, the bit of gray matter that controls impulsive behavior. Having enough of these fatty acids may keep violent impulses in check. Violent criminals may not be the only ones who would benefit from more fatty acids in their diet. In a recent double-blind trial, when omega-3's were given to people with a history of substance abuse, the symptoms of "anger" fell by 50 percent.

Of course, omega-3's are widely hailed these days as a miracle substance, credited with boosting health in dozens of ways. But Gesch warns against what he calls "silver bullets." The state of the evidence, he says, "doesn't allow us to pinpoint which dietary fat is responsible for changes in behavior." In his new study, he will look into whether several interdependent nutrients may play a role.

Gesch further adds that we shouldn't expect nutrition alone to banish violent behavior. "The brain needs to be nourished in two ways. It needs to be educated, and it needs nutrients. Both social and physical factors are important." Simply throwing fish and vegetables at violent criminals is unlikely to have a lasting effect on its own.

Caveats aside, there's something that many people may find unnerving about the idea of curing violent behavior by changing what people eat. It threatens to let criminals evade responsibility for their actions. Think, for example, of the infamous "Twinkie defense," in which an accused murderer's lawyer suggested that junk food was partly to blame for his client's compromised mental state. More controversial, perhaps, is the brave-new-world idea of using diet to enforce docility and conformity to the rules, a sort of state-sponsored version of that timeless parental demand to children everywhere: "Eat your vegetables."

Then again, we already live in a society in which parents have resorted to drugs like Ritalin to quell unwanted outbursts and impulsive behavior. And when you approach it from that perspective, changing what people eat may not be so radical after all.

Stephen Mihm teaches history at the University of Georgia.