Read Me First – A Chronology

by Adrian M. Wenner

This chronology [1A] introduces the publications found in this section of the web site. Numbers in brackets refer to each publication included. One can find a far more complete chronology in our 1990 book, Anatomy of a Controversy: The Question of a “Language” among Bees (Columbia University Press).

1930s and early 1940s

Initially, Karl von Frisch had hypothesized that bee “language” involved only odor-search behavior. He published one popular paper in 1937 (Science Round-Up: The Language of Bees) and a more scientific one in German in 1943 (Experiments on Directing Bee Flight by Odors) to that effect.

Some American and British scientists were dismayed by the shambles left for the German scientific community at the aftermath of World War II and did their best to promote those scientists that they felt had achieved solid accomplishments during the war years. Von Frisch was one such scientist selected for support.

Among other accomplishments, von Frisch had published the results of easily repeatable experiments, the results of which suggested that honey bees had a “language.” Almost anyone could repeat his experiments and gain supportive evidence for the hypothesis that naive bees attending a dancing bee would “fly directly out” (as von Frisch phrased it) to the same productive source of nectar or pollen. Unfortunately, the notion of testing a hypothesis in those days meant little more than a successful replication of the original experiments. At the same time, von Frisch’s earlier odor-search hypothesis disappeared from consideration.


I entered the graduate program at the University of Michigan in 1956. Earlier that decade I had served as an apprentice for several years with two of my uncles (Clarence and Leo), large scale beekeepers in Northern California (Adrian Wenner Retirement Announcement).

During that experience I had become indoctrinated with the notion that von Frisch had “proved” that bees had a “dance language” or that he had “discovered their language.” Only later, after a more intensive scientific training, did I appreciate the fact that he had really only “hypothesized” that honey bees had a language.

While selecting a dissertation topic, I initially avoided working with honey bees, since my bias at the time rested on the naive assumption that “all was known” about their biology. However, other events intervened. While caging thousands of honey bee queens from “baby nuclei” for sale to customers during my California apprenticeship, I had a special opportunity to hear the various sounds produced by bees in their hives.

Earlier, while in the U.S. Navy, I had obtained extensive training in electronics and later obtained a physics minor and a mathematics major in college. At about the same time, the portable audio tape recorder became available, as well as a sophisticated audiospectrograph device that could portray sounds as visual displays. I tape recorded the sounds of individual bees in the hive, analyzed those sounds, and sought to determine whether such sounds constituted communication among bees.

Imagine my surprise: bees engaged in the waggle dance produced a highly structured sound (The Relationship of Sound Production During the Waggle Dance of the Honey Bee to the Distance of the Food Source, Sound Communication in Honey Bees). The earlier training in electronics and physics immediately came to bear; perhaps dancing bees in their totally dark hive were actually communicating by means of sound signals instead of by the visible dance maneuvers that von Frisch had studied.

In the true spirit of scientific openness, I sent von Frisch copies of the tape recordings and audiospectrographs of the bee dance sounds. He replied with an invitation that I become his graduate student. That was not feasible, since I had no financial resources and a new wife and son to support. Little did I realize that von Frisch would then turn the tapes and spectrographs over to a technician in his group to exploit for study, without later acknowledgment of my role.

Further study revealed a correlation between sound production time during the straight run portion of the waggle dance and the distance to the food source visited by regular foragers (The Relationship of Sound Production During the Waggle Dance of the Honey Bee to the Distance of the Food Source, Sound Communication in Honey Bees). At that time I still worked under the assumption that bee “language” was fact. Also, I thought I had found the true means by which foragers communicated the distance of the food source to their naive hive mates in the dark of their hive. Little did I realize at the time that correlations count for little until such hypotheses have survived really critical experimental tests (not merely replication).

A close examination of all of the results von Frisch had obtained before that time revealed some serious flaws – both in experimental design and in interpretation of results. For example, the results von Frisch had obtained about the accuracy of “use” of dance maneuver information by searching bees far exceeded the accuracy of the information contained in that dance maneuver. (Much later, even James Gould admitted that the results of von Frisch’s experiments had not established that honey bees had a “language.”)


While at Michigan, I had learned the important distinction between indirect and direct evidence and realized that all evidence for bee language up to that time had been circumstantial (i.e., indirect). By contrast, if one could build an imitation bee (robot bee) and send naive bees out to a point source in the field, one would have direct evidence that the waggle dance had meaning. Furthermore, if such an artificial bee succeeded when it emitted sounds but not when silent, my discovery of the sounds made by dancing bees would have greater significance (Sound Communication in Honey Bees). To understand the waggle dance more thoroughly, I teamed up with Patrick H. Wells of Occidental College and my first doctoral student, Dennis L. Johnson.

