It is clear that a workable arms control process must be consistent with perceived notions of national security; the problem is that there are very few concrete notions about what national security really is or how to achieve it.
At one end of the security spectrum, more and better of everything seems to be the prevailing logic. Reaction to this disputed position fueled a worldwide antinuclear movement of potentially revolutionary proportions. At the other end of the spectrum are those who support a nuclear ''freeze'' - a concept which has been dismissed by force designers, security analysts, and arms control practitioners as naive. However, suppose a freeze of some kind is imposed by law or practice. What has really been achieved? If the freeze turns out to be one-sided, we will have succeeded only in paralyzing our own flexibility to the Soviets' advantage. Assuming, however, that the freeze is genuinely mutual and includes solid provisions for reducing arsenals, it is not axiomatic that less is always better. Indeed, we must recognize quite specifically that, under certain circumstances, there can be greater inherent risks in maintaining small arsenals than in large ones.
The question ''how much is enough?'' is always difficult. With no doctrine on the initiation of nuclear war, the chief United States concern is to deter the Soviets from attacking by assuring that the US can and will retaliate with a force unacceptable to them.
The danger with small nuclear forces is that even marginal alteration in agreed-upon numbers can create a positive - and possibly decisive - difference in the real balance of power at a time of crisis. Say, for extreme simplicity, that the US and the Soviet Union mutually agree to reduce their ICBM arsenals to 100 missiles apiece; at this level, each believes that it could absorb a first strike and still inflict unacceptable damage on the other side.
Now assume that the Soviets secretly add only 40 missiles to their arsenal - a level of treaty infringement which is virtually impossible to detect. With 10 warheads on each of these missiles, the Soviets would now possess the potential capability of eliminating our entire missile force with a first strike.
With small nuclear forces, therefore, undetected treaty violations or even minor errors in verification can propagate overwhelming strategic advantages for the attacker. The incentives to cheat and the pressures for hair-trigger defensive launches increase even as the ability to deter war through the threat of retaliation decreases. One may suspect but cannot be sure that the other side is cheating; the costs of being wrong are so high that both sides may very well adopt an offensive ''launch-on-warning'' strategy. The smaller each side's forces are, the more the verification of treaty provisions becomes essential . . . but in reality unachievable. Casting aside the usual political rhetoric, the instabilities arising from the wrong kind of agreed-upon arms reduction measures may turn out to increase already deep-seated fears, the kind that may lead easily to thermonuclear war.
This is not to imply that nuclear arms reduction is not possible. Rather it is to insist that the process of mutual disarmament not be permitted to become a source of ad-ditional instabilities. One way is to to make existing US forces more efficient by exploiting the ''complementar-ity'' among the various systems. It is easy to create the illusion of vast force requirements, but effective deterrence should be based on the integrated capabilities of theater and strategic land-, sea-, and air-based systems.
Under current planning, US strategic and intermediate-range nuclear weapons are independently pretar-geted without exploiting adequately the cross-targeting potential among these weapons. If we were to reorganize the decisonmaking chain of command in a way that preserves or improves centralized control, but permits rapid and flexible retargeting, we would in effect multiply the utility of our existing forces; this, in turn, would permit substantial numerical force reductions without diminishing the overall defense of the West.
Another path toward stable force reductions would be some kind of buffering mechanism to extend the options for action short of all-out nuclear conflict. The common and frightening scenario - a massive, all-out attack against our nuclear forces - need not be either the most strategically damaging or the most technically probable. The nuclear perils that lie ahead are no longer limited just to the actions of the superpowers. Nuclear war can spread catalytically as the result of a detonation occurring at any trouble spot in the world; it can result from the effort of a rogue state intent on promoting global carnage; it can occur by accident or as the result of miscalculation at a time of crisis.
Were the unthinkable to occur, the nation must have a way to give hundreds of millions of people a second chance, a way of saying ''stop'' to the parties involved. One such buffer might be a limited - indeed, readily saturable - ballistic missile defense (BMD) designed to cope only with accidental launches or with just the first few warheads of attack against our missile sites.
The US cannot and should not seek to create an anti-nuclear umbrella over the entire US. Even with quantum technological improvements, we will never be able to guarantee near perfect security against nuclear attack. The failure to negate a few incoming missiles targeted against our nuclear forces creates a statistical problem - calculating how many missiles we have left with which to counterattack. The failure to destroy only one incoming warhead aimed at a city can mean the death of tens of hundreds of thousands of people - a risk we ought not be willing to take.
A precisely limited BMD system would allow the US to deal with miscalculations or with nuclear accidents defensively without attracting war to population centers. At the same time, it would deny any advantages of numerical cheating after reaching a stage of having negotiated small nuclear forces. Whether one regards the Soviet Union's intentions as benign or demonic becomes irrelevant; a limited BMD capability on both sides offers both a cooling-off period at a time of accidental or induced crisis as well as some protection in the event that faith in bilateral agreements proves misplaced. (This need not deny either side's ability to retaliate following a first strike.)
The end goal of arms control is stability, not stasis; initiatives like BMD should not be discarded without a serious reconsideration of their potential to achieve dramatic weapons reductions without dramatic reductions in national security.