EVERY TIME Benjamin Bloom talks to people about his book on the development of great talents, some members of the audience come to him afterwards and whisper, "I could have been a great violinist," or "I was going to be a great tennis star."

"Evidently everybody has some sense of having the possibility of greatness, under appropriate conditions," says Bloom. "I guess there's also some kind of guilt: 'My child has a high IQ, and look at what he's doing now!'"

In his quiet, dusty office at the University of Chicago, where he is a professor of education and has conducted groundbreaking studies of how children learn, Bloom seems surprised at having struck such a responsive chord. Since Devleoping Talent In Young People came out a few months ago, he has been deluged by letters and requests for speeches or interviews -- not by his usual readers, who are educators or government officials, but by radio and television stations and hundreds of parents.

The book is based on an extensive study of world-class pianists, swimmers, sculptors, mathematicians, research neurologists and tennis players, undertaken in order to find out how one gets to the extremes of learning. Bloom hoped that the answers would lead to better ways of teaching all children. But in addition, he now realizes, "I've hit on something that seems to involve parent-child relations that I had not anticipated." Nurturing A Child's Talent

Before starting this study, Bloom believed that "exceptional talent would show up in the first three or four years of life, and parents would see it and give their children lessons because they were gifted. But it turned out to be the other way around." First, children were given lessons, he explains, and then -- after much fuss was made about their progress and they put in increasing amounts of effort -- they became talented.

"We couldn't have predicted which children would make it to the top before they reached the age of 12 or so, when most of them had been in their talent field for five or six years," Bloom says. By then they were practicing about 25 hours a week and had been accepted by highly skilled teachers. Their rise depended partly on their families' support. Bloom concludes that "there is an enormous pool of potential talent in our society" -- talent that can either be developed or wasted.

Although he is now 72 years old and grey-haired, Bloom has lost none of his indignation at the horrendous waste of children's potential that he sees throughout the nation's schools. It's not just the failure to produce great talent that worries him, but the destruction of millions of children who are made to feel inadequate and turned against learning -- as well as against society -- by a system that he considers to be "rigged" against thm.

Picking up a piece of chalk and going to the blackboard, Bloom explains that the difference between the achievement of children at the top and bottom of the class at grade two is generally doubled by grade four, and tripled by grade six. "The children at the bottom get further and further away from those at the top," he says, "and nobody helps them get the prerequisites for the next grade. Unless we find some way of interfering in this system, their achievement at grade 11 can be predicted nine years earlier."

Physicians are not satisfied with "merely predicting which patients will live and which will die"; they also cure people, or at least make them better able to survive, Bloom points out. He believes that teachers should take it upon themselves to do the same with respect to learning. Most of his work over the past 40 years has been devoted to this goal. Failure Is Catching

Benjamin Bloom is best known in educational circles for his two-volume Taxonomy of Educational Objectives (1956), which classified learning skills in a hierarchy ranging from simple recall to complex problem- solving. It became the bible of curriculum developers. He is also known around the world as a founding father of the International Association for the Evaluation of Education Achievement, which has compared the educational achievement of students from many countries in such fields as mathematics and science.

In his view, however, his most important book is Stability and Change in Human Characteristics, which he wrote in 1964. Based on an analysis of 1,000 studies that followed the development of children over long periods of time, it showed that each human characteristic, such as physical height, intelligence, or school achievement, has a particular growth curve. Bloom then stated that the power of the environment to affect these characteristics is "greatest during their periods of most rapid growth," which in many cases are the first years of life.

"When the student starts school at age 6, he has attained about one-third of his general learning pattern," Bloom wrote. "At age 4, he is likely to have attained about one-sixth of this pattern." The longer one waits, the more difficult it is to produce any change, he explained. Therefore a child's experiences during the first years of life are critical. This thesis led to the establishment of Headstart centers for children from low-income families throughout the nation.

Many children made rapid progress in the new preschool classes. However, they tended to slip back after a few years in school. It soon became apparent to Bloom that the standard American school system pushed a large number of children into repeated failure. In addition, he says, the experience of failure "infected" many children with emotional difficulties.

Bloom and his graduate students determined to break this cycle. They began by analyzing "the best learning conditions we can devise: one-to-one tutoring." The key features of private tutoring, they decided, were that students got immediate feedback on their work, that the tutor re- taught what they missed, and that they corrected their work before proceeding to the next point. In an experiment comparing such tutoring with a conventional class, they found that only two percent of students in a conventional class did as well as the average student in private tutoring.

How could these key features be provided to large groups of students? After experimenting with several models over a period of years, Bloom devised a tightly- organized system which had these key features but could be used by any classroom teacher: "Mastery Learning." Steps To Achievement

MASTERY LEARNING is revolutionary only in its assumption that nearly all children can learn well.

Otherwise it just brings together techniques which many good teachers already use: frequent diagnostic tests that provide feedback but are ungraded, re-teaching of missed items, corrective work, and re-testing, so that students master at least 80 percent of the objectives of a learning unit before going on to the next. In their final exam they are not graded "on the curve," but against objective standards. Since this results in a very large majority of the class getting A's and B's, the system breeds self-confidence. It also promotes cooperation among students rather than competition.

