“Psycholinguistics” in “PSYCHOLINGUISTICS” | Open Indiana

In the most general sense, we have communication whenever one system, a source, influences another system, a destination, by manipulation of the alternative signals which can be carried in the channel connecting them. The information source is conceived as producing one or more messages which must be transformed by a transmitter into signals which the channel can carry; these signals must then be transformed by a receiver back into messages which can be accepted at the destination. This minimal system, borrowed from Shannon’s discussion of the theory of information¹ and diagrammed in Figure 1, has been applied, with great generality, to information transmission in electrical, biological, psychological and social systems as well as language communication in the strict sense. In a telephone communication system, for example, the messages produced by a speaker are in the form of variable sound pressures and frequencies which must be transformed into proportional electrical signals by the transmitter; these signals are carried over wire (channel) to a receiver which transforms them back into the variable sound pressures and frequencies which constitute the message to be utilized by the listener. The activity of the transmitter is usually referred to as encoding and that of the receiver as decoding. Anything that produces variability at the destination which is unpredictable from variability introduced at the source is called noise.

Figure 1

This model of the communication process, developed in connection with engineering problems, was not intended to provide a satisfactory picture of human communication. For one thing, it implies a necessary separation of source and destination, of transmitter and receiver, which is usually true of mechanical- communication systems but not of human ones. The individual human functions more or less simultaneously as a source and destination and as a transmitter and receiver of messages—indeed, he is regularly a decoder of the messages he himself encodes through various feedback mechanisms. Each individual in a speech community may be conceived as a more or less self-contained communicating system, encompassing in his nervous apparatus, from receptors to effectors, all of the components shown in Figure 1. If we rearrange the components in Shannon’s model in the fashion shown in Figure 2, what might be called a communication unit is described, equipped to both receive and send messages. In the process of human decoding, input of some form of physical energy, linguistically or otherwise coded, is first recoded into sensory neural impulses, operated upon by receiving apparatus, and finally ‘interpreted’ at the destination (presumably as some pattern of activity in the higher centers). In the process of human encoding, an ‘intention’ of the source (presumably some pattern of activity in the same centers) is operated upon by transmitting apparatus in the motor areas, is recoded into physical movements, and becomes the output of this unit. Translating into traditional psychological language, input becomes equivalent to ‘stimulus/receiver becomes ‘reception’ and ‘perception,’ destination and source become ‘cognition’ (meaning, attitude, and the like), transmitter becomes ‘motor organization and sequencing,’ and output becomes ‘response.’

Figure 2

Another insufficiency of engineering models for human communication purposes is that they are not designed to take into account the meaning of signals, e.g., their significance when viewed from the decoding side and their intention when viewed from the encoding side. The research generated by such models has dealt almost exclusively with relations between transmitter and receiver, or with the individual as a single system intervening between input and output signals. This has not been because of lack of awareness of the problem of meaning or its importance, but rather because it is admittedly difficult to be rigorous, objective, and quantitative at this level. Nevertheless, one of the central problems in psycholinguistics is to make as explicit as possible relations between message events and cognitive events, both on decoding and encoding sides of the equation.

Human communication is chiefly a social affair. Any adequate model must therefore include at least two communicating units, a source unit (speaker) and a destination unit (hearer). Between any two such units, connecting them into a single system, is what we may call the message. For purposes of this report, we will define message as that part of the total output (responses) of a source unit which simultaneously may be a part of the total input (stimuli) to a destination unit. When individual A talks to individual B, for example, his postures, gestures, facial expressions and even manipulations with objects (e.g., laying down a playing card, pushing a bowl of food within reach) may all be part of the message, as of course are events in the sound wave channel. But other parts of A’s total behavior (e.g., breathing, toe-wiggling, thinking) may not affect B at all and other parts of the total stimulation to B (e.g., sensations from B’s own posture, cues from the remainder of the environment) do not derive from A’s behavior—these events are not part of the message as we use the term. These R-S message events (reactions of one individual that produce stimuli for another) may be either immediate or mediate—ordinary face-to-face conversation illustrates the former and written communication (along with musical recordings, art objects, and so forth) illustrates the latter.

