Dolphin echolocation and cross-modal shape recognition abilities

Animal Senses

Herman, Pack and Hoffman-Kuhnt performed relatively rigorous experiments to determine the source of dolphin recognition of objects; they wanted to discover, among other things, whether "dolphins attained the shape discriminations (of objects) through associative learning or direct perception" (Herman et al. 1998 292). Fukuzawa, Mills and Cooper sought to determine the mechanism by which domestic dogs responded to commands. Greenberg wanted to discover the facts about depth perception in two species of Asian rodents, the Mongolian Gerbil and two varieties of Spiny Mice.

Dolphins

The experiments run by Herman et al. involved a single dolphin, a female named Elele, and were designed to determine whether echolocation or visual cues were central to dolphin recognition of objects that appeared in their environment. The researchers were extremely rigorous in setting up each experiment, avoiding contamination between visual and echolocation fields; the objects used for the dolphin's recognition tests were never available for both visual and echoic inspection at the same time.

When the trials had begun, Elele chose correctly in 49 of the first 50 trials at recognizing the objects introduced. The researchers noted that their results demonstrated an "impressive capability for direct cross-modal shape recognition" (Herman et al. 1998 298).

These were simple trials, however, with the dolphin being allowed to make her identification of the object immediately upon experiencing it. More advanced tests involved delaying her decision; There too, however, Elele excelled, responding correctly 99.2% of the time in the delayed recognition trials.

The experiments were sufficiently sophisticated that the dolphins reaction/response times were also measured. From the dolphin's responses in this mode, the researchers concluded that the dolphin required a bit more time to reach a decision when the objects introduced were less familiar and more complex than the objects in the original tests.

The results of all the experiments, both the simple ones and the ones involving time delays, suggested to the researchers that "dolphins possess a fundamental ability for shape perception through their echolocation sense." (Herman et al. 1998 303). On the other hand, it also became clear that the dolphin employed her visual sense when needed, leading to the conclusion that dolphins view objects holistically.

Gerbils and mice

Greenberg was at pains to note that it was "experimentally naive" female Mongolian gerbils that were used in this experiment; possibly he wanted to clarify that they had had no other experiences regarding mazes or food or any of a number of other behavioral issues a researcher might have wanted to address.

The investigation itself was concerned with finding out if the gerbils and mice used tactile or visual senses in determining actions to take; in this case, climbing down from a platform. The experiment was designed so that the gerbils and/or mice could use either one or both senses in determining whether to climb down; a failure to climb down was assumed if the animal had not, in fact, climbed down after a period of two minutes.

Some of the conclusions seem almost too simple to note, but Greenberg did note them. One of these is that spiny mice, descent time increased as the height of the platform increased. Another finding that seems so simplistic as to be unworthy of mention is that there were subtle differences between the descent times of gerbils and spiny mice, and between the two different kinds of spiny mice as well.

While the experiment was supposedly designed to determine whether or not gerbils and spiny mice used depth perception as an aid to climbing down from high spots, the conclusion arrived at by Greenberg was that the difference he observed might be due to differences in visual information processing between species and not to depth perception differences per se.

In addition, the experiments were designed so that any natural environmental clues would be removed. In fact, virtually any clues were removed in some cases, with a pure white ground used to prevent the animals from gaining clues from texture and so on; in other experiments, a patterned ground, but one unlike those found in the natural habitats of the animals, was used.

Finally, Greenberg concluded that his experiments demonstrated "a difference between gerbils and spiny mice with respect to her utilization of an dependence on sensory information" (1986 83). This seems a very insignificant conclusion considering the time and effort expended.

Commentary

Of these three experiments, the two that were most like each other were the dolphin and gerbil/spiny mice experiments. In both of these, non-domesticated animals were used as subjects and the experiments themselves allowed for substantial manipulation of the test elements by the researchers, as well as opportunities for gathering quantitative data that could later be analyzed for correspondence to a set of assumptions.

On the other hand, the dog experiment seems very different. First, in involved a domesticated animal with whom humans interact daily. A few humans interact frequently with dolphins; it could reasonably be said that no humans interact frequently -- at least by choice -- with mice and gerbils.

