The Contested Value of Social Science in a Digital Culture

Wednesday 28th October, 2015

Ziyad Marar

Special Section on the Future of Research Governance

Social science seems constantly bedevilled by a mixed press. On the one hand we see its value proclaimed loud and clear such as in the this year’s Campaign for Social Science punchy report ‘The Business of People’ or by figures in the natural and physical sciences who realise that social science is critically important to making progress on a host of problems from global warming to anti-microbial resistance. And it is often said that the work of government is constantly enmeshed in social science themes covering key priorities such as poverty and inequality, housing, crime and policing, education and health, well-being and ageing populations and many more.

On the other hand we see attacks on and mis-understandings around the value of social science proliferate, especially in the US. In April the House passed a reauthorization of the National Science Foundation — the America Competes Act (H.R. 1806) — that cuts funding to the social sciences by 45 percent. This bill is waiting to go to the Senate. And during the summer a 27-3 vote by the Senate Appropriations committee pushed a bill — the Commerce, Justice, Science and Related Agencies Appropriations bill (H.R. 2578) – to the senate as a whole advocating around more cuts to social science funding. And when Education secretary Nicky Morgan advised young people to steer away from careers in the arts and humanities in favour of science and engineering she was in another way signalling the gathering impatience we have for enterprises that fail to provide the right kind of utility.

Why do these attacks have political traction and, implicitly, public support? Social science is a broad, multi disciplinary effort to understand human behaviour and social structures and contains fields that range across quantitative and qualitative domains. The resulting stock of knowledge from this work can be diffuse and long term in its impact. This is a feature of the messy but crucially important problems that are the domain of a typical social scientist. It is harder to identify the complex causes of crime compared with understanding how chemicals interact. People and social structures are too unruly for simple descriptions.

The mixed picture for the value of social science is likely to continue as long as people prefer quick and simple explanations, because it is in the nature of complex problems tackled that a headline grabbing picture seldom emerges from the research. And this has consequences for the way social science is assessed by policy makers and the wider public sphere.

Yet our digital culture may exacerbate this problem by tilting us even further toward speed, simplicity and utility. In what is often called the ‘attention economy’ (where information overload makes attention the scarce resource) there is a clear need to filter our consumption efficiently, therefore inviting the twin enticements of automation and aggregation. The hope is that we can navigate the suffusion of information by using big data, algorithms and metrics of many types and thereby cope with the tidal wave of knowledge without drowning.

But this trend favours certain types of information, the more standardised the better, and relies on a metaphor of knowledge as a process of discovery, yielding up simple enough nuggets that will be grist to the mill of aggregation. This is all very well for simpler, tamer problems that are commonly dealt with by physical and natural science. When it comes to messier and more ‘wicked’ problems, which are to a larger extent the focus of social science and the humanities, interpretation is often a more apt mode of enquiry. And the results are usually more ambivalent and need careful digestion and patience and need to be assessed with a multi-dimensional framing.

This summer’s release of the 2014 journal impact factors came out to the usual wave of ambivalence. No surprise really, given their controversial status; as the phrase goes ‘not everything that counts can be counted, and not everything that can be counted counts’. Yet this truism is routinely ignored when time is short, and all the more in a digital culture that is built for speed.

This summer saw the publication of The Metric Tide a report resulting from an independent review chaired by Professor James Wilsdon, in which metrics such as impact factors and h-indices were assessed in comparison to the system of expert peer review that was used in the last Research Excellence Framework. The report recommends a more responsible approach to these indicators and comments that metrics are no substitute for expert judgement. Interestingly not only was the correlation surprisingly poor but it was particularly poor for fields in the humanities and social sciences. With lower paradigm fields we see a wider gap between what can be counted and what counts in the minds of experts in the field.

The explosion of technological innovation, data sources, open movements and new modes of collaboration has changed the scholarly landscape in many ways, intended and unintended. In many many cases new research modes are thriving (especially those with a quantitative bent). The question is whether fields with a more interpretive mode are losing out in prestige and funding as a result of this shift. This technology led approach to research evaluation and impact has potentially created new norms which are being read across too uncritically into fields that require interpretation and nuance, possibly rewarding narrower and more tractable research agendas.

If the digital republic of letters has squeezed the space for more imaginative, unbounded modes of inquiry, we need a way to re-balance the argument. This summer there was a discussion between the British Academy and the New York based Social Science Research Council as well as publishers, librarians and academics to address exactly this question. These and other organizations such as learned societies in HSS should take a lead in analysing the problem and identifying some remedies. If we don’t then parts of social science (and the humanities) will continue to be under-valued, with the consequence that important, if messy, problems could be neglected in favour of the more tractable. We should find a way to heed Daniel Kahneman’s insight, "when faced with a difficult question, we often answer an easier one instead, usually without noticing the substitution.”

Here's more you may be interested in: