[1999 Rosin, R. Do honey bees still have a "dance language"? Am. Bee J. 139:577-578.]

ABJ recently published a very long article full of praise for the honey bee "dance language" hypothesis, by Taber (1998), and a very long response by Wenner (1998). Wenner's response, however, barely addresses Taber's major points. I shall undertake to very briefly address these points here.

1. Taber claims that recruits in flight cannot sense the wind. What insects in flight cannot sense, nor determine, is the direction of the wind in relation to the still ground. All of them can, however, sense the motion of air in relation to their own body, and this is all they need in order to find sources of attractive odors by use of odor alone, with the aid of directional olfactory sense-organs. (The insects do not respond directly to the wind. Instead, they respond to the effect of the wind on the direction of motion of odor-molecules).

2. Honey bees cannot obtain "dance language" information. The results of v. Frisch's "step" & "fan" tests, touted by Taber, grossly contradicted v. Frisch's expectations, because the accuracy of new-arrivals turned out to be far too high. This obliged v. Frisch (1967) to assume that dance-attendants obtain much more accurate information by averaging the information from many different dance-circuits (each containing a single waggle-run). Averaging requires arithmetic calculations.

Honey bees must perform arithmetic calculation even in order to obtain any distance and direction information from a single dance circuit. This is so, because direction information is presumably obtained from the size of the angle between the direction of the waggle-run and the upward direction (Frisch 1967). Most, and often all dance-attendants, however, attend the waggle-run at an angle. Distance information is presumably obtained from the duration of the waggle-run (Frisch 1967). Taber's statement that honey bees have a time-sense is based on the finding that they have a biological clock with a 24 hours cycles. In order to use such a continuously running clock to determine the duration of a waggle-run, the bees must read the time at the start and at the end of the waggle-run, and then subtract the first reading from the second.

There is, however, no evidence, and no reason to believe that insects can perform arithmetic calculations, or even count, (which is a basic requirement for such calculations). In a recent study Chittka & Geiger (1995) merely suggest that honey bees may have a very primitive counting ability, based on tests that included at the most 5 items to count. This is most unlikely, in view of the fact that even far higher animals like horses cannot count, (as all those who know of the case of the famous horse "Clever Hans" must admit), and in view of the authors' findings that most of their bees, by far, reacted to 2, or 4 items just as they reacted to the original 3 items used in training. At any rate, a counting ability never even tested for more than 5 times could never suffice for the arithmetic calculations required by the "dance language" hypothesis.

3. The typical manner in which recruits invariably arrive at stations, (through a low upwind zig-zag from as far as they can be spotted with the naked eye), fully fits the expectations from use of odor alone, and grossly contradicts the expectations from use of "dance language" information. All users of odor alone and only some users of "dance language" information are expected to arrive in this manner. This is so because many users of "dance language" information are expected to get much closer to stations, and especially to the forager-station, by use of this information from the direction of the hive, and not from a downwind direction. "Dance language" supporters have never been able to fit the typical manner of arrival within the "dance language" hypothesis.

Taber tries unsuccessfully to do so by claiming that the upwind zig-zag is not a response to attractive odors. He proposes that the upwind zig-zig is an attempt on the part of recruits using "dance language" information to cause the wind to slow them down, in order to avoid overshooting their goal. The proposal faces many problems. I shall, however, note only that this upwind zig-zag is the typical manner in which all flying insects invariably arrive at sources of attractive odors, and this includes solitary flying insects who cannot know where their goal is, nor what it looks like, (because they have no one to provide them with such information). Obviously, the upwind zig-zag is a response to attractive odors, and not an attempt to avoid overshooting any goal. Honey bee recruits, too, do not know where their goal is, because they cannot obtain "dance language" information. And they do not know what their goal looks like either (Frisch 1967).

4. Taber believes that the study by Bogdany & Taber (1979), done across a very deep and wide canyon, under presumably crosswind conditions, suffice to confirm use of "dance language" information. This is not the case. The results are explainable by use of odor alone, because even though all stations had the same food-odor, they were expected to each have different natural odors, due to the presence of live vegetation (shown in a photo). The effect of differences in natural odors is expected to be even much stronger than normal in this canyon study, because all other stations were very far from the forager-station, and most were also at various heights that were lower than the forager-station. (More extreme differences in vegetation types between the vicinity of the forager-station and of the other stations, were expected because of the distance, as well as the different heights.)

Moreover, wind-patterns over this complex rugged terrain is expected to have been much more complex, and very different from what the authors naively believed. It is not too difficult to find crosswind conditions in a regular, level area. Had the authors chosen to experiment in such an area, they would have spared themselves many undue problems, (including the very hard labor), and heeded the warning by v. Frisch (1967) to avoid experimenting anywhere except in a regular, level area.


Bogdany, F.J. & Taber, S. The significance of odors for bees orienting across a canyon. Apidologie, 10: 55-62, 1979.

Chittka, L. & Geiger, K. Can honey bees count landmarks? Anim. Behav. 49: 159-164, 1995.

Frisch, K, von. The Dance Language and Orientation of Bees. Harvard University Press, Cambridge, Mass. 1967.

Taber, S. Bees that dance. April: 279-280; Dancing bees-the so called controversy. May: 351-352; More on dance language of bees. June: 457-458. American Bee Journal, 1998.

Wenner, A.M. Odors, wind and colony foraging. October: 746-748; November: 807-810; December: 897-899. American Bee Journal, 1998.

Ruth Rosin
N.Y., NY