27 Knowledge, Skills and Human Potential

Concept Learning

A stimulus class is a collection of objects sharing at least one common property. For example, all circles are geometric objects with all points on the circumference equally distant from the center. Concept learning is inferred when an individual responds in the same way to all instances of a stimulus class. Much of our knowledge base consists of concepts. For example “circle” and “boy” are qualitative concepts. In comparison, “middle-sized” and “tenth” are quantitative concepts. They differ in amount, not just in kind.

Parents usually try to teach such concepts soon after their children speak. How would a parent go about teaching the concept circle and know if the child understands? When I ask my students this question, they usually suggest that the parent say the word “circle” while pointing to circular objects in the environment. You may recall our discussion of the acquisition of word meaning under the topic of classical conditioning. In this instance, the word “circle” is associated with many different stimuli sharing the property. A discrimination learning procedure could also be used to establish and assess conceptual responding to circles. The child would receive an appetitive stimulus for saying the word “circle” while pointing to appropriate examples differing in size, color, etc. The child would never be reinforced for saying “circle” to other shaped stimuli. Eventually, the child should be able to appropriately generalize the response to new instances of circles.

The same procedure could be used with quantitative concepts. When I was a graduate student, the research literature on transposition (i.e., responding to stimuli on the basis of a relationship) suggested that other animals and young children were unable to apply the middle-size relation to physically dissimilar stimuli (Reese, 1968). In my doctoral dissertation (Levy, 1975), I demonstrated near-perfect middle-sized transposition on two very different sets of stimuli by nursery-school children. First they were taught to point to three small squares (such as those shown below in Figure 1) in the order of their height before being required to select the middle-sized one. The placement was changed over trials so the child had to change the pointing order in a manner consistent with the sizes. They were then asked to order and point to the middle-sized member of three much larger squares. The results supported the conclusion that middle-size transposition occurs only when a child sequentially orders the three stimuli in an array prior to choosing the middle-sized one. After being taught to count, it becomes possible to establish relational responding based on any quantity. For example, a child could be asked to point to the fifth-largest triangle. This would require ordering all the triangles in an array based on size, and then, starting from the smallest, counting to five.

Figure 7.1  Example of stimulus arrays presented on each trial of a middle-size problem.  The three stimuli in an array appeared in random order on each trial.

Concept learning, perhaps surprisingly, occurs throughout the animal kingdom. For example, pigeons readily learn visual concepts such as “triangle” and “square” (Towe, 1954), can distinguish between letters of the alphabet (Blough, 1982), and respond to ordinal position (Terrace, 1986). Presenting slides in a Skinner box, it has been demonstrated that pigeons easily learn such abstract natural concepts as “tree”, “water”, and even “person” (Herrnstein and Loveland, 1964; Herrnstein, Loveland, and Cable, 1976). Apparently, excellent vision, not a large cortex (i.e., pigeons have “bird brains”) is necessary for learning such concepts. In a fascinating application of concept learning, Skinner (1960) humorously describes a previously- classified World War II project in which pigeons were taught to identify the defining characteristics of axis-power military ships. The objective was to respond to the invasion of Pearl Harbor with our own squadron of “kamikaze pigeons.” The pigeons became the brains behind the first non-human “smart missile”

We and the Nukak have in common many basic needs (e.g., food, water, shelter, temperature, danger, pain, etc.) and family relationships (e.g., mother, father, brother, sister, etc.). One strategy for describing and contrasting our distinct human conditions would be to study our linguistic concepts. For example, there is no doubt that the Nukak will have a much more extensive vocabulary for types of rain and types of forestry than we will. We will have more extensive vocabularies regarding planes, trains, and automobiles. When I was very young, my mother taught me “red car”, “blue car”, “green car”, etc. My father taught me “coupe”, “convertible”, and “sedan”, and eventually “Chevy”, “Chrysler”, “Ford”, etc.

Unlike the rain forest, some climates and geographies support domestication of plants and/or large animals. Such environmental conditions enabled the development of agriculture and animal husbandry, permitting humans to abandon the nomadic lifestyle. New vocabularies developed related to the essential concepts for these life-transforming activities. When humans were able to permanently settle in a location, larger and larger communities evolved. This created the need for concepts related to increasingly complex interpersonal relations. As food surpluses occurred, there were opportunities for people to dedicate their time and creative efforts to the development of new tools, technologies, and occupations. Eventually, communities, economic arrangements, governments, and formal religions evolved. Along with these developments, the collective human knowledge base and vocabulary expanded. It was after the last ice age, approximately 13,000 years ago, that the agricultural lifestyle became the predominant human condition (see Figures 7.2 and 7.3).  For the great majority, this stage of human history probably had more in common with Stone-Age nomadic cultures than our contemporary conditions. Literacy was not essential and survival needs took up most of one’s daily activities. As noted previously, this changed with the industrial revolution and the institution of compulsory education.

