Being an Iranian nuclear scientist has become a high-risk occupation. The recent death of Mostafa Ahmadi Roshan, the 32 year-old deputy director of one of Iran’s uranium enrichment facilities, brings the total of assassinations to five among what was a small group of people to begin with. There are two aspects to this act – the morality of such assassinations and their efficacy; I’ll talk about the latter aspect first.
One can guess that these killings are aimed, like the Stuxnet computer virus, at slowing down Iran’s progress towards developing nuclear weapons. Targeting critical (if you’ll pardon the pun) individuals is understandable, but one has to wonder how effective a tactic this might be. One way to think about this question is to consider what is required to develop a nuclear weapon and to ask if the tactic of targeted assassination is an effective way to interrupt this process. The minimum requirements would seem to require an aspiring nuclear power to have:
- The basic theoretical knowledge
- A process to produce fissionable materials
- The technological ability to produce fissionable materials and assemble a nuclear weapon
Consider the first point – the basic theory behind nuclear weapons is so well-known that, in 1977 a Princeton undergraduate student wrote a term paper in which he outlined the design for a nuclear weapon. And, for that matter, most of the underlying science is readily available on-line, not to mention in books and other references that are readily available. As one example, I did an on-line search using the terms “nuclear weapons design” (which probably endeared me to whoever keeps track of internet searches) and found a surprising amount of information in less than an hour. At one point in the past – perhaps during the early days of the Manhattan Project, assassinating a few key scientists would have put a crimp in American efforts at nuclear weapons development but we are far beyond those days. Killing Iranian nuclear scientists might delay an Iranian nuclear weapon but it may not be anything more than a speed bump.
Producing fissionable materials is more of a bottleneck. Natural uranium, of which only about 0.72% of the atoms are fissionable U-235, cannot sustain nuclear fission without a lot of coaxing (using heavy water or graphite to help a reactor attain criticality) because there simply aren’t enough U-235 atoms to keep a chain reaction going. In order to maintain a chain reaction for a critical nuclear reactor the amount of U-235 must be boosted to at least 1%, and most commercial nuclear reactors use fuel in which the U-235 concentration has been enriched to at least 3%. Weapons-grade uranium contains as much as 90% U-235 (anything in excess of 20% is outlawed for non-nuclear weapons states). Uranium enrichment is simple enough in theory but, in practice, it is neither easy nor cheap – the current “best practice” is to use a series of high-speed centrifuges to separate the very slightly lighter U-235 atoms from the very slightly heavier U-238. These centrifuges are tricky to build and are even trickier to chain them together in what is called a “cascade,” but the process is merely difficult and the basic technology is decades old. And we have to remember – under the nuclear non-proliferation treaty any nation can legally enrich uranium to 20% – it is only illegal to exceed that level. So Iran is perfectly within its rights to enrich uranium as high as 20% – just as the Dutch and Japanese are permitted to enrich uranium. But we don’t worry much about the Dutch or Japanese developing nuclear weapons – we trust them to uphold their treaty obligations.
The third factor is access to the level of technology needed to produce nuclear weapons. The problem here is that the technology needed to produce weapons-grade uranium is exactly the same as the technology needed to produce reactor-grade uranium – any nation that can do the one can accomplish the other. The other way to produce fissionable materials is by producing plutonium in a nuclear reactor – this technology is also decades old. Virtually any nation on Earth has access to the level of technology needed to produce fissionable materials should they choose to spend the money (and to risk the international opprobrium) to do so. And once the fissionable materials are in hand, assembling them into a nuclear weapon is not simple, but is certainly not overly difficult – again, this is something that the Manhattan Project accomplished nearly 70 years ago.
The bottom line is that developing nuclear weapons no longer requires a Manhattan Project (according to Peter Zimmerman, a former scientific advisor to the Senate’s Foreign Relations Committee). One of America’s most talented nuclear weapons designers, Ted Taylor, shared this viewpoint in a series of interviews with journalist John McPhee that were the basis for a fascinating book, The Curve of Binding Energy. In the Manhattan Project days neither the theory nor the technology were readily available – had our enemies launched an assassination program they could well have derailed the Manhattan Project, or could at least have delayed our first nuclear weapon by a matter of years. But that is hardly the case today – it is almost certainly impossible to assassinate away the Iranian program. Killing crucial people and tinkering with Iranian technology can delay things, but likely no more than that.
But there is a little more to the issue than this – what University of Edinburgh professors Donald MacKenzie and Graham Spinardi call “tacit knowledge.” Tacit knowledge is the knowledge handed down from master to apprentice – think of Yoda teaching Luke Skywalker how to become a Jedi. Tacit knowledge encompasses the secrets of how to turn theory into practice and as MacKenzie and Spinardi point out, there is a LOT of tacit knowledge involved in nuclear weapons design. The question is whether this tacit knowledge is essential to making a working nuclear weapon or a good nuclear weapon. And here, even MacKenzie suggest that making a crude nuclear bomb might not require a high level of tacit knowledge, provided the designers aren’t picky about details such as developing a miniaturized or high-yield device. Bringing this back to the question of assassinating Iranian nuclear scientists we have to ask ourselves if the Iranian nuclear program is at a level of sophistication at which tacit knowledge plays a major role – if there are not yet any Iranian nuclear Jedi masters then whoever is carrying out these attacks is killing apprentices, and they might well be virtually interchangeable. Without knowing the exact state of the Iranian program we might not be able to know whose loss might (or might not) have a significant impact.
