Urban Voices: The Sociolinguistics, Grammar and Pragmatics of Spoken Russian 3631664621, 9783631664629

Citation preview

Urban Voices: The Sociolinguistics, Grammar and Pragmatics of Spoken Russian The volume Urban Voices presents studies, analysing how speakers of Russian convey social meanings across a variety of speech situations. Rooted in quantitative and qualitative methodological frameworks, the contributions show how various linguistic, paralinguistic and pragmatic means relate to sociolinguistic dimensions (e.g. display the social, ethnic, local identity of a speaker, the institutional character of a communicative situation). The analyses are the results of the research network Urban Voices, which focuses on the sociolinguistics, grammar and pragmatics of spoken Russian and investigates its linguistic and communicative diversity. The contributors are Olga Blinova, Nina Bodganova-Beglarian, Beatrix Kreß, Lenore Grenoble, Peter Kosta, Elena Markasova, Ludger Paschen, Nicole Richter, Christian Sappok, Tatiana Sherstinova and Nadine Thielemann.

The Editors

Nadine Thielemann / Nicole Richter (eds.)

Nadine Thielemann / Nicole Richter (eds.)

25

Urban Voices

Potsdam Linguistic Investigations Potsdamer Linguistische Untersuchungen Recherches Linguistiques à Potsdam 25

Potsdam Linguistic Investigations Potsdamer Linguistische Untersuchungen Recherches Linguistiques à Potsdam 25

Nadine Thielemann / Nicole Richter (eds.)

Urban Voices: The Sociolinguistics, Grammar and Pragmatics of Spoken Russian

Nadine Thielemann is a professor of Slavic linguistics at Vienna University of Economics and Business. She is interested in the linguistics and pragmatics of talk-in-interaction, political discourse and multilingualism at the workplace. Nicole Richter is a professor of Multicultural Communication (Slavonic and English Linguistics) at the European University Viadrina, Frankfurt (Oder) and at Collegium Polonicum, mainly interested in contrastive phonetics and prosody.

ISBN 978-3-631-66462-9

PLI_025 266462_Thielemann_JA_A5HC 151x214 fusion.indd 1

10.12.18 17:21

Urban Voices: The Sociolinguistics, Grammar and Pragmatics of Spoken Russian

POTSDAM LINGUISTIC INVESTIGATIONS POTSDAMER LINGUISTISCHE UNTERSUCHUNGEN RECHERCHES LINGUISTI UES POTSDAM Edited by / Herausgegeben von / Edit par

Peter osta, Gerda Ha ler, Teodora Radeva-Bork, Lilia Sch rcks, Nadine Thielemann and/und/et Vladislava Maria Warditz

Editorial Board: Tilman Berger (University of T bingen, Germany) eljko Bo kovi (University of Connecticut, USA) Sarah Dess Schmid (University of T bingen, Germany) Anna Maria di Sciullo (U AM / Universit du u bec Montre l, Montreal, Canada) Steven Franks (Indiana University, Bloomington, USA) Atle Gr nn (University of Oslo, Norway) Holger u e (Dresden University of Technology, Germany) Hans-Georg Wolf (University of Potsdam, Germany) Ghil ad uckermann (University of Adelaide, Australia)

Vol./Bd 25

















Die ualit t der in dieser Reihe erscheinenden Arbeiten wird vor der Publikation durch einen externen, von der Herausgeberschaft benannten Gutachter im Double Blind Verfahren gepr ft. Dabei ist der Autor der Arbeit dem Gutachter w hrend der Pr fung namentlich nicht bekannt; der Gutachter bleibt anonym.



















Prior to publication, the uality of the work published in this series is double blind reviewed by an external referee appointed by the editorship. The referee is not aware of the author s name when performing the review; the referee s name is not disclosed.

Nadine Thielemann / Nicole Richter (eds.)

Urban Voices: The Sociolinguistics, Grammar and Pragmatics of Spoken Russian

Bibliographic Information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available online at http://dnb.d-nb.de.

Library of Congress Cataloging‐in‐Publication Data A CIP catalog record for this book has been applied for at the Library of Congress.

Gefördert mit einem Druckkostenzuschuss der DFG im Rahmen des Projekts Urban Voices / Stimmen der Stadt DFG TH 1506/2-1 Printed by CPI books GmbH, Leck

ISSN 1862-524X ISBN 978-3-631-66462-9 (Print) E-ISBN 978-3-653-05675-4 (E-PDF) E-ISBN 978-3-631-70674-9 (EPUB) E-ISBN 978-3-631-70675-6 (MOBI) DOI 10.3726/978-3-653-05675-4

© Peter Lang GmbH Internationaler Verlag der Wissenschaften Berlin 2019 All rights reserved. Peter Lang – Berlin · Bern · Bruxelles · New York · Oxford · Warszawa · Wien

All parts of this publication are protected by copyright. Any utilisation outside the strict limits of the copyright law, without the permission of the publisher, is forbidden and liable to prosecution. This applies in particular to reproductions, translations, microfilming, and storage and processing in electronic retrieval systems. This publication has been peer reviewed. www.peterlang.com

Table of Content Nadine Thielemann/Nicole Richter Introduction and Overview ....................................................................................... 7 Sociolinguistics Tatiana Sherstinova Studying Linguistic Variation and Communicative Diversity from the One Day of Speech-Corpus ....................................................................................... 15 Christian Sappok Addressing by Voice: Deixis and Design ................................................................ 37 Grammar Nadine Thielemann Constructional Patterns Emerging in Talk-In-Interaction – Some Preliminary Observations on the Grammar of govorit’ in Colloquial Russian ........................................................................................................................ 67 Olga Blinova Побудительные реплики и их компоненты в русском устном диалоге: количественное описание (Imperative Utterances and Their Components in Russian Spoken Dialogue: Quantitative Description) ...................................................................... 95 Elena Markasova Грамматика речевого доминирования (настоящее время) (A Grammar of Verbal Domination: The Present Tense) .................................. 125 Pragmatics Beatrix Kress Relational Work in Conversation .......................................................................... 151 Nicole Richter How Questions Can Affect Replies: Some Insights from Spoken Russian ...... 167

6

Table of Content

Peter Kosta Third Position Repair, Overlaps, and Code-Switching within a Strict Turn-Taking Model....................................................................................... 183 Natalia V. Bogdanova-Beglarian О возможных «выходах» из хезитационного поиска в процессе речепорождения (On Options for the Resolution of Hesitation Phenomena in the Process of Speech Production) .............................................................................. 203 Lenore A. Grenoble Multimodality and Interactional Grammar in Russian Conversation ............. 219 Ludger Paschen On Clicks in Russian Everyday Communication ................................................ 237

Nadine Thielemann/Nicole Richter

Introduction and Overview Face-to-face-interaction in the urban space is characterized by a wide spectrum of differentiation. The communicative situation, the composition of the group and social as well as the ethnic background of the speakers varies. All these aspects involve processes and mechanisms of accommodation and separation or othering. Interlocutors employ various linguistic, paralinguistic and pragmatic means in order to convey their social role (e.g., display the social, ethnic, local identity of a speaker, the institutional character of a communicative situation, etc.) which results in the social meaning. It is this linguistic and communicative diversity characterizing the repertoire of urban Russian dwellers which has been the focus of the scientific network “Urban Voices” funded by the German Research Council (DFG).1 This volume presents the results of research projects which the members of the network – linguists from the State University of Saint Petersburg and several German universities (Hamburg, Greifswald, Potsdam, Hildesheim, Frankfurt/Oder, Bochum, Munich) – have worked on during an intense period of collaboration (2013– 2016). The contributions reflect the variety of methods, topics, and approaches in research into spoken Russian within the network. The network members focus on different aspects and rely on different methods, both qualitative and quantitative. The volume thus reflects the design of the network, which was based on methodological triangulation in order to assess both the benefits and the challenges in applying and combining different analytic approaches. The overall nexus is the work with spoken language, including issues concerning the collection and preparation of data reflecting the linguistic and communicative diversity of urban speakers of Russian both in Russia and abroad. An important database formed by long-lasting recordings of a speaker’s communicative day has been provided by the Russian members who started collecting the One day of Speech-corpus in 2007 (ORD, Saint Petersburg) (Sherstinova 2009, Asinovsky et al. 2009).2 A sample 1

2

All members of the network “Urban Voices – Linguistic and communicative diversity in a face-to-face interaction of Russian-speaking interlocutors in Saint Petersburg and German cities” (DFG TH 1506/2-1) express their gratitude to the German Research Council for the generous funding of their meetings. The first recordings of the ORD corpus were supported by the Russian Foundation for Humanities within the framework of the project “Speech Corpus of Russian Everyday

8

Nadine Thielemann/Nicole Richter

selection from this corpus reflecting the linguistic and communicative diversity formed a basis for the joint work within the network “Urban Voices” in addition to recordings made by the members for the duration of network research activities. In the remainder of this section, we briefly sketch the major research focus of the network, which is also reflected in the structure of the current volume, i.e. the sociolinguistics, grammar and pragmatics of spoken Russian. These three perspectives have been adopted and extensively discussed during three topical workshops, organized in Hildesheim, Frankfurt/Oder, and Potsdam. Prior to this, however, we address the specific database involved in the project. Research into the linguistic and communicative diversity of spoken Russian is particularly challenging due to the limited availability of data. Edited collections of spoken language data for Russian are scarce and have mostly been prepared for specific purposes (e.g. Zemskaja/Kapanadze 1978, Šalina 2011). Although urban styles and registers, as well as situational or functional variation characterizing urban communication, are covered by some of them (e.g. Šalina 2011, Kirilina 2009, Zemskaja (Ed.) 1984, Šmelev/Zemskaja (eds.) 1984), colloquial Russian i.e. the language from casual face-to-face interaction of well-educated urban dwellers features prominently in the research based on these and similar data (e.g. Zemskaja 1973, 1979, 1983, Lapteva 1976, Mills 1990). Several settings and sociodemographic parameters are missing from almost all of these collections and the analyses based on them e.g. migrants’ communication. Most collections do not offer full access to the audio data and include transcripts which provide only very little information especially when it comes to interactional or paralinguistic features. In order to assess all (or at least as many as possible) potentially relevant parameters and features, an important prerequisite is an access to audio data which allows for the preparation of transcripts adequately reflecting the phenomena of interest. The analyses in this volume meet this desideratum in several ways. They are all based on audio data from the ORD corpus (or a sample selection from it) or on recordings taped by the authors. Peter Kosta, for example, recorded Russian migrants living in the greater Berlin area. They grant access to several sociolinguistically relevant parameters and settings that have been under-researched until now, as well as access to linguistic, paralinguistic and interactional features conveying social meanings (e.g. vocal features, clicks, hesitation markers). Lastly, Communication “One Speaker’s Day” (project # 07-04-94515e/Ya). The significant enlargement of the ORD corpus has been funded by the Russian Scienсe Foundation, project # 14-18-02070 “Everyday Russian Language in Different Social Groups”. Currently, further corpus development is supported by the Russian Scienсe Foundation project # 18-18-00242 “Pragmatic markers in Russian everyday speech”.

Introduction and Overview

9

they allow for the preparation of transcripts which conform to the demands of the analytical approach taken (e.g. Conversation Analysis, Functional Pragmatics) or when applying analytic methods that require audio data such as phonetic or prosodic analysis with praat.

Overview of the contributions The contributions reflect the variety of topics and approaches taken by the members of the network which covers a range of social meanings such as gender, situation or dominance and/or cooperation and analysis of how linguistic, paralinguistic and pragmatic resources contribute to their signaling. Contributions are arranged in line with the perspective highlighted – sociolinguistics, grammar, and pragmatics. Tatiana Sherstinova and Christian Sappok analyze how lexical preferences, as well as vocal features, correlate to social as well as situational parameters based on ORD. Both adopt a sociolinguistic perspective, while Sherstinova builds her argument on quantitative and Sappok on qualitative analysis. Tatiana Sherstinova’s contribution familiarizes the reader with the structure and organization of the ORD-corpus. She shows how several dimensions of sociolinguistic variation are made accessible in ORD ranging from age, gender and educational background to communicative situation and genre. ORD further provides a solid database for quantitative analysis in the spirit of variationism which is demonstrated in her frequency analyses. She determines the lexical units most frequently used by men and women and correlates these with other sociolinguistic variables (professional and educational background, client – service encounter). Her results are consistent with gender-linguistic findings based on data from other speech communities which often do not rely on such a solid database. Christian Sappok is interested in voice and variation as a means of adapting to various addressees and communicative situations. Promoting a perceptual approach, he shows how speakers contextualize the mode of communication (e.g. soliloquy, boss-talk) through audible vocal shifts and reveals the tremendous repertoire of voices in merely one speaker. His insights into the adaptive role of voices provide further empirical evidence for Accommodation Theory and Bell’s concept of audience design. Nadine Thielemann, Olga Blinova, and Elena Markasova deal with linguistic aspects of spoken Russian. Their contributions analyze grammatical features of talk-in-interaction ( i.e. the language from face-to-face interaction) and view them as resources in dealing with recurrent tasks in conversations. Nadine Thielemann analyzes the patterns emerging with govorit’ (‘to speak’) in ORD, combining Interactional Linguistics and Construction Grammar. She shows

10

Nadine Thielemann/Nicole Richter

how each constructional pattern serves to highlight a specific participation role. This provides an argument for the grammar of talk-in-interaction being shaped by and tailored to the demands of social interaction. It offers an alternative perspective on register-specific grammar – a perspective that includes pragmatic features in the constructional analysis. Olga Blinova analyzes imperatives in ORD from a quantitative perspective. She determines the lexemes which most frequently occur in 2nd person imperative forms, their aspectual distribution, the units with which they regularly cooccur and their position in the turn and the sequence. The results of her analysis not only support accounts of aspectual linguistics but also demonstrate how frequent linguistic patterns of spoken language such as slušajte, smotri(te) or podoždi(te) undergo semantic bleaching and acquire discourse-specific pragmatic meanings. Elena Markasova approaches grammar, more precisely present tense, as a means to shape the interlocutors’ relationship. She provides examples that illustrate how the present tense of particular verba dicendi, mental verbs, and activity verbs impacts the interlocutors’ relationship in a way which either maintains the balance or which is indifferent towards the maintenance of the social equilibrium. The last section gathers contributions dealing with pragmatic features of Russian face-to-face interaction. They analyze sequential patterns as well as linguistic and paralinguistic means regularly employed by Russian interlocutors to organize their social interaction. Beatrix Kress discusses several concepts and frameworks for the analysis of the interpersonal dimension of face-to-face interaction including functional pragmatics and linguistic politeness. Using an example from ORD, she uncovers the complex interplay of several linguistic and pragmatic features in the signaling of rapport and other interpersonal meanings in interaction. Nicole Richter is interested in adjacency pairs of questions and replies. Her analysis shows how specific question formats combine with specifically designed replies and thus facilitate the emergence of particular sequential patterns. Attention is also devoted to the transition between the first and the second part of these adjacency pairs. Peter Kosta, too, deals with a sequential pattern, namely repair sequences. Using examples from ORD and from his own recordings of Russian-speaking migrants from Berlin and Potsdam, he conducts a fine-grained sequential analysis of the repair mechanisms in Russian including different trouble sources and repair initiators. He shows how migrant speakers of Russian rely on code-switching in their repair sequences.

Introduction and Overview

11

Natalia Bogdanova-Beglalrijan is also interested in repair and repair-related phenomena in Russian interaction. Her starting point, however, is not the sequential repair mechanism but the trouble source and the repair initiator which she refers to in terms of hesitation. She identifies the verbal as well as paralinguistic markers employed by Russian interlocutors and in a next step identifies the sequential procedures by which they resolve the interactional problem. Lenore Grenoble and Ludger Paschen further pursue the line of research into paralinguistic phenomena in Russian face-to-face interaction and thus add massively under-researched aspects to the volume. Lenore Grenoble addresses the role of paralinguistic means such as manual gestures and clicks and analyzes how they contribute to the projection of upcoming units on several levels. Her analyses are based on video data from screened interaction and reveal that the combination of several (linguistic and paralinguistic) resources affects projection at a discourse level, within turns, and across turns. Lastly, Ludger Paschen analyses click and their functions in Russian face-to-face interaction. Though their use displays high inter-speaker variation, his analysis clearly reveals that clicks facilitate the transition at various discursive boundaries. Interlocutors employ them in order to mark a shift of topic or activity and as turn-yielding devices in (competitive) turn-taking. He also finds evidence that clicks occur at moments in which interlocutors accomplish a shift in their interpersonal, affective or attitudinal stance.

References Asinovsky, Alexander / Bogdanova, Natalia / Rusakova, Marina / Ryko, Anastasija / Stepanova, Svetlana / Sherstinova, Tatiana: “The ORD Speech Corpus of Russian Everyday Communication “One Speaker’s Day”: Creation Principles and Annotation”. In: Matoušek, Vaclav/Mautner, Pavel (eds.): TSD 2009. LNAI 5729, 2009, pp. 250–257. Kirilina, Alla V.: “Russkij jazyk v megapolise kak indikator izmenenija jazykovoj situacii”. In: Ždanova, Vladislava (ed.): Russkij jazyk v uslovijach kul’turnoj i jazykovoj polifonii. Sbornik statej. Sagner: München 2009, pp. 75–89. Lapteva, Ol’ga A.: Russkij razgovornyj sintaksis. Nauka: Moskva [1976] 2007. Mills, Margret (ed.): Topics in Colloquial Russian. (American University Studies; Series XII, Vol. II). Lang: New York et al. 1990. Sherstinova, Tatiana: “The Structure of the ORD Speech Corpus of Russian Everyday Communication”. In: Matoušek, Vaclav/Mautner, Pavel (eds.): Text, Speech, and Dialogue. (LNAI; 5729). Springer Heidelberg 2009, pp. 258–265. Šalina, IirinaV. (ed.): Živaja reč’ ural’skogo goroda: ustnye dialogi i ėpistoljarnye obrazcy. Izdanie ural’skogo universiteta: Ekaterinburg 2011.

12

Nadine Thielemann/Nicole Richter

Šmelev, Dmitrij N./Zemskaja, Elena A. (eds.): Raznovidnosti gorodskoj ustnoj reči. Nauka: Moskva 1988. Thielemann, Nadine/Kosta, Peter (eds.): Approaches to Slavic Interaction. Benjamins: Amsterdam/Philadelphia 2013. Zemskaja, Elena A.: Russkaja razgovornaja reč’. Lingvističeskij analiz i problemy obučenija. Nauka: Moskva 1973. Zemskaja, Elena A.: Russkaja razgovornaja reč’. Lingvističeskij analiz i problemy obučenija. Nauka: Moskva 1979. Zemskaja, Elena A. (ed.): Russkaja razgovornaja reč’. Fonetika. Morfologija. Leksika. Žest. Nauka: Moskva 1983. Zemskaja, Elena A. (ed.): Gorodskoe prostorečie. Nauka: Moskva 1984. Zemskaja, Elena A./Kapanadze, Lara A.: Russkaja razgovornaja reč’. Teksty. Nauka: Moskva 1978. Zemskaja, Elena A./Kitajgorodskaja, Margarita V./Širjaev, Evgenij N.: Russkaja razgovornaja reč’. Nauka: Moskva 1981.

Sociolinguistics

Tatiana Sherstinova

Studying Linguistic Variation and Communicative Diversity from the One Day of Speech-Corpus Abstract: The paper discusses a corpus approach for studying linguistic variation in everyday spoken Russian. The methodology of long-term audio recording used for collecting spoken data for the “One Day of Speech” corpus (the ORD corpus) has made it possible to obtain a representative collection of recordings made by volunteers in natural everyday contexts. At present, the corpus contains more than 1250 hours of audio recordings referring to various communication situations (business communication, family conversations, customer-service communication, educational communication, etc.). Moreover, most of the participants had different social roles in one day of recording, providing valuable material to study variants and styles of spoken Russian. All the recordings gathered from the participants are annotated in detail to facilitate data retrieval from the corpus and their further analysis concerning linguistic, sociolinguistic and pragmatic factors. The paper gives an overview of spoken styles observed in corpus recordings, describes approaches to the study of linguistic variation and communicative diversity based on the corpus, and gives an example of studying lexical variation in speech depending on speaker’s gender and the type of communication.

1. Introduction. The “One Day of Speech” corpus of everyday Russian The “One Day of Speech” corpus is a 1-million-word linguistic resource of presentday Russian. The corpus is being created with the aim to study spontaneous Russian speech in natural communicative situations (Asinovsky et al. 2009). Individualsvolunteers of both sexes between the ages of 16 to 83 and of different occupations were asked to spend a day with active audio recorders to record all of their verbal interactions. In most cases, the participants made recordings over the course of a single day, which gave rise to the Russian title of the resource “One Day of Speech” (in Russian — “Odin Rechevoy Den’”), which is abbreviated as the ORD corpus. When recording, participants were asked to go about their day as usual and to do what they ordinarily do (ibid.). They were requested to keep a diary and to complete a sociological questionnaire and several psychological tests. In the journal, each respondent briefly described his/her primary daily events (communication situations) including what he/she was doing, where and with whom.

