Three Dominant Theories of Learning
Before the 1950s, learning theories developed rather independently. As the theories were disseminated across scholars and practitioners, each adopted new features to explain previously unaddressed phenomena. For many learning theories in the first part of the twentieth century, the empiricist/associationist perspective provided the predominant framework (Ertmer & Newby, 2013), and many of the core theory-specific findings were made as a direct result of the particular epistemological framework and culture in which each theory was situated. The purpose of this section is to describe the basic background and elements of each learning theory and to discuss the contributions of various key scholars within each perspective. Let us now start with behaviorism…
Behaviorism
Specifically influenced by objectivist [1]empiricism in which reality is external to the learner and knowledge is attained by experimentation, behaviorism became the leading psychological approach to the study of learning (Schunk, 2012) during the first half of the 20th century. Behaviorism as a school of psychological thought was founded in 1913 by American psychologist John B. Watson (1925). His work Psychology as the Behaviorist Views It (1913) is often referred to as the “Behaviorist Manifesto” (Lattal & Rutherford, 2013; Logue, 1994) and involved a clear rejection of the introspective methods of the structuralists that sought to understand consciousness (i.e., the mind). The precepts Watson discussed in his address formed the dominant perspective for viewing learning for the next 60 years, particularly in English-speaking countries.
The core behaviorist ideas are that behavior is predictable and controllable and only things that can be directly observed and measured are considered acceptable as scientific material. In fact, many behaviorists assert that such concepts like “thought, reasoning, processing, or memory” (Morrison, Ross, Kalman, and Kemp, 2013, p. 353) are unobservable, subjective, and uncontrollable, and therefore of little concern since they cannot produce true knowledge (Morrison et al., 2013). Thus, the focus of behaviorist experiments – usually involving animals – was often rather simple learning processes that produced basic mechanical understanding. Other basic assumptions of behaviorism are that people behave in response to environmental stimuli, learning is “the ability to perform new behaviors” due to response strengthening (Richey, Klein, and Tracey, 2011, p. 52), learning is described as associations among stimuli and responses (i.e. observable events), learning involves a change in behavior, learning is enhanced when stimuli and responses occur closely sequential in time, and many animal species learn in similar manners (Ormrod, 2008).
For the behaviorist, learning occurs within a stimulus-response-reinforcement context in which a learner draws connections between a stimulus or set of stimuli and a reinforced response. A central goal is for the teacher to predict and produce preferable behaviors and reject undesirable behaviors by using environmental conditions in such manner to either reinforce or punish the learner following a response. The core principles of behavioral learning theory have their foundations in Edward Thorndike’s theory of connectionism, Ivan Pavlov’s experiments in classical conditioning, and B.F. Skinner’s research on operant conditioning.
Connectionism
American psychologist Edward L. Thorndike (1905) introduced one of the first theories of connectionism with its three laws – the law of effect, the law of readiness, and the law of exercise. Thorndike’s research on learning theory marks the point when the dominant psychological school shifts from functionalism to behaviorism (Hergenhahn & Olson, 2005). In his doctoral thesis, “Animal Intelligence: An Experimental Study of the Associative Processes in Animals” (1898), Thorndike established the law of effect and concluded that the only way to understand how someone learns is through an experimental approach. Thorndike viewed learning as the forming of associations or connections between perceptions of stimuli or events and responses that are revealed in behavior. This “(emphasis on) associations between stimuli and responses as the basis of learning” (Schunk, 2004, p. 30) places connectionism as an early behavioral theory.
Thorndike viewed learning as a process of selecting and connecting (i.e., trial and error) in which the individual makes a response, receives reinforcement for a correct response, and as a result of the positive reinforcement, makes a connection with the stimulus. According to the law of effect, the type of reinforcement that follows a particular response will determine the strength of a particular connection. To illustrate, if the individual receives a reward, then the connection is strengthened and becomes habitual. However, a punishment will weaken the connection and the corresponding behavior (i.e. learning). Thorndike’s other two laws indicate that the learner needs to be physically ready in order for a new connection to stick (law of readiness) and needs to maintain continuous practice of a new behavior in order to strengthen the new connection (law of exercise), otherwise forgetting will occur due to disuse. Thorndike’s three laws of connectionism provided a basic empirical framework for behavioral psychology and subsequent behaviorist theories of learning, particularly in the areas of behavior reinforcement theory and behavior analysis, but also in the areas of motivation and attitude formation (Hergenhahn & Olson, 2005).However, it is worth noting that Thorndike later modified parts of his earlier theories of learning. Through further research, he found that practice alone is not sufficient to strengthen an association nor does time alone weaken an association (as stated in the law of exercise) “a satisfying state of affairs” can strengthen an association, punishment does not modify behavior, and consequences of behavior often determine what is or is not learned (Hergenhahn & Olson, 2005).
Classical Conditioning
In his studies on canine digestion, Russian physiologist Ivan Pavlov (1927) noticed that each time a person entered the room in which the dogs were being studied, the dogs would salivate. This is because the dogs were associating the researchers with the meat used in the experiments. This led Pavlov to introduce the shining of a light or a bell before each feeding, which led the dogs to salivate whenever they saw the light or heard the bell – even if there was no food present. This then led Pavlov to conclude that a neutral stimulus (i.e., the light or bell) could become a conditioned stimulus that would effect a conditioned response. Pavlov’s classical conditioning experiments laid the groundwork for three key concepts in behaviorist learning theory: reinforcement, experimental extinction, and stimulus generalization (Richey et al., 2011).
