GROUPS AS SYSTEMS

Any set of communicating individuals could be characterized simply in terms of the properties they have in common. We could study the speech of people in a geographic area in terms of a set of individuals, and analyze their similarities and differences in a way resembling standard dialect surveys. Here it would be of interest to know which dialect features, described in terms of properties of the individuals, were common throughout the region, which features varied, and how they varied geographically, socially, and along other dimensions. Historical change could also be treated in this manner. There would be certain advantages in doing this, for our descriptions would be anchored in the people themselves and their similarities and differences, rather than in language. It has been difficult to treat variation using the concepts of language and grammar, which were developed for describing an ideal of perfection, and thus tended to imply uniformity when applied to a community.

But treating a community simply in terms of the distributions of common properties of the individuals would not be sufficiently revealing. A community is more than a set of individuals; it embodies structure reflecting the communicative interactions between the individuals in the community. Taking this structure into account would seem to be necessary if we are to understand more clearly how people communicate. Human linguistics allows us to take this structure into account. We can analyze the structures of groups and communities in terms of systems by means of linkage theory.

Linkages are characterized by properties, as are individuals, but linkage properties and individual properties are different, being related to different sorts of physical objects—to groups of people rather than to single persons. However linkages and individuals, though different, are similarly constituted and structured: They are both analyzed in terms of conditional, categorial, procedural, and foundational properties, and the same general laws hold for linkages as for individuals—the law of componential partitioning, the law of small changes, and the law of restricted causation. This allows a certain symmetry in the theory.

Linkage properties are understood as properties of the linkage as a whole. If two people are in conversation, their joint activity can be represented in terms of linkage properties. If they are talking past each other because there is a mutual misunderstanding, the misunderstanding would be represented in terms of linkage properties, for no person can talk past each other by himself. When the members of a group have agreed on a topic of conversation, the agreement on the topic and the topic itself would be represented in terms of linkage conditional properties, and the coming to agreement would be represented in terms of the executing of linkage procedures. When Jill asks Jack to get an apple for her, the joint communicative activity leading up to Jack getting the apple would be represented as the executing of linkage procedures.

A dynamic state system is appropriate for the analysis of groups because it allows the modeling of such changing group properties as agreeing, disagreeing, bargaining, negotiating a topic of conversation, considering an issue, reaching a consensus, and the like. The standard case in human linguistics is a linkage in which communicating individuals that are similar in some of their properties but different in other properties are communicating with each other.

Recall that a linkage as a system is composed of any number of constituents. These may be communicating individuals as participants, a channel or channels to carry the energy of the sound waves and any other form of energy involved, and possibly props and settings to represent the communicatively relevant aspects of objects and the physical environment.A linkage, then, can be characterized as large or small in terms of the number of constituents it is made up of. It may contain any number of constituents, from only a few to many millions.

Since a linkage represents an object of aggregation that can be set up and dissolved, it can also be delimited in terms of its temporal boundaries. Thus a linkage can also be characterized as brief or long lasting. If two people meet on the street and greet each other, we could analyze what happens in terms of a brief linkage. A linkage covering the extent of a friendship would be longer lasting.

Linkages are also characterized as broad or narrow. A broad linkage involves many properties, as would be the case in a close-knit family. A narrow linkage involves comparatively few properties, as between a housewife and the grocery-store clerk. People in a broad linkage would be said to be well acquainted, having much in common; people in a narrow linkage would be casual acquaintances.

In addition, we recognize analyses in terms of either focused or complete linkages. A focused linkage analysis of a family might consider only one facet of their communicative behavior, such as what names or nicknames they use for each other. A complete linkage analysis would be concerned with the total communicative behavior in the family.

THE RELATION OF INDIVIDUAL AND GROUP

A communicating individual has been defined as an abstraction of a person that includes just those properties required to understand how he communicates, and a participant has been defined as an abstraction of a communicating individual that includes just those properties required to understand how he communicates in a particular linkage. The properties of a participant, therefore, form a subset of the properties of the communicating individual, just as the properties of the communicating individual form a subset of the properties of the person. The properties of both the communicating individual and the participant are set up and justified on the basis of observed similarities and differences between this person and others and between this person and himself at different times. The properties of a participant, therefore, are relative to a person. Except for their selection, they are not relative to a group. They answer to what the person is doing in his own terms when he is participating in the group.

