Abstract
Despite the increasing push towards interdisciplinarity across the physical and social sciences, little is known about the realities of working across such diverse disciplinary boundaries. This article provides empirical insight into the challenges of collaboration from the perspective of a sociologist working on an interdisciplinary project focused on developing a medical device. Findings suggest the effective contribution of sociological research is affected by the framing of interdisciplinary projects. From the beginning, the project pursued a narrow framing focused on scientific development, pushing the sociological research outside the relevance of the project. Reframing is negotiated in shared spaces between disciplines, and fieldwork became important in reframing the project to include the sociological research. However, without commitment to addressing a societal problem, it was impossible for sociology to contribute effectively. Sociologists embarking on similar endeavours should ensure there is shared commitment towards a social issue to prevent the marginalization of sociological research.
Introduction
Interdisciplinary collaboration in academic research is increasingly well resourced, as research funders recognize that complex societal challenges cannot be addressed adequately through traditional disciplinary approaches. Collaboration is a key component of the Research Councils UK’s funding strategy, which suggests: ‘novel, interdisciplinary approaches are needed to solve many, if not all, of the big research challenges over the next 10 to 20 years’ (RCUK, 2014). In this way, interdisciplinarity is positioned by many funders as an essential part of the journey, or in Latour’s (1987) terms, an ‘obligatory passage point’, in the quest to address global challenges. Yet the apparently unequivocal benefits of interdisciplinarity have not been established, particularly for sociology.
Recently in this journal, Geoff Cooper (2013) called for a critical examination of the relation between interdisciplinarity and contemporary modes of academic governance, and its significance for sociology as a discipline. Cooper illustrates that there is no single understanding of interdisciplinarity but rather that many different constructions, with both positive and negative connotations, are employed in different contexts. The multiplicity of the concept adds ambiguity to calls for interdisciplinarity, making it unclear which constructions will be drawn by different actors and at different times. Cooper suggests that understanding the implications of interdisciplinarity is particularly important for sociology because of the unique characteristics of the discipline, which put it in an awkward position in relation to modes of academic governance.
This vulnerability of sociology was the topic of a debate initiated by John Holmwood (2010), who suggests sociology is an ‘exporter’ subject in relation to interdisciplinarity, whereby its rich theoretical and methodological resources are drawn upon by other disciplines. Holmwood argues that the evaluation of research output is problematic for exporter subjects – with their resources on loan, their academic output is at risk of being subsumed by other disciplines. Holmwood suggests that sociology is particularly vulnerable to this risk because of its ambiguity as a discipline. Sociology is a broad and varied discipline, with a lack of internal coherence that enables other disciplines to redefine sociological subjects as their own. This, Holmwood argues, has resulted in the decline of sociology’s contribution to research audits (Holmwood, 2010; Holmwood and McKay, 2015). Indeed, the falling number of submissions to the sociology panel in Research Assessment Framework (REF) 2014 suggests it is unlikely to run again as a separate disciplinary panel. While not arguing against the value of interdisciplinarity per se, Holmwood expresses concern that this vulnerable position weakens sociology’s academic status, which he refers to as sociology’s ‘misfortunes’.
Although still concerned with the implications of interdisciplinarity for sociology, others question the bleakness of Holmwood’s stance. Rosenfeld (2010) stresses the importance to criminology, one of sociology’s main export areas, of maintaining its sociological roots, from which it gains most of its basic concepts. Of particular value to criminology is sociology’s emphasis on the significance of context in understanding social activity, which, Rosenfeld argues, sociology is unique in analysing without prioritizing other factors. Rather than contributing to the decline of sociology, Rosenfeld insists that importer subjects rely on its continued success. Savage (2010) goes further to question the value of making a distinction between importer and exporter disciplines at all. He argues that sociology has always been open to ‘imports’, welcoming academics that have been trained in other disciplines. Rather than an importer/exporter divide, Savage suggests we adopt a version of Andrew Abbot’s argument that ‘disciplines are in constant search for intellectual jurisdiction, which involves complex processes of competition with rivals, and seeking particular kinds of niche expertise’ (Savage, 2010: 662). For Savage, the ‘misfortunes’ of sociology can be better understood through the tension between institutional pressure to demonstrate relevance to practice, and the continued problem of maintaining a position as a specialist discipline.
