This study is concerned with the functioning of language in the establishment of relationships between conversationalists. In my introduction I shall explain the reasons for this investigation into the relational aspects of language use. In part I (chapter 1 - 4) I shall discuss the theoretical basis of the description of such a pragmatic language function. In part II (chapter 5 - 8) I shall elaborate on three relational dimensions. Part III will deal with some relational aspects of communication in intercultural, i.e. interethnic, contexts. The final chapter contains a summary of the relevant relational aspects with reference to the development of a schismogenesis between people. Part I Functional theories in linguistics may relate to either internal or organic functions. In the description and explanation of grammatical structures internal functions may be used, whether external (organic/activity indicating) functions are proposed or not. Without going into the linguistic discussion between formalism and functionalism, concerning the grammatical structure of the language, it will be clear that the study of language use must be seen, by definition, from a functional perspective. The elementary activity indicating functions that may be distinguished, might be based on the external components of the situation, or on the intentional states of the speaker, which may (not) be lexicalized. Previous proposals for functional differentiation have been discussed in relation to one another. I have tried to show that the elementary functions can be best described on the basis of the elementary components of the speech situation. However, we do have to take the interactional aspects of language use into account, and consequently the components referring to the relationship between speaker and hearer. It is argued that the place of relational function, concerning the social relationships between speaker and hearer, which might be settled verbally, is, in addition to the well-known Bühlerian distinctions, an essential aspect of language use. Many stylistic choices appear to be motivated by these social relationships between speaker and hearer, realized and/or presupposed by the speaker. Furthermore, the discourse function of language, i.e. the textual relation between utterances, is considered to be a relevant aspect in the understanding of conversational interaction. As a result I would propose a functional model, which may be seen as a pentagon (or a pyramid, depending on where one places the referential function). I support the pentagon idea. Utterances may be ascribed to one or all of these functions, presupposing a semantic description of the utterance. The Speech Act theory could be considered one of the most prominent functional theories with regard to the functionality of utterances in context, pointing out the propositional and contextual conditions for special illocutionary acts to be executed. However, the taxonomies which are proposed by several authors, do not provide an adequate description of relational functions. No well-known taxonomy has used relational conditions and effects as a criterion. In Searle’s taxonomy for instance, the relational component of illocutionary functions is considered to be just one of the twelve possible criteria mentioned, and is considered to be not important enough to form the basis of the taxonomy. Only Ballmer and Brennenstuhl’s taxonomy (concerning speech act verbs) has made some room for speech acts with a special type of relationship (called the ’struggle model’), but does not provide a description of speech acts that refer to other kinds of relationships. It is argued that this situation is caused by the lexical orientation of Searle and the other authors (even if they deny this) regarding speech acts. When the number of performative verbs denoting relational aspects is small, it might be easy to conclude, in their point of view, that relational acts are not worth mentioning. The lack of empirical analysis of speech acts in discourse may be seen as another cause for the absence of relational acts. It does not seem to be a coincidence, therefore, that the authors dealing with relational aspects within the framework of the speech act theory happened to be linguists, who tried to explain empirical phenomena in discourse. In a discussion of the work of these authors by Labov & Fanshel, Holly, Sager and Adamzik, however, it is concluded that the framework of the original version of the speech act theory (Austin, Searle) is not tenable. All these authors have to make acrobatic moves to fit their description within that particular framework, ignoring some elementary assumptions the framework is based on. All of these authors come across the problem that an utterance is a multi functional unit, whereas the speech act theory does not provide any instruments to deal with this multi functionality. No matter how much these approaches differ from one another with regard to their research methods, their common solution for the observed problem is to postulate speech acts as layered entities, with interrelated strata. Although the idea of a speech act as a layered entity may be seen as a way out, it is argued that the suggested connection of the illocutionary act (in the Searlian sense) with the relational acts is not very convincing. The problems with the speech act theory for the description of discourse are not restricted to relational aspects. We have seen that there are indeed a lot of problems with the speech act theory in this respect. The problem of explaining the relation of the content of the utterance and the illocutionary interpretation, in particular, does not appear to be solved. The description of utterances which are, in a way, not prototypical for a certain illocutionary act, such as indirect speech acts have evoked the question in the literature whether direct speech acts do indeed exist. Due to the problems in describing the relational aspects, in connection with the above mentioned inherent problems of the speech act theory, it may be concluded that the speech act theory should be changed in some essential ways. Referring to the relevance theory, it is suggested that conversationalists