Scientometrics, vol. 12. Nos 5-6 (1987) 339-341 THE SIGNIFICANCE OF BIBLIOGRAPHIC REFERENCES A comment on "TeSting the Ortega Hypothesis: Facts and Artifacts" by M. H. MACROBERTS and B. R. MACROBERTS, Scientornetrics, 12 (1987) 293. H. SMALL Institute for Seientific Information, 3501 Market Street, University City Science Center, Philadelphia, PA 19104 (USA) (Received February 18, 1987) It is not accidental that the MacRobertses cite DavidEdge, the first critic of cita- tion analysis to create the caricature of "references as influence". 1 Whether this reference represents influence or afterthought, however, is irrelevant to my purpose. I consider it as given that some references represent influences, while others do not. To say that all influences on an author are included in his reference list is absurd, and if the MacRobertses have attributed this view to the Coles and others it is a distortion. Moravcsik 2 and his "organic" category of references was the first to quantify the number of "influential" references and many have added to this discussion, including the MacRobertses. 3 The issue is not whether we can rely on reference lists in individual cases as complete sets of influences (we cannot), but rather whether references can be used statistically, in the aggregate, as an indicator of influence, as the Coles have done. 4 There is a great deal of evidence that "influential" papers are more highly cited than "uninfluential"ones. s There is no evidence to suggest that highly referenc- ing papers are highly "influenced", whatever that might mean. Of course much further work could be done on this topic. Vie.wing references only as potential influences, however, misses the main point and the reason citation indexing is so powerful. 6 To understand the significance of references we need to examine the cognitive processes involved in generating written discourse. The most obvious characteristic of references is that they are attached to text and usually to specific points in text. This comes about when we attempt to express ourselves on a topic, and in our train of thought, come to an idea which is in our mind associated with a specific source. We then create a "pointer" to that external source using the accepted conventions. For example, if I am discussing the topic of criticisms of citation analysis I immediately think Of David Edge's paper. I This is literally an act of "information retrieval" in the cognitive psychology, as well as bibliographic, sense of the term, where we retrieve an associated object from Elsevier, A msterdam- Oxford-New York A kadgmiai Kiad6, Budapest