There has been a lot of speculation following the Woolwich attack about how people are 'radicalised', that MI5 'knew' the suspects and therefore could have - should have - stopped this terrible event. It might yet emerge that our intelligence agencies made an enormous operational or strategic blunder, or that crucial information was unavailable to them (hence the renewed calls for the 'Snooper's Charter'). But the truth is, I don't know the details of this case; none of us armchair critics do. But there does seem to be little or no understanding of how MI5 actually works, and quite how difficult it is for them to stop people intent on terror.
In my research work with the think-tank Demos, I spent two years looking into the crunch question that in counter-terrorism: how do you discern between the (fairly many) individuals that hold illiberal, extreme ideas that we permit in a liberal democracy and the (very few) individuals that hold the same ideas, but are also willing to act on them violently. After a lot of painstaking research, I found there is no single answer, no single pathway of radicalisation, no obvious predictive flashing signs that the police or intelligence agencies can reliably and consistently look out for. Such randomness is uncomfortable, but it's true nonetheless.
But because MI5 has limited time and resources, and because it is both counter-productive and illiberal to monitor everyone that holds some beliefs also held by the men of violence, this means that a difficult decision must be taken about whom to focus on. Even with the best evidence in the world, this is inherently probabalistic, and so carries some degree of risk (unless you want to live in a society where everyone with a radical view is under constant surveillance).
But even once a decision is taken, it is not plain sailing. According to some reports, the security services first came across Adebolajo via his arrest in Kenya, or, before that, after being arrested during an extremist demonstration outside the Old Bailey). I imagine MI5 'knows' quite a lot of people. But to undertake sustained and careful surveillance (such as phone tapping or placing an undercover agent which might have offered a clue as to his intentions) MI5 would have required an authoriation under the Regulation of Investigatory Powers Act (RIPA) 2000.
To do that, a senior official - even the home secretary if it involved intercept - needs to be convinced that there is sufficient evidence to suggest that the individual is a genuine security threat and that any proposed surveillance is proportionate.
If there isn't enough evidence, the detailed and careful surveillance of the sort that might have worked here would not be permited. Forget James Bond, it is expensive, time consuming, painstaking and detailed work - with a lot of paperwork to try to ensure any measure is proportinate and necessary: the very halmarks of how intelligence should work in a liberal democracy.
To make things just that little bit more difficult, as I argued in the Times yesterday, intelligence agencies have been a victim of their own success. They have been remarkably good at disrupting major cells and arresting targets, which means those interested in violent activity are often smaller, more amateur, self-starting groups, unable to launch spectacular attacks that require technical know-how and precision planning. This makes it even more difficult to spot, and even more difficult to get the authorisation required to launch sustained surveillance operations.
Whatever is finally revealed about the details of this case, it is at the very least worth bearing in mind the incredibly difficult circumstances under which they work.
Follow Jamie Bartlett on Twitter: www.twitter.com/JamieJBartlett