CEPA eprint 1322 (EVG-032)

Signs, communication, and language

Glasersfeld E. von (1974) Signs, communication, and language. Journal of Human Evolution 3: 465–474. Available at http://cepa.info/1322
Noting the vagueness and confusion in the literature on animal behavior with regard to the phenomenon of communication, an attempt is made to define “sign” and “communication” from the point of view of cybernetics, drawing also on Susanne Langer’s philosophy of symbols. The author argues that neither communication nor language can be successfully investigated as long as the dogmatic behaviorist’s taboo against goal-directed or purposive activities is maintained. In accord with recent developments in feedback theory, he shows that the cyber­neticist’s concept of purpose can be used quite legitimately in the scientific analysis of behavior and that it makes possible a viable classification of signs as well as the discrimination of communication from other types of interaction. In conclusion some of Hockett’s “design features” of langu­age are discussed and an attempt is made to formulate criteria for the form of communcation that should be called linguistic.
In everday usage we are not often bothered by doubts as to what is meant by “com­munication”. We take it to mean that one conveys to another what he feels, thinks, knows, or does – and often also what he would like the other to do. Yet, when the life sciences investigate the phenomenon, the everyday concept quickly disintegrates. Colin Cherry, in one of the most valid introductions to communication science, says (1957, p. 6) : “Scientific usage frequently needs to be more restricted but should not violate common sense …” The study of communication in living organisms has frequently disregarded both the first and the second part of Cherry’s admonition.
In these pages I shall try to point out some conceptual and definitional adjustments that seem indispensable if we want to clear up the vagueness, ambiguities and con­tradictions that pervade the literature on animal communication and language. The stress is on some. It would take more than a few pages to provide the complete set of unequivocally defined terms needed to map such a wide conceptual area.
In a recent paper, of which I became aware only after having finished this one, Donald MacKay (1972) covers a good deal more ground and comes, I believe, to much the same general conclusions. I shall be content to suggest a number of possible moves towards a less confounding use of “sign” and “symbol”, and of the more general terms “com­munication” and “language”. My suggestions are not particularly original. Most of them have been formulated years ago in a branch of science that, in spite of its technicality, is based upon the every-day concept of communication. Cybernetics, as Norbert Wiener (1948) defined it, is the study of “control and communication in the animal and the machine” (my italics). Though Wiener has been cited by authors concerned with animal com­munication (e.g., Altmann, 1967; Hockett, 1960; Sebeok, 1972), most of them limited their interest to two rather specific points while disregarding the general cybernetical framework within which Wiener placed all communicatory processes.
The first of these points concerns the fact that communication events do not conform to the laws of mechanics; i.e. the energy inherent in the communicatory action in no way determines the energy involved in the reaction (Rosenblueth, Wiener & Bigelow, 1943; Haldane, 1955). For instance, the energy of the sound waves emanating from a man shouting “fire!” is in no fixed or calculable relation to the energy of the event triggered by the shout, an event which may be the explosion that sends a six-inch shell on its way, the panic of a crowd in a theatre, or some bystander’s question “where?”. This fact is essential if we want to keep communication apart from mechanical or chemical action-reaction events. If we do not make this distinction, we get definitions of “com­munication” which include every sort of transaction; e.g. “Communication can be said to occur whenever the activities of one animal influence the activities of another animal” (Alexander, 1960, p. 38).
The second point is the one Wiener made in his introduction to the volume Cybernetics (1948, p. 18) : ”… it is completely impossible to understand social communities, .… without a thorough investigation of their means of communication …”
Both these points have to do with the observation of communicatory events, and that, presumably, is the reason why they were readily adopted by professional observers of animal behavior. Taken out of the cybernetical context, however, they do not elucidate the function a communicatory event might have in the organisation of either a social group or the behavior of the individual. In fact, to adopt only these two points is to disregard the very essence of the cybernetical approach.
