Behaviorists define learning as a change in behavior brought about by the environment; some deny the existence of mental events altogether, while others concede that mental events might exist, but that they cannot and should not be studied. Behaviorism spans decades, and many individuals have made significant contributions to its development. Two key individuals in the field, Ivan Pavlov and B. F. Skinner, developed classical and operant conditioning theories which can be applied to education. While behaviorism contributed greatly to our understanding of human learning, most now believe it is insufficient for explaining more complex behavior. Thus, behaviorism has largely been supplanted by cognitive theories of learning which focus on the very thing behaviorists were accused of ignoring – the mind.
Keywords Classical conditioning; Extinction; Operant conditioning; Pavlov, Ivan; Punishment; Reinforcement; Response; Shaping; Skinner, B.F.; Stimulus; Watson, John
Educational Theory: Behaviorism
Although many people associate behaviorism with the work of B.F. Skinner, it was John B. Watson who coined the term, and who first introduced behaviorist principles into mainstream American psychology. Around the turn of the twentieth century, people began putting their faith in science as the way forward to a better future (Harzem, 2004). Watson shared in this optimism, and suggested that psychology – like the natural sciences such as physics and biology – should become a science as well. In order to do so, he argued, psychologists should study only that which is observable, and turn away from the study of consciousness and methodologies like introspection. In a paper published in 1913 called "Psychology as the Behaviorist Views It," Watson wrote:
Psychology as the behaviorist views it is a purely objective experimental branch of natural science. Its theoretical goal is the prediction and control of behavior. Introspection forms no essential part of its methods…The behaviorist, in his efforts to get a unitary scheme of animal response, recognizes no dividing line between man and brute (as cited in Harzem, p. 6, 2004).
The end of the story is well-known. By denying the existence of mental events – Watson even denied the existence of the mind itself – behaviorists left themselves exposed to attack. And inevitably, the 1970s ushered in a new era of psychology – often called the cognitive revolution – whose subject of study was exactly that which the behaviorists had ignored – unobservable mental events, or what behaviorists refer to as 'the black box.' Behaviorism wasn't necessarily 'wrong' in any fundamental sense, cognitive psychologists argued, but it was incapable of explaining complex human behavior. Thus, behaviorism was edged out of the spotlight, but its principles still hold sway, and its impact continues to be far-reaching. As Harzem (2004) writes, "now behaviorism is like a cube of sugar dissolved in tea; it has no major, distinct existence but it is everywhere. It is an essential ingredient of scientific-psychological thought, whether psychologists wish it to be or not" (p. 12).
But any good story has more than a beginning and end. Behaviorism's contribution to human learning and development is immense, and so it is to the 'stuff in the middle' that we now turn – to the insights of behaviorists and to the theorists themselves. Before doing so, however, we need to take one small step backwards, for although behaviorism became largely an American venture, it began not with Watson in America but in what might seem an unlikely place – in the laboratory of a Russian scientist studying salivation reflexes in dogs.
Ivan Pavlov, a Russian physiologist, stumbled across one of the two major principles of learning that now characterize behaviorism. His research was designed to uncover the neural mechanisms associated with digestion; while conducting his experiments, however, he noticed that his subjects, the dogs, began salivating not just in response to the food, but also in response to other environmental cues, such as the lab attendants who brought the food. As Mazur (1994) writes, "Pavlov recognized the significance of this unexpected result, and he spent the rest of his life studying this phenomenon, which is now known as classical conditioning" (p. 58).
Let's look at the components of classical conditioning by dissecting one of Pavlov's first attempts to study the phenomenon. Pavlov began with what he called a neutral stimulus (NS) – in this particular case, a bell. When presented with the ringing of the bell, the dogs did, virtually, nothing. Pavlov then paired the ringing of the bell with the presentation of the food; he referred to the food as the unconditioned stimulus (UCS) because it elicited an unconditioned response (UCR), salivation. After several pairings of the bell and food, Pavlov then presented the ringing of the bell alone, at which time the dogs began salivating. The bell became a conditioned stimulus (CS), the salivation in response to the bell, a conditioned response (CR). This type of learning is also referred to as signal learning, because it is most effective when the conditioned stimulus is presented just before the unconditioned stimulus (Ormrod, 1990). It has been replicated in humans and animals alike with a variety of reflexive responses, such as blinking, galvanic skin response, and taste-aversions (Mazur, 1994).
