(FORTUNE Magazine) – WHEN AT&T was broken up on January 1, 1984, admirers of Ma Bell's deep commitment to research wondered about the fate of AT&T Bell Laboratories -- the great American invention factory. Bell Labs had produced the transistor, the laser, the solar cell, and the first communications satellite, as well as sound motion pictures, the science of radio astronomy, and crucial evidence for the theory that a Big Bang created the universe. Without the vast revenue base provided by the AT&T operating companies, would Bell Labs -- the premier corporate research facility in the U.S. for most of its 62-year history -- wither into just another run-of-the-mill industrial R&D operation? Those worries haven't entirely subsided. For example, Edward E. David Jr., 63, a Bell Labs alumnus who was science adviser to President Nixon, still questions whether any company would continue to support fundamental research if times got really tough. But so far the vital signs are strong at the Bell Labs buildings in Murray Hill and Holmdel, New Jersey, where most of AT&T's basic researchers work. The company's support for its brainy offspring has not wavered. Funding has increased 18% (to $2.2 billion last year), and total employment is growing again (to 21,000 in 1987, vs. 26,000 in 1983 and a post- breakup low of 18,000). Not only has basic research come through the breakup largely unscathed, but Bell Labs is also branching into new commercial areas -- in part by spinning off venture R&D companies -- and looking into licensing to companies as far afield from telecommunications as airlines and shipping firms. Historically, basic research absorbed only about 10% of manpower and spending at Bell Labs, and that proportion remains unchanged. Most of the employees are engineers, and they have always worked on applications, not basic research. Beyond New Jersey, no fewer than 2,700 Bell Labs engineers and software specialists at Indian Hill, Illinois, near Chicago, are developing electronic switches and associated software. In Allentown, Cedar Crest, and Reading, Pennsylvania, scientists and engineers design semiconductor chips and the microscopic lasers that control the flow of light signals through optical fibers. In Atlanta, Bell Labs researchers work side by side with manufacturers in the world's largest optical fiber plant. In Columbus, Ohio, they monitor long-distance telephone traffic and design specialized information systems. In North Andover, Massachusetts, they work on terminals and carrier systems, the telecommunications counterparts of railroad stations and tracks. But basic research at the Labs has always been a huge attention-getter because of its unmatched results and epochal discoveries. The 1947 invention of the transistor set off the world microelectronics and computer revolution. Seven Bell Labs scientists have won Nobel Prizes, more than at any other industrial institution in the world. The honors have continued since the breakup. In 1985, for example, President Reagan awarded Bell Labs the National Medal of Technology -- the only U.S. laboratory ever singled out for it. What happens at Bell Labs is of vital interest to American industry because of the Labs' high quality of research and because it has been strong where the U.S. now finds itself alarmingly weak: in the transfer of research results into products. Says Robert M. White, president of the National Academy of Engineering: ''This country's problem is not lack of basic research but inadequate conversion of scientific discovery to commercialization. Bell Labs does that very well indeed.'' Weathering the breakup wasn't easy. Before it, Bell Labs' support was assured. Nearly half the budget came from a kind of tax on every telephone in the U.S., which amounted to about 50 cents per phone per year in 1983. The rest came from Western Electric, AT&T's manufacturing arm. Since AT&T was < suddenly plunged into unaccustomed competition, many people concerned with the state of U.S. science feared that basic research would be the first casualty. That hasn't happened, although the source of funding has changed. Now AT&T's corporate budget pays for basic research (about $200 million last year), while development (the other $2 billion) is financed by the company's various lines of business. AT&T executives sound reassuring about the future. Says President Robert E. Allen: ''We fund research at AT&T Bell Laboratories with a long-term view, not with our eye on the next quarter's earnings. The products and services our customers want rely increasingly on leading-edge technology, so Bell Labs -- including its research area -- is an important resource for AT&T's future. The 10% of Bell Labs that does research is not a luxury. It's a strategic necessity.'' BELL LABS is striving to help AT&T's businesses