We do not have to look far to discover the reason why strict behaviorists were reluctant to appropriate anything more from cybernetics. The original essay that laid the foun­dations of the cybernetical approach had the title “Behavior, Purpose and Teleology” (Rosenbluth, Wiener & Bigelow, 1943). Both purpose and teleology were outlawed in strict behaviorism. As Marler put it (1961, p. 296) : “This concept (purposiveness), which may also be associated with an anthropocentric viewpoint, has bedevilled investi­gations of animal behavior in the past.” (And he cites Thorpe, 1956.) A page earlier, Marler states that he finds many of the ideas expressed in Cherry’s book (1957) relevant to his investigation and he quotes that author as saying that communication is “the establishment of a social unit from individuals by the use of language or signs”. The quotation is taken from the Appendix to Cherry’s book (p. 305), but it constitutes only half the definition given there for the term “communication”. The other half reads : “The sharing of common sets of rules, for various goal-seeking activities. (There are many shades of opinion)” (Cherry’s italics). Evidently “goal-seeking” was too close to the devil’s activities.
Much the same has happened with references to Shannon’s Mathematical Theory of Communication (1949). Though it is cited by a number of authors (e.g. Hockett, 1960, 1963; Altmann, 1967; Marler, 1961; Sebeok, 1965, 1972), they do not mention the fact that this theory was conceived, developed and formulated within a very specific context, namely the improvement of communication systems with regard to channel capacity and reliability of signal identification by the receiver, and that, for that reason, it implies the purposiveness of the communicated messages. From this point of view, then, it does not make sense to drop, as Emlen (1960) suggests, all “teleological”, “causal”, and “contextual” terms from the description and classification of animal sounds. If this were done, the term and the concept “communication” would have to be dropped, too.
Not all investigators of animal behavior, however, are continually preoccupied with exorcising the devil of purposiveness. Menzel, in a recent paper (1973), applies MacKay’s definition of communication (MacKay, 1972), which makes the requirement of goal­directedness quite explicit. Earl W. Count (1969), one of the very few who, before adopting the use of cybernetical terms, have gone to the sources from which cybernetics developed, introduces the term “information-ecology” and has no qualms about char­acterising a living organism as a “purposeful machine”.
At this point it should be made clear that, in cybernetics, the words “purpose” and “goal” do not mean exactly what they mean in ordinary English usage. “Teleology has been discredited chiefly because it was defined to imply a cause subsequent in time to a given effect. When this aspect of teleology was dismissed, however, the associated recognition of the importance of purpose was also unfortunately discarded… . We have restricted the connotation of teleological behavior by applying this designation only to purposeful reactions which are controlled by the error of the reaction, i.e. by the difference between the state of the behaving object at any time and the final state interpreted as the purpose. Teleological behavior thus becomes synonymous with behavior controlled by negative feedback …” (Rosenblueth, Wiener & Bigelow, 1943, p. 24). The function which was called “feedback” is, in Wiener’s most general definition (1950, p. 47) : “The property of being able to adjust future conduct by past performance.”
With regard to the analysis of behavior, the contemporary development of feedback theory (cf. Powers, I 973b) has made it a good deal easier to see the relationship between the purposive behavior model of the cyberneticist and the stimulus-response model of the traditional behaviorist. The S-R theorist has always sought to describe and predict an organism’s behavior exclusively on the basis of directly observable stimuli. The feedback theorist, on the other hand, sees the organism’s behavior as activity directed towards the minimisation of any discrepancy between a sensory input value (stimulus) and a pre­established “reference value” to which the particular type of sensory input is compared. This discrepancy constitutes “negative feedback”; the larger it is, the more activity the organism has to expand in order to bring the sensory input value to congruence with the reference value. In this sense, and only in this sense, can the organism’s activity be called “purposive” and the reference value a “goal”.
