So what happens, I asked, if adversaries threaten U.S. allies with that kind of an attack? How credible is a U.S. policy of extended deterrence based on full-size strategic weapons? “Not very,” Watts answered. My view: it’s like a police department whose only force option is to blow up the entire block where the perpetrator lives.

Indeed, U.S. extended deterrence is something that not enough people think about when they advocate further cuts in U.S. nuclear forces. The American “umbrella” covers nations such as South Korea, Japan and Turkey, which have the industrial and technological capability to go nuclear very quickly if they feel that they can no longer rely on the U.S. Other nations on the edge of nuclear capability include Saudi Arabia – which is considered likely to get weapons from Pakistan if Iran publicly goes nuclear.

Watts also had some comments on the controversy over Georgetown University Professor Philip Karber’s assertion that China has a much bigger nuclear arsenal than most experts accept, based on his early-1990s role in the Gulf War Airpower Study. “When we got into the Iraqi nuclear program, we found that they had switched after Osirak [the Israeli strike on Iraq’s research reactor in 1981] from a plutonium-based program to uranium,” Watts notes. “That had never occurred to our intelligence or nuclear-weapons people – but we found that instead of six to eight nuclear-related targets in Iraq, there were actually 50 or 60.” And, as Watts points out, the intelligence community has not been focused on Chinese nuclear developments since 2001.

The result: there is no single standard “firebreak” between conventional and nuclear weapons. If the conventional revulsion against their use does not hold, Watts warns, “limited use of low-yield nuclear weapons will become the new normal and give rise to a second nuclear age whose dangers and uncertainties will dwarf those of the first.”