16

Tatiana Sherstinova

The participants obtained permission for recording from their interlocutors with whom they engaged in spoken communication and provided information concerning their social role in the given setting, as well as data on their sex, age, social group, profession, etc. A similar method for obtaining long-term recordings had been earlier used for gathering data for the British National Corpus (Burnard 2007) and the JST ESP corpus in Japan (Campbell 2004). The recordings were made in St. Petersburg in 2007, 2010 and 2014–2015. The last series of recordings was established within the sizeable sociolinguistic project “Everyday Russian language in different social groups” supported by the Russian Science Foundation (Bogdanova-Beglarian et al. 2015). At present, the corpus contains 1250 hours of audio recordings made by 130 respondents (69 men and 61 women). The data refer to various communication situations (business/ professional, familial, customer-service related, educational, etc.) and therefore to diverse spoken language styles. The recordings were made at different places — at home, in the offices, in educational institutions, in cafeteria and restaurants, in the car or outdoors, etc. The conversations cover a wide range of topics (e.g., family matters, work and professional issues, hobbies, leisure, politics, medicine, consumer goods, theatre, sports, arts, and many others). In the course of the recording, most of the participants had different social roles, providing valuable material to study variants and registers of spoken Russian. All recordings gathered from the participants have been annotated in detail to facilitate data retrieval from the corpus and their further linguistic, sociolinguistic and pragmatic analysis. Thus, the ORD corpus has become an unparalleled resource in allowing the study of everyday speech variation and communicative diversity in Russian.

2. Social variables in the ORD corpus The critical social factors which are usually involved in studying linguistic variation are a social group, ethnicity, age, gender, and style (Romaine 2008). The questionnaire filled out by all participants is the primary source of data for the sociological description of the ORD corpus (Asinovsky et al. 2009). The survey solicits the following information: 1) gender, 2) age, 3) place of birth, 4) the native language, 5) the other languages spoken by the respondent, 6) qualification (level of education or technical training/certification), 7) current occupation, 8) previous occupation, and 9) places of long-term residence. Two additional questions related to the respondent’s parents, specifically their social background and nationality, were posted. Filling in the latter one is not mandatory. The same questionnaire was completed for all interlocutors who gave their consent for the

Studying Linguistic Variation and Communicative Diversity

17

recording. A detailed description of the current survey can be found in Baeva (2014). Let us consider what kind of social variation may be analyzed using the material of the ORD corpus: Gender variation. In describing ORD recordings, we have distinguished two gender categories: 1) men, and 2) women. From the beginning of work on the corpus, we tried to balance the participants according to this parameter. Unsurprisingly, these two social groups are the most numerous in the corpus. Age variation. In responding to the sociological questionnaire, each respondent indicated his/her exact age in years. Each participant was then tagged with this number in the database. When conducting research, it is possible to sort the data into different groups, according to the researcher’s aim. For example, there may be two groups —“early adulthood” and “later adulthood” (usually, 18–35 years old, and 36+). There may be three groups — “the young” (18–35), “the middle-aged” (36–59) and “seniors” (60+), and even five groups (18–24, 25–34, 35–44, 45–59, 60+), etc. The borders for the age groups may shift according to researcher’s preferences. Native language variation. Currently, the vast majority of participants were chosen from individuals whose native language is Russian; therefore, this parameter is not valuable for the corpus data. Nationality variation. While it was not obligatory to respond to the questionnaire on the subject of the nationality of their parents, the participants usually did. Most of the respondents identified their parents as Russian, though other conventional answers included Tatar, Jewish, Ukrainian, and Belarusian. Regional variation. With a few exceptions, all recordings were made in St. Petersburg. The corpus is therefore imbalanced regarding regional variation in favor of the St. Petersburg variant of Russian. Many of the participants are native to Petersburg, and many others have been exposed to this variant for long periods of time. However, numerous participants were born in other regions of Russia or/and resided there for a significant part of their lives. This information can be retrieved from two questionnaire items: “Place of birth” and “Places of long-term residence.” We have representatives from various Russian regions (Leningrad region, Moscow, central Russia, Volga region, southern Russia, Eastern and Western Siberia, etc.) and foreign countries that were part of the former USSR (Ukraine, Estonia, Kazakhstan). Some of these internal migrants have been living in St. Petersburg for many — even dozens of — years. Social group variation is currently regarded through the lens of three critical dimensions: (a) professional groups, (b) educational groups, and (c) status groups.

18

Tatiana Sherstinova

Professional group variation. When planning a sociolinguistic extension of the corpus (Bogdanova-Beglarian et al. 2015b), we decided to limit it to the following ten rather large occupational groups, each representing a cluster of professions: 1) Blue-collar workers, individuals who do manual or industrial labor. 2) White-collar workers, individuals working in offices (in particular, those engaged in economics and public relations). 3) Service sector employees (salespersons, cashiers, delivery couriers, waiters, hair stylists, massage therapists, etc.). 4) IT professionals (IT engineers, programmers). 5) Engineering professionals engaged in various industries. 6) Education sector employees (practicing teachers, lecturers, university professors, tutors, etc.). 7) Specialists in humanities (linguists, archeologists, historians, philosophers, etc.). 8) Specialists in natural science (biologists, chemists, astronomers, etc.). 9) Military and security sector employees (officers, cadets, security guards). 10) The people of creative professions (photographers, designers, artists, architects, musicians, stage directors, etc.). The analysis of respondent’s questionnaires has revealed that very often they may belong to two or even more professional groups from this list. For example, specialists in the humanities or natural sciences often work as university professors1, IT or engineering professionals can be involved in the service sector, and so forth. The diversity of individuals’ occupations becomes more evident when considering both previous and current occupations of respondents. Sometimes people’s current professions are rather far from the qualifications they obtained from their universities or other educational institutions. Such professional heterogeneity should be taken into account when analyzing respondents’ speech. For this purpose, we have introduced an additional parameter called “the primary professional group” for each respondent, which is expertly assigned to each participant. Educational group variation. Here, we distinguished the following levels of education: 1) secondary and secondary technical school; 2) incomplete higher education; 3) higher education; and 4) academic degrees (Ph.D. or higher). Currently among the respondents of the ORD corpus a specific prevalence of individuals with higher education is observed.

1

More generally, specialists in any discipline can be engaged in the education sector.

Studying Linguistic Variation and Communicative Diversity

19

Official position or status variation. To define this variation, the five categories are defined as follows: 1) leading officials, managerial employees, people in business or the self-employed, 2) regular employees, specialists, professionals, workers; 3) students and pupils, including those who are working part-time; 4) unemployed, including unemployed pensioners.

3. Variation across styles and registers It is commonly acknowledged that style is a term which is somewhat difficult to define. However, it is accepted that linguistic style marks out or indexes social difference (Coupland 2007) while variation between genres or text types is thought of as a type of stylistic difference (Romaine 2008). A register, or way of speaking, is “a variety of a language that is appropriate in specific situations” (Ottenheimer/ Pine 2018, p. 182). The ORD data allows the study of variations in linguistic style in everyday spoken Russian as well as the investigation of a communicative diversity of everyday discourse (Sherstinova 2015). As the corpus contains real-life recordings of a single person in different discourse situations and the records of different people in similar contexts, it becomes possible with this data to study both inter- and intrapersonal language variation2. Information on linguistic styles and communication situations is noted in the macro episodes annotation, which is obligatory for all corpus recordings. Macro episodes are massive communication episodes of everyday life, usually homogeneous regarding language style (the type of communication), participants and locus (place of discussion) (Sherstinova 2015). In some way, macro episodes are similar to the scenes in dramatic plot structure in that they are confined to a single location. Similarly, the recordings of the ORD corpus were divided on the basis of site or scenes of communication (e.g., at home, in the office, in the hospital, in the fitness club, in the shopping center, at home again, etc.), and then on the basis of language style and participants. Each of these scenes constitutes a macro episode, which has been annotated in detail. In cases where a relatively homogeneous real-life episode lasts a rather long time (e.g., several hours), it is divided into several audio files for practical reasons. According to the ORD data, individuals take part in 20–50 macro episodes every 24 hours with the length of macro episodes varying from 15 to 90 minutes on average (ibid.).

2

The latter reflects differences in the way style person speaks with different interlocutors in different settings (Meyerhoff 2006, p. 17).

20

Tatiana Sherstinova

The principles of macro episode annotation are described in Sherstinova (2015). Expert researchers manually annotate all ORD episodes. Besides indicating the type of communication concerning the particular language style, each macro event is interpreted with a verbal description encompassing three questions: (1) Where does the situation take place? (2) What are the participants doing? (3) Who is (are) the primary interlocutor(s)? Further, these verbal descriptions are normalized to standard codes and may be supplied with additional markers, indicating specific registers. As a result, the annotation of each macro episode includes information on the scene and type of communication, social roles of participants, and a few other factors. These data are used for searching and filtering macro episodes upon user request. Style variation. There are five main styles which are distinguished in ORD data: 1) professional conversations, 2) customer-service communication, 3) educational discussions and lectures, 4) public speeches, and 5) informal conversations. Style attribution of speech is determined by the type of communication in the given setting. 1. The situation is attributed as “professional communication” (business-orientated official communication) if the topic of the conversation concerns speakers’ professional (administrative, business, etc.) duties and if the interlocutors are professionals or specialists in the given sphere. 2. Customer-service communication takes place in various situations, in which one participant (seller, waiter, concierge, social worker, doctor, librarian, etc.) is serving or helping another (customer, patient, client, etc.) on a professional level. 3. Educational communication implies a variety of activities, such as university studies, educational courses, practical training, private lessons, teaching children, etc. 4. Public speeches include public lectures, presentations, etc., which are part of neither educational nor professional duties. 5. Finally, informal communication is assigned to the most episodes not referred to in any of the styles mentioned above. Most private conversations, spoken communication between friends and other relatives as well as discussions with colleagues and acquaintances on informal topics (family matters, domestic issues, sports, hobbies, etc.) fall into this category. Moreover, it should be mentioned that casual conversations are possible between participants of fairly different social roles. According to recent statistics obtained from the ORD data, for a subset of 1854 macro episodes collected from 79 respondents with a total duration of 483 h,

Studying Linguistic Variation and Communicative Diversity

21

informal conversations comprise up to 61% of all macro episodes; professional discussions, 23%; customer-service communication, 8%; educational communication, 6%; and public speeches, 2% (Sherstinova 2015). Thus, despite the fact that in modern society a considerable portion of real-life situations concerns official and semi-official relations of interlocutors (Krysin 2012), the share of informalstyle episodes in the ORD corpus is rather high even in offices. Moreover, one may observe a considerable amount of informal talk between people even in “formal” settings. We can propose several explanations for such surprising imbalance. First of all, it seems that informal language style and idle talk make the relationship between colleagues (even between chiefs and subordinates) more democratic and easier, helping to improve the general atmosphere in the office, where people know each other quite well3. The other explanation for the high share of informal style may be that only those respondents, who do not have strict regulations of personal behavior at their working places could agree to record in their offices and could obtain permission for filing from their colleagues and interlocutors. Thus, some respondents just refused to make recordings at work and preferred to record their day-off. However, the opposite case was also frequent; some respondents favored to record their working conversations but refused to register the private ones. Speaker’s social role variation. Speaker’s social role is a significant factor that influences speech. These data were obtained from the respondents’ diaries for the day of recording and the sociological questionnaire, in which the information concerning main interlocutors should include an indication of their social role about the respondent. When annotating macro episodes, we used the following detailed categorization for this parameter (Sherstinova 2015): 1. Close relatives: (a) husband, (b) wife, (c) father, (d) mother, (e) son, (f) daughter, (g)  sister, (h) brother, (i) grandmother, (j) grandfather, (k) grandson, (l) granddaughter. 2. Other relatives: (a) older relative (aunt, uncle, mother/father-in-law, godfather, etc.), (b) younger relative (son/daughter-in-law, nephew, niece, etc.), (c) relative of the same age (cousins, half cousins), (d) unidentified relative. 3. Friends: (a) male friend, (b) female friend, (c) romantic partner (boyfriend), (d) romantic partner (girlfriend).

3

Here, it should be mentioned that formal style is usually strictly maintained in the customer-service communication, where interlocutors are persons who have never met before.

22

Tatiana Sherstinova

4. Colleagues: (a) colleague, (b) boss or commander, (c) subordinate colleague, (d) colleague-friend. 5. “Customer-service” roles: (a) “customer” (client, patient, etc.), (b) “service staff ” (seller, waiter, repairer, nurse, etc.), (c) student, pupil, (d) teacher, lecturer, trainer, etc. 6. Group members: (a) classmate, schoolmate, (b) partner or companion in some activity (hobby, sports, social work, church, etc.), (c) casual companion, the client of the same cafe, shop or some other public place. 7. Others: (a) master (in communication with pets), (b) talking to themselves. In approximately a third of all communication episodes the respondents assumed more than one social role, because in many real-life settings, more than two interlocutors take part, and their roles may be different. For example, in domestic conversation, one person may be at the same time the father, the husband, the son, etc. When a friend joins the family gathering, the respondent takes on the role of the friend, too. Further, if a telephone or skype rings, the person could take on the role of the colleague, the client, the partner, etc. depending on who is calling. Thus, the variety of social roles may be quite wide irrespective of the setting. If considering only one-on-one communication episodes when the ORD respondents have a single social role, the most frequent roles are the following: colleague (18%), instructor/lecturer (5.6%), friend (female) (5.6%), service staff (5.2%), friend (male) (4.3%), husband (3.1%), customer (2.9%), colleague-friend (2.2%), and classmate (2.8%). Special attention should be paid to the category “Others” due to the unusual styles that frequently accompany it. The role master when talking to pets (usually, with dogs and cats) often implies “baby-talk” style of speaking4. The other interesting “role” appears when the person has no social roles at all, i.e. when he or she is talking to himself. Surprisingly, these are not rare cases as in the annotated part of the ORD corpus they comprise approximately 1.0% of all episodes. Our interest in this style is justified by the fact that we consider this language variant to be the best approximation to “the vernacular, in which the minimum attention is paid to speech” (Labov 1972, p. 112) and which, according to Labov, is “most regular in its structure and in its relation to the evolution of the language” (ibid.). Our own observations show that the talking-to-themselves register usually appears when respondents are relatively alone (at home or walking in the city) 4

The episodes with “non-standard” interlocutors including pets are described in (Ermolova 2015).

Studying Linguistic Variation and Communicative Diversity

23

though occasionally such fragments can occur even in the office when the respondent focuses on his own business. In many cases, such speech resembles inner talk or “stream of consciousness”. Let us now consider the variations in discourse settings. Scene of communication. For all ORD recordings we have information concerning the locus of communication, where the conversation took place. We distinguish the following common scenes of communication: (1) home that is a place of respondent’s permanent or temporal residence, (2) someone else’s home (formally a similar setting to the previous one, but the respondent is here as a guest or service staff), (3) office, (4) cafe (incl. restaurant, snack bar, canteen, pub, etc.), (5) store (incl. shopping center, supermarket, pharmacy, grocery, kiosk, etc.), (6) service center (commercial and government agencies, banks, studios, libraries, service stations, etc.), (7) educational institution, (8) medical center (hospital, first aid room, etc.), (9) military institution (military school, military barrack), (10) public institution or any other public place (museum, theater, cinema, gym, church, etc.), (11) outdoors (on the street, in the park, etc.), (12) public and private transport, (13) workshop (shed, garage, etc.), and (14) unidentified setting (Sherstinova 2015). Despite the fact that traditional sociolinguistics does not consider the place of communication to be a factor having evident influence on speech, we think that in some cases the external environment can matter. Sherstinova (2015) shows that the functional activity of words is different in different scenes. Therefore, the importance and the influence of this factor should be further investigated. Other register variation. More precise communication settings and circumstances are also specially marked in the process of macro episode annotating. In particular, we distinguish: (1) “kitchen” conversations (since the Soviet times, the kitchen is a traditional place for domestic conversations and discussions with family members and friends on various topics, including politics and state affairs), (2) holiday and party communication, (3) “drinking” conversations, (4) conversations, accompanying household or similar activities, (5) talking in motion, when walking or driving, (6) relaxation and leisure conversations, (7) communication by phone, etc. Additionally, the corpus contains information for particular communication phenomena which may be regarded as different styles/registers (e.g., isolated cues, monologues, reading aloud, reciting, singing, prayer, positive and negative emotions, conflicts, humor, etc. At present, the ORD corpus contains 230 (12%) “kitchen” conversations and 54 (3%) “drinking” episodes, phone conversations may be found in 420 (22%) episodes, positive emotions are present in 60 (3%) episodes, negative emotions – in

24

Tatiana Sherstinova

54 (3%) episodes, humor elements characterize 222 (12%) episodes (Sherstinova 2015). In most cases, communication situations presented in the ORD corpus appeared to be complex phenomena. As a result, most annotations include complex codes. Moreover, sometimes one may observe a mix of two general communication types within one episode. These situations are specially marked, too.

4. Some features of the ORD data that matter In this section we have to mention some features of the ORD corpus of which one should be mindful when conducting research using its data. First. The common feature of all spoken corpora is the evident fact that the spoken language differs greatly from the written language, and the transcription of recordings is required (Tagliamente 2006, p. 53). In relation to the ORD collection it implies the following: 1) Only a part of all recordings from this collection have been already transcribed (currently we have reached 1,000,000 words in transcripts). 2) The transcriptions are made based on the ORD conventions (orthographic spelling, the absence of punctuation marks, etc.) (Asinovsky et al. 2009). 3) It is not always possible to determine the exact borders between utterances of the same speaker (Ryko/Stepanova 2008). Second. Real-life conversations are “inherently overlapping” (Tagliamente 2006, p. 53). The situation becomes even worse when several conversations take place simultaneously in the same locus (e.g., in the office). Since multi-tiered transcription significantly complicates further tagging, for the basic ORD transcription we adhere to the rules of linear transcribing (Asinovsky et al. 2009). However, multi-tiered transcription is made at the later stage of linguistic and discourse annotation of recordings. Third. At the stage of basic annotation each utterance is coded with information on the corresponding speaker (i.e., the unique speaker’s code). It is based on this data, that it becomes possible “to build a bridge” between the utterance and the social characteristics of a speaker. In case of the overlapping speech, fragments of recordings (two or more overlapping turns) are left undivided, making it impossible to automatically separate the utterances pronounced by different speakers. However, this drawback is overcome by manually separating them and assigning to each turn the speakers’ codes. Fourth. The ORD recordings are made in natural settings, not in speech laboratories. Therefore, the technical quality of these recordings varies dramatically.

Studying Linguistic Variation and Communicative Diversity

25

Some of them are of good quality comparable to that obtained in professional studios, while others may be characterized by deficiencies of different kinds, e.g. (1) low level of speech, when a person is recorded by means of a remote voice recorder, (2) high level of background noise or (3) the fragments of overlapping speech. Because of these factors some pieces of speech become unintelligible or cannot be properly transcribed. These fragments are specially marked on the phrase-level (Asinovsky et al. 2009).

5. Studying lexical variation with the use of frequency word lists Frequency lists for ORD corpus Frequency is an inherent property of any real-life phenomena. Computing frequencies are actively used in linguistics, especially in corpus-based linguistic analysis. Frequency of any linguistic element (sound, word, grammatical form, etc.) can be an indicator of “markedness” (Baker, 2010, p. 125). Moreover, there is an opinion that “individual texts (authors, styles) can be better distinguished by means of some frequency indicators than by means of qualitative indicators” (Popescu 2009, p. 250). The first frequency list was compiled from the ORD data for the sub-corpus of approximately 205,000 words (33 hours of recordings) (Sherstinova 2010). It revealed the high number of discourse particles in the list of the most frequent Russian word forms. In Sherstinova (2013), one may find the frequency lists of the most common everyday Russian utterances. In a more recent paper (Sherstinova 2015), several high-frequency word lists referring to different communication situations have been compared. In the given paper, we will limit ourselves to the analysis of gender variation. Research data and method. For this exploratory study we have randomly selected 152 macro episodes (below called “episodes”) recorded by 40 participants (20 men and 20 women) and their interlocutors. Total duration of data is 2430 minutes (40.5 hours). The number of word forms in the sample is 232379. The mean number of words in an episode equals to 1550 (SD=1120). Language style variation in the sample is shown by the following distribution: informal conversations — 62.5% (95 episodes; 154051 tokens), professional conversations — 19.7% (30 episodes; 40012 tokens), customer service communication — 9.9% (15 episodes; 21126 tokens), educational communication — 7.2% (11 episodes; 19629 tokens), public speeches — 0.7% (1 episode; 800 tokens).

