B. F. Skinner

The Origins of Cognitive Thought

Recent Issues in the Analysis of Behavior (1989) publ. Merrill Publishing Company.

One Chapter reproduced here.

What is felt when one has a feeling is a condition of one's body, and the word used to describe it almost always comes from the word for the cause of the condition felt. The evidence is to be found in the history of the language-in the etymology of the words that refer to feelings (see Chapter 1). Etymology is the archaeology of thought. The great authority in English is the Oxford English Dictionary (1928), but a smaller work such as Skeat's Etymological Dictionary of the English Language (1956) will usually suffice. We do not have all the facts we should like to have, because the earliest meanings of many words have been lost, but we have enough to make a plausible general case. To describe great pain, for example, we say agony. The word first meant struggling or wrestling, a familiar cause of great pain. When other things felt the same way, the same word was used.

A similar case is made here for the words we use to refer to states of mind or cognitive processes. They almost always began as references either to some aspect of behaviour or to the setting in which behaviour occurred. Only very slowly have they become the vocabulary of something called mind. Experience is a good example. As Raymond Williams (1976) has pointed out, the word was not used to refer to anything felt or introspectively observed until the 19th century. Before that time it meant, quite literally, something a person had "gone through" (from the Latin expiriri), or what we should now call an exposure to contingencies of reinforcement. This paper reviews about 80 other words for states of mind or cognitive processes. They are grouped according to the bodily conditions that prevail when we are doing things, sensing things, changing the way we do or sense things (learning), staying changed (remembering), wanting, waiting, thinking, and "using our minds."

DOING

The word behave is a latecomer. The older word was do. As the very long entry in the Oxford English Dictionary (1928) shows, do has always emphasised consequences-the effect one has on the world. We describe much of what we ourselves do with the words we use to describe what others do. When asked, "What did you do?", "What are you doing?", or "What are you going to do?" we say, for example, "I wrote a letter," "I am reading a good book," or "I shall watch television." But how can we describe what we feel or introspectively observe at the time?

There is often very little to observe. Behaviour often seems spontaneous; it simply happens. We say it "occurs" as in "It occurred to me to go for a walk." We often replace "it" with "thought" or "idea" ("The thought-or idea-occurred to me to go for a walk"), but what, if anything, occurs is the walk. We also say that behaviour comes into our possession. We announce the happy appearance of the solution to a problem by saying "I have it!"

We report an early stage of behaving when we say, "I feel like going for a walk." That may mean "I feel as I have felt in the past when I have set out for a walk." What is felt may also include something of the present occasion, as if to say, "Under these conditions I often go for a walk" or it may include some state of deprivation or aversive stimulation, as if to say, "I need a breath of fresh air."

The bodily condition associated with a high probability that we shall behave or do something is harder to pin down and we resort to metaphor. Since things often fall in the direction in which they lean, we say we are inclined to do something, or have an inclination to do it. If we are strongly inclined, we may even say we are bent on doing it. Since things also often move in the direction in which they are pulled, we say that we tend to do things (from the Latin tendere, to stretch or extend) or that our behaviour expresses an intention, a cognitive process widely favoured by philosophers at the present time.

We also use attitude to refer to probability. An attitude is the position, posture, or pose we take when we are about to do something. For example, the pose of actors suggests something of what they are engaged in doing or are likely to do in a moment. The same sense of pose is found in dispose and propose ("I am disposed to go for a walk...... I propose to go for a walk"). Originally a synonym of propose, purpose has caused a great deal of trouble. Like other words suggesting probable action, it seems to point to the future. The future cannot be acting now, however, and elsewhere in science purpose has given way to words referring to past consequences. When philosophers speak of intention, for example, they are almost always speaking of operant behaviour.

As an experimental analysis has shown, behaviour is shaped and maintained by its consequences, but only by consequences that lie in the past. We do what we do because of what has happened, not what will happen. Unfortunately, what has happened leaves few observable traces, and why we do what we do and how likely we are to do it are therefore largely beyond the reach of introspection. Perhaps that is why, as we shall see later, behaviour has so often been attributed to an initiating, originating, or creative act of will.