Although the formula for classical conditioning is relatively simple, a number of corollary explanations of behavior evolved from it. Psychologists began to investigate how a conditioned response could be extinguished, why certain conditioned responses occurred in the presence of some stimuli and not others, and how classical conditioning could be applied in real-world settings.
Psychologists discovered that the passage of time has little effect on the strength of a conditioned response. That is, if a day, or week, or year passed before a dog were presented with the conditioned stimulus (the bell) again, the dog would still salivate at its sound (Mazur, 1994). What then, they wondered, would cause a subject to 'unlearn' such a response? Through a process called extinction – the presentation of the conditioned stimulus without the unconditioned stimulus, or in this case, the bell without the food – the conditioned response gradually disappears.
The question then arises, is the dog whose conditioned response has been extinguished the same as a dog who was never conditioned in the first place? That is, is the association between the conditioned and unconditioned stimulus permanently erased through extinction? A phenomenon known as spontaneous recovery suggests the association remains intact, although weakened. Dogs who were conditioned on Day 1, for example, and extinguished on Day 2, displayed the conditioned response again on Day 3 even though the conditioned response had been fully extinguished on the previous day. Psychologists disagree about what causes spontaneous recovery, but the phenomenon itself has been well documented (Mazur, 1994).
Rapid reacquisition also suggests that the process of extinction does not return an organism to its pre-conditioned state. Dogs who learn to associate the ringing of the bell with the presentation of food, and whose conditioned response is then extinguished, will re-learn the pairing of the two stimuli during a second phase of acquisition much more quickly than they learned it during the first phase.
Organisms will sometimes display a conditioned response when presented with a stimulus that is similar to, but not exactly the same as, the original conditioned stimulus. Such a phenomenon is known as stimulus generalization. Pavlov's dogs, for example, might salivate at the sound of a second bell that rings at a different but similar frequency as the first bell.
On the other hand, organisms can be explicitly 'taught' to discriminate between two stimuli. If Pavlov repeatedly paired a low-pitched bell with the presentation of the food, but did not pair a high-pitched bell with the presentation of the food, the dogs would learn to salivate at the sound of the first, but not the second.
In some cases, a stimulus that is never directly paired with the unconditioned stimulus can elicit the unconditioned response. For example, after dogs learned the association between the bell and food, Pavlov then began pairing the bell with a light flash, in the absence of the food. Dogs soon began salivating in response to the light flash alone, which they learned to associate with the bell, which they had previously learned to associate with food.
Extinction is sometimes not a reliable way to extinguish conditioned responses (Ormrod, 1990). The rate at which extinction occurs is often unpredictable, and finding opportunities to present the conditioned stimulus without the unconditioned stimulus is often difficult. As a result, psychologists suggest that counterconditioning may be a more effective may to change behavior. In the classic case of "Little Peter" (Ormrod, 1990), a young boy somehow learned to be afraid of rabbits. By giving Peter candy at the same time he was in the presence of a rabbit, the conditioned response elicited by the candy – pleasure – began to replace the conditioned response elicited by the rabbit – fear. Since pleasure and fear are incompatible responses, Peter couldn't experience both at once; gradually, his fear of rabbits disappeared.
Classical conditioning is just one of two theories of learning that characterize behaviorism. The second, known as operant conditioning, was developed by B.F. Skinner in the 1940s. Although both Pavlov and Skinner are considered behaviorists, they disagreed with one another. An editorial review of a talk given by Skinner at the dinner of the Pavlovian Society in 1966, for example, states that "Although very gracious, polite and deferential, Skinner implied that Pavlov was actually riding the wrong horse when he suggested that conditional reflexes could serve as a window to learned behavior. Skinner, of course, held to the unique power of the operant theory" (Skinner, 1996, p. 1).