interpret the utterance semantically and use a lot of utterance cues to activate manifest contextual assumptions, to make a relevant functional interpretation on the levels indicated by the speaker. In that way we could describe the speech act from the speaker’s point of view as a bundle of intentions he tries to realize through the utterance of a parcel of language, in the context of the speech situation. Some of the intentions may be characterized by a performative verb, or by a discourse structuring verb, but in most cases there is no report of unity in the interpretation. That is why I regard this model as a fragmentation model of speech acts. The utterance may be the result of some of the intentional components of the speaker, and may therefore be accounted for in terms of strategies by which the goals of the speakers (their intentions) are achieved. Speakers choose several strategies to work out a particular intention, which has to be recognized by the recipient. These intentions may be based on the functional parameters as have been discussed, without the need to lexicalize. In this way we may account for the nuances in the functional interpretation as is desired by discourse analysts. We may regard this as a radical pragmatic theory, by marginalizing categorial interpretations and intentions. However, this model may be extended to less local goals. It is argued that the verbal communication in toto may be described as goal oriented, where so-called communicative principles (such as the relevance principle and the face saving principle) may be characterized as global, more general goals, whereas situational goals, related to the institutional contexts, for example, evoke special strategies which may explain aspects of language use characteristic for a particular institution. The constructivistic methodology for the description of relational strategies which has been proposed for the interpersonal strategies in a familiar speech community will be discussed in contrast with the strictly inductive methods, used in the ethnomethodological approach of the discourse analysis. It is argued that, although the strict empirical approach of the methodology of discourse analysis has provided us with a lot of interesting insights into the organisation of conversational discourse, this approach does not only lack the unifying and directing character of theory based approaches, but cannot account for intentional aspects of the communication either, because of the refusal to rely on intuitions about langauage use, which are not sequentially underpinned. The constructivistic approach is characterized by a set of theoretical preliminaries, based upon heuristic analysis of pragmatic phenomena, which have been worked out in relation to empirical materials. In this methodology different types of knowledge of language users about contextual characteristics are presupposed, and may bear part of the explanation of the observed phenomena. Besides that, intuitions about (un)marked language use, are exploited for the description of ‘regularities’, i.e. the strategies which are attributed to the speaker. Some theoretical preliminaries have been discussed in prior chapters, some will be discussed later on in connection with socio-psychological issues about relationships. The constructivistic approach in this study (part II) is speaker-oriented. This means that the description is not based on the description of hearer strategies for the interpretation of utterances, but on the description of productive strategies: given the specific goals in the interaction, the kinds of strategies that may be used, and the kinds of restrictions that will be put on the recipient. The relational strategies have been analyzed on three levels of the utterance, the level of performative use of speech act labels (meta-communicative strategies), the level of the propositional content (formulation strategies), and the level of the sequential position of the utterance (sequential strategies). Part III follows a more analytic approach. This part deals with relational strategies which may be used in conversational interaction as well as relational consequences of miscommunication in an intercultural setting and subtle differences indicating the distance to a conversational partner who belongs to an etnic minority. Such a study requires an analytic or a (field-) experimental approach even. Part II The socio-psychological and sociolinguistic literature regarding the relevance of relationships in communication has been reviewed, resulting in the differentiation between three relational dimensions in the description of language use: 1. solidarity - distance, 2. symmetry - dominance/submission, 3. co-operation - competition and agression. It is contented that these dimensions might be regarded as particularizations of a fundamental contrast between relations, viz. the tendency to integration and the tendency to differentiation. People are trying to maintain a balance between these poles in relation to the different dimensions. The first two dimensions have been discussed in past socio-linguistic literature on pronomina, address forms, politeness forms and power forms. But there is no systematic nor an integrative treatment of these phenomena in connection with other relational phenomena. The third dimension I distinguish, is relevant from the perspective of carrying out the tasks, which conversationalists have set for themselves. In this dimension especially, language users may indicate their intention to confirm the basis of the interactional order (as described by Goffman and Grice), or ignore it. These indications may or may not coincide with strategies in the other two dimensions. Therefore, I explore this dimension as a separate resource for language users. For each of these dimensions I have described relational strategies (on the three different levels) as the means by which integrative or differential intentions are realized. The first dimension refers to the amount of solidarity with the conversational partner, which may be conveyed by strategic language use. The meta-communicative strategies in this dimension, in which the speaker uses performative expressions, rarely occur