As the sounds of children getting off to school evaporate into the fall morning, Maj. Robert Redfield, one of the U.S. Army's chief researchers on AIDS, settles himself onto a shaggy beige couch near the tropical fish tank in his family's Montgomery County home. After a moment of talk, his voice lowers, his speech slows. He is afraid for his children, he says.

Redfield is an infectious-disease officer attached to the Walter Reed Army Institute for Research. The 35-year-old physician is one of the key officers in the military's massive AIDS testing program, under which half a million young people so far have been checked for signs of infection.

Redfield's work has made him confront the scope of the AIDS epidemic in a way few others have. He knows that in Manhattan, one in 50 young people may be carrying the disease. He knows that military data show Maryland -- which still ranks relatively low in fully-developed AIDS cases -- to be running second in the nation in the rate of new infections.

Because of the alarming rate of infection among the young, he thinks we may be seeing the end of normal sexual experimentation by American children for some time to come.

"My parents never talked to me about sex -- never, ever," he says. "But now it is going to be critical to talk to our children about sex. Too many kids learn about it too late. There was the case of the girl [on Long Island], a high school senior. She had only three sex partners in her high school career. But she got AIDS ..."

Redfield has already begun to talk to his twin 8-year-olds about AIDS, and has thought through what he will say to all his children as they approach the age of first sex.

"I will teach my children how critical it is to know whether their partners are infected with this virus," he says. "Above everything else, they must realize that having sex with 'Judy,' they are also having sex with anybody 'Judy' ever had sex with.

"This epidemic is personal to me," he says, and not just because "I'm a doctor and have cared for many patients and have had friends die from the disease. It is personal for me because I see my son, my daughter and my two younger sons ..."

He pauses to gather his thoughts and says in almost a whisper: "I mean this very seriously, not to be sensational -- there is a very serious likelihood that one of my four children will get this disease, in the normal development they have between 12 and 25. That's not hypothetical. It's real."

To Redfield, AIDS is personal in another way. His job, and his strongly held views on the subject, have made him one of the most controversial figures in AIDS research. Recently, he has found himself at the center of a bitter controversy about how much AIDS testing should be done and how public health officials should use the information -- the country's largest mass of data on AIDS infection -- that the military has already compiled.

Combating infectious disease is an old and honored tradition in the Army. One of the great deeds in the history of public health was Walter Reed's quick march on yellow fever in Cuba and Panama. Reed -- like Redfield, an army major and a researcher in infectious diseases -- arrived in Havana in 1900, sorted out the microbiology of the disease and produced an effective campaign against it in less than a year, thus eliminating one of the chief obstacles to the building of the Panama Canal.

When syphilis was a raging epidemic in the United States at the turn of the century -- 10 to 15 percent of the population was infected, according to medical historian Allan Brant of Harvard -- the surgeon general, a military officer, began widespread testing and tracing of infected people and their sex partners in an attempt to control that disease.

Screening in that epidemic was important, but unlike the situation up to this year with AIDS, a treatment was available. Syphilis screening was also not without its cost. It used a somewhat inaccurate test, causing thousands to be wrongly identified as infected.

When the military turned its attention to AIDS -- which surfaced in this country as a disease prevalent among gay men, and which many people incorrectly view as a kind of "gay plague" -- it found itself confronting civil liberties issues in a way it had never had to do in the case of syphilis.

Gays have never been accepted by the armed forces: The Uniform Code of Military Justice, whether enforced or not, still declares homosexuality to be anathema. Early in the Defense Department's program to test soldiers and potential recruits for signs of infection with AIDS, the Navy violated its own procedures by trying to dismiss several gay seamen who had AIDS without written notification or military counsel. So when the massive testing program began in earnest in the summer of 1985, it was seen as a campaign to flush uniformed gays from hiding as much as a program to stop a deadly infectious disease.

Redfield, who believes discriminatory treatment of military AIDS patients has stopped, argues fiercely against such a view. "Some people do not do the right thing," he says, but violation of the rights and the privacy of gays is not what the testing program is about.

Though many researchers have now come to worry about the spread of AIDS in the general population, Redfield was the first to study its transmission by heterosexual contact, to track its spread in families. He demonstrated heterosexual transmission of the disease years before other researchers were ready to accept it.

Dr. Samuel Broder, the prominent AIDS researcher at the National Cancer Institute, says that Redfield "has single-handedly changed the whole direction of AIDS research in this country. He has helped us understand the concept of the heterosexual transmission of AIDS and some of the public health measures that could be used to inhibit that."

Redfield is one of the most vocal advocates, in or out of the military, of the traditional public health view: To fight disease, you must test to find out who has it, then try to prevent those who are infectious from spreading it to others. He shuttles from meeting to meeting, from scientific conferences to local AIDS education groups, presenting the military's unique statistical information on AIDS (see box, Page C2) and, when given a chance to explain it, the necessity of widespread voluntary testing.

