The fifth chapter, “Learning” describes and evaluates cognitive processes and learning perspectives. Paradigms of classical and operant conditioning are examples of primitive learning operations in that they specify the temporal configuration of a set of events which result in learning, but they are, in most theories, not derived from other conditioning principles. In specifying primitive learning operations, S-R theories include a response as one of the events that must stand in a particular temporal relation with other events in order for learning to occur. The response appearing in the primitive learning operation is then the response appearing in statements about behavior deduced from this theory. The gap between the theory and behavior is easily bridged, and the deduction of behavior is determinate. A “learning operation” is a set of events resulting in a change in behavior called “learning.” A learning operation is “primitive” if a theory presents it as a given rather than deriving it from some more basic behavioral principle (Flavel et al 2001). When the learner then emits a response that matches that of the model, a reinforcing consequence often follows also. Therefore, through operant conditioning (involving a response), the subject learns to emit the response match-the-behavior-of-the model to instances of the discriminative stimulus a-model-emitting-a-reinforced-behavior. Note that the terms of the contingency are rather abstract functional classes. The functional response match-the-behavior-of-a-model consists of highly diverse movements.
Because virtually all changes in behavior, even the most elementary, involve classes of behavior and classes of environmental events, it is nearly always possible to find systematic relationships among these classes and to construct an intervening variable. Therefore, nearly any behavior change can be conceptualized in cognitive terms. Consider a simple case of operant conditioning. A rat presses a lever, a pellet of food is delivered, and the rat then presses the lever at an increased rate. Although powers of ratiocination are not usually attributed to rats, it is nevertheless possible to interpret the rat’s behavior with cognitive intervening variables because the changes in behavior due to operant conditioning are widespread, and the independent variables are numerous. One question dividing cognitivists and behaviorists is whether these concepts are desirable (Lezak et al 2004). Cognitivists insist that cognitive concepts are important for understanding behavior while behaviorists, for the most part, find these concepts unacceptable. In the terminology of operant conditioning, a problem is the discriminative stimulus for the precurrent response, and the solution to the problem is the reinforcement at the end of the behavioral chain of which the precurrent response is a link. Given the behaviorist interpretation of purpose in terms of reinforcement, this formulation of precurrent behavior can be transformed into purposive language: Precurrent behavior is behavior which occurs for the purpose of solving a problem. It is identified by its role in the adaptation of the organism to its environment (Flavel et al 2001).
These differences in emphasis give rise to disagreements over standards of adequacy in explanation. Cognitivists insist that an adequate explanation must refer to internal features of the organism, features which furthermore are contemporaneous and in correspondence with behavior. Cognitivists thus manifest an “internalism” in counterpoint to behaviorist “externalism”. Behaviorists prefer to explain behavior in terms of the environment, including dispositional properties the environment possesses only by virtue of its previous interactions with the organism. To achieve adequacy for behaviorists, an explanation must ultimately relate behavior to these features of the external environment (Lezak et al 2004).
Because of these disagreements over standards of explanatory adequacy, the difference between behaviorists and information-processing theorists, like all debates between behaviorists and cognitivists, extends throughout all behavior. Explanations within behaviorist molar theories rest ultimately upon a few fundamental behavioral principles, such as the Law of Effect, or the principles of classical and operant conditioning. They refer to relationships among stimuli and responses, and they are “primitives” in that the system offers no further principles to account for them. For cognitivist internalism, such explanations are inadequate because they do not refer to changes in the organism. Therefore, neomentalist theories explain the behaviorist primitive principles as the result of information-processing operations occurring within the organism. Because the information-processing theory accounts for what is left unexplained in molar behaviorist theories, cognitivists claim that the former achieves greater explanatory adequacy. The major difference between Bandura and Gewirtz reduces to the question of whether learning through modeling can be conceptualized as, or derived from, learning by response consequences–i.e., operant conditioning (Thorne and Henley 2004). Because Gewirtz’s operant explanation of modeling can be generalized to all observational learning, the contrast between his theory and that of Bandura illuminates the characteristics of S-R theories of learning. requently people’s reactions to the events taking place about them are heavily dependent on their beliefs, thoughts, and feelings about those events, even when objectively unfounded. A hydrophobic patient, for example, finds water threatening. Here there seems to be a more compelling argument in favor of internal representations (Lahey, 2001).
In learning, thought is a hypothetical construct postulated to consist of covert events occurring within the body. They intermediate between the environment and overt behavior, so as to improve the organism’s adaptation to its world. Behaviorists tend to restrict the properties of their hypothetical constructs in certain respects. Covert events are assumed to conform to the behavioral laws governing overt stimuli and responses and to be lawfully linked to observable dependent and independent variables (Flavel et al 2001). They thus operate under the functional control of external variables. To the extent that these conditions are violated by a theory, mediators acquire functional autonomy. Mediators not under the control of environmental stimuli and obeying laws different from those of observable behavior hardly deserve to be called “stimuli” and “responses.” Long interacting chains of mediators of this sort have only very tenuous ties to the concept of behavior (Lahey, 2001).
In sum, learning processes which have lost their behavioral moorings can be assigned functions isomorphic to the information-processing operations of automata theory, even though they continue to be called “stimuli” and “responses.” The important question is whether precurrent behavior is a sufficient condition for thought; whether behavior itself ever deserves the designation “thinking”. Suppose, for example, in the illustration above, the car owner slams the hood in disgust, and this action somehow exposes the defective engine part so that the person can now fix the car. Surely, the behavior of slamming the hood, although it changes the environment so that later behavior will be reinforced, is not thought. We are tempted to conclude that thought is something that lies behind the behavior and cannot be the behavior itself. Cognitivists within the information-processing approach do not view their differences with behaviorists as a mere disagreement over preferences in theoretical constructs.
References
Flavel, J. H., Miller, P. H., Miller, S. A. (2001). Cognitive Development. Prentice Hall; 4 edition.
Lahey, B. B. (2001). Psychology an introduction. Prentice Hall; 11th edition.
Lezak, M. D., et al. (2004). Neuropsychological Assessment. Oxford University Press, USA; 4 edition.
Myers, E. et al. (2006). Social Psychology (6th Edition). Prentice Hall; 6th edition.
Thorne, B. M., Henley, T. (2004). Connections in the History and systems of Psychology Wadsworth Publishing; 3 edition.