This chapter presents a brief overview of the database and the distribution of adverbial clauses within it.

The data

The database includes thirteen naturally occurring telephone, face-to-face, two party, and multi-party conversations. None of the interaction originates from interview or data-elicitation formats. In this way, special turn-taking formats were avoided. Twelve of the conversations are audio taped, and one particularly long span of talk comes from a video taped multi-party conversation. All the data used for this study are transcribed according to the conventions of conversation analysis (CA). In each instance of an adverbial clause, I have done a careful analysis of intonation. In some cases, this has resulted in the addition of commas (level rising, incomplete intonation) or periods (final falling intonation) to the original transcriptions.

All the conversations are between adults, and all are in relatively casual situations: chatting on the phone, drinking beer on a picnic, visiting over crackers and cheese after a movie, or eating dinner. Eight of the recordings are of two-party telephone conversations, eliminating the question of non-verbal signaling for at least a portion of the data. Five of the conversations are face-to-face, multi-party interactions, in which the contribution of nonverbal cues will not be addressed in this study. There are a total of 33 different speakers in the conversations, 20 women and 13 men. The level of education of the participants has not been controlled for, although there is a preponderance of college-aged young people, as many of the recordings were originally collected for class assignments in CA courses.

In the present conversational corpus, initial adverbial clauses can be described in terms of the information patterns they form, and in terms of the interactional functions they serve. While the dichotomy between the information management or patterning and interactional functions of language is not a discrete one, there is value in approaching the description of adverbial clause usage with this division in mind. Because previous studies of adverbial clause usage have focused on monologue data, we know the kind of work such clauses do in less interactive discourse. That work has been described mainly in terms of information management. Prior studies have consistently pointed to a shift function for initial adverbial clauses. In monologue texts, such clauses set off prior discourse from discourse that follows. An initial adverbial clause uses information that has either appeared in some form in the previous discourse, or that follows sequentially from a point in the previous talk. Such information is either taken directly, negated or put in a contrasting form, or simply introduced as a possible option. The adverbial clause then constitutes explicit background for the following discourse.

With those findings as a source of comparison, we can look at the occurrence of adverbial clauses in conversation to see what such clauses do in encoding and organizing information and, additionally, in managing and maintaining interaction and the social roles of parties in conversations. It is assumed that information patterns, relations of these clauses to their textual environments, exist in conjunction with the interactional work that is being done at any point in a conversation.

Summary

In this study, I have used the framework of conversation analysis, and the body of findings associated with that approach, to examine the distribution and functions of temporal, conditional, and causal adverbial clauses in a corpus of American English conversation. In the present corpus, in line with findings from prior text-based analyses, discourse-structuring functions are realized through initial adverbial clauses, while final adverbial clauses tend to work more locally in narrowing main clause meaning without creating links or shift points in a larger discourse pattern. I have suggested that the pattern whereby conditional clauses are most likely to be initial, and causal clauses final, is related to an interaction between the inherent meanings of these clauses and the discourse functions those meanings are particularly suitable for serving. The common discourse organizational use of if-clauses is likely related to their hypothetical meaning; they are used to create temporary discourse realities, introducing and forming the background for associated modified material. Causal clauses, which present the sources and precipitating states or events that explain other states or events, are well-suited for appearing after the proposition, to be expanded upon, and for introducing background elaboration. They are especially useful as the vehicles for further explanation when problems arise in interaction. Temporal clauses are used most often in post-verbal position, functioning to ground the situation represented by the verb in time. When temporal clauses are placed initially, they are commonly involved in the structuring of discourse involving sequenced events. Temporal clauses are least common after the preceding clause has been finished with ending intonation.

Overview

In the previous two chapters I described the use of adverbial clauses presented in intonationally coherent packages with their associated main clauses. In the present chapter I examine cases involving grammatical connection through adverbial conjunctions, but in which the conjunction introduces a separate intonation unit. These adverbial clauses follow utterances ending in some form of final intonation, either low-falling or high-rising.