"Most school learning is sequential and depends on having learned the pre-requisites," says Bloom. Mastery Learning almost guarantees that students will have the pre-requisites for each step. Therefore the difference between students at the top and bottom of the class tends to narrow, rather than increase, as time goes on.

Mastery Learning was relatively slow to catch on, but by now Bloom estimates that it is being used in about 3,000 schools with over 1 million students. The classes range from kindergarten to university level in such places as New Orleans, Denver, New York City, Chicago and Baltimore, as well as many smaller cities around the country. Although Mastery Learning has proved quite difficult to implement, it has had good results.

One outstanding example is Red Bank, N.J. The students in this largely blue-collar town used to trail one grade below national norms on standardized achievement tests. Since 1979, when the entire school system adopted Mastery Learning, achievement scores have shot up. Last year Red Bank's students scored two or three grades above national norms on Metropolitan Achievement Tests of math and language. At the same time, teachers reported that morale improved and discipline problems decreased.

The 27 schools of District 19 in Brooklyn, N.Y., have also made some progress since they began using Mastery Learning in 1980: the number of students who read at grade level has increased from the low level of 26 percent to nearly 50 percent.

There are several practical problems in shifting to Mastery Leaning. For one thing, it requires teachers to formulate specific objectives. Normally, "teachers don't plan in terms of student outcomes or objectives, but in terms of lessons," explains Dr. Thomas Guskey, a professor of education at the University of Kentucky who has trained thousands of teachers in Mastery Learning techniques and has just written a book on the subject, Implementing Mastery Learning. Another problem is that Mastery Learning requires frequent testing and a great deal of paperwork.

Red Bank's teachers developed their own materials for their Mastery Learning program, which covers all school subjects, but this took a lot of extra time before the program started. Chicago's school system decided to bring in a team of outside experts on reading to develop objectives in this area. The team produced a list of 525 reading objectives (later reduced to 273) which all elementary schools in Chicago were to meet. Then another team of experts designed the instructional package that would help teachers meet these objectives: lessons, guided practice sheets, tests, additional activities to teach certain skills in a different way, enrichment activities, and final tests.

This package (now called the Chicago Mastery Learning Reading Program) was distributed to all Chicago elementary schools in 1981. It has been severely criticized for its emphasis on isolated sub-skills of reading, rather than on reading comprehension. Nevertheless, it is still being used in Chicago, and in at least one elementary school -- the Charles Kozminski Community Academy -- it seems to work quite well in conjunction with reading texts.

"Properly used, it's a great aid," says Allen J. Travis, Kozminski's principal. "It has a good management system -- and we do need one, to be aware of whether the kids are really making progress. We don't want to wait until it's too late and another year has gone by."

On a recent school day, third graders in this school were working with practice sheets that dealt with prefixes and suffixes (happy, unhappy; polite, politely), reading a third-grade level text, and memorizing poetry by Langston Hughes. The school has a good academic reputation. Over the past five years, 20 of the school's graduates, most of them from families on welfare, have received scholarships to top New England prep schools. Making Learning a Priority

BLOOM DOES NOT concern himself with how Mastery Learning is carried out in practice. He had nothing to do with the Chicago reading program and emphasizes that "Mastery Learning is not a curriculum; it's just a way of improving instruction." In general, he says, "I'm interested in the theory of it, and whether it's real and true. I don't want to spend the rest of my life preaching Mastery Learning."

As a result of Bloom's reticence, people interpret Mastery Learning in many different ways. There is some confusion about which is "the real" Mastery Learning, and also a tendency to give almost anything this now-fashionable name.

"A lot of urban districts use the rhetoric of Mastery Learning, but in fact, for the most part, it's not Mastery Learning; they don't overcome some of the organizatonal constraints, such as rigid time schedules," says Dr. Lawrence Dolan, assistant director of a new Prevention Research Center at The Johns Hopkins School of Public Health. The Center will focus on Mastery Learning as a means of improving children's mental health. With funding from the National Institute of Mental Health, it will work with the Baltimore City public schools to improve children's performance in the first grade, on the theory that this will help prevent delinquency, depression and drug abuse in their later years.

In Alexandria, Va., elementary schools all reading programs use Mastery Learning "to the extent that there's no case in which a student is allowed to go on without having mastered previous skills," according to Dr. James Akin, director of elementary education for the city's schools. But "Mastery Learning is quite controversial," he says. "The most controversial aspect is that there are no D's or F's in the grade scale. We couldn't decide whether people were really ready for that! In a regular classroom, you'd assume there would always be some child who'd fail. People want us to sort out students and label them. Mastery Learning makes some very, very different assumptions about learning."

For Bloom, who has seen how much progress students can make under individual tutoring and how many children have the potential for great talent, these assumptions have become second nature.

He points out that "the Japanese are now three inches taller at maturity than they were before World War II. It's a result of special infant-welfare centers that were created right after the war, where women could come during pregnancy to get free supplementary food and advice, and their babies were immunized against disease." Only high-income parents could afford such care before the war, he explains, and only their children grew tall. Today the average Japanese is taller than five-sixths of the prewar population. Bloom believes that similar improvements could be achieved in school achievement, throughout the United States, if we decided to provide "state of the art learning conditions" for our children, especially in the early grades.