26

Tatiana Sherstinova

This distribution is relatively consistent with that in the ORD corpus calculated for 1854 macro episodes or 483 hours of speech (Sherstinova 2015). Transcripts of spoken conversations that were manually made in linguistic annotator ELAN (Hellwig et al. 2015) are exported into database format together with information on speakers. In the database, these data are merged with macro episode annotation (see above), resulting in words having the following description: (1) word unique identifier (serial number); (2) code of the episode; (3) the speaker’s code; (4) text context; (5) type of communication referring to language style; (6) social role of the speaker; (7) locus of communication; (8) other stylistic markers when relevant (e.g., a monologue, an isolated cue; “kitchen” conversation; talk in motion; “drinking” conversation; reciting; conflict; humor; etc.); and (9) textual description of discourse situation. These data are sufficient to compile frequency dictionaries for different language styles and other discourse settings (e.g., locus, social role of speakers, particular stylistic markers, etc.). If we are interested in studying social variation of speech, we have recourse to social data on respondents and their interlocutors, which are kept in the corpus as well. Since the linguistic annotating of the ORD corpus including the lemmatization of words has not been finished yet, in this research, we will analyze the frequency distribution of word forms, not lemmas. Such an approach is used fairly often in quantitative lexical studies (e.g., Rayson et al. 1997, Popescu 2008). The upper zone for the overall frequency list consisting of 50 words is given in Table 1. Each word in this list is accompanied by numeric information: rank, absolute frequency in the given subset, its percentage of the whole and a running cumulative percentage of each word and all the preceding words. For example, Table 1 shows that the 10 most frequently used words (ja, vot, nu, ne, da, a, i, čto, v, ėto) cover approximately one fifth (20,69%) of all spoken communication. As can be seen, in everyday conversations the most commonly-used words are personal pronouns (mainly, ja (I), ty (you), on (he), ona (she), oni (they), my (we), vy (you) in Nom. case, mne (me, in Dat.), menja (me in Acc./Gen.) and the particles, including the so-called discourse markers. The actual meaning of many words on this list differs in real-life conversations from their traditional definition given in the dictionaries (vot, nu, ne, da, net, tak, prosto, voobšče, značit, sejčas, etc.). Many of these words assume pragmatic functions or are used to regulate discourse. Besides, here we have a typical hesitation filler (ė) and one more discourse marker ugu (yeah). As always in frequency lists, a number of auxiliary words may be observed at the top of this list: conjunctions i (and), no (but), esli (if), prepositions v (in), u (at), na (on), c (with), k (to), and the modal verb nado (must). Word frequency data is also affected by clusters of homonymous forms which cannot be automatically attributed

Studying Linguistic Variation and Communicative Diversity

27

without referring to context (e.g., est’ may be a form of the common verbs to be or to have, an infinitive for to eat, a part of conjunction to est’, etc.). Table 1: The 50 most frequently used words in everyday Russian speech Rank Word 1 ja

Count 5950

% 2.56

Cum.% 2.56

Rank Word 26 est’

Count 1256

% 0.54

Cum.% 34.32

2

vot

5411

2.33

4.89

27

menja

1124

0.48

34.81

3

nu

5407

2.33

7.22

28

sejčas

1117

0.48

35.29

4

ne

5400

2.32

9.54

29

oni

1060

0.46

35.74

5

da

5162

2.22

11.76

30

my

1050

0.45

36.19

6

a

4523

1.95

13.71

31

by

1002

0.43

36.63

7

i

4190

1.80

15.51

32

no

989

0.43

37.05

8

čto

4139

1.78

17.29

33

uže

984

0.42

37.47

9

v

4003

1.72

19.01

34

nado

953

0.41

37.88

10

ėto

3905

1.68

20.69

35

eščë

938

0.40

38.29

11

tam

3422

1.47

22.17

36

že

914

0.39

38.68

12

u

3007

1.29

23.46

37

po

889

0.38

39.06

13

tak

2842

1.22

24.68

38

prosto

875

0.38

39.44

14

na

2602

1.12

25.80

39

voobšče

772

0.33

39.77

15

kak

2110

0.91

26.71

40

esli

721

0.31

40.08

16

ty

1877

0.81

27.52

41

vy

712

0.31

40.39

17

vsë

1874

0.81

28.33

42

nas

691

0.30

40.69

18

to

1781

0.77

29.09

43

tože

665

0.29

40.97

19

s

1771

0.76

29.85

44

znaju

664

0.29

41.26

20

net

1727

0.74

30.60

45

bylo

592

0.25

41.51

21

(ė)

1708

0.74

31.33

46

značit

584

0.25

41.76

22

on

1670

0.72

32.05

47

zdes’

577

0.25

42.01

23

ugu

1451

0.62

32.68

48

budet

570

0.25

42.26

24

mne

1293

0.56

33.23

49

k

556

0.24

42.50

25

ona

1275

0.55

33.78

50

tut

555

0.24

42.74

Let us consider, how this list changes, if women’s and male’s utterances are treated separately. Having consecutively applied filters F(emale) and M(ale) to the multitude of words in the exploratory subset, we obtain two new lists (see Tables 2 and 3) referring to the upper zones of the most frequent words for female and male speakers respectively and therefore revealing gender preferences.

28

Tatiana Sherstinova

The first thing that stands out is the difference in the first-rank words of these lists. The personal pronoun ja (I), ranked first in female everyday speech, in male speech gives way to the particle nu (well). As was shown in Sherstinova (2015), this particle is preferably used in informal style (especially in husband-and-wife communications or between close friends) as opposed to vot used in formal settings. Then, it can be seen that women pay more attention to maintaining the conversation (ugu, chorošo), they use fewer hesitations (ė), and are more inclined to use intensifying words očen’ (very), emotional words and interjections (oj). Table 2: The 50 most frequent words in female speech Rank Word 1 ja

Count 3110

% 2.70

Cum.% Rank Word 2.70 26 my

Count 651

% 0.57

Cum.% 34.57

2

ne

2722

2.36

5.07

27

sejčas

571

0.50

35.07

3

vot

2691

2.34

7.40

28

menja

568

0.49

35.56

4

da

2593

2.25

9.66

29

est’

564

0.49

36.05

5

nu

2577

2.24

11.90

30

by

561

0.49

36.54

6

čto

2218

1.93

13.82

31

no

523

0.45

37.00

7

a

2175

1.89

15.71

32

oni

518

0.45

37.45

8

i

2116

1.84

17.55

33

nado

486

0.42

37.87

9

ėto

1925

1.67

19.22

34

eščë

477

0.41

38.28

10

v

1770

1.54

20.76

35

uže

469

0.41

38.69

11

tak

1619

1.41

22.17

36

že

452

0.39

39.08

12

u

1569

1.36

23.53

37

po

430

0.37

39.46

13

tam

1490

1.29

24.82

38

voobšče

423

0.37

39.82

14

na

1176

1.02

25.85

39

vy

423

0.37

40.19

15

kak

1140

0.99

26.84

40

prosto

407

0.35

40.54

16

ty

1004

0.87

27.71

41

nas

395

0.34

40.89

17

vsë

932

0.81

28.52

42

tože

364

0.32

41.20

18

s

913

0.79

29.31

43

znaju

352

0.31

41.51

19

ugu

836

0.73

30.04

44

očen’

342

0.30

41.81

20

to

826

0.72

30.75

45

esli

326

0.28

42.09

21

net

798

0.69

31.45

46

k

298

0.26

42.35

22

ona

783

0.68

32.13

47

budet

294

0.26

42.60

23

mne

749

0.65

32.78

48

chorošo

292

0.25

42.86

24

on

739

0.64

33.42

49

bylo

288

0.25

43.11

25

(ė)

677

0.59

34.01

50

oj

278

0.24

43.35

Studying Linguistic Variation and Communicative Diversity

29

Table 3: The 50 most frequent words in male speech Rank Word

Count

%

Count

%

Cum.%

1

nu

2062

2.47

Cum.% 2.47

Rank Word 26

menja

393

0.47

33.89

2

ja

2000

2.40

4.87

27

oni

387

0.46

34.35

3

ne

1898

2.28

7.15

28

mne

356

0.43

34.78

4

vot

1887

2.26

9.41

29

uže

344

0.41

35.19

5

v

1669

2.00

11.41

30

prosto

331

0.40

35.59

6

da

1610

1.93

13.35

31

eščë

330

0.40

35.98

7

a

1548

1.86

15.20

32

no

325

0.39

36.37

8

i

1530

1.84

17.04

33

nado

320

0.38

36.76

9

tam

1456

1.75

18.78

34

že

319

0.38

37.14

10

ėto

1338

1.60

20.39

35

ona

318

0.38

37.52

11

čto

1335

1.60

21.99

36

po

315

0.38

37.90

12

na

1113

1.33

23.32

37

značit

294

0.35

38.25

13

u

985

1.18

24.51

38

ugu

291

0.35

38.60

14

tak

864

1.04

25.54

39

blin

283

0.34

38.94

15

(ė)

761

0.91

26.46

40

esli

281

0.34

39.28

16

on

696

0.83

27.29

41

by

274

0.33

39.61

17

to

680

0.82

28.11

42

voobšče

269

0.32

39.93

18

vsë

665

0.80

28.90

43

znaju

248

0.30

40.23

19

kak

656

0.79

29.69

44

my

228

0.27

40.50

20

ty

651

0.78

30.47

45

tut

218

0.26

40.76

21

s

609

0.73

31.20

46

govorit

218

0.26

41.02

22

net

537

0.64

31.85

47

zdes’

215

0.26

41.28

23

est’

478

0.57

32.42

48

za

213

0.26

41.54

24

bljad’

422

0.51

32.93

49

kogda

210

0.25

41.79

25

sejčas

410

0.49

33.42

50

tebe

208

0.25

42.04

Men use fewer personal pronouns. Moreover, the ratios of specific pronoun use are also different. If in female speech, we observe practically no difference between the frequencies of on (he) and ona (she), in men’s address the “male” pronoun on (he) is used twice as frequently as compared to ona (she). Finally, as shown more than once by different researchers (Lakoff 1975, Stenström 1991,

30

Tatiana Sherstinova

Rayson et al. 1997, etc.), the expletives are among the most frequent words used by men in everyday conversations. For Russian, in the top list one may find two swear words — bljad’ and its euphemism blin5. A marker of logical conclusion značit ((it) means, therefore) also appears in the male list. In general, these observations are consistent with those described earlier for gender variation in other countries and for other languages (Tannen 1991, Tannen 1994, Hirschman 1994, Coupland 2007, Schmid 2003, Romaine 2008, Potapova/ Potapov 2006, etc.). To demonstrate the possibilities of the ORD corpus, we give here two more tables, containing four lists each, which allow the comparison of women’s and men’s most frequent words for different styles (see Tables four and five correspondently). We present here just the 25 most frequent words in each column, due to space limitations. The leader of the corpus — personal pronoun ja (I) — maintains the highest ranking in informal and professional female conversations. It has the second rank (after vot) in female customer-service communication and yields its leadership only in feminine educational style. In men’s language, the same pronoun most frequently appears in informal talks (moreover, if one separately considers only those episodes in which a man is talking only with his friends, ja ranks first, just as in women’s speech), in other male spoken styles it occupies a stable position of 3rd–4th ranking. Regarding personal pronouns in general, their top frequency lists vary depending on style. Thus, in informal female form the group of individual pronouns is the most numerous, consisting of 7 words (ja, ty, on, mne, ona, my, oni), in informal male style, this list is more concise (ja, ty, on, oni). In professional conversations, women most frequently use ja, ona, mne, ty, while men prefer ja, menja, on, oni. For educational style women use more plural and polite forms (ja, my, vy, nas, vas), and men continue to use the “standard set” of pronouns, consisting of ja, on, ty. Finally, in customer-service communication, women confine themselves to ja, vy, vam, while men prefer to use ja and on. Thus, our observations stress once more the fact that women in conversation are more focused on their interlocutors than men are, mainly if they are speaking with their clients or students.

5

The latter, however, can take on different functions (Bogdanova-Beglarian 2014).

Studying Linguistic Variation and Communicative Diversity

31

Table 4: The 25 most frequent female words depending on language style Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Informal Word ja ne nu da a vot čto i ėto v tam tak u na ty kak s on net vsë mne ona ugu to my

Professional Educational Client-service % Rank Word % Rank Word % Rank Word % 2.87 1 ja 2.92 1 da 3.35 1 vot 3.01 2.61 2 vot 2.59 2 vot 3.05 2 ja 2.28 2.46 3 čto 2.27 3 nu 1.85 3 ne 2.19 2.23 4 ne 2.21 4 tak 1.80 4 i 1.97 2.17 5 i 2.11 5 ėto 1.63 5 da 1.97 2.09 6 da 1.96 6 čto 1.60 6 ėto 1.95 1.88 7 nu 1.87 7 u 1.43 7 a 1.87 1.81 8 v 1.75 8 i 1.31 8 nu 1.81 1.66 9 ėto 1.65 9 ja 1.24 9 čto 1.59 1.57 10 a 1.44 10 ne 1.14 10 ugu 1.52 1.40 11 tam 1.34 11 my 1.06 11 v 1.43 1.40 12 tak 1.30 12 vy 1.01 12 u 1.36 1.40 13 u 1.23 13 ugu 1.01 13 tak 1.29 1.16 14 (ė) 1.13 14 a 1.00 14 vsë 1.10 1.08 15 to 1.05 15 vsë 0.92 15 kak 1.10 1.02 16 vsë 1.01 16 tam 0.89 16 to 0.98 0.83 17 kak 0.92 17 po 0.89 17 est’ 0.94 0.74 18 ona 0.90 18 kak 0.86 18 na 0.88 0.73 19 na 0.82 19 v 0.84 19 vam 0.85 0.70 20 s 0.78 20 (ė) 0.70 20 s 0.78 0.68 21 est’ 0.74 21 na 0.65 21 ona 0.76 0.67 22 mne 0.73 22 s 0.57 22 vy 0.75 0.63 23 ugu 0.66 23 nas 0.56 23 net 0.73 0.63 24 net 0.65 24 sejčas 0.51 24 sejčas 0.72 0.57 25 ty 0.57 25 vas 0.50 25 by 0.67

Table 5: The 25 most frequent male words depending on language style Informal Professional Educational Client-service Rank Word % Rank Word % Rank Word % Rank Word % 1 nu 2.53 1 vot 2.77 1 ėto 4.15 1 vot 2.89 2

ja

2.35

2

nu

2.74

2

vot

3.29

2

nu

2.54

3

ne

2.32

3

da

2.57

3

ja

3.16

3

v

2.18

4

v

2.12

4

ja

2.49

4

da

2.44

4

ja

2.01

32

Tatiana Sherstinova

Informal Professional Educational Client-service Rank Word % Rank Word % Rank Word % Rank Word % 5 vot 2.05 5 a 2.42 5 ne 2.11 5 tam 1.95 6

tam

1.87

6

ne

2.33

6

a

2.00

6

da

1.89

7

da

1.81

7

i

1.87

7

(ė)

1.94

7

ne

1.81

8

i

1.81

8

v

1.72

8

i

1.87

8

(ė)

1.75

9

a

1.77

9

u

1.65

9

čto

1.66

9

čto

1.72

10

čto

1.61

10

ėto

1.46

10

nu

1.45

10

ėto

1.70

11

na

1.43

11

čto

1.43

11

vsë

1.16

11

a

1.68

12

ėto

1.40

12

tam

1.41

12

net

1.12

12

u

1.56

13

u

1.12

13

tak

1.32

13

na

1.10

13

i

1.56

14

tak

0.97

14

na

1.13

14

tak

1.05

14

tak

1.43

15

ty

0.91

15

kak

1.05

15

tam

1.01

15

vsë

1.04

16

on

0.87

16

to

1.04

16

est’

0.97

16

ugu

1.04

17

(ė)

0.78

17

vsë

0.82

17

on

0.95

17

na

0.94

18

kak

0.77

18

(ė)

0.69

18

v

0.93

18

net

0.94

19

to

0.76

19

menja 0.69

19

to

0.91

19

to

0.94

20

s

0.75

20

s

0.60

20

nado

0.84

20

s

0.89

21

vsë

0.75

21

sejčas

0.58

21

zdes’

0.78

21

on

0.87

22

bljad’

0.67

22

net

0.58

22

ponjatno 0.78

22

kak

0.73

23

net

0.60

23

on

0.57

23

sejčas

0.76

23

est’

0.67

24

est’

0.56

24

oni

0.53

24

u

0.69

24

sejčas 0.62

25

oni

0.48

25

ugu

0.51

25

kak

0.63

25

nado

0.54

Both men and women use the particle vot in official styles rather than in informal conversations, supporting our hypothesis concerning the “official tint” of this discourse marker (Sherstinova 2015). On the other hand, two negative particles are more common in informal style than in official ones both for men and women, and in all cases ne is relatively more frequent than net. The usage of another leading particle nu is more complicated. In men’s speech, it ranks first or second except for educational conversations, when its frequency falls to 10th place. In female speech, its frequency variation between styles is expressed more explicitly, and in general, women use this particle less frequently than men. The affirmative particle da (yes) has comparatively similar frequencies (ranked 3rd to 7th) for different styles, both for men’s and women’s speech. However, it unexpectedly ranked first for the female educational style.

Studying Linguistic Variation and Communicative Diversity

33

The discourse marker ugu, which is used to maintain the conversation, is more common in female spoken communication with clients and in education style. In informal and professional women’s speech, ugu is used less. As for male conversations, this discourse marker is observed among the 25 most frequent words twice — for professional and customer-service communication. Hesitation fillers (ė) occur more frequently in the female professional style and in two male styles (educational and customer service). The word sejčas is less frequent in informal style than in the official ones for both male and female conversations. Finally, we would like to mention that men’s educational style introduces three “new” words in the top list, namely nado (must), ponjatno (clear, I see), zdes’ (here), revealing lexical preferences in the male educational style. As for female speech in similar conditions, no special educational words were seen in the upper zone of the corresponding frequency list.

6. Conclusions Evidently, analyzing real-life discourse and studying its variation is an extremely complex and challenging issue. It is clear that any actual conversation is unique in its way regarding particular circumstances, participants, their social characteristics and backgrounds, duration and intensity of conversation, its major and minor topics, the sequence of utterances, their vocabulary, etc. However, the researcher’s aim should be to look for general rules and patterns. In this aspect, a corpusbased approach and quantitative methods (in particular compiling frequency lists) seem to be entirely appropriate, allowing the discovery of many general and typical tendencies. In the presented exploratory research, it was shown that by applying standard filters to the ORD corpus data, one could create a subcorpus of speech of particular styles and registers which can be used to shed light on many topics of sociolinguistic interest. Even with a brief analysis of gender, we analyzed only the 25–50 most frequent words from each list; we were able to show significant differences in approaches to speech in men and women. When we increase the number of words under consideration, we will have a still clearer understanding of the situation. The inclusion of other relevant factors, such as speaker’s age, education, social group, profession, and their social role in the conversation, into the analysis is also required. In addition to the qualitative comparison of frequency lists, it is necessary to conduct multivariate statistical analysis in order to determine the role of particular factors and to use particular statistical criteria to test the significance of the observed variation for different language styles and registers.

34

Tatiana Sherstinova

Acknowledgments Sociolinguistic extension of the ORD corpus and the presented frequency lists were obtained within the framework of the project “Everyday Russian Language in Different Social Groups” supported by the Russian Science Foundation, project # 1418-02070.

References Asinovsky, Alexander / Bogdanova, Natalia / Rusakova, Marina / Ryko, Anastasija / Stepanova, Svetlana / Sherstinova, Tatiana: “The ORD Speech Corpus of Russian Everyday Communication “One Speaker’s Day”: Creation Principles and Annotation”. In: Matoušek, V., Mautner, P. (eds.): TSD 2009. LNAI 5729, 2009, pp. 250–257. Baeva, Ekaterina: “O sposobach sociolingvističeskoj balansirovki ustnogo korpusa (na primere “Odnogo rečevogo dnja”)”, Vestnik Permskogo universiteta. Rossijskaja i zarubežnaja filologia 4 (28), 2014, pp. 48–57. Baker, Paul: Sociolinguistics and Corpus Linguistics. Edinburgh University Press: Edinburgh 2010. Bogdanova-Beglarian, Natalia / Martynenko, Gregory / Sherstinova, Tatiana: “The “One Day of Speech” Corpus: Phonetic and Syntactic Studies of Everyday Spoken Russian”. In: Ronzhin, A. et al. (eds.): SPECOM 2015, Lecture Notes in Artificial Intelligence, LNAI 9319. 2015, pp. 429–437. (2015а). Bogdanova-Beglarian, Natalia / Sherstinova, Tatiana / Blinova, Olga / Baeva, Ekaterina / Ermolova, Olga / Ryko, Anastasia: “Zvukovoj korpus russkogo jazyka kak baza dlja sociolingvističeskich issledovanij”. Proc. of the Int. Conference “Kompjuternaja lingvistika i jazykovye tehnologii”. Available at: < http://www.dialog-21. ru/digest/2015/?type=conference>. Retrieved: February 2, 2016. (2015b).  Bogdanova-Beglarian, Natalia: “Ob odnoj iz samych častych edinic russkoj spontannoj reči: blin s lingvističeskoj i sociolingvističeskoj toček zrenija”. In: V. P. Selegej (ed.): Kompjuternaia lingvistika i intellektual’nye technologii: Trudy meždunarodnoj konferencii «Dialog 2014» [Computational Linguistics and Intelligent Technologies: Proceedings of the International Conference «Dialog 2014»]. Iss. 13/20. RSHU Publ.: Moscow. 2014, pp. 76–82. Burnard, Lou (ed.): Reference guide for the British National Corpus (XML edition). Published for the British National Corpus Consortium by Oxford University Computing Services, 2007. Available at: . Retrieved: February 2, 2016. Campbell, Nick: “Speech & Expression; the Value of a Longitudinal Corpus”. LREC 2004. 2004, pp. 183–186.

