Desperately seeking the cyber shelf

By Mark Field

I owe Bob Bater big time. The information architecture specialist posted a note to the taxonomy list on jiscmail (www.jiscmail.ac.uk/taxonomy) citing an excellent essay by Clay Shirky. If you want to have a little read it’s at http://shirky.com/writings/ontology_overrated.html.

Any old how, Clay proposes that models of information structure based on predefined frameworks – he talks about ‘categorisation schemes’ – are not appropriate models for describing the web. He instances Dewey, Yahoo! and current file system practice, so, to be fair, he’s not picking on librarians, which is a shame because it’s really good fun.

I was very struck by one of his assertions: “There is no shelf”. It’s true that traditional libraries have shelves and that in our current concept of the web space there is no shelf, no way of saying that there is a web place equivalent of where all the Wild West books are kept.

Clay, Clay, Clay. There is always the shelf. Our minds are one flat surface on which we dump stuff, making sense of it all by piling things in groups that work together for us. But, we frequently pile unrelated stuff together because we are imperfect makers of piles of things; but then again, the shelf is scrunched up like the extra seven dimensions in string theory, just so it can fit inside our brains. Or maybe it is a singularity in which everything is just heaped together, like a large shelf in a garage.

The thing about the shelf is that to visualise it, while trying to visualise a truthful representation of a hyperlinked model of consciousness without the sort of CGI that George Lucas has access to, is a tad difficult.

Guy Kawasaki’s incredibly groovy information space navigation tool that I fooled around with on my old Apple LC 475, in 1995, was fun but pointless because there were no bounds. There was potentially endless connectivity, but connectivity for connectivity’s sake is autism.

All good visual navigation tools have bounds. Even if it’s only a receding horizontal surface, which gives you several immediate gauges: distance from self, height above surface or direction. Each of these can be used to convey some quality or measure, and that’s before we add in colour. And I’m only really looking at visualisation here – all of the means of supplying addresses, all the identifiers, naming systems and location systems are implicitly hierarchical and complex. And often imperfect.

Any hierarchy that I can think of can be physically expressed as a set of shelves: maybe not the super-complex twiggy bush that is the model of relationships between species, but maybe if we had a shed big enough...

The point is that we have always strived to find simple and comprehensive ways of making sense of the world, and then accepted compromise because of its complexity, or more probably, complicatedness. Clay has spotted that the Dewey Decimal System is rooted in 19th century American protestant academia. Like, we didn’t know that? Most modern atlases still present a world made for 19th century navigators. I use The Times Atlas when I need an atlas and it still works. I follow links from blog to blog to group to wiki to blog when I need other people’s thoughts. It’s different. It’s the same. It’s just me exploring my world.

Mark Field is group leader of library services, Dstl Knowledge Services (part of the ministry of defence). The views expressed in this article are the views of the author and do not necessarily reflect the views or policies of DSTL Knolwedge Services.