The Honey Bee Odor-Search Hypothesis & DNA Genome Analysis

Bee Culture – June, 2007 – Pages 25-26

Is it conceivable that honey bees do not have a symbolic “language?” Judging from letters and articles in bee magazines (including items in recent issues of Bee Culture and the American Bee Journal), it appears that some people, when pressed, would still reply, “No, that is not conceivable, indicating that their “belief system” remains intact.

We can expect such a ready answer, because “bee language” has now been treated as “fact” in innumerable publications and other media outlets for half a century. Even respectable scientists confuse hypothesis (interpretation) and fact when it comes to “bee language.” The waggle dance is observable fact; the suggestion that attendant bees can use the quite inaccurate distance and direction information present in that dance remains an unproven hypothesis. That is true, despite many claims to the contrary, including claims by those who conducted the radar tracking experiments (see Did Radar Tracking of Bee Flight Paths Resolve the Bee Language Controversy?).

Let us backtrack. For more than 2000 years people have wondered how hivemates of successful foraging honey bees manage to find the same source of food in the field (see Did Radar Tracking of Bee Flight Paths Resolve the Bee Language Controversy?). A somewhat obscure passage in Aristotle’s writings (330 B.C.) suggests that he believed potential recruits may have followed the forager back out to the same food source. Virgil later used the words, “Some lead their youth abroad,” presumably with the same meaning intended.

In 1901 Maeterlinck pondered two hypotheses, as follows:

“Do the comrades who flock to the treasure only follow the bee that first made the discovery, or have they been sent on by her, and do they find it through following her indications, her description of the place where it lies? Between these two hypotheses, that refer directly to the extent and working of the bee’s intellect, there is obviously an enormous difference.”

We can think of those two possibilities as:

1) A “following” hypothesis, as one might expect from other insect behavior studies.
2) A “language” hypothesis (an anthropomorphic explanation), something akin to what we would expect among so-called “higher” life forms, such as Julien Francon proposed in 1939 and von Frish gained supportive evidence for less than a decade later.

Maeterlinck did not consider a third possibility:

3) An “odor-search” hypothesis, such as one insisted upon by von Frisch in the late 1930s.

By 1940 we thus had three competing hypotheses — a situation ripe for exploitation by a “multiple inference” scientific approach to the problem, as eloquently proposed by Thomas Chrowder Chamberlin in 1890. However, instead of keeping all options open, bee researchers after the mid-1940s (including me, at first) sought to substantiate and refine the new language hypothesis proposed by von Frisch in 1946. Chamberlin had recognized the shortcoming of that traditional approach (one subject to “confirmation bias”), as follows:

“Briefly summed up, the evolution is this: a premature explanation passes into tentative theory, then into an adopted theory, and then into ruling theory.”

“There is an unconscious selection and magnifying of the phenomena that fall into harmony with the theory and support it, and an unconscious neglect of those that fail of coincidence. The mind lingers with pleasure upon the facts that fall happily into the embrace of the theory, and feels a natural coldness toward those that seem [not compatible with the theory]. Instinctively there is a special searching-out of phenomena that support it, for the mind is led by its desires. There springs up, also, an unconscious pressing of the theory to make it fit the facts to make them fit the theory.”

Instead, Chamberlin had advocated a working hypothesis approach:

“The working hypothesis differs from the ruling theory in that it is used as a means of determining facts, and has for its chief function the suggestion of lines of inquiry; the inquiry being made, not for the sake of the hypothesis, but for the sake of facts.”

Although the stage was thus set for some serious scientific tests of competing hypotheses (each hypothesis against the other two) by bee researchers, WWII complicated matters. It was during that period that von Frisch had discovered the intriguing correlations between elements of the by-now-famed “waggle dance” and the direction and distance of food sources visited by foraging honey bees in the field. He had concluded, after gaining supportive evidence, that potential recruits had interpreted the information contained in those dances and had “flown directly out” (his words) to the same food source. In so doing, he prematurely abandoned his earlier odor-search hypothesis.

At the end of that war, prominent scientists in England and the United States strove to help German scientists become re-established in the world community. Among other projects, the honey bee waggle dance “language” hypothesis had special appeal (the exotic “sells” in science as well as elsewhere). Very soon the dance language hypothesis became elevated to a “ruling theory” (as in Chamberlin, above). The competing hypotheses (“following” hypothesis and “odor-search” hypothesis) no longer received consideration.

By the 1960s the ruling theory had thus become “fact” in the minds of most scientists (including me) and the lay public. Instead of “dance language hypothesis” we had “the dance language” (considered as “fact”) and ” their language” (interpretation considered as “fact”). Thus, the scientific distinction between interpretation and fact had disappeared. Evidence that didn’t fit that ruling theory became unacceptable — a dismissal practice that largely continues to this day.

However, no one has yet come up with the “extraordinary evidence required” for the “extraordinary” hypothesis (as in Carl Sagan’s famous statement to that effect) for bee language. Equally important: Anyone who wants a particular outcome should not be the person to conduct the relevant experiment. Either that, or such a person should take extraordinary measures to counter confirmation bias (as for example, by use of blind, double blind, and strong inference experimental design – cautions sadly lacking in most confirmation experiments).