Studying Linguistic Variation and Communicative Diversity

35

Coupland, Nikolas: Style: Language Variation and Identity. Cambridge University Press: Cambridge 2007. Ermolova, Olga B.: “Lingvističeskie osobennosti obščenija čeloveka s domašnimi životnymi”. Vestnik Permskogo universiteta. Rossijskaja i zarubezhnaja filologia 4 (32), 2015, pp. 58–66. Hellwig, Birgit / Van Uytvanck, Dieter / Hulsbosch, Micha, et al.: ELAN — Linguistic Annotator. Version 4.9.2, retrieved 2.2.2016, from http://www.mpi.nl/ corpus/html/elan/. Hirschman, Lynette: “Female-male differences in conversational interaction”. Language in Society (23) 1994, pp. 427–442. Krysin, Leonid P.: Jazyk v sovremennom obščestve. Russkoe slovo: Moscow 2012. Labov, William: “Some principles of linguistic methodology”. Lang. Soc. I. 1972, p. 97–120. Lakoff, Robin: Language and Woman’s Place. Harper: New York 1975. Meyerhoff, Miriam: Introducing Sociolinguistics. Routledge: London & New York 2006. Ottenheimer, Harriet J./ Pine, Judith: The Anthropology of Language: An Introduction to Linguistic, Anthropology. Wadsworth Publ.: Belmont, 4 edition 2018. Popescu, Ioan-Iovitz: Word Frequency Studies. (Quantitative Linguistics Series. [QL] 64). Mouton de Gruyter: Berlin 2009. Potapova, Rodmonga / Potapov, Vsevolod: Jazyk, reč’, ličnost’. Jazyki slavjanskoj kul’tury: Moscow 2006. Rayson, Paul / Leech, Geoffrey / Hodges, Mary: “Social Differentiation in the Use of English Vocabulary: Some Analyses of the Conversational Component of the British National Corpus”. International Journal of Corpus Linguistics 2(1) 1997, pp. 133–152. Romaine, Suzanne: “Corpus linguistics and sociolinguistics”. In: Lüdeling, Anke / Kytö, Merja (eds.): Corpus Linguistics: An International Handbook. Vol. 1 Mouton de Gruyter: Berlin-New York 2008, pp. 96–111. Ryko, Anastasia I. / Stepanova, Svetlana B.: “Strategii členenija spontannoj reči na sintaksičeskie edinicy”. In: Kompjuternaia lingvistika i intellektual’nye technologii: Trudy meždunarodnoj konferencii «Dialog 2009» [Computational Linguistics and Intelligent Technologies: Proceedings of the International Conference «Dialog 2009»] 8(15). RGGU Publ.: Moscow 2009, pp. 438–443. Schmid, Hans-Jörg: “Do men and women really live in different cultures? Evidence from the BNC”. In: Wilson, Andrew / Rayson, Paul / McEnery, Tony (eds.). Corpus Linguistics by the Lune: A Festschrift for Geoffrey Leech. Peter Lang: Frankfurt 2003, pp. 185–221.

36

Tatiana Sherstinova

Sherstinova, Tatiana: “Macro Episodes of Russian Everyday Oral Communication: Towards Pragmatic Annotation of the ORD Speech Corpus”. In: Ronzhin, Andrej et al. (eds.): SPECOM 2015, Lecture Notes in Artificial Intelligence, LNAI 9319. 2015, pp. 268–276. Sherstinova, Tatiana: “Quantitative Data Processing in the ORD Speech Corpus of Russian Everyday Communication”. In: Grzybek, Peter / Kelih, Emmerich / Mačutek, Ján (eds.): Text and Language: Structures, Functions, Interrelations. Praesens Verlag: Wien 2010, pp. 195–206. Sherstinova, Tatiana: “Russian Everyday Utterances: the Top Lists and Some Statistics”. In: Thielemann, Nadine / Kosta, Peter (eds.): Approaches to Slavic Interaction (Dialogue Studies, vol. 20). John Benjamins: Amsterdam/Philadelphia 2013, pp. 105–116. Stenström, Anna-Brita: “Expletives in the London-Lund Corpus”. In: Aijmer, Karin / Altenberg, Bengt (eds.): English Corpus Linguistics in Honour of Jan Svartvik. Longman: London 1991, pp. 230–253. Tagliamonte, Sali A.: Analysing Sociolinguistic Variation. Cambridge University Press: Cambridge 2006.  Tannen, Deborah: Gender and Discourse. Oxford University Press: New York, Oxford 1994. Tannen, Deborah: You Just Don’t Understand: Women and Men in Conversation. Ballantine Books: New York 1991.

Christian Sappok

Addressing by Voice: Deixis and Design1 Abstract: We follow the definition of voice given by Trask: “The natural and distinctive tone of the speech sounds produced by a particular person (1996, 378)”, extending his concept of distinctness: The tone of speech is characteristic of an individual speaker and helps a listener to distinguish this speaker from other speakers. In addition, the speaker’s distinct tone shows significant fluctuations, subject to discourse conditions. Listening to a familiar person talking over the telephone to someone absent, we recognize not only a speaker himself, but draw conclusions from his voice as to the nature of his addressee. The data collected in the ORD project reveals rich fluctuation of discourse situations and a wide range of addressing techniques. The object of our analysis is the communicative behavior of two women, S24 and S05, in comparable situations, the bubble of voices between university teachers before a lesson, and the same two women in isolation involved in narration or in soliloquy. An interesting difference in the use of voice parameters can be observed: S24 organizes her addressees in concentric circles where the inner circle, the origo (Bühler 1934), is opposed to an individual or, more generally, to a collective audience. This is the deictic model of voice performance. S05 is more flexible, addressing her superior, a group, a female and a male colleague, but, within the shortest time interval, changing her voice and the nature of this partnership. This initiative modeling of the discursive roles will be described following the proposal of Bell 1984, as audience designing by using voice parameters. Since S24, in her soliloquy addresses her pets and her computer with much the same techniques of type design, a model is proposed that treats this type of design as “imagination oriented deixis” following Goodwin’s 1990 translation of Bühler’s “Deixis am Phantasma”.

1. The role of voice in a multi-source situation In his Theory of Language (Sprachtheorie, 1) 1934, Karl Bühler reports an observation he made with a blind colleague of his: “… nämlich dass er sich im geselligen Kreise, wo die Rede unregelmäßig kreuz und quer von einem zum anderen geht, mit großer Regelmäßigkeit angesprochen fühlt, wenn immer sich irgendeiner speziell an ihn wendet. Versteht sich ohne die Hilfe eines namentlichen oder sonst sprachlich nur auf ihn gemünzten Appells, sondern rein stimmlich“ (i.e. a person feels addressed with great regularity and speech runs irregularly from one to the 1

I am grateful to Alexander and Georgy Krasovitsky for their help in preparing this text for publication.

38

Christian Sappok

other, whenever anyone addresses him specifically) (Bühler 1934, p. 98). Bühler explains this as a mere physical phenomenon, i.e. the directedness of acoustic waves that bears the information of being addressed to the ear of the blind with his trained ear. It will be shown that this phenomenon can be observed in other situations as well, without the preconditions mentioned above: – It is not only the physical properties of directed waves that inform the speaker that he is addressed. There are other properties of voice involved that fulfill the function of directedness. – Addressing functions do not only concern the addressee in his physical presence. They concern the position (both real and constructed in the sense of Harré and van Langenhove1991) and the role of the speaker and hearer involved in spoken discourse. – We do not have a fixed inventory of voice parameters. So we have to rely on perceptual methods to detect them, and this is the best that can be done in cases when the place or the position of addressing changes.

1.1. The linguistic status of voice parameters and their variation Voice parameters are not usually included into the kernel of linguistic categories, being classed as paralinguistic. This is due to the fact that there is no clear-cut set of distinctive voice features (for an overview cf. Eckert and Laver 1994) so we don’t know neither their relationship to one another nor their range of variation or their functional load depending on distinctiveness. Their functioning in oral communication seems to be a subjective matter. Kreiman and Sadtis formulate this idea as follows: “Speakers’ intentions and attitudes within the social context of talk, woven throughout the linguistic prosodic contour of talk, are represented and delivered using kaleidoscopic nuances of voice quality (2013, 301)”. The authors are motivated by the idea that being paralinguistic doesn’t diminish the importance and value of voice parameters. These parameters being inherently characteristic of a speaker continuously build up a major component of his personal identity (this is also a matter of distinctness, as formulated by Trask 1996: p. 378). They “form a characteristic auditory coloring of an individual’s voice” (compare Uni StuttgartProject on Voice Quality, http://www2.ims.uni-stuttgart. de/EGG/page2.htm) where coloring means a fluctuating spectrum of variation within the framework of a constant individual voice. It is this fringe between constant and variable parameters that we are going to focus on in our observation of oral communication. We begin with a perceptive analysis of recordings of speakers we did not know before, neither in person

Addressing by Voice: Deixis and Design

39

nor in voice. We become acquainted with their individual voices so that we can distinguish one speaker from another (cf. problems of speaker recognition in the sense of O’Shaughnessy, 2000, § 11, and the risky use of recognition techniques for forensic aims in Viciano et al., 2016; Bonastre et al. We gradually become aware of the variation in each speaker’s voice and develop an intuition regarding the function of this variation in the vivid course of verbal interaction. What do we expect to find in the realm of the interactional fluctuation of voice parameters? There are expectations formulated by Kreiman and Sadtis 2013, who analyze voice as a vehicle for emotions, attitudes and other components of speaker’s shaping his message in addition to what is expressed verbally. We will start from a different perspective. In oral communication, every participant has voice, his voice makes his message to be understood, supports the identification of the speaker, and variation within his voice parameters may modify his message or his character as a participant and source of the message. Listening to this oral interaction often creates an impression that speaker’s voice characterizes not only the speaker himself as a source, but also varies with respect to the person being addressed. Bühler speaks about the source qualities of voice (defined as “Herkunftsqualität”, cf. 1934, p. 91 et passim), but he also puts forward the notion of the target character (“Zielcharakter“, ibid., p. 98). This aspect of voice functions will be the main focus of our investigation.

1.2. The data from the St Petersburg ORD project What kind of data, what type of recordings will provide us with material suitable for the analysis of the addressing functions of voice? Preferably, it should be communication in a multi-source situation where every participant hears what is said aloud but has a choice as to whom to pay attention and to whom to address. The recordings made within the ORD-project at the State University of Saint Petersburg provide rich data of this type (Asinovsky et al. 2009)). The ORD-project aims to collect lengthy recordings of Russian speakers living in Saint Petersburg. These speakers were equipped with a portable device which recorded all they said in the course of the day along with all what was said around them within the reach of a microphone and with all non-speech noises. This latter fact has to be accepted, although whoever is working with spoken language would be eager to avoid these noisy circumstances. However the study of addressing functions and of voice variation could not find a more ideal field of observation. We pick out two female speakers, coded S24 and S05 in the ORD project, in a comparable situation that allows for comparison of their voice performance within a group of other speakers.

40

Christian Sappok

The multi-source communication in which several acoustic sources, both human and non-human, are involved, will be given the name of kafedra-talk. This means that the persons are assembled in a single room at a university institution, where staff members prepare to leave for teaching sessions held by everybody in the room. In a short stretch of time, a couple of problems have to be solved: Where can I print a handout? Where are the textbooks to be given to the students? What material was covered during the previous session, where from the next session has to begin? Obviously, in both of the kafedra situations, there is coordinate teaching and the roles of both teachers have to be determined. This includes the questions such as which one of the two teachers is to be prepared and has to prepare the students for the final exam? And what should be said to a visiting person who is coming after the teaching session? In a situation like this, stress is predictable. And this stress shapes the communicative situation along with other factors. Not only the teaching staff is affected by stress but also the linguist who is analyzing the resulting bubble. And it could be expected that stress will also affect the readers of this report. They will seldom find a clear succession of complete sentences, a concise narration, a coherent text as may be found in canonical text collections and descriptions. But the result of this laborious work provides an interesting insight into a specific and seldom observed discourse situation in which voice is a crucial factor. Without carefully attending to voice parameters, all the participants would be lost in an acoustic chaos impenetrable not only to those involved in the communication but also to the linguist reconstructing what was going on. Any linguist willing to spare time and energy may investigate this bubble and find an answer to the following questions: What is the role of voice parameters? Where in the text do they play an essential role? What are the methods that allow to bring them to light? Finally, the key question: What is the status of voice parameters in a general model of speech communication? The comparison of various cases of multi-source dialogue communication with its monolog counterpart reveals a rather surprising phenomenon: A good deal of what we observe in the first case will be also in the latter case, even in a standard distribution of speaker and hearer, but also in addressing speechless objects as computer and animal.

2. The deictic model of addressing S24, the head of the department, is the only speaker in fragment 1.1. She brought with her a pack of paper which she offers for common use to everybody within the group of staff members. She waits for a reaction and, as no one reacts, she is uncertain of how she can get rid of the paper which she carries.

41

Addressing by Voice: Deixis and Design

1.1. S24 – levels of addressing in the institute. S24

01 02 03 04 05

Ага … так … сейчас я … так … Так вот это я бумагу принесла, всем да. Пожалуйста! Хорошая бумага, её можно использовать для принтера, 06 Марин, слышите? 07 Да, я не знаю, куда. Я и не знаю, стоит ли сюда класть …

Aha … . well … now I … well… So, here I brought some paper, For all of you. Please! The paper is good, you can use it for the printer. Marina, do you hear me? Yes, I don’t know: where, I’m not sure if it’s ok to put here…

Figure 1: The waveform of 1.1. Numbers in the segmentation correspond to the numbers in text 1.1.

01 0

02

03

04 Time (s)

05

06

07 29.62

Figure 1 shows the waveform of the whole sequence split into short segments. These segments co-occur with a number of intervening non-speech acoustic events. The transcription above depicts only the verbal material which seems to be a continuous flow of a coherent underlying text. Segments 01 and 02 are realized in a kind of murmur while the speaker is handling various objects, e.g. a briefcase. At the beginning of 03 the murmur develops into a clear and intelligible speaking style as the speaker addresses a group of people showing or failing to show attention. This style is interrupted by 06 being a question directed to an individual member of this group. S24 exhibits a threefold change in her addressing voice: segments 01 and 02 are directed to herself, 06 to an individual partner using her name, and the rest to a collective addressee, i.e. to all people present within the acoustic reach. This behavior can be depicted in a three-level model of addressing, self – single – collective. The first level (‘self’) reminds us of what is said by Nikitina 2011 about addressing oneself, with the loudness parameter being a crucial indication of this type of speech. The difference between level two and level three is not the increase

42

Christian Sappok

in loudness, but this difference is audible, and this calls up the observation by Bühler concerning a blind member of his working group who in the overall bubble could easily detect an utterance addressed to him (1934, p. 98). Is it the voice that helps the blind man to distinguish between different addressees in Bühler’s group or is it the ability of S24 to discern level two and level three while producing or receiving the direct addressing in a multi-voice communication? There is a method which allows us to test whether our assessment of the deictic change was correct: Examples 1.1a show the result of an alternative segmentation as compared to 1.1, combining segment 03 with segment 04 and segment 06 with the beginning of 07. 1.2. – a fake combination of turns. 01/02 06/07

*Так … Сейчас я … Так вот это я бумагу принесла, Марин, слышите? Да, я не знаю: куда?

Ah … well … now I … So, here I brought some paper, Marina, do you hear me? Yes, I don’t know: where?

Looking at the results of this combination in the orthographic transcription both 01/02 and 06/07b could be seen as one single turn. The combination of the respective sound segments, though, creates a clear impression of acoustic incoherence (we mark this with an asterisk as deviant). The asterisks mark not an ungrammatical structure, but an audible failure to be perceived as one single turn, although it is realized by the same speaker. Segment 07, where S24 to addresses a group, reveals an additional feature at the level of intonation. Figure 2 exhibits a remarkable amount of stretching in the realm of the two da syllables as compared to the repeated sequence of “ja ne znaju ku-“. This failing balance can be interpreted as a feature of nonneutral intonation as it is described in Yokoyama’s model: it is not a succession of equivalent LH-tones ending with a HL-tone (which is the rule of neutral intonation), but a weakening of tones in favor of one single tone under stress (Yokoyama, 2001, Paschen and Sappok, 2012). This interpretation gives additional evidence for a momentary change in the ongoing development of the discourse situation. Example 1.1 has not to be seen as a single coherent stretch of communication, it is rather three instances. At the beginning S24, all alone by herself and talking to herself, is busy with her briefcase in a murmured soliloquy. Then she collects her thoughts and initiates a new situation of interaction. We don’t have a hint in the recording as to the behavior of the group up to the moment of fragment 03. So we will investigate other situations with the same threefold

43

Addressing by Voice: Deixis and Design

level of addressing in which individuals present at the kafedra participate from the very beginning. Figure 2: Tone and syllable structure of “Да, я не знаю куда. Я не знаю …” (‘Yes, I don’t know where. I don’t know…’)

Да, da

я не знаю куда, j a n ez n aj u

0

ku

я незнаю

da

ja nezna ju 2.542

Time (s)

1.3. S24 with different directions of addressing. S24: 01 Слушайте, всем членам кафедры! Вот придёт К***, кажется, подписать вот эти …

ŽХ:

Listen, this message is for all members of the department! There comes K***, obviously, to sign these…

02 вообще то нельзя подписать до защиты …

in principle, he shouldn’t sign before the viva…

03 Вот. Он сказал что в пять. А в пять мне уже надо (э…э) убегать.

Right. He said at five. But at five I have to be gone.

04 Хорошо. Вот, да оставьте, конечно!

That’s fine. Leave it here, no problem!

S24: 05 Он может заглянуть здесь. Я ему позвоню сейчас, чтобы он заглянул в семьдесят пятую. Слушайте, К*** придёт, вот (…) ему лист, вот чтобы он подписал, и оставил на месте.

He may pop in here. I’ll call him right now and ask him to pop in room 75. Listen, K*** will come, so give him a sheet of paper so that he could sing and leave it here.

In segment 01 of example 1.2 S24 addresses the whole group of those present in the room, segment 02 is her comment to what she had said before, again addressing herself in whisper in presence and within the acoustic reach

44

Christian Sappok

of the group. Segment 02 is addressed to an individual member of the group prompting her to respond, which this individual does in 04. The shift in the quality of the voice, for example the jump from 01 to 02 and then to 03 can be demonstrated visually in figure 3: segment 02 has its own loudness maxima being at the first glance not much lower than the maxima in 01 (middle row), in 02 loudness maxima coincide with hissing consonants, and not, with vowels (at it is the case in 01). Figure 3: The waveform and intensity of segments 01 and 02 from text 1.2.

S24 is the boss and plays her role as an authority within the group of staff members in fluent change connected with individual consideration, individual face to face interaction and as an authority within the group. At the first glance, she as the boss has a well-established right to switch between the three functions, the three styles and the three voices. But alas! There are three minutes between the moment when S24 shows up at the kafedra and the beginning of the conversation in 1.2. And within this short interval there is a constant ongoing communication during which the aforementioned model of the three-level addressing is severely disturbed. The nature of these disturbances will be analyzed in detail, and we will show that they do not indicate an inadequacy of the model but, on the contrary, they yield evidence that the model is legitimate: S24 is confident in situations when the model works and it is a disaster for her when it is in danger of a serious breakdown.

45

Addressing by Voice: Deixis and Design

1.4. S24 and the beginning of the conflict. 01 02 03

S24 Добрый день ! добрый день ! ŽХ (телефон гудит) Трещит? Ž1 Ты же сказала, что тебя не будет.

04

S24 Ну кто сказал? Ну как я сказала ? Ну вот скажите, когда я тебе говорила ? Ž1 Но ты сказала, что ты … отпросилась … получать … S4 Я сейчас уйду, сейчас пришла и сейчас уйду, да, да? Я … я … я … я не ожидала, нет, я сейчас … пришла, и уйду. Я п… по бухгалтериям таскалась.

05 06

Hello! Hello! (lit.: ‘good day’) (The telephone rings) Is the line crackling? You said that you were not coming, didn’t you? Who said this? How could I say this? Tell me, when did I say this? But you said that you asked leave… to receive… I am leaving right now, I’ve just came and will leave straight away, right? I… I… I… I couldn’t expect, no; I’ve just come and will go away. I spent ages at the Finance Office.

S24 enters the room, the kafedra, with a greeting (1.3, 01), “dobryj den’” repeated two times, with rising intonation and breathy voice, making everybody feel that she is in a hurry and wants to attract immediately a high degree of attention (cf. figure 4). Other people in the room are distracted by the telephone ringing exactly at this moment and by a remark made by one of the staff members. ŽX alludes to what S24 said the day before regarding her plan not to come to the kafedra the next day. This is a normal course of events happening in constellations like that quite frequently, and it would be easy for S24 to ignore this remark and to proceed as planned by addressing the collective audience. Instead, S24 reacts in an uttermost emotional way, which becomes obvious from the extra high rise Figure 4: Intonation and the waveform of 1.3, segment 01.

Добрый день! 0

Добрый

Time (s)

день! 2.127

46

Christian Sappok

Figure 5: Intonation and the waveform of 1.3, segment 04 (beginning).

