Relatively late in its history, the human species underwent a remarkable change: its vocal musculature came under operant control. Like other species, it had up to that point displayed warning cries, threatening shouts, and other innate responses, but vocal operant behavior made a great difference because it extended the scope of the social environment. Language was born, and with it many important characteristics of human behavior for which a host of mentalistic explanations have been invented.
The very difference between “language” and “verbal behavior” is an example. Language has the character of a thing, something a person acquires and possesses. Psychologists speak of the “acquisition of language” in the child. The words and sentences of which a language is composed are said to be tools used to express meanings, thoughts, ideas, propositions, emotions, needs, desires, and many other things in or on the speaker’s mind. A much more productive view is that verbal behavior is behavior. It has a special character only because it is reinforced by its effects on people—at first other people, but eventually the speaker himself. As a result, it is free of the spatial, temporal, and mechanical relations which prevail between operant behavior and nonsocial consequences. If the opening of a door will be reinforcing, a person may grasp the knob, turn it, and push or pull in a given way, but if, instead, he says, “Please open the door,” and a listener responds appropriately, the same reinforcing consequence follows. The contingencies are different, and they generate many important differences in the behavior which have long been obscured by mentalistic explanations.
How a person speaks depends upon the practices of the verbal community of which he is a member. A verbal repertoire may be rudimentary or it may display an elaborate topography under many subtle kinds of stimulus control. The contingencies which shape it may be indulgent (as when parents respond to their children’s crude approximations to standard forms) or demanding (as in the teaching of diction). Different verbal communities shape and maintain different languages in the same speaker, who then possesses different repertoires having similar effects upon different listeners. Verbal responses are classified as requests, commands, permissions, and so on, depending upon the reasons why the listener responds, the reasons often being attributed to the speaker’s intentions or moods. The fact that the energy of a response is not proportional to the magnitude of the result has contributed to the belief in verbal magic (the magician’s “Presto chango” converts a handkerchief into a rabbit). Strong responses appear in the absence of an appropriate audience, as Richard III demonstrated when he cried, “A horse! a horse! my kingdom for a horse!” although there was no one to hear him.
Apart from an occasional relevant audience, verbal behavior requires no environmental support. One needs a bicycle to ride a bicycle but not to say “bicycle.” As a result, verbal behavior can occur on almost any occasion. An important consequence is that most people find it easier to say “bicycle” silently than to “ride a bicycle silently.” Another important consequence is that the speaker also becomes a listener and may richly reinforce his own behavior.
The term “meaning,” though closely associated with verbal behavior, has been used to make some of the distinctions already discussed. Those who have confused behaviorism with structuralism, in its emphasis on form or topography, have complained that it ignores meaning. What is important, they contend, is not what a person is doing but what his behavior means to him; his behavior has a deeper property not unrelated to the purpose, intention, or expectation discussed in Chapter 4. But the meaning of a response is not in its topography or form (that is the mistake of the structuralist, not the behaviorist); it is to be found in its antecedent history. The behaviorist is also accused of describing the environmental setting in physical terms and overlooking what it means to the responding person, but here again the meaning is not in the current setting but in a history of exposure to contingencies in which similar settings have played a part.
In other words, meaning is not properly regarded as a property either of a response or a situation but rather of the contingencies responsible for both the topography of behavior and the control exerted by stimuli. To take a primitive example, if one rat presses a lever to obtain food when hungry while another does so to obtain water when thirsty, the topographies of their behaviors may be indistinguishable, but they may be said to differ in meaning: to one rat pressing the lever “means” food; to the other it “means” water. But these are aspects of the contingencies which have brought behavior under the control of the current occasion. Similarly, if a rat is reinforced with food when it presses the lever in the presence of a flashing light but with water when the light is steady, then it could be said that the flashing light means food and the steady light means water, but again these are references not to some property of the light but to the contingencies of which the lights have been parts.
The same point may be made, but with many more implications, in speaking of the meaning of verbal behavior. The over-all function of the behavior is crucial. In an archetypal pattern a speaker is in contact with a situation to which a listener is disposed to respond but with which he is not in contact. A verbal response on the part of the speaker makes it possible for the listener to respond appropriately. For example, let us suppose that a person has an appointment, which he will keep by consulting a clock or a watch. If none is available, he may ask someone to tell him the time, and the response permits him to respond effectively. The speaker sees the clock and announces the time; the listener hears the announcement and keeps his appointment. The three terms which appear in the contingencies of reinforcement generating an operant are divided between two people: the speaker responds to the setting, and the listener engages in the behavior and is affected by the consequences. This will happen only if the behaviors of speaker and listener are supported by additional contingencies arranged by the verbal community.
