Sinan Ulgen has criticized Li Bin and me for taking a "moral" approach to disarmament. He has cast his own views as more realistic than ours. But when Li Bin and I advocate complete disarmament, the main point is to address real threat perceptions and solve real-world security problems. Support for disarmament happens to be the morally correct attitude—but where is the inherent contradiction between moral attitudes and realistic ones?
I can identify several moments in this roundtable when my (and Li's) "moralistic" views have accorded better with global security imperatives than have Ulgen's putatively realistic views. First, Ulgen argues that the world is safer today, when a few states possess nuclear weapons, than it would be if no nuclear weapons existed. But if that's true, why did the nuclear weapon states join the Nuclear Non-Proliferation Treaty (NPT) to begin with? Did they naively accede to a treaty that would make the world less safe? Unlikely. A more convincing explanation is that in order to preserve their nuclear monopoly they gave dishonest (albeit legally binding) disarmament commitments. But a basic requirement for a secure international system is that nations can trust one another to honor their commitments (or can trust, when commitments aren't honored, that viable enforcement mechanisms will come into play). The trust that the international system requires is now eroding where nuclear weapons are concerned. A collapse of trust would threaten global security and ultimately the international system itself. Under such circumstances, is it unrealistic, and merely moralistic, to insist on general disarmament?
Second, Ulgen argues against total disarmament by noting that nuclear weapons, via NATO's nuclear umbrella and US extended deterrence, provide security to non-nuclear weapon states. But even putting aside the illegality of extending a nuclear umbrella to a group of non-nuclear treaty members, many parties to the NPT perceive extended deterrence and NATO's nuclear umbrella as serious security concerns. The problem is that extended deterrence has created a third class of treaty parties. In addition to the nuclear weapon states, there are now two types of non-nuclear weapon states—those that receive the benefits of nuclear deterrence and those that don't. It is this last category of states to which I was referring in Round Two when I wrote that, if nuclear weapons provide security to the nations that have them, "[S]houldn't other nations strengthen their security by obtaining their own nuclear deterrents?" If they do so, the security implications will be quite real.
Third, Ulgen writes that "In a 'zero' world, the incentives for a rogue state to go nuclear would be so powerful that no threat could overcome them—short of the guaranteed destruction of the state and elimination of its leadership." Ulgen believes, in fact, that the disarmament and nonproliferation regime would collapse when the first "rogue state" went nuclear. The term "rogue state" is a subjective, biased classification, not to mention an unpleasant reminder of the George W. Bush era. But beyond that, one state often considered "rogue" has already gone nuclear—North Korea. Has the nonproliferation regime crumbled as a result? It hasn't, so I don't know what lessons can be drawn from a single instance of proliferation.
On the other hand, Israel's possession of nuclear weapons is eroding, in the minds of many Middle Eastern policy makers, the credibility of the disarmament and nonproliferation regime. States in the Middle East perceive Israel's nuclear capability as a direct threat to their security and they have started to question the wisdom of long-ago decisions to join the treaty. If countries across the region withdrew from the treaty, the security implications would be serious indeed. The only workable way to forestall this outcome in the long term is to achieve general disarmament. Once again, Li and I have argued for "zero" on the basis of real threat perceptions and security concerns. Disarmament may be moral—but that is far from the only point.