You witness Bedford Falls save George Bailey for the 20th time in It's a Wonderful Life. You mourn the loss of a parent. Your boss humiliates you in front of your colleagues. You break up with your boyfriend or reunite with your sister after many years. You hold your newborn in your arms.

Chances are, especially if you are a woman, that the overflowing emotions will trigger a switch. You will weep.

"Why the heck do people cry?" asks James Gross, a clinical psychologist at Stanford University. "It is such a weird thing to do. You get upset and water comes out of your eyes."

The answer lies in a complex web of psychology, biochemistry and evolutionary biology. Although many books have been written on the topic, scientists can only speculate on the purpose of emotional tears. These speculations range from no reason at all to removal of stress hormones to a form of communication.

One of the early speculations comes from Charles Darwin in 1872. In one of his last books, The Expression of the Emotions in Man and Animals, Darwin dismissed any purpose to emotional tears.

He had observed in excruciating detail how weeping children contract the muscles around their eyes during prolonged screaming. Darwin attributed this involuntary behavior to a means of protecting the eyes by compressing the engorged blood vessels.

Darwin believed that tears are "an incidental result" of pressure exerted on the tear glands beneath the eyelids due to this tightening of the muscles, like the tears that result from a blow to the eye. Darwin also noted that the involuntary contraction of the same muscles results in tears when you laugh or sneeze hard.

The Right Chemistry

William Frey, director of the Alzheimer's Research Center at Regions Hospital in St. Paul, Minn., and author of Crying: The Mystery of Tears, attributes much more to emotional tears. He speculates that tears remove stress hormones from the body.

Frey spent many years studying the biochemistry of human tears. He made people cry by showing them clips of sad movies and collected a very small quantity of their tears in test tubes. Then he compared the amount of proteins in emotional tears to those you might cry in response to eye irritation by onions.

"Something unique is happening when you cry emotional tears," Frey says. "Emotional tears have more protein in them."

Frey believes that crying may be a way of alleviating stress by removing chemicals that build up as a result of it, although the exact chemistry is unclear. Analysis of tears is very difficult because of the presence of detergent-like molecules that help tears to spread over the surface of the eyes, he says.

Removal of stress-related chemicals by tears would explain why people say they feel better after crying, according to Frey. In studies done through extensive questionnaires, 85 percent of women and 73 percent of men reported that they felt better -- less sad or less angry -- after crying. "When people talk about crying it out, that might literally be true," he says.

Ad Vingerhoets, a clinical health psychologist at Tilburg University in the Netherlands and coeditor of the forthcoming book Adult Crying: A Biopsychosocial Approach, throws in another biochemical speculation.

He wonders whether crying stimulates release of endorphins, substances in the brain that have mood-elevating and pain-relieving effects. "Then it really makes sense to cry when you are hurt," he says, "because it stimulates pharmacological substances that are in the body."

HORMONAL EFFECTS

Women cry five times more than men and do so with tears running down their cheeks. By contrast, 70 percent of male crying episodes are the type in which their eyes fill with tears.

This gender difference appears around age 12 or 13. Frey believes that this is due partly to anatomical differences at the cellular level between the tear glands of women and men. He attributes these differences to hormones, in particular prolactin, which is involved in the menstrual cycle, breast development and lactation.

Alternatively, Vingerhoets says, male hormones such as androgens and testosterone may inhibit crying. Such hormonal effects also could explain individual differences in crying frequency or why women become more prone to crying during menstruation, pregnancy and soon after giving birth.

Marc Hauser, an evolutionary psychologist at Harvard University and author of The Evolution of Communication, also speculates that emotional tears serve a purpose. He believes that tears accompanying human crying signal the honesty of the emotion.

Hauser finds it striking that, even though monkeys and apes share so many of the emotional accompaniments of humans, such as grimaces and fear expressions, they do not weep. Mammals cry only vocally, just like newborn humans in the first weeks of life. They do have tear glands and tear up if you injure them.

There is no good answer as to why the connection between emotions and tears was not made in animals, Hauser says. Instead, he asks why the connection was made in humans.

"It must be telling something about the evolutionary pressures that act on humans for certain kinds of emotional expressions," he says. "Why is it that we have evolved one that leaves a visible trace?"

Nobody has really grappled with this question, Hauser says. He suggests that, since crying with tears blurs your vision and you can't hide the red eyes and nose, it is a signal costly to produce. "Therefore, it's honest," he says. It also is very difficult to cry intentionally with tears. Actors have to work very hard to do it.

Randy Cornelius, a social psychologist at Vassar College and author of The Science of Emotion, thinks similarly. He speculates that "tears are communicating something very important in depth of feeling."

Cornelius came to this conclusion through experiments in which his team showed pictures of crying faces, with or without tears, to two groups of people. Using computers, they erased the tears.

"The tearful faces are seen as much more emotional and much more likely to express sadness," he says. "When you remove the tears from a crying person's face, judges attribute all kinds of emotions to them."

