Skip to content

Preface to the Glossary of terms

In talking with students of behavior on both sides of the Atlantic, I often found myself faced with certain vocabulary problems, even when using English. For a common term, there was not always a common referent; for a common referent, there was not always a common term. Even when both term and referent seemed identical, there were often misleading shades of meaning. Much of the work published by ethologists, in fact, is not readily intelligible to “rat” behaviorists, and many publications by the latter that could be of interest (not to say importance) to ethologists might as well have been written in an obscure Sanskrit dialect. Perhaps it would be better if they had been, since then a need for translation would be apparent. Not only are some writings unreadable, but still others that can be “read” are read erroneously, and both psychological and ethological concepts are sometimes referred to inappropriate observational and theoretical contexts.

Even when the referents of terms agree, the words employed to label them sometimes are misinterpreted because of their connotations in the vernacular or in other scientific contexts. Some scientists react with negativism, if not hostility, to a concept because a particular term is used to designate it; but they may accept without misgivings concepts that are essentially repugnant to them, solely because someone made a happy choice of words to name them. By designating a member of one class of stimuli as a Gestalt, for example, one may alienate the very people who need and use such a concept (under other labels) and may achieve its acceptance by those to whom the objective approach to behavior is a source of moral indignation. Even observations may be questioned or rejected, not on empirical grounds, but simply because a particular word was chosen to label the observation. Terms such as conditioning are not used by some because it implies to them that the Pavlovian model of brain physiology is valid. Along with the term, they ignore or deprecate a whole area of research. Imprinting is rejected by others because “there is no perceptual learning” or because someone wonders what gets imprinted where; obtaining an unpalatable answer, they assume that there is no class of events to which the term can be applied. All this is perhaps unfortunate when one recalls that these problems arise in a scientific context. But scientists are human and behave according to laws that do not always work out toward most rapid scientific progress.
Glossaries that have been prepared are not always as useful as they might be. Most American behaviorists, perhaps oversensitive to the demands of the logical analysts, are not solely vocabulary-conscious–they choose words carefully–but they are also definition-conscious–they like to keep track of what they are talking about. Some of the terms they use are defined ostensively: they point. Others are defined by more elaborate operations. Still others are theoretically defined. Data-language terms (i.e., terms used in describing what happens) are distinguished from theoretical terms and from names for processes (i.e., behavior changes that are functions of time). For this reason, these men become unhappy when they encounter sets of terms including members of all these groups placed together and endowed with literary definitions.
By “literary” definitions, I mean definitions stated in words occurring in common speech for which no clear and unequivocal referent in behavior, for which no coordinating definition or operation that might provide such a referent, is or can be stated. What, for example, does “fuller organization of sense data” mean, and what is “a complex of states”? When should one talk about “sensory integration” or the “apprehension of relations”? Presumably somebody “knows”–but I don’t.
Most sets of definitions, whether carefully assembled or implicitly available to anyone willing to search the indexes of systematic works, give such difficulties. There are “empirical” terms that have no referent or whose referents are so stated that one does not quite know whether, in any one case, to stick an electrode or a qualification into them (center). There are terms that represent laboratory slang apotheosized into scientific terminology (e.g., superstition) and terms that represent the reification of processes (e.g., mimesis). Some are defined adequately, even elegantly in the literary sense; but they turn out to be meaningless, self-contradictory, or contradictory to other definitions in the same system when the definitions of the words used in them are examined. Other terms have been borrowed from physiology to label a phenomenon solely by rough analogy. Still others must be classified as intuitive; if one does not know what the term “means” in the first place, he will never learn. These lack empirical objective referents no matter how careful the attempts to define, to redefine, to determine usage.1 What they mean depends on their connotations for each person who uses them. Yet others are theoretical and incorporate empirically defined terms. Some terms are used in two ways: empirically, as a label for a class of behavior, and theoretically, as the name of a concept relating that behavior to other classes of events. Displacement activity is a theoretical concept, for which Tinbergen (29) has given a definition that is one of the clearest and most precise theoretical definitions that can be found. But in use, it is also an empirical term, and its application as such is independent of Tinbergen’s definition. That one term has two such uses may be misleading, but not seriously so if the user recognizes the two. Otherwise, mere labels become explanatory concepts. It is no wonder that difficulties arise.
