INTRODUCTORY TEXTBOOKS: PROBLEMS
Kohler's Insight Revisited
P. A. Lamal
The University of North Carolina at Charlotte
Wolfgang Kohler's two-stick experiment—involving a chimpanzee joining two sticks and then raking in a bait placed outside its cage—was published about 60 years ago (Kohler, 1917/1925). This particular study is still described in many American psychology textbooks (e.g., introductory, learning, and history and systems). It is almost invariably presented as an example, par excellence, of the Gestalt notion of insight, as opposed to Edward L. Thorndike's idea of trial-and-error learning. The student reading the textbook accounts of Kohler's insight studies might conclude that the concept of insight was well established and thoroughly confirmed. We contend that textbook writers' dissemination of such a conclusion is unwarranted. We also note that the issue of insight is still alive, as in the current reinterpretation of the notion by radical behaviorists (e.g., Epstein, 1981; Epstein, Kirshnit, Lanza, & Rubin, 1984).
An unsystematic survey of 19 recently published introductory psychology textbooks indicated that Kohler's insight studies are referred to in 17 of them; the two-stick experiment was mentioned in 8 of the books. Of the 10 learning textbooks examined, Kohler's insight studies are referred to in 8; the two-stick experiment is cited in 5. Kohler's insight studies are mentioned in every one of the 7 history and systems texts that were sampled, with the twostick experiment cited in 5.
Köohler's two-stick study was first performed between 1913 and 1917 as part of the general research project on the intelligence of anthropoid apes at the Anthropoid Station in Tenerife, Canary Islands. Kohler's lntelligenZPriifung an Menschenaffen was published first in 1917, and its English version, The Mentality of Ages, appeared in 1925.
The description of the two-stick study was mainly based on the observation of the behavior of a chimpanzee named Sultan by Kohler and the animal's keeper. The problem confronting Sultan was to obtain a bait located outside his cage with the aid of two bamboo sticks of equal length but different thickness. Problem solution required that the thinner stick be inserted in the hollow end of the thicker stick making a stick long enough to reach the bait. At first, a number of unsuccessful behaviors were observed. These included pushing a box toward the bars, using one stick to touch the bait, and pushing one stick with another toward the bait. According to the keeper's report, Sultan finally sat on the box, played with the two sticks, held a stick in each hand, pushed the thinner one in the hollow end of the thicker one, moved to the bars, and pulled in the bait with the joined sticks.
Köhler interpreted the solution in terms of an insight that, on a perceptual level, took into consideration the relation of the elements to each other. This interpretation contrasted with Thorndike's understanding of problem solutions as the outcome of a fortuitously successful behavior out of many other behaviors the animal had exhibited. The main behavioral distinction between Kohler's insightful solution and Thorndike's trial-and-error explanation was seen in the relative suddeness with which Sultan solved the problem (about 5 min), as contrasted with the relatively slow solution typically obtained in trial-and-error learning.
The insight interpretation of the two-stick experiment is justified provided there is sufficient evidence to show that chimpanzees solve the problem according to the criteria of insightful learning, such as suddeness of the solution. Before concluding that the chimpanzee had demonstrated insight, Kohler's experiment should be replicated and controlled. Our aim is to consider whether subsequent studies supported Kohler's interpretation of the two-stick experiment.
In an attempt to resolve the issue of whether the solution of complex problems is insightful or achieved by trial-anderror, Pechstein and Brown (1939) replicated the two-stick experiment. The subject, Romeo, was a male chimpanzee, about 4% years of age. The first trial of the two-stick problem lasted 30 min and ended with failure. It was observed that Romeo tried to pull in the bait using either stick separately, and that he played with the two sticks attempting to insert one stick into the other.