In her classification of behavioral situations which can be assumed to constitute contexts for communicatory events, Busnel (1969, p. 35) lists seven types (e.g. meeting of the sexes, search for food, territorial defence, etc.). For each of them, feedback theory would posit at least one reference value. The organism’s behavior, insofar as it can be reliably recognised as belonging to one of the seven situations, can then be described and predicted in purposive terms as striving towards adequation of the sensory input to the relevant reference value. The more highly evolved the organism, the richer is the variety of sensory feedback it receives and the repertoire of behaviors it employs to adequate it. Francisco Ayala (1974, p. 350) has recently mapped the evolutionary beginnings of information processing: “The most rudimentary ability to gather and process in­formation about the environment can be found in certain single-celled eukaryotes. A Paramecium swims following a sinuous path ingesting the bacteria that it encounters as it swims. Whenever it meets unfavourable conditions, like unsuitable acidity or salinity in the water, the Paramecium checks its advance, turns and starts in a new direction. Its reaction is purely negative… . Greater ability to process information about the environment occurs in the single-celled Euglena … it not only avoids unsuitable environ­ments but also actively seeks suitable ones. An amoeba represents further progress in the same direction; it reacts to light by moving away from it, and also actively pursues food particles.” These capabilities of avoiding, seeking and pursuing can be explained as the result of different simple feedback loops, each with its specific reference value.
The transfer of signals from sense organs to a central processor that computes negative feedback, and from the processor to other organs that implement the adequating activities, is often said to involve communication; but this communication takes place within one organism. When we ordinarily speak of communication, however, we intend what Wiener (1948) called “intercommunication”, i.e. the transfer of signals or messages from one organism to another. While the first is explicitly part of a feedback loop, it may not be immediately obvious that the second, too, must be viewed within the framework of a feedback mechanism. Failure to realise this is, I believe, one of the main reasons for confusion in the literature on animal communication.
Considered as an instrument of feedback adequation, intercommunication is a type of activity that appears in the behavioral repertoire of organisms above a certain level of complexity. While all interaction below that level is a matter of “energetics” (Count, 1969), interaction above that level can be a matter of “information”. We can distinguish the energetic type of interaction from the one involving information by means of the criterion formulated by Wiener (1948) and by Haldane (1955). But this, though a necessary condition, does not yet yield a sufficient definition of “communication”. Just as it does not make sense to call all events involving a mechanical transfer of energy between two organisms “interaction”, it is misleading to speak of all events involving the passage of information as “intercommunication”.
If I land with my face in the gutter, because someone rushing along knocked me down, the transfer of energy may have been very much the same as in the case where a well- aimed punch from a mugger dispatches me into a similar terminal position. Yet, the behavioral situation would be quite different. In the first case the event is “accidental”, in the sense that neither I nor my terminal position had anything to do with the rushing person’s reference value nor with the negative feedback he was trying to reduce by rushing; whereas in the second case, hitting me and dispatching me into the gutter was indeed “purposive” on the part of the mugger because he had learned in the course of previous experience that these activities were effective in bringing about a coincidence of his sensory input and his reference signal (e.g. my wallet in his hand). We might call the one “fortuitous”, the other “instrumental”. In ordinary English we should speak of the first event as a “collision” and of the second as an “attack”.
The point I am trying to make, however, has to be made, not on the level of words, but on the level of concepts. We can discriminate the two behavioral situations without reference to will, consciousness, mind, or any other concept “associated with an anthropo­centric viewpoint”; they can be discriminated by specifying the operative feedback loops – and there is nothing particularly anthropocentric about feedback loops, since they are an integral part, not only of Man, but of primitive living organisms and of machines as well.
To specify an operative feedback loop and its reference value may be a difficult task under certain circumstances. Powers (1973a,b) has suggested an experimental approach to this problem that has the virtue of great flexibility. No doubt, more work will have to be done in that direction. But even if we come across behavioral situations where, in practice, we are unable to specify the organism’s reference value with the desired precision, this could hardly invalidate the extrapolation of the theoretical concept, once the “referent” of that concept has been demonstrated and measured in many other cases.
If we apply the discrimination of accidental versus purposive (or fortuitous versus instrumental) activity to the realm of communication between organisms, we find that it is largely compatible with the classification of signs worked out by Susanne Langer (1942). She distinguishes “natural” from “artificial” signs. “A natural sign is a part of a greater event or of a complex condition, and to an experienced observer it signifies the rest of that situation of which it is a notable feature. It is a symptom of a state of affairs.” On the other hand, if “arbitrary events are produced which are purposely correlated with important events that are to be their meaning,” we have “artificial signs” (Langer, 1948, p. 59).