It is clear that interdisciplinarity holds uncertainty for sociology, but although it brings risk might it also present opportunity? As funders and assessment panels alike emphasize ‘impact’ there is increased pressure on the physical sciences to consider the social impact of basic science and technological innovation. Assessing impact requires detailed understanding of the field of use, and thus may create new and exciting opportunities for sociology, for, as Rosenfeld (2010) points out, no discipline has greater resources to understand the importance of context than sociology. So perhaps the increasing focus on impact in research will improve the ‘fortunes’ of sociology, if it can position itself as an essential partner for a widening range of disciplines. This cannot be assumed, however, since achieving successful collaboration across diverse disciplines invokes a new set of concerns and challenges for sociology that warrant exploration.
Collaboration across disciplines necessitates negotiation of hierarchies and power relations as, in Abbot’s and Savage’s terms, disciplines compete for intellectual jurisdiction and the legitimation of particular kinds of expertise (Savage, 2010). As Holmwood (2010) points out, maintaining its position within the social sciences is challenging enough for sociology, but what will happen when it competes in the realm of the physical sciences, where sociology is often positioned at the bottom of the hierarchy (Bourdieu, 1975)? As Rosenfeld (2010) rightly argued, many importer subjects within the social sciences have their roots in sociology, and therefore share many similarities. This is not the case with physical sciences, where epistemological and ontological differences to social sciences may be stark. Yet, if academic research is to achieve societal impact and help address societal challenges of the 21st century, this kind of interdisciplinarity is imperative.
Despite the call for increased collaboration between social and physical sciences, little empirical evidence of the possibilities and realities of this mode of working exists. Reporting on a new study tracing the experiences of a sociologist working on an interdisciplinary project, this article explores the practical challenges presented to the sociologist when divergent disciplines collaborate. The following discussion explores the epistemological nature of the differences between disciplines, moving on to offer an autobiographical account of my experience working on an interdisciplinary project alongside engineers, chemists and medics. Drawing on Actor-Network Theory (ANT), which highlights the role of non-human actors, the article explores the factors affecting the evolving position and value of the sociological component of the project. In conclusion, I reflect on the implications of my experience for both sociology as a discipline and individual sociologists entering into interdisciplinary collaborations beyond the social sciences.
Understanding Interdisciplinary Collaborations
Interdisciplinary collaboration is not a simple matter of knowledge sharing between disciplines. The claims of different disciplines are weighted according to what kind of knowledge counts, and collaborations do not occur on a level footing. Even disciplines that share many epistemological and ontological perspectives – such as engineering, chemistry and biology – employ different approaches to knowledge production. Indeed, disciplines across the physical and social sciences are united in accommodating a variety of epistemological and ontological positions. In this way, Knorr Cetina (2007) suggests that the conceptualization of disciplines as distinct and identifiable categories is superficial, and instead talks of epistemic cultures:
The notion of epistemic culture is designed to capture interiorized processes of knowledge creation. It refers to those sets of practices, arrangements and mechanisms bound together by necessity, affinity and historical coincidence which, in a given area of professional expertise, make up how we know what we know. (Knorr Cetina, 2007: 363)
With differences in knowledge production, science is hierarchically stratified, with sociology often positioned by others at the bottom linked specifically to its treatment of subjectivity in knowledge production (Bourdieu, 1975). Whereas physical sciences often take the perspective that carefully constructed and controlled environments can uncover an objective truth unmarred by human intervention, many areas of sociological enquiry reject objectivity, instead positioning the researcher at the centre of knowledge production. The question this raises is what happens when such contrasting epistemic cultures meet in an interdisciplinary project?
With few empirical studies of such collaborations, presently there is limited evidence to provide an answer. What research exists is informed largely by philosophical debate rather than empirical investigation (e.g. Pohl and Hadorn, 2008), and the few empirical studies are limited by their methodological approach, focusing on retrospective interviews with researchers with previous experience of interdisciplinarity (Broto et al., 2009; Oughton and Bracken, 2009) and the ethnographic study of collaboration between disciplines within physical science (Duncker, 2001). Consequently, these studies serve to highlight the difficulties of working across epistemological divides without offering empirically grounded insight into how we might move forward.
If collaboration with diverse disciplines is to be fortuitous for sociology, we need to understand what the challenges and opportunities of collaboration may look like empirically, and the implications of this for individual researchers and the discipline, which requires the longitudinal study of such collaboration as it unfolds. This article provides such evidence, presenting an autobiographical account of my experience as a sociologist working on a diverse interdisciplinary project, focused on developing a healthcare technology.
My experience is particularly pertinent, as epistemological differences between the social and physical sciences are most apparent in their different understandings of technologies. Traditionally, physical science adopts a determinist perspective in which, as Bijker (1993: 129) describes:
technology is conceived as a separate entity that follows a linear path. Technology is like a train, with a track that is fixed, although not known in detail. One cannot hope to change the train’s direction, only to check its speed and the safety of the crossing.
