I spent most of last week at the 2013 Web Science Conference. Although I’ve been working on my PhD in Web Science for about 18 months (now part-time, as I work on the Exploring the Emerging Impacts of Open Data in Developing Countries project with the Web Foundation), it was the first big Web Science centred event I’ve made it along to. And I left feeling quite conflicted about how to respond to the event (I can tell this is the case, as it’s taken me about 45 minutes to write this sentence). As I started writing up reflections for my research journal on the Eurostar back home yesterday I came to find three broad themes at the route of my frustrations with the event. Firstly, the lack (or divergent nature) of clear shared framing of what Web Science is about. Secondly, the conference format itself, and the lack of spaces for dialogue. And thirdly, around which both of these pivot, questions over the value of discipline building as opposed to developing a problem-centred field. Rather than my initial fairly negative reflections, I’ve tried in the post below to work through some a set of constructively critical reflections that explore these points.
Two kinds of web science?
From a problem-centred perspective, there were a number of different kinds of papers offered at WebSci 2013.
These might be divided between what I will call ‘inward‘ and ‘outward‘ Web Science. The first set, papers with ‘inward’ focus, were seeking to deploy insights from one or more disciplines to either describe, or suggest interventions in, ‘The Web’ as the focal point of their study. Those with an ‘outward’ focus sought to draw on content or concepts from ‘The Web’ to illuminate a wider societal phenomena or issues, which may (or may not) be intertwined with the web, but where the description, or intervention, offered in the paper is primarily directed to some existing communities of academics and actors who do not take the web as their primary object. The former set deal in ‘problems’ with, or of, the Web; the latter with a wide range of theoretical and practical problems articulated in other disciplines.
The notion of ‘The Web’ in use here (hence the inverted commas above, which I’ll now drop), is a socio-technical one. Insight from, problems of, and interventions in, the web, may be identified at any number of levels: from the most general or wide reaching components that include technical standards and infrastructures, laws and regulations, and strong norms – through intermediate levels that exert wide-reaching influence, such as major content platforms or net cultures – to localised and specific aspects of the Web such as individual websites or particular online practices and communities.
While the ‘butterfly diagram’ of Web Science offers one route to see the range of disciplines involved in Web Science, it fails to capture these twofold distinctions: between the use of disciplines to address the Web as the object of intervention, and the use of phenomena of the Web as source of data to illuminate other objects of study; and between the different aspects or levels of the web in focus. Inwards and outward web science may well be complementary – but I suspect being clearer on the distinction between the two may help in shaping productive Web Science conferences and debates. It may be that these two kinds of web science have different needs of a conference – on the one hand having the opportunity to bring in diverse insights and strategies for better conceptualising, describing, and realising, visions of a pro-human web; and on the other being nourished, challenged and inspired to take methods, approaches and insights from web science research out to a wide range of fields and fora.
Web building? Discipline building? Field building? Building capacity?
There are big challenges ahead for the web. Anil Dash captured some of them in his recent Berkman talk ‘The Web We Lost’, and Tim Wu’s warnings about how past technologies of communication have generally had their radical and open potential blunted, should give us significant cause for concern. The founding myth Wendy Hall offers for Web Science, of a conversation in 2005/6(?) about why the Web of Data had not taken off, and the need to bring different discipline in to address that question, highlights also how much of the original vision for the web remains unrealised. And still, the majority of the worlds population are neither online, nor using an open web – nor is it inevitable that an open web will win out over proprietary mobile and computing platforms in future. Given the web clearly has a profound role in the ongoing transformation on our politics, culture and economics, there is a clear need for study that can contribute towards addressing these challenges.
Yet – much of the narrative around Web Science right now seems not to be about the problems that should act as a core for interdisciplinary (or transdisciplinary) research, but about discipline building. On the one hand, energy expended discipline building may considerably distract from energy spent on shared problem solving – as very different disciplines grapple with intercultural communication in the absence of articulated shared problems or objects of study. On the other hand, discipline building may be a long-term strategy, to ensure that, within the deeply flawed incentive structures of contemporary academia there will be space for the broad based study of the web, and that choosing to engage in web studies across conventional disciplinary boundaries is not a career risk.
Personally, I would rather attack, or find ways to work around, the broken incentive structures of academia, instead of trying to work with the tension of shared publish-or-perish, yet different publishing models, of different disciplines. However, if Web Science does want to discipline build, then is most likely needs to become a lot more creative and bold in formulating the field, and in using a variety of formats and facilitation techniques to construct future conferences.
