Is Gottman himself "kitchen-sinking"--attempting to do too much at once--and in the process endangering his own research? His methods are certainly eclectic, combining basic and applied research, peer-reviewed journals and Oprah Winfrey, mathematical vectors and marketing savvy. The systematic study of human relationships, he argues, is the new frontier of psychology, a major shift away from the long-standing focus on the individual. But will the popularization of his findings distort the integrity of his work, congealing tentative data into how-to certainties? Or is the psychologist simply doing what a good scientist should do: making his findings accessible to a public that can use them, in language it understands?

To my left are three computers, each equipped with a white plastic coding strip above its keyboard. I lean over and read the fine print on the strips: Disgust, Contempt, Belligerence, Domineering, Criticism, Anger, Tension, Tense Humor, Defensiveness, Whining, Sadness, Stonewalling, Neutral, Interest, Validation, Affection, Humor, and, finally, Surprise/Joy. ("We don't actually see much of that," remarks Carérre dryly of the last.) It's a disturbing selection, like the spice rack of the biggest fight you ever had. And it raises a question: Can even the best numerical analysis reveal the formula for the bond that the essayist Lynn Darling once called "an intricate pattern of consideration and savagery that only two human beings moving together through time can produce"? Is committed love in fact a set of skills, like whipping up a tangy risotto--something that can be learned by mimicking the gestures of highly skilled chefs?

Unlike the questions explored by most academic research, this is one to which almost everybody is eager for an answer. Despite its nasty historical provenance in property ownership, Western egalitarian marriage has weathered everything from psychoanalysis to the sexual revolution, surviving the divorce-happy 1970s as central as ever to the majority of people's emotional experience. Even today, more than 85 percent of the public still ends up tying the knot. If anything, the debate over gay and lesbian marriage has only made the institution's status as a basic human experience even more apparent.

Meanwhile, right-wing politicians call for tighter divorce regulations; evangelical Christians counsel wives to submit lovingly to their promise-keeping hubbies; and feminists suggest men stop shirking the "second shift" of housework. Writers such as John Gray and Deborah Tannen pose the problem as one of communication: Men and women, they suggest, speak separate languages.

This is a tough atmosphere in which to maintain a disinterested perspective. And that's especially so for Gottman and his wife, Julie, who, like many academics of their generation, come from activist backgrounds. John, a conscientious objector during the Vietnam War, developed an educational counseling program for migrant high-school dropouts; Julie worked with drug addicts and schizophrenics and did some of the earliest research on the long-term psychological health of adult children of lesbian mothers. But their very different approaches--Julie is an empathic, dangly-earringed therapist; John a former mathematician with a trim beard and a way of jumping up and down when he talks about vectors--have made their collaboration an unusual one. Their shared project, the Gottman Institute, combines his theory with her practice. To critics, the institute is the progeny of a shotgun wedding, performed under the gun of publicity. But to Gottman, it is "a legacy, so that after I die, there's really a place where research gets blended with clinical work... that builds theory, that does really good science, that provides good treatment."

Originally trained in hard-core quantitative methods, Gottman's earliest interest was not in marriage at all but in finding new methods for evaluating human interactions. An important breakthrough, he said, came early in his career when he designed a study that debunked popular assumptions about how children make friends. Conventional child therapists, says Gottman, based their approach on misguided adult assumptions: Go up to Jimmy, they suggested to shy kids, say hi, and ask him if he wants to play. But according to Gottman's videotapes of children interacting, the "master" friend makers sidled in slowly, picking up cues rather than awkwardly interjecting themselves. Drawing attention to oneself, it turned out, was the worst possible way of making a connection--a finding that emerged from studying the kids who were successful at the task rather than just the ones who were having trouble. A major influence Gottman cites is the work of child psychologists Fritz Redl and Haim Ginott, who advocated studying emotional interactions as they happened: analyzing a fight in real time, say, rather than asking for participants' opinions after the battle is over.

