Chris Dent

L505 Essay 6



More Structure, More Constraint:

A ramble on job security in LIS


Computerized automated tools provide many of the same advantages and disadvantages for indexing as they do for abstracting. Neither term frequency nor the presence of a term gives a faithful indication that a particular term should be included in an index. In their current state, computers are unable to make the same contextual considerations that a trained reader can make.

A solution used by some publishers is to either have the author or the editor(s) of a document indicate as they work that a particular section should be indexed. In other words, for example, as the author writes, if s/he feels that a particular section deserves to be reachable from the index they can flag it with something along the lines of an “index this” tag. Computerized tools later aggregate and organize these flags. This is certainly better than nothing but it comes with a cost. As Bonura implies, the author has an intention for their work that may be different from the intention of the reader. It is the job of the indexer to bridge that gap, to effectively provide access shortcuts (Bonura, p. 7).

In my essay for session 5 I suggested that a structured approach to document creation and editing would help to allow easier automated abstracting. The same approach could be used to improve indexing. If the document is highly structured, the process of identifying terms, sections and their interrelations could be eased. This can be done, but at what cost? Highly structured documents are not necessarily good things. Take for example the text from Bonura. It is highly structured; terms and sections are pointed out with great rigor. It is also dry, tasteless, impossible to read with any sense of joy or discovery and downright insipid. It is all of the vitamins without any of the sugar. Properly designed computer programs could abstract and index this baby in a flash.

Conscientious authors and readers want to write and read art. Educational material can be written to be both densely informative and elegantly expressive. Therefore, I submit that although structured document technologies may play very important roles in the future of publishing I do not think they will devastate the market for talented editors, abstractors and indexers. The skills that are valuable for those fields are the same skills that are required to metastasize[1] a document into a structured form that will make it usable by both human and computer researchers. When I write the Great American History of the Cupcake my florid prosing shall not be interrupted by mundane thoughts of structure: Someone else can do that--and they can get paid for it.



Bonura, Larry S. The Art of Indexing, New York: Wiley, 1994.

[1] I’m making up this (probably unfortunate but full of connotative deliciousness) usage: of a document: to change the form of, to modify by adding or extending; giving structure by breaking apart: putting in the meta-data and thus causing change.