Lone Wolves, Black Swans, and a Surveillance Llama

So, yesterday two murderers who murdered someone in plain view of video cameras, CCTV and a good number of shocked passers-by were convicted of having done just that. British law showed itself to not be a banana. But the commentary that immediately followed this most open-and-shut case focused upon (as always) whether the policing and security agencies had somehow missed a trick; just as they did around the Norwegian lone-wolf atrocity too.

Our zero tolerance for attacks is of course founded on our individual desire not to be caught up in such an event, and the way that we secure this is to make sure that it doesn’t happen to others either. All of that seems perfectly rational to me.

But in this seasonal time of gift giving (and the Snowden / NSA story is the gift that refuses to stop giving) are the problems of identifying, containing and rolling back lone wolves and surveillance connected? Or are lone wolves the stuff of Taleb’s black swans?

The immediate commentariat last night pointed out the following (and it will only be in time that we know if this is accurate or not, more precisely whether there is an evidential underpinning for these assertions):

Only one text message was exchanged between the two assailants, so not a developed pattern of signals to intercept.

That radicalisation occurred over a long period – some 10years or so – and thus if one was looking to a pattern from these two it was not towards the sudden cachophony of violence that then came forth. As one of my students quite wisely put it, perhaps the choice we make is about where we set the bar for more intrusive levels of interaction – speech acts might be enough, so long as we’re happy to resource it. Peter Neumann (of this KCL parish) made the very good point on Channel 4 news that there is no formula for this radicalisation. The 7/7 bombers also had 10year pre-event histories, whilst others might go live in a matter of months. The precise mechanism or transition from radicalisation to violence is simply not known, or not well enough known.

That they were known associates of known radical voices – so, again a targeted widening of the net, with all the resource implications that come with it.

That one of them had mental health problems, from adolescence, whilst the other had a long history of gang membership and low-level but reasonably intensive violence in their day to day lives. But it’s an easy and slippery narrative to say ‘ah, they were just mentally ill and maladjusted’ because the consequences of those thoughts and those actions are disproportionate. This would have a poor public policy outcome.

So, the radicalised lone wolf is a tricky character. If these two had avoided being arrested overseas they would have been loud and radical voices in a crowd. That would make them difficult to priortise – certainly from an outside perspective. But the dovetail into the NSA surveillance stories, is that one of the utilities of these techniques (which have also now run into legal problems via the 4th Amendment) is that they both highlight relationship details within degrees of separation, which would have been relevant here, and that they are capable of tracking worrying/dangerous narratives (eg narratives that lead to actual violence). So, that they did not work in this instance does not mean that they do not work. The transparency called for in the open letter from the internet giants (see here) should be around whether this kind of mass data approach is effective in picking up actively dangerous radicals (eg those capable of inflicting physical harm). Because that’s a risk assessment dialogue that the public and the relevant governments have not had.

Governments have assumed that the public is entirely intolerant of intelligence failure (which has become synomyised with attacks taking place), and that the technology was available to tackle some of these issues. Running alongside this is that the technology was ‘usefully’ outstripping the pace of the legal and political frameworks in place.

So tying it all up is the need, I think, to discuss the range of threats or problems being tackled. Is it all about physical violence or upheaval, or is it also around managing the message. The limitations of the dragnet – eg on lone wolves – is also important to acknowledge. Yesterday saw a final piece of the jigsaw put in place around the murder of a decent family man and soldier. It also opened a tension between the perception of the all-seeing eye and the analogue threat of lone-wolves with easy to acquire weaponry.

Yesterday was a rare day of no winners and no clear positives.


4 thoughts on “Lone Wolves, Black Swans, and a Surveillance Llama

  1. Quintin says:

    Great piece Rob,

    As an extension: though Taleb authored The Black Swan in 2007, this concept was borrowed from one of his favourite philosophers: Karl Popper. In his The Logic of Scientific Discovery, first published in 1935, Popper states under a heading of The Problem of Induction: ‘…no matter how many instances of white swans we may have observed, this does not justify the conclusion that all swans are white.’ This ‘problem of induction’ is of course, the most significant problem that any intelligence agency will face – and the very problem that such agencies attempt to mitigate by this ‘over-gathering’ approach that Snowden continues to expose.

