IT TOOK THE MANAGERS of American science much of the postwar period to persuade their political patrons that knowledge really is power. But now that the maxim is firmly established in federal policy -- as reflected in bountiful research budgets under Ford, Carter and, yes, even Reagan -- there's a new protectionist twist that's stirring axieties amoung the triumphant scientists.

If laboratories are the intellectual powerhouses for modern industrial and military strength, a rising breed of scientific protectionists says, let's selectively guard their doors and output against prying competitors.

For military applications, the secrecy process has always been the case. But what is being talked about now is an extension to certain kinds of heretofore wide-open civilian research.

The thought is so in conflict with science's traditions of easygoing international traffice that several of the reigning mandarins are sending out alarms.

Prominent amoung them is Frank Press, White House science adviser under Carter and president of the elite National Academy of Sciences, who fears that science's intellectual lifelines are threatened by amateurish delusions about guarding the processes and products of research.

His concerns are shared by his successor in the White House advisory post, George A. Keyworth, who has tentatively agreed to back an Academy inquiry into the man Howard Johnson, the as-yet-unannounced study originates in concerns that scientific freedom and openness my easily be infringed upon.

There are already signs of this: The Pentagon, for example, has prodded big research-oriented universities to be aware that some of their on-campus electronis research might -- even though unclassified -- come under the Arms Export Control Act. In such cases, foreign students and researchers would have to be fenced out, which could cause some turmoil for American academe, given that foreigners constitute a big chunk of student enrollments and treaching staff.

Further, the National Security Agency has cajoled academic computer scientists into a voluntary arrangement for consulting with the agency on prospective research publications that might be of value for code making and breaking. That step toward prior restraint, taken after much agonizing by the researchers, is an unprecedented peacetime concession by civilian scientists. But NSA was putting on the pressure, and the scientists finally settled for what looked like the least intrusive bargain.

The argument for close playing of our scientific cards is deceptively attractive to the new protectionists, but unfortunately, it rides on antiquated notions about American scientific preeminence.

Emerging unscathed from World War II, science here boomed because of an influx of superstar refugee researchers and great amounts of Cold War research money.

If took Europe a long time to catch up, but at last it has, while other nations -- notably the Soviets, Japan and India -- have taken strong positions on many sectors of the frontiers of science. We can't shut them out from our labs without risking reprisals that would cost us the benefits of their considerable scientific prowess.

But even more important, the erection of barriers around non-military science can choke the simple communications processes that underlie the vigor of the American research enterprise.

That's the main point of a newly published Rand Corporation report to the Defense Department on the risks and benefits of trying to deny the Soviets advanced American technology.

Titled "Selling the Russians the Rope?" -- a reference to Lenin's remark about greedy capitalists selling the rope for their own handing -- the Rand study concludes that "if the national purpose is to maintain the United States' technological lead, our first concern should be to remain good innovators ourselves. We should beware lest we hobble ourselves, as the Soviet system has so clearly succeeded in doing in the greater part of its industry."

That's good advice.