As described in chapter 4, final adverbial clauses following continuing intonation serve information-completing functions, narrowing clause meaning. Final adverbial clauses that follow ending intonation, though displayed through conjunctions to be extensions of previous units, also represent separate units in their own right. As will be seen in this chapter, these added-on adverbial clauses have clear interactional origins.

Adverbial clauses that are added after ending intonation have been referred to as “afterthoughts” (Chafe 1984). This term suggests that a unit of talk was originally planned not to include the adverbial clause, but that, after the unit was completed, the speaker decided to add another element of modification or elaboration. In the present chapter, I argue that, in addition to representing the editing of a speaker's talk based on her/his own thought process, such final adverbial clauses may also be the products of speaker–recipient negotiation specifically aimed at achieving interactional ends. There are specific conversational contexts in which speakers present a main clause plus adverbial elaboration in separated intonation units. The manner in which a speaker becomes aware that more might be added to an already completed unit often involves feedback from the other participants in the conversation.

In this chapter I will analyse two spoken texts extracted from longer conversations. The first is taken from Arnold and Tooley (1972) and the second from the Corpus. The analyses will also include some statistics. Special attention will be paid to the frequencies of the types of relationship between the non-prosodic CD distribution and the PP distribution.

In the diagnoses concerning the relation between the two distributions, the following new abbreviations are introduced: PERF (perfect) CORR (correspondence), N-R (non-reevaluating) INT (intensification), R (re-evaluating) INT (intensification), and DESH (deshading). The abbreviation N-R INT always denotes the selective type of non-reevaluating prosodic intensification (see p. 159).

In each case the abbreviation is followed by one of the following symbols: o (indicating absence of post-IC prosodic shade); 1 (indicating the presence of a post-IC prosodic shade made up of one or more context-dependent communicative units); 2 (indicating the presence of a post-IC prosodic shade made up of one or more context-independent communicative units occurring in the shade on account of their semantic character); 3 (indicating the presence of a post-IC prosodic shade produced by re-evaluating prosodic intensification); 1 + 2 (indicating the presence of a post-IC prosodic shade made up of one or more communicative units of type 1 and one or more communicative units of type 2); 0.2 (indicating the presence of a post-IC prosodic shade that arises within a communicative unit that provides a distributional subfield and stands last within the basic distributional field; the shaded units are of type 2).

A number of colleagues have asked me to prepare a synthesis of my publications on functional sentence perspective (FSP), which have appeared in various periodicals but are not always easily accessible. Among the British scholars it is especially Professors Sir Randolph Quirk, Sidney Greenbaum and Geoffrey Leech who have urged me to present such a synthesis. This is why I have gratefully accepted Professor Greenbaum's invitation to publish my book on FSP in the present series.

Inspired by the work of Vilém Mathesius, Josef Vachek, František Daneš and Maria Schubiger, I published my first paper on FSP in 1957, at a time when it was particularly Czechoslovakia where the problems of FSP were considered to be of special importance for a better understanding of how the sentence functions at the moment of communication. Since then, I have been gradually developing a theory of FSP covering both the written and the spoken language. I have been doing so predominantly on an empirical basis, concentrating on English but frequently comparing it with other languages, especially Czech, German and French. In the meantime a veritable explosion of interest in FSP has taken place (its problems being treated also under such headings as ‘theme–rheme structure’, ‘topic–focus structure’, ‘information structure’).

My endeavour to accompany my arguments by as many examples as possible and also limitations of space have prevented me from making my synthesis as comprehensive as I should have wished it to be and from presenting the reader with an all-inclusive and adequate survey of the state of the art.

The first part of the present chapter will amplify the preceding discussions by concentrating on the FSP function of transition proper in connection with (i) the boundary between theme and non-theme both in verbal and in non-verbal distributional fields and the related question of bipartition, tripartition or pluripartition, and (ii) the FSP of questions, negative sentences and commands. The second part will amplify the preceding discussions by returning to (iii) the concept of CD and (iv) the phenomenon of potentiality, and by dealing with (v) the FSP contextual applicability of a semantic and grammatical sentence structure and (vi) the signals of the distribution of CD degrees.