Research done with eyes wide open often provides curious twists. While trying to construct an imitation bee, we stumbled onto the disconcerting notion that bees learn quickly (the conditioned response phenomenon, as with Pavlov and his salivating dogs). That realization perhaps should have come as no surprise, but for the fact that by then bee researchers and others had considered bee language an “instinctual signaling system” and thus would not involve learning.

Despite strong resistance by anonymous reviewers, we published the results of our experiments on learning in honey bees (see summary in Chapter 7 of our book, Anatomy of a Controversy). However, our find had far more serious implications – the experiments described in von Frisch’s classic 1950 Cornell University Press book dealt only with the re-recruitment of experienced bees, a success that could be explained solely by their reliance on odor and conditioned response. If bees had a language, such an ability would then only apply to the flight out of the hive by inexperienced bees.

A 1966 event at the Salk Institute in La Jolla, California (Excursus SI: The Question of a “Language” Among Bees) led to a series of experiments. Our results later appeared in the journal Science (but again, only against the strong objections of anonymous reviewers). The first set of experiments relied on a rigorous double control design, in which inexperienced bees would either use information they had obtained from the waggle dance or search for the odor of the food source exploited by experienced foragers (Honey Bees: Do They Use the Direction Information Contained in Their Dance Maneuver?, Honey Bees: Do They Use the Distance Information Contained in Their Dance Maneuver?).

A second set of experiments relied on a more rigorous strong inference design. In both sets of experiments, searching bees used odor of the target source and ignored any information they might have obtained from the waggle dance before leaving their hive (Honey Bee Recruitment to Food Sources: Olfaction or Language?).

Without realizing it, we had arrived at the same conclusion that von Frisch had published in 1937 (Science Round-Up: The Language of Bees) and in 1943 (Experiments on Directing Bee Flight by Odors).


Despite the compelling evidence we had gathered with the use of more rigorous experimental designs than ever employed earlier, James Gould (initially with fellow undergraduate co-workers) reverted to a single controlled experimental design and obtained evidence in support of the language hypothesis. Despite the extreme paucity of evidence that he gathered (much of which contradicted von Frisch’s and his own conclusions), most language advocates embraced his results as a “final solution” to the controversy.

After a few more publications (e.g., (Do Honey Bees Have a Language?), the door largely closed for us. Major scientific journals would neither accept manuscripts by us nor permit us to respond to challenges of our work (Excursus EXC: The Question of a “Language” Among Bees). With a premonition that such censorship would occur, I had already moved into the field of marine biology as a “sabbatical” of sorts, while waiting for the bee research community to become more open to free expression.


Articles in support of the bee language hypothesis continued to become published without adequate critical review. R. Rosin, though, maintained a challenge of the language hypothesis during the 1970s and 1980s in a series of articles and letters in journals. Her contributions remain largely ignored by bee language advocates.

In 1987 Joe Graham (editor of the American Bee Journal) provided our first break in the stalemate by publishing a summary article against the advice of antagonistic reviewers (The Honey Bee Dance Language Controversy: the Search for “Truth” vs. the Search for Useful Information). The outcry he had anticipated never materialized. Apparently, tempers had cooled considerably by then.

I then re-entered the field of bee research in collaboration with Robbin Thorp of the UC Davis campus. In 1988 we began to locate and remove all feral honey bee colonies from Santa Cruz Island, a 96 square mile (25,000 hectare) island offshore from Santa Barbara – part of the Channel Islands National Park. Our intent: to remove honey bee competition so native bees could rebound and pollinate native plants more effectively, thereby hastening island recovery toward a more natural state.

For more than a decade we recruited scores of volunteers and learned much about colony foraging patterns and many other aspects of bee biology (Efficient Hunting of Feral Colonies, The Honey Bees of Santa Cruz Island, Swarm Movement: A Mystery Explained, Recruitment, Search Behavior, and Flight Ranges of Honey Bees, Foraging, Recruitment, and Search Behavior of Honey Bees).

Also, by deliberately introducing the varroa mite we could study the rate of mite spread and time elapsed before colony demise occurred over the entire island. That project now seems near completion.

During that decade language advocates still continued to try to “prove” that naive bees could use the direction and distance information contained in the waggle dance. Such attempts, of course, only highlighted the fact that those researchers unwittingly thereby acknowledged that no one had really succeeded earlier. Researchers in Germany and Denmark gained much publicity about their “mechanical bee” experiments. However, a close inspection of their results revealed that they had not succeeded (Recruitment, Search Behavior, and Flight Ranges of Honey Bees). The notion of robot bee success now seems to have largely disappeared.

Columbia University Press agreed to publish our book, Anatomy of a Controversy: The Question of a “Language” among Bees. We had spent 20 years studying the broader scientific question (psychology, sociology, history, and philosophy of science) and spent five years writing the book. It survived two levels of severe reviewer scrutiny at the press and received rave reviews from almost everybody, with reviews by bee language advocates a notable exception. People in other academic disciplines grasped the meaning of our message and wrote their own articles about the controversy (e.g., The Honey Bee Dance Language Controversy).