Ну кто сказал? Ну 0

кто

ска-

Ну как я сказала? зал? Ну

Time (s)

как

я

ска-

зала? 1.716

of the tonal contours in the first two questions within 04. What is it that makes S24 to react in such an unexpected and inadequate manner as she does in the turns to follow? We could assign to this responses in 03 and the utterances to follow the attribute of conflict or threat (in the sense of Ohala’s 1996, p. 1813 formulation: “I am large and a threat to you”); this, however, would only be justified if ŽX can be shown to be an antagonist, an adversary of S24. The role of S24 as an authority should be shown to be attacked by ŽX, the role of the former as a superior figure questioned, with an unresolved rivalry between S24 and ŽX finally coming into light. But this is not the case. ŽX, as it can be seen from her reaction, doesn’t even understand the outburst she has provoked. In cooperative intention the speaker tries to refresh the wording of what she is relying on in uttering 03, and we have no reason to believe that she was misled or misunderstood the formulation heard on the previous day. So it is not a conflict between the interacting parties, and we are confronted with an eruptive succession of turns the motivation of which would stay a riddle – unless we rely on the model of three-level addressing presented earlier and apply this model to the situation without any trace of conflict. S24 came into the room to play the role of a superior addressing the staff. Being addressed by an individual member of staff before she had a chance to realize this model makes it necessary to start the model anew and put it into motion addressing the community once more. But this is not possible for S24 in the present moment. She activates all her authority to refute the adversary who claims that she heard S24 saying: Tomorrow I will not show up. So S24 has to deny that there was

Addressing by Voice: Deixis and Design

47

someone who said this (Kto skazal?), the way it was formulated (Kak ja skazala?). She admits her words, but she questions the time of their utterance (Kogda ja tebe govorila?). This demolition of time, place, person and manner is an attempt to reconstruct her role of the superior addressing the collective listeners using the complex mixed address: Skažite, kogda ja tebe govorila! There is a second invasive turn against S24, not against her personally, but against her inner model of a superior addressing her staff. But this time she is better prepared to defend her model. 1.5. S24 – discussion about joint teaching. 01

02 03 04

05 06

ŽХ: Здравствуйте. Можно вас побеспокоить? Я … я достала папку по первому курсу вечерников, S24: Так? ŽХ: вот, и хотела бы узнать, в какой форме мы сдаём экзамен. И по поводу вопросов … S24: Вопросы я вам (…) где-то там, я не знаю, во второй половине ноября или в середине ноября только дам. ŽХ: Угу. S24: Вопросы, да, какой ва… , вот два вопроса там будет. Вот по (э…э) в основном по (…) тому материалу, который мы с вами будем проходить, вот, но это не исключает (э…э) учебник, и как-то нам на… вам надо о учебнике поговорить.

Good morning / afternoon. Sorry to disturb you. I… I got a folder on Year 1 evening classes… And? …and I’d like to find out how we are going to sit the examination. And with respect to questions… I will give you questions only…. I don’t know, maybe by the second half of November or mid November. Uh-huh. Questions, right, there will be two questions. Mainly based on the materials we will cover [in the class], but this doesn’t mean that you won’t need a textbook; we need somehow… you need to have a conversation about a/the textbook.

As in 1.2, an individual speaker approaches S24 who would like not to be addressed, but to address. ŽX uses a quiet voice, begging pardon in advance trying not to be intrusive. We can conclude that ŽX and S24 are bound by conjoint teaching, planning and preparing for final exams. They need to develop a standard approach concerning the questions to be asked at these exams. ŽX, who obviously does most of the teaching, is eager to enter a discussion and to find a solution for the problems related to teaching, examination and to discuss specific questions for the forthcoming exams. The best idea would be to solve these problems which are relevant both for S24 and ŽХ means of

48

Christian Sappok

a dialogue, i.e. a communication between two individuals having a common task, equal goals, equal rights, and equal status in addressing each other with questions and answers. Again, as in 1.3, this is not in line with the model on unidirectional addressing followed by S24. And this time, S24 doesn’t experience any threat to her preferred model of communication which she maintains in her reactions, even when her answers don’t match the questions asked by ŽX. For example, as response to the question “What will we do …” she says: “I will do …”, and the obligation “nam nado … / we should …” is instantly changed to “vam nado / you have to”.

3. Speaker 05 and the design model of addressing Speaker 05 is one generation younger as compared to S24, but there is a couple of common traits: she also works at the university, she is member a department’s teaching staff, though she is not a professor. So we could choose out of her daylong recording a portion where she is at the kafedra, in the room where the teaching staff assembles shortly before everybody goes to their classrooms to begin a lesson. The recording in question is not a structured dialogue or polylogue between speakers involved in a well-bred conversation, but a fluctuation of changing and intermingling interactions between people concerned with pending educational and administrative matters. S05, equipped with the ORD recorder, enters the room with a greeting, talks to different people who are present there, and changes her addressees within very a short time interval, varying her voice with respect to different goals and topics. But her technique is quite different from the one used by S24, and we will try to reconstruct it by means of aural analysis. As in the previous case, we will use the detailed transcription made by the ORD group at the Saint Petersburg State University (Asinovsky et al., 2009). Without this written version it would be utterly difficult to understand the wording, and we would like to express our deepest gratitude to the whole group. Here, as in the case of S24, the study of the written text alone would not enable us to understand what is really going on, and we would be helpless if we could not rely on the audio recording, and most of all on the voice quality. Voice quality allows us to distinguish between individual speakers (recognition by stable parameters) and to capture changes in the direction of addressing (“Zielcharakter” of Bühler 1934, p. 98). To give an idea of the technique used by S05 and of her intelligent behavior in this stressful situation we make use of a virtual dialogue between A and B, virtual in that it does not match exactly what we find in the recording.

Addressing by Voice: Deixis and Design

49

2.1. construction: A entering the kafedra and exchanges turns with B. 01

A:

Здравствуйте, B., еще раз!

V., hello once again!

02

B:

Здравствуйте!

Hello!

03

A:

Какой компьютер с принтером?

Which computer is connected to the printer?

04

B:

Вот этот.

This one.

05

A:

Слушайте, я прямо не могу. Что мне с этим семинаром?

You know, I’m really fed up. What am I supposed to do with this seminar?

06

B:

Там какие-то нарушения движения и речи. Я сейчас тебе покажу.

There are movement and speech impairments there. I’ll show you.

07

A:

Вот. И что бы придумать такого чтобы они самостоятельно сидели?

Right. And is there anything at all which would allow them to sit without assistance?

2.1a is a sequence of turns organized in adjacent pairs of turns beginning with a greeting and continued with a sequence of information-oriented steps of interaction. For the construction of dialogue 2.1a, we used portions of 2.1b (below) that match the speech material of the original recording. 2.1a. S05 entering the kafedra (original). 01

02 03

S05

Здравствуйте О. А. еще раз! Какой компьютер с принтером? Вот этот. Слушайте, я прямо не могу. Что мне с этим семинаром? Там какие-то нарушения … движения и речи.

O. A., hello once again! Which computer is connected to the printer? You know, I’m really fed up. What am I supposed to do with this seminar? There are movement and speech impairments there. ŽX.: Я сейчас тебе покажу. I’ll show you. S05: Вот. И … что бы придумать такого Right. And is there anything at all which чтобы они самостоятельно would allow them to sit without Сидели? assistance?

Except for turn 04, S05 is the only person speaking in this passage. We can’t say that S05 is performing a monologue. Instead, she is varying her voice and changing her addressing directions with admirable ease and helps us, though we are not present in the original situation, to find the exact reconstruction of the changing constellations. No video recording could be of better help to analyze different directions of addressing.

50 01 02 03 04

Christian Sappok - directed to OA, S05’s superior, with a polite, neutral voice with slight satisfaction about the repeated encounter. - directed to the present community, it is a question. But the answer is given by S05 herself, without the slightest trace of waiting. - again, addressed to all, with a tone of helpless attitude towards the pending lesson and what it should be about. Here, one of the colleagues promises help. -again addressed to all and again with a tone of complaint and a new object of complaint: how is it possible to cope with a herd of unable students?

2.1b. Collective hearers of a narrative. S05: Вот … и что бы придумать такого, чтобы они самостоятельно си… Я вчера их встречаю на улице, ну в пятерочку. Шла как раз …

In the following fragments, 2.1c and 2.1d, S05 addresses two colleagues. In the first one she speaks to a female coworker with whom S05 is involved in a conjoint seminar. The second is a conversation with a male colleague, barely understandable in the surrounding bubble of voices, but it becomes clear that he asks for an object (file) that is saved on a USB stick. 2.1c. S05 addressed by a female colleague, ŽX. 01 02

ŽX: S05

03

ŽX:

Лен, я по поводу … Я поняла. Я сейчас, сейчас сeйчас сейчас сейчас. Я подождать могу.

Lena, I wanted to see you about… I understand. Hold on, hold on, hold on. I can wait.

2.1d. S05 addressed by a male colleague, MY. 01 02 03 04 05 06 07 08

MY S05 MY S05 MY S05 MY S05

09 10 11 12

MY S05 MY S05

Лен! Да! А там нет (НРБР.)? Нету. А зачем он вам нужен? Нужен. Вообще-то вам я нужна. Вы мне всегда нужны (смеётся) Она там такая. Ну она ещё ничего не сделала, после третьей пары. Aaaa … На чем она? На флешке. На какой? Откуда ж я знаю? Ну идите спросите!

Lena! Yes? Is (unclear) there? No. Why do you need him? I need him. Actually, it’s me you need. I always need you. (laughs) Like, she hasn’t done anything since the third period. Aaah … Where is it? On the flash drive. Which one? How do I know? Go and ask!

Addressing by Voice: Deixis and Design

51

The sequences from 2.1a to 2.1d reflect the wording of what was said, but they do not fully exhibit the essence and the contents of the interaction between S05 and her interlocutors. But the strategy of switching addressee and the position on the recorded utterances where a switch takes place are clearly signaled by voice modifications. There appears a succession of addressees: superior O.A., the collective addressee, the female colleague, the male colleague. This seems to be similar to the switches which we observed in the case of S24. One obvious similarity between the two speakers is that both change the quality of their voice as they change an addressee. But this is not the whole truth. S05 not only performs changes in the choice of addressees, she also changes her voice parameters within the interaction with one and the same partner. To give it a simplified equation: S05 has four partners but eight voices. It is difficult to believe, but it leads us to one central aspect of voice variation. The parameters of voice determining the perceived tone of a stretch of speech must not stay identical over this whole stretch in question. A change in voice is perceived at the beginning of the stretch as opposed to the preceding passage, as a momentary shift, a signal of discontinuity. It is shown to function in the examples of misssegmentation, see above, examples 1.1a with asterisk. This jump, as we named it, within a stretch addressing one and the same partner is best to be seen in the case of S05 speaking with her superior partner, B. She begins with a tone that might not be overtly classed as submissive or humble, but it is a tone with the appropriate respect; you hear it and you notice: this is how a superior has to be addressed. But then comes 3.1: 3.1. S05 loses superiority. 01 02 03 04 05 06 07 08 09

S05: Слушайте, там к нам К* на кафедру просится, представляешь, B? B: Какая К***? S05: Ну которая какая К*** … такая … B: учебник? S05: Да, которой учебник … B: О-о-о! S05: Вот! Но ее не хотят брать. Так, а что мне надо? Мне… B: На кафедру – это ку-куда? S05: Сюда, сюда, к нам в это … в П***.

Listen, apparently K* wants to join the kafedra, can you imagine? Which K***? Well, the K*** who … the one… textbook? Yes, the one who wrote the textbook … Oh! There! But they don’t want to hire here. Right, what did I need? I… The kafedra … Where’s that? Here, here, that’s us – P***

Reading the text, one notices the following: The respectful use of name and patronym (cf. Dickey 1997 on the use of names in addressing) used in 1.1b, S05 changes into the confidential predstavljaeš’; the syntax becomes colloquial and elliptic. This

52

Christian Sappok

is what can be observed with an overall look at the text of 3.1. But what you hear is a change from the very beginning of 3.1. It is a voice void of the preceding distance, it is the adequate tone for two interlocutors of equal status, of equal position, of common conspiracy. What is it that motivates S05’s momentary downgrading of her superior in tone and style? It is the appearance of a third person, not at present, but in near future: the shared negative attitude towards this person unites the two and promotes the leveling of the difference in status. There is a noticeable difference between S24 and S05. In 1.4; ŽX addresses S24 as her superior and asks for a dialogue at the level of “us both” uniting the partners in the face of a common project, but it is not accepted by S24 who does all she can to restore and maintain the distance. 3.1 presents a similar case of intrusion by an inferior, but it is accepted and does not affect the mutual communication. The opposite is true in 3.3, the direct continuation of 2.3, where S05 converses with her female colleague and partner in a joint teaching process. 3.2. Colleague treated as inferior 01

02 03 04 05

S05: Да, там семинар, в общем сущность какая? Семинар, тема да? Там по сути дела теория … концепция Выготского и концепция Леонтьева. Там у меня все расписано, то есть вопросы я им уже да.. заранее выдала. Список литературы выдала и результат по которому оценивать я тоже там написала, да? То есть у каждого в результате семинара был ли он на семинаре – не был ли он на семинаре, должна быть заполнена табличка, форма таблички прилагается. ŽX: Уже отдала да? Ну там сравнение двух концепций? S05: Да. Это все отдано. ŽX: То есть, я должна у них это все собрать, да? S05: Ты должна у них вообще семинар провести. Вы должны все это дело все вопросы обсудить и совместно заполнить эту табличку. У каждого напросто должна в тетрадке …

Yes, the seminar, what is it about? The seminar topic, yes? Basically, it’s theory … Vygotsky’s concept and Leontiev’s concept. I have it all planned out, so I’ve already given them questions … in advance. I’ve handed out the reading list and I’ve written up the marking criteria too, yes? So, after the seminar, everyone, whether they were at the seminar or not, should have filled in a table. I’ve attached a sample table.

So you’ve already handed it out – the comparison of the two concepts? Yes. That’s been handed out. So I need to collect that from them, yes? You need to lead the seminar. You need to discuss all this, all the questions, and fill in the table together. Everyone should simply have in their notebook …

Addressing by Voice: Deixis and Design

53

The confidential tone comes to an abrupt end. S05 and ŽX are no longer equal partners in conversation; S05 is no longer in the position of asking for help or advice (cf. 2.2 01 “What shall I do with this seminar / Что мне с этим семинаром?”). Rather, S05 is a superior: she is dispensing advice in a cool and matter-of-fact voice that is best seen in the adjacent pair of 3.2, 04 and 05. ŽX’s question concerning the details of joint work with the students is not answered but instead prompts a reprimand: “You just have to lead the seminar.” A request for help by a male colleague in 3.4 prompts a similar reaction, but includes an element of joking and laughter. 3.3. Male colleague as a flirting partner. 04 05 06 07

S05: MY: S05: MY:

Нету. А зачем он вам нужен? Нужен. Вообще-то вам я нужна. Вы мне всегда нужны.

No. Why to do you need him? I need him. Actually, it’s me you need. I always need you.

S05’s provocative question, “What do you need it for?” is answered by MY with a simple “I just need it”. This gives rise to a little flirt: “It is me you need!”; the male partner reacts as he must react as a man in such a situation: “It is always you that I need”. The flirt ends as quickly as it began. The location of the required object remains unclear and this is not resolved until the end of this interlude, where S05 not only refrains from answering MY’s question, but also refrains from making it her own question: “How should I know?”. The resulting tone is one of rudeness, invoked not only by voice, but also by the general effect of answering a question with a question (cf. Yokoyama 1990). In the three fragments presented in this paragraph, 3.1, 3.3 and 3.4, we observe a clear momentary change of tone, a switch of voice, a discontinuity of addressing. This does not involve a change of partners, but it is realized in interaction with the same partner: S05 with OI in 3.1, with the female colleague in 3.3, and with the male colleague in 3.4. We feel a certain conflict in our formulation: S05 does not switch from one addressee to another, but rather from one addressing style to another. In a certain sense we could formulate this the other way around: S05 switches from one addressee to another, however, the two addressees are embodied by the same person. We will, at the end of our study, propose a solution to this problem by drawing on Bühler’s concept of imagination-oriented deixis (Deixis am Phantasma): If we activate our imagination to take over the role of S05’s partners, we surely would feel the effect of being played with. ŽX expresses

54

Christian Sappok

this feeling overtly in her turn directed at S05: «Ну ты меня тоже огорошила!» – „Well, you also stupefied me“. 3.2 is an exception from this game of constructing two addressees from one physical partner. S05 does not change voice and style in addressing a group. 3.4. The collective hearers of a narrative. S05 Вот (…) и что бы придумать такого, чтобы они самостоятельно си … Я вчера их встречаю на улице, ну в пятерочку.Шла как раз в магазин. А был вечер, пол десятого. Я такая: «О-о! Вы приехали? А у вас завтра занятия будут?» Они такие: «Будут!» Я говорю: «Да-а, не повезло мне ! А что такое? А у меня завтра у вас три .. два семинара.» А они приехали всего в восемь вечера только из Кронштадта они при.. их привезли, представляете?

Right (…) and is there anything at all which would allow them to sit without assistance… I met them in the street yesterday – I was going to Pyatyorochka, to the shop. And it was about half past nine at night. I said, like: “Oh! You’ve arrived? Do you have classes tomorrow?” And they went: “We have!”. And I said: “Yes, bad luck for me! I’m teaching three … two seminars tomorrow” And they only just arrived from Kronstadt at eight, they were brought here … can you imagine?

Such behavior does not stem from the fact that a group is more difficult to play with. Rather, the reason is that in this brief narration depicts S05’s encounter with a group of students of hers in a shop the previous evening. And, in the reported speech she brings to light this scene, she uses her voice for other purposes, i.e. for characterizing her own voice and the voice of the students. Here and elsewhere we see good reasons to follow Tannen 2007 and use her term constructed (instead of reported) speech or dialogue (Tannen, 2007, chapter 4). This is an interesting phenomenon concerning our position as listeners of a recording with a wide spectrum of voices and voice modifications. The change of voice and roles without accompanying changes of persons present in the discourse situation is easy to interpret for a person present in the situation. In a listener not present in the situation and relying only on the acoustic signal, however, this can lead to a conflict in orientation. An interesting case can be found in paragraph 4, the monologue of S05 where she is involved in a narration, not at the kafedra, but chatting with her friends.

4. Voice parameters and their modification in monologue We started our observations of voice parameters with recordings made in a specific situation, which we named kafedra, where an unordered bubble of voices provokes its participants to do their best in using their acoustic facilities. The absence

Addressing by Voice: Deixis and Design

55

of a canon ruling turns and turn-taking, as we know from a wealth of studies on conversational analysis, forces speakers to create rules of their own and to initiate signals of their acoustic presence in an already occupied field. This requires awareness of their own qualities as a speech source and an effort to make it heard in a setting where everything is heard by everybody, requiring one to identify oneself as the addresser as well as one’s addressee. A comparison of two participants in the erratic field of intersecting sounds, 024 and S05, illustrates how difficult it is to find a stable point of departure and an adequate voice technique to present oneself and to signal directedness. Both S24 and S05 make use of a signaling technique based on social hierarchy. S24 has difficulty in defending her role of the superior addressing her inferiors; S05 performs much better by employing the same hierarchy but makes use of her voice to change her position within the hierarchy. Two questions arise once we accept these observations made possible by using the unique data in the ORD corpus. I. Is it only the battlefield situation of type kafedra that forces speakers to evoke and exploit voice as a strategical means not necessary in normal conditions? If this is the case, we could return to the well structured forms of spoken communication where voice again has to occupy its marginal status of paralinguistic ornament of linguistic categories in the sense proper? Or, to bring forward an alternative, is there a functional need outside kafedra for S24 and S05 to use their voice facilities in a different context and field of communicative reality? II. The other issue is the dualism of models we proved to be adequate for our comparison of S24 and S05. We proposed, following Bühler 1934, the deictic model for S24, although it does not always function with uninhibited ease but rather in the defective cases where her addressees fail to fit into the expected scheme. What we observed was, essentially, the effort to regain the temporarily lost model. We saw that S05, though using the same model of social hierarchy, switches her position of source and the hearer’s position her message is directed at. We proposed in this case the design model (Bell 1984) highlighting the variable status of audience: In the use of variants a speaker feels free to model his audience according to his own initiative and creative audaciousness. Once we adapt these two concepts, the following question arises: Are we confronted with two individual cases that can be multiplied in an ad hoc manner observing other individuals using their voice and varying its components, or are we likely to find a common model for deixis- and design-type of using voice? To resolve this, we propose to shift our attention to the narrative fragments below: S05 recounts her first driving lesson that took place the day before.

56

Christian Sappok

4.1. S05 and her driving instructor 1. 01

02 03

04 05

06 07 08

S05: S05: Там меня дядька посадил, короче говоря, за машину, и спросил ну типа:

So, long story short, the guy put me behind the wheel and asked me: “Can you drive at all, or not? Have you been behind the wheel?”

«Умеешь вообще кататься или нет? Сидела за рулём?» Я говорю: «Ну как? Пару раз было дело, конечно, ну знаю просто, где тормоз, газ, сцепление, всё, больше ничего не знаю!» Он такой: «Ну ладно, говорит, давай!» Отъехали мы, вот, от эт.. (э-у) площади. Ну он сам отъехал там на такую вот дорогу уже (э) как сказать, закрытую. Выехали, там где машин мало. Ну вот, и он меня посадил за руль. Я вначале тронулась такая дын-дын-дын вообще, знаешь! Ну вон такой: «Я сказал: сцепление по-другому!» А я думала, что ну этот (э) мужик, что он такой, как бы сказать, что он псих такой. Я так не хотела к нему идти.

I said: “I mean, I have a few times – I know where the break, the gas and the clutch are, but nothing else!” He was, like: “OK, well, go!” We drove away from this … ah … square. He drove to, like, a closed road, where there are few cars. When I first moved off, it was, like, dun-dundun you know. He was, like: “I said you need to use the clutch differently.” And I thought that this, ah, guy, was, how should I put it, a psychopath. I really didn’t want to learn with him.