As examples of natural signs, Langer gives: wet streets as a sign that it has rained (past event) ; the smell of smoke as a sign of fire (present event); a ring round the moon as a sign that it is going to rain (future event) ; thunder as a sign that there has been lightning and lightning as a sign that there will be thunder. In every case, the item we choose to consider a “sign” is an event which, in our prior experience, has been reliably correlated with another event; and since, as Hume put it, we believe that future experience will largely resemble past experience, the occurrence of the sign-event leads us to anticipate its correlate. If the perceiving organism has associated a behavioral response with the correlate-event, the behaviorist calls the sign-event a “discriminative stimulus”; if the anticipation, in a cognising organism, leads to an internal representation or thought, the philosopher calls it “inductive inference”; on both levels it constitutes the simplest and most general form of learning and knowledge.
As samples of artificial signs, Langer lists: a whistle means that the train is about to start; a gunshot. means that the sun is just setting. The difference between these and the natural signs is clear. Neither the whistle nor the gunshot were (or could be) corre­lated with the train’s starting or the sun’s setting before they were introduced qua signs, simply because they did not happen within those contexts; they were made to happen in order to signify (cf. Hockett’s discussion of Paul Revere, 1960, p. 420).
Langer, unfortunately, blurs the issue a little by placing what she calls ’’fortuitous signs’’ on the side of the artificial signs. Fortuitous signs are, for instance, the warning cries of animals, such as the various vocalisations of the vervet monkey that alert other members of the group to the proximity of a large or small mammalian or avian predator respectively (Struhsaker, 1967). Similarly, some of the various sounds “produced incidentally to other behaviors’’ by aquatic organisms (Tavolga, 1960, p. 94) seem to function as natural signs. It would, indeed, be anthropocentric to assume that the vervets invented these cries in order to signal the presence of specific predators; but such an assumption is unnecessary. The distinction Langer presumably had in mind is this: w bile the signs she calls ’’natural’’ are events which, from the human observer’s point of view, are in some causal connection to other elements of the event or state of affairs which they signify, the signs she calls “fortuitous” are not. This distinction is the same that creates some difficulty for Hockett in his discussion of the design feature he calls “specialisation” ( 1960, p. 408). Whether the vervet’s utterance of a cry can be causally linked to its flight effort, or whether it is to be considered merely a fortuitous corollary of flight, depends on the sophistication of our causal explanations and does not affect the semiotic status the cry has for vervets. For them it has become a sign and not an “artificial” one, to be sure – because it happened frequently enough in flight situations to have been associated by them with flight and the presence of specific predators. In other words, when a vervet hears such a cry, it responds to it on the basis of its previous experience of the associated situation and, therefore, as though it actually perceived the predator.
To conclude this investigation of the sign-concept, I should like to stress that all the items which, according to our definition, could be called “sign”, must have one feature in common: they stand in a one-to-one relationship to the occurrence of the event or situation they signify. They are “natural” signs if they co-occur so frequently with an event or situation that they come to signify it because they become associated with that event by the receiver. Alternatively, they are “artificial” signs if, and only if; their association with the signified event or situation has been contrived, in that the sign-event is made to happen by the sender in order to be interpreted as a sign by the receiver.
This differentiation between natural and artificial signs, in many cases, is quite irrelevant to the interpreter of the sign. From the point of view of the traveller lingering at the bar of the railroad station, there is no essential difference between hearing carriage doors being slammed, the station master’s shout “All aboard!”, or the whistle of the locomotive. All three are signs that the train is about to leave. The traveller does or does not respond to them, regardless of any knowledge as to the fact that the first of them is an integral part of the departure event, while the second and third were contrived and added for purely semiotic reasons. Yet, it is precisely this distinction which enables us to differentiate between events that have the instrumental function of communication and events which have no such instrumental function.