Still, Gottman says, he found it rough sledding when he began studying married couples. "When I first started doing this research and submitted articles, they wouldn't even review them," he recalls. "They said, 'This isn't psychology, this is sociology.'" After his first year as a professor at Indiana University in 1973, his promotions committee warned him that he was unlikely to get tenure because the field of marital research was "too soft." Gottman grins. We sit in his office, which overflows with books; there is also a microwave oven, a treadmill, a jam-packed year 2000 wall calendar, and colorful drawings by his and Julie's nine-year-old daughter. "You can't really do science about marriage," he remembers his advisers saying. "Personality theory hasn't even solved the problem of the individual."

Although marriage counselors had been trying to patch couples together for decades, their work was not grounded in any established theory about how happy marriages worked. Freud himself advised against working with couples.

"YOU CAN'T REALLY DO SCIENCE ABOUT MARRIAGE," GOTTMAN REMEMBERS HIS ADVISORS SAYING. "PERSONALITY THEORY HASN'T EVEN SOLVED THE PROBLEM OF THE INDIVIDUAL."

It wasn't until about seven years ago that Gottman's research really began to pick up steam. In collaboration with the UC-Berkeley psychophysiologist Bob Levenson, Gottman combined the be-here-now videotape approach with the measurement of bodily responses. According to Levenson, people experiencing fear, for example, have cold hands; those who are disgusted, hot ones. Anger triggers a rapid heart rate, disgust a slower rate. The researchers also began using the UC-San Francisco psychologist Paul Ekman's well-known Facial Action Coding System, a.k.a. FACS, which distinguishes a physical signature for each human emotion. Ekman's data have shown such physiological patterns to be universal across cultures, and the use of his techniques to analyze videotapes has yielded remarkable results: In one study, researchers were able to distinguish suicidal from nonsuicidal depressed patients by freeze-framing their fleeting expressions of contempt and disgust.

Using FACS, Gottman and Levenson scrutinized their videotaped couples for nuances of expression--picking out a genuine smile from a fake one, say, by watching for an associated eye squinch. Contempt, one of the emotions Gottman watches for at the Love Lab, is detectable through the "dimpler muscle," or buccinator, which pulls the lip corners to the side and creates a distinctly uncute dimple; it's often combined with an upward eye roll. From his FACS data, Gottman arrived at some startling conclusions. For instance, he found that the frequency with which husbands displayed contempt during an interaction was highly correlated with the number of infectious illnesses their wives would contract within the following four years. (The prediction didn't work when the genders were reversed, however, unless loneliness was factored in: Lonely husbands of contemptuous wives had a higher rate of illness than other husbands.)

To collect more data, Gottman put together the room the media later termed "The Love Lab," with its ear-clip and finger-clip heart monitors, respiration testers, electrocardiograph, and charmingly low-tech "jiggle-ometer"--a device that measures how much a nervous participant is jiggling about in his or her chair. A video monitor faces each chair. Couples in the lab are asked to discuss distressing issues or to tell the story of how they met; a box of tissues sits by each chair for teary moments. Gottman also uses a version of this setup in his clinical office, where he counsels couples.

As with the friendship research, Gottman notes, his emphasis is on distinguishing the masters of matrimony from the disasters--learning the elements of successful wedlock from those folks who are getting it right. He has tested newlyweds, couples in their first seven years of marriage, forty- to sixty-year-old couples, and those in violent marriages. What he found was surprising and sometimes counterintuitive. The presence of anger, he deduced, did not indicate an imminent divorce or marital unhappiness--indeed, some of the most happily married couples went at it hammer and tongs. Nor was repression necessarily problematic, as long as both members of the couple abided by a tongue-tied approach. Instead, he found that the best markers of a bad marriage were certain familiar--and often gender-linked--dynamics in marital disputes. A "harsh start-up" (the abrupt and negative introduction of an issue) by the wife was associated with bad outcomes, for example. In response, men tended to become physiologically "flooded," closing down emotionally and physically as a self-soothing mechanism. Ultimately, Gottman found that four behaviors (which he termed "the four horsemen of the apocalypse") were most associated with long-term unhappiness: criticism, defensiveness, stonewalling (retreating into silent unresponsiveness), and contempt. Contempt, he found, was especially corrosive and highly predictive of marital disintegration.