This depth of feeling can elicit social support from other people, Cornelius says. Even when we cry alone, he says, we have ourselves as an audience, as well as the internalized views of other people. "It's social support that makes us feel better after we cry," he says, "If we don't get that support, it's unlikely that we're going to feel better."

Also, a negative response may have the opposite effect. For example, crying in the workplace can be disastrous when others perceive you as weak and not up to the job, he says. Hence, people often cry after work, between 7 to 10 p.m., when they are with their significant other. They are more fatigued in the evening, Frey says, and have a lower threshold for releasing their emotions.

The Stress Factor

Stanford's James Gross, editor of Functional Accounts of Emotion, also speculates that crying is a social signal. He studied how crying affects the nervous system. He and his collaborators wanted to see whether "crying it out" also released tension built by stress in the nervous system.

Gross showed three- to four-minute clips of the movie Steel Magnolias to 150 women. The clips portrayed a mother crying after her young daughter dies. As the women watched the film, Gross's team measured how their autonomic nervous systems responded.

The autonomic nervous system controls involuntary events such as heart rate, sweat gland activity and skin temperature. One branch, called the sympathetic nervous system, stirs things up; another, the parasympathetic nervous system, calms things down.

The film clip upset most of the women and affected their sympathetic nervous system. Thirty-three of these women cried. The researchers expected things to calm down at that point. Instead, the opposite occurred. "Crying seems to be associated with heightened sympathetic activation," Gross says.

Perhaps this makes sense because people don't feel calm while they are crying, he says. "We didn't, of course, test whether an hour later, or five hours later or 10 days later, they felt better in some sense."

These results made Gross and colleagues advocate the socioemotional view of crying. You can't understand the function of crying just by looking at what it does to the person who is crying, he says.

"Crying is about sending a powerful social signal to others around us to say that things are not going okay," Gross says. This signal is so upsetting that it motivates others to help, he says. "It is sort of a social glue." People go out of their way, for example, to stop children crying on an airplane, he notes.

Children certainly seem to know how to make use of the crying signal. As infants, they cry to communicate hunger, pain or other distress. "That's a very effective means to mobilize people to comfort the baby and to offer help," Vingerhoets says.

The signal seems tricky with older children. Most people are familiar with the scenario of the child who stumbles and falls and then cries only if someone is nearby.

Hauser says that's not because children can automatically turn their tears on and off. They are often right at the edge of crying, he says. "If they see someone is around, they just let their emotions go."

Hauser agrees that this type of crying may seem manipulative but says all communicative signals are manipulative. You are asking for a response to your needs.

On the other hand, not all adults cry as a signal for help. Vingerhoets says the presence of others may in fact inhibit crying in some cases.

"There is a lot of stigma attached to portraying yourself as being weak," especially for men, Gross says. "They feel very uncomfortable expressing that kind of vulnerability and need."

Suppressing emotions, however, may have physiological consequences. Whether you cry or don't cry with sadness, your sympathetic nervous system shows increased activity, Gross says.

If you cry, people nearby will help you, and the parasympathetic nervous system can take over and calm things down, he says. If you suppress crying, however, the sympathetic nervous system will continue stirring things up for a longer period of time.

"You may need to do that in a business interview or some other context," Gross says, "but if you inflexibly and chronically suppressthese emotions such as anger or sadness, I think that can get you in trouble."

The Bottom Line

Vingerhoets and Cornelius have devised a model to explain weeping: It is a way to cope or bring difficult emotions under control. "It facilitates recovery after you have been in distress," Vingerhoets says.

This recovery can be through psychobiological pathways such as release of endorphins, he says, or through a social or psychological channel because it induces emotional support.

"There is an evolution of society's attitudes," Frey says, "moving towards acceptance of crying and a view that it is not a bad thing, not a sign of weakness and loss of control and self-indulgence."

For example, Maine senator Edmund Muskie lost the Democratic presidential nomination in 1972 after he cried in public, Frey says. But Americans accepted President Clinton crying at the funeral of commerce secretary Ron Brown in 1996 and Bob Dole's eyes filling with tears after his eulogy at the funeral of Richard M. Nixon in 1994.

The researchers themselves exercise the unique human ability to cry. Frey says he cries about once a month. "I only get watery eyes and do not have tears running down my cheeks, except on very rare occasions," he says, "perhaps once every two years."

Cornelius says he cries often. "I almost always tear up when I see a film or story about childbirth," he says. "I cried when my son was born."

Hauser says he cries when he watches touching movies, hears Bach's Double Violin Concerto and says goodbye to his daughter.

Vingerhoets says his crying frequency matches closely the male average of once a month. He says he cried when a colleague suddenly passed away and when he wanted to thank his daughter in his inaugural speech upon becoming a full professor.

Haleh V. Samiei is a freelance science writer and producer of a Web site on nutrition news at www.NutriBytes.com.