Flicking through twitter over the weekend, I was struck by a peculiar contradiction between two articles that were being widely linked to. The first (which was really a cluster of articles) related to the revelations about the US government's surveillance activities, which, we now know, exploit the affordances of ubiquitous digital media, to record and analyse billions of informal communications between members of the public. As various wags have quipped, it turns out that Big Data really is very hot in Washington right now. The UK government has also been implicated, with the Foreign Secretary astonishingly blasé about the activities being alleged.
The second was an article in The Economist, complaining that the UK government has "a worrying disregard for data". Evaluations and censuses are being cancelled, statistics are being manipulated, evidence is being ignored. As a result, policies are being built on intuitions and prejudices, and New Labour's glorified 'evidence-based policy' agenda has fallen by the wayside. This is odd, given the recent flurry of excitement surrounding 'what works centres', in which the new 'gold standard' for policy evaluation, the RCT, has been promoted as the ultimate arbiter of rational policy-making.
Enthusiasm for RCTs, alongside the rise of Big Data and neuroscience, represents the emergence of a neo-positivism, in which the data is professed to 'speak for itself', unpolluted by theory or causal metaphysics. No doubt the National Security Agency's (NSA) enthusiasm for social media is that, unlike the subjects of Stasi and KGB surveillance, Facebook users are happy to speak for and surveil themselves, sharing all sorts of banal details of their lives, locations and activities voluntarily. The genius of this new surveillance state is that it, like the liberal market state before it, works with the grain of individual freedom, rather than against it.
But which is it to be? Do we want states to become more interested in data or less? The positivist answer is that we want governments to become more interested in some data, and less interested in other data. The tricky thing here is that positivists must also vehemently deny that there is any politics of data in the first place: repeat, data must be allowed to speak for itself. And yet only certain individuals, using certain techniques, are able to truly listen to what the data is trying to say to us. Positivists may end up making the somewhat underwhelming critique, that security agencies and marketing companies simply aren't good enough listeners. If only they had read their Karl Popper a bit more closely, and then watched The Lives of Others, they'd represent the perfect audience for Verizon customers.
What this reminds us, should any reminder be necessary, is of the contradictory nature of positivism generally. The primary problem with positivism isn't epistemological, but political: it seeks a monopoly on the legitimate means of truth, to match (and occasionally complement) the state's monopoly on the legitimate means of violence. It then abstains from any political engagement with the critics of this monopoly, for fear that this would be to abandon it. Hence, perfectly empirical and empiricist histories and sociologies of science (with or without post-structuralist baggage) must be simply ignored, a tacit recognition that they also carry some truth, which (like the IRA in the 1980s) must be starved of the oxygen of publicity. The same holds in the relationship between economics and the history of economics, not to mention economic history. A huge amount of political work goes into keeping politics at bay.
In Science in Action, Bruno Latour discusses the 'janus-faced' nature of modern science, which attributes its real-time 'discoveries' to individuals in certain times and places, but its past findings to some non-human thing called 'nature'. Something similar is going on with data right now. On the one hand, it is becoming hotly politicised, rarely more so than over the last few days in the context of the NSA revelations. Even aside from surveillance or governmental issues, it appears that social and economic life is becoming lived dynamically in terms of production, manipulation and interpretation of data, with (for example) live events being compulsively recorded, tweeted and connected to. Data is now part of lived experience, or rather lived experience is being 'datafied'.
On the other hand, data is being icily naturalised, with its institutional and methodological preconditions being marginalised from discussion. Arguably this is a rhetorical ploy on the part of those who simply want to collect more data. The idea of an entirely neutral methodology, which allows data to 'speak for itself', will be most appealing for those who don't wish to justify their scientific activities.
Every wave of positivism eventually comes undone, when its central contradiction emerges. The fact that the resolutely 'anti-political' behaviorists of the 1920s, for example, produced techniques of psychological control that could serve the interests of military and corporate power, is one demonstration of this. The NSA revelations arguably herald a similar moment of reckoning for digital data collectors and analysts. Google's childish political mantra of being anything other than 'evil' ignored the historical lesson, that apolitical, purposeless knowledge is often one of the greatest assets to the strategies of the powerful.
What would be the positivist (or neopositivist) response to any of this? For those who believe fMRI's reveal the 'truth' of who we are, that RCTs prove 'conclusively' 'what works', or that, with enough data, correlations start to 'speak for themselves' allowing us to abandon our causal theories, is the task now to build an even higher wall around what counts as 'science'? While Google, Facebook et al cosy up with the US security services, will Dawkins, Ridley et al still turn their backs on the increasingly obvious Nietzschean insight, that power and knowledge are entangled, rather than ontologically separate? And if they shrug, and quietly accept the role of money and violence in making knowledge possible, what will they have to say about it? What notion of democracy will they appeal to, once the boundary of the science lab is accepted as arbitrary and constructed? No doubt they fear that the choice is between positivism and a complete postmodern free-for-all, echoing the Hobbesian choice between the all-powerful Leviathan and the deathly state of nature. But if this is the case, they will find it increasingly difficult to criticise the actions of anyone who claims an exclusive right to observe and analyse, including security agencies.