by tailoring basic research more closely to company needs without sacrificing the scope and sweep of investigations. Bell Labs executives scrutinize research activities for their contribution to the parent company's business, defined as ''management and movement of information.'' The economics and psychology departments have been cut drastically while robotics and computer science have grown, but that shift in emphasis involved only about 40 of the 200 or so scientists who pursue the purest kind of pure research. ''To an outside observer it may seem that we've gone techie, but the intellectual content of the work is the same,'' says vice president Arno A. Penzias, 54, the scientist in charge of research at Bell Labs. Penzias, an ebullient astrophysicist, made his mark soon after he arrived at the Labs in 1961. He was asked to join a committee of older scientists who were trying to devise the best way to calculate the precise positions of communications satellites by triangulation. The scientists were talking about setting up tall, expensive radio masts when Penzias piped up with the suggestion that nature's own radio masts -- radio stars, which emit characteristic frequencies from fixed positions in the sky -- would serve equally well at no cost whatever. Penzias's idea was accepted and the committee disbanded on the spot. Later, Penzias and his colleague Robert Wilson built precision measuring devices for Bell Labs' radiotelescope in Crawford Hill, New Jersey, as part of their effort to track down the source of static that often interfered with their studies of radio waves from the Milky Way. The noise they studied turned out to be the residual radiation from the Big Bang; for their discovery, the two men shared a Nobel Prize in 1978. BEFORE THE BREAKUP, Penzias predicted that without the operating companies as a base, Bell Labs would become a ''sinking ship.'' He feels better now. Of the Labs' direction, he says: ''None of the things we do are really unconnected to our business, but sometimes the connection is hard for the average person to see. We've added to the big part of research where the connection to the business is clear.'' Since AT&T views basic science as a competitive tool, the scope of research at Bell remains wider than at most other industrial labs and even some universities. The staff includes 3,430 Ph.D.s -- more than the total research staff of the closest rival corporate lab, at IBM (see box). They and the other scientists are spread among physics, chemistry, computer science, mathematics, electronics, and sundry fields. Bell Labs' method has always been to assemble a huge mass of diverse specialists who interact closely. The sprawling building at Murray Hill is an immense beehive. It houses 3,049 researchers, developers, and support staff along quarter-mile-long corridors lined with hundreds of small labs crammed with the latest instruments. The physics research division alone employs 250; it's larger and more diverse than most university physics departments. Investigations range from basic studies of the nature of matter, including such hot topics in theoretical physics as instabilities and chaos, to building ceramic superconductors and creating so-called neural networks in silicon chips that mimic rudimentary animal brain pathways. Electronics and optics are two other large areas of emphasis. The latest work includes development of optical amplifiers that boost the intensity of light particles, or photons, so that they can carry voice and data signals along optical fibers without the need to convert them to electrons and then back to photons, as current systems require. Recently AT&T began to install the world's most advanced fiber-optic transmission system, developed at Bell Labs, which can speed 24,000 telephone calls simultaneously through a pair of fibers, each twice the thickness of a human hair. It has 40% more capacity than any other commercial system. The most basic work at Bell Labs has a way of merging into development, though that's not immediately apparent from the activities of some of the basic scientists. One recently reported on the activities of ants in the jungles of Brazil; another observes faint galaxies at the edge of the universe from observatories in Chile and Hawaii. The student of ants, Thomas Gradel, 49, working with scientists from Cornell, reports that a major cause of acid rain in the Amazon is formic acid, a pungent, colorless substance released by the decomposing bodies of ants. His investigation provoked a letter from an AT&T stockholder who wondered why the money could not be better used to bolster AT&T's skimpy dividends. But Gradel's interest in the Amazonian ants is highly practical: He is a corrosion chemist, and part of his job is to find out why telephone equipment can fail in