Psychology’s Movement Toward Cognitive Science

The progression between psychological schools of thought in this brief history emphasizes three trends: First, psychology progresses through the development, evaluation, and integration of experimental methods into psychology.  Second, psychological schools of thought move from emphasizing the conscious, qualitative aspects of mind and mental processes to emphasizing characterizations of the mind and mental processes in terms of informational processing.  Third, psychologists develop and adapt experimental techniques in order to more reliably explore those elements of their information processing models not directly observable by experimenters.  The development of experimental methodologies, the refinement of animal and other models, increased knowledge of human mentality, development, and physiology, and the development of technical ideas such as information and computation ultimately coalesce, allowing for the conceptual framing of cognitive phenomena as well as its systematic experimental investigation.

In our discussion the development of physiology, I emphasize that physiology introduces a strong empirical and experimental emphasis to psychology. Indeed, many early psychologists receive their training from physiologists and anatomists. Consistent with that idea, I begin the discussion of the development of psychology towards cognitive science with the development of an experimental technique that has proven quite central to psychology—reaction time.  

Reaction Times and Mental Chronology

Reaction time is the time it takes from the initial presentation of stimulus until the subject reacts.  For instance, the time between a light flashing and a subject pressing a button would be that subjects reaction time for that stimulus-response pairing.  Reaction time measurement begins with Friedrich Wilhelm Bessel (1784-1846), a German astronomer and Gymnasium drop-out. The controversy regarding disparate observations of transit times between the astronomers Maskelyne and Kinnebrook, and later between Enckes and Gauss intrigues Bessel.  Transit times are measurements of how long it takes stars to cross hairlines in telescopic observations, indicating the time it takes the star to pass over the meridian of the observatory.  Astronomers use transit times for a star to determine the coordinates of the star.  Precise measurements of star coordinates are becoming increasingly important in the 1700s and 1800s, both for the astronomers and also because nautical tables of the time rely upon coordinate information for stars.  Twenty years after the original dispute (1821), Bessel both analyzes the observations of different astronomers and performs the simple experiment of comparing observations between observers using the same equipment.  Bessel determines that skilled astronomers will vary consistently in their observations of transit times.   As a result, Bessel introduces the notion of an “involuntary constant difference,” in describing his findings in the preface to the eighth volume of his Astronomische Beobachtungen¹(1823).

In astronomy, the phenomena now commonly goes under the name introduced by John Pond (1767–1836) in 1933², as the “personal equation.” It is not an exaggeration to describe this work as the first experimental quantitative measurement of reaction time, and it eventually results in the first attempt to control both for reaction time and also for individual differences in scientific observation. The German physiologist, Hermann Ludwig Ferdinand von Helmholtz (1821-1894) published his “On the Rate of Transmission of the Nerve Impulse”³ in 1850. The Swiss astronomer Adolph Hirsch (1830-1901) publishes his “On the Speed of Mental Processes,” (1862) in which he ⁴

Franciscus Cornelis Donders (1818-1889) is usually cited as the first researcher to use differences in human reaction time to infer differences in cognitive processing time.  Building on the work of his graduate student, Johan Jacob De Jaager,⁵ and with an awareness of earlier work by Helmholtz and Hirsch, Donders uses the same subtraction method employed by Helmholtz, to make inferences about the times of various mental processes.  In 1868, Donders publishes “On the Speed of Mental Processes,”⁶ in which he shows that a simple reaction time is shorter than a recognition reaction time, and that the choice reaction time is longest of all.  Using these times, Donders makes inferences as to the speed of mental processes through subtraction: recognition = (recognition reaction time - simple reaction time).  Donders' results are an instance of mental chronometry, i.e., the study of the relative speed and temporal sequencing of mental process under some specified set of conditions.  The ideas of subtraction, mental chronometry, and reaction time are now part of the central methodological framework of cognitive psychology.  The next experimental technique, introspection, has a less venerable history.

