For a published departure from conventional scientific professional etiquette, see the survey article “Mathematics and the Internet: A Source of Enormous Confusion and Great Potential,” in which Walter Willinger, David Alderson, and John C. Doyle criticize scale-invariant network models of the Internet. The article is unusual for its polemic, insulting tone. While it is not unusual for researchers to insult other researchers in private conversation, it is unusual to see this in print. Its authors spare no opportunity to criticize their competition, as well as mathematicians and physicists generally, whom they regard as foppish, insular ivory tower aesthetes, whose nostrils are unacquainted with the bracing scent of an expertly soldered electrical connection.
The authors deploy a literary reference to insult their competition:
“What about replacing power-laws by the somewhat more plausible
assumption of high variability in node degrees? While the answer of
the scale-free modeling approach consists of tweaks to the PA
mechanism to enforce an exponential cut-off of the power-law node
degree distribution at the upper tail, the engineering-based approach
demystifies high-variability in node degrees altogether by identifying
its root cause in the form of high variability in end-user bandwidth
demands (see [33] for details). In view of such a simple physical
explanation of the origins of node degree variability in the
Internet’s router-level topology, Strogatz’ question, paraphrasing
Shakespeare’s Macbeth, “… power-law scaling, full of sound and fury,
signifying nothing?” [52] has a resounding affirmative answer.”
The authors seem to suggest by this literary reference, which would not be lost on readers of the AMS Notices, that a model of the internet that predicts a power law node degree distribution is a “tale told by an idiot.”
The authors suggest that mathematicians and physicists must get their hands dirty, do some engineering and then contemplate the authors’ HOT models of Internet connectivity, which they assert, will be more mathematically interesting “… and certainly more relevant and hence more rewarding than that of the scale-free models of the PA type.” This sentence combines a dubious claim about what mathematicians should find interesting with a swipe at scale-free preferential attachment models of the Internet.
The authors conclude with these remarks:
“In this article, the Internet has served as a clear case study, but
the issues discussed apply more generally and are even more pertinent
in contexts of biology and social systems, where measurement is
inherently more difficult and more error prone. … Although the
Internet story may seem all too obvious in retrospect, managing to
avoid the same mistakes in the context of next generation network
science remains an open challenge. The consequences of repeating such
errors in the context of, say, biology are potentially much more grave
and would reflect poorly on mathematics as a discipline.”
Why would mathematics be at fault? The authors do not cite the literature on the independent history of debate over the applicability of power law models in biology and the social sciences, e.g., A Brief History of Generative Models for Power Law and Lognormal Distributions by Mitzenmacher.
Again I mention this as an unusual example in print of what appears to me to be a departure from conventional scientific etiquette.
