Abstract :
The author examines patterns of productivity in the
Internet mailing lists, also known as discussion lists or
discussion groups. Datasets have been collected from
electronic archives of two Internet mailing lists, the
LINGUIST and the History of the English Language.
Theoretical models widely used in informetric research
have been applied to fit the distribution of posted messages
over the population of authors. The Generalized
Inverse Poisson-Gaussian and Poisson-lognormal
distributions show excellent results in both datasets,
while Lotka and Yule–Simon distribution demonstrate
poor-to-mediocre fits. In the mailing list where moderation
and quality control are enforced to a higher degree,
i.e., the LINGUIST, Lotka, and Yule–Simon distributions
perform better. The findings can be plausibly explained
by the lesser applicability of the success-breedssuccess
model to the information production in the electronic
communication media, such as Internet mailing
lists, where selectivity of publications is marginal or
nonexistent. The hypothesis is preliminary, and needs to
be validated against the larger variety of datasets. Characteristics
of the quality control, competitiveness, and
the reward structure in Internet mailing lists as compared
to professional scholarly journals are discussed