The distinction is parallel to the one suggested for “interaction”. It can be made only by reference to the sender of the sign, i.e. by an analysis of the sender’s situation, his operative feedback loops, and his activities as instruments for the adequation of his input signals to a specific reference value or goal.
In this context, Tinbergen’s concept of “incipient movements” (1952 ; cited by Altmann, 1967) or incipient behavior in general, may be helpful in explaining the evolutionary development of artificial signs out of natural signs. It would be absurd to consider a bite the sign for a bite; but the biter’s opening his mouth may become a sign that a bite is forthcoming. With more experience, an organism could learn that another’s raising his head or getting up in a particular way may also herald a bite. Carpenter (1969, p. 46), for instance, describes the development of such an anticipatory sign in the interaction of an infant howler monkey and its mother. Thus the chain of connected events becomes extended and the interval between the sign-event and the signified event becomes greater. The interpreting organism obviously gains an advantage from this extension. But so, does the sending organism. If biting was an activity that had become instrumental in keeping the immediate environment clear of other organisms (implying a reference value with regard to which an intruder would constitute a disturbance), the biting organism will soon learn that getting up or opening his mouth is sufficient activity to eliminate his negative feedback. From there to “ritualised” behavior is a small step; and, once certain acts or incipient acts have taken over the instrumental efficacy that originally belonged to the whole chain of which they were merely the beginning, a similar chain may form at the response side of the transaction, replacing the interpreter’s im­mediate response by a subsequent one, and so on until there is no longer any obvious experiential connection (from the observer’s point of view) between the sign and the receiver’s ensuing activity.
Once we reach such complex levels of semiotic exchange, the sign itself; no matter how we analyse it, gives us no clue as to its signification, nor can we be sure to infer its signifi­cation correctly from the receiver’s response. In many situations, however, there is a fairly reliable clue and it has been intuitively used by every observer of animal communication, though, to my knowledge, it has never been made explicit. It is the simple fact that the sender will cease to emit the sign whenever his environment, which often includes the receiver of the sign, has changed in such a way that the discrepancy between his sensory signals and his reference value has been eliminated. (Note that this approach necessarily puts relatively permanent incidental changes of the environment, e.g. an animal’s tracks or odoriferous deposits, into the class of fortuitous or natural signs.) Marler’s compendium of aggregation and dispersal signs (1961) is an example in which the relevant reference values can be easily and accurately specified. These reference values are distances between the individual and other members of his group or between the sender’s group and other groups. The heard (received) vocalisations provide feedback as to the actual distance. Whenever there is a discrepancy between that and the reference value, a “distance-increasing” or a “distance-reducing” vocalisation is emitted and the instrumental function, i.e. the purpose, of these vocalisations is the elimination of that discrepancy.
In other situations it may be much more difficult to infer the specific reference value that gives a sign its purposive aspect. It may, indeed, require some form of experiment­ation, i.e. deliberate, methodical interference of “disturbance” on the part of the observer (Powers, 1973a,b). But by varying the sensory input to an organism and by observing both the kind and the amount of activity with which the organism responds to the disturbance, we can eventually establish what the reference value is, to which the organism strives to adequate his input.
This isolation and specification of the reference value or goal is essential if we want to make a distinction between behaviors that have a semiotic function and all the other behaviors that have no such function and are significant only in the sense that they happen to have been experientially associated with certain events or states of affairs. In other words, it is only when we are able to specify the instrumental relation between a sign and the goal towards which the sender of the sign is striving, that we would speak of “communication”.
Once we adopt the feedback model of behavior, the importance of communication in the development and maintenance of social organisation can be easily explained. What welds a number of individual organisms into a social group is not merely the similarity of certain behaviors, but rather the commonality of certain reference values or goals. Because of this commonality, the activity of one individual will frequently eliminate not only his own feedback discrepancy but also the feedback discrepancy of others. This creates a mutual or reciprocal instrumentality: others are instrumental in the adequation of my input to my reference, just as I am instrumental in the adequation of their input to their reference. This mutual instrumentality, which has nothing whatever to do with the ethics of altruism, makes for the cohesion of the group.