From this perspective, social implementation is an outcome of developing successful technologies and, with the understanding that this can be achieved in isolation from user settings, the technological determinist perspective sees a linear path from invention, to innovation, to diffusion (Godin, 2006).
In contrast, sociology, and in particular the field of Science and Technology Studies (STS) within which my research is situated, takes a socio-technical approach, viewing the construction of technologies as the outcome of a range of inseparable social and technological influences. Technologies are shaped just as much by their use outside the laboratory as by scientific development within it, or as Heath et al. (2003: 77) put it: ‘technologies … gain sense and significance within everyday activities and ordinary experience’.
The contrast between these two perspectives is evident and provided the backdrop for a difficult collaboration as I sought to have my expertise about the field of use legitimized by the other project members, when often it sat in contrast to their understandings of what makes technologies ‘work’. Yet my challenge was not to reconcile these different perspectives – which would be a futile task – but to find a way to work within them. As Turner (2001) explains, rather than it being a fixed concept, people are persuaded of particular expertise through mutable and shifting conventions. To accept particular knowledge claims in one instance does not mean the acceptance of the ‘immaculate conception of expertise’, implying that what is sanctioned as expertise at any time is contextual. With this in mind, my objective was to uncover the circumstances within which my expertise would be valued. To understand the influences on the position and perceived value of sociological knowledge within the project, I draw on ANT.
ANT is the most widely accepted STS approach (Law, 2008), and proposes that outcomes are produced through the relations between networks of heterogeneous actors. Focusing on the relational ontology of actors, ANT illustrates that rather than owning a single identity, actors become what they are because of their position within the network; actors change as networks evolve, and acquire new identities in other networks. In describing their actions and influence, ANT makes no distinction between human and non-human actors; the same agency is ascribed to all. Indeed, Latour (2005) talks of actants in recognition that agency is not located within individual actors, and may transcend superficial boundaries of human or non-human entities.
In the context of this research, the unique insight into the complex relations between networks of actants ANT enables is valuable in exploring interdisciplinary relations within the project and the multifaceted influences on the position of the sociological research. To be clear, this research does not seek to make a theoretical contribution to the ANT literature, but draws on the themes promoted by the theory to gain a nuanced understanding of interdisciplinarity. The analysis that follows offers insight into the realities for sociologists of working beyond the social sciences, with implications for improving future collaborations.
Methodology
This article reports on a project commissioned by a UK research council that aimed to take basic science research through to technological application. A team of engineers led the application, recognizing the opportunity it presented to further develop their work. The funding call stipulated the use of an interdisciplinary approach to develop a device to target a specific healthcare need, and the engineers drew together a team spanning chemistry, biochemistry, medicine and sociology, to develop a point of care diagnostic test (POCT) that would detect infections in renal patients.
The funding body stipulated the inclusion of a sociological component to ensure that the research would achieve social impact, and this was constructed as a PhD project with the objective of ‘exploring paths into use’. This meant exploring the potential use of the device within specific settings and moving this information upstream to inform technical development, thereby guiding the construction of a device that was appropriate to and workable within user settings. This demanded a prospective research approach capable of assessing the workability of developing technologies within organizational settings, details of which are reported elsewhere. Utilizing the results of such research in the design of the POCT device required effective collaboration, and therefore an important part of my research was to study the dynamics of interdisciplinarity within the project.
Methods
I explored the dynamics of interdisciplinarity through participant observation of team meetings and semi-structured interviews with individual team members throughout the duration, as detailed in Table 1.
Interviews with project team.
Within the project I held the complex position of being both a team member contributing to the development of a new technology, and a researcher studying interdisciplinarity. The duality of my role meant I could only explore the dynamics of interdisciplinarity through my experiences as a team member, and therefore this component of my research utilized an autobiographical approach, focusing on the construction of social realities through the personal experiences of the researcher (Merton, 1988). Autobiographical research is often criticized on the grounds that it emphasizes the importance of one individual over others, and can only ever offer localized accounts of reality (Bryman, 2008). Conversely, autobiographical approaches are prominent within feminist research precisely because they do not purport to offer objective accounts. Feminist scholars have long argued that all research and knowledge is situated, and call for greater recognition of this subjectivity in all areas of research (Cosslett et al., 2000).
Furthermore, Stanley (1993) insists that autobiographical accounts need not be individualized simply because they are constructed through the experiences of an individual; through the individual it is still possible to see social structures and networks. In this case, my position did not preclude me from developing a comprehensive understanding of the workings of the team, but instead added authenticity by situating it in my own experiences and affording me access to knowledge that would not be accessible to an external researcher. For Merton (1988: 18), autobiographical researchers ‘are the ultimate participants in a dual-participant-observer role, having privileged access – in some cases, monopolistic access – to their own inner experience’. In this way, my position as a researcher and member of the team provides a depth and richness to the data that would otherwise be impossible. However, the duality of my position also presented some ethical challenges, particularly in obtaining informed consent and retaining the confidentiality of the team members.
