Harvard psychologist Jerome Kagan argues that inborn temperament stays with us through our lives. But critics say his own ideas are less than consistent.
ATTENTION, NEW PARENTS. Jerome Kagan, a professor emeritus of psychology at Harvard, has devised a fun little test for you. At 4 months old, plop your baby into a bouncy seat and present him with a series of colorful new toys - ones he's never seen - one after the other, for 20 seconds at a time. Does he cry madly and shake his arms and legs? If yes, be forewarned: Your baby may be at higher risk for "developing serious anxiety over social interactions" a decade down the road.
If he screams at 4 months, he'll be more likely to stay home from junior-high dances. If he screams, he'll be more likely to answer "no" when a psychologist asks, at age 11, "Are you happy most of the time?"
It won't really matter if you cuddled your child as an infant or showered him with play dates as a toddler. He'll probably never be a brash CEO or politician, although he might become a brilliant solitary researcher or a melancholy poet.
On the other hand, if your baby just stares calmly at the toys, he will be calm on dates but also slightly more likely to become a delinquent, because parental threats won't faze him.
Although he would hate to hear it put this way, biology does sound an awful lot like destiny in Kagan's new book, "The Long Shadow of Temperament" (Harvard), which he co-wrote with his colleague Nancy Snidman. It comes as the much-honored professor - who in 2002 ranked 22d on one psychology journal's list of the top psychologists of the 20th century, one notch above Carl Jung - prepares to shut down his famous child-development lab at Harvard after four decades. (Kagan retired four years ago and Harvard needs the space. Another infant lab remains open.)
The ancient Greeks were right, Kagan believes. There is such a thing as temperament - although his discussion of innate personality traits relies on EEG probes and brain-stem activity, not any musings on the four humors. The book centers on studies that Kagan and Snidman began in 1986 with 500 infants. Roughly 20 percent of the babies who screamed at toys and other unfamiliar stimuli grew into 11-year-olds who were shy with interviewers and who showed biological signs of alarm in stressful situations. By contrast, 33 percent of the calm, cool tykes grew into composed, sociable preteens - "Clint Eastwood types," as the authors put it, in the case of some boys. "A 45-minute lab observation of 16-week-old-infants revealed dispositions that were preserved in some children for over 10 years," the authors write.
Kagan's revival of the old idea of temperament - one he has written about, with less data, over the last decade - doesn't always go over well in developmental psychology, where the shaping power of the environment is still king. In particular, he's caused some backs to stiffen with his staunch rejection of "attachment theory" - the idea, pioneered after World War II by British psychiatrist John Bowlby, that the bond between mother and infant, as measured in the first year, plays a key role in later emotional and even intellectual growth.
Attachment theory holds great sway in the public mind, thanks to popularizers like the bestselling "attachment parenting" guru William Sears. It's one reason some doctors rush to get infants into their mothers arms immediately after birth, and why lactation consultants tell moms to breast-feed skin-to-skin. Yet Kagan says what's being measured in the most famous attachment studies is often just temperament, which he says is largely inherited, or shaped in the prenatal environment, and only moderately alterable by parents.
An old friend of Kagan's, Lewis P. Lipsitt, professor emeritus of psychology and human development at Brown, echoes many when he objects to "the extremity of his position" - his "overemphasis" on hereditary and biological factors. (In the new book, Kagan even endorses the idea that introverts are more likely to have certain physical characteristics, like lean builds and blue eyes.) Yet whether or not Kagan is right about the importance of temperament, Lipsett says, "he's gotten the dialogue going and has incensed enough people to get them working harder on the issue."
Indeed, attachment theorists are not the only ones to bear the brunt of Kagan's polemical energy. He condemns evolutionary psychologists (who see Darwinian selection at work in many human psychological traits) as apologists of "selfishness" and has directed withering fire at Judith Rich Harris, author of "The Nurture Assumption" (1998), who argues that parents play little role in shaping their children's basic personality traits. (Peer influence, she believes, mold what's not shaped by genes.) "I am embarrassed for psychology," he told Newsweek when her book came out in 1998.
Yet not everyone thinks Kagan's various polemical attacks are consistent. Robert Plomin, a professor of behavioral genetics at King's College London (whose work Judith Harris cited approvingly), says: "Although he was very anti-genetic for quite a long time, he seemed to switch to a pro-genetic stance in the 1990s. We published several papers together showing genetic influence. But now he seems to be becoming less genetic."
"It's hard to know," Plomin concludes, "if he is a bellwether or just blowing in the wind."
To some, Kagan is an iconoclast, to others the ultimate defender of the psychological establishment. But whichever it is, the arguments that swirl around him serve as proof of just how unsettled the field of developmental psychology is.
y Kagan's own account, the study of infant psychology was a "wasteland" when he joined the field in the 1950s. Not long after graduate work at Yale, he found himself at the Fels Research Institute in Yellow Springs, Ohio, where he became involved in one of the first grand longitudinal psychological studies. Kagan's task was to study a group of just over 100 white males and females, born between 1929 and 1931 - tracked afterward by the Fels center - and to chart the consistencies of personality traits and the effects of parenting styles. The resulting book, published in 1962, is still cited by scholars for, among other things, its finding that young boys and girls with the most stereotypically masculine and feminine styles carry those traits into adulthood.
Harvard recruited Kagan in 1964 to start up a human-development program in its psychology department, and the nearly 400 papers he has produced since then pepper every developmental psychology textbook. Kagan has shown, for example, that young children have distinctive thinking styles - some leap to conclusions, while others ponder - and that even babies in their second year show signs of grappling with their consciences. In 1978, he produced one of the first studies showing that good day care did not harm the development even of children who entered at 3 months of age.