This is not the place to discuss how concerted efforts, division of tasks, and, eventually, reciprocal dependencies develop out of an original commonality of reference values, but it should be clear that signs, signals, and communication play an indispensable role in the development of any such organisational feats. Sebeok (1972, p. 79) speaks of members of a species becoming “integrated into ’interest communities’, joined together, for instance, for mutual protection – like a school of porpoises.” Teleki (1973) observed and described the cooperative behavior of chimpanzees in predation and meat-sharing. Malinowski, who half a century ago investigated communication among primitive tribes, said (1923, p. 310) : “Speech is the necessary means of communication; it is the one indispensable instrument for creating the ties of the moment without which unified social action is impossible.” He used the terms “speech” and “language”, and though he discriminated clearly between language functioning “as a link in concerted human activity, as a piece of human action” and language as “a condensed piece of reflection, a record of fact or thought,” (loc. cit., p. 312) the two categories have been confounded in almost everything that has been written about language and linguistic communication since then.
The difference between the use of a significatory item as an instrument of action and its use as an instrument of reflection was stressed once more by Susanne Langer. She contrasts “signs” with “symbols” (1948, p. 61) : “To conceive a thing or a situation is not the same as to ’react toward it’ overtly or to be aware of its presence. In talking about things we have conceptions of them, not the things themselves; and it is the con­ceptions, not the things, that symbols directly ’mean’ ” (Langer’s italics).
This is a paraphrase of the basic semantic relation Ogden & Richards (1923) illustrated by their famous triangle. The fact that the symbol is associated with a conception, and only in some cases, and then indirectly, with a “thing”, is what releases the symbol from the one-to-one correspondence with the occurrence of an experiential situation. Thus a sign becomes a symbol when it can be used without direct connection to an experiential context. The linguistic items we normally call “words”, though they can, of course, be used as signs, are primarily symbols. We can use them in operational loops where neither the input nor the reference values are perceptual and where the activities performed to eliminate discrepancies between the two are not observable behavior but private reflective operations that never become directly manifest in public results. We can use them to modify another’s reference values, to supplement his repertoire of instru­mental activities, to change his representation of the world he experiences, i.e. his “knowledge”.
This removal from the actual experiential situation is not quite the same as the removal Hockett characterised with his design feature of” displacement” (1960, p. 415). The proto­hominoid, who spotted a predator, took fright, and “did not immediately burst out with his danger call, but first sneaked silently away towards the remainder of his band,” may, indeed, have achieved a kind of displacement of the danger call that had survival value, but the mere delay does not change the one-to-one correspondence between the sign and the actual perceptual situation. The danger call remained wholly instrumental in a feedback loop that was disturbed by the sensory input provided by the actual predator. The danger call will become a symbol only when a later hominoid can utter it without having spotted a predator and when the other members of the band can hear it without taking flight. The displacement characterised by Hockett may well be a plausible step in that development, but it is at best a preparatory step and, as such, not a criterion for the symbolicity that is an essential feature of “language”.
Hockett subsequently (1963, p. 10) introduced an additional design feature that implies symbolicity, i.e. “prevarication”, and he remarks: “It ought to be noted that without the property here labelled ’prevarication’, the formulation of hypotheses is impossible.” It seems unlikely that we could ever establish by direct observation that another organism’s sign is uttered in the hypothetical mode, unless the organism already has a conventional sign for indicating this. What primatologists have called a “play-face”, i.e. a sign indicating that a subsequent aggressive behavior is not to be taken seriously, is a case in point. It should also be noted that in Piaget’s theory of cognitive development the sym­bolic function first arises in play (Piaget, 1967, p. 90).
Though the meaning of symbols is conceptual and, therefore, by definition inside an organism and not directly observable, their function can be demonstrated. Premack in his artificial communication system with the chimpanzee Sarah, introduced questions and negation at an early stage (Premack, 1971). Both require conceptual representations that come very close to what we call an hypothesis and the successful handling of a combination of the two proves the presence of some symbolicity.
