Skip to main content
SearchLoginLogin or Signup

Chapter XII: Interchange of Messages

Published onDec 01, 1953
Chapter XII: Interchange of Messages

Chapter twelve of Erving Goffman. Communication Conduct in an Island Community. PhD diss., University of Chicago, 1953.


In terms of the framework of this study, the minimal concrete unit of communication consists of the sign behavior of a sender during the whole period of time through which a focus of attention is continually directed at him. This unit of communication may be called a “natural message,” or, where no confusion is likely, simply a “message.” Short messages such as words or single sentences apparently have distinctive phonetic features; long messages, such as an uninterrupted thirty-minute talk, seem less neatly describable in phonetic terms. A single natural message may, of course, involve different pieces of information, but these differences are not relevant in terms of this study.

The concept of natural message has been employed in other studies of social interaction and has apparently been independently hit upon. One student gives the following definition for the basic unity of his study:

The unit of verbal behavior chosen was arbitrarily defined as the entire statement a person made that occurred between the statements of individuals immediately preceding and following the person’s expression.1

Two other students of interaction, in a work on attention quota or what they call “participation rates in small groups,” give the following statement:

The basic unit of participation labeled by the observers is the word, sentence, or longer statement of an individual that follows such a participation by one member and continues until it is terminated by an appreciable pause or by the participation of another member. In other words, an individual’s uninterrupted contribution is taken to be one participation.2

It is interesting to note that the acting profession employs a similar unit, called “a speech.” In Mencken’s definition, a speech is “A unit of an actor’s spoken part; it may be one word or a thousand.”3

Our commonsense view of linguistic communication, especially of the conversational kind, leads us to expect that when one message terminates, one of its recipients will take over the role of sender and convey another message. We expect a statement made by one person to be given an answer by another person. It is also to be expected that any particular message, whether statement or answer, will have two components, an expressive one and a linguistic one. Recipients will be concerned with what the sender says and also with the way in which he says it. However, when we examine conversational interaction closely, we find that a more complicated process frequently occurs.

First, we find that the recipient, in paying attention to the message, expresses the fact that he is doing so, by means of posture and facial expression. Also, as a larger and larger fraction of the message becomes transmitted, the recipient comes to be more and more in a position to know what the message will contain and what consequence it will have for him. The state of being in possession of this information seems to flow over, impulsively and spontaneously, into expressive behavior. In addition, the recipient seems to conduct an incipient internal monologue, conveying to himself or to anyone who is close enough or perceptive enough to hear, a rehearsal of how he is going to respond to the message when it is finally terminated, or (and this would seem to be more frequently the case) how he would like to respond to it were there no reason to exercise forbearance and restraint. We shall refer to this responsive expressive impulse and this self-communication as “take,” following a Hollywood usage which has precisely recognized this element in communication. Of course, wittingly or unwittingly, the recipient’s take is conveyed to the sender as a source of impression, giving the sender an opportunity of constantly checking up on the probably consequence of his message. When the sender has just terminated his message, the recipient perhaps reaches a culminative point in the judgments and considerations which he addresses to himself regarding the message. Once the recipient starts transmitting his considered reply to the previous sender, his take may diminish. Thus, if we modify our original decision of message to include the take along with the reply that emerges from it, we see that the expressive weighting in a message may decline in importance as the importance of the linguistic component increases.

In conversational interaction where messages are very brief and where participants feel they need not exercise much control over the expression of their responses, the take and reply in a message may become merged and may overlap considerably. In other kinds of communication situations, the two components—take and reply—may be distinctly separated. For example, we may explode when we receive a letter and an hour later answer it in a friendly and polite way. In cases where an individual overhears others talking and is not himself in a position to answer them, we may get from the eavesdropper a take and no reply.

In general, then, a take is expected to be genuinely expressive, although in fact it is often feigned. It is not officially directed to the sender.4 Sometimes it involves words of the class that are called expletives. A reply is expected to be more linguistic in nature. It is officially directed to the previous sender, not to oneself. This differentiation is also a temporal one; a take precedes (whether partly overlapping or not) a reply.

It is interesting to note that when a message is long, recipients frequently employ their take in a fairly open way as a signal of their attitude toward the message. Thus, during a lengthy political speech, cheers, hoots, and boos may be quite openly conveyed by recipients while they are still in the process of receiving the very message to which their take is a response.

On the island, adults in talking to other adults attempted on the whole to suppress any signs of their take to a message, except signs conveying the fact that they were attending to the message and were generously receptive to it. A frank take to the message seemed to be indulged in only when the recipient could have assurance that the sender would not see it. Among young people, who were presumably not yet obliged to treat each other with the delicacy required between adults, take was often not suppressed and was often given an important role in the communication process. Temper tantrums and sulking, while apparently relatively rare among island children, illustrate this. Further, on occasions when a person was being teased or was having his leg pulled, it was thought cooperative of him to evince as great a surprise take as possible, both in response to the lie he was being told and in response to being told that he had been told a lie. Among pre-adults, an explosive take in teasing situations seemed spontaneous and not put on; girls, especially, had the bait of attacking male teasers with their fists and feet out of desperation at not having any other means of response under control.5

It may be noted that in Dixon, as apparently elsewhere, unaddressed recipients sometimes neglect to prevent a frank expression of their valuation of the sender from appearing on their faces; feeling in observed, they take the opportunity of spontaneously expressing for their own private consumption what they really feel. Persons in Dixon sometimes suddenly turn on their unaddressed recipients in order to catch them for a moment in inadvertent sincerity.

