An Abstract Concept
Editor • ONLINE
Leafing through some old issues of ONLINE, I came across this sentence in a letter to the editor in the May 1983 issue: “Full text has its merits, but it is basically a gimmick.” The librarian wrote her letter to the editor to protest the distribution deal that gave LexisNexis (then Mead Data Central) exclusive distribution rights to the New York Times Information Bank (now INFOBK in both the NEWS and INFOBK Libraries in LexisNexis). She went on to praise the “solid, intelligent, systematic indexing” of the abstract database, which is what she believed was missing from LexisNexis.
Two things stand out—the assertion that full text is a “gimmick” and the belief in the power of quality indexing. Advances in search technology and the near universal expectation that web search engines search the entire text of webpages removed full text search from gimmickdom. The corollary that abstracts were unnecessary artifacts of a bygone age enjoyed several years of popularity. Abstracting and indexing, however, are experiencing a rebound in perceived value.
The problem with full text today is that there’s too much of it. Search engines claim to return millions of results to simple queries. Full text could mean a 10,000-word scholarly article, a corporate annual report filled with numeric data, a patent with diagrams, a blog post of any length, or a 140-character tweet.
We also now confront the most basic of basic issues. What is full text? Does it include the comments to a newspaper article or only the bylined article? Does it include supplementary data to a scientific paper that is available on the web but not in the print journal?
For some years, popular wisdom held that abstracts were anachronistic, no longer needed given the advances in search technology. Most information professionals disagreed. Now we are seeing a revival in interest in abstracts. I speculated at first that the Google snippet was responsible. Then I realized the snippet was more of a KWIC (keyword in context) entity than a carefully written abstract.
Information overload sparked the abstract revival. At the Academic Publishing in Europe conference held in Berlin in January, one speaker said that a working scientist could easily be bombarded with thousands of relevant scientific articles per month. Nobody can read that many. What was the speaker’s solution? The abstract. You can scan the abstracts and select from them the papers that you want to read in full text. Much as I like abstracts, I’m not sure anyone is going to scan hundreds every day.
Abstracts are one antidote to information overload. Indexing is another. Major information aggregators—EBSCO, Factiva, LexisNexis, ProQuest—have invested significant resources in their indexing and taxonomies. Today, these taxonomies are even more solid, intelligent, and systematic than they were in 1983. The value added by well-written abstracts and appropriate thesauri are more important than ever. Understanding their use in our research and applying them in production of information is a vital information professional skill.