Hypertext 2017 and the Festival of Narrative Automata

I spent the last week in Prague at the 28th ACM Conference on Hypertext and Social Media. As usual it was a pleasure to catch up with friends and colleagues, and Prague is a beautiful city. The conference venue used to be a church, and the vaulted ceilings and huge wall frescos certainly added to the event!

This was also the sixth workshop on Narrative and Hypertext, co-chaired by myself and Charlie Hargood, which this year was an interesting mix of position papers and plenary discussion. I presented work on transmedia modelling by my PhD student Ryan Javanshir; transmedia is a huge area that has not really been examined by Hypertext researchers (with ARGs reaching hundreds of thousands of people) so its a great area to be exploring. Another of my students, Callum Spawforth, also presented his work on multi-particiaptory narratives. Callum has been looking at how models of interactivity in games might inspire approaches to interactive narratives, stories which are experienced by more than one reader. Most interestingly for me is that Callum is proposing that you can model this in sculptural hypertext, by making the state space of one reader accessible to another.

Charlie attempted to address a problem that in the workshop plenary was identified as a grand challenge for narrative research, which is how we can evaluate narrative research. Charlie suggested an approach based on analysing the delivery of information in the story (narrative payloads) which is definitely an interesting way to understand how the story functions, and could produce alternative models of how well a narrative works according to the definition of what the information in those payloads actually is. Delivery of the Fabula is the obvious thing to look for, but filtering for elements that signify themes, or are linked with a particular type of reading, could also be enlightening. So a feminist reading of a story is likely to identify a different set of payloads to a marxist one. I am doing some work at the moment in the direction of narrative analytics, and I think building a probabilistic model of how a particular set of payloads are delivered could be an interesting direction.

In the main conference we were lucky enough to again be nominated for the Douglas Engelbart best paper award (which we won last year for our paper on sculptural patterns). This year our paper was ‘Tiree Tales: A Co-operative Inquiry into the Poetics of Location-Based Narrative’ this was an account of our trip to the Isle of Tiree last year, leading to a number of observations about the decisions made by locative authors, and also the realisation that while sculptural patterns can describe the narrative logic well, you need calligraphic patterns to understand the narrative constraints inherent in the landscape.

Claus Atzenbeck was also nominated for his paper on Revisiting Hypertext Infrastructure a great piece of work that coupled his research into spatial hypertext systems with architectural ideas that referred back to the work of the OHSWG (something that I was involved with in the late 90s, so it was great to see it talked about again). Claus’ notion was that component architectures could act as a way to incorporate AI systems into hypertext, and he gave an interesting demo where the AI made suggestions of elements to be included into a spatial hypertext.

The ultimate winner of the award this year was Ujwal Gadiraju for his impressively thorough paper ‘Clarity is a Worthwhile Quality: On the Role of Task Clarity in Microtask Crowdsourcing’ where he examined the clarity of task descriptions on crowdsourcing platforms, and looked at their effect on people’s decisions to participate. Quite a lot of work in the social space seems to lack a firm motivation, so it was good to see that here the analytics was directly useful to inform task uploaders on how best to describe their tasks to maximise take-up. For me Ujwal also wins best Hypertext Hat award!

There were a number of other interesting social media papers, including Erick Elejaldes work (nominated for the Ted Nelson Newcomer award) on ‘The Nature of Real and Perceived Bias in Chilean Media’ where he used various analytical techniques to position the chilean media within a political views framework (choosing one created by Marshall Fritz in 1985). What I thought was interesting was that they seemed to have identified the Overton window, which they then normalised to show (for example) left/right bias. Normalising is an interesting decision, but does show how an organisation (*cough* BBC) might be simultaneously accused of bias by both the left and right. Could it be that in the UK our print media is so right wing, that actually an objectively unbiased organisation appears to the left of the Overton window?

I was a bit disappointed that there was not more work that sat in the intersection of hypertext and social media. For example, another student of mine, Mark Anderson, presented his really nice analysis of transclusion within Wikipedia (or rather the lack of it). Mark’s work really questions whether a huge long lived hypertext like Wikipedia can afford to ignore hypertext mechanisms (such as transclusion) that are designed to make maintenance of the wiki easier. His conclusions are that Wikipedia does include transclusions, but that they are used in a tiny percentage of pages, and seem to be limited to particular sub-communities of authors.

One of my favourite parts of the conference was Mark Bernstein’s Festival of Narrative Automata, a workshop that Mark intends to work up into a new book, where he considers the ways in which various narrative automata (include a wide range of paper based RPGs) work as narrative systems. As a broad review of the state of play it is already interesting. But I think it also raises interesting questions about the potential role of a GM within a digital interactive story, and also role-playing itself, as very few interactive stories (either in the IF or Games space) actually ask the reader to make choices in character.

I was also struck by a quote in Mark’s draft manuscript.

‘Much of the Twine world equates hypertext research generally (and this writer specifically) with The Man; I think this, too, is an unfortunate misperception on their part. But, then, I would’ – Mark Bernstein, Narrative Automata

If nothing else we need to bridge this divide that has grown up between the various communities doing interactive storytelling. Whether we are storytellers, publishers, technologists, critics, or teachers (or all of the above), there is no real reason we cannot learn about this new phase of storytelling together.