When we examine an interplay we often find that the messages which occur within it are not evenly spaced out in time but occur, rather, in temporal clusters or groupings. Messages within one of these temporal groupings or spurts of communication are usually more closely related in content and culture than are messages which occur in different temporal groupings of the same interplay. Frequently the first message in one of these groupings presents a “statement” of some kind and the following messages in the grouping provide a reply, then a reply to the reply, and s on. A communication spurt of this kind may be called an interchange.6

An interchange may involve several persons but ordinarily it is restricted to two persons who alternately take the role of sender and addressed recipient while all the other participants in the interplay restrict themselves to the role of unaddressed recipient. The two persons may question or answer each other, or engage each other in parries and thrusts, while the unaddressed recipients merely watch. The two persons who are actively engaged in the interchange may not, of course, have equal sending rights. It should be added that in large formally-organized interplays of the actor-audience kind, the audience not only limits itself to conveying a few kinds of recognized messages but also comes to serve for the performer as a single addressed recipient. We thus tend to get interchanges between two actors, one of whom is the whole audience.7

After a particular interchange is completed, certain communication behaviors are possible: two other participants may provide the subsequent interchange of the interplay; the same participants may initiate a new interchange; one of the original participants may initiate a new interchange with a previously unaddressed recipient; or the interplay may be terminated.

It should be noted that a sender and his addressed recipient seem to accept greater obligation towards each other with respect to mutual responsibility, forbearance, and accommodation than do a sender and his unaddressed recipients. When there are no more than two recognized participants in an interplay, then, of course, the heightened responsibility between sender and addressed recipient necessarily applies to both participants.8

The unit of the interchanges has been considered so far chiefly in reference to its physical characteristics, namely, a rapid exchange of messages between two participants. In this chapter an attempt will be made to account for the nature of the unit by reference to two explanatory principles having to do, first, with communication as a ritual system, and, secondly, with communication as an informational system.

Explicitly or indirectly, any message involves, or may be taken by participants to involve, an evaluation or judgment of all persons who receive it. Hence the sender necessarily runs the risk of giving offense to the image that his recipients have of themselves and of him or to the image they have of things with which they feel identified. The rapid sequence of messages which follows after the initial message of an interchange either conveys to the sender an acknowledgment that his valuations are provisionally acceptable or modifies these valuations until provisional agreement is reached by all participants. The lull which follows an interchange is permissible because the working acceptance conveyed by the interchange ensures that a brief silence will not be taken as a sign that someone has been offended. A silence during an interchange usually conveys the fact that the recipients cannot frame a reply that is workable consistent with their own valuations and the valuations projected by the sender.

In Chapter VII it was suggested that a ritual model might well be fruitful in the study of interaction. Instead of employing a rationalistic bias, claiming that we perform our tasks strategically taking into consideration the probable response of others, we can employ a ritualistic bias, claiming that we interrupt our tasks in order to worship and placate the gods around us. Offerings and placations are, of course, a consequence of our taking into consideration the gods’ likely response to us, but this likelihood is not established by indications that the gods make concerning their future effect upon us but rather by the religious tenets and norms which guide our treatment of them and the idols which represent them. In human interaction, however, the idol which we are ritually careful of is also ritually careful of us. If we offer him up a prayer or perform a gesture of obeisance, he, unlike other kinds of idols, can answer us back, blessing us, or returning the compliment of worship. Thus, instead of a single act by which a devotee expresses his attitude toward a graven image, we get a double act, a statement and reply, for the graven image is in a position to respond to the offering that has been made to him.

The ritual model of the interchange is suggested by Park in his introduction to Doyle’s Etiquette of Race Relations in the South:

Etiquette is concerned primarily with personal relations. It grows up in the first instance, perhaps, as the spontaneous expression of one person in the presence of another, of a sentiment of deference. Under ordinarily circumstances such an attitude of propitiation of one individual implies and is likely to evoke a corresponding expression of benevolent recognition on the part of that other. Expression and response rather than stimulus and response are the natural termini of every instance of social interaction.9

By employing this model we can see that an interplay is not a continuous flow of communication; it proceeds by discontinuous jerks or steps, an interchange at a time.

Perhaps most interchanges—as Park implied—are limited to only two messages, expression and response, as he called them. An actor usually has a fairly clear understanding of the expectations of his recipients and is sensitive for many reasons to the rule that requires persons to be treated with tactful concern for their attributes or sacred qualities. We may therefore expect him to restrict his messages, in most cases, at least, to ones to which it is possible for a recipient to express acceptance. The recipient, in turn, must express something or else become responsible for the ambiguities conveyed by silence. In this way we can understand the prevalence of two-message interchanges.