Figure 7.2  Sickle from 3000 B.C.

Figure 7.3  Ancient Egyptian hoe and plow.

We have seen how the ability to use speech to communicate a continually-expanding vocabulary of concepts has enabled humans to live very differently and control their fates far more than the rest of the animal kingdom. Until relatively recently, however, only the privileged acquired the ability to read, write, and perform mathematic operations (i.e., learn the 3 “R”s). This meant that the great majority of humans, even in the relatively-advanced western societies, were unable to profit from or contribute to the accumulating knowledge recorded on the written page. John Adams, one of America’s founding fathers, stated “A memorable change must be made in the system of education, and knowledge must become so general as to raise the lower ranks of society nearer to the higher. The education of a nation, instead of being confined to a few schools and universities for the instruction of the few, must become the national care and expense for the formation of the many” (McCullough, 2001, p. 364). As Adam’s call for universal education was eventually realized, an increasing number of people became literate over the past century. This created an expanding pool to contribute to the ever-evolving knowledge-base. The resulting technologies continue to transform the human condition at an accelerating pace.

Learning to Learn

The ability to transform the human condition involves more than the knowledge and skills acquired in school. This knowledge must be converted into the action necessary to solve problems and to create tools. We will begin our discussion of problem solving by describing Harlow’s (1949) classic research demonstrating chimpanzees’ acquisition of learning sets. The term “ learning set ” may be interpreted to refer to either an independent or dependent variable. It can refer to a number of experiences that have something in common or to the effect of those experiences (as in being “set up”). Harlow provided his chimpanzees with over 300 two-choice visual discrimination problems. For example, the first problem might require choosing between a circle and a square; the second problem, between a red triangle and a green triangle; the third problem, a large diamond and a small diamond, etc. Different stimuli and dimensions were relevant across the different problems. Since each problem includes only two possible choices, the likelihood of being correct on the first trial by chance was always 50 per cent. The result on the first trial provides the necessary information for an alert subject could be correct from trial two on. If correct, one would continue to choose the same stimulus; if incorrect, one would switch to the other possibility. Harlow and others described this ideal performance pattern as a “ win-stay, lose-shift ” strategy.

The chimps’ performance improved gradually over the first 30 problems, suggesting an incremental learning process. This appears qualitatively, rather than quantitatively different from the sudden, discrete win-stay, lose-shift strategy characteristic of human adults. However, over the remaining trials, the win-stay, lose-shift strategy emerged so that the performance of the chimps on the last 55 trials was human-like with perfect performance on the second trial. It appears that just as pigeons are able to learn concepts by “abstracting out” the common characteristics of a collection of visual stimuli, chimpanzees are able to “abstract out” an approach to solving two-choice visual discrimination problems regardless of the stimuli involved. They have been “set” (i.e., have learned how to learn) to solve a particular type of problem.

Problems

We frequently describe challenges in life as problems. A problem exists when there is a discrepancy between the way things are and the way one would like them to be. The solution consists of acquiring the information and ability to eliminate the discrepancy. As described in Chapter 3, many animals appear to engage in behaviors which do not appear survival-related. Kittens and infants play with toys for extended periods of time with no apparent external reward other than the sensory stimulation. Monkeys will learn a response in order to gain the opportunity to look through a window (Butler, 1953). Human adults appear to find intrinsic reinforcements in solving complex problems. How else could we understand the creation of crossword puzzles and recreational games such as chess?

Two-choice discriminations are as simple as problems get. One piece (i.e., bit) of information is all that is required to solve the problem and obtain the reward. Crossword puzzles and chess are far more complicated. Perhaps we seek complexity because such experience is adaptive. Unfortunately, many problems in life are extremely difficult to solve and to address. Issues related to health, interpersonal relationships, and finances often top the list. It would be good preparation to acquire skills and strategies that apply in such circumstances. Many have likened life to a game of chess, posing problems having many possible options and requiring extensive planning for future possibilities. In fact, some have described life as consisting of one problem followed by another.