Figure 6: Provocative questioning of the driving instructor with stress on the verb.

Умеешь

вообще

кататься

или нет?

LH 0

сидела

за рулём?

LH

Time (s)

2.54

Addressing by Voice: Deixis and Design

57

4.2. Driving instructor 2. 01

02

03

S05 А такой нормальный. Он так ласково разговаривает, такой: «Ну вы успокойтесь, у вас всё будет нормально, всё получится, всё хорошо, знаешь»! По имени сразу такой, так спросил как по имени: «Вот Леночка давай, так отжимай, так тихонечко, да да да да да, вот так, вот знаешь?» Он то есть … чувствуется, что он с такими лохушками как я, знаешь, работает.

But he’s pretty normal. He speaks, like, really softly: “Calm down, everything will be OK, you’ll manage just fine, you know”! He addressed me by name straightway, he asked for my name straightaway: “There, Lena, release slowly, yes, yes, yes, yes, just like that, you know?” So he’s … you can see that he’s used to working with muppets like me.

The extracts 4.1 and 4.2, which give rise to S05’s narration, describe the beginning of her driving lesson and her first contact with her instructor. The events of this introductory scene are passages of reported speech, citations of instructor’s speech along with S05’s reactions and comments. The speech of these passages is vivid in tone, rich in coloring and emotions. The majority of this scene reports what the instructor said, brought to the listener’s attention through the rhetoric of S05 known from the kafedra situation. I listened these two passages over and over again to in order to establish the sequence of events: What happened between the two extracts that gave me the impression that we are dealing with two different instructors? Upon asking the ORD colleagues, I was reassured that there was only one instructor involved in the lesson in question and in the reported scene. We are confronted here with an interesting phenomenon yielding insight into the functioning of our acoustic perception. Listening to S05 in the kafedra situation we performed without difficulty to receive her voice as constantly belonging to her even while she drastically changes the tone; there was never any doubt that S05 remains the same person and that it is her constant individual voice that undergoes a change, a modulation which modifies her voice. The addressing functions provide a good starting point for interpreting the motivation of this kind of using voice parameters. In 3.1, the impression given of the driving instructor is one of a rough, unfriendly guy; then, after a very short interval, we hear this same person change his tone, he treats his pupil with a voice that can be described as normal, nice, even with a hypocoristic note. So why is it that I fell victim of this technique of modifying voice parameters and didn’t perceive it as a change in tone or style, as one variant against the other?

58

Christian Sappok

My solution is of a rather preliminary nature. Our ear’s perceptive and operational capacities are of a high level (unsurpassed by any invented technology). The human ear admirably distinguishes with masterful ease the two levels of speaker identity and speaker variation. In cases such as 4.1 and 4.2 there enters an additional level, the reporting voice of S05. We use our auditory capacity to discriminate between her speaker identity and her variation, but we do not have to our disposal an additional device for dealing with a two-level individual. This is not a task the auditory apparatus is made for; it is designed to be used in cocktail- party situations (Cherry, 1953) to single out an individual as opposed to the background noise, but not to make a true assessment of all the acoustic sources in the auditory scene (in the terminology of Bregman, 1995). So, we infer from the two additional tones, those of S05 and of the driving instructor, that there have to be two sources and two individual voices not sharing the same identity. S05 has a voice to address different people, and she has two (or more) voices to address the same addressee. This was our observation of S05’s behavior in the kafedra situation (that parallels Cherry’s cocktail party). In the narrative monologue situation, she makes use of the same art of voice modulation, not in the function of addressing proper, but in the function of reporting addressing speech. This is relatively close to a situation where S05 uses her modulation of speech for reporting not only individual voice (not present in the actual situation of narration), but to address objects or individuals not entering into communicative interaction at all, whether actual or in preceding past. This can be demonstrated in our next example. 4.3. Babus’ki S05

Там сидят бабуськи такие, на остановке, у них глаза такие по полтинику, ну представляешь, они сидят, ждут автобуса. Там площадка, а я там такая, наяриваю, эти восьмёрки, вообще, ещё знаешь, ну не уверена могу там, хоп, там, в один момент, там чуть в яму не въехала. Ну он там вырулил, конечно, потом поехали.

There were these grannies at the bus stop. Their eyes almost popped out, can you imagine? They’re just sitting there waiting for the bus. There’s a parking lot there, and I’m doing these figures of eight – you know, I’m still shaky, almost drove into a pothole one time. He steered away, of course.

S05: Там сидят бабуськи такие, на остановке, у них глаза такие по полтинику, ну представляешь, они сидят, ждут автобуса. Там площадка, а я там такая, наяриваю, эти восьмёрки, вообще, ещё знаешь, ну не уверена могу там, хоп, там, в один момент, там чуть в яму не въехала. Ну он там вырулил, конечно, потом поехали.

Addressing by Voice: Deixis and Design

59

S05 recounts in 4.3 a dramatic scene of her first driving experience. She is in full stress and has to pay attention to the dangerous street conditions and to unwelcome observers. These enter the scene in the form of a group of old ladies sitting on a bench, waiting for the bus. S05 describes them not in a neutral descriptive tone, but in a special tone of addressing non-serious people. I could class this type of speech as baby-talk, with the exception that the modulation is not addressed to a baby but to old ladies. We observe a similar modulation again in the passage of 4.4 where S05 uses the same style to refer to the car she used for her first lesson, this car bearing witness to her status as incompetent driver (cf. 4.1, 08). 4.4. Vosmjorka. 01

C:

Машина ? Машина какая ?

Car? Which car?

02

S05

Машина, да! такая она хорошенькая, такая красненькая ну правда.

Car, yes! It’s quite nice – red, though.

In extract 4.4, the same baby-talk-like modulation is not used for a direct or reported addressee, but to refer to a car, i.e. to an inanimate non-interactional object. This seems to be the highest level of S05’s variation of her voice parameters to characterize partners, third persons or non-human objects out of the reported scene.

5. Two models of addressing or one? We now come to the final question concerning the two cases of S24 and S05: Are we dealing with two individuals each using an addressing mode and a voice variation of his or her own? This would open the field of observations to an infinite collection of cases like S24 an S05. Or, is there one underlying common model with predictable alternatives? The second solution is, of course, seems preferable since it would bring to light a general connection between voice parameters and discursive functions. Let us reassume our comparative observations: The two protagonists, S24 and S05 (we give them this title not because they are destined to the focal point in the situation, but because they bear the ORD microphones on them and are heard and understood best), have the following traits in common: – They share the so-called kafedra situation. For the parties entering into a verbal interaction, this is a complex and difficult task compared to the so-called cocktail-party situation described in the study of Cherry 1953. The nature of the difficulty lies in the fact that there are no obligatory rules in the kafedra

60

Christian Sappok

situation – rules, such as those of turn taking and turn sequencing known from the literature on conversational analysis. This does not mean that there are no traces of those rules, but they cannot be expected to be followed, because the obligation “Go and enter your class room!” is predominant. – There is in both cases a clear social hierarchy with three layers, superior – lecturer – student, the last group only mentioned or reported by S05. – Both S24 and S05 show variations in their use of voice parameters in addressing. These variations can be perceived if one focusses on places where there is a sudden jump or break in the discursive coherence. These discontinuities are best observed by employing the false segmentation technique: A passage before the break and a passage after the break are extracted from the original context and examined in isolation. The results in a clear impression of a lesion, of a discontinuity in discursive coherence. Along with these common traits, there is a considerable number of differences. – S24 enters the kafedra with a preset strategy of addressing: She – as a superior – addresses the present group of kafedra members, allowing her interim addressing of others, but not allowing others to address her. The consequences can be seen in the extracts 1.2 and 1.3. They show a collapse of the prepared strategy. These are fatal for S24, but they yield considerable evidence in support of our addressing model. We term it, following Bühler 1934, a deictic model of addressing, giving the field of addressees a threefold structure, i.e. the self (addressed by whispering or speaking aside), the individual next to the speaker, and the group of all those within earshot. – S05 enters the kafedra under similar conditions as S24. First, the superior is addressed, then the group as a whole, then the female colleague, and finally the male colleague. The lowest level in the social hierarchy, the group of students, is addressed only in passages of reported speech, but with a clear stylistic marking. – In contrast to S24, S05 has no pre-elected strategy to maintain and defend throughout the interaction. She accepts her position on the middle level addressing the high level (i.e., her superior); she switches as an individual addressing the group, from addressing her female colleague to the male colleague. Up to this point, her addressing technique and the use of voice does not differ, in principle, from that of S24, only in that she is not bound by a fixed addressing strategy, but instead switches without conflict within the deictic field. – She makes an important further step, however, abandoning the normal deictic use of addressing. Keeping the social hierarchy intact, she switches her position within it. This only occurs in short passages, but is marked clearly by her voice characteristics, obvious even for a listener not present at the scene.

Addressing by Voice: Deixis and Design

61

Her position with respect to her superior remains unchanged but, for a short stretch of the conversation, she transforms this distance to a conversation between partners of equal standing. She leaves the same-level constellation with her female colleague intact, but, again, for a short interval, she switches to an addressing of type superior to substitute, giving her instructions “how to give a seminar”. She speaks with her male colleague about a missing object, leaving intact the constellation of same-level colleague, but changes this constellation, again for a very short period, to a flirting between man and woman. We feel a strong tendency to describe the addressing behavior of S05 within the audience-design model proposed by Bell 1984. S05 accepts her partners as they are, but she feels free to modify it, at least for a short stretch of interaction, according to a shift within the spectrum of variation or, in our case, of voice parameters. She creates a modified constellation between herself and her addressee which is clearly audible to the listener. These modifications are created and activated for a certain interval of time after which there is the possibility to exit the imaginative scene and to return to the situation before the shift. S05 uses the realistic scheme of constellations, modifies them in a way recognizable by her voice, and returns to the point of departure. The above considerations open the path to a common model for S24 and S05. Both speakers use their voice to signal addressing within their field of verbal communication. Both of them use the same deictic field, structured as self – near – distant, to which in acoustic terms produces soliloquy, individual in near distance and open to all present in the acoustic field. The additional field occupied by S05 with the help of voice parameters is a kind of dislocation, not only in her own mind, but also understood by listeners, guided by her voice. This is a technique described in the deixis paragraph of Bühler 1934 as a special case of “imaginationoriented deixis – Deixis am Phantasma”. Deixis oriented on the real coordinates of the speaker and his addressees and the dislocation of the coordinates (cf. Versetzungsdeixis in Bühler’s terms) is a legitimate means of employing deictic signals. It is widely used and accepted by the addressees, be it in written or in spoken contexts. It takes place in a sphere of freedom, accepted and welcomed as a means of creative language use and voiceparameter use. If this is true, we should wonder why S24 is not familiar with these functions. And, unsurprisingly, she makes widely use of these as soon as she is out of the bounds of the kafedra situation. To give a single but impressive example, we follow her coming home and sitting at her desk. She enters into vivid communication with her computer. She transforms it, an object unable to react, but addresses it as a fully-fledged communication partner in fragment 5.

62

Christian Sappok

5. S24 addressing the computer. S24

01 02 03 04 05 06 07 08 09 10 11 12

Где панель управления? Покажите пожалуйста! Так. Не то! Да что ж это такое! Так. Сейчас. Так. Молодец! Это хорошо. Так, ну хорошо, давай давай давай. Ну давай ну … такая хорошая! Готово. Ой хорошо как! Ой какой ты хороший, такой маленький мой. Ну что такое то? Ну то … ну я не знаю! Чудесно! Чудесненько! Молодец! Отправить!

Where’s the Control Panel? Show me, please! Right. No, that! Come on! Right. Now. There. Well done! That’s good. OK, good, come on now, come on. Come on … good girl! Ready. Oh, good! Oh, you’re so good, little one. What’s this? Oh … I don’t know. Wonderful! Wonderful! Well done! Send!

Figure 7: Two versions of kakoj, (1) is the neutral style, cf. 1.4, 03, (2) is taken from 4. 10, i.e. from the baby-talk style S24 addressing the computer.

The reader of the written version of the interactions reported here may feel uneasy at their inability to examine the acoustic material. Ethical considerations, however, do not allow for open access. The reader is therefore confined to the poor devices of graphical presentation. Figure 7 compares two versions of the word “kakoj” in

Addressing by Voice: Deixis and Design

63

the kafedra and in the home context, the latter version is realized in the baby-talk voice we reported in the case of S05 talking about the old ladies sitting on a bench and talking about the little car used in her driving lesson. The values of F1 in the left-hand (baby talk) version serve as a solid indication of lip rounding causing a decrease in the formant position. Varying voice parameters as addressing device is shown to be used not only in real interaction, but also in narration, reporting, and in simple referring. It is oriented toward the individual as it is, and transforms the addressee in his or her social (or local) position. A wide field of creative use of voice in oral communication opens for analysis.

References Asinovsky, Alexander / Bogdanova, Natalia / Rusakova, Marina / Ryko, Anastassia / Stepanova, Svetlana / Sherstinova, Tatiana: The ORD Speech Corpus of Russian Everyday Communication “One Speaker’s Day: Creation Principles and Annotation, 2009, retreived 15.03.2018, from: https://www.academia.edu/2359835/ The_ORD_Speech_Corpus_of_Russian_Everyday_Communication_One_ Speakers_Day_Creation_Principles_and_Annotation. Bell, Allan: “Language style as audience design”. Language in Society 13, 1984, pp. 145–204. Bonastre; Jean-François., Kahn, Juliette, Rossato, Solange, Ajili, Moez: Forensic speaker recognition: Mirages and reality. Susanne Fuchs, Daniel Pape, Caterina Petrone, Pascal Perrier (eds.), Individual Difference in Speech Production and Perception. Peter Lang: Frankfurt/M. 2015. Bregman, Albert S..: Auditory Scene Analysis. MIT Press: Cambridge, Mass. 1995. Bühler, Karl: Sprachtheorie. Die Darstellungsfunktionen der Sprache. 3. Auflage, UTB: Stuttgart 1999 (1934). Cherry, Colin E.: “Some experiments on the recognition of speech, with one and with two ears”. Journal of the Acoustical Society of America 25, 1953, pp. 975–979. Dickey, Eleanor: “Forms of address and terms of reference”. Journal of Linguistics 33, 1997, pp. 255–274. Eckert, Hartwig / Laver, John: Menschen und ihre Stimmen. Aspekte der vokalen Kommunikation. Beltz: Weinheim 1994. Goodwin, Donald Fraser: Theory of Language: The representational Function of Language (Translation of Karl Bühler 1934). Benjamins: Amsterdam 1990. Harré, Rom / van Langenhove, Luk: “Varieties of positioning”. Journal for the Theory of Social Behaviour (21) 1991, pp. 393–407.

64

Christian Sappok

Kreiman, Jody / Sidtis, Diana: Foundations of Voice Studies: An Interdisciplinary Approach to Voice Production and Perception. Blackwell Publishing Ltd., Chichester, West Sussex, 2011. Nikitina, Sima E.: “Adresaty v avtokommunikacii”. In: Arutjunova, Natalja D. (ed.): Adresacija diskursa. Indrik: Moskva 2012, pp. 402–410. Ohala, John J.: “Ethological Theory and the Expression of Emotions in Voice”. Proceedings 4th International Conference on Spoken Language Processing, Philadelphia, Wilmington, University of Delaware 3, 1996, pp. 1812–1815. O’Shaughnessy, Douglas: Speech Communication: Human and Machine. (2nd ed). IEEE Press, New York 2000. Paschen, Ludger / Sappok, Christian: „Drei Konzeptionen der russischen Intonation“. Wiener Slavistischer Almanach 70, 2012, pp. 247–293. Tannen, Deborah: Talking Voices. Repetition, Dialogue, and Imagery in Conversational Discourse. (2nd ed). Cambridge University Press: Cambridge 2007. Trask, Robert L.: A Dictionary of Phonetics and Phonology. Routledge: London and New York 1996. Uni Stuttgart Project on Voice Quality, Uni Stuttgart Project on Voice Quality, http://www2.ims.uni-stuttgart.de/EGG/page2.htm. Viciano, Astrid / Hummel, Philipp / Catanzaro, Michele / Tola, Elisabetta: „Verdächtige Stimmen vor Gericht“. Neue Züricher Zeitung, Jahrgang 2016, Ausgabe 15. Januar, p. 58. Yokoyama, Olga T.: “Responding with a question in Colloquial Russian”. In: Harvard Studies in Slavic Linguistics Vol. 1, Pusey Library, Harvard University: Cambridge MA, 1990, pp. 175–195. Yokoyama, Olga T.: “Neutral and Non-Neutral Intonation in Russian. A Reinterpretation of the IK System”. Die Welt der Slaven XLVI 1, 2001, pp. 1–26.

Grammar

Nadine Thielemann

Constructional Patterns Emerging in Talk-In-Interaction – Some Preliminary Observations on the Grammar of govorit’ in Colloquial Russian Abstract: The paper adopts a usage-based perspective on the grammar of Russian talk-ininteraction and analyzes specific patterns relying on govorit’ (‘to say’) in the ORD corpus. Combining ideas from Construction Grammar and Interactional Linguistics, we determine and analyze patterns which meet recurrent communicative needs and which are associated with specific usage contexts in terms of constructions. The focus is on constructions with govorit’ (‘to say’) which indicate a shift of footing (Goffman). Shifting the footing covers various ways in which speakers detach themselves from their utterances and manipulate the degree of responsibility for their talk. In doing so, we pursue two aims, a theoretical and a practical one. On a conceptual level, we test Construction Grammar as a framework for the analysis of linguistic patterns in talk-in-interaction. From this perspective, we scrutinize it as a grammar for the description of a spoken language. On the level of application, the analysis of the constructions in which one of the most frequent verbs in Colloquial Russian occurs reveals ongoing processes of language change. The results of this analysis nurture the argument that grammar (at least grammatical structures of talk-in-interaction) emerges from and is adapted to discourse routines. Our analysis underscores this argument in revealing the network of constructions which allow speakers to detach themselves from what they say. We can show that each construction within this network is tailored to a specific usage context.

1. Introduction Talk-in-interaction, i.e. language as it occurs in casual face-to-face conversation, is the variant to which users of a language are most often exposed. It is this variant of language which interlocutors use in order to solve recurrent tasks in face-to-face interaction relying on the appropriate linguistic patterns. It is also the starting point of processes of language change, which can result in the emergence and routinization of novel patterns or units tailored to specific interactional tasks. The One Day of Speech-corpus (further ORD corpus) provides a unique opportunity to conduct research on both aspects of spoken language. It allows for an analysis of linguistic units and structures of Russian talk-in-interaction as resources for social action in their discursive environment, and for detecting ongoing processes

68

Nadine Thielemann

of language change which may result in the emergence of new linguistic patterns and units. In this paper, we adopt such a perspective on constructions with govorit’ (‘to say’) in Russian talk-in-interaction. Govorit’ is an interesting test case for mainly three reasons. First, it is the second most frequent verb in the variety of colloquial Russian according to statistical analysis conducted by Ljaševskaja / Šarov (2009) based on the Russian National Corpus. Second, it occurs in a variety of constructions. Third, as a complement-taking predicate, it is prone to undergo grammaticalization (grammatikalisierungsanfällig in Imo’s (2007) terms). The first two arguments secure that a sufficient number of examples can be extracted which represent each linguistic pattern employed for the realization of a specific social action or interactional task. The third argument has already been attested for in other complement-taking predicates (znat’ ‘to know’, ponimat’ ‘to understand’) of Russian. In this vein, Freidhof (1996) shows how their 2nd person present forms (znaeš’ ‘you know’, ponimaeš’ ‘you understand’) in specific positions in the turn and the sequence in Russian talk-in-interaction evolve into discourse markers for the segmentation of utterances and for the management of interaction. As a grammatical framework for the analysis and description of the grammar of govorit’ in spoken language, we promote a combination of Interactional Linguistics and Construction Grammar. This approach allows for the integration of pragmatic aspects (e.g. interactional, discursive, or sociolinguistic features) in the analysis and description of linguistic structures of talk-in-interaction. Construction Grammar can easily be turned into a grammar which is “semantically, pragmatically and discursively informed – a discourse grammar of sorts” (Östman 2015, p. 18). In the main part of this paper, we analyze the discourse grammar of govorit’ in Russian talk-in-interaction from this perspective. Due to the limited space, however, we will focus on a particular group of constructions, by which interlocutors indicate that they render speech from another source or situation and thus shift footing (Goffman 1981, pp. 144–146). These constructions form a kind of network within which each one has a specifically nuanced communicative function (i.e. highlights a specific participation role) and occurs in a specific discursive environment (e.g. genre, speech activity or social action, sequential context). The ORD corpus provides rich data in this regard since it also grants access to the sequential and wider discursive context in which a construction occurs. The data allow for the identification of the specific and recurrent communicative task for which interlocutors employ a construction in talk-in-interaction from an emic point of view. ORD-data also reveals that there are current shifts occurring in the network: ‘new’ constructions are emerging in order to meet specialized

Constructional Patterns Emerging in Talk-In-Interaction

69

communicative needs in interaction. These shifts testify to ongoing processes of language change which cannot (yet) be classified in terms of grammaticalization, pragmaticalization or lexicalization. Both the data and analytic perspective thus also introduce a fresh perspective on a long-term discussion in Russian linguistics, namely whether shortened 3rd person present forms of govorit’ such as garit, grit or even gyt inserted in rendered speech develop into evidential markers or merely count as allegro forms. Against the background of our analytic approach, we argue that these forms occur in constructional patterns which emerge primarily as interactional solutions to meet specific communicative needs. The structure of the paper follows this line of argumentation. In section (2), we discuss the capability of Valency Grammar and Construction Grammar to account for discourse-pragmatic aspects influencing verb grammar in talk-in-interaction. In this context, we show that a version of Construction Grammar which includes ideas from Interactional Linguistics provides a promising framework for the analysis of the discourse grammar of govorit’. In section (3), we present the results of a preliminary analysis of a group of constructions with govorit’ based on the ORD corpus. Prior to this we will shortly introduce Goffman’s concept of footing since this serves as the umbrella under which these constructions can be gathered. We do not raise the claim, however, that the discussed patterns cover all constructions for shifting footing in Russian talk-in-interaction. In the closing section of the paper, we point to the potential of this analytic perspective in order to appropriately describe ongoing processes of language change in terms of emerging constructions.