Assorted ideas on hacking the conference format
I often feel like I spend my life in workshops and conferences, although as it turns out this was my first experience of a large scale academic conference. Hosted as part of the European Computing Research Congress alongside a number of other conferences, and also co-located (at least for it’s first day) with the much larger CHI conference, WebSci had a very formalised agenda – of keynotes, paper sessions, a poster session and a panel. Even the workshops before the main conference started were driven by paper presentations, with less than 20% of the time given over to discussions ‘from the floor’, and no small group discussions. It may be simply the experience of culture-clash that drives the following reflections, but I’m pretty sure a lot of the potential of Web Sci was lost as a result of following disciplinary defaults, rather than taking a critical approach to ‘hack’ the academic system so that it better supports the creating of a new discipline, or interdisciplinary field building.
WebSci did seek to introduce a few ‘innovations’ this year: firstly by asking for a number of papers to be presented Pecha Kucha style (20 slides, 20 seconds per slide – 6m 40 overall); and secondly, disassociating the mode of paper submission from the mode of presentation, so extended abstracts, short and long paper submissions all ended up variously as posters, short, long and Pecha Kucha talks – depending on the opinion of the reviewers. However, in the main, the formats used seemed ill-adapted to fostering interdisciplinary dialogue. Little attention was paid to creating physical environments conducive to dialogue, with presenters setting up behind the presenters who had just been on – making question asking and answering an awkward experience – and a long-thin room creating distance between presenters and audience. And somewhat surprisingly, there was almost no use of the web to enhance the conference experience – with only a quiet twitter backchannel – and papers provided on USB stick, rather than accessible as social and sharable objects on the web (although I understand they will become available later when proceedings of the conference go online).
I expect that some of these format decisions can be explained by some of the disciplinary defaults, and pressures, mentioned above. In some disciplines, the conference functions as the primary publication venue for work, and conference publications are a valuable part of an academic CV; in others, giving a paper is a chance to test out ideas that will later be written up and published in a journal. Within a disciplinary conference, presenting a paper might be a chance for people with a shared knowledge based and set of conceptual frameworks to pick out particular parts of the argument in need or more development; but in an interdisciplinary conference the role of post-presentation Q&As requires more thought. So, what can Web Sci do to take seriously the different incentives structures of different disciplines and to work with these, whilst still building a coherent and credible conference that pushes forward the study of the web.
Some assorted ideas to consider:
- Asking authors what they need when the submit papers – as well as asking reviewers to offer an assessment as to whether a submission is best dealt with as a short or long talk, or a poster – ask the authors what they are after from submitting a paper. If it’s a chance to test out ideas, then scheduling such papers in panel sessions may be appropriate. If an author needs to give the paper to be able to participate in the conference then, if it is of the standard required, it can be scheduled in that way.
- Open up peer-review – reviewing papers for an interdisciplinary conference is hard work. One move would be to move to much more open review – giving authors the option to place their papers into an online open review process. Conference chairs would still need to ensure each paper had reviews from a range of perspectives, but by opening both the original paper when submitted, and the reviews, negative review dynamics could be disrupted. I don’t suggest making open review the only option, but I would love to see WebSci experiment with making it the default, with the fall-back option of submitting papers through a conventional process for those whose disciplines or fields react badly to early publication of papers.
- Provide platforms for pre-conference collaboration. There must be many opportunities to make better use of the web in shaping the conference – from opening up spaces where people can discuss paper, panel and workshop ideas long in advance of submission deadlines, through to using the web to support collaborative work on scheduling papers. Curating diverse papers can be tricky, but there may be ways to transfer the challenge of spotting relationships with other works onto authors, which may create the opportunity for a series of presentations followed by a panel, rather than the token space for a brief Q&A after each paper. The use of the web for shaping the conference could also be extended to the event itself (if there was working WiFi… alas not always the case this week), inviting input from delegates on which papers they would like to use plenary session time to discuss, or collecting questions on each paper so that even if not all asked in speech, authors could respond later to interlocutors.
- Mixing in some multi-track time – Whilst in the closing of WebSci it was argued that the single track nature of the conference was important for building the community, I fear in practice it means there are few opportunities for sub-fields within Web Science to develop, which would potentially strengthen the community as a whole. Mixing single and multi-track elements of the program (e.g. one day of multiple tracks) with smaller paper sessions may well help. It may be even be possible to see these tracks as ‘feeders’ into plenary sessions, with rapporteurs from the tracks bringing their different insights together at the end of the day.