Introspection and Introspection-Based Psychologies

The technique of introspection enters psychology through the work of Wilhelm Maximilian Wundt, (1832-1920) a German physician, psychologist, physiologist, and university professor.  Wundt, Titchener, and Brentano are often portrayed together under the title of introspectionist psychology.  However, as we will see, these theorists differ significantly in their theoretical perspectives as well as their use of introspection.

Introspectionist Psychology: Voluntarism

Wundt creates the first laboratory explicitly dedicated to psychological research (as opposed to labs for teaching demonstrations) at the University of Leipzig in 1879. He also begins the first journal for psychological research in 1881. If Freud is the “father of Psychiatry,” Wundt is likely the “Father of Experimental Psychology.”

Wundt studies under Müller and Helmholtz prior to writing his first book, Beiträge zur Theorie der Sinneswahrnehmung (Contributions to the Theory of Sense Perception)⁷ (1862). He follows this book with his second, Vorlesungen über die Menschen -und Tierseele (Lectures on Human and Animal Psychology)⁸ (1863). Wundt has founded the first school of psychological thought by 1986 and 1897when he publishes Grundriss der Psychologie (Outline of Psychology) and Grundzüge der physiologischen Psychologie (Principles of Physiological Psychology).⁹  Ironically, Wundt’s success at training a new group experimental psychologists leads to a distortion of his own views.  One Wundt’s students, Edward Titchener (1867-1927), actively, but misleading, associates Wundt’s view with Titchener’s own view, structuralism. Wundt names his view voluntarism, and as we will see, it differs somewhat from the typical views attributed to structuralism.

Wundt equates mentality with consciousness--to have a mind requires that one has conscious experiences.  Wundt holds that experimentation can help one to understand simpler conscious phenomena, but not more complex phenomena.  Essentially, Wundt limits psychology to studying emotions and what contemporary cognitive scientists call sensation and perception.  For Wundt, one must employ historical analysis and naturalistic (non-experimental) observation to understand higher mental functioning such as reasoning, language use, religious experience, etc..

As regards simpler conscious phenomena, Wundt distinguishes and studies two kinds of sensation and emotion.  Wundt studies perception--passive and involuntary combining multiple sensations--and apperception--active voluntary perception involving attention.  He seeks the fundamental constitutive elements of conscious mentality and the rules by which these elements combine into more complex experiences.  In this way, Wundt follows the mental chemistry model of the mind that one finds in the British Empiricists.  However, for Wundt attention and the will act as a sort of catalyst making apperception active.  Thus, Wundt differentiates himself from the British Empiricists, in part, because of his introduction of an active, volitional component to the mental chemistry account.  In recognizing the role of attention in how people perceive the world, Wundt anticipates the sorts of perceptual phenomena that later psychologists (e.x. Gestalt Psychologists) come to recognize as crucial to perception.  In contrast, Wundt’s exclusion of higher mental functioning from the domain of scientific studies of the mind differs from other atomists.

Wundt breaks mental elements into two categories; sensations and feelings. Sensations are the result of stimulation of the sense organs.  Each sensation has an intensity (ex. bright vs dim) and a modality (touch, taste, etc.).  Modalities have associated qualities such as sweet and sour for taste.   Feelings are distinct from sensations but co-occurring.  Wundt proposes a tridimensional account of feelings in which feelings have values along three orthogonal dimensions; pleasantness-unpleasantness, excitement-calm, and strain-relaxation.

In order to study perception and apperception, Wundt adopts two primary experimental techniques for studying these simple conscious phenomena; introspection and reaction time.  Reaction-time enters Wundt’s experimental repertoire from Helmholtz and Donders.  However, he eventually abandons the use of reaction time as too unreliable for gathering data.  For Wundt introspection provides immediate observation of mental phenomena—direct observation unmediated by measuring and recording devices.  Unlike many other psychologists who follow him, Wundt uses introspection in a highly constrained fashion in keeping with its use in physiology and psychophysiology. Wundt often presents stimuli to experimental subjects, asking them to  provide only yes or no answers to signify whether they are experiencing a sensation.  However, Wundt does insist that subjects must be trained in introspection to learn the appropriate categories.   