With the “logjam” broken (2000 Years of Uncertainty, History and Status of the Two Von Frisch Recruitment Hypotheses, Recruitment to Food: Tangible Odor-Search or Mystical Language?, Why Not Give All the Facts?) we could once again publish freely and also received invitations to write lead articles for scientific journals (e.g., Recruitment, Search Behavior, and Flight Ranges of Honey Bees, Science as a Process: The Question of Bee “Language”, Is the Touted “Language” of Honey Bees Real?; also The Role of Controversy in Animal Behavior).


I again attended regional and national meetings and found that the near universal hostility encountered 20 years earlier had dwindled to an occasional cool reception. Many attendees privately voiced great support for the persistence we had shown through the decades.

The increasingly widespread use of the Internet had a perhaps unanticipated influence on the bee language controversy. No longer could language advocates suppress expression of divergent viewpoints, as they had done so successfully while serving as anonymous reviewers of manuscripts submitted to journals. During this past decade this controversy has surfaced on several e-mail networks (e.g., BEE-L, social insects, entomology, comparative psychology, history and philosophy of science), once each year or two.

Each such episode provided an open platform for expression of divergent views, with such exchanges lasting a month or more. Whenever someone pressed the case for naive bee use of dance maneuver information, I could cite one publication or other that contained hard evidence to the contrary and could mail photocopies to interested parties. That opportunity contrasted sharply with the near total absence of mention of our research in all beekeeping and animal behavior books.

This newly open atmosphere also permitted me to publish summaries of work overlooked or ignored earlier, as well as digests of material important to beekeepers (e.g., Swarm Movement: A Mystery Explained). For instance, in 1973 Larry Friesen published results of a study that documented the importance of wind direction for recruitment of naive bees to food sources, the only such study this past half century. Now beekeepers and researchers have ready access to that information, including newer finds on this subject, (e.g., 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). Perhaps others may now pursue this promising avenue of research.

Meanwhile, R. Rosin continued to publish challenges of the mindset that has been with us these past few decades (e.g., A Note on the Decisive “Proof” for Use of “Dance Language”, More on the Honey Bee “Dance Language” Controversy, Do Honey Bees Still Have a “Dance Language?”).

Although flare-ups still occur occasionally (e.g., Bee “Language” Again?), the notion of bee “language” continues to recede in scientific discourse.

In time, those interested in honey bee biology may appreciate the notion that the waggle dance maneuver may well be only a symptom of what a foraging bee has experienced as it flies between hive and food place, not a signal for other bees. If so, the millions spent to study “bee language” may have been largely for naught (The Elusive Honey Bee Dance “Language” Hypothesis). It will not be the first such case in science, though; we also have “cold fusion,” “polywater,” “water with a memory,” etc.

A problem similar to experiments on presumed function of the dance maneuver repeatedly surfaces with respect to another behavioral trait, exposure and presumed function of the Nasanov gland. Under specific circumstances a bee raises its abdomen and flexes the last segment down, releasing a fragrance from a pouch opened by that motion. For more than three-quarters of a century “conventional wisdom” has held (in part) that the Nasanov gland pheromone attracts searching bees to food sources exploited by other bees.

In the June 2004 issue of Bee Culture Larry Connor published a review of Nasanov pheromone. In that article he repeated the prevailing assertion, initially based upon some simple experiments conducted by von Frisch in the 1920s, that successful foragers expose that gland when at a profitable food source and that searching bees then become attracted to that same source. However, von Frisch published some compelling and far more extensive negative evidence in 1947 (see Table NG.1 in Excursus NG: The Scent Gland (Nasanov Gland) of Honey Bees). In those later studies of his, searching bees had shown no preference for food at stations with scent glands open as against those with scent glands closed. At the time von Frisch dismissed those negative results that he himself had obtained.

As a result of extensive experimentation in the 1960s and 1970s, we became disillusioned with the established doctrine of Nasanov gland pheromone attraction during our “crucial” or “strong inference” test of the dance language hypothesis (see Table 1 in Honey Bee Recruitment to Food Sources: Olfaction or Language?). By controlling the amount of odor in the food, we learned that decreased amount of food odor coincided with increased exposure of the Nasanov gland, increased dancing in the hive, but decreased recruitment of bees to a food source. Clearly, Nasanov gland pheromone did not attract bees to a food source (see also Table NG.2 in Excursus NG: The Scent Gland (Nasanov Gland) of Honey Bees).

