The recent Contadora "peace scare" made it clear that Central America's three years of talks have almost become an end in themselves, serving the interests of all sides merely by continuing.

While there were a few moments of excitement last month when it looked as though a peace agreement might be at hand, they faded before a self-imposed June 6 signing deadline, and the talks have since returned to their plodding pace.

Further false alarms are predictable as new deadlines come and go, with one player or another drawing back at the last moment. It is unlikely that the Contadora process will produce regional peace, but equally unlikely that the talks will be abandoned.

In May, U.S. conservatives withdrew in alarm over the pact then taking shape, convinced it would force U.S. abandonment of anti-Sandinista rebels without disarming Nicaragua. Last fall, the leftist Sandinista rulers of Nicaragua walked out, charging that that draft would disarm them while leaving the rebels intact. This week, El Salvador's foreign minister announced that "the guidance of Contadora has disappeared" and that the four Central American nations -- excluding Nicaragua -- will take the lead from now on. That, of course, sparked new speculation that Nicaragua will soon sign the latest treaty draft.

This dance occurs because any comprehensive, enforceable peace pact would fundamentally change the political equation in the region in ways no one is sure would be desirable. The fundamental issue -- whether the Sandinistas will remain in power in Nicaragua -- may not be subject to resolution by treaty. But talking about it at least delays a military decision.

A pact would require the United States and its allies -- El Salvador, Honduras, Costa Rica and Guatemala -- to accept the Sandinistas as legitimate, even though they would then be under an obligation to become more democratic.

There is tremendous resistance among U.S. conservatives to that change. They are determined not to legitimize Nicaragua the way they claim President John F. Kennedy legitimized Cuba by promising in 1962 that the United States would not invade.

The conservatives also say they think that Nicaragua would never honor a pact, no matter how hedged with guarantees. In addition, a pact would force the Reagan administration to abandon its hope that the anti-Sandinista rebels, known as contras, might somehow force the Sandinistas from office.

That would mean the end of what has become a closely watched field test of the Reagan doctrine, the theory that the United States can roll back communist governments by backing domestic insurgents instead of sending in the Marines. No Reagan decision-maker has publicly given up on the contras.

The Sandinistas, in turn, want a pact that would rid them of the contras. But they would be forced by an accord to end any concrete expressions of solidarity with leftist guerrillas in El Salvador, Colombia and elsewhere, and to accept the kind of press and political system they regard as tools of an exploitative economic elite. They would no longer have the contras to blame for Nicaragua's repressive state of emergency and its ailing economy, and they would have to face the possibility of being voted out of power.

In short, the Sandinistas agree with U.S. analysts that the revolution is not yet consolidated and is subject to reversal. But the United States sees reversal as obligatory, and the Sandinistas see it as out of the question. No treaty is likely to bridge that gap.

Both sides, however, benefit from continued negotiations.

President Reagan's request for $100 million in new contra aid has failed to pass because moderate members of Congress want to exhaust diplomacy first. The Sandinistas know they are not popular on Capitol Hill and do not want the stigma of breaking off talks. But the contras, or counterrevolutionaries, are not popular either, according to public opinion polls, and the administration knows a signed treaty would doom the aid program.

While Contadora continues, neither side gains nor loses and both appear to be open to persuasion. In addition, Congress avoids having to make a hard choice. The only ones frustrated by more discussions are the contras.

The talks are also Nicaragua's principal evidence that it is not isolated within the otherwise heavily democratic Latin American community. But the eight Latin nations directly pushing a pact have different interests from the Central American nations that would sign it, and more talks benefit them, too.

Colombia, Venezuela, Mexico and Panama, which launched the Contadora process in 1983, want above all to avoid a U.S. military intervention in the region that would fire up leftist elements even as it forced governments to take sides.

Argentina, Bolivia, Uruguay and Peru -- mobilized by Peru as a "support group" when talks flagged last year -- also want the Marines to stay home but are more interested in ending the U.S.-Soviet dimension of the conflict.

Central American unrest has escalated hemispheric tensions and deferred U.S. and other foreign investment. The eight Latin nations have massive debt and development problems that they say dwarf Central America's squabbles. Many Latin diplomats say any Contadora pact would defuse things long enough for more pressing issues to get the U.S. attention they deserve. They are pushing harder for an accord than any of the five principals.

At the same time, Contadora has provided virtually nonstop contact for Latin leaders with U.S. officials, who are constantly being reminded of wider regional needs. Continued talks are almost more valuable in this context than an accord, which would turn U.S. attention from solving the problems to enforcing the agreement.

All sides agree that enforcement is crucial to any pact, but they know it is impossible unless all sides want to abide by the treaty terms. In the absence of that fundamental accord, the rewards of talk appear to be enough for now.