In the opinion of many experts, the foremost center of technology in the world, for most of this century, was the Bell Telephone Laboratories.

Before the court-ordered breakup of American Telephone & Telegraph Co., its Bell Labs complex was a fabulous hothouse of discovery and invention, with 26,000 scientists, engineers and other employes, 20,000 patents to its credit and an annual budget of $2 billion.

"Before divestiture, a young engineer driving up to Bell Laboratories in Holmdel, N.J., for his first day of work might have spotted the vast, glittering building across an immense stretch of grass fields and felt as if he were approaching a scientific Oz," wrote Paul Wallich, an associate editor of Spectrum, the magazine of the Institute of Electrical and Electronics Engineers. "In a sense, he would not have been wrong," Wallich added.

The breakup of AT&T's monopoly in January 1984 and its transformation into a competitive high-technology manufacturer have meant profound changes for AT&T's Bell Labs -- the loss of 8,000 employes and new marching orders for those who remain.

The fear among some experts and onlookers alike is that AT&T's unique commitment to research will be eroded by the change, weakening one of the nation's greatest technological outposts just as the telecommunications industry is entering a new, crucial round of worldwide competition.

An impressive investigation of this issue, along with other repercussions from the breakup, is contained in this month's issue of Spectrum.

The partitioning of Bell Labs actually occurred in two cuts. A 1980 ruling by the Federal Communications Commission permitted AT&T to enter the computer field -- but required it to create a separate, arm's-length subsidiary for this business. About 4,000 Bell Labs employes joined the new subsidiary. A roughly equal number transferred from Bell Labs in 1984 to a new, independent research organization serving the regional telephone companies. Some of AT&T's scientific brainpower has defected, to start up new ventures.

The 18,000 Bell Labs people who remain must cope with a new kind of R&D institution. The commitment to research has not slackened, said Ian Ross, president of Bell Labs. But the goal of R&D has shifted, he added in an interview this summer.

"When the transistor was invented in this company [in 1947], we had to wait 10 years before we got it reliable enough to go into switching and transmission equipment in the network, which were the only businesses we were allowed to be in," Ross said. "And in that period, people like Texas Instruments came out with the transistor radio. We weren't allowed to sell transistor radios."

Now, the overriding mission at Bell Labs is to connect scientific work with products that AT&T can sell, as it swings into competition with International Business Machines Inc. and a raft of other computer and telecommunications companies here and overseas.

"Next time around, when we have a breakthrough invention, we can do the equivalent of a transistor radio. We can capitalize on our technology with much more freedom in the future than we could in the past," said Ross.

AT&T's gain is likely to be its competitors' loss, however. As Spectrum noted, until the 1980 FCC ruling and the breakup, Bell Labs had provided a rich lode of technology for other companies. "As part of the resolution of an earlier antitrust case, Bell Telephone Laboratories published its work readily and let other companies use its patents -- among them the transistor -- for reasonable license fees," the magazine said. It's no wonder that Bell Labs was revered as a national treasure.

Ross denied Bell Labs is closing its doors to outsiders. Its focus on the commercialization of its technology will bring more of its discoveries more quickly into the marketplace. Competitors can license the technology from AT&T -- and some undoubtedly will copy it, Ross said.

The concerns of the scientific community, however, are demonstrated by the case of AT&T mathematician Narendra Karmarkar, who devised a radically new technique for solving complex equations that promises to speed up large computer-programming tasks.

Spectrum noted that such equations also could be used for airline scheduling or other such uses. While other mathematicians praise Karmarkar's theoretical breakthrough, they have been unable to duplicate the results because details of his work aren't available, said Spectrum. "Some observers outside AT&T Bell Labs point to this as an example of how competitive interests and scientific research may collide," Wallich said.

It remains to be seen whether the new Bell Labs can keep alive the powerful creative culture that flourished in the old institution. Spectrum quoted the concerns of Chanchal Amanta, a member of the AT&T Information Systems Laboratories technical staff, who worried that as a competitive company facing harsh marketplace issues every day, AT&T may not have enough patience with long-range basic science.

Ross agreed about the danger: "If ever ATT did what the press said it's going to do -- focus down on the short term and neglect the long term -- I think it's going to be asking for trouble." Ross said he doubts that will happen.

But as Spectrum noted, the deregulation of AT&T and the transformation of Bell Labs is an experiment whose outcome is not clear. If the problems created by that experiment are not effectively dealt with, warned Continental Telecom's vice president and chief scientist John C. McDonald, "our children will inherit a second-rate telecommunications infrastructure."