In another time or with any other disease, he says, the military's specific information on who has been infected would be put to use. Where it was practical, sexual contacts of those infected would be traced. Those people would be informed of their risk, and offered testing and counseling or treatment. In places where very few infections have shown up, such "contact tracing" can sharply limit the spread of the infection. But in the case of AIDS, public health officers have openly rejected the traditional tracing of partners and screening of people -- such as those who turn up at VD clinics -- with a higher than average chance of having been exposed.

So why would anyone not want to know if he was infected? The answer is fear, mainly. Fear of disease and death, which shows itself first as a denial, a refusal to take the possibility seriously. And fear -- far from groundless, as experience has repeatedly shown -- that news of one's affliction would find its way to insurers, employers, landlords and others who could make life difficult for people with AIDS.

Gays have long understood that their lives could be ruined by a little information reaching the wrong hands. All Redfield's assurances notwithstanding, says Jeff Levi, head of the National Gay and Lesbian Task Force, "gays found to be positive may not be thrown out [of the military] today, but they will never be allowed overseas assignments, and their careers will be cut short." (The State Department, as if to support Levi's point, recently proposed a policy whereby Foreign Service employes testing positive for AIDS would be restricted from service abroad, and applicants testing positive would not be hired.)

The tests now widely used do not actually determine if a person has AIDS; the disease itself is defined by a set of symptoms.

But the tests indicate that the body has made antibodies to defend itself against the AIDS virus. Since the virus is believed to remain in the body permanently once it infects, the tests are shorthand indications that a person is infected and probably infectious, whether or not the full symptoms are present.

The symptoms themselves may appear only years after infection, thus creating the dangerous situation in which hundreds of thousands of people are now infectious and may pass the virus for years before they are aware of it.

Levi and other gay activists, including some physicians treating AIDS patients, argue that testing is not nearly so crucial as Redfield claims.

"It's a very individual choice," Levi says. "Since there is no good medical intervention for those who are positive on the test, only those who feel the need for intervention to change their behavior" will find the test useful.

Wouldn't testing stop people from infecting others? No, Levi argues, because anyone at risk should alter his sexual practices anyway. "Whether you have tested positive or not, you need to practice safe sex. You need to use condoms, mutual masturbation and withdrawal." Millions of gay men, he says, "have changed their behavior without tests, and we have some disconcerting information about some people who knew they were antibody positive but still did not take precautions." As for Redfield's alarming statistical data -- showing that 1.5 in every 1,000 would-be military recruits is infected, and that the disease is now passed widely through heterosexual sex -- Levi says that "all of his research is inherently suspect." Military figures for infections among those who come to recruiting centers, he says, may derive mainly from a small class of young people who are gay or use drugs intravenously, but lie about these things because of the military's strict policies against them.

Redfield, meanwhile, finds arguments like Levi's enormously frustrating. "People don't understand why our public health officials have been paralyzed by these civil liberties issues," he says. "They are important, but knowledge is our only tool right now."

When the military decided to do its testing, he says, he met with officials of numerous state health departments about the program. After outlining the plan, "I would say, 'You guys will probably get the names of those who are positive and go from there ...'

"Then, I would notice there was a little friction in the room. Someone would say, 'Well, why are we going to get the names?'

"I said, 'You are the state health department! If we found the one in a thousand in your community, we could do what anyone would do with a deadly infectious disease. Interview the individual. Follow up his contacts. Interview the contacts. You could get a handle on these micro-clusters of this epidemic.'

"But they said, 'Well, I don't know ...' "

All 50 states, Redfield says, refused to take the names.

Already bald and a little overweight, Redfield doesn't look like the stereotypical military officer, and his positions on civil rights and sexuality are not stereotypical either.

He believes gays have been maltreated by society, have the right to their own ways of loving and should be free from discrimination based on their sexual habits. He believes it is normal and acceptable for children over 12 to have sexual relations.

"Those of us who have advocated testing to the gay community have been represented as people who have homophobia," he says. "In reality, I care very deeply about the men who die from this infection. I think most doctors do."

Born and raised in suburban Maryland, Redfield grew up in a scientific household. His father was a doctor at the National Institutes of Health and his mother still works as a biochemist at a Hoffman-La Roche research lab in New Jersey.

He had no intention of making a career in the military, but his family couldn't pay for his college education, so he funded a bachelor of science degree from Georgetown University through the Army's ROTC program. After he graduated, the Army put him through Georgetown Medical School as well. He emerged from his medical training owing five years of service, but says he will stay with the Army when his obligation is up next June.

