Philosophical approach 

    It is a problem to know how to start thinking about nuclear weapons. Whereas other fields have standard concepts and shared assumptions, nuclear weapons have none. Or the ones they have are wrong.

    Economics, for example, has common concepts and shared assumptions. All economists talk about marketplaces, supply and demand, and the centrality of money. There is an agreed intellectual framework that can serve as the jumping off point for those with new or diverging ideas. What are the main ideas of nuclear weapons? Escalation and deterrence, both of which, on closer inspection, are hollow men.

    Escalation is presented as a process of rational decision pushed inexorably upward by the characteristics of the weapons. A moment's thought about real escalation scenarios, though, will convince most people that revenge is a crucial part of escalation. "You hit me and I'm going to hit you back harder." Escalation is an emotional savage masquerading in a suit. A real exploration of escalation would include the latest work by psychologists and anthropologists on revenge. But those scholars are never invited to sit down at the escalation table.

    Deterrence (as I've said elsewhere) is the result of the characteristics of missiles. Conflate the characteristics of missiles with nuclear weapons and you'll end up with a conceptual jumble. How can that serve as a conceptual foundation for an entire subject?

    The real giveaway is the number of important works on nuclear weapons that have history as their backbone. Chronology, typically, is the organizing principal of last resort. It works well for stories, but not as a tool for conceptual analysis. Imagine studying love chronologically. A conceptual study would organize the topic by kinds of love: sexual, familial, romantic and so on.  Chronology works for "telling the story" of nuclear weapons. Real intellectual analysis is not story telling, however.

    There are some works, of course, that are not chronological. (Schelling's books come to mind.) However, the structure of these few is not the same from book to book. They have no common form or shared assumptions. 

    My original analysis of this problem was that nuclear weapons were either a military, political or moral problem. Initially I thought they were military. They were used in war, after all. But the military sphere is subsumed under the political and the importance of nuclear weapons seemed too large for such limitations. Also, the vast majority of cases in which nuclear weapons have played a role in human affairs are instances of diplomatic threatening, not use in war.  This led to the second possibility.

    So I scrapped the first outline and created one that used politics as an organizing tool. This was in keeping with those people who urged the notion that deterrence was at the center of things nuclear. But it is possible to do more than just threaten with nuclear weapons. A political approach did not give enough emphasis to war. And political approaches are regularly criticized for not including enough moral discussion.

    Of course, making morality the underlying framework didn't work either. Too much of the debate, strategic and political, ignores morality for it to be the central conceptual underpinning. 

    I was visiting my friend Kevin in Nashville and doing some reading in the Vanderbilt library on the Cuban missile crisis. In one of those moments of realization that you read about in books, I suddenly understood that what I wanted was to focus on use. I've read a good deal of both William James and Ludwig Wittgenstein, two philosophers who focus on everyday use and actual practical outcomes as a way of thinking philosophically. I realized that it would be possible to underpin the entire nuclear topic by employing use as the organizing principal. And the practical outcome of that use, the "usefulness" of the weapons in various situations, would provide answers to important questions.

    The large divisions would be uses in war, political uses and other uses (to destroy asteroids, etc.) Use in war would be subdivided into use on the battlefield, use against rearward military targets, use against economic targets, and use against civilians. Use as threats would be divided into deterrence, compellence and existential threats and influence. Since moral reactions are a practical outcome of real actions in the real world, the study could be shot through with considerations of morality.  In every case the practical end result would be the criterion, not emotions about the means or theoretical models built on little experience.
    Almost right away this new approach began to pay dividends. I saw relatively quickly that the theoretical approach of one important segment of nuclear strategists was seriously flawed. I realized that it would be possible to use history to illuminate discussions of the use of nuclear weapons in war. I noticed that much of the discussion about nuclear weapons focuses on them as means, rather than paying attention to the outcomes - the ends - they create.

    Usefulness, I decided, was the key.