There is no need to review here the history of "psycholinguistics" but only to point out, that it is only in this decade, that psycholinguists have begun the study of natural talk. We were trained as graduate students in the North American intellectual climate of a pragmatized structuralism and functionalism. BYG was indoctrinated in the descriptive and applied linguistics of Bloomfield and Fries, and LAJ was trained in the psychology of neurophysiological behaviorism (Lambert, Hebb, Osgood). The psycholinguistics of the 1950's was preoccupied with conditioning and semiotics: the acquisition of sign-function and its semantic features. The word was the methodological unit of inquiry as shown by the topics of the investigations of that period: the effects of word-frequency; similarity and synonymy; wordassociation clustering effects; verbal learning of pained assocites; tachistoscopic perception of words and their emotionality value; Atlases of semantic differentiation; stimulus generalization; phonetic articulation; semantic satiation; and many others where the word was the unit for investigating the psycholinguistic laws of language behavior (see Jakobovits & Miron, 1967, for representative articles).

However, two separate developments in the 1960's helped clarify the idea that the laws of language behavior are to be found beyond the unit of the word. One development was the spread of the Ethnomethodology School, and the second, the importation of ideas from the British School. Both introduced a new paradigm which allowed a methodology of natural talk.

In our own history, we encountered the ethnomethodological school through the work of Goffman, Garfinkel, and Sacks (see References). From their writings we acquired conceptual tools for studying language behavior as a sociopsychological phenomenon; that is, that language behavior in natural situations was a spontaneous, reactive phenomenon. Discourse production was to be seen as a natural biological phenomenon, hence, resposive to environmental effects. Sociologists Goffman, Garfinkel, and Sacks took social organization on the daily round as the basis for defining the functional units of language behavior. As sociologists, they viewed the analysis of natural talk as a task in uncovering the effective social stimuli that made possible the successful accomplishment of ordinary transactional routines. Discourse production thus became a medium for transactional exchanges. The unit of analysis thus becomes the transaction; not the word, but an 'exchange of words'!!

The second development that shifted our theoretical position from the word to the exchange, was the work of the British Ordinary Language School (see Steinberg & Jakobovits, 1971, for representative articles). Here we acquired the idea that discourse and talk operate through ethnosemantic conventions
--the meaning of the message is carried along with the un-verbalized support of presuppositions and implications. Thus, the unit of language behavior must relate to the social context of the situation. There exist operational rituals for transacting topical exchanges, and these commonly held operations provide a motivational direction to talk.

In summary, then, we started with the word as the unit, and we moved beyond it, to the transactional exchange. We would like to discuss some theoretical and practical issues involving this new focus for psycholinguistics. The accompanying table presents the focal issues in our own theoretical development.

TABLE 1
Focal Issues in Our Theoretical Evolution

Shifting focus from the word to the exchange is accompanied by the realization that all discourse is interactional discourse, i.e., discourse is produced by more than one individual. This is of course most obvious in the study of conversation where it appears that participants take turns at talk: it is clear that the discourse visible in a transcript is interactionally produced. But this is equally true in both writing and in interior dialog where there appears to be only one person producing the discourse. However, it is a matter of common observation that writers change their discourse in response to the intended or imagined audience, showing that it too is a form of interactionally produced discourse. In self-talk (or interior dialog), the person acts as if there is an audience: reports of interior dialog produce transcript-like segments in which the person addresses himself or herself using the pronouns [I, you, we] along with the appropriate verb form, thus indicating that the discourse produced in self-talk is also a derivative form of interactional discourse.

The analysis of interactional discourse hinges on the recognition that discourse production is a spontaneous reactive phenomenon. By analyzing the organizational structure of interactional discourse one in effect investigates the structure and operation of a social psychological phenomenon. The laws of social interaction are uncovered through a close analysis of the setting in which the discourse is a by-product. Discourse is thus seen as a medium within which interactions are transacted. This presupposes the notion that a discourse intervention by a participant counts as a move. The functional significance of discourse derives therefore from its significance as a transactional move. Transactional moves are organized by pre-established rituals of talk. These procedural operations are acquired as part of one's ordinary social competence on the daily round. We would like to explore some theoretical issues that arise from this perspective.