SENSING

To respond effectively to the world around us, we must see, hear, smell, taste, or feel it. The ways in which behaviour is brought under the control of stimuli can be analysed without too much trouble, but what we observe when we see ourselves seeing something is the source of a great misunderstanding. We say we perceive the world in the literal sense of taking it in (from the Latin per and capere, to take). (Comprehend is a close synonym, part of which comes from prehendere, to seize or grasp.) We say, "I take your meaning." Since we cannot take in the world itself, it has been assumed that we must make a copy. Making a copy cannot be all there is to seeing, however, because we still have to see the copy. Copy theory involves an infinite regress. Some cognitive psychologists have tried to avoid it by saying that what is taken in is a representation perhaps a digital rather than an analog copy. When we recall ("call up an image of") what we have seen, however, we see something that looks pretty much like what we saw in the first place, and that would be an analog copy. Another way to avoid the regress is to say that at some point we interpret the copy or representation. The origins of interpret are obscure, but the word seems to have had some connection with price; an interpreter was once a broker. Interpret seems to have meant evaluate. It can best be understood as something we do.

The metaphor of copy theory has obvious sources. When things reinforce our looking at them, we continue to look. We keep a few such things near us so that we can look at them whenever we like. If we cannot keep the things themselves, we make copies of them, such as paintings or photographs. Image, a word for an internal copy, comes from the Latin imago. It first meant a colored bust, rather like a wax-work museum effigy. Later it meant ghost. Effigy, by the way, is well chosen as a word for a copy, because it first meant something constructed-from the Latin fingere. There is no evidence, however, that we construct anything when we see the world around us or when we see that we are seeing it.

A behavioural account of sensing is simpler. Seeing is behaving and, like all behaving, is to be explained either by natural selection (many animals respond visually shortly after birth) or operant conditioning. We do not see the world by taking it in and processing it. The world takes control of behaviour when either survival or reinforcement has been contingent upon it. That can occur only when something is done about what is seen. Seeing is only part of behaving; it is behaving up to the point of action. Since behaviour analysts deal only with complete instances of behaviour, the sensing part is out of reach of their instruments and methods and must, as we shall see later, be left to physiologists.

CHANGING AND STAYING CHANGED

Learning is not doing; it is changing what we do. We may see that behaviour has changed, but we do not see the changing. We see reinforcing consequences but not how they cause a change. Since the observable effects of reinforcement are usually not immediate, we often overlook the connection. Behaviour is then often said -to grow or develop. Develop originally meant to unfold, as one unfolds a letter. We assume that what we see was there from the start. Like pre-Darwinian evolution (where to evolve meant to unroll as one unrolled a scroll), developmentalism is a form of creationism.

Copies or representations play an important part in cognitive theories of learning and memory, where they raise problems that do not arise in a behavioural analysis. When we must describe something that is no longer present, the traditional view is that we recall the copy we have stored. In a behavioural analysis, contingencies of reinforcement change the way we respond to stimuli. It is a changed person, not a memory, that has been "stored."

Storage and retrieval become much more complicated when we learn and recall how something is done. It is easy to make copies of things we see, but how can we make copies of the things we do? We can model behaviour for someone to imitate, but a model cannot be stored. The traditional solution is to go digital. We say the organism learns and stores rules. When, for example, a hungry rat presses a lever and receives food and the rate of pressing immediately increases, cognitive psychologists want to say that the rat has learned a rule. It now knows and can remember that "pressing the lever produces food." But "pressing the lever produces food" is our description of the contingencies we have built into the apparatus. We have no reason to suppose that the rat formulates and stores such a description. The contingencies change the rat, which then survives as a changed rat. As members of a verbal species we can describe contingencies of reinforcement, and we often do because the descriptions have many practical uses (for example, we can memorise them and say them again whenever circumstances demand it) but there is no introspective or other evidence that we verbally describe every contingency that affects our behaviour, and much evidence to the contrary.

Some of the words we use to describe subsequent occurrences of behaviour suggest storage. Recall-call back-is obviously one of them; recollect suggests "bringing together" stored pieces. Under the influence of the computer, cognitive psychologists have turned to retrieve-literally "to find again" (cf. the French trouver), presumably after a search. The etymology of remember, however, does not imply storage. From the Latin me or, it means to be "mindful of again" and that usually means to do again what we did before. To remember what something looks like is to do what we did when we saw it. We needed no copy then, and we need none now. We recognise things in the sense of "recognising" them responding to them now as we did in the past.) As a thing, a memory must be something stored, but as an action "memorising" simply means doing what we must do to ensure that we can behave again as we are behaving now.

WANTING

Many cognitive terms describe bodily states that arise when strong behaviour cannot be executed because a necessary condition is lacking. The source of a general word for states of that kind is obvious: when something is wanting, we say we want it. In dictionary terms, to want is to "suffer from the want of." Suffer originally meant "to undergo," but now it means "to be in pain," and strong wanting can indeed be painful. We escape from it by doing anything that has been reinforced by the thing that is now wanting and wanted.

