WHEN THE SUPREME Court ruled against me in 1980 and upheld the enforceability of government secrecy agreements, my father -- who is a conservative superior court judge -- predicted that "one of these days some patriot in the White House will realize the power the Brethren have given him," and saddle us with a system of censorship such as we've never seen in this country.

My father has been proven right. President Reagan, citing Snepp v. U.S., has decreed that every bureaucrat "with authorized access to classified information shall be required to sign a nondisclosure agreement. . . ."

This order will obligate some bureaucrats to submit all work-related writings for government censorship for the rest of their lives. And the Supreme Court made clear in my case that these government workers won't even have to sign secrecy agreement to become censorship candidates. All they have to do is get assigned to an official "position of trust" with "conceded access to confidential sources and materials." From that point on, they're implicitly obligated not to publish anything, classified or not, about their work, without official approval. Forever.

In a "fact sheet" attached to the Reagan order, the Justice Department reminds all bureaucrats of this implicit "fiduciary duty." This clears the way for a censorship system that is virtually open-ended.

Steven Garfinkel, the official responsible for monitoring governmentwide security programs, has conceded to Congress that though he can't say for sure how many bureaucrats traffic in classified information, at least 65 official departments and agencies do. The mind boggles at the potential number of gag victims this estimate implies. The 11 agencies that make up the U.S. intelligence community, by themselves, are 200,000-strong.

In practical terms, if the Reagan order is enforced, many of the turnstile bureaucrats who come and go with each administration will be out of business.

Will the New York Times, for example, look favorably on rehiring Richard Burt, a former reporter who, as a current State Department official, will be forever subject to the State Department's censorship?

Would the disciples of academic freedom at Georgetown or Harvard find it easy to embrace such returning prodigals as Jeane Kirkpatrick or Richard Pipes?

Would the conservative experts from the Heritage Foundation who've served the Reagan White House be happy about being censored by the liberal constituents of a Mondale, Glenn or Cranston administration?

You don't need a definitive answers to these questions to view the Reagan order as ill-conceived and dangerous.

Predictably, the administration has had a problem selling its scheme. Deputy Assistant Attorney General Richard Willard, principal author of the Reagan directive, initially claimed that the secrecy agreements were needed to stem a flood tide of leaks which "has increased in severity over the past decade." But then Garfinkel, the government's designated auditor of leaks, conceded to a congressional subcommittee that only a half- dozen leaks had been reported to his office in the past three years.

Willard tried to recoup. In a TV interview, he said it wasn't the quantity or severity of leaks that necessitated the gag rule. Rather, it was the worries of our allies -- their "lack of confidence in our government's ability to keep secret important information. . . ."

Since most of our allies (witness the British) have far more stringent secrecy regulations than we do -- and far more serious security problems -- Willard's attempt to justify the Reagan directive is a token of how desperate his case has become. And no wonder. Numerous authoritative voices have been raised against its assumptions.

Writing in Foreign Policy last fall, former CIA Director Stansfield Turner declared: "Fortunately, while several leaks about actual espionage in the past six or seven years have involved serious breaches of security, very little information harmful to U.S. intelligence interests has been revealed. In short, the impression that intelligence agencies cannot keep secrets is highly exaggerated."

Former Deputy CIA Director Bobby Inman has also cast doubt on the wisdom of the administration crackdown. Last winter he told U.S. News & World Report that the "most damaging" leaks come from "senior officials" who "gossip" excessively. To judge from his remarks, the most serious leaks wouldn't be staunched by any sort of prepublication review system. A far better antidote would be a program to educate presidential appointees about classified information.

Despite the problems raised by the president's decree, the administration apparently intends to implement it at once. It has said that it will use the CIA censorship system as the model for the entire government. If that news was meant to reassure anybody, it shouldn't. The CIA's "system" -- which was not formalized when I wrote my book -- has been administered with none of the sensitivity to First Amendment values which its architects claim for it.

For anyone laboring under the CIA gag regime, the first question is: what must be submitted to the censors? The answer, I have discovered, is: nearly everything you write, before you show it to another living soul.

Admittedly, CIA regulations generously exempt "a manuscript of a cookbook" or "a treatise on gardening." But the rules require the surrender of novels, articles, nonfiction books, speeches, movie scripts and even slide shows which "contain or purport to contain . . . materials on the subject matter of the agency and its activities," or on "intelligence matters generally" or which "may be based on information classified or classifiable. . . ."

With such ambiguous guidelines, only anninveterate gambler would fail to submit just about anything he or she authors. The penalties for denying the agency its "due" are nothing short of life-threatening: punitive monetary damages; the confiscation of all authors' profits, and a permanent gag order aimed at increasing the hazards of guessing wrong again. (After publishing "Decent Interval," my next "violation" could bring a contempt-of-court citation and a jail sentence.)

Nor is there any time limit on the censors. Say, for example, that you resign from the government, enroll in a night course on international relations, and decide to write on the crisis in Central America. Even if your paper deals with nothing that transpired during your stint in government, it must be submitted for predisclosure screening. That means before it goes to typist, Xeroxer, or teacher. Only active-duty government experts, the CIA argues, can determine if you've poked your pen into something "classifiable."