In high school, he had spent one summer working in his mother's lab, then two more at Columbia University in New York, helping out in the lab of Dr. Sol Spiegelman. One of Spiegelman's earnest beliefs, for which he had no proof at the time, was that an odd class of viruses called retroviruses would be found to cause human disease. No such cases were known then.

As work in science often does, Redfield's curled back on itself. In 1983, working out of a Walter Reed lab, he was one of the minor players on the team led by Dr. Robert Gallo of the National Cancer Institute that found the virus that causes AIDS. It was the retrovirus now called human immunodeficiency virus (HIV).

He has worked on AIDS for four years now. At first, "more or less because I was told to." Next, because he fell into the thrall of the scientific mystery. Then, because the epidemic began to grow so grotesquely large that it became a threat to society as a whole. "The final step is personalizing it," he says. "I began to think of my own children."

Redfield met his wife Joy when the two were in medical school. Within a few months of meeting, they married. They both wanted a big family, and they were pleased when Joy soon became pregnant.

Then came a major lesson about medicine, errors and human feeling. Because of complications at birth, the infant could not draw enough oxygen and died.

The boy's death was unnecessary, Redfield believes. The doctors, some of whom were his friends, were not quick enough. Following the death, he found they were also unable to help with his grief.

"Lots of doctors run away when their patients die," he says. "I maybe have started to do that a little, because I have been through so many deaths in this epidemic. But I try to make a point of not running away; I want to help them die, and to work the family through it."

Two years after their son's death, he and Joy had twins. Two other children followed later. But the Redfields usually have more than their own four underfoot. Their house is a gathering place for at least four other small ones from the neighborhood.

Every three months, Redfield has himself tested for signs of AIDS. The tests are not pro forma.

In working with virus-infected cells taken from patients, he used to use a needle to extract the cells and then put them into a flask for tests. Twice he stuck himself with a needle full of virus-loaded cells while transferring the cells from one container to another. One of those times, he not only stuck himself but accidentally injected a large number of the cells into his arm.

For 14 months after that, he says, he refrained from sex, waiting till he was certain the virus had not taken hold.

Indeed, AIDS inhabits Redfield's life like a persistent, unwelcome house guest. His daytime hours are absorbed by it, he travels often to give talks at conferences, he has patients to treat. He has missed his children's soccer games and school picnics. The declining population of neon tetras in his fish tank must remain untended for now.

His 8-year old son has made a deal about the soccer games. "You go ahead and cure this AIDS, Dad," he says. Redfield demurs, but the boy's simple formula continues: "If you can't do it now, I'll do it when I get bigger."

In these ways, Redfield is not unlike other researchers visited by the urgency of AIDS. "There is a time in your life when you can actually see your work in history, and this epidemic is not going to be on Page 2022 of the history of medicine," he says. "This is the first epidemic in human history in which we can watch a new microbe becoming endemic in the species. And we have the [biological] tools to do something about it.

"This virus has the potential to cause major changes in population, at least in some parts of the world, to change the makeup of society," he continues. It can erase a significant portion of the young, sexually active people on the globe. In some Central and East African nations, the infection has already gone so far that it is not unreasonable to predict the death of a quarter of the entire population."

As Redfield talks, the family collie, Misty, peers in accusingly from another room.

"She doesn't like me," Redfield says. Some time ago the dog's sister got a flea treatment that killed her. On the night she was dying, as Redfield worked to save her, Misty watched, and apparently never understood. "She thinks I killed her sister."

Redfield tells the story with some humor; he has resigned himself to suspicion from his dog. But on a more serious question of trust, he is not resigned. He believes people must summon enough trust to begin more widespread testing for AIDS.

Some of the problem is in overcoming general fears. "We are all afraid to find out if we have something bad," he says. "Women who have to go for a breast biopsy -- it's a terrible experience. You really want to know. But you really don't want to know.

"The trouble [with AIDS] is, if you don't know for sure, the person who is having sex says to himself, 'I'm probably not infected.' Then, if he has risky activity, he says to himself, 'Hell, I'm probably infected anyway.' And he has a 50 percent chance of being right in each of those situations."

Two kinds of AIDS testing are needed, Redfield says -- one in which names would be kept on file, the other completely confidential. "I would like to be notified if I slept with a woman who turned out later to have AIDS. That opportunity will not be given if we have anonymous testing. But a few people will be afraid and want confidentiality."

The policy of not testing and not telling, he believes, "is threatening the health of the whole community. And ultimately it's going to threaten [gays'] freedom. They don't understand it, but a lot of people are going to be angry when they learn that the public health authorities of our country have been paralyzed because of this concern about confidentiality ... "

Anyone who tries to persuade people not to get tested, Redfield insists, summarizing his position as bluntly as he can, "has the blood of more gay men on his hands."