In terms of the group, however, different people may do the same thing in different ways. For example, a husband and wife in a clothing store may have selected a particular dress from several they like. Either the husband or the wife might go to the clerk and close the sale by saying, “We’ll take the blue one.” These would be equivalent, for in either case the clerk would write up the order for the selected dress, and in this sense the husband and the wife would be doing the same thing. But their participant properties might be quite different. The husband may use “blue” as selecting one from fifteen or twenty colors, whereas the wife may use “blue” as selecting one from twenty or thirty colors. Everybody is different. It would be unusual if two persons always did the same things in exactly the same way.

So we also want to be able to say that although they are doing it differently, still they are doing the same thing from the point of view of the group. For this we use the concept of a role part. A role part is a functional part of a linkage, and its properties are relative to the linkage that it is a part of, rather than to a communicating individual. Thus role-part properties are set up and justified on the basis of observed similarities and differences between different groups and between the same group and itself at different times, rather than between different persons and between the same person and himself at different times as is the case with participants. The relation of participant to role-part is that of form to function. A participant, as a constituent or formal part of a linkage, fills a role part as a functional part of the linkage.

In the example, the linkages are similar in that each has a customer as a role part, but they are different in that one of the linkages has the husband and the other the wife as a participant. In terms of role-part properties, we are able to say that the husband and wife are doing the same thing when they say to the clerk, “We’ll take the blue one,” for it fills the same function by making the same selection in a customer-clerk linkage—closing the sale—and thus it affects the linkage properties in the same way.

Since the constituents of a linkage include not only participants but channels, props, and settings as well, it is appropriate to define other functional parts of a linkage. These are called channel parts, prop parts, and setting parts. The obvious analogies hold.

A linkage is more than the sum of its parts. The organization or arrangement of the parts is also important. As an example, consider a linkage between a concertgoer and a ticket taker as participants, the concert-hall entrance as a setting, and the ticket as a prop. The communicative interaction when the concertgoer presents his ticket and enters the concert hall involves the physical arrangement of both participants with respect to the setting, and the prop with respect to the two participants. As another example, consider a teacher standing in front of a class as one arrangement. Then if a student is to give a report, the student may take the place of the teacher in front of the class in a different arrangement. The communicative phenomena taking place involve not only the teacher and students as participants, but also their arrangement in the classroom as a setting.

We see, then, that the properties of the various functional parts of a linkage—the role parts, channel parts, prop parts, and setting parts—are all set up on the basis of the same evidence used to set up linkage properties, namely, observed similarities and differences of different groups and of the same group at different times. The evidence is just interpreted differently—in terms of linkage properties in one case and in terms of the properties of the functional parts of the linkage and their arrangement in the other case.

We can conclude that a linkage can be described alternatively in terms of linkage properties or in terms of the properties of its role parts and other functional parts, and their arrangement. The two descriptions are of the same thing, a communicatively coherent group, but they are at different levels of analysis.

GENERALIZING OVER GROUPS

Although we suspect that every linkage is different from every other linkage, there are often similarities between linkages that go beyond single linkage properties. Thus there is a possibility for further generalizations.

Suppose we are studying the members of a family from the point of view of what names or nicknames they call each other. We could set up brief focused linkages dealing with situations of direct address between family members. Each instance of one family member directly addressing another would count as a separate brief focused linkage. If the focused linkages were sufficiently limited in scope, there would be cases where a number of focused linkages were identical in their categorial, procedural, and foundational properties, since the differences between the corresponding complete linkages would have been removed from consideration when we focused narrowly on the phenomena of interest. For example, most of the linkages where a member of the family is addressing the youngest son might be identical in involving him being called “Tommy,” while the other aspects that show differences would have been set aside.

In cases like this, where two or more focused linkages are identical in their categorial, procedural, and foundational properties, these properties define a linkage type. A linkage type is not a linkage, since it does not include the conditional properties, which, being changeable, make any linkage different at different times, nor does it include specified temporal boundaries. It does, however, characterize each of the similar focused linkages in terms of the static properties they hold in common, and these focused linkages will all operate the same dynamically. Thus a linkage type is a generalization over focused linkages that have identical static properties.