Introspectionist Psychology: Structuralism

Edward Titchener (1867-1927) is an English student of Wundt who comes to Cornell University where he continues Wundt’s general project of trying to identify the elements of simple human consciousness and their interactions.  However, Titchener’s views and methods differ significantly from Wundt’s. Unlike Wundt, Titchener seeks to apply the lens of experimental psychology to higher order mental phenomena as well as simpler conscious phenomena. Titchener views experimental psychology as generating a morphological account of mental experiences—that is, an account of the elements and composite structure of conscious mental experiences.  In his article, “ The Postulates of a Structural Psychology"¹⁰ (1898), Titchener tells readers that,

Likewise, Titchener differs from Wundt in that Titchener rejects the idea of an active volitional component of mental experiences.  Titchener rejects as unscientific volitional, functional and teleological descriptions of mental processes at the psychological level.  Instead, Titchener follows the British Empiricists in supposing that psychology should create an account of mental experiences in terms of structures created via the combination of basic elements through the mechanism of association.  Similarly, Titchener holds that mental elements can only be known through their attributes.  Titchener distinguishes three kinds of mental elements; sensations, images, and affections. Perceptions are composed of sensations, ideas are composed of images, and emotions are composed of affections.

Titchener’s use of introspection differs from Wundt’s in that it requires subjects to actively probe or analyze their experiences to formulate reports.  This requires extensive training. Titchener intends his training to cultivate an ability to observe and describe conscious experiences without the tincture of “stimulus error.” Stimulus error occurs when subjects report their perceptions--reporting the meaning of the stimulus (or it’s conceptualization).  For instance, if a subject saw an apple, the subject must report the hues, shapes, etc. of their experience--they should not report seeing an apple or a fruit. This combination of indoctrination into descriptive categories and active, even retroactive, analysis by subjects renders introspection even more methodologically problematic as an experimental tool.  In addition to the difficulties surrounding Titchener’s use of introspection, he also ignores as irrelevant numerous research areas where significant progress is occurring.  For instance, Titchener discounts animal behavior and evolution, abnormal behavior, learning, development, and inter-subjective variation.

Introspectionist Psychology: Act Psychology

If you wish to find a true introspectionist villain, Franz Clemens Honoratus Hermann Brentano (1838-1917) probably best fits that description.  In his major work, Psychologie vom Empirischen Standpunkte¹¹ or Psychology from an Empirical Standpoint (1874), Brentano coins the term “intentionality” to characterize his view that every mental act has an object to which it refers.  For example, when someone sees an apple, they see it as an apple, an object, and not merely as a qualitative experience.  Brentano tells readers:¹²

Brentano also eschews the study of static, simple conscious experiences, framing mentality in terms of acts, that is, in terms of the mind being directed towards an object in order to perform some function.  Indeed, he holds that psychology should study mental processes in order to determine their function.

Though Brentano never practices experimental psychology, he does employ and advocate “phenomenological introspection,” in his theorizing about the nature of the mind and its processes.  In employing phenomenological introspection, the researcher either asks the subject to analyze temporally extended processes such as inferences, or performs such an analysis themselves.  Though Brentano publishes very little, he influences many people, for instance, Freud.

The Downfall of Introspectionist Psychology

It was the reliance upon introspection as their sole experimental methodology, together with the implicit biases introduced by training subjects, and allowing extended and retroactive introspective analysis that ultimately doomed the methodological side of introspective psychology.  Its practitioners make little to no effort to assess introspection’s calibration, or control for subject and experimenter bias.  Instead, perhaps naturally, they assume introspection is perfectly reliable across all of its methodological uses.  Quite to the contrary, introspection turns out to be relatively poorly calibrated and susceptible to massive subject and experimenter bias.  The behaviorists heavily criticize introspection as part of their rejection of the various forms of introspective psychology.  The contemporary uses of introspection, as a result, are highly constrained and cross-validated.   Indeed,  Richard Nisbett and Timothy Wilson publish an influential literature review in 1977,¹³ which still serves to highlight the perils of introspection.