in Dutch. As far as they are indicating social distance, it is shown that they function in the development of conflicts and are therefore, relevant for the co-operation dimension as well. The formulation strategies refer to both the code and the content of the utterances. With regard to the code speakers may use intonation aspects (pitch, speaking rate) to indicate the distance or solidarity towards the recipient, e.g. by forms of assimilation towards the intonation pattern of the recipient. The content strategies may be divided into strategies, by which the speaker expresses the distance to or familiarity with the recipient in a direct or more indirect way. In the direct way predicates denoting the measure of alliance between speaker and hearer, or discourse modal particles, indicating confidence and ‘togetherness’ in emotion and knowledge are used. The indirect strategies consist of the use of evaluative lexemes, referring to aspects of the recipient; but also the use of utterances which refer to a matter of course, or which presuppose the self-evidence of the content of the utterance, through the use of tag questions or particular modal particles, may be seen as indirect strategies conveying solidarity intentions. Finally the possibility of stylistic assimilation has to be mentioned. Sequential strategies that are relevant in the solidarity dimension are first of all related to the turn taking machinery. I determined the conditions on which interruption and simultaneous speaking are functioning to indicate togetherness. But more importantly in this respect is the preference structure as described in the discourse analysis. I argued in a discussion of some CA literature on this point, that response utterances, which follow initiative moves with a preference concerning the next utterance, may indicate an integrative or a differentiating intention in both the solidarity dimension and the co-operative dimension. In this dimension, we do not only have to account for the response moves, as I have pointed out, but for the closing third moves in sequential structures, which may confirm the mutual relationship as well. Finally I characterized the relational strategies in the solidarity dimension, used in the opening and closing sections of a conversation: the use of routines in the question/response sequence, and the relational routines in the pre-closing section of both parties, all of which are motivated by intentions in this dimension. The second dimension refers to the symmetry or asymmetry of a relationship. This dimension has one integrative pole, but two differential poles on the asymmetrical side, viz. dominance and submissiveness. Due to the unmarkedness of language on the symmetrical pole, I will refer to this dimension as the dominance dimension. The unmarkedness of the symmetry is shown in the absence of performative expressions, and in the relatively small amount of formulation strategies indicating such a relation. The symmetrical intention has to be realized mainly in the mutuality of certain forms of solidarity and social distance, and in the absence of strategies indicating an asymmetrical intention. Meta-communicative strategies may be used to convey intentions on the dominance and the submissive poles of the dimension; not only are illocutionary acts used in that way, but discourse structuring acts, especially those conveying dominance, as well. However, there is no indication about the basis for the asymmetry in the meta-cognitive strategies. Many formulation strategies in this dimension do not show such an indication either. These formulation strategies, conveying dominance over the recipient consist of depreciating terms and downgrading forms, regarding certain qualities or properties of the recipient, solidarity forms which cannot be returned, intensifying and upgrading forms regarding the speaker, direct forms to tell the recipient what to do, or a differentiation from the recipient in the other two relational dimensions, attention getting signals and relatively long turns. Formulation strategies conveying submissiveness are in some respects complementary: hesitant and careful speech, realized through hedges, tags, hesitations and (unilaterally used) deference forms. Besides these more general formulation strategies, there are certain formulation strategies which underly the special power bases as has been observed in the literature. The sequential strategies in this dimension are first of all found in the organisation of the turn taking in conversations. Dominance may show itself through turn assignment and self selection, as well as trough the use of unsupporting forms of interruption, while submissiveness is conveyed by carrying out the turn obligations, which are set by the prior speaker. Secondly, we observe sequential strategies in this dimension in the opening and closing sections of the conversation. It is argued that certain formulations in the opening sequence are relevant in this respect, and that the distribution of the speakers in the closing section may indicate the intended relationship in this dimension. Thirdly, we have to mention the strategies which are directing the recipient in a way he cannot easily avoid. Although one could say that speaking always directs the conversation in a way, there are strategies which have a special status in this respect. Many of these strategies make use of the question form of utterances. So we find the chaining of questions, the questioning in sub-sequential position (’arching’), and the indication of preferred answers. But in institutional interactions we may find a number of specific directing strategies. In doctor/patient conversations, for instance, we may observe talking down strategies in recapitulations, the absence of expressive responses, the absence of the diagnosis and the explanation of the prescription, whereas in court room interaction the special strategies, used by the judge, consist of the elicitation of conclusions, which means self accusation, through the use of a question series with presupposed answers. However, a possibility still remains for some contra-directing on the part of the patients and the suspects, which confuses the directing