The run-up to the Reagan-Gorbachev summit is producing some high drama within the administration, and some high stepping in respect to the official reading of the Anti-Ballistic Missile Treaty of 1972.

Successive administrations, including this one, had always read the treaty to ban testing and development of futuristic missile-defense systems. But the other day, National Security Adviser Robert McFarlane popped out with a new reading permitting testing and development of President Reagan's Strategic Defense Initiative.

This is no matter for small quibbling. In some quarters inside the administration and in many quarters outside, it is accepted that if some testing and development limits are not applied to defensive arms, the Russians will lose most if not all of their taste for limiting offensive arms. This is what the summit is likely to be most about.

At this point, things get lawyerly. Having been taken around the several tracks, I submit that the new administration position is, at the least, painfully labored. "Agreed Statement D" can be read, as the administration now reads it, to countenance steps short of deployment for space-based ABM systems and to subject only deployment proper to mutual "discussion . . . and agreement." But Article V flatly bans attempts to "develop, test, or deploy" such space- based systems -- SDI clearly is one.

How could an Agreed Statement seem to open a loophole for something outlawed in an Article? It turns out that the dispute is not simply over the text but over the negotiating record, which happens to be secret. Those who negotiated the treaty insist that the Agreed Statement -- though admittedly it doesn't say so -- was meant just to keep open a way to exotic variations of the fixed land-based systems permitted by the treaty, not to welcome exotic space-based SDI-type systems. The administration's new reading supports its intent to give SDI lots of running room.

Not for the first time, one is struck here by the reversal of Soviet and American positions. In the early 1970s, it was the Americans who wanted to restrict antimissile work, on the theory that it made little sense to limit current models and leave the door open to future models. The Soviets, either plodding or calculating, were wary of writing limits without specifying what was being limited.

A number of times, evidently, the Soviets listened without taking exception to expressions of American interest in a broad future ban. The record is said to show only one occasion, whose significance is disputed, where the Soviets indicated (but didn't explicitly declare) assent.

In sum, Article V ("Each Party undertakes not to develop, test, or deploy . . .") is clear-cut, but the stuff of dispute and distrust is still there. This is, alas, the Soviet-American pattern. For all of that, politically the new position is a bomb. The administration looks two-faced in surfacing on summit eve a brand new treaty interpretation that liberates SDI, whose containment is Moscow's summit priority. The explanation that officials didn't get around to looking hard at the treaty until recently is embarrassing. The new position cuts across the professed American eagerness to gain firmer Soviet compliance with the ABM treaty. It will be tough to sell to the American public and the European allies.

It is quietly suggested that, notwithstanding the words of McFarlane, Reagan has not personally signed off on the testing-and-development issue. This is -- keep in mind -- not just one more skirmish with his liberal critics but possibly the make-or-break issue of the summit.

If the president is convinced that pushing the pace on SDI is vital to long- term American security, then he must be prepared to risk certain interim costs: frustration at Geneva, more tension abroad (although not the end of the world) and a new surge in the foreign policy debate that he largely stilled in the last year by turning a diplomatic face to Moscow.

I say this on the presumption that Gorbachev does not have unlimited political license and must produce results -- in the form of some progress toward an arms agreement -- if he is to sustain his experiment in dealing with Reagan, whom his predecessors had proscribed.

I wonder if anyone has made the case to Reagan that he could live with the old American reading of the ABM treaty without undercutting his SDI. The argument is that the research must necessarily go on for a long time and the decision to test and develop can prudently be deferred; meanwhile, pocket the gains from deep cuts in offensive arms.