various environments. THE STARGAZER, astrophysicist J. Anthony Tyson, 47, has his feet on the ground as well. He is trying to improve another Bell Labs invention, the charge coupled device -- in effect a chip that can see. It has revolutionized astronomy because it collects light up to 1,000 times more efficiently than film, but it also has potential uses as the eyes of robots and in the precision manufacture of semiconductors. Tyson is one of a handful of Bell's basic scientists who ''couple us to the universe of science,'' as Penzias says. ''It's a small but vital part of our business strategy to have a few scientists do work that gives Bell Labs a connection to the universities and the rest of the scientific community that it couldn't get otherwise.'' Among other things, such connections help attract young scientists. Recruitment has been unaffected by the breakup; 80% of the students offered jobs at Bell Labs accept. It pays competitive or somewhat higher salaries than other major corporate labs, such as those at IBM and Du Pont. (A new physics Ph.D. starts at about $45,000.) And although Penzias says that some scientists earn $100,000 and more a year at the Labs, money is not the main draw for most of them. The freedom, the facilities, and first-class colleagues come before that. HARVESTING the fruits of research happens faster at AT&T than it did in the good old days. Bell Labs' president, Ian M. Ross, 60, is a subdued British- born Ph.D. in electrical engineering with several advances in semiconductors to his credit. He cites the emergence and the rapid adoption by AT&T's businesses of a remarkable mathematical shortcut to the solution of immensely / complex versions of the celebrated traveling salesman problem, which requires devising the shortest possible route connecting a given number of destinations. Indian-born mathematician Narendra Karmarkar, 31, described this new insight in 1984. Where programmers and mathematicians once took days to solve a problem with thousands of variables, the Karmarkar algorithm allows them to do so in minutes. AT&T is already using the algorithm to design a vast and complex phone network among the 20 nations of the Pacific Rim that involves more than 60,000 variables. The algorithm is useful in other fields as well; Bell Labs is getting ready to apply it to airline and shipping businesses. A 1956 consent decree that AT&T signed with the Justice Department to avoid a threatened breakup forbade AT&T -- and Bell Labs -- to operate in any field outside telecommunications. Moreover, Bell Labs had to share any discoveries it made with the world at large. Now that the breakup has finally taken place, Bell Labs no longer has to give away its discoveries for a pittance -- as it famously did with the transistor. Sony secured the rights in the 1950s with a $25,000 down payment on royalties and thereby ignited the consumer electronics revolution with transistor radios. That may well have been history's richest commercial windfall. If Bell Labs invented the transistor today, it would be completely free to produce transistor radios, provided that made business sense. The breakup of AT&T has opened other windows into the marketplace for Bell Labs. Pixel Machines, one of 11 venture companies started so far by Bell Labs scientists and engineers under AT&T's aegis, offers a computer terminal with striking capability to analyze and display computerized data in some of the sharpest pictures available commercially. The company is so new that it's still housed in its incubator, the huge green-glass Bell Labs building in Holmdel. While Bell Labs people working for the new companies can't expect to become millionaires -- AT&T gives them no stock -- they can qualify for bonuses for exceptional work and at the same time satisfy entrepreneurial urges. ''When these businesses get started, they are very carefully nurtured and have freedom from a lot of the AT&T bureaucracy,'' says Pixel Machines President Kathryn M. Sullivan, 45, a former AT&T marketing executive who supervises a staff of 30 scientists, engineers, and marketers. ''I like to look at it as the best of both worlds. You have all the resources of AT&T and yet you have a lot of freedom to do things the way a small company can.'' Initially the ventures are sponsored by AT&T divisions that pay their overhead. If a venture is successful, it could become a business in its own right, tied to AT&T. Or it could be merged into AT&T mainstream divisions, or spun off completely. Competing against the rest of the world is teaching Bell Labs product developers to couple R&D even more closely to both manufacturing and market needs. In the past technology drove Bell Labs development; now the customer does. A classic example of a technology-driven product from the pre-breakup days: the