Now the Pentagon is toying with the MX Deceptive Dense Pack Basing Concept. D2P for short. The idea is that a better way of protecting our retaliatory ability may be to cluster our MX missiles in fairly dense groups instead of spreading them out, which was the old thinking. The Washington Post's Pentagon reporter, George Wilson, explained it a few days ago: "The first Soviet warhead to come down on a pack will not only destroy its target, but, in a kind of sympathetic detonation, will blow up the warheads flying right behind it on the way to other U.S. missiles in the cluster. These other U.S. missiles may then survive."

D2P would make a swell new cartridge for your home video game. Tricking your opponent into zapping his own warheads should rack up lots of points. And if it doesn't work, you push the reset button and start again. Unfortunately, real live war doesn't have a reset button. Once the nuclear exchange is under way, it's all over for both sides. Survivability, in the terms contemplated by D2P, is a dangerous fiction. Which is what is wrong with D2P. The technical thought behind it is admirable, its cleverness impressive. But it is the answer to a question fewer of us are asking. What growing numbers of us want to know is not how to survive nuclear war but how to avoid it. And in this regard, D2P may be worse than useless. It seeks to answer the question by technology: by fashioning a counter to whatever the Soviets come up with, which means that the Soviets, to the extent that they too are locked into nuclear Atari, must come up with something to counter our counter.

The only way that can enhance anyone's security is if one side or the other takes the next logical step: take advantage of its temporary edge and wipe out the other side. Indeed, it is our assumption that the Russians would do exactly that or at least blackmail us by threatening to do it that leads us to such breakthroughs as D2P. But both sides have long since reached the point where neither can launch even a theoretically winning attack against the other without itself being destroyed. Some of us recognize, even if our leaders don't, that war is not an arcade game.

Nuclear war, as a result, seems increasingly unlikely --except as the result of some hideous chain of errors. And our budget-busting expenditures in preparation for nuclear war in fact weaken our capacity to carry out the conventional war that is far more likely. Suppose Russia moved--conventionally--against Western Europe. And suppose that, even against the full array of NATO forces, the Russians were winning.

Would we launch nuclear war against the Soviets, knowing their nuclear retaliation would be not against Europe but against us? I don't think so. If I am right, doesn't it follow that we would be much better advised to start upgrading our conventional capability?

And, by the way, I don't mean exotic new versions of conventional armaments. The Post's Wilson had another piece, the day after his report on D2P, quoting leaders of the Army and Air Force national guards as calling for simpler, cheaper weapons It noted, for instance, that $3 million could buy a better antitank attack plane than the highly sophisticated F15 fighters that go for $40 million each. Their report cites "a fundamental contradiction between what we have and what we need." They've got it exactly right.