Security correspondent, BBC News
A new report from a respected British nuclear weapons scientist warns that a military strike on Iran could speed up rather than slow down Iran's production of a nuclear bomb.
The US has not ruled out the use of force against Iran
It claims it would bolster domestic support and increase the country's willingness to use all means possible to attain a weapon.
In his report, Frank Barnaby argues that an attack might not destroy all of the nuclear programme. In its wake, it would be much more feasible for Tehran's political leadership to pull out of the Non-Proliferation Treaty and launch a crash programme, devoting maximum resources to developing one or two bombs as quickly as possible.
This, it is argued, means that a nuclear-armed Iran might end up occurring sooner rather than later as a result of military intervention.
The Iranian government denies it is seeking nuclear weapons and insists its interest in nuclear technology is for peaceful purposes only.
The UN atomic watchdog, the IAEA, has been unable to either rule out or confirm a weapons programme based on what it has seen. But international proliferation experts generally agree that Iran is, at the very least, seeking to develop technology which could quickly be diverted towards weapons if required.
The US has examined the possibility of military strikes on other countries' nuclear facilities in the past.
It came closest in 1994, when a White House meeting discussing whether to strike North Korea was interrupted by news of a possible deal over the country's nuclear programme.
The option of military strikes against Pakistan's Kahuta plant were also examined in the late 1970s but ruled out because the chances of success were rated too low when compared to the consequences of going ahead.
But there is one important precedent for an attack on nuclear facilities.
In June 1981, eight Israeli fighter jets took only 90 seconds to destroy Iraq's Osirak reactor in an audacious bombing raid. It is sometimes cited as a precedent for a US or Israeli (or joint) attack on Iran, but is it really a useful parallel?
In that case, Israel had come to the conclusion that the US and the international community were not willing to take sufficient action after Iraq purchased a reactor from France. Once it became clear that diplomatic pressure and covert methods had failed to stop Iraq - and also that an attack by commandos was too difficult - Prime Minister Menachem Begin ordered an aerial assault.
But did the Osirak raid stop - or even significantly slow down - the Iraqi nuclear programme?
The evidence is not conclusive. In terms of intent, the raid did not stop Saddam Hussein, it only forced him to change tactics for achieving his goal of a nuclear bomb and also intensify his work.
In the wake of the Israeli raid, Saddam Hussein personally summoned an experienced British-educated scientist Dr Jafar Dhia Jafar from jail. "He told me we must develop a deterrent," Dr Jafar recalled in an interview with me shortly after he fled Baghdad in 2003.
It was in the wake of the raid that Saddam Hussein moved far more definitively towards an active weapons programme rather than a latent programme which could be diverted towards weapons at a later stage.
Iran's Natanz nuclear plant is largely underground
And the attack failed to have any deterrent effect within the country. "After the Israeli bombing of June 1981, many scientists came and applied specifically to work on the programme," Dr Jafar recalled.
The number of scientists increased from 400 to 7,000 and Saddam Hussein poured far more resources into the programme - something like $10bn over the coming years. He was also far more careful in hiding the programme from the outside world.
The result was that when inspectors scoured Iraq after the 1991 war they found that it had made much more progress than anyone had realised (although what many failed to spot, according to Dr Jafar, was that in 1991 Saddam gave orders to destroy the programme).
Iran - and other countries - have also learnt from the Osirak raid by dispersing their nuclear research over a number of sites and by building plants such as Natanz deep underground covered by layer upon layer of earth and concrete, making the effectiveness of even bunker-busting bombs questionable.
There are also far more sites in Iran now than there were in Iraq back in 1981 and there are real questions over whether US and Israel can be confident enough that their intelligence has sight of all of the programme. Because of the way in which states learnt from the Israeli raid on Osirak, that strike may well be a one-off in terms of effectiveness which cannot be easily replicated.
So a strike against Iran would risk leaving more of the programme and knowledge intact than was the case in 1981 but could have the same political impact in terms of increasing the determination to develop nuclear technology as fast as possible. Of course, much would depend on the intensity of the strike - but a prolonged strike might lead to many more civilian casualties and a much greater international backlash.
Iran insists that its nuclear programme is for energy only
As with many of the recent reports on the options against Iran, the latest report by Frank Barnaby and the Oxford Research Group emphasises the negative consequences of taking action.
But what is less fully analysed or debated are the consequences of failing to act and of Iran actually developing a nuclear arsenal (if indeed that is what it is seeking).
Even if traditional deterrence made it unlikely that Iran would use the weapon, it could embolden Iranian behaviour across the region, directed not just against Israel through allies like Hezbollah but also against other states in the Gulf.
In turn it would almost certainly lead other Gulf states like Saudi Arabia and Egypt at the very least to consider whether they too required a nuclear option.
If the choice was between the status quo and military action then there would be no real need for a debate but, in the minds of many experts and policymakers, that is not the choice that is being faced.