Boundary between theme and non-theme

In verbal fields

In the foregoing, emphasis was laid on the different roles played in communication by the notional component of the verb and its categorial exponents. Let me concentrate on the latter and outline their operation at the syntactic, semantic and FSP levels. Doing so, I shall contribute to the question of the boundary – and link, for that matter – between theme and non-theme.

At the syntactic level, the categorial exponents establish a link between the grammatical subject and the grammatical predicate. In this way, they serve as a centre within the sentence viewed as a field of syntactic relations. (‘Centre’ is not used here as a word order, but as a relational term.)

The syntactic meaning of the categorial exponents is rooted in their semantic content (see, p. 14).

The scales of dynamic semantic functions

The analyses of the texts in the introductory chapter and the discussion of the successful competitors of the verb show that, in relation to the information conveyed by the subject, the information conveyed by the verb, or rather by its notional component (see p. 70), participates in the development of the communication in one of two ways. It perspectives the communication either (i) towards the phenomenon presented by the subject, or (ii) towards the quality ascribed to the phenomenon expressed by the subject or beyond this quality towards its specification. In other words, it performs either (i) the dynamic semantic function of presentation (Pr), or (ii) that of expressing a quality (Q). In consequence, the subject either (i) performs the dynamic semantic function of expressing the phenomenon to be presented (Ph), or (ii) the dynamic semantic function of expressing the quality bearer (B). The discussion of the competitors of the verb has dealt with two other dynamic semantic functions: that of expressing a setting (Set) and that of expressing a specification (Sp).

The qualification ‘dynamic’ is necessitated by the fact that the semantic content concerned is not viewed as unrelated to the flow of communication, but as linked with definite contextual conditions and as actively participating in developing the perspective of the communication.

Word order as a system

The preceding chapters have demonstrated that in the interplay of the FSP factors determining the distribution of CD over the written sentence, linear modification manifests itself – if not worked counter to by the contextual factor and/or the semantic factor – through the actual linear arrangement of the elements. In this way it asserts itself as an important factor (principle) of word order. This raises the problem of the relationship between word order and FSP. This problem will be taken up by the present chapter. Let me first offer a few thoughts on word order in general and word-order principles in particular.

The literature on word order is extensive, but of specific relevance to my research is Mathesius' view that word order is a system constituted by the mutual relations of word-order principles (Mathesius 1942; 1975; 153–63). These principles are essentially valid for all Indo–European languages and possibly even for languages outside the Indo–European sphere, but may differ in the extent of their operation from language to language, or even from one period to another in the development of one particular language. Like other systems of language, the system of word order is not viewed as closed and perfectly balanced. Mathesius emphasizes that the word-order system of a language can be understood in a more comprehensive way if it is compared with that of another language, preferably one of different structure. He refers to this approach as the method of analytical comparison (Mathesius 1936: 95).

The verb and its successful competitors

The preceding chapter has shown that in determining degrees of CD, linear modification cannot assert itself if the contextual factor operates counter to it. Another factor capable of operating counter to linear modification is the semantic factor. Its operation will be discussed in the present chapter and in the chapter to follow. In the spoken language, the interplay of factors is joined by intonation. But intonation does not operate in the written language. It will be dealt with in the second part of the present study – in the part devoted to FSP in spoken communication.

The designation ‘semantic factor’ has been chosen for the sake of brevity. What it actually covers is the impact that the semantic character of a linguistic element, as well as the character of its semantic relations, has on the distribution of degrees of CD. The present chapter will concentrate on the role played by the semantic content of the verb and its semantic relations.

We already know that it is only in the absence of certain elements that the verb completes the development (see p. 7) of the communication within a distributional field. In other words, there are elements that, if present, take the development of the communication further than the verb and so come closer to, or even effect, the completion of the communication. In consequence, they prove to be dynamically stronger; they carry a higher degree of CD than the verb. In regard to the dynamics of the communication, they can be looked upon as successful competitors of the verb.