Models and Narratives (2)

Attendees:
Ilan Chabay, Heinz Gutscher, David De Roure, Sarah Wolf, Armin Haas, Achim Maas, Vittorio Loreto, Filippo Addarii, Steven Bishop, Trista Patterson, David Chavalarias, Patrik Jansson, Ralph Dum, Kurt Dopfer, David Tuckett, Jason Greenlaw, Zhangang Han, Ralph Dum, Laszlo Pinter, Merijn Terheggen, Joan David Tabara.

– We are required to create a narrative for GSS. But this is more the objective of the conference. Most often scientists tell the world why they are going to build a model (although they may have a different narrative in private). A narrative created to get funding for building models is not the same as a narrative to make models useful to the public. How do we explain models to a big global community? Pragmatically, we need a narrative first to understand how what we do, as scientists, fits into the global systems science narrative. Different scientists have different processes. Two examples of different processes might be:
1. problem -> narrative -> model
2. model -> narrative -> problem
In this workshop however, the focus is more to consider narratives as a subject of study in their own right.
– Narratives should be used to understand limitations and context dependency of models. They can package research for general audiences and can convey complex ideas. They can reach globally through local contexts. The audience of a narrative needs to be considered.
– ICT and narratives are now intimately connected; new media produces new images and narratives very cheaply. For example, YouTube creates narratives with just a camera, which, if they get taken up, can shape the views of future generations. Online games are another source of creation and interaction of narratives (see Insite project).
– Automatic translation of narratives as an ICT contribution. Also semantic analysis (computational linguistics). Perhaps the a task is to not only analyse the grammatical structure in a given sentence, but the ‘grammatical structure’ of the narrative as a whole, in order to map an argument.
– It’s not just a matter of getting feedback from a narrative after the modelling process. We need to engage with stakeholders during the research phase to understand that we are studying the right question.
– Narratives are traditionally enacted in a theatre. What is the theatre for these types of narratives? We need to present them in the right context (place, time, format)
– Narratives can be descriptive or prescriptive (esp with policy-makers). We can render a lot of good science useless by being too prescriptive.
– We need to get serious and engage professionals who already study and create narratives. How do we start connecting with these people?
– Narratives are the primary means of humans making sense of the world. But we don’t know how they work, how they spread and how they stop spreading. A large fundamental study is required to gain insights into all of these. It’s intuitive that narratives should be linked with heuristics but this needs to be better understood. We should begin to focus on how narratives really matter. Not ‘how do people think about what scientists think?’, but ‘how do people think about climate, finance, etc?’ ICT provides us for the first time with tools for studying narratives. You can’t impose narratives, we need to understand them on a fundamental level.
– We are far from understanding the right science and the right model – how can we construct the right narrative?
– We don’t always need to create narratives – there are many out there already. How can you find a narrative that underpins (or even contradicts) what your analytics show you?
– There is a connection between this discussion and the future of scholarly communication. What is ‘beyond the pdf’? Can we define knowledge sized chunks? What chunks are required? A human readable component surely but what about a machine readable component? If we only work in units the size of a human narrative, will this restrict us in terms of what we can comprehend?
– Narratives can be right and wrong. Sometimes it is just about power – which narratives are stronger than others? It seems vital who is first, how they spread etc. What are the decisive factors to determine the dominant narrative in the end? What makes a narrative powerful? Can we understand these under the context of an ecology of narratives?
– We sometimes believe counter narratives as fact. Sometimes we know they are counter factual, but they may retain function so we keep using them. There is a large literature in social psychology on debunking myths. There are other bodies of work that we should be paying attention to from a range of diverse backgrounds.
– There is a danger of self-fulfilling narratives: Narratives can create models and then those models confirm the narratives (c.f. economics).
– If there is a feedback process (empirical testing process) then you can keep calibrating. If you don’t do this, you might end up in a dead end. What are the mechanisms for doing this?
– A good problem analysis attempts to capture multiple narratives about what the problem is. At what point do you decide when you have collected enough narratives?
– Cartoons in newspapers: fascinating to see how some people can condense often many narratives into one picture.
– It is not always about the content. You may pick up that people are concerned about something which they are not aware of by understanding their narrative.
– There is a useful distinction to make between creating narratives and analysing narratives. In analysing narratives, the role of ICT is quite clear. How can ICT be used in the process of creating narratives? Digital artwork, ICT environments where rich narratives can emerge (e.g. YouTube), visualisation.

Leave a Reply

Your email address will not be published. Required fields are marked *