II. Theoretical Issues

The primary theoretical issue that arises as one moves beyond the word to the social exchange, is the explanation of connectedness at two separate levels of operation. The idea of 'having an exchange of words' implies the minimal diadic arrangement, and we follow the ethnomethodologists in the technique of dividing conversational exchanges into turns at talk, or talking turns. The talking turn (TT) is one level of organization for interactional discourse, i.e., discourse produced by more than one talker.

Another level of organization for the connectedness of natural discourse is what we would call the "within turn" organization, i.e., the discourse produced by a single individual during his turn at talk. The following diagram depicts this description and points to some implications.

CONNECTEDNESS IN INTERACTIONAL DISCOURSE
(= NATURAL TALK)

(DYNAMIC LEVELS)

We shall explore here only the theoretical implications of between-turn connectedness and present the direction of our current work dealing with the social psychology of language behavior or, "sociopsycholinguistics."

III. Some Syntactic Properties of Conversational Interaction

A first practical issue to be resolved about the common phenomenon of conversation is represented by the question, "What's going on in the conversation?" We follow here the ethnomethodological dictum that the answer to this question must exclude anything which cannot be demonstrated to be a feature to which participants are oriented. This stricture insures that the theoretical explanation objectively matches the actual units that govern the organization of verbal exchanges. By "actual units" we mean to designate the features of the social exchange which conversationalists are oriented to notice by virtue of their common socialization training. In other words, the phenomenon of conversation is viewed as a managed exchange
--managed by the participants according to shared rituals of operation. The question then arises as to how the interactional exchange is successfully managed by the participants; more specifically, what mechanisms are there for regulating the sequence of turns in a conversation? The mechanisms to be described must be mechanisms that are actually used, and these are perforce dependent on the noticings of the participants--when to talk, when to say what, when to acknowledge, disagree, change topics, and so on.

To proceed with this task, then, we begin by defining the minimal unit of exchange as a situated episode. An "episode" is a sociopsychological concept. It derives from Goffman's work on the nature of ritual behavior on the "daily round." All social behavior is situated in time and place. A "setting'' is defined as a time/place specification for routine activities in a community. For example, our students in social psychology are given the exercise of recording, minute by minute, the course of a day. Here is a sample:

With this technique, which we call logging activities, one can arrive at a local ethnography of community settings. It is, in other words, a daily round map that empirically specifies the available settings in a community. With such a map as a reference point, episodes may now be investigated as a function of the setting within which the exchange occurs. The following diagram depicts these relationships:

SITUATED EPISODES

We shall present a transcript segment, prepared by a student, and illustrate some techniques that are possible for investigating the functional components of situated episodes (Winskowski, 1977).

Transcript segment: A and B are friends in their early twenties. B is A's boyfriend and has come to pick up A at her house. As the doorbell rings, A opens the door, holding a taperecorder in her hands.

1. B: Hi. [opening front door.]

2. B: What's up? [gesturing to the tape recorder.]

3. A: I'm taperecording you.

4. B: Are you kidding me?

5. A: Nope.

6. B: But what am I supposed to say?

7. A: Whatever you want. [walking in the kitchen.]

8. B. Well, what a nice bunch of groceries you've got. [said in the exaggerated tone of Little Red riding Hood, exaggerated tone, seeing two shopping bags on the table]

9. A: How 'bout that.

10. B: That's very nice.

11. A: Amusing, eh?

12. B: Uh, huh.

13. A: I got most of my gear together except I gotta get something of. . . .

14. C: Randy, you're on time. What's wrong with you? [C is A's mother who just walked into the kitchen]

15. B. No, I'm not. I'm fifteen minutes late.

16. A: You're fifteen minutes late, you know. Did you know that.

17. D: Hi, Randy. [D is A's father who walked in with C in (14)]

18. B: Hi.

etc. etc.

The accompanying table presents a first-order analysis of the structural and functional components of this transcript segment.

Note that we've arranged the interactional discourse recorded on the tape in terms of six parts which we call bracketed segments of the conversation. The first segment comprises the structural units of talking turns 1 and 2, which occurred at the front door. This setting is familiar to the participants: both share common ritual sequences known as "greeting"--a ritual that occurs when two acquainted participants suddenly find themselves in face-to-face contact. In other words, TT's ("talking turns") 1 and 2 are structurally localized (episodal openers) and functionally standardized (greeting). To demonstrate the-functional properties of this syntactic 'slot' one need only consider alternative forms that might occur there without change of function: e.g.,