Yet words defined in so many and such arbitrary ways are often used consistently and intelligibly. This suggests that there is something wrong with the definitions that have appeared. Without specifying exactly what is wrong, beyond their sheer incapacity to communicate information about behavior or to tell one how and when to use a word, one can propose alternative definitions based on the fact that the terms are used with some success. If usage is watched, if one watches and listens to people using terms, if the events to which the terms are applied are observed, the fact becomes clear that many such terms have clear-cut referents and are hence capable of objective (i.e., empirical) definition–they are labels. This would be surprising if it were not true that the empirical observations with which the vocabulary has to do, and which it purports to describe and explain, are made on the same subject matter: the behavior of living animals, scientifically observed.
The philosophical analysts, or logical positivists, and other methodologists have proposed methods for the elaboration and clarification of vocabularies that have helped make communication more precise. Three classes of terms, for example, can be isolated. The first class falls into what we shall call the data-language, the terms and syntax used in describing observations. If words are to be used in the data-language, they must be defined so that anyone after a minimum of training can use them consistently. The terms must refer to objective events so that statements embodying them can be verified directly by other people. Terms that do not lead to such agreed upon statements fall outside of this class. Theoretical statements cannot appear in this language, but only statements such as: “The rat turned left and ran into the goal-box.” “The man in the brown coat is eating a hamburger.” or “Three chaffinches are mobbing an owl.” The words that appear in the data-language are typically defined by simple (and perhaps repeated) empirical designation, that is, by pointing.
A second class of terms is what I shall here call “empirically” or “symptomatically” defined terms. They are labels applied to classes of statements in the data-language; their application depends upon the verification of one or more statements in the data-language. The term extinction, for example, can be applied if and only if certain observations are made. We must state the operations which lead to the observation, and we must state an outcome of the operations too. Agreement among observers that such and such an observation has been made is implicitly required before the word can be applied. Theoretical statements are not necessarily made when members of this class of terms are employed in a statement, although, as we shall see, theoretical statements necessarily employ them.
“Operational” definitions include all that fall into these first two classes. They demand agreement, and they make it possible for anyone who is able to read to reconstruct the observations to which the terms apply. It does not follow that all terms that are operationally defined will be equally useful to the behavior scientist; many such empirically defined concepts may prove to be of limited value and interest. The scientist can take them or leave them. The operational definition of a term may be modified at a later date. Two operationally defined concepts may prove to be equivalent, and one term will then cover both. Such definitions state the observational conditions under which an observer may apply a particular term. They are, as it were, a description of the discriminative stimuli, releasers, or sign stimuli for a piece of verbal behavior.
The third class of terms includes all those defined by the exclusive use of members of the first two. Such terms are theoretical in that they may propose relationships among members of the earlier classes and in that they may introduce new concepts that hypothesize classes of events, states, structures, or mechanisms that have not yet been directly observed in connection with the behavioral events to which they are theoretically related. Statements incorporating theoretical terms are subject to verification, either through experimental test of their consequences or by direct observation. Thus, theoretical terms may graduate, as it were, into the class of empirical terms.
Heuristically, a set of definitions stated with these principles in mind may serve some useful purposes. It should permit the behaviorist to learn what ethology is about in terms of behavior and at the same time make the behaviorist literature intelligible to ethologists. It should make it possible to translate statements using terms familiar to one group into statements using terms familiar to the other.
