Theories of Informetrics and Scholarly Communication. Cassidy R. Sugimoto, Editor. Berlin: de Gruyter Mouton, 2016. 426pp. $112.00 (hardcover). (ISBN:978-3-11-029803-1)

It gives great satisfaction to publish an article about a nice piece of research with results that are relevant and useful for the scientific community. In that sense a publication is in the first place a report or communication to colleague-researchers. This, however, does not automatically imply that the author is seeking communication with colleague-researchers. Perhaps an author may behave like an artist: this is my work, I show it to you, and you can think about it what you want. This remark is a preamble, to show that there is a certain ambiguity in the concept of communication. Publications are by definition stored (in databases, curricula vitae, and so on) as separate building blocks. Almost invariably discussions on and subsequent “theories” of citation behavior are based on the phenomenon of a single publication. In reality, a specific publication is usually not the central object. Scientists often write, as their research progresses, several publications on the same or related topics. Instead of an isolated publication, we are dealing with a scientific “oeuvre” (Moed & van Raan, 1986). The way scientists cite will be influenced by this multiplicity of publications. Particular references are related to specific aspects of work discussed in one publication; for instance, in a historical introduction. Other references will be related to aspects dealt with in other publications in the oeuvre. Thus, authors have the choice to allocate references according to different aspects of their work discussed in related publications. A model of citation behavior should take the oeuvre concept into account. I speak about “model” because in my view “theory” is an exaggerated term in the context of informetrics. A theory is a body of knowledge that can both describe as well as explain the causal elements responsible for a phenomenon in such a way that it can make falsifiable predictions. I regard the definition of theory given on the first page of the book—after all, the book is about theories—as too weak for a theory. Therefore, in most cases I will speak about models rather than about theories.