Page 149 - Becoming Metric Wise
P. 149

140   Becoming Metric-Wise


          5.11 TRI-CITATIONS

          In (Small, 1974) the author defined tri-citation as follows. Let S(X)be
          the set of papers citing document X, S(Y) the set of papers citing docu-
          ment Y and S(Z) the set of papers citing document Z, then the tri-
          citation frequency of documents X, Y, and Z is #(S(X) - S(Y) - S(Z)).
          The relative tri-citation frequency can then be defined as

                                 ð
                                # SðXÞ - SðYÞ - SðZÞÞ
                                                                      (5.11)
                                # SðXÞ , SðYÞ , SðZÞÞ
                                 ð
             Similarly one may consider author tri-citation where S(A) denotes all
          authors citing author A. Marion (2002) and McCain (2009, 2010) applied
          this approach with one specific restriction: they hold one author fixed,
          leading to a specific context (provided by the author who is kept
          unchanged) for the cocitation of the two other ones.



          5.12 HIGHLY-CITED DOCUMENTS BECOME CONCEPT
          SYMBOLS

          Garfield (1970) and Small (1978) have shown that highly-cited articles
          or books become symbols for the idea (or ideas) contained in them. In
          this way the act of citing becomes a form of labelling. The fact that this
          is actually possible is another argument in favor of citation analysis (and
          hence of citation indexes). Small (1978) investigated to which extent
          highly-cited articles in chemistry were cited for the same reason. He
          found many examples of high uniformity (the article was almost always
          cited for the same reason). He also found that such highly-cited articles
          are often included in groups of redundant citations. He concluded that
          such highly-cited articles become symbols for the concept with which
          they are associated: they become concept symbols or concept markers.
          From this investigation we may conclude that references can be seen as
          a part of the symbol language of science. We finally note that if an arti-
          cle and hence its author is always associated with a certain concept, this
          may lead to eponymisation.
             Moreover, Susan Cozzens (1982, 1985) found that the meaning of a
          scientific document can be different for different groups, and may even
          change over time.
   144   145   146   147   148   149   150   151   152   153   154