Figure 3

Figure 3 presents a model of the essential communication act, encoding of a message by a source unit and decoding of that message by a destination unit. Since the distinction between source and destination within the same communicator (as shown in Figure 2) seems relevant only with respect to the direction of information exchange (e.g., whether the communicator is decoding or encoding), we substitute the single term mediator for that system which intervenes between receiving and transmitting operations. The ways in which the various sciences concerned with human communication impinge upon and divide up the total process can be shown in relation to this figure.

1.2 Disciplines Concerned with Human Communication

Microlinguistics (or linguistics proper) deals with the structure of messages, the signals in the channel connecting communicators, as events independent of the characteristics of either speakers or hearers. Once messages have been encoded and are “on the air,” so to speak, they can be described as objective, natural science events in their own right. In an even stricter sense, the linguist is concerned with determining the code of a given signal system, the sets of distinctions which are significant in differentiating alternative messages. The term exolinguistics (sometimes called metalinguistics) has been used rather loosely by linguists to cover all those other aspects of language study which concern relations between the characteristics of messages and the characteristics of individuals who produce and receive them, including both their behavior and culture. Whether or not the grammatical structure of a language influences the thinking of those who speak it is thus an exolinguistic problem. The social sciences in general, and psychology, sociology, and anthropology in particular, are concerned with the characteristics of human organisms and societies which influence the selection and interpretation of messages—attitudes, meanings, social roles, values, and so forth. The rather new discipline coming to be known as psycholinguistics (paralleling the closely related discipline termed ethnolinguistics) is concerned in the broadest sense with relations between messages and the characteristics of human individuals who select and interpret them. In a narrower sense, psycholinguistics studies those processes whereby the intentions of speakers are transformed into signals in the culturally accepted code and whereby these signals are transformed into the interpretations of hearers. In other words, psycholinguistics deals directly with the processes of encoding and decoding as they relate states of messages to states of communicators. The terminal aspect of human speech encoding, production of speech sounds, is the special province of phonetics. Similarly, the initial aspect of human speech decoding, whereby sound pressures and frequencies are transformed into impulses in auditory nerve fibers and relayed to the cortex, is a special field of psychoacoustics. Finally, the science of human communication would be concerned with relations between sources who select messages and destinations who interpret and are affected by them. In the broadest sense, therefore, human communications as a science includes the other disciplines that have been mentioned; in a narrower sense—and one more in keeping with contemporary activities—students of communications research have usually worked at grosser levels of analysis, concerning themselves with sources such as radio and the newspaper and destinations such as the mass audience, members of another culture, and so on.

1.3 Plan of This Report

Psycholinguistics is that one of the disciplines studying human communication which is most directly concerned with the processes of decoding and encoding. What are the major divisions within psycholinguistics itself? Mapping of this area was one of the tasks of the seminar, but it was done in a casual manner and appears as a spontaneous clustering of the research problems the participants found significant. In other words, the organization of the field of psycholinguistics followed here is one that the members of this seminar found fruitful.

Section 2 of this report provides brief and non-technical orientations to the three approaches to language study, linguistics, learning theory, and information theory, in which we were particularly interested. The members of the seminar spent the first few weeks in such orientation as a means of providing themselves with a more homogeneous background, and most readers of this report are probably in the same position we were in, e.g., perhaps trained in linguistics but not psychology and only remotely conversant with information theory, or possibly familiar with both learning and information theory but entirely vague about linguistics. During the course of this orientational work, discussion by the seminar repeatedly devolved upon the problem of psycholinguistic units—the need for clearly defined units in quantitative research, the relevance of available linguistic units, and so on. Although we have been able to do no more than set up the problem and suggest possible ways of attacking it, the prior importance of this issue justifies a separate treatment, given in section 3.