2. Verb grammar meets discourse In this section, we discuss two frameworks for the analysis of the discourse grammar of verbs which mainly differ in their perspective. Valency Grammar starts from the verb (or another lexical unit, e.g. noun or adjective) and analyzes its capacity to bind arguments. Construction Grammar, in contrast, works on the assumption that constructions are the core units of grammar and considers argument structure to be such a construction (in which verbs are inserted). Both approaches are scrutinized for their potential to account for the impact of pragmatic aspects on verb grammar and in a wider perspective also for their applicability in the analysis of the linguistic structures of talk-in-interaction.

2.1. Valency Grammar and Pragmatics Approaches based on the idea of valency aim at describing the patterns with which verbs can be combined. They determine the semantic role and the syntactic

70

Nadine Thielemann

form of verbal arguments, and they are interested in how both are mapped onto each other (argument linking). This concept has inspired valency dictionaries such Apresjan and Pall (1982) as well as the FrameBank project () which gathers lexical constructions (Lyashevskaya / Kashkin 2015) of verbs based on the Russian National Corpus. Both have entries for the patterns combining with govorit’ but do not pay special attention to pragmatic or frequency aspects influencing argument realization. They mainly aim at providing comprehensive lists of valency patterns for each entry. Interactional Linguistics reinterprets verb valency, i.e. the capacity of verbs to bind specifically realized arguments in terms of projection (Auer 2005). Once delivered in an utterance, a verb (form) raises expectations towards specifically shaped arguments to be produced. Auer argues that interlocutors orient to these projections and, for example, time and possibly suspend turn-taking until the projected unit is delivered. In this sense, argument structure provides a linguistic resource for the organization of interaction. Pragmatic or communicative valency, however, seems to be an underdeveloped concept. The scant research work that deals with pragmatic aspects influencing the realization of arguments mainly pivots on issues of information structure or management. Pioneering work on pragmatic valency by Růžička (1978), for example, deals with conditions for the omission of actants. He argues that, if an actant can be retrieved from discourse context since it is included in a scene evoked by a prior sentence, the argument can be skipped in order to avoid communicative redundancy. Welke (1989) links pragmatic valency with issues of information management, and points to its impact on the serialization of arguments. For him, pragmatic valency refers to (a) the order in which the arguments are arranged (with special focus on the argument serving as subject and having the status of the topic) and to (b) the strength with which an argument is presupposed (which lends the scenario a specific perspective). Lastly, DuBois (2003) argues that there is a systematic pattern of correlation between the grammatical realization of an argument (lexical vs. non-lexical) and its information value (new vs. already introduced and accessible referent). His concept of preferred argument structure is spelled out into four discourse constraints on the realization of arguments: – – – –

Avoid more than one lexical core argument Avoid more than one new core argument Avoid lexical A (i.e. transitive subjects) Avoid new A (i.e. transitive subject)

(Du Bois 2003, p. 34)

Constructional Patterns Emerging in Talk-In-Interaction

71

These constraints trace back to observations made in corpora of spoken language. They have been determined on the basis of frequently observed patterns and thus represent a kind of discourse norm or pragmatic standard (Proske 2013). The analyses included in DuBois / Kumpf / Ashby (eds.) (2003) show that these patterns can be observed in discourse data from typologically quite different languages. In doing so, they provide backing for DuBois’ (2003) claim that preferred argument structure is a universal discourse constraint. Turk’s (2000) analysis of casual conversations conducted in the variety of colloquial Russian – though not explicitly referring to the framework and terminology of preferred argument structure – provides empirical evidence that these constraints also regulate information management and argument realization in Russian talk-in-interaction. In summary, approaches to verb grammar based on the idea of valency predominantly pursue lexicographic aims (i.e. gathering valency patterns for each entry). Pragmatic aspects affecting the realization or serialization of arguments in discourse do not feature prominently, though information management is acknowledged as an important factor. Especially the concept of preferred argument structure and latest work based on data from Russian (Turk 2000) or German (Proske 2013) talk-in-interaction show a direction in which ideas inspired by Valency Grammar and discourse pragmatics can cross-fertilize each other in the service of discourse grammar. However, other pragmatic factors (e.g. activity, genre) and their possible impact on the argument structure are beyond the scope of interest of analyses that rely on the idea of valency. In the next subsection, we shift the perspective to approaches which focus on constructions as core units of grammatical analysis and show how they can be fruitfully combined with ideas from a distinctly pragmatic framework for the analysis of linguistic structures of talk-in-interaction (i.e. Interactional Linguistics).

2.2. Construction Grammar and Interactional Linguistics Construction Grammar (further CxG) – while appearing in various versions (cf. Ziem / Lasch 2013; Croft / Cruse 2004) – essentially relies on the idea that constructions, defined as conventionalized form-meaning pairings, form the core unit of grammar. Constructions are conceived of as cognitive gestalts. They represent holistic units, the meaning of which cannot be derived in a compositional fashion from their parts. Consequently, it is assumed that grammatical knowledge is organized construction-wise. Constructions can vary both in size and in complexity. They can range from schematic constructions (such as the formation of the plural in the English noun N-s) to idiomatic constructions. In her seminal work introducing CxG, Goldberg (1995) argues that argument structure, too, forms

72

Nadine Thielemann

such a construction. Goldberg argues that constructions such as the transitive or ditransitive construction provide the base pattern and superimpose the meaning. Accordingly, verbs which meet the semantic coherence principle can be inserted in these constructions. A construction as a symbolic unit is characterized by a form pole including syntactic, morphological and phonological features and by a meaning or function pole including semantic, pragmatic – and lately also discourse-functional features of a construction (Croft / Cruse 2004, p. 258). Constructions owe their role as cognitively entrenched units of grammar essentially to their non-compositional character. This criterion was later mitigated by Goldberg (2006) herself who now also considers usage-based entrenchment for highly frequent though rather regular form-meaning pairings (cf. Bybee / Beckner 2010). Nevertheless, CxG understands itself as a cognitive framework explaining how grammatical knowledge is mentally represented. The format of this mental representation is referred to as constructicon, a blend from construction, and lexicon. The structure of this inventory or network of constructions remains unclear at the given stage of theory development, however (Imo 2015, p. 71). CxG has by now developed into a comprehensive model of grammar, although it began as a framework mainly interested in semi-idiomatic constructions and other phenomena situated on the ‘periphery’ of grammar (cf. malyj sintaksis in Russian linguistics). Lately, CxG has also been adopted for the analysis of discourse grammar, especially for the grammar of talk-in-interaction (e.g. Deppermann 2006; Fried / Östman 2005; Günthner 2009; Östman 2015; Imo 2015). The papers collected by Bücker / Günthner / Imo (eds.) (2015), or Fischer / Nikiforidou (eds.) (2015), for example, show that CxG can be developed further in a way which allows for the analysis of conventionalized patterns ‘above the sentence’ in terms of constructions. They also illustrate that pragmatic functions and sociolinguistic meanings can be included in the semantic and functional features of a constructional analysis, and that multimodal as well as interactional features add to the form pole. In doing so, they turn CxG into a grammatical framework for the analysis of communicative patterns, in other words, into a “full-fledged model of communication” (Östman 2015, p. 15). This line of research has been enabled by a combination of CxG and Interactional Linguistics, which is also termed “Interactional Construction Grammar” (Imo 2015). In the remainder of this subsection, we show how and why both approaches complement one another so well. Interactional Linguistics (further IL) is a label for research located at the interface of Conversation Analysis and Linguistics. It refers to empirical analyses of

Constructional Patterns Emerging in Talk-In-Interaction

73

linguistic structures of talk-in-interaction in terms of resources for the solution of recurrent interactional tasks and studies “grammar as sets of practices adapted to social action” (Ford / Fox / Thompson 2014, p. 119). These recurrent tasks can refer to discourse organization, the management of turn-taking, the accomplishment of speech actions or they can play a crucial role in the constitution of genres. Analyses can start from a grammatical format and detect the social action regularly accomplished by it; analysis can also start from a recurrent interactional task and determine the linguistic structures which interlocutors repeatedly employ for their solution. Irrespective of the direction, linguistic structures or grammatical formats of talk-in-interaction are analyzed as conventionalized communicative patterns for the solution of recurrent tasks, or as “a set of forms, patterns and practices that have arisen to serve the most recurrent functions that speakers need to fulfil” (ibid., p. 122). Both frameworks, CxG and IL, reveal a number of commonalities regarding the phenomena focused upon and the data relied on to facilitate their combination. The most important points of intersection as highlighted by Imo (2015, pp. 71–72) and Fried / Östman (2005, pp. 1754–1755) are as follows: – Both approaches deal with ‘pre-patterned chunks’ and can in general also be applied to chunks above the sentence. – Both treat them as conventional patterns of usage in which specific formal features are associated with specific meanings or functions. – Both are usage-based, i.e. they rely on or stress the role of corpora. – Both share an interest in non-standard linguistic structures and make no difference between core and periphery. – Both are surface-oriented. The main difference between both frameworks lies in the status of their units. “While construction grammar focusses on the psychological entrenchment of linguistic units, Interactional Linguistics focusses on the description of empirically attested routines in interaction” (Imo 2015, p. 72). This, however, does not prevent them from cross-fertilizing each other. An interactionally enriched CxG provides a framework for the analysis of possibly complex communicative patterns which are conceived of as “conventional pattern[s] of usage in which form and meaning / function are associated in particular ways” (Fried / Östman 2005, p. 1754). In order to arrive at a “communicatively and grammatically adequate treatment ((…)) in a CxG framework” (ibid., p. 1753), however, pragmatic and discursive as well as multimodal and even sociolinguistic features characterizing constructional patterns that emerge in talk-in-interaction must be systematically included in the

74

Nadine Thielemann

constructional analysis and description. Östman (2015, pp. 25–35) provides a very detailed list of such “external features” or “contextual attributes” and even integrates them in the matrix or box notation typical of most grammatical analyses within CxG proper (cf. Fried / Östman 2005). Here however we restrain ourselves from this practice and follow Imo’s (2007) and Deppermann’s (2006) lead. This means we provide examples from our data which are particularly representative of and clearly illustrate each constructional pattern, and we describe the characteristic formal and functional features of the construction in words. This description also includes prosodic, stylistic and expressive features, genre and activity related attributes and information about the position in the turn, the sequence or the discursive episode. In doing so, we arrive at a pragmatically and discursively informed grammatical analysis which “includes knowledge of communicative patterning” (Fried / Östman 2005, p. 1754) and integrates it in the constructional analysis.

3. Constructions with govorit’ in order to shift footing The ORD corpus provides a unique source in order to analyze the discourse grammar of govorit’, and Interactional Construction Grammar offers a suitable toolkit for the analysis of the patterns in terms of constructions. The constructions attested for govorit’ in ORD range from semi-idiomatic ones such as govorit’ tuftu ‘talk shit’ or govorit’ erundu ‘talk rubbish’ to discursive patterns in which forms of govorit’ meta-communicatively frame reformulations of prior claims. Due to limits of time and space, however, we are not able to present the entire discourse profile of govorit’ in Russian talk-in-interaction including all constructions typically associated with this verb. Instead, we focus on a particular family of constructions with govorit’ by which interlocutors detach themselves from talk and attribute it to another source or situation. In traditional terminology, some of these patterns could be described as direct or indirect speech. Some patterns, however, do not match these categories because they, for example, merely relate talk to another source in a very general sense, or mark it as general knowledge. Due to its inherent semantics as a verbum dicendi and the overall function of relating information to a source, an alternative common denominator for this group of constructions could be evidential marking (hearsay). Yet this concept seems to be too narrow and specific. Yet lately several scholars interested in discourse (including talk-in-interaction) have developed similar yet more comprehensive concepts such as stance (Fox 2001; Thompson 2002; Clift 2006; DuBois 2007; DuBois / Kärkkäinen 2012) or epistemological positioning (Bednarek 2006). These concepts combine evidential

Constructional Patterns Emerging in Talk-In-Interaction

75

meanings with affective or emotive and evaluative meanings and also include non-grammaticalized, i.e. discursive and sequential strategies for the expression of evidential stance. Bednarek (2006) develops an analytical grid which distinguishes the source of knowledge (self vs. attribution to other) and the basis of knowledge (perception, general knowledge, proof, obviousness, unspecified, hearsay, mindsay). Her model is very elaborate and allows for capturing more fine-grained functional differences than those discriminating the constructions focused on here. For the given purpose, Goffman’s (1981) concept of footing provides an appropriate framework to characterize the specific function or meaning of each of the analyzed constructions. Footing refers to the stance speakers take towards the content and form of their talk. In this context Goffman (1981, pp. 144–146) distinguishes several participation roles which correspond to different forms of responsibility for one’s talk. The principal is the originator of the ideas promoted and takes social responsibility for the talk. The author accounts for the phrasing, i.e. for the words in which the ideas are couched. The animator merely lends their voice in order to articulate the principal’s ideas and the author’s words. Per default, all participation roles are unified. The dissolution of this unity represents a deviation from this rule and requires specific marking. The constructions under scrutiny all allow interlocutors to shift footing. Within this functional domain, each construction serves for highlighting one particular participation role (principal, author or animator) and occurs in a distinct discourse environment (activity, sequence, genre etc.). The group thus forms a “usage-based network” (Imo 2015, p. 71) within which each construction is tailored to meet a specifically nuanced communicative need. In the remainder of this section, this network will be reconstructed by thoroughly describing the formal features characterizing each construction and by identifying the social task or practice regularly dealt with by each construction.

3.1. Matrix clause [with a form of govorit’] + CP [čto + clause] The first construction is traditionally referred to as indirect speech. According to the Academy Grammar (II, 1980), it is conventionally analyzed as a matrix clause with a form of govorit’ and a subordinated (dependent and embedded) clause introduced by the subjunctor čto. Syntactically, we thus deal with an instance of complementation, i.e. “the syntactic situation that arises when a notional sentence or predication is an argument to a predicate” (Noonan 1985, p. 42). This analysis has been challenged from several perspectives; this no doubt has been nurtured particularly by analyses of complementing in talk-in-interaction from

76

Nadine Thielemann

different languages (e.g. Thompson / Mulac 1991, Thompson 2002, Englebretson 2003, Hansen / Letučij / Błaszczyk 2016). Verhagen (2001) argues that we do not deal with semantic-conceptual subordination in these cases and suggests that the matrix clause should rather be analyzed as a space builder, framing the following clause as something that is said (thought, claimed etc. – depending on the semantics of the verb). Based on observations made in English talk-ininteraction Thompson / Mulac (1991) and Thompson (2002) similarly dispute the conceptual independence of the matrix clause and suggest a reanalysis in terms of epistemic/evaluative/evidential fragments. As for Russian talk-in-interaction, Hansen, Letučij and Błaszczyk argue that “when the complementizer is omitted, the (originally) embedded clause becomes the pragmatically main one for the speaker, and the matrix predicate serves as a parenthetical / modal word” (2016, p. 116). In a more general fashion, Hopper / Thompson favor an interpretation of such biclausal constructions “as relatively open-slot ‘pre-fabs’ which strongly project certain types of social actions, whose grammatical manifestation takes many different forms” (2008, p. 116). ORD data, too, support these objections. In the remainder, we suggest an analysis in terms of semi-schematic constructions for the shift of footing. We will show that in Russian talk-in-interaction two such biclausal constructions emerge and that each of them focusses a specific participation role. The first construction is used in order to highlight the participation role of the principal. It allows interlocutors to render talk originating from another person which usually consists of one statement. Neither wording nor stylistic or expressive features of the quoted utterance are preserved. The utterance is rendered by the current speaker in a neutral way and deictically aligned with their perspective. The construction consists of two parts – a clausal unit with a form of govorit’ and a clausal unit introduced by čto. The first part introduces and projects the rendition of an utterance and attributes it to a source, and the second part delivers the rendition. Both parts can be realized within one prosodic contour as illustrated in examples (1a) and (1b) or with a cesura and thus in two separate prosodic units. The cesura can follow the complementizer as in example (2) or precede it as in example (3). Irrespective of the prosodic packaging, interlocutors suspend turn-taking until the delivery of the rendered statement. The insertion by M1 in example 3 merely offers a repair candidate which is triggered by I1’s word search in the end of line 1. It does not prevent I1 from completing the initiated construction. This confirms that interlocutors orient to the format; it also serves as interactional evidence for the constructional status of the pattern (cf. Günthner 2014).

Constructional Patterns Emerging in Talk-In-Interaction

77

(1a) ORD S20_02, Olja (Ol), Lena (Le)1 1 2 3

Ol :

4 5

=hehehehe; (2.2) ((phone rings)) onamneteper’govoritčtoves’piterboleettuberkulёzom;  she says to me now that the whole of Saint Petersburg suffers from tuberculosis (1.2) nu;=i sootvetstvenno, (-) well and accordingly

(1b) ORD S20_03, Lena (Le) 1

Le :

2 3



äh vo:t;= äh well =i ( )onagovoritčtouneёtamtakojkarmašekpodrulёm,= and ( ) she says that she has such a case under the steering wheel =i ja vsё vremja v čechol tuda kladu;= and I always put in that case

(2) ORD S03_04 1

F1:

2



3 4 5

I3:

6



o čto u nee nu muž /(a)/starše ee byl v dva raza / t that she has well a husband/ (and)/ he was twice as old / to est‘ takoj ( ) / that is such a ( ) aonagovorilačtovoobšče/kak-to/ and she said that in general / somehow / neočen‘prijatnyjmomentbyl// it wasn’t a very pleasant moment// ponjatno // da / uvy / voobšče situacija I see // yes / oh my god / generally the situation dostatočno ščekotlivaja / (…) is sufficiently delicate / (…)

(3) ORD S2_01 1

I1:

(…) vyeščetogdagovorili/čtoontakoj/*P (…) still then you said / that he is such a / *P somnitel’nyj dubious složnyj/ complicated/difficult da / on složnyj / ja pomnju ego / ja pomnju yes / he is complicated / I remember him / I remember

2

M1:

3 4

I1: M1:

1

The ORD team has transcribed the speech data according to the transcription conventions established by Zemskaja (1978) and her team in compiling a first corpus of colloquial Russian. Recording ORD_S20 has been transcribed by the author according to the transcription conventions established by Selting at al. (1998) for Conversation Analysis and Interactional Linguistics given at the end of the paper. It has not been included in the set of transcribed recordings which was available at the beginning of the Urban Voices project.

78

Nadine Thielemann

The construction is semi-schematic since it imposes very little formal restrictions of the realization of its parts. The first unit or the matrix clause, however, is characterized by several morphosyntactic features: It mostly contains a present form of govorit’, though past tense also occurs in the data. The subject is predominantly realized by a pronoun, though there are view instances with names or other nouns. The matrix clause can include an indirect object or an adverbial modification. The second unit is introduced by the complementizer čto and contains the neutralized statement projected by the previous unit. In providing what has been said, it fills the valency slot of govorit’ in semantic terms. The data clearly show that the information conveyed by the CP represents the main point which is an essential pragmatic feature of this construction. In contrast to observations by Hansen, Letučij and Błaszczyk (2016, p. 116), this effect does not seem to depend on the omission of the complementizer. In examples (2) and (3), it provides the information to which interlocutors react, and in examples (1a) and (1b), the information which is elaborated on in the rest of the turn. The CP is not subordinated, at least in pragmatic terms (cf. Auer 1998). Moreover, it maintains the discourse continuity (cf. Auer 2002, Boye / Harder 2007). The overall function of this construction as derived from its usage in talk-in-interaction is the marking of a statement as something that has been said and the attribution of this to a particular principal.