For each linkage type we can also define appropriate role types, channel types, prop types, and setting types. These are defined in an analogous fashion in terms of the identical categorial, procedural, and foundational properties of the role parts, channel parts, prop parts, and setting parts that are the constituents of the focused linkages of the linkage type. For example, one of the role types would be an appropriate generalization over the role parts representing any family member addressing the youngest son as “Tommy.”

INTERACTIONS BETWEEN GROUPS

When studying several people interacting with each other communicatively, it may be worthwhile to consider the phenomena not in terms of one overall linkage, but instead in terms of two or more linkages interacting with each other.

Linkages that interact with each other are said to be coupled. When the coupling between two linkages does not involve any third linkage, we say that the two linkages are directly coupled. Thus a focused linkage with a man and his employer as participants might be directly coupled to a focused linkage with the man and his wife as participants, even though the employer and the wife may never have met. For example, the employer may suggest to the man that he invite his wife to an office party that is being planned.

In this example the focused linkages are coupled through an individual. The man is a participant in both linkages—he is the employee in one linkage and the husband in the other. And since the linkages are coupled, these two participants in different focused linkages must have some properties in common that account for the coupling. The properties that they have in common approximate what is called the contact in the individual between the two linkages. We can say that the two linkages are coupled through their contact in the man. Although the properties in the contact between these two participants in different linkages are identical, being subsets of the properties of the same individual, it must be remembered that the properties of the corresponding role parts in the two linkages will be different, being structured relative to different linkages.

We can distinguish several structural types of direct coupling between linkages.

One type of direct coupling is the messenger type. For example, a skilled negotiator may conduct sensitive negotiations between two countries by acting as a messenger. The messenger moves back and forth between two linkages, being a participant in each, with the linkages being coupled through contact in these two participants in the same individual.

Another type of direct coupling between linkages is the dinner-table type. Consider the case of two friends of long standing who are dining with each other and their spouses. There would be two long-lasting husband-wife linkages and a linkage for the two friends. There would also be a brief linkage of the four of them around the dinner table. What goes on communicatively might best be described in terms of these four linkages and the interactions between them, which are of the directly coupled type through contact in the various participants in the several linkages.

An interesting type of direct coupling is the linkage-creating type. When a person brings one of his friends up and introduces her to another of his friends, and then excuses himself, leaving the two to talk, a linkage has been created between the person’s two friends. Here the new linkage is created by contact in the participants in the two initial linkages and in the new linkage across its initial boundary.

A frequent type of direct coupling is the broadcast type of coupling between linkages. A lecturer is a participant in a separate linkage with each member of the audience. Each of these linkages is directly coupled to all the others through contact in the lecturer. There may, of course, be other linkages between different members of the audience, but not necessarily. The result of the contact is commonality of properties in the members of the audience, all having heard the same lecture.

Linkages can also be directly coupled through contact in props, channels, and settings. For example, a family expecting guests in the evening may leave the porch light on to mark the house for the guests. As the guests come up the street looking for the house, the linkage of family and porch light would be coupled to the linkage of porch light and guests through contact in the porch light as a prop.

Writing can be analyzed as a prop. The writer would be a participant and what he is writing would be a prop in one linkage, and the reader would be a participant and the writing a prop in another linkage, with contact between the two linkages in the props. The properties of the props in the contact would be the same in each prop if the reader reads the original, but the prop parts in these two linkages would be different, being relative to different linkages. In this way we can account for misinterpretations on the part of the reader. We can handle in a similar manner other signs, such as cairns of stones, tracks in the snow, and the phenomena of nature.

Alternatively, writing can be analyzed as a channel—a constituent of a linkage standing between the writer and the reader as participants. Here there would be delay in the channel, and the reader may not even be contemporaneous with the writer. The telephone and other electronic means of communication are important in modern culture and do not normally involve much delay. They can easily be analyzed as channels.

Another way in which linkages can be directly coupled is through contact in an arrangement. In a doctor’s office with several examining rooms, the linkage between the patients and the nurse is coupled to the linkage between the patients and the doctor through an arrangement of the participants when the nurse ushers the patients into the examining rooms in a certain order so that the doctor will know who is next. Furthermore, she may place each patient’s chart on the desk in the examining room so that the doctor will know which patient is waiting in the examining room by the arrangement of this prop.

