Can serendipity be harnessed and used in learning and research?

This is a guest blog article by Jim Hensman from Coventry University.

I have been involved over a number of years in various projects relating to collective intelligence and collaborative working. Organisational, social, psychological and cultural factors have to be taken into account when working in these areas, but technology can play an important role. One focus of this has been developing techniques to find non-obvious research connections, particularly cross-disciplinary ones. Finding connections we don’t expect is considered a key aspect of serendipity, so it is worth looking at some of our experience with this. The tools we have developed start from looking for common terminology in information about the work of researchers – taken from their profiles, publications etc. This can be considerably enhanced by identifying connections at a higher level, creating appropriate ontologies and other knowledge structures to do this. Again, in the way that search engines such as Google have to rank and prioritise matches, we have to do the same, using concepts such as semantic distance as a guide and displaying results to reflect this.

There are many important issues that arise and have to be dealt with. We have been engaging particularly with groups and communities, such as the EPSRC Spires community. Finding ways of linking individuals both within and between communities has to go beyond just aggregating the links between their individual members and requires developing methods to generate group ontologies. Different communities often talk about similar things in different ways and represent information differently. Extending the scope of the work to include finding links between research and business, part of current work on Open Innovation for instance, requires tackling this issue. Particularly interesting and challenging has been to ask whether we can go beyond finding connections between work that already exists to identifying new areas for research – something that would often be associated with serendipity. A number of techniques have been experimented with in this area that have worked in practice and show considerable potential. Putting these different methods together has revealed some startling connections – from ones between research on wallpaper and cancer research, to links between petroleum engineering and the Large Hadron Collider. Systems developed have been deployed in many different contexts, from long-term strategic work with research projects and communities to use at conferences and workshops. One area in particular in which significant work has been done has been in developing spaces to facilitate collective thinking and discussion for physical groups, which can also connect to virtual participants. This work, which uses immersive 3-D interactive environments with word clouds and other representations of ideas, utilises some of the techniques described above to connect ideas being discussed and augment them using information about the participants in the discussion. You can see a video of some of this work at http://vimeo.com/49155054 and relevant websites for wider work carried out as part of JISC funded projects are project-inspires.org and project-brain.org.

Could developments in techniques like this potentially include and explain in principle at least, all of what we call serendipity? I don’t believe so. Let me relate two personal incidents just from events I attended last week. At one, I was sitting in the audience at a presentation when I overheard someone behind me say a few words which described exactly something I was working on. Chance clearly played a major part in finding this connection. In another case, I was at an event for businesses in the arts sector. The event organisers had used a simple matching technique to suggest potential connections between attendees, and I was able to compare this against some of the methods my projects had been developing using myself as an example. Although these worked considerably better, they didn’t find the best connection I established, which developed out of the interaction from a casual conversation. How can we investigate serendipity from this angle? I have been involved in improvisational performance of different kinds for many years, which provides some interesting insights. Even here, where what happens appears to be completely unstructured and random, things aren’t always what they seem. The kind of improvisation one sometimes sees on the stage or on TV, where random words suggested by the audience are woven into a meaningful narrative by the performers, of course requires considerable skill and experience. What is perhaps not often realised is that improvisation without something specific to base it around is normally even more difficult, and performers involved will usually introduce their own details to deal with this. I have worked with several ways of adding structure to activities of this kind, and when these techniques are carefully designed they can enhance the collective creativity which arises – something that the Riders project on Interactive Drama Environments has also been investigating.

Can methods of this type contribute to serious research? Creating suitable conditions for investigating this area is not easy, but I was able to organise a workshop as part of one project to try some ideas out. This brought together about 40 learning researchers and the aim was to identify new areas for research and define projects around these. Although some of the techniques described earlier were used in various ways, this was before much development had taken place with them and they were only used in rudimentary form. In the main the session was designed to encourage interaction, including using principles that worked with dramatic improvisation. In less than two hours a significant number of new project concepts were thought up and detailed – sometimes to the level even of the project name. A number were taken forward and I believe at least one of these is still running today. A number of the participants came to me afterwards asking me how I had “fixed it”, whether I had planted people etc. The truth of it was that it was just a good example of creating the right conditions and then letting collective intelligence and serendipity take over.

In finishing, I would just like to suggest one tentative hypothesis about serendipity that has come out of the work I have mentioned. The reality our senses detect around us in the physical world we know is not as it seems. It is in fact an emergent phenomenon that arises from something that quantum physics tells us is far different. We may visualise an atom as a miniature planetary system – but the truth is that this is just a convenience on our part. But there is also the potential of misunderstanding in doing this. Knowledge is also an emergent phenomenon. It arises through a whole number of complex processes, which often we try conceptualise using what we see as the structure and rationality of the finished product. There is a similar danger when we do this in both learning and research, for instance in defining rigid required outcomes and impacts, which we need to take account of. We need to have scaffolding and structure, but also allow space for serendipity as a valid and necessary element. The work I have been involved with, even though a great deal more needs to be done, has certainly demonstrated the massive untapped potential for research and knowledge generally in breaking out of the silos into which they are often divided. Putting this together with the developing understanding of serendipity and how it can be used, which the Serena project is a key part of, would I believe constitute an extremely powerful combination which could unleash huge possibilities for learning and research.