The listener’s belief in what the speaker says is like the belief which underlies the probability of any response (“I believe this will work”) or the control exerted by any stimulus (“I believe this is the right place”). It depends on past contingencies, and nothing is gained by internalizing them. To define interpersonal trust as “an expectancy held by an individual or a group that the word, promise, verbal or written statement of another individual or group can be relied on” is to complicate matters unnecessarily.
The meaning of a response for the speaker includes the stimulus which controls it (in the example above, the setting on the face of a clock or watch) and possibly aversive aspects of the question, from which a response brings release. The meaning for the listener is close to the meaning the clock face would have if it were visible to him, but it also includes the contingencies involving the appointment, which make a response to the clock face or the verbal response probable at such a time. A person who will leave for an appointment upon seeing a certain position of the hands of a clock will also leave upon hearing a response made by a person whose responses in the past have been accurately controlled by the position of the hands and which for that reason control strong responses now.
One of the unfortunate implications of communication theory is that the meanings for speaker and listener are the same, that something is made common to both of them, that the speaker conveys an idea or meaning, transmits information, or imparts knowledge, as if his mental possessions then become the mental possessions of the listener. There are no meanings which are the same in the speaker and listener. Meanings are not independent entities. We may look for the meaning of a word in the dictionary, but dictionaries do not give meanings; at best they give other words having the same meanings. We must come to a dictionary already “provided with meanings.”
A referent might be defined as that aspect of the environment which exerts control over the response of which it is said to be the referent. It does so because of the reinforcing practices of a verbal community. In traditional terms, meanings and referents are not to be found in words but in the circumstances under which words are used by speakers and understood by listeners, but “used” and “understanding” need further analysis.
Verbal responses are often said to be taken by the listener as signs, or symbols, of the situations they describe, and a great deal has been made of the symbolic process, some examples of which we shall consider in the following chapter. Certain atmospheric conditions may be a “sign of rain,” and we respond to them to avoid getting wet. We usually respond in a slightly different way in escaping from the rain itself if we have had no sign of it in advance. We can say the same thing about the weatherman’s verbal responses, which are no more a sign or symbol of rain than the atmospheric change.
Metaphor. We have seen that a stimulus present when a response is reinforced acquires some control over the probability that that response will occur, and that this effect generalizes: stimuli sharing some of its properties also acquire some control. In verbal behavior one kind of response evoked by a merely similar stimulus is called a metaphor. The response is not transferred from one situation to another, as the etymology might suggest; it simply occurs because of a similarity in stimuli. Having come to say “explode” in connection with firecrackers or bombs, a person may describe a friend who suddenly behaves in a violent manner as “exploding in anger.” Other figures of speech illustrate other behavioral processes.
Abstraction. A characteristic feature of verbal behavior, directly attributable to special contingencies of reinforcement, is abstraction. It is the listener, not the speaker, who takes practical action with respect to the stimuli controlling a verbal response, and as a result the behavior of the speaker may come under the control of properties of a stimulus to which no practical response is appropriate. A person learns to react to red things under the nonsocial contingencies of his environment, but he does so only by emitting a practical response for each red thing. The contingencies cannot bring a single response under the control of the property of redness alone. But a single property may be important to the listener who takes many kinds of practical action on many different occasions because of it and who therefore reinforces appropriately when a given object is called red. The referent for red can never be identified in any one setting. If we show a person a red pencil and say, “What is that?” and he says, “Red,” we cannot tell what property evoked his response, but if we show him many red objects and he always says, “Red,” we can do so—and with increasing accuracy as we multiply cases. The speaker is always responding to a physical object, not to “redness” as an abstract entity, and he responds “red” not because he possesses a concept of redness but because special contingencies have brought that response under the control of that property of stimuli.
There is no point in asking how a person can “know the abstract entity called redness.” The contingencies explain the behavior, and we need not be disturbed because it is impossible to discover the referent in any single instance. We need not, with William of Ockham and the Nominalists, deny that abstract entities exist and insist that such responses are merely words. What exist are the contingencies which bring behavior under the control of properties or of classes of objects defined by properties. (We can determine that a single response is under the control of one property by naming it. For example, if we show a person a pencil and say, “What color is this?” he will then respond to the property specified as color—provided he has been subject to an appropriate history of reinforcement.)
Concepts. When a class is defined by more than one property, the referent is usually called a concept rather than an abstract entity. That concepts have real referents has been pointed out by saying that “they are discoveries rather than inventions—they represent reality.” In other words, they exist in the world before anyone identifies them. But discovery (as well as invention) suggests mental action in the production of a concept. A concept is simply a feature of a set of contingencies which exist in the world, and it is discovered simply in the sense that the contingencies bring behavior under its control. The statement “Scientific concepts enable certain aspects of the enormous complexity of the world to be handled by men’s minds” is vastly improved by substituting “human beings” for “men’s minds.”
