Experimenting beyond science

Joseph Wright's 'An Experiment on a Bird in the Air pump' 1768

I gave a talk to the other STS fellows here at Harvard last Tuesday on the topic of experimentation. My presentation covered emerging approaches in STS and related disciplines which use the metaphor of experimentation to study social processes, rather than the term's more common association with the scientific method. This blog post will cover some of the basic points from the review work I carried out - what are the different disciplinary approaches to such a study? why might the metaphor of experimentation be a useful one? and where are the conflicts, disagreements and problems? Over the next few weeks I'm planning to explore the potential for linking approaches to experimentation with my work on public engagement practices in the UK, so watch this space for further developments with this project.

I originally became interested in the idea of 'collective experimentation' when I read the European Commission report 'Taking the European Knowledge Society Seriously' from 2007, which was written by a group of prominent STS scholars. The authors of the report were charged with coming up with solutions for what the European Commission labelled 'public unease' with certain developments in science and technology.  In a challenge to this framing, the report described the currently dominant European regime of the organisation of innovation and technological development was one of the 'economics of techno-scientific promise'. Within this regime state actors are generally enthusiastic about the promise of innovation and tend to assume that the 'new' will always be better than the 'old'; however, the same actors have increasingly realised the need to consider broader societal challenges and public reactions in innovation processes. The report contrasts this regime to the newly emerging regime of the 'economics and socio-politics of collective experimentation'. This regime is best illustrated by approaches where innovation is much more broadly distributed across different kinds of actors, such as technology users or particular communities like patient associations, open-access software developers, or peasant farmer plant-breeding networks. The report refers to such processes as 'experimentation' as they can all be characterised as responses to particular situations which have emerged allowing societies to try out new things and learn from them.

Related approaches have variously considered: the experimentalisation of society through the development of new technologies such a biotechnology whose risks cannot adequately be tested in controlled laboratory conditions alone; the 'real-world' experiments which occur as the result of unwitting human interventions in the world due to the inherent risks of technologies like nuclear power; how processes around ecological design or restoration can be understood as 'real-world' experiments due to the close interactions between social and natural processes that they necessitate, and therefore the integration of wider social contexts and actors which is required within this particular scientific practice; the 'collective experiment' that is climate change due to the complex relationships it engenders between processes of knowledge-making and interventions int the world, highlighting the provisional nature of scientific knowledge; or finally how practices of public engagement might be considered as 'institutional experiments' which are reconfiguring relationships between science, society and politics, through the constant need to perform and achieve certain democratic values. 

The cover of Thomas Hobbes' 'Leviathan' from 1651

It is also interesting to look at how the idea of experimentation has been approached differently in different disciplines (though of course there has also been a lot of dialogue and cross-pollination between disciplines). The American Pragmatist philosophers like John Dewey were concerned with challenging the commonly drawn distinction between practice and theory; thus Dewey's claim that politics is experimentation suggested a messy and iterative relationship between political interventions and the knowledge claims they were supposedly acting upon, and between experience and learning. In 1983 the philosopher of science Ian Hacking accused fellow philosophers and historians of science of failing to take experimental scientific practice seriously at all, instead producing accounts of the historical development and logical derivations of prominent scientific theories. In contrast Hacking suggested that far from being wholly determined by dominant theoretical commitments, experiments and their outcomes were also influenced by material conditions and apparatus, and broader social contexts. Since then historians of science like Peter Dear and Steven Shapin have focused on the relationship between societal developments in Europe and the development of specific forms of the experimental scientific method in early modern and modern times, to situate the assumed credibility and objectivity of the experimental method in its social context. In a more universalist enterprise, Peter Galison has considered how experiments end, highlighting that it is always a social judgement to assert that enough convincing evidence has been accrued to end an experiment.

Geographers, sociologists and science and technology studies (STS) scholars have drawn on such insights to relate the idea of experimentation to contemporary events at the interface of science and society. Geographers like Gail Davies have been concerned with how forms of experimental life are increasingly overflowing their boundaries in time and space, with consequences for forms of social ordering and social life. Sociologists like Matthias Gross have drawn upon a long history of approaches to experimentation in the discipline (linked to the Chicago School in the early twentieth century) which in contrast to popular understandings of scientific experimentation, stress the uncontrolled nature and unboundedness of experiments. Thus Gross seeks to emphasise how experimentation involves the looping back of interventions in the world into new knowledge claims which in turn call for new interventions. Yet there is an equally strong vain of work in sociology (linking back to the Columbia School) which seeks to replicate the controlled and replicable nature of scientific experiments, relying on quantitative analyses and randomised control trials in relation to topics like education and healthcare. STS scholars have used the notion of experimentation to further their project of understanding the interactions between social and technical systems, and their subsequent organisation and reorganisation, resulting in both descriptive and normative accounts. 

The Mp3 Experiment, from Improv Everywhere

There are also some broader themes and fault lines to be found amidst these different works on experimentation. Firstly, authors differ on whether they consider social scientists to be the initiators of societal experiments (as is the understanding amongst quantitative sociologists) or whether they consider the role of the social scientist as one of making visible certain practices and processes of experimentation (as the historians of science and most geographers and sociologists do). Or the debates about the role of the social scientist could be stretched even further by the normative calls of the European Commission report for a more experimental approach to organising processes of innovation and technological development as iterative and partial processes of learning. Approaches to experimentation also differ between those which explicitly seek to draw on the experimental method, or those which seek to redefine notions of experimentation. Those drawing on the scientific experimental method might do so because of its supposed rigour (like John Dewey or the quantitative sociologists) or they may do so from the point of view of many historians of science which would highlight the messiness and impossibility of total control within the scientific method. On the other hand, the sociologist Mathias Gross has suggested a radically new definition of experimentation as being uncertain, uncontrolled and unbounded, whilst the geographer Harriet Bulkeley has suggested that experimentation refers to novelty, experience and trying out new things. A final interesting fault line to consider is that between approaches which are primarily concerned with tracing the societal effects of emerging science and technology, or those which see blurring boundaries and processes of co-production between scientific and social experiments. 

I'm still working through how this little review will link into my own work on public engagement around science and technology in the UK, but I hope that some of the relevance of experimentation to understandings of learning and reflexivity is starting to become apparent. More to come...