16 June 2009

The Tables and Chairs of Database Origins

A friend was over from the States this weekend. She is the one who invented boundary objects, and she always gets me thinking. She gets me thinking largely because she is so smart, but also because she puts things in a way that gets you thinking. That is a rare gift.

Anyway, I got to thinking about databases. There is so much rubbish written now about databases, both from the techie side - knowledge as semantic first order logics, metalanguages, and meta-descriptions, and from what we could call media anthropology - database as metaphor for social relationships, surrogates for ontologies, etc. I was also reading Lev Manovich's softbook Software Takes Command, where he talks about the early days of Ivan Sutherland, Ted Nelson, Douglas Engelbart, Alan Kay, Nicholas Negroponte and their personal computer as new media. Rather, the personal computer as a universal media machine. What got me thinking was how my friend talked about databases, and how databases are missing from Lev Manovich's book. She talked about them in relation to ever more global indicators, not as mere data repositories. We can talk about how publishing, writing, TV, video, radio, sound, letter writing, the business ledger, etc. etc. have been transformed into software, and how they have been translated, mixed and extended as new media, as Manovich does. But where are databases? I thought that if word processors are the media machine's typewriter and now publisher, if spreadsheets are its ledgers, email its letters, and now we have TV, film, radio and photography on-line, what is the genealogy of the database?

I do realize that the genealogies for all of these systems are complex, and that what is most important is the transformation from media to media machine to integrated new media, but that is not what I am talking about here. What I am talking about here is where is the genealogy for databases, they seem to be the odd one out.

Some histories place the database within the genealogy of information management, within the history of the Archive and the Library. But this can't be right. Databases are not repositories, though they may now function as repositories, but this certainly is not their heritage. Others place it within command and control systems. Though this certainly was the origin of the word itself, from tabulations coordinating command and control systems, this does not seem right either as it is to specific a locale for a genealogy. It is as though we were to say that John Smith was the first to explicitly acknowledge his name, therefore he is the origin of the Smiths.

They genealogy has to lie with tabulation. It has to be found in the 19th century rise of tabulation, census, and indicators, and, technologically, with the tabulation machines (Herman Hollerith's machines, IBM, etc.). Now I don't have the time, or the space, to explore this here. However, what I would like to ask, as a way forward, is "Why has some centralized, imperial and state instrument like tabulation come to be a dominant application on the Web and underlie the functioning of just about everything?"


  1. I think that the big peak that databases had in the 1970s and 1980s was a consequence of the positivist currents that dominated many other fields in those decades and earlier ones. Computer science let it percolate through and this trend of wanting everything systematised, classified, taxonomised and organised fuelled database theories. Hence relational, OO and the like.
    Makes sense?

  2. I think you are certainly right about the role of the 1970s and 1980s information theory that drove the rise of, in particlar, relational databases, but what I find interesting here is that the use of databases, from the 1960s has been far more diverse, massively more diverse, than the theory would, and could, suggest. In fact, I think, if we were to look at what databases were used for today, we would find very few examples, relatively, of what we would see as a prototypic database. We would find massive numbers of flat file databases doing all sorts of tasks that were not strictly classification or taxonomy. I don't think that the world is necessarily going to be changed by this insight, but it is very interesting.