August 2, 2008 Ben Peters

Who Made the Watchmaker? Dewey and Mendeleev on Information Organization

The note below crudely illustrates a distinction in two modes of information organization predating the twentieth-century invention of automated data management: flexible classification models demanded constant care and intelligent intervention from the outside, while stable models precluded human management except in its original design (the watch-maker).

To Mendeleev’s periodic table of the elements and Dewey’s decimal classification (used in library card catalogs), the first offers a system that is discretely limited but self-contained in its parameters; the second is continuously expandable but bound to some form of external regulation. The first presents a discrete (though theoretically infinite) number of elements arranged by internal characteristics (e.g., atomic weight patterns); the second allows for any number of items to be organized in relationship to each other along a thematically-grouped alpha-numeric line (000 Computer science, 100 Philosophy, etc.).

Whither the human manager? Whence the human user? Once established, Mendeleev’s periodic table is relatively self-regulating and tautologically autonomous: it is so because it says so. To shift the order of elements would require a concomitant shift in the governing principles of the periodic table, which otherwise remains inflexible in its topological order (meaning, no matter what representation one may choose for the elements, they remain in the same relative order); librarians, on the other hand, are free to rearrange the order or place of titles according to their interpretation of the title’s content to pre-given topic categories (000 Computer science, 100 Philosophy, etc.), and they can add new titles between preexisting titles without disrupting the information order. Card catalogs work by relying on an external set of commonly held symbols (numbers and letters) and topics allow items to reordered continuously. Yet the freedom of data reshuffling comes at the cost of the system requiring a governing body external to it (e.g. librarians). These distinctions can be applied across modern society (Deweyian accountants and lawyers interpreting Mendeleevian spreadsheets and codebooks).

In a later note, I intend to push thinking how, if at all, automatic information systems (e.g., databases, brains as circuits, and search engine algorithms) breakdown these two caricatured systems for information organization. The search engine algorithm, for instance, is a mixed-system: it provides a relatively flexible (Deweyian) system of information organization but requires relatively (Mendeleevian) minimal intervention in the system.

This note draws on Glenda Claborne’s Linnaeus, Mendeleev, Dewey, and Ranganathan: What can they tell us today about the organization of information?” A presentation at the 2005 ASIS&T-PNC Annual Meeting, May 14, 2005, Seattle, WA. [See options for viewing presentation].

Comments (5)

Comments are closed.