Ethical Issues
All team members gave written consent for the research; however, given the ambiguity of my role, I questioned how informed their consent was. I was not just a researcher but also a colleague, and afforded the trust and confidence associated with this. As my interactions with the team were framed by my position as a colleague rather than a researcher, my research agenda was not as prominent as it would be in a traditional researcher–participant encounter. Therefore, while my position afforded me access to rich data, it simultaneously made informed consent more difficult to establish. To address this, I was transparent about the aims of my research, regularly reminding team members that I was studying the nature of interdisciplinary working.
The importance of gaining informed consent was exacerbated by difficulties in maintaining confidentiality. To conceal the identity of the team in the wider academic community, I have assigned pseudonyms, anonymized job titles and changed scientific details (consequently descriptions of materials and methods may not be scientifically accurate). However, it is difficult to ensure anonymity within the team, as individuals will be recognizable to each other through the idiosyncrasies visible in direct quotes. Consequently, I sought explicit consent from each individual to use direct quotes.
Conceptual Framework
ANT provided a methodological and theoretical tool through which to explore how the actants in the network affected the nature of the collaboration, and subsequently the development of the POCT device. In what follows, I combine this with MacMynowski’s (2007) conceptual framework for understanding multidisciplinary collaborations, which I use to interpret the evolving relations between the disciplines. MacMynowski (2007) proposes four modes of interdisciplinary working: conflict; tolerant ambivalence; cooperation; and transformation. Conflict refers to the clashing of irreconcilable differences, and the hostile maintenance of boundaries. In tolerant ambivalence researchers co-exist; boundaries are maintained, but tolerance is exercised. In a cooperative mode, analytical tools and theories are shared with a similarity of purpose. Transformation marks the reorientation and recombination of knowledge, whereby researchers work together to address a problem that transcends academic boundaries.
If sociology is to be effective in facilitating the societal impact of academic research, such interdisciplinary projects must adopt a transformative mode of working, offering an environment that allows all disciplines to contribute equally. MacMynowski (2007) suggests that different approaches may exist on many levels, and proposes that collaborations could move towards transformation by engaging in a process of differentiation, clarification and synthesis, whereby the differences in disciplinary approaches are delineated and a new interdisciplinary approach is constructed. Yet this is a theoretical proposition, rather than an empirical observation, which presupposes some form of effective collaborative working that would enable interdisciplinary teams to collectively engage in these reflexive processes in the first instance. Given the disparity between the social and physical sciences, it seems unlikely that this practice would be successfully established, at least at the beginning of a project. The following analysis uses empirical evidence to explore how interdisciplinary collaboration is negotiated in practice, which suggests it is significantly affected by the framing of the project.
Framing Collaboration
The framing of research is the process of drawing boundaries around what the project is, and is not, concerned with. Knorr Cetina (1999) talks of framing as the design strategy for knowledge, suggesting that it continues throughout the course of a project as different actors come in and out of play. Knorr Cetina (1999) emphasizes the importance of the framing at the beginning of a project, which creates the script for the subsequent unfolding drama throughout the research, although recognizes that framing is continually negotiated as the research progresses.
In what follows, I demonstrate that, in contrast to the rationale in the original proposal, from the beginning the POCT project was framed around the nanoconductor. This framing privileged the position of the physical scientists and positioned the sociological research outside of the ‘real work’ of the project, thereby undermining a transformative mode of collaboration. However, as the project progressed, the framing was continually renegotiated with significant impact on the positions of the disciplines and the nature of the collaboration. Using the changing positions of sociology and chemistry as an illustration, I argue that it is in the establishment of shared spaces between disciplines – a common ground that eludes disciplinary dominance – that the framing of an interdisciplinary project is negotiated. Therefore, the ability of sociology to play a meaningful role in interdisciplinary endeavours rests upon its success in bridging the void that separates it from physical sciences.
Privileging Disciplines from the Beginning
Although the original funding proposal included the idea that the POCT would be used to diagnose renal infections, this did not remain a fixed goal. Research funding was delivered in two stages, and qualification for the second phase required partnerships with commercial investors. The physical scientists believed that a device with multiple potential applications, rather than one defined use, would be an attractive commercial investment and therefore, for them, maintaining a flexible end-use became an important strategy to enrol investors into the network. Consequently, the potential application for the POCT device was broadened to almost unlimited possibilities within and beyond healthcare; its use in renal disease became a target to achieve proof-of-concept for the technique being developed. In this way, the nanoconductor itself became the centring object (Knorr Cetina, 1999) for the physical scientists – their focus was on developing the science that would transform the nanoconductor into a biosensor.