The set of alternatives a is a form class whose items distribute themselves over variable topical dimensions, but remain functionally equivalent, being each a greeting response, thanks to their structural location as episodal openers. This observation leads us to a fundamental principle of interactional discourse production, namely, that transactional function and topical content are independent. For example, TT (11), has a form class aa, as follows:

The sum of the sets of alternatives available in a talking turn, i.e.,

E[a, aa, aaa ... ] TT may be called the display repertoire of the verbalizing community. Through socialization, training, and experience, a conversationalist acquires a certain portion of the culturally available display repertoire. Individual variations exist in available responses to a talking turn. Skilled conversationalists have available a greater range of the verbalizing community's display repertoire than those who are less skilled. We may speak here of social or transactional competence. One strategy in becoming a better 'transactional engineer' would presumably be to (i) catalogue the set of alternatives available in particular conversational slots. and (ii) learn them. Step (i) amounts to making a local ethnography of social settings by some natural history technique (as in 'logging activities", discussed above). Step (ii) amounts to becoming acculturated, i.e. learning standardized patterns of topical interventions. The latter may he recognized as the perennial problem of the young, the visitors, and the foreigners: "What shall I say, When, and How?!?" Assimilation and re-educational training can be viewed as attempts to enlarge an individual's set of alternatives in interactional discourse--but enlarge in a particular direction, namely, the direction of greater overlap between the individual's current performances and the target 'norm'. This element is not fixed, but operates within a 'range of normalcy' (Goffman's term). Episodal exchanges are progressive, i.e. they wind down from 'openings' to 'closings' (Sacks; Shegloff). At any point within this sequence a participant may find himself "at a loss for words." This is a "normal" occurrence on the daily round. Yet because episodes perforce wind down, there must be mechanisms for re-starting so that the closing exchange may ultimately occur. This mechanism also operates when one is 'at a loss for words', at which time one re-starts, viz., begins a new bracketed section of the conversation.

We can summarize the above considerations by stating the following empirical hypotheses about the character of conversational interaction:

Hypothesis 1: Verbalized utterances in a talking turn are treated by participants as particularized surface variations of a form class the items of which have the same transactional significance or function.

Hypothesis 2: The first pair of talking turns, i.e. the first "adjacency pair" (Sacks' term), in a conversational episode is treated by participants as a greeting opener.

Hypothesis 3: The transactional or functional significance of a talking turn is jointly recognized by participants in accordance with a transactional code which they share by virtue of their common membership in a verbalizing community.

We may state a fourth hypothesis upon considering TT (2) in which participant B says "What's up?" while gesturing towards the taperecorder held up by A upon opening the front door:

Hypothesis 4: Any publicly noticeable change in the "normalcy" status of the environment (Goffman's term), is a routinely available candidate for being made the topic of an utterance in a talking turn.

When we consider that the utterances in one talking turn have a functional relation to utterances in adjacent turns (before and after), we are led to a notion, discussed by Sacks and others, known as "the setting-up move."

Hypothesis 5: There is a class of utterances, known as general purpose inquiries, that when uttered in a talking turn, will be treated by participants as serving to set-up the immediately next alternating talking turn such that it will contain a move that will constitute a justificatory comment on a readily noticeable environmental event.

Hypothesis 6: If a talking turn is made up of an utterance of the class belonging to a general purpose inquiry, it will be seen by participants to serve as a setting-up move for the utterance in the immediately next alternating talking turn such that it, in turn, will be seen as a directed response, a supportive move, a reply, a remedy.

The fifth and sixth hypotheses are general formulations that have the merit of showing up the common structural basis of a large class of conversational events that relate to the sequential aspects of alternating talking turns. Next, when we consider talking turns (3) - (12), we note that they constitute a bracketed segment of the episodal exchange. This leads us to the formulation of the next hypothesis.

Hypothesis 7: There is a class of conversational events in the form of an exchange of alternating talking turns that is bracketed from other parts of the conversation, and where the bracketed exchange is seen by participants as specifically different from adjoining conversational material, this difference being that it is to be seen as semi-serious or "playtalk" in contradistinction to the rest which is seen as serious.

