Architects, Social Networks and Hypertext

28th September 2004 @ 4:03 pm
Web, Architecture, Design, Thoughts, Social-Networks, Graph-Theory, Hypertext and Space-Syntax
I've recently re-read Christopher Alexander's essay A City is not a Tree (found via City Of Sound) as research for a lecture I'm giving which will tentatively be titled Turning Architects into Programmers (or perhaps less aggressively, Programming for Architects). The first time I read it, in relation to a social network visualisation project, it triggered a brief exchange with Alasdair Turner, a lecturer in architectural computing at UCL, who was lecturing me in Methods of Synthetic Construction 2 from the Bartlett's MSc Virtual Environments at the time.

As part of the course, students were asked to visualise the social networks existing between past and present course staff and students, with an emphasis on the alumni database which was becoming increasingly difficult for one person to carry around in their head. The alumni visualisation was difficult due to incomplete information, but because it has always been online sites like the Wayback Machine at archive.org helped a little. Questioning the staff from the course also helped, though it became apparent that not everyone agreed how long they had been involved - some people even got their own involvement wrong!. It's worth pointing out that effective and general purpose social network visualisation is incredibly rare (please correct me if I'm wrong).

A lot of the social network visualisations started out by assuming that the social network of the course was a tree, i.e. that each person was a 'child' of a year group. I also did this, but I did make an attempt to bridge the gaps between years by developing a way to show that faculty members and part-time students are members of more than one year group. This was designed to emphasise the continuity provided by the course team, and also to illustrate how weak the ties were between year groups.

Christopher Alexander's essay is not just relevant to architecture. He is mainly talking about hierarchical trees, which Dan Hill notes are "the technical and experiential structure of most sites on the web." For me, the key point we can draw from Alexander is that despite the intuitive manner with which we can arrange and consider data in tree-form, these forms don't occur when things (in this case towns and cities) develop organically. It could be for this reason that the original pioneer of 'space syntax' methods Bill Hillier says, "I wouldn't design a city … I'd grow one."

As Alasdair pointed out to me when I first read the article, it is unfortunate that the web has come to be dominated by hierarchical trees, when the original concept of hypertext and http was about navigating through complex networks. (Note to self: Alasdair also mentioned the post-structuralists and the notion of Finnegan's Wake as the first 'hypertext' book.)

Alasdair was right, of course, that the original hypertext aim was not to have hierarchies of documents, but to cross-reference and interlink to your heart's content. Hence "world wide web", and not "world wide tree". This distinction is explicit in Tim Berners-Lee's initial hypertext proposal for CERN (for the uninitiated, this marks the birth of the world wide web). The brilliant thing here is that Berners-Lee actually begins by describing the web of social contact and collaboration which transcended CERN's organisational hierarchy in the late eighties.

Quote from Tim Berners-Lee's "Information Management: A Proposal" follows, apologies for length but it all seems relevant.

"CERN is a wonderful organisation. It involves several thousand people, many of them very creative, all working toward common goals. Although they are nominally organised into a hierarchical management structure, this does not constrain the way people will communicate, and share information, equipment and software across groups.

The actual observed working structure of the organisation is a multiply connected "web" whose interconnections evolve with time. In this environment, a new person arriving, or someone taking on a new task, is normally given a few hints as to who would be useful people to talk to. Information about what facilities exist and how to find out about them travels in the corridor gossip and occasional newsletters, and the details about what is required to be done spread in a similar way. All things considered, the result is remarkably successful, despite occasional misunderstandings and duplicated effort.

A problem, however, is the high turnover of people. When two years is a typical length of stay, information is constantly being lost. The introduction of the new people demands a fair amount of their time and that of others before they have any idea of what goes on. The technical details of past projects are sometimes lost forever, or only recovered after a detective investigation in an emergency. Often, the information has been recorded, it just cannot be found."

The sad thing of course, apart from the increasingly hierarchical structuring of large sites, is that the web as we know it suffers from a high turnover of documents, much as Berners-Lee described a high turnover of people at CERN. As I pointed out in our crit session after the project, this problem afflicts the MSc too, since by design there is a yearly turnover of 90% of the people involved.

Back to alumni databases and social networks then, and to the defense of the tree, for a moment. I actually think that from an egocentric point of view a social network is most usefully considered a tree. That is, if I know two people already, it is of little consequence that they know each other. The only connections that matter to me are the ones which form the shortest paths between people I already know, and the people I want to know next. This is social networking in order to get ahead in business, or to make new friends, I admit.

On the other hand, considering the loops inside of who knows who, as well as the tree of who knows me, might allow a certain amount of insight to be gained into the nature of interactions across the whole social network. The interconnectedness of it all is what everyone was talking about in the crit, and what we're all stuck with trying to visualise and interpret in a meaningful way. Do self-organising structures hold the answer? I would argue not, but I'll leave that for another time.

I hope that we can be rid of the hierarchical straight-jacket that much of the web is in right now, and I think a combination of search engines and weblogs will get us there in the end (not to mention tagging systems like del.icio.us and Flickr which have emerged strongly since I first wrote this). Weblogs aren't just trendy, it's practically their whole raison d'etre to link and be linked, and we're seeing big businesses cotton on to this fact in a big way. If everyone had one, and used it (more than I use mine!), then maybe we would be able to map out social networks as we go, instead of trying to construct them after the fact.