978-1-5386-2366-4/17/$31.00 ©2017 European Union Transfer or Translation? The Actor-Network Theory Approach to the Social Impact of Science Seweryn Rudnicki AGH University of Science and Technology Faculty of Humanities Krakow, Poland sew.rudnicki@gmail.com Abstract—The question of the social impact of science has become a crucial issue in the current debates on the role of science in society and, importantly, one of the criteria upon which academic institutions and scholars are assessed. However, in both the public debate and evaluation procedures there is a limited understanding of the nature of the social impact of science. The aim of this paper it to deconstruct the predominant views as based upon the simplistic model of knowledge transfer, show its theoretical shortcomings, and contrast it with the proposed model of translation, derived from actor-network theory. Keywords—social impact of science, knowledge transfer, translation, actor-network theory I. INTRODUCTION Arguably, all scientific discipilnes have been experiencing the growing pressure from governments, industries and general public to increase their social impact. Universities and scholars have become routinely assessed not only on the basis of their scientific merits, but also on the basis of their social impact understood as success in dissemination and commercialization of knowledge. Concepts like ‘Mode 2’ production of knowledge and triple helix model have become popular as notions reflecting the emergence of new science systems in their social contexts [1, 2]. It can be argued, however, that the evaluation of the social impact of science (SIS) is often done according to rather conventional, taken-for-granted understanding of the process, and thus fails to reflect the actual complexity and multiplicity of SIS. Given the importance of SIS for societies, economies, and academia itself, there is an urgent need to propose new analytical frameworks for its understanding, capable to grasp the complexity of the field and still useful for policy makers. This theoretical article attempts to offer such a conceptual alternative to the predominant and over-simplified model of knowledge transfer by contrasting it with the proposed model of translation derived from actor- network theory (ANT). Additionally, this paper shows that some of the concepts from ANT – that has been developed as a descriptive and purely scientific approach – may also have high practical value as informing policy-making. II. THE MODEL OF KNOWLEDGE TRANSFER It is argued here that the predominant understanding of SIS is now based upon a set of assumptions that may be summarized as a model of knowledge transfer (MKT), prevalent both in the public discourse and many science polices. MKT may be also identified as a cognitive foundation of some traditional concepts of SIS: ‘the-trickle-down’ view and the linear innovation model [3, 4]. Importantly, MKT – as many other influential concepts [5] – is based not upon a scientific understanding but upon a cognitive metaphor, specifically the metaphor of transfer. It assumes that: • Knowledge is a stable entity that can be separated from the environment in which it is produced (i.e. academia) and passed on to other domains relatively intact. • Knowledge transfer is essentially linear – from knowledge production (basic research, applied research) to its application in the field of practice (innovation, implementation and diffusion processes). • The production of scientific knowledge is generally more difficult and requires more effort than its application. In some cases scientific results will be just taken up by the practitioners without an extra effort. If not, the transfer may be deliberately facilitated and the gap between science and practice bridged. • Scientific and codified knowledge is superior to other forms of knowledge (e.g. everyday knowledge, context knowledge, practical knowledge). • Transfer of knowledge is predominantly dependent on the characteristics of knowledge (its newness, reliability, usefulness etc.), not on its social and institutional context. • All scientific disciplines are capable (or should become capable) of delivering transferable knowledge. • The difficulties in SIS stem from insufficient motivation or competences of scientists and scientific institutions,