lazykoblog   occasional musings about knowledge organization

I’m reading a relatively new book (Nate Silver The Signal and the Noise: Why So Many Predictions Fail–but Some Don’t. Penguin, 2012). It’s a good book, a good read and informative, and I can recommend it to everybody reading this blog, although it isn’t particularly germane to KO alone. However, the breadth of the meta-analysis that Silver brings to bear is truly impressive. I wish we had more work like that in KO. Hjørland’s “Theory and Meta-theory of Information Science: A New Interpretation” was published in Journal of Documentation in 1998 but very few scholars in KO have ever taken up the methodology. The virtue of meta-analysis is meta-theory, which is to say theory that emerges from a broad view of what might otherwise have seemed unrelated findings. In a domain based on the concept, in which we don’t seem to agree on what a concept is, I’d say we ought to avail ourselves of this technique more often.

In “Epistemology of Domain Analysis” I tried, really I did, whether it looks that way or not (this is chapter 6 in Cultural Frames of Knowledge, ed. by me and Hur-Li Lee, Ergon 2012, launched so dramatically at the opening of ISKO 12 in Mysore!). I used a synthesis of prior writing to propose a definition of “domain”–this was not itself meta-analysis, merely typical synthesis. Then I used the definition to analyze all of the formal domain-analytic papers I could find from KO, especially if they seemed to be in response to Hjørland and Albrechtsen’s 1995 “Toward a New Horizon in Information Science: Domain-Analysis” (JASIS 46 (1995):400-425). (Apparently I didn’t count the papers, I can’t now find their number, although I remember it to be around 70 at one point–perhaps I narrowed it in the writing … hmmm.) Of course, I was looking for epistemology, so meta-theory doesn’t emerge explicitly, except in the empirical alignment of domain-analysis. Oh well.

I have written a few other meta-analyses, probably the most useful is “A Meta-Analysis of Instantiation as a Phenomenon of Information Objects” in Culture del testo e del documento 9 (2008): no. 25: 5-25. My current students in Comparative Bibliography have just analyzed a dozen vastly different instantiation networks, from anarchist pamphlets to legal decisions to French symbolist poetry. The theory of instantiation arises over and over in the empirical demonstration of the presence of absence of canonicity on the evolution of an instantiation network. That’s fascinating, and it’s fascinating to watch it develop more or less along the lines I predicted in that paper.

The other thought I had as I began reading Silver’s book was that we almost never engage in prediction in KO. I suppose that’s because so little of our research is directly empirical. It seems we look either at the present or the past, but rarely attempt to suggest what might emerge. There is enough domain-analytical work to begin to suggest at least hypotheses about the development of domains, enough to suggest (for example) that an emergent ontology uncovered in a CWA analysis in a given workspace might evolve in certain ways given the presence or absence of certain conditions. I won’t go further here, I just want to suggest a direction for research in our domain.