Unearthing Archaeology: A Study of the Recent Coverage of Selected English-Language Archaeology Journals by Multi-Subject Indexes and by Anthropological Literature

Abstract

Librarians, faculty, and professional researchers, and students already encounter difficulties in locating journal articles for the field of archaeology, yet, in the current budgetary climate, librarians needing to reduce subscription costs may be tempted to cancel smaller, discipline-specific indexes in favor of large multi-subject indexes with broad coverage. This study examines and compares the coverage provided to 208 archaeology and archaeology-related journals and magazines by six multi-subject indexes and by anthropology's primary index, Anthropological Literature, over a twenty year period (1988–2007).

KEYWORDS:

• archaeology
• indexes
• databases
• evaluation
• comparison

Notes

1. For ease of discussion and to avoid tedious repetition, the authors will be using “index,” “database,” and their grammatical variants synonymously throughout.

2. That anthropologists desire a comprehensive database that includes current references, abstracts, the full texts of articles online, pictorial materials, and maps was raised earlier by Hartmann in his study of anthropologists’ information needs and information seeking practices (1995).

3. Unfortunately, that AnthroSource has not yet met anthropologists’ desire for an easily searchable and coherent universal portal can be deduced from Nardi et al.'s own article (surveyed anthropologists’ expressed desire for AnthroSource to cover “publications beyond AAA publications” [under “Expanded repository”]) and a later reviewer's complaints about its falling “short when it comes to depth of content” (Wheeler 2005, 36). Another reviewer further characterized the database's centaur-like arrangement with JSTOR as being potentially “maddening” (LaGuardia 2006, 28).

4. Those inclined to quibble may well note that IngentaConnect is not an index or database in the same sense that Academic Search Premier or Web of Science are: the content provided via IngentaConnect is not selected by indexers; rather, Ingenta offers a platform for content that is provided by a range of publishers (Ingenta 2009a). However, from the perspective of its users, this point is largely academic. Certainly, IngentaConnect would meet the definitional criteria for “database” employed by the Chicago Manual of Style (University of Chicago Press 2003, 753).

5. As with these prior studies, the journal Celestinesca (ISSN: 0147–3085) was not included because the authors could not see what connection, if any, that it had to archaeology (Tyler et al. 2006a, 25). Additionally, the Archaeological Conservancy's American Archaeology (ISSN: 1093–8400), Dushkin/McGraw Hill's Annual Editions. Archaeology (ISSN: 1092–4760), and the British School of Archaeology in Jerusalem's Report and Accounts were cut from this study for not having appropriate content. The North West Archaeological and Historical Society's Templemore was disallowed because the authors were unable to discover any evidence of its having been published after 1987. Mesolithic Miscellany, which appeared as a title from the United States in the previous studies, has been reassigned to the UKI subgroup as a result of its current association with the University of York (Milner 2007).

6. Of course, this study's methodology would, if applied to article-level titles rather than to journal-level titles, provide an even more accurate picture of how well the selected databases covered the field of archaeology, but a number of the association, society, state, and county journals selected were so poorly indexed and/or narrowly distributed that this more accurate approach, if pursued, would have quickly proven itself to be logistically unfeasible. Those hoping for a somewhat better sense of the selective nature of the databases’ indexing practices should consult the individual journals’ listings in the appendix.

7. Those interested in similar studies conducted for other social science disciplines and topics are invited to peruse the review of literature in Tyler et al.'s (2008) article, “EBSCO's Communication & Mass Media Complete: An appreciable improvement over previous communication studies indexing?”. Those more generally interested in the checklist method of coverage study should read Thomas E. Nisonger's (2008) “Use of the checklist method for content evaluation of full-text databases: An investigation of two databases based on citations from two journals.”

8. As noted, eLibrary is not intended for academic researchers, and so the authors had no great expectations for its performance and would very much caution against employing this study to influence subscription decisions for public and/or elementary, middle, and high school libraries. eLibrary was included in this study primarily because our institutions do provide academic researchers with access to it and because the authors hoped to obtain some sense of its utility for our anthropology students and faculty.

9. There is, of course, pre-1990 Web of Science coverage, but as the database is prohibitively expensive, the authors were not able to include coverage from the backfiles in this study.

10. In lay terms, Bradford's Law describes a bibliometric regularity wherein, for a given field of study in the sciences, the literature for that field may be divided into three roughly equal “zones” of published articles that have been produced by three groups of journals of descending productivity, with the number of journals in the groups increasing dramatically but proportionally from one group to the next (Bookstein 1990; Bradford 1934; Diodata 1994; Wallace 1987). Bradford expressed the ratio between the nucleus (or “core) of highly productive journals and the peripheral groupings as follows: 1:n:n² (86). As noted by Bookstein, Bradford's Law is “often used to refer to citations received by journals rather than actual articles appearing in them” (370). Pareto's Law, which sometimes is referred to more commonly as the “80/20 rule” and which was developed by Vilfredo Pareto to describe the distribution of wealth and income in Italy, is popularly used to describe unequal distributions in which roughly 20% of a given population produces 80% of the population's effect and the other 80% of the population produces just 20% of the effect (Bookstein 1990; Diodata 1994; Pareto 1971).

11. The term “Matthew effect” refers to rich-get-richer/poor-get-poorer types of advantage processes and, according to Merton, alludes to a passage from the Gospel according to St. Matthew (Diodata 1994; Merton 1968).