Psychologists have studied problem-solving in humans and other animals almost since the founding of the discipline. As described previously, Thorndike studied a few different species in puzzle boxes, describing the problem-solving process as involving trial-and-error (or success) learning. In his classic, The Mentality of Apes (translated in 1925), the Gestalt psychologist Wolfgang Kohler argued that the puzzle-box, by its very nature, requires a “blind” (i.e., trial-and-error) learning process since the required behaviors cannot be determined by observing the environment. Kohler created a number of problems for his subjects, primarily chimpanzees, in which the solution could be grasped by observing the environment. He considered such problems to be more representative of those we confront on a day-to-day basis.

One famous example of Kohler’s problems required that the chimpanzee insert a thin bamboo stick within a wider one. This created a tool long enough to reach a banana outside the cage. Another problem required stacking boxes high enough to reach a banana. A third required combining sticks to reach a banana hanging from above. The following classic video of Kohler’s research demonstrates individual and collective (i.e., social) problem-solving by his chimps with these tasks. Kohler amusingly anthropomorphized, attributing human characteristics to his subjects. He described the chimp’s initial frustration resulting from unsuccessful attempts.  Under circumstances where the necessary components of a solution are observable, Kohler characterized the problem-solving process as requiring “ insight .”  You may recall that Gestalt psychologists primarily studied perceptual phenomena (e.g., the phi phenomenon). It is not surprising that Kohler considered insight to be a perceptual process requiring reorganization of the perceptual field in order to attain closure.  Presumably, the chimp continued to scan the environment until attaining the specific insight required to solve the current problem. Wertheimer (1945) later published a “how to” book based on Kohler’s work, extending Gestalt concepts to childhood education.”” when the chimps performs the behaviors necessary to obtain the banana.

Facilitative Effects of Prior Experience

Other researchers believed that Gestalt psychologists under-emphasized the role of prior experience in problem-solving. The subjects in Kohler’s primate colony were reared in the wild, not in captivity. Since bamboo sticks were prevalent in that environment, it was likely that the chimpanzees had handled them previously, increasing the likelihood of solving the two-stick problem. Birch (1945) provided five previously unsuccessful chimps with sticks to play with for three days. They were observed to gradually use the sticks to poke, shovel, and pry objects. When again provided with the two-stick problem, all five chimps discovered the solution within 20 seconds, demonstrating the importance of prior experience.

Based on Harlow’s observation of learning to learn, one can imagine Kohler’s chimps entering their cages, looking for the banana, and asking themselves “OK, what does Kohler want me to do today?” In a humorous simulation of the box climbing problem (Epstein, R., Kirshnit, C. E., Lanza, R. P., & Rubin, 1984), pigeons needed to move a box under a plastic banana and then step on the box in order to peck the banana to receive food. Some pigeons were shaped to move the box to wherever a spot appeared on the floor, others were shaped to stand on the box and peck the plastic banana, and a third group was taught both responses. Only the group taught both components of the required behavior displayed “insight”, confirming the importance of prior learning experiences in problem-solving (see video).

Interference Effects of Prior Experience

Prior experience can impede, as well as facilitate, problem-solving. Luchins (1942) gave college students a series of arithmetic problems to solve (see Figure 7.4). They were asked to provide the most direct way of obtaining a certain amount of liquid from jars holding different quantities.

                                          Figure 7.4  Luchin’s water jar problems.

In an example, subjects were first shown how it was possible to obtain 20 units of water by filling a 29-unit container and spilling 3 units into a separate container, three times (i.e., 29–3*3, or A-3B). After the example, a control group was administered problem 6 that could be solved using two (the direct solution, A-C) or all three jars (the indirect solution, B-A-2C). An experimental group was provided with five problems that could be solved with the B-A-2C formula prior to being administered the last problem. The experimental subjects were much more likely than the control subjects to use the less-efficient indirect method. This sort of “rigidity” is counter-productive. Thus, although it is often helpful to rely upon past experience in approaching problems, there is also value in considering each problem separately. Otherwise, we may be very unlikely to “think outside the box.” Figure 7.5 shows the solution to the well-known 9-dot problem in which one is instructed “Without lifting your pencil from the paper, draw exactly four straight, connected lines that will go through all nine dots, but through each dot only once.” Here, the solution literally requires thinking outside the box.

Figure 7.5  “Think outside the box.”