Narrowing the framing around this technical work created power imbalances within the team; positioned as the centring object, the nanoconductor effectively distributed power in its relations with other actants. Knorr Cetina (1999) refers to the relations between actants as object-circuits. By establishing legitimate object-circuits with the nanoconductor a hierarchy of expertise was ordered between the disciplines, based on who had the knowledge to develop the biosensor. This framing privileged the position of the engineers who, as custodians of the nanoconductor, had the power to validate other disciplines’ claims to object-circuits. In this way, the project was framed as an endeavour of engineering to be assisted through collaborations with other disciplines.
For the engineers, the work to develop a functioning biosensor was reduced to a set of chronological steps, which aligned the other disciplines along a temporal trajectory according to the particular junctures at which their expertise was required. The first step was to develop a functionalization procedure – the process of attaching antibodies to the surface of the nanoconductor – which requires expertise in chemistry. Because of their relations to the actants involved in functionalization, specifically the nanoconductor and attachment chemistry procedures, from the beginning the engineers and chemists were positioned as the ‘core technical team’. The next stage was the validation of the POCT device in biological samples, which requires expertise in biochemistry and medicine. As validation required a functionalized device, this was positioned as downstream work for the second phase of the project. Within this linear approach the sociological research, which was seen to deal with end-use, was positioned at the end of the process.
This framing challenged the value of the sociological research. In contrast to the other disciplines, which were understood to have legitimate object relations with the nanoconductor, the sociological component could now only be explained as a requirement of the funding body. Without this stipulation, the natural end point for this project would have been the validation work, and there was a general feeling among the team that it was too early to explore the end-use:
‘[y]ou need to have the technical capability and then you look at how you implement it … there’s no point in trying to figure out how you’ll deal with point of care devices unless you’ve got point of care devices’ (Lisa, biochemist, year 1).
Time pressure was an influential actant within the project, and enhanced the downstream positioning of the sociological research. Although the project had been successful over competitors in gaining funding, the commissioning body was an unstable ally and it was clear that to maintain its allegiance for a second phase the project must demonstrate good technical progress. Within this context, the sociological research represented a time burden for the physical scientists; it was something that might be useful to think about, but doing so would take precious time away from the ‘real work’ of the project:
To be honest we are all extremely focused on getting the prototype to work, because if the prototype does not work there’s nothing … it would be nice to think about when it works how we proceed, and technologically we know the next step. But to think about a step after that, who’s going to use it and how are they going to use it … it’s kind of outside of our league to think about this. (Christopher, engineer, year 1)
In response to the time pressures the engineers adopted a problem-focused approach, seeking quick solutions to specific problems that would temporarily stabilize relations between specific actants and enable work in other parts of the project to begin. This strategy necessitated the division of the group into smaller sub-teams to work on individual components of the research. As the project progressed, sub-team meetings that focused on specific problems were favoured over whole team meetings. With these problems framed as ‘technical issues’, the sociological research was of little value to any of the sub-teams, and consequently I struggled to access the meetings – with no direct role, people typically forgot to inform me of their occurrence. This made remaining up to date with technological development difficult, and reduced the possibilities for others to feed into the sociological research, furthering its separation from the ‘real work’ of the project.
With sustained commitment to addressing the specific healthcare need outlined in the original proposal, the project could have adopted a broader framing (beyond technical considerations) that would have allowed space for the sociological research to contribute. Instead, as a compromise between the generic tool the physical scientists wanted to develop and the defined end-use originally agreed, I redesigned the sociological research to explore the potential use of the device within three different clinical settings. This approach was agreed for the purpose of the sociological research, yet the wider project maintained a focus on developing a generic device. In this way, although the sociological research had been included to ensure social impact, the physical scientists did not accept the challenge to the linear approach to technical development that this brought. Thus from the beginning, my collaboration with the physical scientists adopted a position of tolerant ambivalence, rather than a transformative approach.
If sociology is to be an effective partner in ensuring scientific research achieves societal impact, collaborations must adopt a framing in which its contribution can be valued. Although the framing at the beginning of a project has important implications for the nature of the collaboration, framing is not a discrete event but is negotiated and renegotiated throughout the duration of project (Knorr Cetina, 1999). I argue that it is in the establishment of shared spaces between disciplines that the continued framing of a project is negotiated. Yet shared spaces are fragile and difficult to maintain, even between disciplines with similar perspectives and approaches. With a vast epistemological and ontological divide to bridge, sociology has an even greater challenge in establishing shared spaces with physical sciences. Below, I illustrate this by comparing the changing positions of chemistry and sociology within the POCT project. With the privileged position the engineers occupied, it was by establishing shared spaces with them that the framing of the project was renegotiated.
