A recurring theme in the use of science and technology indicators, as well
as in the construction of new ones, is the interpretation of these indicato
rs. Given the dependence on citation data in the majority of interesting sc
ience and technology indicators, a general citation theory would make the m
eaning of S&T indicators more transparent. Hence the continuing call for a
citation theory in scientometrics. So far, such a theory has not yet been a
ccepted by the experts in the field. This paper suggests an explanation for
this. It also tries to sketch the outline of a general indicator theory by
discussing new implications of an earlier proposal (Wouters, 1998) in rela
tion to existing citation and indicator theories.