It may do more than provide for direct translation. When each concept is examined in its experimental and theoretical context, when the causal statements summarizing and defining it are analyzed following the criteria which we have stated, other possibilities arise. Analysis enables one to distinguish between observational and theoretical levels of discourse and so may indicate areas in which empirical data are not yet adequate to provide the basis for theoretical advance; hence, it may pinpoint areas where research is badly needed. Certain concepts, often treated as purely speculative, may prove to have sound empirical bases. Analysis may show that some “empirical” terms have no referents; the definition of a term in the data-language may include a statement about events that cannot be observed–there are no unicorns, and perhaps there is no sensory preconditioning either. Other terms, used as though they were empirical and frequently referred to as “things” that belong to the second set of terms, may turn out to refer, not to empirical goats, but to theoretical sheep. The status of various other theoretical concepts may also be clarified. Some may prove to be “intuitive,” that is to say, undefinable without the introduction of nonempirical, nonobjective sets of referents, and hence to have no place in scientific statement.2
Finally, two apparently different terms may prove to have the same, or almost the same, empirical content so that the number of theoretical entities (and hence theoretical problems) can be reduced, the body of experimental data related to the survivor can be amplified, and new experimental hypotheses can be generated. One can state with some precision the theoretical relations between Lorenz’ action-specific energy and Skinner’s reflex reserve, and between Tinbergen’s motivation and Hull’s sEr. What is more important, one can state the relationship of both to experimental observations of behavior.
Attempting to follow the principles and the objectives that have been stated, I have prepared this glossary for the use of those who work with behavior as an objective datum. In defining terms, I have tried to follow usage in stating definitions, in preference to following or restating literary definitions however well put. It follows that the definitions will not always correspond to those which have appeared in other publications and, hence, that not everyone will agree with them. I have tried to avoid arbitrary redefinition and to restate concepts, in so far as possible, on the basis of usage in the literature, in the laboratory, and in field observations. Occasionally, and especially with theoretical terms, a usage definition had to be arrived at inductively on the basis of numerous instances of the appearance of the word.
The major rules followed in preparing the glossary are given at the end of the Preface.
The glossary is not an exhaustive one–it cannot pretend to be. I have tried to include as many useful terms as possible, but some important ones have doubtless been omitted, just as some trivial ones have crept in. It arrives at a thoroughly Skinnerian and ethological point of view toward behavior. (The Hullian terms, fitted into place, will mislead no one; their usefulness rests upon the redefinition of the response concept.) But this, I submit, is where any such attempt at clarifying psychological terms will land us. We always have to go back to pieces of behavior that we can identify, and we have to build from them.
I am the first to recognize that writing this glossary is a rash enterprise and that I have been presumptuous indeed. Someone, I suspect, has to be. Like all glossaries (back to Dr. Johnson’s dictionary), it is a personal one and admittedly exhibits certain biases, not all of which I am probably aware of. Certainly, not everyone will be happy with many or all of the definitions given. This is so if only because not all behavior theorists and not all ethologists agree on the exact usage of several terms. I have tried to follow what seemed to be the most common usage, even when it did not conform with the usage intended by the men who first introduced the term into the vocabulary of behavior. No doubt, some of the definitions are based on misunderstandings; but, if this is so, others too will have misunderstood. I hope that those who read and use these definitions will offer emendations or alternative correct definitions so that my errors will not (or cannot) be repeated by me or by others who have been led into the same misunderstanding.
The glossary, then, is intended to provide an empirical vocabulary that can be used by anyone in the science of “human” or “animal” behavior. It may serve to familiarize its readers with some of the very great developments in the study of animal behavior and to suggest their relevance to human behavior. Perhaps it will clarify the status of some of the concepts that have been used in both areas in the past. To some it may suggest, as it has to me, new lines of research on humans. To others it may recommend itself as a supplementary text.
But if this glossary provokes thought, discussion, and argument, if it increases mutual intelligibility between ethologists and other behavior students even a little, it will have served its purpose.


