Unlearned behaviors are "built-in" in some sense. Either they unfold via the process of maturation, or they manifest themselves at some stage in life in all or nearly all members of the same species. In other words, unlearned behaviors are more consistent in form and timing of development than learned behaviors. Learned behaviors must be acquired. Learned behaviors will exhibit a wider variability than unlearned, and will not be as universally distributed as unlearned behaviors. Some examples of each will help to clarify these distinctions.
The three-spined stickleback is a small, freshwater, European fish. It lives in small streams. Males of this species build nests during the mating season. They then defend the cylinder of water above the nest from other males. Males develop red bellies during mating season, and that is the primary cue that other males use to distinguish males from females. In fact, Tinbergen showed that even crude fish models with red bellies were attacked by resident males, but accurate models without red bellies were not. Females do not develop red bellies. When a female enters a male's territory, he engages in the "zig-zag" dance. If a female is receptive, she will follow the male to his nest, lay her eggs inside, and leave. The male will then fertilize the eggs, fan water through the nest with his tail to prevent rotting, and continue to defend the nest. The point of this long-winded example is that ALL normal male sticklebacks will engage in this kind of behavior WITHOUT having to learn it. (Papa stickleback does not have to explain to Junior stickleback about the birds and the bees.....) The above is an example of a complex unlearned situation.
The digger wasp demonstrates a situation in which there are both unlearned and learned components. These digger wasps live near the coast in the Netherlands. They hunt for caterpillars, stun them, lay eggs in them, and bury them in small burrows. All of those behaviors mentioned are unlearned, meaning that all normal digger wasps will engage in such behavior without having to learn it. By the way, the caterpillars are put into a state or torpor or immobility by the sting. The young wasps hatch and then eat their way out of the live but immobile caterpillar. The learned part of this example comes in navigating from the area in which the caterpillars are found to the various burrows each wasp maintains. Each burrow is visited several times before it is finally sealed, and each wasp maintains several burrows. The Baerends were able to demonstrate that the wasps were navigating by learned cues. They did so by moving the landmarks near burrows. When the wasps returned, they landed where the burrow should have been. Eventually, however, they learned the new location. This example gives us an idea about a function of learning, namely, to help animals deal with environments that change. The beach will change daily, but caterpillars have been the same for millions of years.
The final example involves human and monkey parenting. An interesting question is, "How much of human parenting is learned?". Unfortunately, we cannot run the experiment to answer that question for ethical reasons. (e.g.,. Take a number of children, raise them without parents....) Harlow, however, was able to run just such an experiment with rhesus monkeys. Harlow separated some mothers and their offspring shortly after birth. Some monkey infants were reared in total isolation, others with "surrogate mothers," others with other monkey infants, and still others were not separated from their mothers. The infants raised in total isolation were completely abnormal. The infants reared with surrogate mothers fared somewhat better, and they developed an attachment to the soft, warm surrogate mother. Later, the males reared in this condition could not become fathers because they were unable to copulate with females. Females here did copulate successfully, but made horrible mothers. Both males and females in the group that was reared with other monkeys could copulate and made adequate, if somewhat below average, parents. In rhesus monkeys, then, parenting skills are learned. They are in humans as well. Many of us have had experience with infants prior to giving birth to our own. However, the experiences surrounding the birth and raising of the first child are the major time of learning for most parents. Subsequent children are treated very differently because of those experiences. Interestingly, our society requires no demonstration of capacity to parent. But, society requires demonstration of capacity to drive an automobile. Lately, some have argued that some formal training in parenting should be required. Informal avenues for training exist, such as LaMaze training and intervention programs for some parents. The research on child abuse is instructive here. Child abusers are much more likely to have been abused children themselves, giving rise to a cyclical hypothesis of child abuse. Intervention programs aim to break that cycle.
So, understanding learning provides insight into much of human behavior because so much of human behavior is learned.
- Lectures in Learning--index, basic, short, links, graphics
- From the home page of an instructor; provides basic, referenced lectures with graphics: evolution and animal intelligence, the reflex from Descartes to Pavlov, basic concepts in classical conditioning, cognitive models of associative learning, what is learned in classical conditioning, trial and error, from the rise of Thorndike to the fall of J.B. Watson, operant conditioning, generality, constraints and concepts in learning. http://www.dur.ac.uk/~dps1rwk/