"Offense is defense, defense is offense. Killing people is good, killing weapons is bad." This grisly piece of Orwellian doublethink may seem a decade premature, but it has already served 20 years as the basis for U.S. nuclear strategy. Judging from the historical record, it has served rather well. There has been no nuclear war, and with one significant exception, we have shown commendable restraint in brandishing our nuclear sword.

At the heart of the nuclear balance of terror is the concept of mutual assured destruction. Each side maintains nuclear attack forces sufficient to inflict unacceptable damage on its opponent even if the other side strikes first.

The greatest deterrent to nuclear war is a strategic force which can weather an opponent's first strike and still retaliate in kind. In such a situation neither side has an incentive to strike first. That temptation would arise only if one side could destroy the other's weapons in a pre-emptive strike. Then a tremendous advantage would accrue to the side striking first. Similarly, if one side had some means of protecting its population from nuclear attack, it might also be tempted to begin nuclear warfare, because it would be reasonably safe from retaliation.

So threatening to kill people is good: It deters nuclear attack. Threatening to kill weapons is bad because it challenges the retaliatory capability of the other side. An invulnerable nuclear force is the best deterrent, so offense is the best defense, and an effective ballistic missile defense--an ABM--upsets the balance: Defense is offense.

Critics of mutual assured destruction think the strategy's acronym is well-deserved; undoubtedly, it is an M.A.D. basis for national survival. There is something particularly ironic about Nixon and Brezhnev, all smiles and champagne, signing an ABM Agreement which prohibits either side from defending itself against a missile attack by the other. You and I and millions of other Americans are Brezhnev's hostages in the game of nuclear strategy. President Nixon gave us to the Russians in return for the right to hold a hundred million Soviet citizens as his hostages. So it goes.

A number of questions follow from this strategic paradox. Is mutual assured destruction the best strategy we can manage? If it isn't, what strategy should we adopt? And if it is, what strategic forces does the strategy require? We should address these questions in reverse order because illogical as it might seem, that is how they are considered within the government.

How many nuclear bombs does it take to deter a Soviet nuclear attack? One U.S. Poseidon submarine carries 16 missiles. Each missile carries ten separate warheads; each warhead has about twice the destructive capability of the atomic weapon which destroyed Hiroshima. So one American submarine can destroy 160 Soviet cities, a threat which is surely sufficient to deter a Soviet leader with any shred of rationality.

The United States has 41 ballistic-missile carrying submarines with over 600 missiles, and a fleet of over 400 long-range bombers. In the missile force alone, the United States has over 8000 separately targetable nuclear warheads. Ten years ago we had only one-tenth that number.

The dramatic increase in the number of nuclear warheads is the result of several factors, few of which have anything to do with mutual assured destruction. Interservice rivalry, pressures from the military-industrial complex, and the imperative of weapons technology all played a part. Whatever the reasons for the increase, we now find ourselves in the unsettling situation of having enough warheads to destroy every Soviet city many times over. Even the Pentagon admits that the force is more than enough for assured destruction. Still, we are told, we need new and better weapons.

Why? What strategy could possibly justify the maintenance of so many nuclear weapons, much less additions to the force? The answer is a strategy known as counterforce: the ability to attack and destroy an opponent's military forces.

Counterforce is based on an appealing premise: If a nuclear war breaks out, we should have options other than attacking our opponent's cities or doing nothing at all. The problem with counterforce is that it is strategically destabilizing. If one side has a capability of destroying the other's missiles, then it has an incentive to strike first in a crisis. And if the other side realizes its missile force is vulnerable to a pre-emptive attack, it might decide to strike before it is hit.

Assured destruction on the other hand is relatively stable. It matters little which side strikes first. The other side will have enough nuclear punch left to inflict a prohibitive level of damage.

The debate over assured destruction and counterforce has raged, with varying degrees of intensity, for a number of years. But while theoreticians and government officials argued the pros and cons of different nuclear strategies, weapons technicians continued the job of programming the available warheads to strike targets in the U.S.S.R. if a nuclear was should ever occur.

As the number of warheads in the strategic inventory increased, the targeters found themselves running out of assured destruction targets. Once all the population and industrial centers were covered, the targeters found themselves with excess warheads. As the surplus continued to grow, there was nothing left to do but target military installations, including missile fields. So while the theoreticians continued their great and noisy debate, the technicians quietly gave us the beginnings of a counterforce capability. No policy decision was ever made in favor of counterforce, but the capability continued to grow. The forces began to determine the strategy; the tail began to wag the dog.

Today there are two ways of looking at our strategic forces. Seen through the lens of assured destruction, there are far too many strategic nuclear forces and we ought to look for ways to reduce the force level. Seen in counterforce terms, however, the force is woefully inadequate both in quantity and quality. We need more missiles with more warheads, with higher yields and better accuracy. Not surprisingly, the Pentagon has supported this latter interpretation.