Dear BEE-L subscribers,
After my short input about confirmation bias on the 19th, Bob
Harrison raised some points related to that concept. Let me clarify.
Confirmation bias is one of the most important problems in
scientific research, but all too many scientists and others fall into
that seductive trap. As eminent philosopher Karl Popper wrote,
“Confirming evidence does not count except when it is the result of a
genuine test of the theory.” He also pointed out that confirming
evidence is very easy to obtain. (Bob: you can find Popper’s
comments on p. 22 of our 1990 ANATOMY OF A CONTROVERSY book.)
Consider now that beekeepers have endured (?) quite a few claims
about techniques that might control/eliminate varroa mites. We have
had promotions of mineral oil, powdered sugar, small cells, screen
bottom boards, etc., all mostly based upon confirmation-type
experiments. A problem then surfaces: anyone who puts a lot of stock
in one of those treatments and loses a great many colonies will likely
remain silent about the loss. By contrast, should one’s colonies
survive, the natural tendency is to alert others to the success (even
if the technique itself might not have been responsible). That’s human
nature.
But anyone who really WANTS a particular outcome should not be the
person to do the experiment. (That is why we have blind, double
blind, and strong inference experimental designs, techniques usually
avoided by those locked into dogma.)
Likewise, anyone committed to dogma should not be the person to
write a review about a controversy; yet, that is done all the time.
For example, Emily Smith and Gard Otis epitomized that flaw in their
“review” of the dance language controversy, as published in the March
and April 2006 issues of the AMERICAN BEE JOURNAL. They rounded up all
the confirmation evidence and positive commentary they could find. In
doing so, they omitted a great deal of evidence that backs up the 1930s
von Frisch odor-search hypothesis and counters the dance language
hypothesis.
Pat Wells and I addressed the Smith and Otis omission in our letter
published in the subsequent July issue of ABJ (p. 561). For an
expanded version, see
http://www.beesource.com/pov/wenner/abjmar_aprreply.htm (“Resolving a
controversy or shoring up a belief system?”)
Bob also included a sentence from James and Carol Gould’s book: "
Wenner and Wells were misled by their assumption that if bees
communicate by odor under one set of circumstances , then they must use
odor in all cases." Gould rose to stardom by reverting to single
controlled, confirmation-type experiments, consistently misrepresenting
our conclusions (as above), and omitting evidence contrary to his
belief system.
Bob also wrote, “I gathered from reading the book the Goulds had
done experiments to prove Wenner & Wells hypothesis incorrect?” No,
Gould misrepresented the odor-search hypothesis and instead gathered
confirmation type evidence for the existing dogma, evidence welcomed by
those who wished to believe. Then he claimed that he had proven us
wrong. However, journals did not permit us to respond to his claims.
For example, see: http://www.beesource.com/pov/wenner/EXC.htm (Bob: on
p. 274 of our book).
The genome sequencing of the honey bee DNA, completed in 2006,
provided an opportunity to resolve the controversy. Recruitment
communication, if an “instinctual signaling system” as claimed, would
require the presence of genes not shared with other insects. No genes
for “bee language” surfaced. Researchers instead found a total of 170
odor receptor genes (most not shared with other insects), indicating “a
remarkable range of odorant capabilities.” The language hypothesis had
thus failed another test and the odor-search hypothesis gained more
support; see: http://www.beesource.com/pov/wenner/jib2002.htm
Can we now expect an O.J. type defense tactic (DNA isn’t necessary
for the instinctive dance language, after all)?
Adrian
Adrian M. Wenner (805) 963-8508 (home office phone)
967 Garcia Road [log in to unmask]
Santa Barbara, CA 93103 www.beesource.com/pov/wenner/index.htm
"Having one view prevail is harmful; it becomes a belief system, not
science."
Zaven Khachaturian — 2006
-- Visit www.honeybeeworld.com/bee-l for rules, FAQ and other info ---
|