The seven hypotheses outlined above imply the existence of three types of functional mechanisms in conversational interaction. First, we may mention the mechanism of sequencing devices. Given that talk proceeds through time, the events that take place in it must be ordered in some way, and a description of these ordering procedures is what's called for in the elaboration of conversational sequencing devices. Second, we may mention the mechanism of boundary limits in bracketed sections of conversation. The function of boundary limits is to indicate to participants where some event begins and where it ends. We've reviewed some of the elements of boundary limits, namely, talking turns, adjacency--pairs, opening sections, closing sections, and playtalk sections. Third, we may mention the mechanism of transactional moves. The function of transactional moves is to indicate to participants the significance of a conversational display (utterance or gesture) for their relationship, i.e. for their behavioral or interactional implications. Thus, sequencing devices, boundary limits, and transactional moves are theoretical mechanisms that are available to participants for ordering the sequence of talking turn utterances within a conversational episode. The elaborations of these mechanisms in Hypotheses 1 through 7, stated above, attempt to show that conversational displays (in gestures or in utterances) are treatable as transactional moves whose significance derives from their structural properties, that is to say, their locus of occurrence in the conversation. By "locus" we mean such things as sequence of talking turns, boundary limits of the bracketed section they belong to, place of that section within the overall episode, and type of relationship of participants as implied by previous history of joint conversational episodes.

We have reached here a crucial stage in our theory building. Since the list of empirical hypotheses (as proposed above) is potentially open-ended or indefinitely large, we need an explanatory mechanism that accounts for the occurrence of an indefinitely large number of transactional sequences, as one observes regularly in the continuing round of episodes on the daily schedule. In other words, the listing of empirical hypotheses about natural talk is a descriptive stage of data processing (taxonomy?). We now need a process-valued function that sets talking matters in motion, and directs movement towards an objectively identifiable goal. We shall refer to this post-taxonomic phase as the psychodynamics of talk.

Goffman's elaborated notions on "face work" serves as the starting point for our proposal relating to the motivational dynamics of talk. The dialectic of offense and remedy is posited as the generating mechanism. Utterances and gestures are displays, or display presentations, or performances. Displays are organized as moves in transactional sequences. Moves have direct 'face work' implications, i.e. moves are indices to a person's transactional reputation. Each utterance or gesture displayed within a talking turn carries a transactional function. The transactional value of a talking intervention is either positive or negative. when positive, the move counts as a remedy; when negative, it counts as an offense.

The dialectic of remedy and offense Provides us with the theoretical starting point for evolving an explanatory account for the connectedness between talking turns in a conversational episode. We have pointed out at the outset of this section that the answer to, "What's going on in this conversation?" must allow only elements which pertain to what the participants themselves orient to, as the episode winds down to a closing. We need therefore a notation system for recording the occurrences of noticings during talk. Let us call this type of record a relationship history.

It is intuitively valid that the transactants to an episode have a coding system for keeping track of the episode's evolution. Thus, topics that occur earlier are pre-supposed in subsequent bracketed sections of the conversation. As well, there is left an impression of the quality of the transactional face work--whether pleasant, friendly, involving, or their opposites. And finally, there is a sense of the episode's context, in time, place, and schedule on the daily round.

It is, then, intuitively valid to presuppose, in natural talk, the existence of methods participants use to keep track of the directionality and cumulative value of each other's face work. This cumulative record serves to reify the mediating mechanisms of interactions, i.e. the reputation and identity of co-tranactants.

Relationship history is the cumulative record of episodal interactions between two individuals. Individuals use standard methods of keeping track. Garfinkel calls these methods "accounting practices" while we refer to them as "community cataloguing practices" or "CCP's" (see Jakobovits & Nahl, 1975-77).

To recapitulate, we are proposing that the directional syntax which generates the organizational sequencing of transactional moves in social episodes, is the dialectic of offense and remedy in face work. This motivational dynamic gives a goal-orientation to conversational exchanges, and accounts for the natural winding down of all episodes. Participants have standardized methods for keeping a cumulative record of each other's relationship history. These standardized methods for keeping track reify reputation and identity. Now we are proposing to investigate these cataloguing-practices that must form part of ordinary social competence on the daily round. By generating such data we are preparing the components needed for explaining the connectedness of utterances in interactional discourse. The explanation of this connectedness will involve explaining, (i) how participants keep track of what's going on, i.e. the features of the setting they orient to, notice, and code into a cumulative record; (ii) how the cumulative record reifies the natural phenomena of reputation and identity; and (iii) how reputation and identity provide a mechanism for relating moves to each other, i.e. for explaining the connectedness of utterances in a conversation.

We can only present here a brief account of our work thus far, but in any case, that work is only initial, our interest here being, to offer some possible directions which we are finding fruitful (Jakobovits & Nahl, 1978; 1979).