However, in some ways, the dog and dolphin experiments are more similar to each other than either is to the gerbil/mice study. While dolphins are not domesticated animals, for thousands of years, mankind has been fascinated by dolphins and, in the past half century especially, has come to regard them as sort of the 'humans of the sea,' or, in other words, an animal life-form with a high degree of intelligence. At times, discernment is even attributed to them, and other 'human' mental qualities. Likewise, we have certainly anthropomorphized dogs, despite their apparent lesser intelligence than dolphins. We depend on canines to perform certain tasks, in fact, from bringing in the newspaper to guiding blind humans to guarding cattle. Humans have been fascinated by the prospect of having dolphins carry out tasks for us in the sea, as well.

Taking each experiment by itself, however, makes it clear that none of these studies offers as much information as it probably intended to do.

The dolphin experiment, perhaps the most thoughtful of the three, did conclude -- however shakily -- that dolphins are capable of holistic recognition, that it isn't only visual clues or only their ability to perform echolocation that allows them to correctly identify objects introduced to their environment. The research noted that dolphins displayed and ease of recognition of objects. Moreover, they found that prior experience of any particular type was not necessary for the dolphin -- this dolphin, at least -- to perform well on new tests. In short, the dolphin turned out to be smart enough to assess new situations instantly and use whatever skills/abilities it needed to obtain correct information and further, to use its prior training to let the researcher know its answer. Still, the question must be asked: Is there a point to this experiment? Will knowing how a dolphin 'sees' objects bring us any closer to an understanding of dolphins that is usable for human purposes? Or was this pure science?

Perhaps a continuation of the experiment could involve the dolphin -- and one assumes it would have to be the same dolphin -- performing some actual task that used its combined visual and echolocation abilities. Still, the dolphin experiment seems to have merit because the dolphin has already proven its usefulness in certain Naval experiments, and in myth and lore, and it would be helpful to ascertain whether the dolphin's abilities hold up under the stress of human direction for human purposes.

The canine experiment was ludicrous. Dogs are and have been for centuries helpers for humans, if only as companions in some cases. To attempt to find out whether the dog is responding to a particular set of sounds, the entire word, the setting of the word, the environment in which the word is used, or the voice of a known or unknown person giving the command seems to be the most completely meritless investigation of the three studied here. In fact, dogs are virtually always commanded by a human who either knows the dog, or is in an emotional condition in relation to the dog, for example, if a stranger tells a dog to come so that it won't be hit by a car. Thus, experimenting in the absence of the ordinary and normal environment of the dog seems completely useless. This seems like weird science, for perhaps no more reason than to obtain grant money and spend it. It is doubtful that the lives of dogs or humans was enhanced by the experiment, and the only logical sequel to it would be no sequel at all.

The gerbil/spiny mice study was only marginally more useful than the canine experiment. Whether a mouse or gerbil uses his vision or his sense of touch, or both (or neither) in climbing down from a platform is difficult to explain as either a psychology or physiology investigation. Even the researcher noted that his results were probably relatively meaningless in view of the fact that gerbils, who live mainly underground, would have precious little need for the skills he was testing. In that respect, it might be concluded that gerbils succeeding at all is noteworthy.

The researcher himself admitted that it is difficult to make "valid generalizations about he behavior of theses species in their natural habitats on the basis of experiments that present the animals with artificial tasks." (Greenberg 1986 83). He suggest that a subsequent experiment be conducted in the natural habitat, keeping "behavioral repertoires" in mind (Greenberg 1986 83). If one were to design a further experiment, then that would seem the best avenue to explore; the problem is, of course, that it would probably eliminate the gerbils from the experiment, as they do live primarily underground and are not presented naturally with the need to climb down from platforms.

Two of these three experiments are consistent: the gerbil/mice and dog experiments are superfluous, telling us very little about either the lives of the animals in question or anything germane to human life. Two of the experiments are similar in methodology; the gerbil/mice and dolphin experiments both pose artificial tests, although at least the dolphin experiment promises some small shred of useful dolphin knowledge might be developed. Of the three, the only one that seems at all worthy of continuation, expansion or development into other investigations is the dolphin experiment although, at the very least, similar tests would need to be conducted with a number of other 'domestic' dolphins and, preferably, with a control group of wild dolphins as well.