As we’ll see, the failure of introspection and the various schools of psychological thought that rely heavily upon it also serves to shift the emphasis from understanding the mind through conscious experience towards understanding the mind through behavior and eventually cognition

Ebbinghaus: The Quantified Study of Memory as a Process

Hermann Ebbinghaus (1850-1909), begins one of the first systematic studies of memory in 1879.  He studies only the ability to memorize nonsense syllables by rote.  He selects nonsense syllables since previous learning will not influence learning of these meaningless sounds.  This choice represents one of his contributions to memory research, namely, that ease of memorization is increased by meaningfulness and relevance to the memorizer, and vice versa.  Ebbinghaus might have used some of his students as subjects, but he seems primarily to use only himself as a subject.  Ebbinghaus publishes his results in his book, Über das Gedächtnis Untersuchungen zur Experimentellen Psychologie ¹⁴ or On Memory (1885), which is later translated and published as Memory: A Contribution to Experimental Psychology¹⁵ (1913).

In On Memory Ebbinghaus reports results that are the basis for the “learning curve” and the “forgetting curve.” The learning curve shows that learning time, measured as number of repetitions, increases exponentially with the number of items memorized.  Likewise, the increase in retention for each repetition decreases exponentially so as to approach complete retention asymptotically.  The forgetting curve is similarly exponential, showing the forgetting decreases at an exponential rate, so as to approach complete failure of retention asymptotically.  Specifically, R = e ^-t/s (where R = retention, t = elapsed time, and s = strength of original memory). Ebbinghaus also documents the serial position effect, viz. the recency and primacy effects (subjects are more likely to remember the last item in a series [recency] and the first item in a series [primacy]).  Likewise, Ebbinghaus documents “savings."   If one memorizes a list and then waits until recall is zero, one will still generally relearn the list at a faster rate despite the seeming lack of recall.  Ebbinghaus terms the difference between the first and second memorization the savings.

Though Ebbinghaus is not prolific, does not identify himself with any psychological school of thought, and does not seek out pupils, his work spurs research on memory.  Ebbinghaus’ careful, well-designed experiments, his rigorous quantified results, statistical analysis, and systematic presentation prove extremely influential.

Functionalism

Most historians consider William James (1842–1910) the first figure in the first school of psychology in the United States--functionalism.  Functionalism overlaps significantly with both structuralism and behaviorism.  James’ text, The Principles of Psychology¹⁶(1890), actually predates by two years Titchner’s arrival a Cornell.  Principles, a two volume, twelve hundred page text, makes James as famous and widely studied as Wundt.  Functionalism, much like pragmatism, the philosophical movement with which James is also associated, has no central figure, nor a clear-cut doctrine.  However, functionalists make important criticisms and contributions to other schools of psychological thought, and do share several common general commitments. (1) Functionalists oppose the search for the basic elements of thought that characterizes Wundt’s and Titchener’s views.  In fact, Principles portrays mentality as a stream of consciousness incapable of analysis into elements.  (2) Functionalists, in contrast to voluntarists and structuralists, think of the mind as dynamic, and mental processes as serving functions.  (3) They view the function of mind through the lens of evolution. Thus, functionalists understand mental processes and behavior in terms of the general goals of adaptation and selectional advantage.  (4) Unlike the rather rigid determinism of structuralism and behaviorism, functionalism tends to emphasize adaptation and differential responses driven by motivation.  (5) Methodologically, functionalists tend to accept both introspection and behavioral observation as methodological tools.  Though few early functionalists conduct experiments, later functionalists do conduct experiments.  Functionalists also support psychological research on animals, children, and abnormal populations as a means to understand normal human mentality. (7) Finally, unlike Wundt, who views psychology’s mission as pure basic research, functionalists tend to see psychology as a means to improve society and people’s lives.