of the dominant speaker. The third dimension is called the co-operation dimension; the integrative pole is cooperation, while the differentiating pole of the dimension is competition and even agression. I have started by challenging the presumption that co-operation is to be considered an elementary condition for communication. In this dimension it is argued that people choose to define their relationship in a conversation, without blowing up the communication. Communication presupposes indeed a mutual orientation of speaker and hearer (which may be characterised in terms of the relevance principle), but that does not imply that their goals cannot be competitive or even agressive. Co-operation is defined in this study in terms of collective goals of the participants, with regard to the referential or the other functions in the conversation. Metacommunicative strategies in this dimension prove to be mostly responsive in character, presupposing a prior discourse. Speakers using these strategies may convey the measure of acceptance or non acceptance of the preceding utterances. On the level of the formulations most strategies convey non co-operation. Those kinds of strategies may be positioned on the continuum between the integrative and the differentiating poles, indicating that the goals of the speaker and the recipient are more or less oppositional, or that the participant is more or less disqualified as a co-operative conversation partner. These strategies vary from the use of comparision and contrast constructions in the typifying of the position of the conversation partner, to different forms of blaming and accusation, and to all kinds of abusive words and curses. On the sequential level we may observe co-operation strategies which prevent disorder in the turn taking machinery, by saving silences or interruptions. Another kind of such a sequential co-operation strategy is found in some introductory formulations, which precede non-accepting moves in the conversation. Through those kinds of strategies the speaker conveys his positive orientation to the contribution of the prior speaker, even when he is not accepting that contribution. In case of interactional incidents, the co-operation relationship is in danger; it may be restored by remedy strategies, but there may also be an escalation of the incident, as has been shown. Some types of escalation strategies are worked out. The end of an incident is the return to an interactional balance, which may be redefined, or the termination of the interaction. Part III When the relationship between participants of the conversation is determined by the social groups to which they supposedly belong, one may often observe relational phenomena in the interaction which refer to the fundamental opposition of the integration - differentiation continuum. The categorisation of the conversational partner as belonging to another group, and the treatment of that categorisation as important, makes that continuum more salient than the particularizations in the three dimensions. In this part of the study I concentrate on the intercultural communication and on the relational phenomena and strategies which supply the strategies which are discussed in part II. One of the most important aspects of intercultural communication between speakers of different cultures and languages is the possibility of assimilation on the part of the native speaker, showing an orientation towards the language proficiency level of the conversational partner, or on the characteristics of the ethnic variety, by converging to or diverging from that level or those characteristics. I discuss the way in which forms of foreigner talk and conversational adaptation may be described as relational strategies. My conclusions are first of all that the relational interpretation of foreigner talk is rather complex, dependent on the level of proficiency of the conversational partner, and secondly, that a convergence on the accent level (of a non-native speaker) is positively related to the level of co-operation of the (native speaker) conversational partner, and to the unmarkedness of the turn taking procedures, and finally, that the assimilation on a conversational level can only be described in relational terms in connection with certain conversational goals as well as the other relational strategies which are used in the conversation. Another aspect of the intercultural interaction which is related to relational work, is the existence of miscommunication. I have discussed the origins and the consequences of miscommunication regarding the relationship between the participants. The consequences of different cultural schemes for a certain context and the consequences of differences in conversational rhytm in particular have been discussed and worked out in a few examples. The phenomena which are connected with misunderstanding are not used in a strategic way to gain particular goals; but in most cases the recipients cannot distinguish it from intentional verbal behaviour either. So these phenomena of language use will be interpreted using the relational parameter, just as they were intended to be. The most important aspect of the relational aspects of intercultural communication is the way stereotype beliefs (or prejudices) have their influence on the conversation. The negative stereotypes may be conveyed, but because of the social taboe, they are often masked. In a comparative study of telephone conversations of two callers, one with a white ( (eastern Dutch) accent, and one with a black (Westindian) accent, with police officers, I have analysed the way in which certain stereotypes, connected with relational aspects, become visible in the language use of the police officers. The strategies which indicate differentiating intentions in this context (dominance, social distance and non co-operation) are a.o. an indication of outgroup position, downgrading, suspicion and interrogation. In the confirmation processes of these prejudiced beliefs I have analysed the way in which complementary schismogenesis could evolve in the relationships between people from different cultures. A discussion on that issue on a general level, completes this book.