A special case of being blinded by past experience has been demonstrated with the use of physical objects, a phenomenon called “ functional fixedness .” Dunker was one of the pioneers investigating functional fixedness in humans. One of the tasks he created required using several common objects in an unfamiliar way to create a “candle box” (see Figure 7.3).

Figure 7.6  Functional fixedness.

In another example of functional fixedness, Maier and Janzen (1968) found that college students were much more likely to use some objects rather than others to tie strings suspended from the ceiling together. For example, they were more likely to tie a ruler to the bottom of a string than a bar of soap. Presumably, the usual function of soap interferes with consideration of it for another use, even within a different context. This effect was demonstrated experimentally by Birch and Rabinowitz (1951). Two groups of college students were provided experience using two different objects to complete electrical circuits. Subjects were far more likely to use the unfamiliar object as the weight when they were given the two-string problem to solve

The Gestalt psychologists emphasized the tendency to perceive objects as meaningful wholes. Functional fixedness appears to be an inevitable result of this tendency. An implication of this perspective is that requiring individuals to describe the parts of objects should reduce the likelihood of functional fixedness. This was found to be the case when college students were asked to engage in a task similar to the introspection procedure employed by the structuralists to analyze conscious experience (McCaffrey, 2012). Subjects were asked to break down objects into their component parts without consideration of how they were used. This reduced the occurrence of functional fixedness.

The Unusual Uses Test (Guilford, Merrifield, and Wilson, 1958; Guilford and Guilford, 1980) is a popular assessment of creativity based upon the concept of functional fixedness. One is asked to list as many uses as possible for different objects (e.g., “What can you do with a brick?”). Responses may be counted or scored for originality. It is conceivable that encouraging test-takers to break down objects into their component parts could increase creativity scores on this test.

The General Problem-Solving Process

A general problem-solving process including five distinct stages has been described. The stages are: (1) general orientation; (2) problem definition and formulation; (3) generation of alternatives; (4) decision making; (5) verification (Goldfried and Davison, 1976, p. 187). The general orientation stage encourages individuals to approach situations eliciting unpleasant emotions as problems. Problems relating to health, interpersonal, and financial matters can be devastating, possibly resulting in debilitating anxiety and/or depression.

Weight-control is a common health concern. When I consulted for a medically-supervised weight clinic, I encouraged a self-control approach to fitness and health. Frequently, emotionality related to unrealistic societal ideals for appearance interfered with a client’s adhering to a prudent lifestyle. It was helpful to reduce the emotionality related to one’s appearance by adopting a problem-solving approach to weight control and body shape (Stage 1). The problem was defined as a discrepancy between one’s current weight and dimensions and a more desired profile (Stage 2). This permitted a relatively-detached brainstorming discussion of different nutritional and exercise modifications designed to affect caloric input and output (Stage 3). The likely benefits and drawbacks of implementing the different approaches were discussed with the goal of deciding upon a strategy that could be sustained (Stage 4). The decided upon strategy was implemented, with objective (weight and measurements) and subjective (ease of implementation, satisfaction, etc.) progress consistently monitored (Stage 5). A TOTE (Test-Operate-Test-Exit) approach was implemented to determine the need for fine-tuning or changing the strategy (Miller, Galanter, and Pribram, 1960). Similar to a thermostat, the individual would test the environment (i.e., determine current weight and measurements), operate on the environment (i.e., “turn on” the nutritional and exercise program), and continue to assess progress until achieving the desired objective. This same process would be sustained in order to maintain the desired end state.

The same “thermostat” approach could be applied to financial matters. The problem could be defined as a discrepancy between a family’s income and expenditures. Brainstorming would be conducted to list possible ways to increase income or reduce costs. A strategy would be decided upon, implemented, and continually assessed. Adoption of a problem-solving approach is particularly helpful with interpersonal problems, which are almost always emotionally charged. It is difficult, but possible, to teach individuals or couples to respond objectively to the substance of what someone says while ignoring provocative language. Once this is achieved, difficulties and solutions can be mutually defined and strategies for addressing them can be negotiated prior to implementation and assessment (D’Zurilla and Goldfried, 1971).

Attributions

Figure 7.2 “Sickle from 3000 BC” by Wikimedia Commons is licensed under CC BY 2.0

Figure 7.3 “Hoe and plow” is in the Public Domain, CC0

Figure 7.5 “Nine dots” by Marmelad is licensed under CC BY-SA 3.0

Figure 7.6 “Functional fixedness” by Karl Dunker is in the Public Domain