The traditional notion of meaning and referent runs into trouble when we begin to analyze larger verbal responses under the control of more complex environmental circumstances. What are the referents of sentences—not to mention paragraphs, chapters, or books? A sentence surely means more than its separate words mean. Sentences do more than refer to things; they say things. But what are the things they say? A traditional answer is “Propositions.” But propositions are as elusive as meanings. Bertrand Russell’s view has been paraphrased as follows: “The significance of a sentence is that which is common to a sentence in one language and its translation into another language. For example, ‘I am hungry’ and ‘J’ai faim’ have in common elements which constitute the significance of a sentence. This common element is the proposition.” But what is this common element? Where is it to be found? A dictionary that gave the meanings of sentences would simply contain other sentences having the same meanings.
A translation can best be defined as a verbal stimulus that has the same effect as the original (or as much of the same effect as possible) on a different verbal community. A French translation of an English book is not another statement of a set of propositions; it is another sample of verbal behavior having an effect upon a French reader similar to the effect of the English version on an English reader. The same interpretation may be made of a translation from one medium into another. It has been said that the Prelude to Tristan and Isolde is “an astonishingly intense and faithful translation into music of the emotions which accompany the union of a pair of lovers.” Rather than try to identify the feeling, let alone the proposition, which is thus translated, we may say simply that the music has something of the same effect as physical union.
The concepts of expression and communication may be treated in a similar way. A speaker or a listener responds to conditions of his body which he has learned to call feelings, but what he says or hears is behavior, due to contingencies of which the felt conditions may be by-products. To say that music expresses “what is inexpressible in cognitive, and especially in scientific, language” is to say that it has an effect that verbal behavior cannot have. Verbal behavior does not communicate feelings, though it may result in conditions similarly felt. It does not communicate propositions or instructions. To “instruct” a mother cat to desert her young by delivering an electric shock to a part of her brain does not communicate an instruction that was first held in the mind of the scientist; the shock simply has an effect (a dash of cold water would have produced the same result). Von Frisch’s account of the language of bees (an account which is becoming increasingly suspect) did not make him a Champollion, reading a Rosetta stone.
The concept of stimulus control replaces the notion of referent with respect not only to responses which occur in isolation and are called words (such as nouns and adjectives) but also to those complex responses called sentences. Possibly “fact” could be said to describe a referent of the latter, although its suggestion of truth versus falsity raises difficulties. The child responds in sentences to events in his environment—events involving more than one property or thing, or relations among things, or relations of actor and acted upon, and so on, and his responses contain elements which he never has any occasion to emit alone. The linguist assigns these elements to syntax or grammar. He does so as part of an analysis of the practices of a given verbal community, from which he extracts rules which may be used in the construction of new sentences, as we shall see in Chapter 8.
Structuralism has been strongly encouraged in linguistics because verbal behavior often seems to have an independent status. We are inclined to give special attention to its form because we can report it easily, and rather accurately, simply by modeling it, as in a direct quotation. The report “He said, ‘hammer’ ” gives a much more complete description of the topography of his behavior than “He was hammering.” In teaching a child to talk, or an adult to pronounce a difficult word, we produce a model—that is, we say the word and arrange contingencies under which a response having similar properties will be reinforced. There is nothing especially verbal about modeling (in teaching sports or the dance, the instructor “shows a person what to do” in the sense of doing it himself), but with the invention of the alphabet, it became possible to record verbal behavior, and the records, free of any supporting environment, seemed to have an independent existence. A speaker is said to “know” a poem or an oath or a prayer. Early education in China and Greece was largely a matter of memorizing literary works. The student seemed to know the wisdom expressed by the work, even though his behavior was not necessarily under the control of the conditions which induced the original speaker or writer, or an informed listener, to respond in a given way.
Verbal behavior has this kind of independent status when it is in transmission between speaker and listener—for example, when it is the “information” passing over a telephone wire or between writer and reader in the form of a text. Until fairly recently, linguistics and literary criticism confined themselves almost exclusively to the analyses of written records. If these had any meaning, it was the meaning for the reader, since the circumstances under which the behavior had been produced by the writer had been forgotten, if they were ever known.