The Daily Round Archives or "DRA" as our students refer to it, can be viewed as a local ethnography of a community's daily round cataloguingpractices. What gets kept track of provides the investigator with the data needed to map the social settings and indicate their 'sociodynamic value', i.e. their influencing effects on the behavior of participants. The DRA files at the University of Hawaii are prepared and catalogued by students who are trained in psycholinguistic techniques applied to the natural history description of community life. We have been using ethnosemantic sampling techniques for building a cumulative catalog of social occasions on the daily round of our students. A portion of the current taxonomy is shown in the accompanying table:

go to DRA Classification Scheme

The DRA Index is a cumulative record of the cataloguing-practices of the community, as witnessed by the contributors to the daily round data bank. This catalogue serves to identify the ordinary noticings of participants engaged in social episodes, whether monadic (by onself), or diadic, or public. Contributors use a specified format of reporting which we provide and which evolves as we understand more about what's to be done! We find one particular technique useful, what we call annotations. Contributors make reports which involve the form of transcripts, microdescriptions, and interior dialog or discourse thinking (inner arguments and comments made to the self). These form the raw data. Next, the raw data are processed two ways through 'witnessings': first, the contributor annotates the raw data; second, "readers" annotate them as well, and in some cases, annotations are annotated by subsequent readers. This generational processing of the data bank creates a cumulative catalogue of what social stimuli are noticed and are being kept track of in a community. Assuming progress in a successive approximation towards the solution of the cataloguing issue (above), the next phase of this theoretical program is to ascertain the transactional value of the noticings on the daily round that participants keep track of. We do this by obtaining particular kinds of annotations of the recorded noticings by both the participantcontributor to the DRA and as well, by readers or users of the DRA. These 'readers' are successive generations of students so that the data in the DRA gets processed cumulatively in the form of annotations on annotations.

References

Aarons, A., Gordon, B. Y., & Stewart, W. (Eds.). Linguistic-Cultural Differences and American Education, Special Anthology Issue of The Florida FL Reporter, Vol. 7, 1969.

Garfinkel, H. Studies in Ethnomethodology. Englewood Cliffs, N.J.: PrenticeHall, Inc., 1967.

Goffman, E. Relations in Public: Microstudies of the Public Order. New York: Basic Books, 1971 (Harper Colophon Books, 1972).

Gordon, B. Y. "An Application of the Findings of Structural Linguistics to the Teaching of English in the Lower Elementary Grade: An Exploratory Study." Language and Literature, Linguistics, Columbia University, Ed.D., 1962.

James, Leon. "Comparative Psycholinguistics in the Study of Culture. International Journal of Psychology, Vol. 1, 1966, pp. 15-37, reprinted in translated version (French) in T. Slama-Cazacu (Ed.) La Psycholinquistique, Paris: Editions Klincksieck, 1972. - .3

James, Leon and Diane Nahl.  Society's Witnesses: Experiencing Formative Issues in Social Psychology. Department of Psychology, University of Hawaii, 1979 (mimeographed lecture notes).

James, Leon. Workbook for the Study of Social Psychology (2nd Edition). Department of Psychology, University of Hawaii, 1978.

James, Leon. Community Cataloguing Practices (Series I through VI). Department of Psychology, University of Hawaii, 1975-1977 (mimeo). (Abbreviated as J & G, 1975-77.) 24

Jakobovits, L. A. and Gordon, B.Y. The Context of Foreign Language Teaching. Rowley, Mass.: Newbury House Publishers, 1974.

Jakobovits, L. A., & Miron, M. S. Readings in the Psychology of Language, Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1967.

Sacks, H. Unpublished Lectures. University of California at Los Angeles, Davis (mimeo), ca. 1966.

Sacks, H., Shegloff, E. A., & Jefferson, G. A. Simplest Systematics for the organization of Turn Taking for Conversation. Language, 1974, 50, (4), 696-735. (Also reprinted in Schenkein, 19 78.)

Schenkein, J. (Ed.) Studies in the Organization of Conversational Interaction. New York: Academic Press, 1978.

Steinberg, D. S., & Jakobovits, L. A. (Eds.). Semantics: An Interdisciplinary Reader in Philosophy, Linguistics, and Psychology. New York: Cambridge University Press, 1971.

Winskowski, C. An Empirical Investigation of Topicalization. Unpublished Master's Thesis, Department of Psychology, University of Hawaii, 1975.