We have attempted to account for the character of interchanges by reference to the fact that persons are ritually delicate objects which must be treated with care, with ceremonial offerings and propitiations. Working acceptance often marks the termination of an interchange, and a working acceptance is required in order to keep in control the eddies of insult and offense, of reprisals and counter-reprisals that persons can involve themselves in. A second explanatory principle for the nature of the interchange may be suggested now.

No matter what it is that a sender wishes to communicate, it would seem that his object is to communicate successfully; he wants the recipient to receive the message and to receive it correctly. If a recipient replies to the message he has received and gives some form of answer, then the original sender can use this reply as a test of whether or not his original message has been correctly received. Whether the reply is one of agreement or disagreement may be of secondary importance, as long as it is a meaningful reply, for if the reply is meaningfully related to the original message, then the original sender can be sure that the line of communication between himself and the other is in effective operation. The working acceptance that is achieved by the time the interchange terminates—whether the achievement of this measure of consensus has required only one exchange of messages or a long series of exchanges—may serve chiefly to signify that the participants understand one another, not that they agree with what they understand. And, in fact, many working acceptances seem to be limited to a meager consensus of this kind.

We have attempted to account for the fact that an interplay proceeds by steps, an interchange at a time, by reference to two factors, a ritual factor and an informational factor. Taken together, these two factors seem to supply a partial explanation for why an interchange takes the form that it does. The aim of this study, however, is to describe, not to explain, and in the next and following chapters many varieties of interchanges will be illustrated. Before proceeding to this, it will be convenient to raise four questions about the nature of interchanges.

A message has been defined as the sign behavior of an individual during the continuous period when he is the focus of attention. However, sometimes during a long message a sender will pause, obtain a momentary take of approval from his recipients, and then launch into what appears to be a different message. Interestingly enough, both Steinzer, and Stephen and Mishler, qualify their definition of a message in this way.

If the person stopped talking for five or more seconds, then continued, the statement was counted as two units.10

However, if there was a clear change of content during the course of a lengthy contribution, it was taken to be the beginning of a new unit of participation.11

One way to account for these double messages is to say that they are in fact separated by a fleeting message on the part of the recipient. Another explanation is possible. Often a particular sender contributes the last message in an interchange and then happens to be the sender who contributes the first message in the next interchange. We would then expect that there would be little if any connection between the two parts of the sender’s message and that at the same time there would be little if any interruption of them on the part of others. Thus we have an analytical explanation of what originally appeared to be an awkward qualification that commonsense observation forced upon the definition of a message.

Secondly, it is apparent that while many interplays are wholly made up of clearly articulated interchanges, this is not always the case. For example, when men in Dixon gather to talk about their exploits as seamen in distant ports, one participant will tell his tale and the moment he has finished (or a moment before he has quite finished) another participant will tell his, and then another participant, and so on. In these conversations, one message will cling to the topic established by the previous one, but the statement-and-answer character of communication will be muted. In a certain sense it would be more realistic to say that each participant was merely waiting out his turn to take the floor and obtain a share of the attention indulgence. Thus, while this study is limited to a consideration of the interchange as a concrete unit, it should be understood that there are other natural units in interplay as well.

Thirdly, in making a distinction between interplay and interchange, it must be admitted that it is possible for an interplay to contain only one interchange and hence be coterminous with it. Thus, there are many kinds of interchange taking the form of courtesies that are coterminous with the interplay that incorporates them. This fact may be placed alongside the fact that a social occasion may be coterminous with the interplay it incorporates.

Finally, as suggested in a previous chapter, an entire interplay may function as one message in a prolonged series or exchange of messages. An interplay may take a form other than the rapid exchange of concrete messages. The overall treatment of an individual during an interchange, or interplay, or social occasion may function as a single message in an extended exchange of messages. The individual may respond by appropriately adjusted behavior during the next interchange, or interplay, or social occasion, and this may in turn give rise to retaliatory or compensatory behavior on the part of others at the next meeting. Similarly, the tendency of one set of persons to come together and form a conversational cluster may be taken as a message and responded to by other persons forming their own cluster or by other persons attempting to disrupt or expand the one that is formed. Exchanges of messages of this kind may take the form of interchanges, the moves or messages of which are themselves complex interactive systems. These protracted or higher-order interchanges often rely partly upon messages which are solely expressive, but this need not be the case. Thus, a linguistic message on the part of one participant may carry implications for a particular recipient and not be responded to or answered by him until several interchanges have occurred or even until another social occasion. These protracted interchanges are often less neatly brought to a conclusion than are the simple concrete ones we have been discussing but they none the less provide an important area of study. For practical reasons, however, the present study is only concerned with the interchange of concrete messages.


Chapter twelve of Erving Goffman. Communication Conduct in an Island Community. PhD diss., University of Chicago, 1953.

No comments here
Why not start the discussion?