Successful handling, in this context, is not a question of behavior, if by “behavior” we intend the responses which the behaviorist dogma has sanctioned as observable. That an organism handles a symbol successfully can be shown only by the use of other symbols, because the reference values as well as the “inputs” that have to be adequated to them are not on the perceptual level. If, for instance, we show an orange to a chim­panzee and then ask him, by means of any signs he has learned to use, “Is ’apple’ the name of this?”, the correct response is to say “No”. What makes it correct is not the perceptual difference between the apple-sign and the orange (any sign would be different from an orange), but the difference between the concept associated with the apple-sign and the concept by means of which the orange is recognised as an orange. The kind of item we mean by “concept of an orange” has been discussed with admirable lucidity by Bronowski & Bellugi (1970). In order to respond correctly, a comparison has to be carried out on the conceptual level, which is not the case in the usual discrimination tasks given to experimental animals. In this very specific sense the combination of question and negation is a test for symbol manipulation.
When we ordinarily use the word “language”, we certainly take it for granted that we are dealing with a system whose elements are capable, not only of displacement, but also of functioning as symbols. But there is also a more easily discernible feature which we take for granted. What we call “language” must be a combinatorial system of symbols which, to quote Hockett (1960, p. 418) once more, ”… provides certain patterns by which these elementary significant units can be combined into larger sequences, and conventions governing what sorts of meanings emerge from the arrangements.” Hockett judiciously avoids the term “syntax” and speaks of the “sorts of meanings that emerge from the arrangements”. This is important because linguists have traditionally under­stated, if not disregarded, the semantic dimension of syntax. The crucial feature, then is not that language has rules for stringing symbols together, but that the stringing together adds another level of meaning. “To paint a red house” contains the very same elementary word-symbols as “to paint a house red”, but some of the relational concepts indicated by the two strings are different. It is this feature of combinatorial meaning that leads to the “openness” of linguistic communication systems and gives the user the possibility of “productivity” or “novelty” of expression. Combinatorial meaning and symbolicity of the elements, thus, yield a reliable criterion for distinguishing between language and other communication systems whose elements have a necessary one-to-one correspondence to the occurrence of events or states of affairs in the experiential context in which they are used.
Preparation of this manuscript was supported by National Institutes of Health grants HD-0616 and RR-00165, by the Georgia Follow Through Program, and by the Department of Psychology of the University of Georgia. The ideas expressed in it do not necessarily coincide with the views of the sponsors or of my colleagues.
Alexander, Richard D. (1960) Sound communication in orthoptera and cicadidae. In: Lanyon W. E. & Tavolga W. N. (eds.) Animal Sounds and Communication: 38–92. Washington D. C.: American Institute of Biological Sciences.
Altmann, Stuart A. (1967) The structure of primate social communication. In: Altmann S. A. (ed.) Social Communication among Primates: 325–362. Chicago, Ill.: U. of Chicago Press.
Ayala, Francisco J. (1974) The concept of biological progress. In: Ayala F. J. (ed.) Problems of Reduction in Biology: 339–354. London: Macmillan (in press)
Bronowski J. & Bellugi U. (1970) Language, name, and concept. Science 168: 669–673.
Busnel, Marie-Claire (1969) The psychophysiological study of acoustic communication in animals. In: Sebeok T. A. & Ramsay A., (eds.) Approaches to Animal Communication: 31–39. The Hague, Mouton.
Carpenter C. Ray (1969) Approaches to studies of the naturalistic communicative behavior in nonhuman primates. In: Sebeok T. A. & Ramsay A., (eds.) Approaches to Animal Communication: 40–70. The Hague: Mouton.
Cherry, Colin (1957) On Human Communication. Cambridge, Massachusetts: M. I. T. Press (2nd edition, 1966)
Count, Earl W. (1969) Animal communication in man-science: An essay in perspective. In: Sebeok T. A. & Ramsay A., (eds.) Approaches to Animal Communication: 71–137. The Hague: Mouton.
Emlen, John T., Jr. (1960) Introduction. In: Lanyon W. E. & Tavolga W. N. (eds.) Animal Sounds and Communication: Ix-xiii. Washington D. C.: American Institute of Biological Sciences.