ELSEWHERE ON THE WEB If you're not too keen on Gottman's methods, there are other ways to read your lover. In the Victorian era, being able to read fan-semaphore was the key to romantic insight; The Language of Love reveals what a woman really meant when she opened her fan wide in front of someone else. For those seeking a more superficial take on possible love, The Love Calculator offers instant compatibility analysis based on names and nothing more. And for the most suspicious, Check-A-Date is a way to find out about your potential mate before any nasty emotions get involved.

On a happier note, he discovered that successful repair attempts--tiny bridge-building gestures in the midst of a dispute, ranging from an apology, to a "hey, we're getting off-topic!" to a goofy grin--were crucial chords in composing marital harmony. No matter what style of interaction a couple favored, Gottman found, a simple five-to-one ratio of positive to negative interactions indicated a marriage that would last.

One of Gottman's most widely publicized conclusions was also one of his most widely satirized--the "yes, dear" hypothesis. For a marriage to work, he concluded, the husband had to be willing to accept the influence of his wife when it came to working out marital disagreements. Women, whatever their other problems in relationships, had little trouble allowing themselves to be swayed by their spouses, but men sometimes did, and male recalcitrance doomed a relationship to disintegration. Although the conclusion might seem innocuous, the press ran with it, and Gottman was mocked everywhere from the Dallas Morning News editorial page to Saturday Night Live for seeming to advocate a mass pussy-whipping of the male population.

Within his own field, however, Gottman's most controversial claim lay in his critique of one of the staples of marriage therapy, the "active listening" technique. In active listening, a therapist or group leader encourages one spouse to speak (often using nonblaming "I" phrases) while the other silently listens; the listening spouse then echoes what was said. For example, when a wife says, "I feel ignored and unattractive when you flirt with your ex-girlfriend in front of me," the husband responds, "When I flirt with Lola, you feel ugly and ignored." The goal is to create empathy and increase communication, enabling couples to solve problems by understanding each other's point of view.

Healthy couples, one might assume, can listen and echo each other's feelings naturally. But in fact, claims Gottman, married couples--whether miserable or giddy with happiness--almost never interact this way. "People said what they thought about an issue, they got angry or sad," he wrote in The Marriage Clinic: A Scientifically Based Marital Therapy (Norton, 1999), "but their partner's response was never anything like what we were training people to do in the listener/speaker exercise--not even close." Such exchanges occurred in less than 5 percent of marital interactions, he wrote, and they predicted nothing about whether the marriage would do well or badly. What's more, Gottman noted, data from a 1984 Munich study demonstrated that the exercise itself didn't help couples to improve their marriages. To teach such interactions--whether as a daily tool for couples or as a therapeutic exercise in empathy--was a clinical dead end.

When Gottman's conclusions were published in the peer-reviewed Journal of Marriage and the Family in 1998, they were immediately picked up and disseminated in a wide variety of formats--in part because the journal sent out an accompanying press release, the first time it had done so for a published paper. Soon Gottman was profiled in a Los Angeles Times article that featured (and, Gottman says, exaggerated) the "yes, dear" hypothesis; he appeared on Oprah and began giving keynote addresses at conferences for clinicians. Within a year, he had become the default go-to guy on marital dysfunction, overshadowing other researchers in the field. And his findings seemed to be moving willy-nilly into the clinical canon.

Criticism was not far behind. "Because the findings from this one study had such far reaching effects (which is not typical)," write University of Denver psychologists Howard Markman and Scott Stanley and UCLA psychology professor Thomas Bradbury in the February issue of the Journal of Marriage and the Family, "the methods and interpretations bear close scrutiny."