The body of this report presents theoretical analyses and suggested research within specific areas. At the time of presentation of these research problems for preliminary discussion by the seminar, it became clear that we could not organize this field in terms of the three methodological approaches, linguistics, information theory, and learning theory, since each problem seemed to require combinations of techniques drawn from all three aproaches. Rather, the various problems suggested by members of the seminar seemed to fall quite naturally into clusters based on similarity of content and underlying theory. During roughly the last half of the seminar period, its members worked in overlapping groups of about three or four people on such clusters of related problems, reporting back to the seminar as a whole for general discussion. These work-group reports, as written up by the chairman of each group, form the basis for the remainder of this published report.

The organization of content in psycholinguistics developed by the seminar can perhaps best be seen by reference to Figure 4. The temporal dimension runs, as usual, from left to right. Brief sequences of time are indicated by the banded arrows. Periods A and B may refer to either two different stages within the development of an individual speaker or two different stages in the development of a language within a speech community. The upper half of the figure represents diagrammatically the interacting levels of behavioral organization within the individual; this is the special province of psychology and, more remotely, of the other social sciences. The lower half of the figure represents the various levels or bands of the message; this is the special province of linguistics and, programmatically, kinesics (study of facial and bodily gestures) and, more remotely, all disciplines concerned with media (content analysis, aesthetics, etc.).

Figure 4

The levels within the communicator are here labeled cognitive states, motive states, anticipational and dispositional states (or sets), and sensory and motor skills—these labels are intended to be suggestive, not limiting. Synchronic Psychology would deal with organization both within these levels and between them in decoding and encoding. The various synchronous bands which comprise messages are here labelled linguistic, kinesic, situational (e.g., manipulation of significant objects, arrangement of the social or physical situation in which communication takes place) and ‘other’ (e.g., odor, warmth, touch, and other modalities which may contribute to communication)—signals in any of these bands may be either naturally or arbitrarily coded. Synchronic Linguistics in the broad sense would deal with both organization within these bands (e.g., descriptive linguistics deals specifically with the structure of linguistically coded stimuli) and between these bands (e.g., relations between linguistic, kinesic codes and the like). Synchronic Psycholinguistics deals with relations between momentary psychological states of communicators and momentary states of messages. Since a large number of problems fall in this area, the seminar divided them into two groups: Synchronic Psycholinguistics I.: Microstructure (relations of phonemic units of messages to perceptual and motor discrimination in communicators, for example) is discussed in section 4; Synchronic Psycholinguistics II.: Macrostructure (problems of meaning, of relations of language to thought and culture, for example) is discussed in section 7. This distinction between microstructure and macrostructure is probably not a happy one, but it seemed to serve our purposes.

Over short periods of time, at least, events at any psychological level are to some degree predictable from antecedent events at either the same level or other levels. Principles of association, for example, are concerned with the dependence of one cognitive state upon another. Similarly, enforced regularities in either input or output events (e.g., grammatical regularities) may give rise to sequential neural organization. Study of problems of this order could be called Sequential Psychology. On the message side, likewise, events at one point in time can be shown to be dependent to varying degrees upon antecedent message events—presumably such phenomena could be studied within kinesic or other bands as well as the linguistic. Such study could be called Sequential Linguistics. The relations between transitional sequences in messages and transitional sequencing mechanisms in the communicator is the field of Sequential Psycholinguistics, and problems in this area are discussed in section 5.

When the psychologist deals with changes in organization, either through maturation or through learning, he makes comparisons between two stages of performance in time (e.g., pre-training and post-training) and this might well be termed Diachronic Psychology. The same term could be applied to differences in organization between two stages in culture, e.g., comparison of the habits or associations between S and R for two sets of individuals at two discrete times. When the linguist compares the structures of messages produced by members of the same speech community at two discrete periods in time, this is called Diachronic Linguistics. In this report, we are extending the same term to cover the linguist’s comparisons between the structures of messages produced by the same individual at two discrete periods in time—that is, to the study of first language learning, second language learning and bilingualism. Diachronic Psycholinguistics would be concerned with relations between the changing behavioral organizations of either the individual or the group and the changing structures of messages they produce, particularly the application of learning principles to these problems. This area is discussed in section 6.

¹ Shannon and Weaver, The mathematical theory of communication (University of Illinois Press, 1949). Mathematical aspects of Shannon’s theory of signal transmission are discussed in section 2.3. of this report.

Psycholinguistics