An "epiphany," Kagan says, came in 1972, when he studied infants in a poor village in northwest Guatemala, where children were kept from almost all social contact the first year of their life because parents feared evil spirits. At 1, they were pathetic in appearance and - to American eyes - retarded in physical and mental development. "But when you walked around the village, you saw that at 5 they were as lively as Cambridge kids," he says. Kagan started to think the way infants were treated in the first months couldn't be as important as psychologists thought.
Now he looks back in embarrassment at the degree to which the notion of the blank slate held sway only a half-century ago. "In my first academic job I taught hundreds of students that you could produce an autistic child if a mother was cold and unresponsive," he says, an idea discredited a generation ago by genetic studies. "That is incredible."
His first experiment on temperament came in 1979, but the present project began seven years later, when he and Snidman began dangling those toys in front of the 500 babies. Twenty percent of the babies showed distressed "crying and vigorous pumping of the legs and arms, sometimes with arching of the back" on at least 40 percent of the trials. These were classified as "high reactive." Forty percent showed little motion or emotion and were dubbed "low reactive." The rest fell somewhere in the middle.
Many of the children returned to the lab at 2, 4, and 7 for follow-up tests. Finally, between ages 10 and 12, 237 of the original subjects got a full battery of brain scans, heart-rate analyses, and body-temperature readings, both at rest and during moments of stress, as when they were asked out of the blue to give a speech.
The key finding was that about 20 percent of the high reactives preserved both their dislike for the new and strange, manifested now by shyness, and also gave off biological signals that they were stressed. A third of the low reactives remained calm and collected in the face of strange situations. Many infants drift toward the gray middle of behaviors - a sign of the importance of parents and the environment, Kagan thinks (though many high reactives who appeared outwardly calm betrayed anxiety when hooked up to electrodes). Yet only 5 percent in each category switch from one to the other extreme. You can't turn a screamer into a cool customer.
hile these findings are interesting in themselves, the study may gain more attention for the theoretical arguments Kagan draws from it, for it constitutes yet another weapon in his longstanding battle against attachment theory.
The classic attachment experiment, developed by Bowlby's protege Mary Ainsworth and known as the "Strange Situation," examines how distressed a child gets, in the lab, when he or she is left alone, or with a stranger, by her mother. Children who cry the hardest, and who later resist their mothers' attempts to comfort them, are graded as less securely attached than those who cry less and settle down quickly.
Yet it's possible, Kagan says, that those poor kids "could have a wonderful mother, and made a very secure attachment with her, but they may have a high-reactive temperament, and so they couldn't be soothed." Attachment theory, he says (echoing its feminist critics), places unfair blame on moms.
Everett Waters, a professor of psychology at SUNY-Stony Brook, says that attachment theorists knocked down Kagan's arguments years ago. If attachment studies measured temperament, he says, researchers would not have found that the bonds between mother and child change when Mom experiences a stressful life change, for example.
Waters believes temperament is a real thing that interacts with parenting styles to determine personality - and wishes Kagan had left it at that. "That's important," he says, but "it's not as interesting as throwing over John Bowlby and Mary Ainsworth," alluding to Kagan's own combative temperament.
In a way, it is curious that Kagan has taken such a consistently all-or-nothing view of attachment theory, since he has been highly critical of scholars who take similarly absolute positions in other areas. Judith Harris, for instance, a successful writer of child psychology textbooks who received a master's in psychology at Harvard before Kagan got there, argues that modern psychologists have failed to prove any influence of parents on their children's basic personality, beyond the genetic. Invariably, she says, the studies that Kagan cites to demonstrate parental effects fail to sort out genes and local environment.
Harris's work is "total nonsense," Kagan says today, reiterating his earlier harsh words. Her book, he says, "was a media event. It had nothing to do with science. If the media had not hyped it - partly because it was so crazy - nobody would know about her."
Yet Harvard psychologist Steven Pinker, author of "The Blank Slate: The Modern Denial of Human Nature," calls Kagan's attacks on Harris "a classic example of what the historians of science talk about when they say the establishment tries to dismiss radical challenges." He adds, "I actually found Jerry's reaction puzzling. There's no reason for him - himself having been an advocate of the effect of genetics - to insist so dogmatically on the importance of parents. I think the data are against him there."
Harris herself, reached by e-mail at her home in Middletown, N.J., says that studies of twins reared apart "fail to support Kagan's belief that the right kind of parenting can make a child less timid." Twins reared apart, she says, are no more different in basic temperament than twins reared together.
Nathan Fox, a former student of Kagan's, now a professor of human development at the University of Maryland, says that a recent study by Avshalom Caspi of the University of Wisconsin proves the greater wisdom of Kagan's position. The study, which Fox says represents "the future of psychology," demonstrated that mistreated boys who inherited a highly active version of one gene crucial in brain chemistry later proved to be more resilient than boys with a weak version. "That is what Kagan has been talking about," Fox says - the delicate interaction between parental behavior (abuse, in this case) and genes.
Harris has her own problems with that study and says she has never claimed that abuse doesn't matter. But whatever further research shows, Kagan is unlikely to be shaken by the brickbats that come his way. He had some anxious days as a kid, he recalls. But this Harvard legend has since been self-diagnosed as a low reactive - the cool, composed, Clint Eastwood type.
Christopher Shea writes the biweekly Critical Faculties column for Ideas. E-mail: firstname.lastname@example.org.