Book Review

De Bellis, Nicola. Bibliometrics and Citation Analysis: From the Science Citation Index to Cybermetrics; Scarecrow Press: Lanham, MD, 2009, $55.00 (Paperback). 394 pp. ISBN: 978-0-8108-6713-0.

This rather massive monograph is the outgrowth of a research project on a related topic and is an English translation, encouraged by Eugene Garfield, of the Italian original. The history and philosophy behind citation indexing and other bibliometric measures is documented in chapters one and three. The empirical basis, the literary antecedents, and comparisons with concept indexing and other full text retrieval, are described in chapter two, including some discussion of Salton’s work. The work of the giants in these portions of the information industry, Bernal, Merton, Price, Garfield, and Small, are documented in detail. The mathematics of bibliometrics are described in chapter four including skewness, Lotka’s Law, Bradford’s Law, Zipf's Law, and the work of Mandelbrot. Chapter five is titled "Maps and Paradigms" and discusses involvement of bibliographic citation with the history and sociology of science using co-citation analysis and other methods. Chapter six, titled “Impact Factor and the Evaluation of Scientists: Bibliographic Citation at the Service of Science Policy and Management”, probably has the most relevance to the most scientists. The various metrics are discussed including the Hirsch (h) index. Chapter seven, with the intriguing title, “On the Shoulders of Dwarfs: Citation as a Rhetorical Device …”, describes reasons for citation, professed and actual. Chapter eight evolves the discussion into cybermetrics including the involvement of citation or linking in the performance of search engines.

Errors and omissions do occur. Reference 5 for the introduction is missing. Recall and relevance are only discussed briefly in the context of "improvement" of results from concept indexing (manual) and retrieval by means of Salton’s geometric machine indexing and other full text indexing methods. No mention is made in chapter seven (or apparently elsewhere) of the relevance of citation retrieval since it should be commonly known among searchers that authors don’t always cite other references for the same reasons that the searcher is interested in. Due to the multitude of topics and concepts that can appear in a single article, many of us searchers can cite instances where a citation was made for a non-relevant concept. Curiously, the discussion of citation searching in patents is the last section of chapter six and has no discussion of the validity of the bibliometric value of citations in patents. The work of Narin is described and referenced, but that of critics of the method, including Edlyn Simmons, Stu Kaback, and Nancy Lambert, is missing. The existence of other uses of citation indexing and searching, e.g. in the CA file on versions of STN, are not mentioned.

In the Conclusions, the author provides an either/or summation of the evaluation controversy. Either you believe that citations are “Mertonian” or you don’t. If the former, researchers, organization, and journals can be evaluated. If you don’t, none of the evaluations can be made and the Citation Index itself may not be of value. This reviewer instead takes an intermediate attitude. Bibliometric evaluations can be a valuable supplement in a larger, more personal evaluation scheme. As for searching, use of citation indexes is a valuable supplement to other methods of searching (index, full text, etc.) and all methods should be used, none exclusively. This book is of interest to those interested or researching in the fields of information science or history of science. Chapter six should be made available to the management of academic and other organizations that use citation analysis for personnel evaluation.

Robert E. Buntrock
Buntrock Associates
16 Willow Drive
Orono, ME 04473

Software to publish news website