Markman, Stanley, and Bradbury take issue with Gottman's work on three levels: First, they point out what they perceive as methodological and conceptual flaws. Gottman's Love Lab analyses, they contend, rely too heavily on physiological responses. They contend that some of the Love Lab's observations may in fact be interpretations: To label a particular marital exchange as a "husband rejecting influence," for example, is to build into the data "a gender-power conclusion that may reach beyond the meaning of the behavior being coded." They also question Gottman's conclusion that anger does not correlate with bad marriages, whereas other negative behaviors do. "Simply put: after the codes for defensiveness, belligerence, criticism, contempt, and stonewalling absorb their variance in this coding system, what is really left to label as anger?" Finally, they point out potential problems in statistical methods and choice of subjects, claiming, for example, that Gottman erred in creating a pattern of subjects that emphasized masters and disasters (especially good and bad marriages) rather than forming a representative bell curve of marital satisfaction. By doing so, they claim, he exaggerated his predictive abilities.

Gottman's rejection of the active- listening approach comes in for particularly sharp criticism. "As nowhere else, the effects of the interpretation of these data have been far reaching," huff Stanley, Markman, and Bradbury. "We have heard from marital therapists and marriage educators from all around the world wondering if they would be wise to abandon such strategies based on reports from this study...we believe the conclusion to abandon active-listening strategies is entirely unwarranted." For one thing, they argue, Gottman overstates the use of such therapies, making them sound more rigid and universal than they actually are. Moreover, they point out, just because couples don't naturally conduct active-listening doesn't mean such an exercise won't help them: "The argument Gottman et al. put forth is tantamount to concluding that aspirin would be an ineffective treatment for pain relief because we do not see elevated levels of acetylsalicylic acid in people without headaches."

Nonsense, Gottman shoots back. In his reply to Markman et al.'s critique, he defends his masters-and-disasters approach as especially well suited for "meaningful comparisons"; lists biological and emotional distinctions between anger and other negative emotions; and clarifies the basis for his definition of rejecting influence. Gottman and his critics quote different studies to back up their views on the effectiveness of active listening. But Gottman pointedly notes that "nothing in the area of psychological treatment ever dies easily, and we clinical researchers usually do not seem to have the courage to say that the Emperor has no clothes."

AGAINST GOTTMAN, MARKMAN AND HIS COLLABORATORS INSIST THAT ANGER IS A SIGN OF A FAILING MARRIAGE.

There's nothing unusual, of course, about professional disputes over proper methodology and study design. They're a natural (and desirable) outcome of the scientific method. But what aggravates this debate is that almost everyone involved has something at stake financially--not just research funding, the usual academic sticking point, but book contracts, workshops, and brand-name marriage-saving techniques. Even hard-core researchers in this field offer therapy to couples, and many are affiliated with one self-help program or another. Markman and Stanley, for example, conduct the Prevention and Relationship Enhancement Program (PREP), which, as they acknowledge in their critique, uses the active -listener exercise. (In PREP this component involves passing back and forth a piece of linoleum labeled "the floor.") Other marriage-manual authors include SUNY Buffalo psychology professor Frank Fincham, whose cognitively based tome, co-authored with Keith Humphreys and Leyan Fernandes, is titled Communicating in Relationships: A Guide for Couples and Professionals; Seattle Pacific University psychologists Les and Leslie Parrott, who offer a Christian marriage mentor program that matches young couples with older ones; and UCLA's Andrew Christensen and the late Neil Jacobson, whose Reconcilable Differences was released in February.