Haldane J. B. S. (1955) Animal communication and the origin of human language, Science Progress 43: 383–401.
Hockett, Charles F. (1960) Logical considerations in the study of animal communication. In: Lanyon W. E. & Tavolga W. N., (eds.) Animal Sounds and Communication: 392–430. Washington D. C.: American Institute of Biological Sciences.
Hockett, Charles F. (1963) The problem of universals in language. In: Greenberg J. H. (ed.) Universals of Language: 1–22. Cambridge, Massachusetts: M. I. T. Press.
Langer, Susanne K. (1942) Philosophy in a New Key. New York: Mentor Books (paperback edition, 1948)
MacKay, Donald M. (1972) Formal analysis of communicative processes. In: Hinde R. A. (ed.) Non-Verbal Communication: 3–26. Cambridge, England: Cambridge U. Press.
Malinowski, Bronislaw (1923) The problem of meaning in primitive languages. In: Ogden C. K. & Richards I. A. ) The Meaning of Meaning: 296–336. New York: Harcourt, Brace & Co. (8th edition, 1946)
Marler, Peter (1961) Aggregation and dispersal: Two functions in primate communication. In: Jay P. C. (ed.) Primates, Studies in Adaptation and Variability: 420–438. New York: Holt, Rinehart and Winston.
Marler, Peter (1961) The logical analysis of animal communication. Journal of Theoretical Biology 1: 295–317.
Menzel, Emil W. (1973) Leadership and communication in young chimpanzees. Proceedings IVth International Congress of Primatology. Vol. 1: 192–225. Basel: Karger.
Ogden C. K. & Richards I. A. (1923) The Meaning of Meaning. New York: Harcourt, Brace & Co. (8th edition, 1946)
Piaget, Jean (1967) Six Psychological Studies. New York: Random House.
Powers, William T (19736) Behavior: The Control of Perception. Chicago, Ill.: Aldine.
Powers, William T (1973a) Feedback: Beyond behaviorism. Science 17:: 351–356.
Premack, David (1971) Some general characteristics of a method for teaching language to organisms that do not ordinarily acquire it. In: Jarrard L. E. (ed.) Cognitive Processes of Nonhuman Primates: 47–82. New York: Academic Press.
Rosenblueth A., Wiener N. & Bigelow J. (1943) Behavior, purpose, and teleology. Journal of Philosophy of Science 10: 18–24. http://cepa.info/2691
Sebeok, Thomas A. (1965) Animal communication. Science 147: 1006–1014.
Sebeok, Thomas A. (1972) Perspectives in Zoosemiotics. The Hague: Mouton.
Shannon, Claude E. (1949) The Mathematical Theory of Communication, (with introduction by Weaver) Urbana, Ill.: U. of Illinois Press (paperbound edition, 1963)
Struhsaker, Thomas T. (1967) Auditory communication among vervet monkeys. In: Altmann S. A. (ed.) Social Communication among Primates: 281–324. Chicago, Ill.: U. of Chicago Press.
Tavolga, William N (1960) Sound production and underwater communication in fishes. In: Lanyon W. E. & Tavolga W. N., (eds.) Animal Sounds and Communication: 93–136. Washington D. C.: American Institute of Biological Sciences.
Teleki, Geza (1973) The omnivorous chimpanzee, Scientific American 228 1: 33–42.
Thorpe W. H. (1956) Learning and Instinct in Animals. London: Methuen.
Tinbergen, Niko (1952) ’Derived’ activities: Their causation, biological significance, origin, and emancipation during evolution. Quarterly Review of Biology 27 1: 1–32.
Wiener, Norbert (1948) Cybernetics. Cambridge, Mass. MIT Press (2nd paperback edition, 1965)
Wiener, Norbert (1950) The Human Use of Human Beings. New York: Avon (paperback edition, 1967
Found a mistake? Contact corrections/at/cepa.infoDownloaded from http://cepa.info/1322 on 2016-09-16 · Publication curated by Alexander Riegler