A pessimistic view of the history of arms control with suggestions for a more positive future. Berkowitz, who has roamed the corridors of the intelligence community as well as those of the Brookings Institution and the Council on Foreign Relations, has previously authored American Security (1986). In a nutshell, his stance is that arms control, despite good intentions, has rarely saved money and has often led to deadlier as well as costlier weapons. Berkowitz is driven by a cynical view of human and international relations, and no policy seems to win his approval. ""Strong decisive leadership alone is not a solution to an arms race. Trying to control an arms race through quantitative limits is much like trying to escape from a bog of quicksand: decisive action just pulls one in deeper. . ."" Even seeming progress is only a red herring, as progress brings out ""the pesticide phenomenon"": just as insects develop stronger breeds resistant to pesticides in following generations, ""as arms control progresses, the remaining issues become more complicated and more controversial, not less."" As for verification, Berkowitz is most persuasive in demonstrating how technology has far outpaced the probabilities of accurate accounting. Why can't scientists just stop working on military technology? Berkowitz argues that much technological experimentation is value free, having spin-off applications to the military only after the fact. In all of arms control negotiations, the author finds only two models for success--the 1972 Incidents at Sea Agreement and the 1963 Washington-Moscow Hotline Agreement. These two programs were successful, he states, because they were practically costless, verification was simple, and they dealt with practical issues on a day-to-day basis. Like a deathbed watch--negative expectations with only a glimmer of hope.