Reframing in Shared Spaces
The role of chemistry within the POCT project was to develop the functionalization process, for which Neil was responsible. Neil’s previous success in functionalization indicated his expertise for the project; however, the nanoconductor was developed from graphene, a substance with which Neil had no previous experience. Ironically, collaboration in the POCT project severed the object-circuits with successfully functionalized surfaces that had secured Neil’s involvement, and demanded he form new object-circuits with graphene. This would determine the nature of collaboration with the engineers, and the subsequent position of chemistry within the project.
The nanoconductor was an important actant in the work of both the chemists and engineers, and with their relations to the nanoconductor providing a shared space for collaboration, the relationship began in cooperation. In response to the pressure for rapid results, the chemists quickly developed a manual functionalization process. This was adequate for the purposes of securing proof-of-principle, but a viable commercial product required an automated system. At this point, the engineers continued to work with the manual process for the purposes of developing the wider system, while the chemistry team focused on developing a more stable process for the prototype. However, the required chemistry had not previously been shown to work on graphene. The relations between the actants involved, including the graphene, antibodies, chemical processes and electronic systems, ultimately proved too fragile and unpredictable for the chemists to control within the confines of the project, and eventually hopes for an automated system were abandoned. Reflecting on the progress of the chemistry work, Neil felt that the difficulties came from the instability of graphene in relation to chemistry:
It’s been very difficult. I had not realized when we set out to do this project, what a difficult material graphene was to work with … it’s horrible, dirty, filthy stuff. I mean it’s great for the engineers because they can use lots of really hard chemistry to do their engineering … but if you want to involve biology with it then it’s much more tricky. (Neil, chemist, year 3)
Neil’s description of graphene suggests its multiplicity: in relation to producing the nanoconductor, graphene is a stable and hardy material, yet in relation to biochemical processes it is unstable and difficult to control. The more experience Neil gained with graphene, the more unstable it became, and ultimately he could not forge stable object-circuits. With their different relations to graphene, the chemistry and engineering teams lost the nanoconductor as a shared space; they were effectively working with different materials.
The dissolution of this shared space was not just a conceptual observation, but inspired a physical separation that changed the nature of the collaboration. Whereas the engineers worked with the nanoconductor, the chemistry team was given pieces of graphene with similar properties to the nanoconductor to work with. They became two separate teams, working on two separate problems. Instead of the cooperative approach with which the collaboration had begun, the instability of the actants involved in chemistry work inspired a retreat into tolerant ambivalence. Indeed, segregation arose as a strategy to deal with conflicts:
‘[t]here were lots of arguments and conflicts, and these self-resolve in the end by people not necessarily working together as one big team’ (Alex, engineer, year 3).
The changing position of chemistry within the project illustrates the location of power within relations to the nanoconductor; as the relations between chemistry and the nanoconductor became unstable, so too did the position of chemistry. The experiences of the chemistry team show the difficulty of maintaining shared spaces in interdisciplinary projects, even between disciplines that share much common ground. The challenges are even greater when shared spaces must be created across epistemological divides.
Bridging Epistemological Divides
From the beginning of the project there was little shared ground between sociology and engineering, and the collaboration began with an approach of tolerant ambivalence. Making progress with the research required that the value of sociological knowledge was legitimized within the project. Initially, I attempted to introduce socio-technical understanding of innovation into the project, using sociological studies to illustrate that the success or failure of technology is a consequence of more than technical capability. However, the epistemological and theoretical void between our perspectives proved too vast to allow debate. The team found it difficult to relate to the research I described, perceiving it to be focused on fundamentally different technologies to the POCT device, and therefore of little interest. Instead of inspiring debate about how to use the sociological research, these discussions highlighted the disparities between our approaches. The nature of the collaboration made a noticeable change with the introduction of empirical knowledge from the field as an actant.
