In science policy, it is often important to track emerging development
s: new fields, fast-changing areas that are the focus of special fundi
ng efforts, or areas of growth or decline. This article presents metho
ds to produce literature-based indicators for such areas using journal
-to-journal citations. Using case studies of AIDS, superconductivity,
and oncogenes, we posit that the inclusion of a new journal can be use
d as an indicator of structural change if the addition indicates the e
mergence of a new journal category. Using the cases of robotics and ar
tificial intelligence, we illustrate the development of areas chosen f
or priority funding. Again using artificial intelligence, we demonstra
te the importance of constructing even such simple measures of scienti
fic performance as publication counts using dynamic rather than consta
nt journal sets. Change in performance within a subfield can be system
atically distinguished from change in the delineations among subfields
over time.