Even more extensive results about Nasanov pheromone lack of attraction appear in a 1993 paper (P. Wells, et al., “Does Honey Bee Nasanov Pheromone Attract Foragers?” (see Tables 1-3 in Does Honey Bee Nasanov Pheromone Attract Foragers?). Unfortunately, those and von Frisch’s extensive negative results have not appeared in books and articles written by contemporary researchers. That omission has led many researchers to repeatedly “start from scratch” in attempts to use artificial chemical components of the Nasanov gland secretion to enhance visitation of crops – even though Gordon Waller published negative results from similar experiments in 1970 (See p. 315 in Excursus NG: The Scent Gland (Nasanov Gland) of Honey Bees).

We summarized Waller’s conclusions (in part): “None of these fragrances regularly increased bee populations in experimental alfalfa plots when applied in water, but each of them singly or in mixtures did so when applied in sucrose solutions. Apparently these odors had thereby merely served as marker stimuli when coupled with a food reward in a conditioned response situation.” (See A Parade of Anomalies: Learning)

Various claims of success in crop visitation enhancement experiments during the last two decades have seemed unjustified, leading Connor to conclude: “The results have not been exciting.” In short, available evidence indicates that the Nasanov pheromone fails the test as an attractant. Instead, it appears to be an orientation and settling pheromone – most notably with respect to the location of the hive entrance, to swarm movement, and while that swarm settles into a new location, as explained in Connor’s article and more fully in Swarm Movement: A Mystery Explained.

Despite evidence accumulating counter to the dance language hypothesis, bee language advocates press on with dogged adherence to that hypothesis. The recent development involved British and German researchers who have employed a radar tracking technique – as in their own words (12 May 2005):

“We have used harmonic radar to measure the flight trajectories of bees recruited after observing the waggle dance, this has enabled us to settle (hopefully once and for all) this controversy in favour of Von Frisch.”

One can easily recognize in that statement, and in other statements of theirs, that the researchers started with the conviction of bee language and then set about to prove their bias. That type of biased experimentation – attempting to prove a hypothesis true – has been criticized repeatedly by scholars of scientific method. Furthermore, the amount of positive evidence they obtained was meager. For a critique of their most recent contribution in Nature (see Did Radar Tracking of Bee Flight Paths Resolve the Bee Language Controversy?).

Increasingly, beekeepers and researchers around the world are awakening to the realization that the dance language hypothesis has proved of little or no use for understanding colony management or recruitment to crops for pollination. Our work with odor and importance of wind (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), among other contributions, led to an invitation for me to be the Keynote Speaker at the Third European Congress on Social Insects in St. Petersburg, Russia (August 2005; see Odor and Honey Bee Exploitation of Food Crops) and a year later as Plenary Speaker at the Eight European Congres of Entomology in Kusadasi, Turkey. At about the same time, I was elected President of the Western Apiculture Society.

At that social insect congress in Russia, a Plenary Speaker from Australia reported on experiments with honey bee learning, independent work in parallel with our 1960s studies.

Mapping of the Honey Bee DNA Genome

Yet another breakthrough emerged as a result of the mapping of the honey bee DNA genome. The Washington Post summarized some of the findings on 10/30/2006:

A. mellifera has 170 genes for “odorant receptors,” of which 157 are in a gene family so far found only in honeybees. This is far more smelling apparatus than either fruit flies (with 62 receptor genes) or mosquitoes (with 79) possess. It probably reflects the extreme importance of smell in helping bees find flowers and communicate with one another, including with their queen, through pheromones.

That research provided us with a noteworthy find with regard to the bee language controversy — those researchers found no genes that would support the dance language hypothesis as the touted “instinctual signaling” system that it had become known as (see The Honey Bee Odor-Search Hypothesis & DNA Genome Analysis).

If history repeats itself, though, we can expect that bee language advocates will not likely abandon their belief in their favored hypothesis but will continue to seek confirmatory evidence that may support that belief system.


After every lecture that outlines the evidence against bee language, the first question is almost always, “But, then, why do bees dance?” Our Judeo-Christian heritage instills in most of us the teleological notion that everything must have a purpose – presumably due to actions of a creator. However, the scientific approach to problems works best without such an attitude. We treated this topic at length in Excursus TEL of our 1990 Columbia University Press book (see Excursus TEL: Teleology).


Despite a deep entrenchment in the public psyche, the bee language controversy continues to reach an ever-wider audience (e.g., The Role of Controversy in Animal Behavior) and promises to become an object lesson in how science progresses – not so much by “proofs” and “discoveries” as by the generation and replacement of hypotheses (see following):

…as well as genuine fun at times, as illustrated by Joe Traynor Bee Controversy Breakthroughs: Odor-Search vs. “Language”.

Barry Birkey’s web site now provides a selection of some of the publications so long ignored and/or suppressed by bee language advocates. In fact, this well may be the first opportunity for former and current graduate students of those language advocates to know that such information exists.