Behaviourism is a theory of stimulus and response. The emphasis here is on modifying behaviours; and internal mental states or states of consciousness are considered to be of little importance. They are not considered relevant to the idea or practice of learning. The learner is passive, and behaviours are understood as being caused by external stimuli as operant conditions. B.F. Skinner, a leading proponent, argued the following. Pleasant experiences are positive reinforcers. If experienced by a learner, they establish connections between stimuli and response. On the other hand, unpleasant experiences are negative reinforcers. They have the effect of causing learners to avoid undesirable responses to stimuli. If learning is continuously reinforced, this increases the rate and depth of that learning. Both positive and negative reinforcement can shape behaviours immediately and in the long-term. If the learner does not receive any reinforcement, then this can also shape behaviour. If learners do not receive any response to their behaviour, they may change their behaviour to induce or encourage some kind of external reinforcement.
An example of a programme of learning underpinned by a behaviourist meta-theory is the Keller method. This method, or, more accurately, teaching and learning approach, has been influential, if not decisively successful, in the education of the professions in Brazil (Mota, 2013). The Keller Plan (Keller, 1968) was launched in the early 1960s and it is an early attempt to use new technologies in teaching and learning environments. The Plan, also called the Personalized System of Instruction (PSI), was developed by Fred S. Keller with J. Gilmour Sherman, Carolina Bori and Rodolpho Azzi, among others, in the middle 1960s as an innovative method of instruction for the then new University of Brasilia. When the Keller Plan was launched, the new digital technologies were in their infancy and this meant that content delivery, the development of learning environments, and their capacity to deliver a deep learning experience, was limited. In addition, its reliance on a behaviourist approach meant that it was operating with a severely restricted pedagogy, and consequently its impact on learning was less than originally hoped for. However, it is worth examining because it constituted an early attempt to use the new technologies to create productive learning environments.
The Keller Plan is a type of personalized instruction in which learning materials are presented in small units. When a student feels ready, they take a test prior to completing the unit and, if they pass at an appropriate level, are allowed to continue on the unit. This test is also diagnostic in that it provides a description of the capabilities of the student, which allows the subsequent programme to be adjusted to the needs of the student. It is in this sense that the programme can be described as personalized. The student completes each of the subsequent units at his or her own pace. This indicates one of the benefits of this form of learning: the capacity of the system to accommodate students who wish to progress through the programme rapidly, as well as those who wish to take their time. This is one element of the inherent flexibility in these types of teaching and learning approaches. Under the Keller Plan, instructors (or teachers as we would know them) serve only as facilitators, administer no punishment at any stage of the learning, and award only pass or fail grades.
The Keller Plan is underpinned by a behaviourist philosophy (Zimmerman, 2002). The primary presentation of new content was through written texts. Given the forms of media available at the time when the Keller Plan was developed (e.g., lectures, movies, audio records, television, radio, paper-based text, etc.), paper-based texts gave students the greatest freedom; books and texts are portable, can be read at one’s own pace, can be started and stopped at any time, can be easily reviewed, and can be written upon by the reader. As an application of behaviourism, the Keller Plan was designed to maximize the number of operant behaviours that could be reinforced; this could best be done with written materials rather than learners being passive observers of other media.
Subject matter material was broken down into separable, meaningful units. These units could have various kinds of relationships; for example, one unit could provide learning which forms a prerequisite for understanding another, or a later unit could be an elaboration of an earlier one. Indeed, these forms of learning, because they allow flexibility, are able to accommodate different progression modes. A number of these progression modes have been identified. The first is prior condition. In the acquisition of particular knowledge, skill and dispositional elements, there are pre-requisites in the learning process. An example might be mathematical where knowledge of addition is a pre-requisite of multiplication. A second form is maturational. A maturational form of progression refers to the development of the mind of the learner. There are some mental operations that cannot be performed by the learner because the brain is too immature to process them. A third form is extensional. An extensional form of progression is understood as an increase in the amount, or range, of an operation. Greater coverage of the material is a form of progression, so a learner now understands more examples of the construct, or more applications of the construct, and can operate with a greater range of ideas.
