If the United States is re-examining its nuclear weapons arsenal, shouldn't it first reconsider why it has such a stockpile?

No sane person would argue that the proliferation of nuclear weapons across more national borders would make the world a safer place, or that diminishing the psychological barrier against a renewal of nuclear testing, or the first use of nuclear weapons, is a good thing. Yet policies in Washington seem to be encouraging all of the above.

While railing against the possible development of nuclear capabilities in Iran, and refusing to negotiate with North Korea, Washington officials apparently kept secret their knowledge of a Pakistani initiative to build a new plutonium reactor that will be capable of producing up to 50 nuclear weapons a year. At the same time Washington has been pushing hard for a non-military nuclear cooperation pact with India that would help promote India's bomb-making capacity. Besides the incendiary possibilities of further arming two military rivals, it is important to remember that neither country has signed the Nuclear Nonproliferation Treaty. What message does that send to countries, such as Iran, that have signed the Treaty?

If this weren't bad enough, the US Congress is likely to nearly double funding for the Reliable Replacement Warhead programme, to develop and deploy new warheads to replace all existing weapons in the US nuclear arsenal. It is claimed that these warheads will not need testing, but most physicists familiar with the programme recognize that there will be incredible pressure to do so, before deployment. This would put pressure on China, Russia, India and Pakistan to test their weapons and make it more difficult to convince Iran and North Korea to abandon their nascent programmes.

If the US is re-examining its nuclear weapons arsenal, shouldn't it first reconsider why it has a stockpile, in this post-cold-war world, of nearly 10,000 such weapons? There is no rational reason for preserving this, at incredible cost, in today's world. But then, rationality doesn't seem part of nuclear policy of late. Next month: first use, and what physicists should do about it.