FROM THE START of the nuclear age, two conceptions of strategy have vied. Deterrence, emphasized by the United States through the years of its clear strategic preponderance, threatens all-out war against any Soveit nuclear attack or ultra-high-risk non-nuclear challenge. Behind this strategy lies the premise that the way to make nuclear war unthinkable is to rule out in advance that any Soviet nuclear-level threat -- large or small, against us or an ally -- would be met by a less than total American response.

But some strategists have always wondered about deterrence. They have feared that, with the advent of strategic parity or perhaps a Soviet advantage or the prospect of one, the Kremlin might not be deterred, or our allies might not think they would be. They have feared, too, that the growing Soviet capactiy to destroy American land-based missiles in the ground might embolden Moscow to attempt a first strike on the silos and then sit back and dare Washington to open a countercity nuclear exchange. They have argued for a war-fighting strategy to spare the United States such a "suicide or surrender" choice. Behind this strategy lies the premise that it is not the size but the certainty of a nuclear repsonse that would actually deter Moscow, and that the threat of a total response to a limited challenge, or to a challenge in disadvantageous circumstances, is simply not credible.

This is the context in which to assess President Carter's reported new directive to the Pentagon to make certain changes in the direction of a war-fighting strategy. The changes evidently involve less targeting on Soviet cities and more targeting on Soviet military forces and political and military command centers. The new directive is said to take into account the political fact of widespread doubts that, in a condition of Soviet parity and perhaps parity-plus, the United States would put its cities at risk by aiming at Soviet cities: to gain credibility, you aim at military targets and you let that be known. The directive also reflects the technical fact that both great powers are well along in developing the capacity to destroy each other's missiles in the ground: in these matters, once you have a certain capability, it customarily follows that you change your declared strategy in order to use it.

If it all sounds unreal, that is because it is unreal. No nuclear weapon has been fired in anger since World War II. The destructive power of nuclear weapons makes demands upon the imagination that are very difficult to fit back into reality. It is impossible for a country that possesses nuclear weapons not to be preoccupied with how it could use them in a crisis, but it is a fair question whether any strategy conceived in the calm would guide decisions in the storm -- though strategy certainly does affect the building of forces, the planning of their employment, the foreign policy of which they are a central part, and other pre-crisis preparations. It is all of these, apparently, that Mr. Muskie, who became secretary of state just as the president's directive was being put into final form, wishes now to look over.

Perhaps the best one can hope for is that, whatever this or that directive says, the argument will go on. No riddle is worthier of constant and total attention. The debate may seem and be unreal. The objective is to keep nuclear war unreal.