The functionalists act as a counterweight to both structuralism and behaviorism.  For instance, historians often cite John Dewey’s “The Reflex Arc Concept in Psychology"¹⁷ (1896) as the beginning of functionalism.  In this article Dewey criticizes the notion of the reflex arc as consisting of discrete stages; stimulus and response.  He also anticipates challenges facing behaviorists by noting the difficulties in specify a context-free notion of stimulus and response, i.e., specifications capturing the relevant features of particular stimuli and responses as well as spcifications which predict the generalizations to future cases.  Furthermore, the functionalist emphasis on practical applications in psychology, evolution, and diversity in methodology and research areas help to plant the seeds for cognitive science.

Historians classify Edward Lee Thorndike (1874-1949) as a functionalist.  However, much of his work is arguably the first research on conditioning and is certainly the first work on operant conditioning.  Thorndike publishes his dissertation, “Animal Intelligence: An Experimental Study of the Associative Processes in Animals,” in 1898--predating Pavlov’s first public reference to conditioned reflexes by approximately a year.  Thorndike republishes this work in 1911 as Animal Intelligence.¹⁸

Thorndike studies a number of animals, but he is most famous for his studies of learning in cats using homemade puzzle boxes (above).  Thorndike puts a cat in a box, and lets it behave randomly until it stumbles upon the release mechanism.  He repeats this procedure until the cat can release itself in negligible time.  Thorndike then plots the decline in time to escape relative to times in the box, using this ratio to characterize learning by “learning curves.” He formulates his generalized results in terms of the law of effect.¹⁸

Behaviorism: Pavlov’s Discovery of Classical Conditioning

Ironically, it is Ivan Pavlov (1849-1936), a Russian physiologist studying digestion, who provides the key finding around which Behaviorism evolves--classical conditioning.
 

Pavlov’s research includes the physiology and the neurophysiology of temperament, conditioning, and involuntary reflex actions; but the bulk of his work focuses on digestion.  Pavlov’s experimental research on digestion was innovative and sophisticated.  Pavlov’s techniques include surgical removal of components of the digestive system from animals to facilitate observations of their functions, lesioning nerve fibers to trace their function by observing the lesion’s effects, and implanting fistulas (tubes or holes) draining into pouches to examine the organ's contents.  In the 1890’s Pavlov’s lab is performing experiments on digestion using dogs.  Specifically, Pavlov’s group is studying the salivatory functions of dogs by surgically externalizing a salivary gland so that the saliva could be collected and analyzed.  During their research Pavlov notices that the dogs begin to salivate before receiving food.  He calls this phenomena “psychic secretion”¹⁹ (p.7), and the lab begins to investigate this phenomena.  The researchers realize that these “psychic secretions” result from associations between the food and other stimuli.  These investigations eventually reveal what Pavlov calls “conditioned reflexes,” and we now call classical conditioning.

Pavlov first mentions his discovery in a lecture to the Society of Russian Doctors of St. Petersburg in 1899.  Printed accounts of the research appear in a dissertation by Pavlov’s student, Wolfson, and in a report to the 1903 Congress of Natural Sciences by Ivan Tolochinov,²⁰ Pavlov’s collaborator.  However, the discovery does not receive significant attention until Pavlov discusses it in his Nobel Prize acceptance speech in 1904. Pavlov’s own account does not emerge until he publishes, Conditioned reflexes: An Investigation of the Physiological Activity of the Cerebral Cortex¹⁹ in 1927.

As illustrated in the diagram above, classical conditioning works by associating a stimulus that triggers a specific response with a novel stimulus. Specifically, the unconditioned stimulus is a stimulus that elicits a particular response called the unconditioned response.  The conditioned stimulus is paired with the unconditioned stimulus repeatedly so that the pairing elicits the unconditioned response.  This repeated pairing increases the association between the conditioned stimulus and the unconditioned response, as reflected in the increasing likelihood of the conditioned stimulus eliciting the response in and of itself, making the unconditioned response a conditioned response.