    It seems odd that yet again, western societies want their toast buttered on both sides: on the one hand, we require omniscience of our intelligence agencies, on the sole condition that such omniscience should exclude any information about ourselves. Of such a futile desire, it strikes me as equally futile that intelligence agencies would even attempt such omniscience in the face of the ‘problem of induction’.

    Shrug, walk away and hope that Popper was wrong – unlikely as that may be.

  2. davidbfpo says:


    Is there ‘zero public intolerance of attacks’ ? I have my doubts, even after media reporting – often after government “leaks” – polling in the UK does not give terrorism a high ranking. The public are aware there is a risk,, possibly a constant one, but know that attacks are rare – even more so outside London and some English cities.

    State failure to deliver public goods, such as public security (very different from state security) is seen all too regularly in the UK, Western democracies and elsewhere. Yes the UK government has ‘assumed public intolerance of intelligence failure’, but is that real. Look at the repeated individual and institutional failure over child protection, with regular serious case reviews for at least two decades finding the same errors and the same lame excuses “we’ll do better next time” from those institiutions involved.

    The media reporting in this case, of which I have read a little, appears to be a court day only phenomena for nearly all UK media outlets. Even then we are assured there will be no full public explanation; incidentally in marked contrast to the post-Breivik public investigations in Norway. See: http://22julikommisjonen.no/en/Report and alas only in Norwegian the report by the PET (security police): http://www.pst.no/media/43446/evaluering22072011_PST.pdf

    “Lone wolves” do pose a challenge to public security and Breivik is a well known example of the price to pay when they are successful. They can be identified prior to an attack, as in the case of the Polish professor who plotted to blow up their parliament: http://nationalinterest.org/commentary/polands-agrobomber-european-terror-7762

    Note this polot was disrupted by human intelligence, his own wife and students reporting their concerns.

    (More another time).

  3. Mike Wheatley says:

    Actually, I had been expecting the pair to mount a defence of:
    “it wasn’t murder, it was warfare, legally equivalent to your drone strikes, just lower-tech”.

    Of course, that would have failed, since the UK isn’t doing drone strikes in Pakistan or Yemen, so we can find them guilty of murder without looking like hypocrites.

    But if/when it happens in the US, and the attackers say “it was a drone strike, treat us as enemy combatants,” then what happens?

  4. davidbfpo says:

    Part 2

    One of the weaknesses for UK CT, esspecially the Prevent option, is that radicalisation can be very fast. Steven Simon in his book when examining the Atocha (Madrid) train station attackers found they went from indifference to an attack in sixty days.

    Some of those involved, from the community, involved in Prevent work, including Channel Project, have found a marked reluctance amongst officialdom and media to consider the impact of mental illness as an issue.

    In the current learning environment created by the Snowden revelations, we are now far more aware of the emphasis given to the collection of ‘metadata’. In particular communicatiosn data which enables association to be charted. That does not mean the collectors and analysts know the content of the messages exchanged.

    It is a moot point if such communications data can alert the ‘watchers’ to a person or group moving beyond being ‘loud & radical’. It may identify patterns, it may identify those involved – including those who prefer to remain “off the radar”, the so-called “radical speakers”.

    Far better as a guide to an individual changing is when those around him / her notice their new stance on their faith, politics, violence and more. Now whether they wish to argue, intervene and more is very unclear. There is certainly anecdote a plenty that some of those recently convicted outside London aroused concern long before their arrest, so that friends and others put distance between them and the person of concern. Oddly there is an expectation in some communities that officialdom will notice this in time and be able to respond.

Leave a Reply

Your email address will not be published. Required fields are marked *