For all the disagreement between Gottman and his interlocutors on questions of quantitative methodology, their therapy programs have much in common. What Gottman calls "the four horsemen," Stanley and Markman term "danger signs." While Christensen and Jacobson emphasize acceptance and integration, Gottman talks about improving the underlying friendship of married couples. Some programs emphasize cognitive tasks; others focus on behavior; still others on psychodynamic family history. But as several studies have found, most counselors are instinctively eclectic, whatever their training: Behaviorists may still ask about family history, and psychodynamic therapists model better cognitive habits. (Diane Sollee, a lobbyist for the Coalition for Marriage, Family, and Couples Education, flippantly terms the options in the field "a hamburger with different fixings.")

Western marriage counseling itself is a relatively young field. In the 1930s and 1940s, popular marriage guides offered succor for those making the transition to a "companionate" (in other words, emotionally intimate) marriage ethic. Gynecologists, educators, and clergymen offered conjugal advice to individuals. But it wasn't until the 1950s and 1960s that therapists began to work with both members of a couple. As feminism and the self-help movement elbowed their way into the culture, marriage counseling adjusted to the changing times. Whereas in the 1950s a therapist might have considered it his role to salvage the marriage no matter what, often by encouraging the wife to accept her traditional role, in the 1970s a marriage therapist was more likely to see a potential for personal liberation in divorce. In the last twenty years the cultural pendulum has swung back, toward a more commitment-friendly approach. Meanwhile, quantitative evidence that any of these approaches can actually help married couples is still scarce.

Indeed, the numbers do not always go the way marriage therapists would like them to. Although Neil Jacobson, one of the most prominent figures in outcomes research, originally thought couples counseling paid off, when he reanalyzed his data in 1984, he found that he had in fact massively overstated therapy's effectiveness. Only 35 percent of couples, it turns out, were in the nondistressed range at the end of the sessions; and two years down the line, 30-50 percent of that number had relapsed into a state of distress. Only 11-18 percent of couples, in other words, had a better marriage as a result of marriage therapy. Much of the data had been muddied by confounding factors, such as the patients' desire to pat the therapist on the back for trying--couples would say exercises had been helpful even if their marriage was just as miserable as ever. Jacobson found that the largest statistical difference involved the contrast between counseled couples and those in the control groups; those in the control groups deteriorated at such a rapid pace they created the false impression that therapy was improving marriages--when it in fact just held them at the same unhappy level.

The problem, as Gottman acknowledges, is that knowing how marriages work doesn't necessarily help clinicians--or couples. "In marriages that are happy and stable, for example, there's a lot of positive affect when people are resolving conflict. They laugh a lot, they're teasing each other, they give appreciations of one another, even while they're fighting," he says. "That's a stable difference between happy and unhappy couples: It predicts divorce and marital happiness." Unfortunately, he adds, it's also a useless finding, because it's unteachable: "You can't take an unhappy couple and say, 'Have more affection and humor while you're fighting.' People have tried to do that, with all kinds of methods. They'll say, 'Pretend that you're happily married! Be nice to each other! Remember, you love each other, right?' 'Well, yeah.' 'Well, act like it!'" Alienated couples simply can't pick up these skills, says Gottman, unless they have already built an underlying friendship.

Given this depressing conclusion, why even step out of the lab? Gottman's turn to doing real-life marriage treatment came in part because he was willing to accept influence from his clinically oriented wife. "His interests were primarily research," recalls Julie. "Observing as opposed to fixing. And here I was this fix-it woman." (John had done a bit of marriage therapy early in his career and had found it a miserable task: "The couples weren't happy to be there, and they weren't grateful to me or my students. And then you look at the data, and you have this temporary effect. I found it very unpleasant.")

From the earliest point in their marriage, Julie said, the couple spoke frequently about bridging the gap between his research and her clinical work. As his research progressed, John became increasingly persuaded that his findings suggested strategies that could help prevent couples from spiralling into patterns of ever-escalating conflict.Then in 1996, while on a canoe trip, the Gottmans began to talk in earnest about taking John's findings to the public. "He had made a significant shift in really wanting to take his research and do something to help couples," recalls Julie. "He was beginning to conceptualize a therapy which was based upon his predictions about what helps couples stay together. I jumped in with my ideas as well. And it was like, The Happy Collaboration!"