The next issue to be grappled with as your manuscript makes its way along the censor's conveyor belt is: what may be deleted? The Supreme Court in my case declared "harmful disclosures." CIA regulations specify "classified" or "classifiable" material obtained during employment (though I've seen this temporal bracket ignored in the censoring of one of my speeches and a novel).

Last year Reagan, by executive decree, empowered the intelligence community to reclassify material already in the public domain. This means the censors can now justify deleting anything from an ex-bureaucrat's manuscript simply by declaring it reclassified.

And lest you consider this an absurdly hypothetical proposition, ponder the sad case of Ralph McGehee. The former CIA field hand shepherded his manuscript through months of scrutiny by the censors, only to discover at the last minute that numerous passages previously "cleared" had been reclassified and struck out.

The Justice Department's Willard has attempted to defend such excesses as simple "mistakes" by CIA censors. But he gives them too much credit. Several years ago, when agency officials got hold of Victor Marchetti's "The CIA and the Cult of Intelligence," they deliberately over-censored it, snipping out perfectly harmless items (171 in all, including a reference to Richard Helms' difficulty in pronouncing certain foreign words), so they would have bargaining chips in case Marchetti challenged any of their 339 deletions in court.

Not only was this tactic personally approved by then-CIA Director William Colby; the CIA's top lawyer, John Warner, was so taken with its ingenuity he bragged about it in an official (unclassified) memo.

Under law, CIA censors must complete their initial review of a manuscript within 30 days. But as Marchetti and McGehee discovered, dickering over deletions can stretch on for years and may ultimately turn on whether the author can produce enough news clippings or other "public references" to prove the disputed information is so widely known that its suppression would be purposeless.

Footnoting a piece of fiction in this way is nearly impossible, and even if a novelist can do it, the censors may not relent. For it is the CIA's position that if the reference material itself was originally classified and published without authorization, it's still classified.

In the end there may be only one means of salvaging a manuscript: through a court battle. While there's nothing inherently wrong with having the federal courts resolve a dispute between censors and their victims, there's nothing salutary about it, either.

For one thing, the Supreme Court in my case left unclear how the victim of censorship gets his case to court. Former CIA Director Turner, whose forthcoming memoir was recently trashed by the very censorship board he created, has said that even he's confused on this point. He has mentioned suing the CIA to try to force the censors to change their minds. But he has also acknowledged that it may be necessary simply to defy them, as I did, to prompt a showdown in court.

Then there's the legal representation problem. If you're shackled by a secrecy agreement, you'd violate it if you discussed the censors' deletions with any lawyer who does not have a security clearance. Thus the censors have the right to veto your legal counsel.

Finally, if you manage to challenge the censors' handiwork in court, there is little you can argue except that the material at issue isn't properly classified. And all the government has to do to defeat it is to pull out the "reclassified" stamp. End of dispute.

Thus, those who wind up facing Willard's secrecy agreements will find themselves with a vexing problem. Once they leave government service, they will have difficulty parlaying their professional experience into a job. Rare is the newspaper, TV station, university, think tank, law firm or consultancy that will hire anyone who must submit his every written thought to government censors.

Nor are book publishers likely to rush to contract a work by such a cripple. Aside from the incredible editorial problems of having any substantive manuscript change await the censors' okay and a 30-day review, the publisher himself can be made liable even for an inadvertent slip the author may commit.

Not only does the Reagan order impose an intolerable burden on individual liberties to no positive effect; its aim could be achieved in more constitutionally benign ways.

There are laws on the books which criminalize the disclosure of intelligence agents' names and critically sensitive intelligence derived from radio intercepts, nuclear research, or spy-in-the-sky satellite photography. If these aren't sufficient to deter harmful leaks and the betrayal of intelligence "sources and methods," nothing will.

Besides its ambiguity, superfluity and a potential for overkill, the new directive cannot be made to work. Even the CIA's more modest prepublication review system has proved unmanageable. With 15,000 potential authors on the CIA's active-duty roster, and many more on its retiree lists, the agency's censors have found it impossible to enforce their regulationfs uniformly.

With Justice Department help, they have engaged instead in the most pernicious kind of censorship -- the use of their powers in a political manner. Of countless ex-agents who've refused to submit to prepublication review (Tom Braden, William F. Buckley, E. Howard Hunt, to mention but a few), only six have been chastened by the law. Five of these are real or presumed "critics" of the agency (Marchetti, myself, John Stockwell, Philip Agee and Wilbur Eveland).

Indeed, only after a federal judge found the agency to be flouting the rules of fair play was the prosecution list extended to a noncritic, William Colby, and even then, he wound up with a sweetheart $10,000 fine for failing to clear his memoir, "Honorable Men."

Given the vast number of bureaucrats affected by Reagan's order, there is no way censorship can be administered equitably. Nor do the compromise proposals suggested by some civil libertarians offer a way out. They would limit prepublication review agreements to the intelligence community -- and exempt, say, some State Department officials and some ex-bureaucrats from other agencies. Their ultimate effect would be to damage the First Amendment twice over, by institutionalizing both censorship and discrimination.

Notwithstanding all the defects of the Reagan order, there is something to commend it: its timing. Nothing could be so fitting a curtain-raiser on 1984 as this attempt by the administration to establish a censorship system which belongs, appropriately, in George Orwell's imagination.