Pavlov also discovers and studies extinction and spontaneous recovery. When one elicits the conditioned response using only the conditioned stimulus, the association between the conditioned stimulus and conditioned response weakness over time.  This weakening is called extinction.  After an association between a conditioned stimulus and a conditioned response reaches extinction, the conditioned stimulus can elicit a conditioned response at a later; such cases are called spontaneous recovery.

Behaviorism: John Watson

John Watson (1878-1958) adapts the work of Pavlov into a general approach to psychology, which he presents in his 1913 paper, “ Psychology as the Behaviorist Views It”.²¹ Watson embraces the idea of classical conditioning and sets the goal of psychological investigation as the prediction and control of behavior.  Additionally, Watson explicitly rejects the project of analyzing conscious experience, and the methodological tool of introspection.  Watson describes his view as follows:²¹

While Watson emphatically distances behaviorism from introspectionist psychologies, there are strong similarities under the surface.  Watson limits the appropriate phenomena of psychology to behavior, and sets the goals of behaviorism as prediction and control of behavior.  However, in practice Watson’s treatment of behavior is quite atomistic.  Watson divides behavior into four classes; explicit learned behavior such as riding a bicycle, implicit learned behavior such as a rumbling stomach when smelling someone else’s dinner cooking, explicit unlearned behavior such as pulling your hand away when it is hurt, and implicit unlearned behavior such as sweating when it is hot.  For the purposes of this class, we’ll focus upon Watson’s theoretical framework for the prediction and control of learned behavior.  Watson hypothesizes that explicit, complex learned behaviors--such as chess playing or language—can be understood as a series of simpler learned and unlearned behaviors performed in a sequence cultivated by imitation and classical conditioning.  Simple learned behaviors are explained in terms of the environment, unlearned behaviors, and conditioning history.  For instance, implicit learned behaviors (like a fear response) develop from simple conditioning and appropriate environmental ques.  As Watson tells readers of his text, Behavior: An Introduction to Comparative Psychology²²(1914):

If this general approach seems familiar, it should. Watson, in effect, proposes an atomistism for behavior; a set of elemental behaviors—unlearned behaviors—from which all learned behaviors are generated and combined through a process of association based upon contiguity and frequency.

Neo-Behaviorism

Historians classify the behaviorists that follow Watson as neo-behaviorists. One can find the general motivation behind the neo-behaviorist research in the above quote from Watson.  On the one hand, Watson seeks to provide a highly mechanistic/deterministic account of the generation of behavior.  On the other hand, Watson wants to use his account to explain complex, flexible behavior, including behavior in novel circumstances.  Neo-behaviorists seek to expand the basic behaviorist framework to allow for increased flexibility and complexity.  For instance, suppose that a researcher trains a rat to associate food with a blue box. Sometimes when presented with the blue box, the rat does not try to eat.  Why doesn't the rat respond all the time?  Will the rat respond to the box when the researcher changes the color or the shape slightly?  How much change to the box can occur before the conditioned response is no longer triggered?

Neo-behaviorists continue to view overt behavior as the central phenomena for psychology.  They also hold that the prediction and control of behavior is the central goal of psychology.  Learning reamins central to psychology.  Finally, neo-behaviorists share Watson’s conviction that animal models of learning and perception are easily and robustly transferable to humans.

Neo-behaviorists commonalities go beyond Watson as well.  For example, neo-behaviorists share Watson’s general commitment to grounding psychology in observation.  However, unlike Watson, neo-behaviorists seek to tie all theoretical terms to experimental operations for the measurement and/or application of those terms.  Researchers call this view about the treatment of unobservable theoretical terms operationalism.  Operationalists refer to the specification of a set of operations for a theoretical term as an operational definition.  Watson and the neo-behaviorists see animal experimentation as essential to psychological research because of the continuity of animal learning and perception with human learning and perception.  However, neo-behaviorists see additional value  in animal experimentation because it allows for more rigorous controlled experiments.  Lastly, neo-behaviorists differ from Watson and one another in the manner in which they seek to extend behaviorism.  Radical behaviorists like B.F. Skinner hold that the prediction and control of behavior must eschew internal, unobservable mental and physiological events.  Methodological behaviorists allow for appeal to internal states so long as those terms are tied to observation.

