Commentary is a regular feature of the Crimson editorial page that provides a forum for opinion from members of the Harvard community. Those interested in contributing pieces should contact the editorial chairman.

THE REYKJAVIK SUMMIT raised hopes about an agreement for the abolition of nuclear weapons, but it appeared to founder on the President's intransigence over his Strategic Defense Initiative. One proposed utopia appeared to defeat another. How should we think about SDI and the long-term future of nuclear deterrence?

Since humans are fallible, it seems reasonable to assume that nuclear deterrence may someday fail. This is not a certainty and if we can continually reduce possibilities, failure is not inevitable in any meaningful time frame.

Moreover, even if the failure of deterrence were "inevitable," this proposition does not necessarily entail the apocalyptic conclusions that are often drawn from it. Failures need not be large ones that lead to extinction of the species. Even if a large failure is inevitable, that inevitability may come after thousands of years. What we do now can affect these outcomes.

A technological solution has received the greatest attention in the past three years. President Reagan has called his SDI a "moral imperative." But the language of morality has been bandied about loosely and inappropriately by both sides in the so-called "Star Wars debate." It is a good example of stunted moral reasoning in the discussion of nuclear issues.

IN HIS MARCH 1983 speech, the President suggested that strategic defense might provide an exit from the dilemmas of nuclear deterrence. The goal of SDI was to make nuclear weapons "impotent and obsolete." But to escape from deterrence requires a leak-proof defense not only against ballistic missiles, but also against bombers, cruise missiles and weapons smuggled into our cities. Such a perfect defense seems unlikely.

Few scientists or defense officials believe in perfect defense. Some who doubt the technological feasibility of such a task have urged that the Strategic Defense Initiative Organization in the Pentagon concentrate on the lesser task of defending our weapons rather than our cities. The rationale would be to enhance rather than replace deterrence. That goal is very different from the President's original speech. In a sense, the public is being sold a product that does not match its packaging. An elementary concern for truth in advertising should lead us to distinguish "SDI 1" (The March 1983 brand) and "SDI 2" (the partial defense being developed).

"SDI 2" is no more a moral imperative than are alternative ways of enhancing deterrence. While a case can be made (and debated) for such defense, it has to surmount the obstacles of feasibility and cost in terms of competing strategies and competing moral claims on resources. It is not moral simply because of good intentions.

Ambassador Paul Nitze has set forth three criteria that must be met if "SDI 2" is to be technically feasible. 1) It must work as a system. This means more than fancy weapons. It means an enormously complicated system has to work without ever being tested under the stress of nuclear war. 2) It must be cost-effective. If it is much cheaper for the Soviets to add offensive missiles than for us to add defenses, we will merely provoke an increase in the offense without being more secure. 3) It must be relatively invulnerable. If it is easy to destroy or foil an SDI by surprise attack, it will be a tempting target creating instability in times of crises.

These questions do not mean that there should not be a long-term research program on strategic defense. Ouite the contrary. A modest program is a hedge against Soviet work in this area. Moreover, the prospect of enhancing deterrence, and perhaps over the long haul of being able to save a large number of lives in the case deterrence fails, justifies considerable efforts. The value of such an initiative will depend on the consequences, not the motives.

It may be "better to defend than to avenge," but only if the consequences of trying to defend do not increase the risk of nuclear conflict in the meantime. Those consequences are likely to be determined not by our intentions, but by the technology chosen. How will it affect crisis stability and the political state of American-Soviet relations (will the introduction of defense be in a cooperative or antagonistic setting)?

THE POLITICAL PATH lacks the glamour of the technological one, and, as many technologists have discovered, politics is harder than physics. But political change is the key factor in the long-range future. Further political change may take a number of forms. There may be changes in relations between the United States and the Soviet Union; there may be changes in the growth of international institutions and cooperation among states; there may be changes in domestic political and social attitudes toward the sovereign state and its defenses.

Theorists of international cooperation point out that warm relations and trust are not necessary for cooperation. Competition does not prevent cooperation, nor is a central authority necessary. What is needed is a more sophisticated and long-range perception of self-interest, an ability to learn from experience and the realization that a relationship will continue over time. When games of "Prisoner's Dilemma" are played over long periods, a strategy of reciprocity proves to be most effective. On such realist premises, it is quite possible to expect a gradual evolution of American-Soviet cooperation.

Certainly relations in the 1980s are vastly different from the 1950s. Despite frequent allegations of return to the Cold War, there is far more contact and cooperation now than 30 years ago. There is far greater awareness of each other's perceptions of vital interests and internal processes. It is highly plausible to expect further change in the relationship over the next 30 years and more.

In addition, there may be changes inside the Soviet Union, for while it is unlikely to change quickly, it does evolve slowly and unevenly. It has opened up somewhat over the last 30 years. There are more contacts as well as more sophistication in the perception of outside reality. There are more pinholes letting light into the "black box."

Of course, this view of the political future is only one of many possible outcomes. Other benign, or horrendous, outcomes are possible. What is important is that we should not let our imaginations be captured solely by images of imminent holocaust or by cynical views of the immutability of our dependence upon nuclear weapons, for both tend to stifle the modest efforts we can and should make now so that the next generations have more opportunity to develop a better future.

Similarly, we should not let the current debate about a single technological utopia such as SDI so dominate our thinking that we fail to evaluate its various dimensions in relation to alternative visions of the long-term future. A responsible moral approach to the nuclear future requires us to avoid large risks to the crucial values of survival and freedom that we must pass on to future generations, and to make continual efforts to reduce reliance on nuclear weapons whenever that does not increase risks. Such an approach requires hard thinking about uncertain probabilities and proportionate risks rather than succumbing to the escapism of popular utopian solutions.