A fourth form is intensification. Related to the idea of extension is a deepening or intensifying of the construct or skill. Whereas extension refers to the amount or range of progression, intensification refers to the extent to which a sophisticated understanding has replaced a superficial understanding of the concept. Then there is a notion of complexity. In relation to the knowledge constructs, skills and dispositions implicit within a learning environment, there are four forms of complexity that allow differentiation between units. These are: behavioural complexity, symbolic complexity, affective complexity and perceptual complexity. There is also a type of progression, abstracting, which involves moving from the concrete understanding of a concept to a more abstract version. A further measure of progression is an increased capacity to articulate, explain or amplify an idea or construct (i.e., learners retain the ability to deploy the skill and in addition, they can now articulate, explain or amplify what they are able to do and what they have done). A final form of progression is pedagogical, and this refers to the way that learning is also influenced by its means of delivery. An example could be moving from an assisted performance to an independent one. Students are allowed to advance through the course material at their own pace and in an order which suits the type of progression that is most appropriate for them. Learners then move through a programme as quickly or slowly as they choose, as long as they finish the whole programme within a determined period of time.
Students are required to satisfy a mastery requirement in one unit before proceeding to the next. Typically, a unit in the programme would have more than one equivalent form of assessment; for example, three quizzes of equal difficulty or three primary sources or data sets to be analysed. Students are required to demonstrate mastery of a unit’s objectives at a certain level. If the student does not reach the threshold, he or she is redirected to unit materials (or supplements if provided) and can then take an equivalent form of the unit assessment. From a behaviourist perspective, demonstrating mastery, and being allowed to continue to a subsequent unit, was presumed to be reinforcing.
Teaching assistants or proctors were an important element of the Keller Plan. They could have been external to the programme (adults or peers recruited from external sources) or internal (advanced students on the programme who were doing well, had completed all the units to date, and had good interpersonal skills). They acted as the arbiters of unit mastery; they certified mastery, identified areas of weakness, and directed students to the next units. The Keller Plan was used extensively in the Brazilian higher education system, particularly as a more personalized form of instruction, but there is nothing inherent in the Keller formulation to restrict its application to particular grade levels, contents or types of programmes. There has been some research on the effectiveness of the Keller method which suggests that it has had robust, significantly positive effects on learning when compared to more traditional lecture-based formats (Pear and Crone-Todd, 1999).
The Keller Plan, as we have suggested, is underpinned by a behaviourist meta-theory and this may have contributed to its relative lack of success. Behaviourism can be contrasted with the two alternative meta-theories that we have already made reference to: cognitivism or symbol-processing and constructivism. The main focus of cognitivism is the role played by inner mental activities. The learner is viewed as an information processor, passively receiving information from an external source. Cognitivist perspectives on learning are a paradigm example of a symbol-processing learning philosophy.
On the other hand, constructivism entails an active process of learning and is generally associated with the work of the Russian psychologist, Lev Vygotsky. Moore (2012:18–19) summarizes Vygotsky’s views on development, instruction and consciousness. Cognitive development is achieved most effectively by elaborating ideas and understandings in discussion with teachers or pedagogical experts and peers. Learners perform and develop better with help than without help, and are given tasks that will test what is developing in them rather than what has already developed (the notion of stretching not just “able” students, but those who may be perceived as under-achieving in comparison with any accepted developmental or positional norm). Learners aim to develop “conscious mastery” over what they have learned rather than merely being able to recite facts which may have little meaning for them. The development of such expertise is not subject-specific, and once acquired becomes a tool through which all learning is facilitated and enhanced. Student–teacher relations therefore are dialogic rather than monologic, involve collaborative learning, both with peers and the teacher, recognize learning as an active and interactive process concerned with the provisional nature of the student’s knowledge, and emphasize articulation and meta-processes of learning.