A near synonym of want is need. It, too, was first tied closely to suffering; to be in need was to be under restraint or duress. (Words tend to come into use when the conditions they describe are conspicuous.) Felt is often added: one has a felt need. We sometimes distinguish between want and need on the basis of the immediacy of the consequence. Thus, we want something to eat, but we need a taxi in order to do something that will have later consequences.

Wishing and hoping are also states of being unable to do something we are strongly inclined to do. The putted golf ball rolls across the green, but we can only wish or will it into the hole. (Wish is close to will. The Anglo-Saxon willan meant "wish," and the would in "Would that it were so" is not close to the past tense of will.)

When something we need is missing, we say we miss it. When we want something for a long time, we say we long for it. We long to see someone we love who has long been absent.

When past consequences have been aversive, we do not hope, wish, or long for them. Instead, we worry or feel anxious about them. Worry first meant "choke" (a dog worries the rat it has caught), and anxious comes from another word for choke. We cannot do anything about things that have already happened, though we are still affected by them. We say we are sorry for a mistake we have made. Sorry is a weak form of sore. As the slang expression has it, we may be "sore about something." We resent mistreatment, quite literally, by "feeling it again" (resent and sentiment share a root).

Sometimes we cannot act appropriately because we do not have the appropriate behaviour. When we have lost our way, for example, we say we feel lost. To be bewildered is like being in a wilderness. In such a case, we wander ("wend our way aimlessly") or wonder what to do. The wonders of the world were so unusual that no one responded to them in normal ways. We stand in awe of such things, and awe comes from a Greek word that meant "anguish" or "terror." Anguish, like anxiety, once meant "choked," and terror was a violent trembling. A miracle, from the Latin admirare, is "something to be wondered at," or about.

Sometimes we cannot respond because we are taken unawares; we are surprised (the second syllable of which comes from the Latin prehendere, "to seize or grasp"). The story of Dr. Johnson's wife is a useful example. Finding the doctor kissing the maid, she is said to have exclaimed, "I am surprised!" "No," said the doctor, "I am surprised; you are astonished!" Astonished, like astounded, first meant "to be alarmed by thunder." Compare the French etonner and tonnere.

When we cannot easily do something because our behaviour has been mildly punished, we are embarrassed or barred. Conflicting responses find us perplexed: they are "interwoven" 6r "entangled." When a response has been inconsistently reinforced, we are diffident, in the sense of not trusting. Trust comes from a Teutonic root suggesting consolation, which in turn has a distant Greek relative meaning "whole." Trust is bred by consistency.

WAITING

Wanting, wishing, worrying, resenting, and the like are often called "feelings." More likely to be called "states of mind" are the bodily conditions that result from certain special temporal arrangements of stimuli, responses, and reinforcers. The temporal arrangements are much easier to analyse than the states of mind that are said to result.

Watch is an example. It first meant "to be awake." The night watch was someone who stayed awake. The word alert comes from the Italian for "a military watch." We watch television until we fall asleep.

Those who are awake may be aware of what they are doing; aware is close to wary or cautious. (Cautious comes from a word familiar to us in caveat emptor) Psychologists have been especially interested in awareness, although they have generally used a synonym, consciousness.

One who watches may be waiting for something to happen, but waiting is more than watching. It is something we all do but may not think of as a state of mind. Consider waiting for a bus. Nothing we have ever done has made the bus arrive, but its arrival has reinforced many of the things we do while waiting. For example, we stand where we have most often stood and look in the direction in which we have most often looked when buses have appeared. Seeing a bus has also been strongly reinforced, and we may see one while we are waiting, either in the sense of "thinking what one would look like" or by mistaking a truck for a bus.

Waiting for something to happen is also called expecting, a more prestigious cognitive term. To expect is "to look forward to" (from the Latin expectare). To anticipate is "to do other things beforehand," such as getting the bus fare ready. Part of the word comes from the Latin capere "to take." Both expecting and anticipating are forms of behaviour that have been adventitiously reinforced by the appearance of something.
(Much of what we do when we are waiting is public. Others can see us standing at a bus stop and looking in the direction from which buses come. An observant person may even see us take a step forward when a truck comes into view, or reach for a coin as the bus appears. We ourselves "see" something more, of course. The contingencies have worked private changes in us, to some of which we alone can respond.)

THINKING

It is widely believed that behaviour analysts cannot deal with the cognitive processes called thinking. We often use think to refer to weak behaviour. If we are not quite ready to say, "He is wrong," we say, "I think he is wrong." Think is often a weaker word for know; we say, "I think this is the way to do it" when we are not quite ready to say, "I know this is the way" or "This is the way." We also say think when stronger behaviour is not feasible. Thus, we think of what something looks like when it is not there to see, and we think of doing something that we cannot at the moment do.