There were also pragmatic concerns. The canoe brainstorm began, says John, when he and Julie were discussing their dream of buying a country house on idyllic Orcas Island. Indeed, John had begun to recognize that there was a lucrative market for scientifically based marriage advice--a market in which his research could be a mouthwatering commodity. One eye-opening experience came when he was asked to speak at a local Christian college; too busy to accept, he attempted to put the school off by asking for what he viewed as the exorbitant sum of one thousand dollars. The school accepted. After the workshop, he ran into a clinician who revealed that she had paid $190 to attend the 180-person session. "I did the arithmetic. And I thought, you know, this is the operational definition of a moron!" he recalls. "Someone who makes one thousand dollars while they make all this money."

Still, leaping into the marketplace involved an element of risk, Gottman knew. His colleagues might view him as a sellout, and he might risk his reputation as a disinterested academic--and with it, his funding. "My last book was on the New York Times best-seller list three times. And I've been on all these TV shows. And it pisses people off!" he says. "They think you're arrogant, and you're wealthy, and you don't need it, and you're building an empire, and why aren't they doing it, and why aren't people calling them? Not everybody, just a minority of people." Still, he says, "one person's enough to put a kibosh on a project."

What's more, as Gottman was aware, linking an entrepreneurial project to his academic work inevitably raised the problem of conflict of interest. "There's a danger in marketing a program and then having a commitment to that program," he concedes. "That you end up selling the program regardless of what the data show. But all you have to do, I think, is to recant what you've proposed whenever the data show that you're wrong. And I've done that, several times." In fact, he says, when he went back and looked at the hypotheses he's made since the 1970s, he found that 60 percent of the time he had been proven wrong. The most striking example: In his first self-help book, Gottman recommended the active-listening technique. Being wrong but clear is central to science, he says. "As long as you're interesting, and as long as your hypotheses are testable, you make a contribution. Because even if the data show the theory's not correct, you can still modify the theory, and then that's going to be interesting. So it's a no-lose situation, I think."

Even if Gottman's academic work sticks to the careful language of the hypothesis tester, however, his self-help books offer easier answers. For instance, whereas Gottman cautiously qualifies his findings in his article in the Journal of Marriage and the Family, establishing which points require further study and pointing to potential weaknesses in his methodology, the approach he takes in The Seven Principles for Making Marriage Work is an entirely different matter. "It used to be that couples could achieve [intimacy] only through their own insight, instinct, or blessed luck," crows the first chapter. "But now my Seven Principles make the secrets of marital success available to all couples." Of the questions that people ask about marriage--including "How can you prevent a marriage from going bad, or rescue one that already has?"--Gottman writes, "After years of research I can finally answer these questions."

Of course, kvetching about a how-to book's catchphrase certitude is like complaining that a sonnet has too many end rhymes. Self-help books are by definition supposed to offer help for the self, not establish the groundwork for fresh research. And Gottman's excitement about helping people is clearly genuine; he especially relishes the give-and-take with clinicians, whose emotional engagement he says he finds refreshing after decades of grad-student hauteur.

Gottman acknowledges that there are dangers in linking his expert status to a how-to program. "One of the problems is that if the couple fails, if the tapes or the workshop doesn't help their marriage, then they blame the marriage, because the expert said this was surefire. And I think that's a real issue." Nonetheless, he says, the difficulty of translating one's findings into popular language is worth it. In the nineteenth century, he points out, the culture was quite different: "Darwin's books were best-sellers. Scientists were expected to speak to the public." If you want to get your data out there, Gottman says, and do some good, that may mean presenting your findings in a palatable, and significantly simplified, form.

USING EKMAN'S CODING SYSTEM, GOTTMAN WAS ABLE TO DISTINGUISH A GENUINE SMILE FROM A FAKE ONE.