- Make use of open spaces. It’s something I would suggest for virtually any conference – but I think Open Space methods are incredibly valuable in providing a light structure in which conversations can be fostered and taken forward. Adding a short open-space slot to the program, even if only over an early extended lunch break (but not tacked on the end as a last-day add-on, as this rarely works as well), would I think provide a lot more opportunities for dialogue building around the conference.
In closing
The reflections in this post are initial thoughts – merely offered as I try and find constructive reflections that might contribute to thinking about managing some of the tensions within Web Science as an evolving academic space. Whether working within the constraints of such academic structures, largely forged in pre-web eras of information scarcity, is an important part of delivering social change or not will have to remain the subject for a future blog post.
This is a thoughtful and helpful summary of key issues Tim. Will have to go away and reflect on this.
Joseph
Thank you for this thoughtful post, which I just ran across. A few comments, in no particular order:
I’m not quite sure what you mean by ‘discipline building’ — would you be able to elaborate?
I can understand what you mean about the lack of time for discussions, although personally I didn’t feel that discussions were stifled (I guess I’m more used to this style of conference!). Certainly the room was an odd shape… I don’t know if you attended any of the workshops but some of those had *crazy* rooms.
I like your ideas about platforms for pre-conference collaboration. Indeed, this was preciesly the goal of Web Art and Science Camp, an event I co-organised years ago, was aimed at engaging with and broadening the WebSci and Hypertext communities. I don’t want to lead you astray from your PhD, but perhaps you could chat to your supervisor about the possibility of a future similar event, or — to be way more lightweight — a WAIS-based thing maybe half a year before the WebSci’14 deadline to throw ideas around. It could go in the regular seminar slot.
What’s the Open Spaces method? Is it related to unconference formats, or birds of a feather sessions?
Last but not least, I understand why you’re calling for multi-track time but I do strongly disagree on that. The single-track thing is essential to WebSci: sub-fields can develop in coffee sessions, lunch breaks, dinners, workshops, paper collaborations and perhaps these Open Spaces or birds-of-a-feather sessions. As soon as the conference has a multi-track format, it will split along disciplinary lines, and be very difficult indeed to reform. Don’t get me wrong, I understand single-track has disadvantages, but I maintain that it’s important. (Not that I’m in any way involved in any of these decisions, I would add!)
You’ve clearly thought deeply about all of this. Thanks for the post!
Thanks Clare for the detailed comments.
To try and respond briefly:
– Building a discipline might be thought of as creating a body of knowledge and methods, and a collection of journals, conferences and other fora, such that the value of activity can, to a greater or lesser extent, be measured in relation to these reference points, and promotions, careers and discussions are structured around these. Disciplines generally seek to reproduce themselves, training people in that same set of tools and skills, and working on problems primarily under that disciplinary identity.
This can help secure funding, scale-up training around needed skill sets, and increase the security of careers for those who choose to be ‘web scientists’ (i.e. avoiding people needing to split their identity – publishing in web science journals on the work they’re interested in, but having to publish back into ‘home disciplines’ to get the academic credit needed for promotion etc.).
I’m personally relatively ambivalent about discipline building: asides from nothing that it can risk creating distractions from problem-centred research, and that in web science any fixing of boundaries for web science as a discipline would necessarily involve a degree of exclusion of some of those currently within the broader ‘field’ of web science.
– By ‘field’ I understand an interdisciplinary space, where people are not only addressing particular problems, but are also contributing to a shared interdisciplinary knowledge base consisting in overlapping collections of knowledge and theory.
– There is more on ‘Open Space’ conference formats here. It is an unConference approach.
– I’m already fairly astray from my PhD
but already talking with other students and staff on possible approaches to support more cross-boundary collaboration ahead of WebSci 2014. In the post I was particularly thinking about digital platforms for pre-conference collaboration – such as, for example, opening up the agenda drafting process (which could be as simple as sharing a google doc through which the agenda is being put together for a 5 day period of comments from the authors and delegates to help put together the best possible sets of papers).
– Whether multi-tracks lead to the conference splitting I think depends on facilitation, chairing and curation: not on the existence of multiple tracks or not. If the ‘community’ has to be ‘forced’ to share as space in order to exist as a community, then we’ve got problems.