We have had three generations of citizens and fledgling scientists exposed to the bee language ruling theory as fact from cradle to grave (children’s books up through advanced biology textbooks and encyclopedia entries).

However, hypotheses do not become facts but always remain hypotheses. New evidence emerged in the 1960s and 1970s that countered expectations of the language hypothesis and supported the von Frisch odor-search hypothesis of the 1930s. Searching bees did not “fly directly out” to the “intended” food location. Most that left the hive failed to find the food source. Searching bees that succeeded took far too long in flight. (For access to much of that body of counter evidence, see: Bio page).

In simple experiments (a design that any objective beekeeper or researcher can conduct with help from a body of student volunteers), that pitted odor against dance maneuver information, searching recruits had ended up where the odor was present and had obviously ignored the direction and distance information present in the waggle dance (See: Honey Bee Recruitment to Food Sources: Olfaction or Language? for details). In that experiment the search behavior of more than 2000 bees were monitored with use of a blind experimental design. Those unwelcome odor-search results were then either dismissed or ignored by proponents of the ruling theory — as Chamberlin had emphasized, they apparently felt “a natural coldness toward [results] that seem refactory.”

A further complication emerged. Experiments revealed that searching bees do not find food that has no scent, even when regular foragers imbibe and return to the hive with their loads of pure sucrose solution (which has a vapor pressure of zero — hence no odor). Von Frisch had come to the same conclusion, as in the following passage:

“In performing this experiment, I succeeded with all kinds of flowers with the exception of flowers without any scent. And so it is not difficult to find out the manner of communication. When the collecting bee alights on the scented flowers to suck up the food, the scent of the flower is taken up by its body-surface and hairs, and when it dances after homing the interested bees following the movements of the dancer bee, and holding their antennae against its body, perceive the specific scent on its body and know what kind of scent must be sought to find the good feeding-place announced by the dancing bee. That this view is correct can be proved easily.”

A dilemma then arises; to have recruitment, one must use odor in or near the target food source (any odor will suffice – even incidental odors, such as use of sun tan lotion by assistants – can influence results). But, if some odor is necessary for success, then one can never be sure that successful searchers had not used odor alone rather than any direction or distance information obtained from waggle dances.

The recruitment controversy has now existed for nearly half a century, with a quite specific starting point (see: Anatomy of a Controversy, Excursus SI). Millions have now been spent by those locked into their belief system, in repeated efforts to “prove” the “reality” of the “instinctual signaling system” (as it had come to be known). Language advocates seem unaware that each such attempt at proof constitutes an inherent admission that all previous such attempts had failed (see: The Elusive Honey Bee Dance “Language” Hypothesis). Either that, or such attempts have failed to come up with the necessary “extraordinary” evidence for supposed final “proof.”

The latest development in the controversy emerges from the genome sequencing of the DNA in honey bees. A total of 170 odorant receptor genes were found, indicating “a remarkable range of odorant capabilities.” By contrast, the DNA sequencing study found no unique cluster of genes that would indicate that searching recruits could use direction and distance information found in waggle dances. Such evidence of extensive odor reception capability constitutes a powerful endorsement of the odor-search hypothesis.

Does this all mean that many bee researchers will now conduct research with odor-search as a primary working hypothesis? Can they break free of the restrictions imposed by the ruling theory (bee language) so long adhered to? That is, can they now alter their approach and no longer conduct experiments “for the sake of the hypothesis but [instead strive] for the sake of facts”? An overriding question also emerges: Is it even possible for scientists and others to abandon their allegiance to the honey bee dance language hypothesis. (Sociologists who study the process of science have found that scientists are the last to abandon a favored hypothesis.)

Whereas the language hypothesis has not proved of practical benefit for beekeepers in its half century of existence, I see great potential for beekeeping advances in future studies of the importance of odor during recruitment to crops. While some have accused me of being stubborn, I simply trust what I have seen from the “unmolested” behavior of thousands of searching bees – rather than rely on what theory might dictate.

Let us hope that serious bee researchers will now turn their attention to the promise of studying odor use by bees and its great potential for improving crop pollination and understanding colony foraging patterns. A decade ago I provided some leads toward that end in a series in the American Bee Journal, now readily available as follows:

Odors, Wind and Colony Foraging (1 of 3): The Need for Odor
Odors, Wind and Colony Foraging (2 of 3): The Role of Wind Direction
Odors, Wind and Colony Foraging (3 of 3): Insights from Bee Hunting

For a few years, we can expect writers of popular literature and authors of reviews will continue to extol the “remarkable bee language” in their publications for decades to come – focusing on confirmatory evidence instead of focusing on the beauty of this example about how science is a process and not a search for absolute truth.


Many thanks to Barry Birkey, who provided a platform for results and ideas so long suppressed.

Adrian M. Wenner*
967 Garcia Road, Santa Barbara, CA 93103
*Professor Emeritus (Natural History), UCSB