As with the rest of the project, my fieldwork was discussed at smaller meetings, which provided an opportunity to discuss both findings and the future direction of the work. Within these meetings the fieldwork, rather than the nanoconductor, became the centring-object, and moved the spotlight from technical development to reveal the construction of the device that continues after it has left the laboratory. The presence of my empirical work within the network changed the physical scientists’ perception of both the field of use and the nature of the POCT device, and at this point they began to engage with the sociological research. My fieldwork provided us with a shared space and, from their perspective, gave relevance and legitimacy to the sociological research:
I think you’ve proved yourself very useful. So I go back on my initial scepticism! … Some of the things in your presentations give very clear messages from the people you were talking to that they’re embedded in an organization, and they’re only interested in innovating in that organization if it really does demonstrate something that is a big improvement from what they’ve already got. (David, engineer, year 3) If I was really honest, at the outset I was probably quite sceptical … I think what you’ve done, certainly from my perspective, is you’ve made me think more about not just how do we make a device that works, but how do we make a device that goes from a prototype that we think is actually pretty clever, into something that can make a difference in the context of a patient setting. (Lisa, biochemist, year 3)
For Lisa and David, engaging with the sociological research highlighted the difference between developing a working technology and a workable device. Indeed, by the end of the project most of the team members had come to value my research in some way. Although, not everybody’s perspective had changed – Neil in particular continually struggled to see its relevance:
I think it was far too early to be doing that sort of research. I’m sure in its own context it has its own value. But I don’t know how much it’s informed what we’ve had to do … The first stage of actually getting something that works is so difficult that if we take any of our attention off that target it becomes impossibly difficult … [Your research is] kind of like the salad with the steak: it’s there, it’s green, and because the funding body say you have to have vegetables, we eat it. But does it really make any difference? No, it’s the steak that you’re after. (Neil, chemist, year 3)
Neil’s difficulty in understanding the sociological research emphasizes the influence of my fieldwork as an actant in reframing the project. Throughout the project Neil had little involvement with my research, avoiding most of the feedback meetings. He was extremely focused on developing the chemistry, and saw anything that distracted his attention from this as a hindrance. For Neil, the project was always framed around the nanoconductor; he had no relations with the sociological research and did not experience the reframing that it could bring. As a result, I was unable to move beyond conflict in collaboration with Neil. By contrast, reframing the project around the sociological fieldwork in interactions with the other team members enabled us to move from the shadow of tolerant ambivalence into something more like cooperation. However, although the team had engaged more with my research, a transformative mode of working was still significantly hindered by our clashing approaches to the research.
Undermining Transformation
Maintaining a flexible application for the technology remained strategically important for the physical scientists throughout the project, which effectively pushed decisions about end-use outside of the relevance of the project:
Because we are in the luxury position that we know that if we get it to work for one or two it will work for any protein, we haven’t so far thought about [applications] … If we had been developing a sensor that would only work for one specific protein, we would have had to think about [end-use] much earlier; now would be way too late. (Christopher, engineer, year 3)
With the perceived flexibility of the device, the sociological research lost any sense of urgency: it is nice information to have, but we can do something about it later. Consequently, there was no coherent approach for dealing with the results of the fieldwork. Indeed, it was not even necessary to agree on the implications of the research. For example, David interpreted my research as evidence that the POCT device was unlikely to be implemented within the NHS and should instead be marketed as a home-testing device, although he knew that Alex disagreed:
I would see it being sold to the general public … Because of all the issues that you were raising … I think it’s quite difficult to get a sensor accepted in the NHS, because they’re quite happy with their current sensors … Alex thinks that these things will be used in a clinical environment, rather than going to the general public. (David, engineer, year 3)
In relation to the linear approach to technological development the project adopted, these disparities in goals for the device posed no problem, as decisions concerning end-use are made later down the line. However, for the sociological research the lack of a cohesive goal prevented the effective use of information about the field of use.
The completion of my fieldwork marked the dissolution of the shared space between me and the other disciplines. I was no longer required as a spokesperson for the actants in the field, since securing their allegiance was of little concern for the engineers at that time, and the collaboration fizzled out. Indeed, we did not meet again as a team before the project ended. Without a clear and shared goal for the POCT device, the collaboration was unable to move towards the transformative mode of working needed for the sociological research to be fully utilized. In conclusion, I reflect on the insights this research gives into the dynamics of interdisciplinarity across the social and physical sciences.
Shaping Interdisciplinarity
The focus on ensuring academic research achieves ‘impact’ is encouraging greater interdisciplinary collaboration. Rather than being detrimental to the academic standing of sociology, as Holmwood (2010) fears, interdisciplinarity may conversely open up new opportunities. Although by no means the only role for the discipline, sociology is uniquely positioned to understand the social impact of research, and therefore is arguably an essential component of such work.