Clinician and pop psychologist, academic and self-help guru, John Gottman is engaged in a complex juggling act. And it is about to get even more unwieldy as the Love Lab concludes its most controversial study yet: the first observational study of the interactions of gay and lesbian couples, a longitudinal project twelve years in the making. The findings, Gottman says, are both fascinating and tentative. (He and Levenson have not finished analyzing their data--but Gottman has divulged some of the study's findings in speeches.) "Initially, we weren't interested in gay men and lesbians at all," Gottman explains. "We were really doing it to see whether the differences we were seeing between husbands and wives were gender differences or role differences. Do lesbians stonewall? Do two guys get critical when they fight?"

But while he found a few gender distinctions (gay or straight, women tend to become sadder than men), Gottman was most startled and excited by the strikingly upbeat dynamics in homosexual relationships. "There's a 'reverse economy' in the diagrams," he exclaims. "Not only do they start off more positive, but the positivity has a much greater impact, and the negativity a much lesser impact." Gay relationships, in other words, did not need five positive interactions to blot out each negative one. On the other hand, Gottman says, these couples had more trouble repairing things once they did go sour, a built-in fragility that diverges from straight relationships. Gottman was also startled by the gay couples' candor in talking about sex--"they can really talk to each other openly about what they like and dislike."

Gottman is clearly exhilarated by this fresh subject matter and by the opportunity to explore a new population of couples. (He was reading The Gay Metropolis when I arrived.) But for a while, the gay and lesbian data created a dramatic internal debate at the institute: Julie, for one, was worried that the research might make their family a target for the right wing and potentially endanger the fledgling institute. After hashing it out (with a fair bit of active listening!), the couple resolved their differences, and John seems eager to finish the mathematical analysis. The conceptual issues, he says, are particularly interesting: Without simple husband and wife slots, the researchers decided to code the interactions depending on which partner "initiated" and which "received" marital issues.

Still, it's only one study. And Gottman has numerous qualms about its statistical design and sample selection. (Couples were required not to have been married before, to be within the same age range, and not to have children, among other things.) Not to mention the cohort concerns: Is it possible to isolate distinctive aspects of gay and lesbian relationships when the cultural tide shifts daily? Nonetheless, the Gottman Institute has already begun a speeded-up cycle of data-to-therapy: It has established a task force of gay and lesbian therapists and is preparing to offer workshops for homosexual couples. "We're about helping people to have good relationships," says John. "We take a moral stand only on that. We honor love when we find it."

Gottman's next step? Infomercials--and a call for an on-campus Relationships Department. He's met with a marketing consultant on the first. The second, he says, will have to wait until enough support can be gathered. The new discipline would bridge the gap between psychology and sociology, focusing neither on the individual psyche nor on big-picture social structures. The intricate exchanges between people, says Gottman, have been sorely neglected in theorizing human experience. He cites the Australian zoologist Karl von Frisch's "dance of the bees": Put one bee in the lab, and you learn nothing; gather many bees, and you witness the figure-eight dance that enables bees to learn how to fly with respect to the sun. Courses would cover families, friendships, parent-child relationships, even workplace dynamics. Gottman's goal, he says, is to create "relationships experts." "Some would wind up being labor-management mediators. Some would wind up flying to trouble spots and acting as mediators between warring parties. Some might be marriage counselors, child therapists. There are all kinds of possibilities: Some could be business consultants."

In the end, Gottman hopes, the gradual accretion of data will produce that most-wished-for result: a bona fide tool kit for domestic happiness. "I've always thought of what we were doing as designed to be helpful to people," Gottman muses. "And I think that a lot of what's wrong with relationships is that this information is just not known. Eventually everyone is going to know this stuff. And nobody's going to know where it came from. It's sort of like brushing your teeth or flossing--who thought of flossing? Who thought of this stuff? Nobody remembers who thought of it. It's just a good idea. And it gets absorbed by the culture and everybody does it and you know"--he smiles warmly--"people have better teeth."