So – with all this in mind – we have to ask ourselves if these assassinations are ethical and if they are effective.
We never questioned our development of nuclear weapons – we were the good guys after all in both World War II and the ensuing Cold War while the Nazis and the Soviets were the bad guys. We didn’t worry about the British and French nuclear weapons either – they were also the good guys. And when the Soviets and the Chinese developed these weapons we weren’t delighted but, as musician and humorist Tom Lehrer pointed out, “the balance of power’s maintained that way.” It wasn’t ideal, but it was something that we could live with because we felt we could at least trust these nations to act rationally. Not only that, but we had no way to throw a monkey wrench into those nations’ progress – at least not without starting a war that we were loath to undertake. The reason that Iran poses a more difficult problem is that we don’t trust them and we don’t know if we can count on the Iranian leaders to act as rationally as the Soviets and the Chinese.
As far as the ethics of assassination go there are a few considerations. One is whether or not such a program is likely to achieve its desired ends. If the desired end is to put a halt to Iranian nuclear weapons work, then assassinations will almost certainly not work. But if the intent is simply to delay such a program, then they might have some utility, but without knowing how much depth the Iranians have on their bench we cannot know how much of a delay – if any – there might be.
Whoever is ordering these assassinations also has to weigh what is at stake compared to the killing that is taking place. The answer here seems simple – a nuclear attack might kill hundreds of thousands of people so killing a handful to prevent such a high toll seems a reasonable trade. This argument – utilitarianism – is simply a matter of mathematics; what provides the greatest good for the greatest number. But is it legitimate to wonder if ethics can be boiled down to simple math. What if, for example we were told that a terrorist group was going to detonate a nuclear weapon that could kill a half-million people, but they would forgo their attack if we agreed to execute a hundred thousand citizens. Clearly the mathematics would argue in favor of the executions – but we have to ask ourselves if this is an ethical course to follow. Even with a life-saving ratio of 5:1 I suspect most of us would say that purposely executing 100,000 people is unethical – even if doing so would save a half million lives. So numbers alone do not make a decision ethical. But do the ethical books ever balance taking this approach? Is it OK to execute 1000 citizens to stave off a nuclear attack? What about 100? Or 10? What if those to be executed are the citizens of a neutral nation – would this make the executions more or less acceptable? What if they were prisoners? Or in the case we’re talking about – what if they are nuclear scientists helping to develop a weapon you are worried about? It’s a tough question – one for which there is likely no answer to which we would all agree. But the fact remains that Iran is not presently at war with any other nation – it might be ethical to assassinate key weapons scientists in an enemy nation, but does Iran’s relationship with any other nation fall to this level? Again, there is likely no universal answer to this question.
The other question of course is – ethics aside – the goal of such a program and whether or not a program of targeted assassination is likely to be effective. In this case if the goal is simply to delay Iran’s development of nuclear weapons, I would suggest that the delay is likely to be inconsequential. As noted earlier, the information and technology are all “out there” and they are simply too widely disseminated to stuff this particular genie back into the bottle. And if the goal is to halt Iranian nuclear weapons research, then the goal is even less likely to come to pass. Destroying machinary– the goal of the Stuxnet computer virus – or delaying technology development through economic sanctions and restricting trade in crucial items can really put a crimp in the process of developing nuclear technology, but it is unlikely that any crucial knowledge will be held by only a single person. At the same time, targeted assassinations might actually be counterproductive – at some point Iran is likely to realize that it has to produce more nuclear scientists, to protect them better, and to urge them to progress faster. Similarly, if the goal is to dissuade other nuclear scientists from making progress, this goal is almost certain to fail- in fact, the remaining scientists may well feel that their safety is best protected by the rapid development of a working nuclear weapon.
The bottom line is that the ethics of these assassinations is murky at best – balancing the possibility that Iran might develop nuclear weapons (and the use to which these weapons might be put) against the lives of the scientists and engineers who are involved in supporting this development process. And even putting aside ethics for a moment and asking only whether or not such a program is likely to achieve its ends we can only say “I don’t know” because the technology and basic science are already so well-known and so well-established. The fundamental question is whether the certainty of several small deaths in Iran balance the possibility of large-scale death elsewhere – this is a question that must be asked, even if there is no simple answer.
Dr Y is a certified health physicist, trained in nuclear power plant design and operations, with experience in nuclear power, environmental science, and planning for radiological and nuclear emergencies. He has 30 years of experience in the areas of nuclear and radiation safety.