Libraries & the Cultural Record

Memory, not only in the sciences, is Geoffrey C. Bowker's touchstone in his far-ranging consideration of memory practices of the past two hundred years. Memory requires classifications—slots into which things to be memorized can be put for storage, whether in an individual mind, a shared cultural archive, or an electronic database. The present classifies and maintains its information with an eye to what the future might need, so memory is deeply involved with ways of knowing and with perception of the things to be remembered and of self and with time. Time can be and has been conceived of in a variety of ways—the synchronic and diachronic do not exhaust the possibilities. Technologies for memory shape storage, maintenance, and recall. Recall is not the inevitable or even desirable end of stored memory; forgetting can be the goal, enabled by storage. [End Page 105]

It is not clear who the audience is for this book. Bowker's language and style of organization—the accumulation of abstraction deployed in place of reasoned argument—does not show evidence of anyone questioning the presentation of a great volume of data that refuses to become concrete. There are interesting and thoughtful ideas here, but they are not clearly presented or argued.

Bowker returns to memory as he gathers a vast number of ideas, methods, and data that he would join together as being pertinent. Here, in a loose order from abstract to concrete— (loose in part because Bowker has not created a structure of argument or hierarchy of importance and in part because the relationships between items exist in context and change with changing context)—is some of the book's stuff: time, philosophical inquiry, the human mind, knowledge, structure of knowledge, science, scientific practice, technology, culture, the synchronic and diachronic, the human genome project, databases, distributed databases, classification systems, "Geology and the 1930s," "Cybernetics and the 1960s," "Biodiversity and the 2000s" (these three are chapter subtitles that promise but do not deliver an orderly presentation of information), the necessity for forgetting, clocks based on mitochondrial DNA or the sedimentary record or biblical generations, Linnaeus's limit on the number of botanical genera—matters general and particular, hypothetical and proven. About a book so filled with detail it may be perverse to complain that there is too much abstraction, too little that allows the reader to think along with the book, but one difficulty of reading it is that the examples themselves are abstract.

Bowker, for instance, supports his statement about scientific arguments and stored scientific data—that in the past twenty years there has been "a partial disarticulation of these two features of scientific work"—by saying that the human genome databank, among others, might be seen "as archetypical of a new kind of science in which the database is an end in itself." Arguments made from the data, he says, are separate matters (119). The reader has to take these things on fait h. Is Bowker saying only that scientific projects have become so complex in the past twenty years that specialists will work on only one aspect of the project? If this is what he is saying, has not there always been division of interest as well as labor on large scientific projects? Is it simply the use of massive databases that makes the past twenty years special? Because Bowker does not seem to have imagined that a reader might not understand what he is writing about or, if understanding, object, he does not make any effort to support or convince, at least not at the point he makes his claims. He simply continues with more assertion, not slowing down for the reader by showing the steps in his thinking or by anticipating where his ideas might be challenged and then meeting the challenge. This is frustrating, for the book may actually provide pertinent evidence, but it is dispersed, and the reader is left to make the connections with the aid of a simplified and...