Neo-Behaviorism: Radical Behaviorism

Historians classify Burrhus Frederic “Fred” Skinner (1904-1990) as a neo-behaviorist, though Skinner is very strongly associated with behaviorism in the popular mind.  Skinner’s association with behaviorism is due in part to his theoretical approach.  Like Watson, Skinner insists that theorists focus on predicting and controlling behavior.  Likewise, Skinner also supposes that conscious mental states have no part to play in psychological theorizing, insisting instead that environment and conditioning history provide sole basis for the psychological understanding of human conduct.

However, Skinner’s theoretical perspective diverges from Watson on several key points.  Skinner adopts an approach called “functional analysis” that he traces to Ernst Mach.  Functional analysis has a quite different meaning for Skinner than for the theorists we will discuss later.  For Skinner functional analysis characterizes dependencies--not meticulously detailed step-wise causal relationships--between observable phenomena, i.e., between conditioning histories and environments.  Specifically, Skinner understands functional analysis as a method for establishing relationships between stimuli and responses through the application of operant conditioning.  Skinner's analysis is often called a "three-term contingency" analysis in that it characterizes the environmental features that act as a trigger for the behavior (sometimes called the discriminative stimulus), the response (the specific rigorously characterized behavior), and reinforcement (the consequence of the behavior that positively or negatively influences the probability of the behavior in the eliciting conditions).  While Watson rejects Thorndike’s law of effect as too subjective, Skinner creates a systematic, objective formulation of the learning paradigm--operant conditioning.  Unlike Watson, Skinner does not understand behavior as something elicited by environmental stimuli.  Rather, Skinner views behavior as active operations on the environment.  Likewise, Skinner sees patterns of behavior emerging as a result of those behaviors being selected by contingent environmental reinforcement.

Skinner writes his famous book, The Behavior of Organisms²³ (1938), while at his first job at the University of Minnesota.  In that book Skinner reformulates Thorndike’s law of effect so that it describes environmental selection of behavior through the reinforcement resulting from the behavior. Skinner makes no reference to subjective states like desires, drives, etc. in characterizing reinforcement.

In operant conditioning a creature’s behavior--often random behavior--is either positively or negatively reinforced (rewarded or punished).  The probability of the behavior occurring again in the relevant eliciting conditions increases or decreases in proportion to the number of behavior/reinforcement pairings. For example, as depicted in the animation above, a cage might be divided into two sections.  Whenever a rat wanders into one section, experimenters administer an electric shock. Over time, the probability that the rat will, for instance, leave the section of the cage where it recieves shocks whenever placed there increases.

Operant conditioning together with classical conditioning broaden the range of learned organism-environment interactions. Classical conditioning provides a mechanism whereby stimuli from the environment can elicit a response, i.e., stimuli cause organism responses. Operant conditioning provides a mechanism whereby behavior becomes part of the organism’s repertoire as a function of its consequences, i.e., consequences elicit behaviors. 

Part of the lore of cognitive science and behaviorism involves Skinner. In 1957 Skinner publishes a book, Verbal Behavior,²⁴ based upon lectures originally given at the University of Minnesota, and further refined at Columbia and as the lectures William James. Skinner supposes that verbal behavior has no significant and essential differences from other sorts of behavior. For instance, he denies that verbal behavior results from an innate capacity. Skinner proposes to treat verbal behavior using his functional analysis method.

In 1959 Noam Chomsky publishes a review of Skinner’s book.²⁵ Often accounts of the development of cognitive science portray Chomsky’s review as a refutation of behaviorism and the beginning of cognitive science. Chomsky’s review represents an informed, articulate, and forward-looking indictment of the promise of Skinner’s functional analysis as a methodology for investigating language. Few authors could hope improve upon Chomsky’s articulate formulation of the task facing anyone who seeks to understand language and language acquisition:²⁵