Cambridge, Massachusetts
February, 1955.
1 To risk a pun, the writer remains ignorant of what a cognition is. So far as he knows, he has never had one, and no one has ever been able to correct him on this, or tell him how to have one or how to recognize it if he did.

2 Some will question this cavalier treatment of “intuitive” concepts; some may even suggest that these are more important than others. There is certainly a place for intuitive concepts, which are often a major tool in an experimenter’s pre-experimental and pre-theoretical speculations, as are his casual and quite uncontrolled observations. Indeed, I suspect that a major part of the behaviorist’s problem is to provide a scientific account of how and why intuitive concepts are developed and used, that is, to “explain” them in terms of human verbal behavior and its discriminative control. But intuitive concepts do not belong in statements of fact or in statements of theory set up to deal with fact, just as casual observations have no scientific status as fact. In passing, ethologists may be interested in the treatment accorded to the (often intuitive) concept “emotion” in a recent 800-page advanced text of experimental psychology (19). The word simply does not appear in the book; two derivatives appear once each.

Major Rules Followed in Preparing the Glossary

1. Terms having empirical, symptomatic status are defined in the data-language or in other empirical terms and are designated emp. Such definitions are given to some terms that seem to warrant it but that have not previously been endowed with empirical status.
2. Many terms having sound theoretical status are defined in terms of the empirical concepts on which they are based. For others, such definitions are overlong and they are explicated elsewhere (e.g., in Hull’s Principles of Behavior), therefore their content is stated in general terms and their full definitions are omitted here. In both cases, their definitions are designated th. Terms that pretend to theoretical status, but that are not reducible to empirical terms, share with their intuitive, literary, and conversational fellows the label con.
3. Where terms have several usages, each is defined separately.
4. This glossary has omitted all but a few data-language terms, including the names of specific responses that have been studied extensively. From the psychological vocabulary such self-explanatory terms are omitted as “turning right,” “pressing a bar,” “eating a pellet,” “jumping to the left,” and “pecking”; and from the ethological vocabulary, such terms as “mobbing,” “preening,” “egg rolling,” “brooding,” “forward threat posture,” “greeting ceremony,” and “long call,” to name but a few. These are all ostensively defined. The terms retained define apparatus that may be unfamiliar to ethologists (e.g., Lashley jumping-stand).
5. Laboratory slang is designated lab slang.
6. The meaning of some terms has evolved considerably; in some instances, words that have been widely used are dropping out now. I have tried to indicate such terms and to state the current usage. (See, for example, innate and its cross references.)
7. To save space, and to obviate the coining of new terms, the definitions are stated in the broader, more widely used vocabulary of behavior theory.
8. When the usages of ethologists and behavior theorists (see glossary) differ, each is labeled appropriately (eth and bt). Similarly, terms used by only one of the two groups are designated appropriately.
9. Although many of the words have been borrowed from physiology, only a few of their definitions in physiology are given.3 Physiological definitions (designated physiol) appear only when it seemed interesting to compare the two usages and when the comparison might suggest the experimental and theoretical steps to be taken before the behavioral and physiological definitions of the same word can state the same concept.
10. Whenever it seemed desirable, brief discussion of the concept and its definition, and examples of usage, are appended.
11. Insofar as possible, all behavioral terms appearing in a definition are themselves defined in the glossary. The reader, I hope, will find occasion to check this statement.
12. No attempt has been made to incorporate the whole system of terms that various theorists have developed. For these, the reader is referred to the works themselves, which appear in the list of references.

3 This absence of physiological definitions should be interpreted as showing, not an “anti-physiological” bias, but rather a distaste for “physiologizing,” i.e., for explaining behavior in terms of the properties of a mythical nervous system with handy (but unverifiable) characteristics. A greater volume of experimental work on the physiology of the behavior that behaviorists deal with will presumably make the inclusion of physiological definitions necessary in future glossaries such as the present one.

Creative Commons License is licensed by Admin under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.