On the question of bee "language":
 
******** SECOND COMMENT ("Compromise" does not lead to progress in science)
 
   In this recent interchange we were once again asked to believe:  "1)
sometimes the bees do dances, and their nestmates use the information in
these dances to influence where they themselves seek forage, and 2) [at]
other times the bees still do the dances but their nestmates make little or
no use of the dance information [as to] where they end up foraging."
 
   That position seems to serve as a "security blanket" of sorts.  If I do
an experiment with only single controls (as von Frisch, Gould, and other
language proponents have done) I can get results supportive of the language
hypothesis --- that is, condition #1 (above) prevails.  However, if I do
strong inference or double controlled experiments (as Wenner and co-workers
did) and obtain evidence not in agreement with the language hypothesis
(whatever that might be now), then condition #2 (above) prevails.  We thus
have a peculiar circumstance --- any set of results is acceptable, and we
need not concern ourselves with hypothesis testing.
 
   However, where would we be today if geneticists in the 1940s and 1950s
had compromised, as follows:  "sometimes DNA carries the genetic
information and sometimes protein carries that information."   (Would we
have "genetic engineering" today?)  Or, what if Pasteur and fellow
scientists had compromised:  "sometimes life arises by spontaneous
generation and sometimes life can only come from life."   (Would we have
pasteurized milk today?)
 
   I suggest that bee language proponents now get together and agree upon a
concise scientific statement of the hypothesis they believe in, one that
can be tested experimentally.  (Language proponents now seem to agree that
the conclusions of von Frisch were not justified on the basis of the
evidence he had gathered.)
 
   On this point I am reminded of a recent statement by Harold B.
Hopfenberg (1996.  "Why wars are lost."  AMERICAN SCIENTIST.  84:102-104):
 
 
   "The biased hypothesis...provides valuable scaffolding for sincere
inquiry as long as the methodology and subsequent interpretation of data do
not result in the confusion of hypothesis with conclusion."
 
   Look for the THIRD COMMENT  that follows shortly (Why scientists
necessarily rely more on experimental results than on "consensus").
 
                                                      Adrian
 
 
Adrian M. Wenner                         (805) 893-2838 (UCSB office)
Ecol., Evol., & Marine Biology           (805) 893-8062  (UCSB FAX)
Univ. of Calif., Santa Barbara           (805) 963-8508 (home office & FAX)
Santa Barbara, CA  93106