Many thought processes, however, have nothing to do with the distinction between weak and strong behaviour or between private and public, overt and covert. To think is to do something that makes other behaviour possible. Solving a problem is an example. A problem is a situation that does not evoke an effective response; we solve it by changing the situation until a response occurs. Telephoning a friend is a problem if we do not know the number, and we solve it by looking up the number. Etymologically, to solve is "to loosen or set free," as sugar is dissolved in coffee. This is the sense in which thinking is responsible for doing. "It is how people think that determines how they act." Hence, the hegemony of mind. But again the terms we use began as references to behaviour. Here are a few examples:

When no effective stimulus is available we sometimes expose one. We discover things by uncovering them. To detect a signal does not mean to respond to it; it means to remove something (the tegmen) that covers it. When we cannot uncover a stimulus, we sometimes keep an accessible one in view until a response occurs. Observe and regard both come from words that meant "to hold or keep in view," the latter from the French garder Consider once meant "to look steadily at the stars until something could be made of them" (consider and sidereal have a common root). Contemplate, another word for think, once meant "to look at a template or plan of the stars." (In those days all one could do to make sense of the stars was to look at them.) We not only look at things to see them better, we look for them. We search or explore. To look for a pen is to do what one has done in the past when a pen came into view. (A pigeon that pecks a spot because doing so has been occasionally reinforced will "look for it" after it has been taken away by doing precisely what it did when the spot was there-moving its head in ways that brought the spot into view.) We search in order to find, and we do not avoid searching by contriving something to be seen, because contrive, like retrieve, is from the French trouver, "to find." We bring different things together to make a single response feasible when we concentrate, from an older word concentre, "to join in a center." We do the reverse when we separate things so that we can more easily deal with them in different ways. We sift them, as if we were putting them through a sieve. The cern in discern (Latin cernere) means "to separate or set apart." We mark things so that we shall be more likely to notice them again. Distinguish, a good cognitive term, once meant "to mark by pricking." Mark is strongly associated with boundaries; animals mark the edges of their territories. To define is literally "to mark the bounds or end" (finis) of something. We also determine what a word means by indicating where the referent terminates. We compare things, literally, by "putting them side by side" so that we can more easily see whether they match. The par in compare means "equal." Par value is equal value. In golf, par is the score to be matched. We speculate about things in the sense of looking at them from different angles, as in a specula or mirror. Cogitate, an old word for think, first meant "to shake up." A conjecture is something "thrown out" for consideration. We accept or reject things that occur to us in the sense of taking or throwing them back, as if we were fishing. Sometimes it helps to change one mode of stimulation into another. We do so when we convert the "heft" of an object into its weight, read on a scale. By weighing things we react more precisely to their weight. Ponder, deliberate, and examine, good cognitive processes, all once meant "to weigh." Ponder is part of ponderous, the liber in deliberate is the Latin libra, "a scales," and examine meant "the tongue of a balance." We react more precisely to the number of things in a group by counting. One way to count is to recite one, two, three, and soon, while ticking off (touching) each item. Before people learned to count, they recorded the number of things in a group by letting a pebble stand for each thing. The pebbles were called calculi and their use calculation. There is a long, but unbroken, road from pebbles to silicon chips. After we have thought for some time, we may reach a decision. To decide once meant simply to cut off or bring to an end. A better word for decide is conclude, "to close a discussion." What we conclude about something is our last word. It is certainly no accident that so many of the terms we now use to refer to cognitive processes once referred either to behaviour or to the occasions on which behaviour occurs. It could be objected, of course, that what a word once meant is not what it means now. Surely there is a difference between weighing a sack of potatoes and weighing the evidence in a court of law. When we speak of weighing evidence we are using a metaphor. But a metaphor is a word that is "carried over" from one referent to another on the basis of a common property. The common property in weighing is the conversion of one kind of thing (potatoes or evidence) into another (a number on a scale or a verdict). Once we have seen this weighing done with potatoes it is easier to see it done with evidence. Over the centuries human behaviour has grown steadily more complex as it has come under the control of more complex environments. The number and complexity of the bodily conditions felt or introspectively observed have grown accordingly, and with them has grown the vocabulary of cognitive thinking.

We could also say that weight becomes abstract when we move from potatoes to evidence. The word is indeed abstracted in the sense of its being drawn away from its original referent, but it continues to refer to a common property, and, as in the case of metaphor, in a possibly more decisive way. The testimony in a trial is much more complex than a sack of potatoes, and "guilty" probably implies more than "ten pounds." But abstraction is not a matter of complexity. Quite the contrary Weight is only one aspect of a potato, and guilt is only one aspect of a person. Weight is as abstract as guilt. It is only under verbal contingencies of reinforcement that we respond to single properties of things or persons. In doing so we abstract the property from the thing or person.