Pavlov discovered that reinforcement could be either positive or negative. In positive reinforcement, the likelihood of a desired response rises when a stimulus is provided but the likelihood of an undesired response decreases once the stimulus is removed. The second concept, experimental extinction, occurs once the researcher discontinues reinforcement and introduces a conditioned stimulus (i.e. a light/bell) without the presence of an unconditioned stimulus (i.e., food). Then the conditioned response should reoccur with “spontaneous recovery” without being prompted – although such a response may be noticeably weakened (Bower & Hilgard, 1981). Finally, stimulus generalization occurs whenever a learner begins to respond to similar stimuli in a similar manner. For instance, behaviorist John Watson performed a rather controversial controlled experiment with a child called “Little Albert” (Watson & Rayner, 1920). Through classical conditioning that paired a white rat with loud, scary sounds, “Little Albert” learned to repeatedly cry whenever the rat was present even in absence of the loud sounds. Thus, Watson concluded that the fear generalized to other fuzzy “rat-like objects” as well. Incidentally, Watson and Rayner (1920) were among the first to show empirical evidence of the validity of classical conditioning in experiments with human learners. Also interesting about classical conditioning is that it is the stimuli which produce the behavioral response, whereas in operant conditioning (discussed next), it appears to be the reinforcement that holds the greater ability to strengthen or weaken behaviors (Bouton, 2016). The learning theories of Thorndike and Pavlov were later synthesized by behaviorist scholar Clark Hull in his book Principles of Behavior (1943), in which Hull expressed animal learning and conditioning in biological terms and established behaviorism as the dominant theory of learning for the first half of the twentieth century (Hilgard, & Bower, 1975).
Operant Conditioning
Another key scholar influenced by Thorndike and Pavlov is the most widely known behaviorist, B.F. Skinner. Skinner’s (1938) work on operant conditioning built upon stimulus-response-reinforcement theory with the added goal of “strengthening a response by following it with reinforcement” in an “effort to identify, predict and control behavior” (Richey et al., 2011). Simply stated, the concept of learning for Skinner (cited in Spector, 2012, p. 71) is defined as “a change in overt behavior” (Skinner, 1954) and the core precepts within his operant conditioning theory “emphasize the effects of external conditions such as rewards and punishments in determining future behavior” (Morrison et al, 351). Noteworthy is that Skinner also asserted within his philosophical perspective – termed radical behaviorism – that even thoughts and feelings can be subjected to the same environmental variables as observable behavior is (Chiesa, 1994; Dillenburger & Keenan, 2009; Schneider & Morris, 1987).
The first significant contribution of B.F. Skinner and operant conditioning to behavioral theory is a refined notion of reinforcement in which both positive and negative reinforcement (i.e., reward) can increase the likelihood of the reoccurrence of a desired behavior reoccurring while positive and negative punishment can decrease the likelihood of a desired behavior reoccurring. Specifically, Skinner’s use of the term positive reinforcement indicates that something the learner wants is given as a reward for producing a particular desired behavior. Similarly, the term negative reinforcement indicates that something the learner does not want is taken away as a means of rewarding him for producing a particular desired behavior. For instance, in a classroom context at the most basic level, a teacher’s giving stickers and small treats often serves as a reward for small children for performing acts of good behavior (positive reinforcement). Likewise, the removal of a mundane chore from a child’s list of weekly responsibilities can also serve as a reward for good behavior (negative reinforcement). Skinner was careful to point out that punishment is not the same as negative reinforcement (Skinner, 1954). As for punishment, positive punishment in a class setting might take the form of the teacher’s requiring extra clean-up duties during lunchtime or simply marking a bad grade on the child’s daily progress record. Negative punishment would therefore involve the teacher’s taking or removing something that the learner wants as a punishment for undesirable behavior. For instance, some schools allow teachers to take cellphones from students who use them for illicit purposes (i.e., texting, cheating, visiting questionable websites).
Other key concepts that Skinner uses in contrast to Pavlov’s classical conditioning theory include extinction, or “the cessation of reinforcement used to maintain a behavior” (Driscoll, 2005), intermittent reinforcement as “impacting the conditioning and resulting strength of the response” (Richey et al., 2011), and the conditioning paradigm of shaping as “rewarding segments of behavior to gradually change an existing response across consecutive experiments until the desired behavior is attained” (Skinner, 1938). Finally, a crucial principle of operant conditioning involves “the importance of [the teacher] providing timely and informative feedback to learners” (Spector, 2012). Such feedback is provided through a system of timed intervals or schedules of reinforcement.
Tomorrow we will continue discussing the major theories of learning by introducing…Cognitivism.
[1] In the objectivist perspective, “Knowledge is regarded as existing independently of any human experience and the role of the learner is to acquire it. Objectivists place a strong emphasis on defining learning objectives and implicitly assume that the learner is an empty vessel, to be filled by the instructor.” (Reeves, 1992)