From Thomas Mann's book A Guide to Library Research Methods Chapter 9, "Computer Searches":
Computer searches thus have both advantages and disadvantages. To use them intelligently you must recognize their limitations as well as their strengths. The important thing to remember is that they are only one weapon available to the researcher; they are not the whole arsenal. A thorough review of the literature of any subject requires a combination of the approaches discussed in this book.
The versatility of computer searches is so dazzling that a large humber of researchers are failing to note or heed their limitations. And, contrary to the popular saying, what you don't know can hurt you. I have seen this problem repeatedly, particularly with graduate students who want to do a literature review in preparation for writing a dissertation. ...
What is just as bad is that professors who direct doctoral dissertations allow computer searches to pass for complete literature reviews --- for the professors are usually just as naive about their limitations as the students are.
What this amounts to can only be described as cultural lobotomy on a grand scale. When a significant percentage of our most educated people (prospective Ph.D.s) relies almost exclusively on computer searches for "in-depth" research, then we are fostering the growth of an intellectual system with very shallow roots. Since so much of the written memory of humanity before the 1970s is not in the computer in the first place --- or is only superficially indexed --- it is likely to be ignored by immature scholars if it isn't as easily retrievable as the more recent material. The older material --- especially the older journal articles --- "does not compute"; and to many graduate students this tends as a practical matter to mean "therefore it is not important." There tends to be a similar neglect of certain forms of literature, especially single-author books, because machines more readily retrieve journal articles and research reports. A moment's reflection will indicate that the computers are no better than the material that is loaded into them; and yet a surprising number of researchers expect them to be omniscient.
Saturday, June 30, 2001 at 06:34:10 (EDT) = Datetag20010630