Picturephone of the mid-1950s. It worked fine, but market studies of the potential demand for it failed to make clear just who could afford to use the thing. Today Bell Labs would let the market determine whether it would develop a Picturephone. Specialists from the Labs now accompany AT&T marketing people on calls on customers. The pressure for faster payoff has ended the days when Bell Labs could wait 20 years for a product to pay for itself. ''When Ian Ross comes here now, he asks, 'How are you helping our business?' '' says David J. Meskell Jr., 52, director of the transmission media lab at the 151-person Bell Labs outpost at the Atlanta fiber-optics plant. ''This is a change from what he used to ask us four years ago: 'I want to hear about your newest technology here.' '' But with basic research, Ross's emphasis is just the opposite: He tells researchers to keep their eye on the future, five years and beyond. Into the competitive world today Bell Labs developers are bringing such impressive products as a gigantic computerized electronic switching system, the 5 ESS, which can cost several million dollars and handles up to 300,000 telephone calls an hour. The 5 ESS allows customers to transmit both voice and data directly in digital form, which is more compact and accurate than traditional analog transmission. In combination with minor hardware additions to existing phones, digital switches improve signal flow enough that AT&T can now offer three transmission channels instead of one on a single line. When a customer has this service, managers in different locations can talk to each other and simultaneously exchange documents over the second channel. The third, less capacious channel controls the transmissions and can also be used to display a caller's phone number. When a client calls his insurance ! agent, for example, the client's telephone number would trigger electronic retrieval and display of his file on the agent's computer screen. The feature could also be used to screen or trace incoming calls. Such new services are part of the rapidly evolving Integrated Services Digital Network (ISDN), which Bell Labs is pioneering. ISDN can link a company's far-flung locations; Bell Labs is helping install an ISDN system that will eventually connect McDonald's 7,500 hamburger outlets and the company's administrative offices. As high-capacity optical fibers come into wider use, ISDN will greatly increase the amount of information that can be transmitted into homes and offices -- spreading the use of facsimile terminals and high-resolution television and probably bringing the Picturephone back to life in an inexpensive new version. In all these activities Bell Labs people think they have a competitive advantage because research has been integrated into the work of the parent company better than at any other industrial lab. From the outset Bell Labs elevated ''co-location'' -- the stationing of its people at manufacturing plants -- into an art. Scientists and engineers are scattered among AT&T factories in six states to provide a direct link from research to development to manufacturing. TYPICAL OF THIS coupling is the work of Mark Melliar-Smith, 42, executive director of the electronics and photonics devices division. (Photonics includes light-processing devices such as tiny semiconductor lasers that drive photons through optic fibers.) Melliar-Smith has offices at Bell Labs in Murray Hill and at the AT&T Technologies semiconductor and laser plants in Allentown and Reading. ''Here there's always tremendous interaction,'' he says. ''Basic researchers may discover something of interest to us. We in turn provide materials to them. We call it a seamless product realization process.'' A Ph.D. in chemical physics, Melliar-Smith has worked during his 17 years at AT&T in basic research, in development, and as director of engineering at AT&T's Kansas City, Missouri, electronics components factory. Bell Labs' basic scientists insist that competition is nothing new for them, that they have always competed against the world at large. As Arno Penzias puts it, Bell Labs traditionally has been a place that ''made its own future happen.'' Just as it opened the new world of microelectronics by inventing the transistor, Bell Labs is now far along in harnessing the electron's ephemeral ! cousin, the photon, for the task of information movement and management. In Bell Labs' bag of surprises there even could be an optical computer superior to its electronic counterpart. Progress in that field in recent months has been exceptionally rapid. The optical computer, using laser beams instead of electrical connections, would work 1,000 times faster than today's electronic variety -- an almost unimaginable boon to everyone from theoretical physicists to weather forecasters.