One may still argue that at some point the term is abstracted and carried over, not to a slightly more complex case, but to something of a very different kind. Potatoes are weighed in the physical world; evidence is weighed in the mind, or with the help of the mind, or by the mind. And that brings us to the heart of the matter.

MIND

The battle cry of the cognitive revolution is "Mind is back!" A "great new science of mind" is born. Behaviourism nearly destroyed our concern for it, but behaviourism has been overthrown, and we can take up again where the philosophers and early psychologists left off.

Extraordinary things have certainly been said about the mind. The finest achievements of the species have been attributed to it; it is said to work at miraculous speeds in miraculous ways. But what it is and what it does are still far from clear. We all speak of the mind with little or no hesitation, but we pause when asked for a definition. Dictionaries are of no help. To understand what mind means we must first look up perception, idea, feeling, intention, and many other words we have just examined, and we shall find each of them defined with the help of the others. Perhaps from people who did not know precisely what we were talking about, and we have no sensory nerves going to the parts of the brain in which the most important events presumably occur. Many cognitive psychologists recognise these limitations and dismiss the words we have been examining as the language of "common sense psychology." The mind that has made its comeback is therefore not the mind of Locke or Berkeley or of Wundt or William James. We do not observe it; we infer it. We do not see ourselves processing information, for example. We see the materials that we process and the product, but not the producing. We now treat mental processes like intelligence, personality, or character traits-as things no one ever claims to see through introspection. Whether or not the cognitive revolution has restored mind as the proper subject matter of psychology, it has not restored introspection as the proper way of looking at it. The behaviourists' attack on introspection has been devastating.

Cognitive psychologists have therefore turned to brain science and computer science to confirm their theories. Brain science, they say, will eventually tell us what cognitive processes really are. They will answer, once and for all, the old questions about monism, dualism, and interactionism. By building machines that do what people do, computer science will demonstrate how the mind works.

What is wrong with all this is not what philosophers, psychologists, brain scientists, and computer scientists have found or will find; the error is the direction in which they are looking. No account of what is happening inside the human body, no matter how complete, will explain the origins of human behaviour. What happens inside the body is not a beginning. By looking at how a clock is built, we can explain why it keeps good time, but not why keeping time is important, or how the clock came to be built that way. We must ask the same questions about a person. Why do people do what they do, and why do the bodies that do it have the structures they have? We can trace a small part of human behaviour, and a much larger part of the behaviour of other species, to natural selection and the evolution of the species, but the greater part of human behaviour must be traced to contingencies of reinforcement, especially to the very complex social contingencies we call cultures. Only when we take those histories into account can we explain why people behave as they do.

That position is sometimes characterised as treating a person as a black box and ignoring its contents. Behaviour analysts would study the invention and uses of clocks without asking how clocks are built. But nothing is being ignored. Behaviour analysts leave what is inside the black box to those who have the instruments and methods needed to study it properly. There are two unavoidable gaps in any behavioural account: one between the stimulating action of the environment and the response of the organism, and one between consequences and the resulting change in behaviour. Only brain science can fill those gaps. In doing so it completes the account; it does not give a different account of the same thing. Human behaviour will eventually be explained, because it can only be explained by the cooperative action of ethology, brain science, and behaviour analysis.

The analysis of behaviour need not wait until brain science has done its part. The behavioural facts will not be changed, and they suffice for both a science and a technology. Brain science may discover other kinds of variables affecting behaviour, but it will turn to a behavioural analysis for the clearest account of their effects.

CONCLUSION

Verbal contingencies of reinforcement explain why we report what we feel or introspectively observe. The verbal culture that arranges such contingencies would not have evolved if it had not been useful. Bodily conditions are not the causes of behaviour but they are collateral effects of the causes, and people's answers to questions about how they feel or what they are thinking often tell us something about what has happened to them or what they have done. We can understand them better and are more likely to anticipate what they will do. The words they use are part of a living language that can be used without embarrassment by cognitive psychologists and behaviour analysts alike in their daily lives.

But not in their science! A few traditional terms may survive in the technical language of a science, but they are carefully defined and stripped by usage of their old connotations. Science requires a language. We seem to be giving up the effort to explain our behaviour by reporting what we feel or introspectively observe in our bodies, but we have only begun to construct a science needed to analyse the complex interactions between the environment and the body and the behaviour to which it gives rise.