However, this new frontier can only be valuable for sociology if its contribution is valued by other collaborators. As illustrated in the POCT project, if sociological work is marginalized to give priority to other forms of knowledge its capacity to facilitate societal impact is constricted, which offers neither a worthwhile pursuit for sociology nor an effective route to impact. For interdisciplinarity across the physical and social sciences to be successful, it must be understood as more than the meeting of disciplines. To use Knorr Cetina’s (1999) terminology, interdisciplinarity should seek to construct new epistemic cultures – new fields of knowledge – that are free from disciplinary dominance. This is achievable only through a transformative mode of collaboration that transcends the power imbalances and hierarchies of knowledge that are incubated by other modes of working.
As the mode of collaboration is inextricably linked to the framing of a project, to produce new epistemic cultures research must be framed by the intended societal outcome – a goal to which all disciplines can contribute – rather than the technical details, which privileges certain types of knowledge. Pohl and Hadorn (2008) call this transdisciplinary research, whereby the collaboration of disciplines is seen as the necessary means to pursue solutions to real-world problems. My experience offers insight into how the framing of future collaborations might be managed to enable sociology to capitalize on the opportunities that interdisciplinarity presents, and ensure that academic research achieves social impact, with implications for both individual researchers and funding bodies.
Individual Researchers
The framing of the POCT project around the development of the nanoconductor provoked many challenges for the sociological research. To enable a more favourable framing, I offer two pieces of advice to researchers embarking on similar projects:
Ensure that interdisciplinary projects are committed to addressing a societal problem, rather than advancing scientific processes.
Reframing occurs in the shared spaces between disciplines – work to establish common ground as a priority.
Addressing Social Problems
The importance of ensuring interdisciplinary projects aim to address societal problems must not be underestimated. Although the original funding proposal outlined a specific application for the POCT device, from the beginning it was clear that the team were not committed to this end-goal, instead intending to develop a generic device. With reference to the implications of the end-goal for the framing of the project, I urge social scientists embarking on interdisciplinary projects to ensure that the aims and aspirations of the project are explicitly discussed and agreed early on, ideally before committing to the project. As important as it is for all parties to have a shared vision, it is imperative for sociologists that this is directed towards addressing a societal problem; without this, sociological knowledge is vulnerable to being pushed outside of what is relevant by others in more powerful positions. Arguably, a shared commitment to a defined and applied end-goal should be a condition of sociologists’ participation in interdisciplinary endeavours.
Establishing Shared Spaces
Collaboration across diverse disciplinary boundaries is notoriously difficult, as languages and perspectives clash. This research suggests that the establishment of shared spaces is imperative to achieve effective collaboration, as it is within these spaces that the continued framing of the project is negotiated. My findings resonate with Oughton and Bracken (2009), who also emphasize the importance of shared spaces for the framing of interdisciplinary projects, although our findings suggest that these spaces function in slightly different ways. Oughton and Bracken (2009) suggest that shared spaces enable effective communication, which allows the framing of the project to be negotiated. However, in the POCT project it was the shared space itself that acted as a centring object to frame the research, and within this framing the nature of the collaboration was determined.
I argue that sociologists, who are likely to share less common ground with disciplinary perspectives outside of the social sciences, must work urgently to establish shared spaces. The objects, subjects or discussions that constitute shared spaces are likely to vary across disciplines and project teams. Given their importance, sociologists would be wise to work to establish these spaces from the start of a project, understanding that it is likely to require a degree of trial and error.
Funding Bodies
Although the nature of the collaboration on the POCT project was affected by the framing inside the project, it was also influenced by the agenda of the funding body, which exacerbated power imbalances. The eligibility criteria in the original funding call stipulated the inclusion of social science, yet scientific development was prioritized over the applicability of the device in securing further funding, which placed sociology in a vulnerable position.
My experience carries an important message for funding bodies: if the intention is to commission transdisciplinary collaborations capable of addressing societal issues, then it is not sufficient to ensure that applicants ‘tick the social impact box’ at the time of applying. They must be judged on the integration of this into the project when it comes to re-funding and evaluation. The disconnect between what the funders asked of the project at the beginning and the criteria against which success was measured, undermines the value of sociological research. Furthermore, if research funders are committed to addressing societal problems, they must ensure that projects have an appropriate end-goal that remains fixed throughout the project.
This article has drawn out some of the practical challenges posed to sociologists as they embark on interdisciplinary collaborations beyond the social sciences, offering empirically based guidance as to how future projects might be improved. The increasing push for interdisciplinarity offers exciting opportunities for sociology to apply its rich resources in new ways, but for these opportunities to be valuable for sociology we must find effective ways to ensure that such collaborations are truly interdisciplinary.
Footnotes
Acknowledgements
I would like to thank Susan Halford, and the two anonymous peer reviewers for their insightful comments and suggestions.
Funding
The research on which this article is based was undertaken as part of a PhD funded by the Engineering and Physical Sciences Research Council.