THE STRUCTURE OF COMMUNITIES

When linkages interact only through one or more other linkages, they are said to be indirectly coupled through a chain of directly coupled linkages. The employer’s wife may ask her husband to ask several employees to ask their wives to a party that she is planning. Here the first husband-wife linkage is indirectly coupled through the employer-employee linkage to the second husband-wife linkage.

Chains of directly coupled linkages can also provide the indirect coupling characteristic of the spread of rumors. A rumor can change in two ways. It can change in each linkage because the properties of the participants are relative to different individuals, who may have different understandings. And it can change in each individual in the contact between linkages, for the properties of the role parts are relative to different linkages, allowing for the possibility of paraphrase rather than verbatim repetition.

One way, then, in which the structure of a community can be understood is in terms of the direct and indirect coupling of linkages. Each individual is a participant in many linkages. Some of these will be directly coupled to others, and some will be indirectly coupled to still others. An individual may participate in linkages involving other members of his family, neighbors, friends, and acquaintances. He may participate in linkages with sales and service people, people in his school or place of employment, in government and political organizations, and with other members of a religious organization, lodge, union, or club. He may also participate in linkages with newscasters and public figures through radio and television, and with writers and reporters through his reading of newspapers, magazines, and books. In short, for most people there is a very large circle of others that he participates with in linkages. And these linkages are potentially directly and indirectly coupled to still other linkages in the community. The individual may be linked in this way through word of mouth, and especially through the media, to virtually everyone else in the community.

The picture is that of a complex network with individuals at the nodes. Each individual participates in many linkages with many other individuals. Some of these linkages are directly coupled to each other through contact in his being a participant in each. Some are directly coupled through other individuals to other linkages, and some are indirectly coupled to still other linkages reaching to still other individuals in the community. This network would extend throughout the community, with probably no individual or part of the community completely isolated communicatively from the rest. For some sorts of communicative phenomena, this picture of a community may be especially revealing, but for other phenomena its extreme complexity would be a disadvantage.

At the other extreme one could treat the community simply as a single very large linkage covering the whole community. Some structure could be handled through concepts of the arrangement of the participants, and through statements about the arrangement of properties and changes in the arrangement of properties as news and information flow through the community. This would eliminate the extreme complexity of the network analysis, but at the expense of not allowing for the handling of important elements of structure.

Fortunately there is another possibility. An analysis is possible in which people can communicate with groups, and groups can communicate with other groups. For example, a person may communicate with an organization such as a unit of government, a financial institution, or a utility company, and the unit of government may communicate with the utility company. This possibility is completely implied by the theory as it stands, because linkages as well as individuals are defined in terms of conditional, categorial, procedural, and foundational properties.

A linkage, then, can have among its constituents not only participants, channels, props, and settings but also other linkages. We simply have to introduce appropriate terminology and work out the details. A linkage which has one or more linkages among its constituents will be called a compound linkage. This is distinguished from a simple linkage, which has no linkages among its constituents. When a linkage participates in a compound linkage, it will be called a compound participant in the compound linkage. The functional part that it plays in this compound linkage is called a compound part.

When a person communicates with a group, we can analyze what he is doing at several levels. For example, at one level there may be a simple linkage having the person and a department-store clerk as participants. At a higher level there would be a compound linkage having the person as a participant and the toy department, which is a simple linkage, as a compound participant. And at a still higher level there would be a larger compound linkage having the person as a participant and the department store, which is a compound linkage, as a larger compound participant. The several analyses will be systematically related and useful for different purposes.

And when two groups communicate with each other, as a city government and a public utility company, this can also be analyzed on several levels down to the two individuals talking to each other on the telephone.

A community can be analyzed from this point of view as a complex multiple hierarchy of linkages and compound linkages. An individual is seen as participating in very many of these linkages and compound linkages, which cover all of the significant structurings of the community. This view does not have the tangled complexity of the linkage network view, or the extreme oversimplification of the covering linkage view. It models realistically the ways in which the community is actually structured communicatively, and thus allows us to develop a fuller understanding of how people communicate. This view of the relation of individual and community in human linguistics is thus quite different from the concept of an ideal speaker-listener in a completely homogeneous speech-community, which will be familiar to some readers.

On the basis of the present discussion it should be clear that a human linguistics is indeed possible, and that the second alternative is indeed a feasible option. We are now ready to move on and test some further implications of the theory against observational evidence, and in so doing to compare the prospects of human linguistics with the past performance of business as usual.