The following message is sent by Adrian Wenner ([log in to unmask]).
Please submit all replies to him (not REPLY function). Thank you.
 
 
 
Over some of these networks, I earlier (1/24/96) provided a list of 16
facts that counter
claims that honey bees have a "language," facts that (by contrast) mesh
well with the
competing odor-search hypothesis.  Later (1/31/96) I provided a dissection
of the bee
language controversy into some five of its essential elements so the
various points
could be addressed without too much confusion.  Nevertheless, questions
arose, and
several individuals asked me to provide more input on some items.
 
   In an earlier transmission about the bee language controversy I had
written:  "One can
neither prove a hypothesis true nor prove it false."  Despite that caution and
reservations expressed by other respondents,  language proponents have avoided
addressing the 16 troublesome facts posted earlier.  They instead
emphasized vague
notions such as "compromise" hypotheses, "overwhelming evidence,"  and
"definitive
experiments," comments that can be addressed easily, as follows:
 
COMPROMISE HYPOTHESES
 
   During the 1940s and 1950s, geneticists debated whether protein or DNA
carried the
genetic information between generations.  However, where would genetic
engineering
be today if they had settled on a compromise hypothesis (e.g., sometimes DNA is
responsible and sometimes protein)?  No, they turned to STRONG INFERENCE
experiments and resolved the issue in favor of DNA (protein is involved in
other
processes).
 
DEFINITIVE EXPERIMENTS:  "ROBOT" BEES.
 
   When the robot bee episode began (after many had lost faith in the
significance of
"misdirection" experiments), it received a great deal of media attention
(as did "cold
fusion").  We studied the experimental results as published in the original
1989
NATURWISSENSCHAFTEN article but found that the results actually did not
differ from
random (see Wenner, Meade, and Friesen, 1991, "Recruitment, Search
Behavior, and
Flight Ranges of Honey Bees." American Zoologist  31:768-782).  When one of
those
investigators visited my laboratory three years ago, I showed him that
results in one of
his papers indicated that searching bees (after leaving a robot bee)
required about two
HOURS to reach a target station located less than a MINUTE flight time
away.  His
response:  "Here, let me show you some other results."
 
DEFINITIVE EXPERIMENTS:  OCELLI-BLINDED "MISDIRECTION" EXPERIMENTS.
 
   We treated a basic flaw in the experimental protocol of the misdirection
experiments
on pages 252 and 253 of our book:  Wenner & Wells, 1990, Anatomy of a
Controversy:  The Question of a "Language" Among Bees (Columbia University
Press).  In there we described how one cannot reconcile the results of the
ocelli-blinding
experiments (nor the experimental design) with other experimental results
published by
the same author five years earlier.  Neither do the protocol or conclusions
mesh with the
results of other ocelli-blinding experiments published by Renner and
Heinzeller (bees
with ocelli totally blinded could not function normally).  I also draw
attention of the reader
to two other relevant porrtions of that book (with respect to this issue):
Excursus NEG
("Negation of the Dance Language Hypothesis") and Excursus GT ("The Method of
Training Bees").
 
   When we executed the STRONG INFERENCE design experiments in a second series
of experiments (summarized in chap. 10 of our book), we published  ALL the
results of
three-hour runs for each of 24 consecutive days.  [In those experiments,
recruits did not
arrive at target stations supposedly indicated by dance information but
instead at a
remote station scented with the odor brought in by bees the day before.]
 
   By contrast, the first blackened occeli misdirection report (NATURE,
1974) relied for
its conclusion on results obtained during three one-half hour runs (perhaps
out of 33 of
that type run that season) of SINGLE-CONTROL design experiments.
Furthermore, (if I
read the available material correctly and as an example), those same three
one-half
hour data sets, augmented by an additional half hour set of results, were
published
again (in different formats) in the 1975 SCIENCE and 1976 QUARTERLY REVIEW OF
BIOLOGY articles (cited earlier in this exchange by others); those other
displays thus do
not represent replicates of the experimental results published earlier in
NATURE, as
one might suppose.
 
   Neither NATURE nor SCIENCE would permit us a simple reply to those
challenges of
our work, but I can furnish a revealing critique of those experiments by
regular mail to
anyone interested.  One can also read chapter 13 of our book for a summary
of Rosin's
independent critical analyses of those experiments.