The availability of verbal behavior in this apparently objective form has caused a great deal of trouble. By dividing such records into words and sentences without regard to the conditions under which the behavior was emitted, we neglect the meaning for the speaker or writer, and almost half the field of verbal behavior therefore escapes attention. Worse still, bits of recorded speech are moved about to compose new “sentences,” which are then analyzed for their truth or falsity (in terms of their effect on a reader or listener), although they were never generated by a speaker. Both logician and linguist tend to create new sentences in this way, which they then treat as if they were the records of emitted verbal behavior. If we take the sentence “The sun is a star” and put the word “not” in the proper place, we transform it into “The sun is not a star” but no one has emitted this instance of a verbal response, and it does not describe a fact or express a proposition. It is simply the result of a mechanical process.
Perhaps there is no harm in playing with sentences in this way or in analyzing the kinds of transformations which do or do not make sentences acceptable to the ordinary reader, but it is still a waste of time, particularly when the sentences thus generated could not have been emitted as verbal behavior. A classical example is a paradox, such as “This sentence is false,” which appears to be true if false and false if true. The important thing to consider is that no one could ever have emitted the sentence as verbal behavior. A sentence must be in existence before a speaker can say, “This sentence is false,” and the response itself will not serve, since it did not exist until it was emitted. What the logician or linguist calls a sentence is not necessarily verbal behavior in any sense which calls for a behavioral analysis.
The transformational rules which generate sentences acceptable to a listener may be of interest, but even so it is a mistake to suppose that verbal behavior is generated by them. Thus, we may analyze the behavior of small children and discover that, for example, part of their speech consists of a small class of “modifiers” and a larger class of “nouns.” (This fact about verbal behavior is due to the contingencies of reinforcement arranged by most verbal communities.) It does not follow that the child “forms a noun phrase of a given type” by “selecting first one word from the small class of modifiers and selecting second one word from the large class of nouns.” This is a linguist’s reconstruction after the fact.
The analysis of verbal behavior, particularly the so-called discovery of grammar, came very late. For thousands of years no one could have known he was speaking according to rule. What happens when rules are discovered will be considered in Chapter 8.
Development. An undue concern for the structure of verbal behavior has encouraged the metaphor of development or growth. Length of utterance is plotted as a function of age, and semantic and grammatical features are observed as they “develop.” The growth of language in a child is easily compared with the growth of an embryo, and grammar can then be attributed to rules possessed by the child at birth. A program in the form of a genetic code is said to “initiate and guide early learning … as a child acquires language.” But the human species did not evolve because of an inbuilt design: it evolved through selection under contingencies of survival, as the child’s verbal behavior evolves under the selective action of contingencies of reinforcement. As I have noted, the world of a child develops, too.
A child does seem to acquire a verbal repertoire at an amazing speed, but we should not overestimate the accomplishment or attribute it to invented linguistic capacities. A child may “learn to use a new word” as the effect of a single reinforcement, but it learns to do nonverbal things with comparable speed. The verbal behavior is impressive in part because the topography is conspicuous and easily identified and in part because it suggests hidden meanings.
If the structuralists and developmentalists had not confined themselves so narrowly to the topography of behavior at the expense of the other parts of the contingencies of reinforcement, we should know much more about how a child learns to speak. We know the words a child first uses and the characteristic orders in which they tend to be used. We know the length of utterances at given ages, and so on. If structure were enough, that would be the whole story. But a record of topography needs to be supplemented by an equally detailed record of the conditions under which it was acquired. What speech has the child heard? Under what circumstances has he heard it? What effects has he achieved when he has uttered similar responses? Until we have this kind of information, the success or failure of any analysis of verbal behavior cannot be judged.
In verbal behavior, as in all operant behavior, original forms of response are evoked by situations to which a person has not previously been exposed. The origin of behavior is not unlike the origin of species. New combinations of stimuli appear in new settings, and responses which describe them may never have been made by the speaker before, or heard or read by him in the speech of others. There are many behavioral processes generating “mutations,” which are then subject to the selective action of contingencies of reinforcement. We all produce novel forms—for example, in neologisms, blends, portmanteau words, witty remarks involving distortion, and the mistakes of hasty speech.
A great deal has been made of the fact that a child will “invent” a weak past tense for a strong verb, as in saying “he goed” instead of “he went.” If he has never heard the form “goed” (that is, if he has associated only with adults), he must have created a new form. But we do not speak of “creation” if, having acquired a list of color words and a list of object words, he for the first time says “purple automobile.” The fact that the terminal “-ed” suggests “grammar” is unnecessarily exciting. It is quite possible that it is a separable operant, as a separate indicator of the past tense or of completed action in another language might be, and that “go” and a terminal “-ed” are put together, as “purple” and “automobile” are put together, on a novel occasion. The so-called creative aspect of verbal behavior will be mentioned again later.