Complexity Sciences: Towards an Alternative Approach to Understanding the use of Academic Research Lemay, M. A. and Sá, C. (2012). Complexity sciences: towards an alternative approach to understanding the use of academic research. Evidence & Policy, 8(4), 473-494. Abstract Academic research is increasingly linked to a range of socioeconomic benefits. With increasing investments in academic research, there is growing pressure to improve the uses of research and demonstrate its impacts. From theoretical and methodological perspectives, the use of research is not well understood. This paper examines the limitations of the widely accepted conventional framework for understanding research use and explores the potential for conceptualising research use as a complex social process. The paper argues that properties and behaviours of complex systems are relevant to building a realistic and more complete understanding of the research use process. First off, this is not a research paper. It is published in the Debate section of Evidence & Policy so there is a lot of opinion without the necessity for backing it up by empirical evidence. The opinions are certainly backed up by the literature so this is not an uninformed opinion. And a heads up, this is not written in clear language. For example, « An appropriate ontology for such a (complexity-based) theory is not served by the deterministic, reductionist perspectives usually adopted from neoclassical theories of innovation« . As an editor recently said of a sentence I wrote, « I would love to make this more simple if I only understood what this means ». Amen to that. The paper is based on the premise that traditional theories of research use are reductionist and deterministic and involve linear causality. That means they are based on assumptions that reduce (=reductionist) concepts to their most simple forms that enables us to determine (=deterministic) outcomes from those simple systems that are based on linear models of research instead of models that involve multidirectionalknowledge flows. The authors call this linear and simple model the « conventional framework for understanding research use ». I don’t actually think this is a conventional framework at all. I think this is an outdated model of research use that has been abandoned in favour of more iterative and social models that are widely embraced. It is these iterative and social models of research use that are today’s conventional model. Nonetheless, the rest of the thinking on complexity is useful for challenging how we think about modelling our research and services to enhance research use. The authors discuss the roles of the « user » in research use frameworks and models. The authors indicate that the so called « conventional models » of research use do not place the user at the centre of the research use process and few consider the roles that users lay in the process. Giving research users active and integral roles in research user makes the system inherently more complex….hence the need for a complexity lens. We are now on the fourth page of the paper and the authors finally provide us with some background on complexity science. First a little on systems thinking. See an earlier journal clubpost on the work of Best & Holmes for systems thinking in research use. « Systems thinking within the context of research use considers the entire environment (system) within which research is conducted, disseminated, translated, used and applied….Context is a dimension of systems thinking that appear to be critical to the process of research use in policy and practice….Research use is influenced by the specific conditions that exist at the particular time and place in which it occurs. Context suggests that standardised research use strategies designed to be applied broadly are unlikely to be effective. » Essentially, because context changes between different research use settings then what works in one setting/context will not necessarily work in the next. This is part of what drives complexity and challenges making broad conclusions about « what works ». Context is one of three factors (along with evidence and facilitation) that are the basis of the popular PARiHS framework. This is reflected in our own knowledge mobilization practice. Locally York University’s Knowledge Mobilization Unit employs six practices to support knowledge mobilization but implements these differently if we are working with community partners in York Region versus working with policy makers in the Ontario Public Service. The types of conversations at research forums or KM in the AM are different when partnering academic researchers with community expertise versus PhD policy makers but both happen in the context of in person meetings where academic and non-academic expertise are provided equal voice. The role of the knowledge broker is to be able to adjust style and manage expectations of partners appropriate to these different contexts. This respects the complex nature of research use. Similarly, while the ResearchImpact universities share a common philosophy of connecting research and expertise to inform policy and practice, we all implement differently. « Work on complex systems emphasises the interaction and interdependencies among system components. Separating the components eliminates the relationships that define or give the system characteristics, properties and behaviours. Because of the inter-relatedness of the components, even though they are autonomous, a complex system cannot function if it is separated into its individual parts. » Looking at research use through a complexity lens the collaboration or co-production of knowledge in each instance needs to be considered as a whole and placed in context, not isolated and broken into its component parts. Take a look at the list of characteristics of complex systems and see how many align with characteristics of research use: many interacting entities (yes); dynamic (yes); self-organising (maybe but yes with the help of knowledge or innovation brokers); systems within systems (yes); far from equilibrium (yes); interdependent (yes); co-evolving (especially if this can be interpreted to mean co-production); path dependent (not necessarily); contingent (yes, contingent upon context); situated in larger environments (a university is in a community that is in a region that is in a province…although I think this is the same as systems within systems); adapting (yes) emerging (yes, as in the outcomes of any innovation system are continually emerging). Based on this analysis research use fits nicely within a complexity framework. An interesting observation of complex systems, « It is virtually impossible to predict how a complex system will evolve by studying the individual elements of the system or past behaviours. » This is linked to a discussion of strategic planning within complex systems. Basically, you can’t. You can’t predict future outcomes based on past actions in a complex system. This has real implications for planning and evaluation. As soon as you start to seek one set of outcomes by planning them and trying to measuring them you miss the opportunity to observe unintended consequences arising from a complex system; therefore, evaluations may need to be based more on process than on outcomes. But what does this do for all our evaluation colleagues who are charged with maximizing the impacts of a research use strategy when the outcomes cannot be predicted? How can a system of knowledge mobilization share common evaluation metrics when the contexts are so different? Complexity theory suggests that the best we can do is learn from the past in order to understand the dynamic and emerging nature of a complex system. These learnings can inform our actions but not predict outcome. At the end of the day this paper creates a feeling of being fairly overwhelmed with complexity and yet underwhelmed with what we can do about it. « A possible strategic outcome….could focus on building awareness of the implications of complexity and creating the conditions necessary to effectively operate as part of a complex system. » That’s nice but it doesn’t help me decide what to do to improve research use. The paper would have benefited immensely from a real world example of how complexity was embraced in research use and how it was used to inform planning to maximize the impacts of research. It might be a debate paper but the debate needs to be grounded in real world experience. Questions for brokers: If we can’t apply strategies across different contexts why does our « industry » seem to have a focus on developing and disseminating tools? Shouldn’t we focus on building blocks or principles and determine how to apply or implement those on a case by case basis? What is your experiences adopting and adapting tools from one context to your context? Do you think you’re working in a complex system? If so, then how do you manage complexity in your planning processes? See a recent post about knowledge mobilization not being rocket science. It really isn’t. But this doesn’t reconcile with knowledge mobilization operating within a complex system. What elements of your practice are generalizable to other contexts? RIR is producing this journal club series as a way to make the evidence on KMb more accessible to knowledge brokers and to create on line discussion about research on knowledge mobilization. It is designed for knowledge brokers and other knowledge mobilization stakeholders. Unfortunately this article isn’t available in an open access format. If you’re a community member seek a colleague at your local university to obtain this article for you. Read the article. Then come back to this post and join the journal club by posting your comments.