My team are halfway through developing an approach to computational reasoning that represents interactions between systems as systems. I’ve led other interdisciplinary projects but never with members that have such dramatically different knowledge: neuroscience, narrative analysis, computer science, biology and design. As a visual example, consider the different modes of modeling we each use, below. One of our goals is to bring the strengths of these methods together:
So as well as addressing our funded problem, we’ve also been working on the meta-issue of how to meaningfully synthesize our ideas. Without this, we can’t even discuss the project. It is also the only way we can go a step further, and innovate together.
To di this, we have to build a new space of discussion, in which we know how to relate our diverse knowledges. I’ve started to refer to this synthesized space as a shared story. I’m thinking of something particular from narrative analysis in using this phrase. I make animated models of how meaning is woven from fragments of existing knowledge, step-by-step, in an unfolding story. In this example from Red Riding Hood as a Dictator Would Tell It, the process of integration is gradual and piecemeal, and eventually produces a new frame of reference, one that shows how fairytale and dictator contexts can be connected in this one instance. This piecemeal assembly is the job of narrative – to reveal and connect information in such a way that unexpected circumstances or novel ideas can be comprehended. It forms a new context.
I believe this same process is needed in collaboration, especially when its disciplinary spread is extreme. Otherwise, the exchange ideas between fields is going to be tokenistic. And confusing.
For example, in our group, the word system meant something different to each of us – for our biologist, it’s a bodily process, in computer science it’s an implemented technology, in narrative analysis it refers to a regular way of doing things. We didn’t need to translate every single term. But we did need to figure out what terms were going to be important in this context, and through them, understand how our knowledge would fit together. Our group needed a shared story to translate everyone’s individual knowledge into some sort of common format, one that can be accessed by all participants.
Department of Energy dumping toxic waste at Hanford, Oregon.
When a group doesn’t succeed in coming together properly, the results can be catastrophic. William Lawless, now a research fellow at the Navy Research Lab, tells a true story about how a thousand people with doctorates made a very stupid collective decision on behalf of the Department of Energy: they decided to bury nuclear waste in cardboard boxes in the ground (Lawless, 1985) (Ledford, 1984) (Mayell, 2002).
Anyone who has been part of an ineffective committee might understand how this is possible. Without good leadership, a group is capable of making dumber decisions than any one person could usually manage. This poor decision-making reflects a compromised process: instead of being guided by a shared story, the project is instead driven by factors that are non-intrinsic to its goals, such as peer or societal status (Cohen and Lotan, 2014, p. 36). Decisions become informed by a struggle for dominance. One way to think about collaboration – especially interdisciplinary collaboration – is that a shared story enables issues related to the target problem to dominate, rather than the personality dynamics of the group itself.
I’m currently writing a chapter about the relationship between collaborative frameworks and narrative processes, for a book titled Computational context: Why it’s important, what it means, and can it be computed?, supported by the Navy Research Lab. It discusses interdisciplinary collaboration in general but also how my diverse team tackled the building a common perspective.
In short, we wrote three peer-reviewed papers, which served as stepping stones to gradually integrate our work. In the first paper, I recast my previous work towards the new project. In the second paper, I linked my research to the collaborator whose ideas were closest to my own – this was Larry Sanford, neuroscientist at the Eastern Virginia Medical School. We found a common feature across our two fields – the reorganization of memory structures. In a third paper, everyone came together for a special issue of Progress in Biophysics and Molecular Biology. Again we found a common structure across all five disciplines – in this case, it was the stages of emergent organization.
There are surely many ways to incrementally integrate a group, not just by writing successive papers. This worked for me as a writer, because it was a larger version of the process of weaving fragments of information in fictional stories. And just like writing a fictional story, this process sometimes felt like a free-fall. The scariest step was making the below model, in which information from Patric Lundberg’s herpes research was presented using the operations of my narrative modeling technique.
(If this is too small to read, try making the browser window larger)
The method used was one I had previously only used to track narrative inference in literature, film and TV shows such as Game of Thrones. I had the strangest hesitation about drawing this biology-related version. It was worse than facing a blank page as a writer – something I have tricks to work around. I sat with computer scientist, Ted Goranson, and Patric’s work, and we painstakingly built the new description over three days. Each step felt precarious. Why? Perhaps it was simply because there were no rules or firm ground. We stepped into empty air, with no accompanying theory or literature, and only the rope of our ideas connecting us to anything at all.
The result is the first ever representation of biological and neurological behavior using operations of context, as usually observed in narrative. Our task of creating shared story isn’t finished. My team are building it as we go, and probably won’t be finished until the project is over.
To better support this process, I’m collecting qualities from the arts that enable the development of a shared story. My favorite so far is a tolerance for ambiguity – as described by Cindy Foley at the Columbus Museum of Art. A sense of ambiguity, of spaces becoming less tangible and more dynamic, comes with moving beyond the edge of your discipline. The feeling can be disorientating and I’ve seen promising projects implode from the fear this inspires. When that sense of freefall is feared, everyone races back to comfortable scripts – their discipline or social status. No-one wants to wheel through empty space. I’ve seen a tug-of-war for dominance ensue as a result, in the fight for ones own disciplinary terms to be chosen as dominant.
In the next stage of this work, I think my team will need to define anew where the innovation is going to be in this project. Every field takes its own scope of innovation for granted. In creative writing, the novelty comes from casting a new perspective over reality and ironsmithing language to hold it; in neurobiology, it is making a new theory that can be instanced in brain or body tissues and reproduced; in design, the novelty comes in a sublime act of simplification that also reveals keys paths of formation.
We are bringing modeling methods together to capture system-level behaviors. Some of the operations of context I’ve developed for narrative need to be used, but not in their current form. Deciding how the design of this technique should be redeveloped depends on finding a fit between essential and disposable elements in each of the participating fields’ modeling methods. So far this discovery has come to depend entirely on working through examples.
Each of us describes an example, such as PTSD, from our different perspectives. The example becomes a rosetta stone, the grand underpinning analogy that brings us together. It can be jarring to hear something familiar in a foreign format. I remind myself to have patience and be willing to know very little, at least temporarily. Through this process, I’m starting to see where our innovation is, one that is anchored firmly to the example itself. We will take parts of my narrative modeling of analogy, it’s most high-level aspect, and use it to model the accumulation of fear learning. This might make it possible to illustrate why one mouse develops PTSD after a traumatic event, and another one doesn’t.
When writing a story, you discover the rhythm of the piece as you work through the actual text, making new associations. Each new idea you form feels tentative and temporarily, but you still jump to it, to shift the horizon line. Eventually, turning back, the emerging structure comes into view, turning freefall into freedom.
Cohen, E., and Lotan, R., 2014. Designing Groupwork: Strategies for the Heterogeneous Classroom. Teacher’s College Press, New York.
Lawless, W., 1985, Problems of Nuclear Waste, Bulletin of Atomic Scientists, 41, 10, p. 38-42.
Ledford, J., 1984. Engineer claims DOE has problems with radioactive waste. United Press Int.
Mayell, H., 2002. Idaho, U.S. Battle Over Nuclear Waste Dump. National Geographic News, April 2, 2002, accessed online November 10 2017, https://news.nationalgeographic.com/news/2002/04/0402_0402_nuclearwaste.html