3.2. Matrix clause [pronoun + present form of govorit’] + [presented speech] This construction is traditionally referred to as direct speech. From the perspective of Interactional Construction Grammar, it is a semi-schematic construction for the presentation of speech, i.e. a construction by which interlocutors highlight the participation role of the animator. It is used in order to render possibly longer stretches of talk that mimicks stylistic, expressive, and occasionally even vocal features. The current speaker adopts the role of the talking figure, deictically shifts the origo to this figure and stages their utterance(s). Formally, this construction also consists of two parts. The matrix clause which introduces the presented speech and ascribes it to a figure, and the actual speech presentation. The matrix clause includes a present form of govorit’ and an overt (pronominal) subject. Other elements such as indirect objects or adverbial modifiers are mostly absent. The matrix clause usually also forms a prosodic unit of its own and is followed by a cesura. As for the second part, there are almost no formal restrictions except for the pragmatic (deictic shift), stylistic, expressive, and prosodic (shift of vocal deixis) means employed in the presentation. Interlocutors,

Constructional Patterns Emerging in Talk-In-Interaction

79

however, tend to mark the beginning of the presented talk in specific ways as illustrated in examples (4a), (4b) and (5). Very often typical turn-initial elements such as forms of address (e.g. marin, olja, vova), discourse markers (e.g. a nu-ka) or routine formulae (e.g. dobroe utro) indicate the shift. The rendered talk itself can consist of several turn constructional units of presented speech as in examples (4a) and (4b). The examples also show that interlocutors orient to the format since they suspend turn-taking until the end of the presented speech. Potential transition relevance places (TRPs) are contextualized within the speech presentation. Due to the specific footing shift (highlighting the participation role of the animator and focusing the act of uttering) and the formal character of the second part (possibly longer stretch of presented speech which is mimicked), the matrix clause in this construction projects the second part, and does so merely pragmatically and not so much semantically or syntactically. In this construction, the matrix clause (including the form of govorit’) much more resembles the evidential fragments (Thompson 2002) or pre-fabs (Hopper / Thompson 2008) into which similar constructions in talk-in-interaction from other languages are claimed to develop. The first part of the construction including a present form of govorit’ merely indicates that speech will be presented. The syntactic and semantic format of what is following is not further specified or related to the introductory part but definitely conveys the main point, i.e. the information which is relevant to the interlocutors. (4a) ORD S12_24 1 2 3

F1: I12:

4



5



6



(…) v čem delo? (…) what’s up? to est‘ mne ka… (…) jagovorjumarin / that means it seems to me.. (…) I say Marina / umenjavot(ää)mneočen‘neponravilas‘vot I have well (ä ä) I very much did not appreciate well ėtasituacija/da? the situation / did I? u menja k nej očen‘ sil‘no (…) otnošenie izmenilos‘// my relationship with her strongly changed// ((tells about the their better relationship in the past))

(4b) ORDS S13_01 1 2

F1:

3 4 5

I13:

(…) vidno ej delat‘ bylo tože nečego / (…) obviously she had nothing to do / ionagovorit/ and she says / anu-kaolja/jatebedavalaspisok// and come on Olja / I’ve given you a list/schedule nu-kaotčitajsja// come on report // no ona takaja viš‘ / vse-taki strogaja // da? but she is such a you see / anyhow demanding // isn’t she?

80

Nadine Thielemann

The construction can also be used in order to stage a whole dialogue and thus to present the speech of the involved characters. This is illustrated in example (5) when the current speaker re-enacts something that has happened to her earlier. The matrix clause introduces the speech presentation and attributes it to each figure. (5) ORD S30_15 1 2 3

I30:

4 5 6 7



ėtot (…) vova / kotoryj pozvonil mne / v dvenadcatom času/ this ( ) Vova / who called me / at twelve o’clock/ a mne-to vstavat‘ / v pjatom *P / and I have to get up / at five *P on govorit / čto ty delaeš‘ ? he says / what to you do? jagovorju/ vova / ja splju // I say / Vova / I sleep // on govorit / dobroe utro // he says / good morning // *P parazit // sponger //

The data show that interlocutors feel the need to present speech in specific discursive contexts from which further pragmatic features of the construction can be derived. As illustrated in the examples above, Russian interlocutors employ this format as an evidential strategy in order to back an otherwise disputable claim or assessment. The social “practice of giving […] evidence for what one claims”, however, can be realized in different ways “where a crucial source of difference has to do with claims to authority, responsibility, and entitlement” (Fox 2001, p. 174). In highlighting the role of the animator and in re-enacting an utterance (or dialogue) instead of merely rendering the information (or re-telling the story), the speaker displays what they want to convey. Couched in the terminology of Clark / Gerrig (1990), they provide a demonstration of it. Clift therefore refers to reported speech as a “powerful evidential display” which allows speakers to display (not claim) epistemic priority (i.e. “having reached [an] assessment first”) (2006, p. 580). Russian interlocutors rely on this construction format in order to grant their audience immediate access to the evidence for their claims. In (4b), the demanding character of the staged figure Olja is displayed by mimicking her habitual way of delivering orders. The interlocutor in line 5 immediately aligns with this and explicitly confirms the assessment (no ona takaja viš‘ / vse-taki strogaja // da?) illustrated in the prior turn. In (5), the staged or cast dialogue displays Vova’s nasty character or way of acting in staging a scene in which he behaves inappropriately (instead of re-telling the incident). Here, the current speaker (I30) herself in line 7 delivers the thusly-illustrated assessment – he indeed

Constructional Patterns Emerging in Talk-In-Interaction

81

is a ‘sponger’ (parazit). As an evidential strategy, the given construction allows for the creation of involvement which favors emotive alignment.

3.3. 3rd person present forms of govorit’ inserted rendered speech In Russian talk-in-interaction interlocutors have yet another linguistic resource at their disposal which allows them to shift the footing of an utterance. These are often shortened 3rd person present forms of govorit’ which are inserted in rendered talk in a seemingly unsystematic way. As already indicated in the introductory section, there is disagreement whether these forms have already developed into an evidential or quotative marker (e.g. Hinrichs 1989, Grenoble 1998, pp. 142– 143, Kopotev 2014, Kozinceva 2007) or whether the phenomenon is due to the fast speech rate and thus merely represents allegro forms (e.g. Bogdanova 2008, Wiemer 2008). Several features speak in favor of the first position. Next to govorit forms such as garit, grit or even gyt with an erosion of the stem morpheme occur. The forms behave like clitics which are attached to a preceding unit. They are syntactically not integrated in the utterance in which they are inserted. Govorit’ in these usage contexts loses its external syntax. Alternatively, these forms can be interpreted as parentheticals which meta-communicatively indicate that the surrounding utterance is speech by someone else (cf. Hinrichs 1983, Grenoble 2004, Sonnenhauser 2009). In this usage context, the verb form mainly has an indexical function that testifies to semantic bleaching. At least in ORD, the inserted verb forms are, however, still marked for flexion which prevents them from being reanalyzed as pragmatic markers (quotative marker or evidential particle). The overall picture therefore points to an ongoing process of language change and not to a result of grammaticalization, lexicalization, or pragmaticalization. In this section, we adopt a different perspective on these forms and suggest an alternative account of the ongoing process of language change. ORD data show that these forms of govorit’ are not always randomly inserted, but that patterns emerge in Russian talk-in-interaction which are tied to specific communicative contexts. We argue that these patterns represent linguistic formats by which speakers stress that they are not socially responsible (principal) for their talk. In the following subsections, we sketch two patterns and analyze the discursive and sequential contexts in which they are frequently met and which allow for the identification of further functional features that characterize and distinguish them from each other. In doing so, we adopt a perspective on the above-sketched process of language change which conceives of the process as the entrenchment of constructions (including communicative patterns at the discourse level) – and not as the re-analysis of single units. At the same time, it

82

Nadine Thielemann

should be stressed that these two patterns do not cover all uses of the shortened and inserted forms of govorit’.

3.3.1. [pronoun + govorit’] / [unit of presented speech]=[govorit’] / [rest of presented speech] The first pattern is typical of conversational narratives in which tellers unfold their stories in a dialogic fashion. The teller re-enacts and does not re-tell the plot. The presentational mode resembles mimesis rather than diegesis. When telling stories in conversation, interlocutors stage dialogues, adopt the perspective of each figure and render their talk. In doing so, the teller displays their own involvement and creates interpersonal involvement with their audience (cf. Tannen 1989). In extract (6), speaker I19 relies on this pattern (or construction) in presenting or re-enacting a dialogue with the tech support at her office. It essentially contributes to the vivid display of the problems and lends the presentation an emotive flavor. In constructional terms, we deal with a pattern ‘above the sentence’ or a communicative pattern describing the arrangement and form of several turn constructional units. The first part is represented by an introductory clause containing a pronominal subject and a present form of govorit’, followed by a first chunk of presented speech to which a present form of govorit’ is immediately attached. This part is then followed by the remaining parts of the rendered speech. The first segment of presented speech often consists of typical turn-initial elements such as vocatives (marin) or form of address in general, discourse markers or tags (vy znaete) as illustrated in example (6a). This pattern is similarly also attested in German talk-in-interaction (Imo 2007, pp. 91–93). Both constructions only differ in the realization of the second form of to say/sagen/govorit’. Colloquial Russian allows for pro drop, whereas German does not. (6) ORD S19_04 1 2

I19:

3 4



((…)) 5 I19: 6



on ves‘ na odnom ajpi adrese i poėtomu it is all on one ip address and that’s why ona govorit marin govorit / to čto u vas she says Marina (she) says / that in your group v pomeščenii sejčas nikto ne skačaet ne fakt in the office now nobody downloads is wrong čto tam / Šarena / kto u vas ešče možet skačat‘ / that there / Šarena / who else from your group can download / ta govorit vy znaete govorit / ja tam ne znaju she says you know (she) says / I don’t know there nikakoj mariny / nikakoj nadi / ((laughter)) i tak dalee // any Marina / any Nadja / ((laughter)) and so on //

Constructional Patterns Emerging in Talk-In-Interaction

83

3.3.2. [first constituent of a rendered utterance]=govorit / [rest of rendered utterance] Conversational narratives can as well be less emotive, but their plot, too, unfolds in a dialogic fashion. In high involvement story-telling, interlocutors rely on speech presentation (including the reproduction of prosodic and stylistic features) and often also display their emotive stance as in (6). In this mode, there are sufficient signals that talk does not originate from the teller but that it is attributed to a specific character, even if markers of deictic shift are absent. When the mode of telling is rather detached, such cues are missing. Then interlocutors meet the communicative need to indicate that a stretch of talk does not originate from them (i.e. that they are not the principal) as well as the need to attribute it to another source. This is illustrated in (7). F2 stages a dialogue between herself and a colleague. She starts her turn in line 1 rendering an utterance which she has produced herself in the staged situation (“jagovorju / a čto ty chotel ?”). Then (line 2) she adopts the role of the colleague who wanted to get access to free medical treatment (“god govorit besplatno chodit‘ / k zubnomu lečit’sja //”). There are hardly any prosodic or expressive cues in these turn constructional units and also no signals of a deictic shift which facilitate the interpretation as a replica which is attributed to another figure of the re-enacted story. F2 thus communicatively needs to indicate that she is not socially responsible for this line, and that this is rendered talk by her colleague. This is accomplished by govorit inserted after and attached to the first constituent of the turn constructional unit rendering what the colleague has said. (7) ORD S09_06 1

F2:

2



3

Fch:

jagovorju/ a čto ty chotel ? *P I say / and what did you want ? god govorit besplatno chodit‘ / k zubnomu lečit‘sja // a year (she) says get free treatment at the dentist a počemu imenno god? and why exactly/of all things a year ?

Even if speakers start their narrative by strongly relying on speech presentations and extensively using prosodic, stylistic/pragmatic, and/or expressive cues, the need to indicate that they are not the originator of their talk (i.e. principal) may arise from time to time. It is often difficult to maintain this intense mode of telling across several sequences, since it requires constant mimicking of speech features. As shown by Bolden (2004), in conversational Russian the end of such extended quotes or reported speech is often not clearly marked, especially when the rendered talk is longer. She labels this phenomenon as fade out. Longer narratives thus form another discourse environment in which interlocutors insert shortened 3rd person present forms of govorit’ in order to mark that they still render talk attributed to

84

Nadine Thielemann

someone else, i.e. talk for which they are not socially responsible. This is the case in the next extract. Example (8) nicely illustrates how a longer conversational narrative is actually performed and not just told (cf. Bauman 1986). Daniil, the current speaker, tells the story mainly in mimicking what the main character, a former KGB-officer who is now their teacher, has said in the reproduced or staged situation. At the beginning, he highlights the role of the animator and puts much effort in adopting the formal register (officialese) of the teacher’s speech, depicting him as a typical representative of the cadre. Especially at the initial stage of the performance, this even includes a shift of the vocal deixis. Daniil presents speech attributed to this man with a much deeper voice. These features, however, become weaker in the course of telling. As shown by linguistic anthropologist Richard Bauman (1986) the point of an oral narrative is as a rule couched in direct speech attributed to and delivered by a character of the story world. This is also the case in (8). At this stage of the telling, however, other signals marking the shift of footing are already fading out. Additionally, there are recipiency tokens inserted by which Daniil’s interlocutors Ol’ga and Lena signal that they actively listen and align with the presentation (lines 15–17). They do not claim the floor but they stop the flow of the narrative at least for a moment. Upon the delivery of the point (or punch line) in lines 18–21, Daniil, thus additionally indicates that this is talk originating from the officer and not from himself – “, 

((lautes Aus- und Einatmen))  (0.06) 

2

As a matter of fact, the difference between assumptions, presuppositions and conversational, conventional implicatures and indirect speech acts cannot be addressed here in detail for space, but I refer the reader to Kosta (2011) where these differences are somewhat well explained.

190

Peter Kosta

[38] [04:28] [04:28] KS

[04:29] [04:30]

[04:31] [04:31]

(üb'ėrTRAGbarnyj/üb'ėrTRAGbarnoj); 

DA; =ėto:_u SAM aban'im'

VA

da ėta ZNAč'it;  (0.37) 

(0.1) 

(0.32) 

[39] ..

[04:35] [04:35]

[04:36]

KS ėnt id'ot at fau bė bė– 

[04:38]

[f'ėrK'ĖR b'ėrl'in brand'ėnburk; (.) NO ] :

VA

N'ĖT nu (.) ŠT [O ty (šč'itaiš); kak ty JĖchat’ bud'iš_ta; ] (0.08) 

[40] [04:39] [04:39] KS

[04:42]

[04:42] [04:44]

ta!R'IF! (.) b'ėRAJCH u m'in'a b'ėrl'in a bė; 

VA

tar'ifb'ėRAJCH b'ėrl'in a bė; A; (.) okėj; 

(0.38) 

(1.27) 

5. Multiple self-repair and other-repair In our example (5), we want to show another lovely example of so-called multiple self-and other-repair. The daughter (VA) seems to have difficulties with the Genitive Plural form of the word noskí, but in fact, it is the mother who is wrong and initiates after the first self-repair an other-repair which leads to a very paradoxical situation. (5) Multiple ‘self-repair’ and ‘other-repair.’ [41] .. [04:45] [04:46] [04:49]

[04:49]

KS VA

[ėta !TO!žė m] ai nask'i;  (0.38) 

[42] .. KS

[04:51]

[04:52]

[04:52]

!N'ĖT!; =ėta MAi na [sk'i; ]

VA

[ėta MA] i– = ja kup'ila SĖryi z'ėl'onyi i s'in'i:; = v adnoj !PAČ'!k' (0.18) 

[43] ..

[04:58] [04:58]

KS

[05:00]

((lacht)) °h 

VA i; = za dva !JĖV!ra; =vot d'iv'atnacat' vos'imd'is'at dva v aL'ĖKs'ė;  (0.17) 

(0.53)

191

Third Position Repair, Overlaps, and Code-Switching [44] .. [05:01] KS

[05:03]

[05:04]

[05:05] [05:07]

va:bšč'ė_ta ėta pa_moimu !MA!i;  [no už jėsl'i ėta t] vai to zab'iR'I ich s'ib'ė; 

VA

[ėta od MAi. ]

ja i 

 

(2.04) 

[45] .. VA TAK pastajan: t'ėr; (.) ja naD'Ėjus' što v abšč'ižyt'ii bud'it pa_druGOmu; tam ja s'ib'ė kupl'u  [46] ..

[05:13] [05:15]

KS VA iznaČ'AL'na (.) n'ė znau; (.) d'ės'at' naSOK, 

[05:15] [05:16]

[05:17]

I, 

[naSOK; ]

nasKOF; (.)

i pasmaTR'U  [kuDA an'i ]

(1.49)  [47] [05:17]

[05:18] [05:18]

[05:20] [05:20] [05:20] [05:20]

KS

[naSOK. 

VA budut d'ivaca; 

naSOK? (.) fs'o_tak'i byla PRAv'il' [na, 

a paČImu n'ė naskof; 

(0.21) 

(0.08) 

[48] [05:21] [05:22] KS

[05:26] [05:26]

[05:27] [05:28] [05:29]

patamu št m::: nask'i N'Ė sklan'ajuca; 

VA

pač'iMU,  (0.67) 

[05:29]

NU: vot [ (.) ėm– ] (0.49) 

[aD'IN nasok; ] dva  (0.9) 

[49] ..

[05:30] [05:30] [05:31] [05:31]

KS

[05:32] [05:32]

N'ĖT; 

[05:33] [05:33]

VA nasok; 

DA; 

dva nasKA,  (0.14) 

[05:34] [05:34] [05:35]

dva nasKA;  (0.17) 

TR'I naska;  (0.13) 

(0.11) 

(0.56) 

[50] [05:36] KS charašo n'ė spr'aGAjuca– = ė ja n'i ZNAju ėta sklan'ėn'iė spr'ažėn'ijė; = v_n'ė sprašyvaj m'in'a [51] ..

[05:42] [05:42]

KS  gramaT'Ič'ėsk'i nazvan'ija; 

[05:43]

[05:44] [05:45] [05:45]

°hh ja gavo!R'U!  [pa_rusk'i a n'ė– ]

VA (0.05) 

(0.64) 

(1.37) 

192

Peter Kosta

6. Cooperative overlaps, completion and smooth turn-taking In this section, I am using the term ‘cooperative overlap’ to describe a face-toface interaction in which one speaker talks at the same time as another speaker to demonstrate an interest in the conversation. The term cooperative overlap was introduced by the sociolinguist Deborah Tannen in her book Conversational Style: Analyzing Talk Among Friends (1984). Overlaps, which seem to fulfill the function of mutual co additions between partners can be seen in the next talk. In the following example (6), we will see that the two speakers are a male spouse and his wife who address the issue, namely the relationship with their cat. In turn 002, first, the man (M) speaks and asks himself whether he has turned on the recording device correctly (Jėl’ė fkl’uč’il …), the woman (F) does not interrupt him but instead confirms and completes in an entirely harmonious way his question. The overlaps here have to be classified as a completion of the turn of her husband rather than an interruption; they are used to answer the question or to introduce or make a stance on the topic of the dialogue. The subsequent discussion revolves around the relationship of the man to the male cat, his wife confirms that the cat apparently feels better with the/a man that it is the men who understand each other better and that the cat runs away from her. This dialog also does not seem to mark any conflict, but rather a cooperating and harmonious complementarity of turns of each other, so the apparent overlaps are not of the same type I have discussed in my previous sections. In this sense, here, I would not talk about a repair mechanism as a result of real overlaps but instead of a harmonious completion between partners without real duplications and interruptions. (6) = ORD (35–1) [1] [00:00]

[00:06]

M

[00:07] [00:10]

[00:11] [00:13] [00:13] [00:14]

JĖl'ė fkl'uč'il;  (2.14)  d'iktaFON mam; 

SYŠ'? 

F

zaČ'ĖM;  (6.53) ((Nebengeräusche)) 

(2.21) 

(0.86) 

[2] [00:15] [00:16] M

[00:17] [00:19] [00:19] [00:20]

a VOT– (.) štoby raBOtal; 

F

zČ'ĖM;  (1.48) 

[00:21] [00:21]

[00:22]

zaT'ĖM; (.) št [oby| ] (1.26) 

[a:– ] (1.02) 

t'ib'ė Ėtat,  (0.09) 

Third Position Repair, Overlaps, and Code-Switching [3] [00:22] [00:22] [00:24] M DA:; 

[00:27] [00:27]

nu pr'iV'ĖT; (.) pr'iV'ĖT; (.) ty moj SLA:Tk'ij t;  (0.34)  uch ty moj chaRO:šėn'k'ij; (1.49) 

[4] ..

[00:32] [00:32]

M  (.) ty moj chaRO:šėn'k'ij; ((lacht)) 

[00:33]

[00:33]

DAL nam  [spako–| ]

F

[V'Id'iš; (.) a] t m'ėn'a skar'ėj b'ėžyt k t' (0.53) 

[5] ..

[00:36] [00:37] [00:37]

M

kaN'

F iB'Ė =|= v'id'ima ty imu d'ėlaiš pr'iJATn'ėė; 

[00:38] [00:38] [00:39]

[ĖŠna; ]

°hhh kaN'

[(za/s) uŠ] Am'i gl|  (0.58) 

(0.43) 

[6] ..

[00:40] [00:41]

M ĖŠna; hh° 

[00:42]

[00:42]

[00:44] [00:44]

°hh (skžy) my žė m [užyK'I, ]

F

(ved'/i) my žė 

[toka V'] Id'it t'ib'a– (.) +++ ++++++  (0.4) 

(0.24) 

[7] ..

[00:46]

[00:47]

[00:49] [00:49]

M mužyK'I, (.) druk dru [ga pan'iMAim; ] F

mama n'ė uM'

[b'ėžyt k t'iB'Ė; ] (.) ZNAit; (.) tak znač't tam LUČšė;  (0.12) 

[8] ..

[00:51] [00:51] [00:51] [00:52]

M Ėit tak d'ėlat'; 

[00:54] [00:55]

mama n'ė uM'Ėit; (.) DA? 

F

my ZNA:im č'o nam nada; 

DA:;  (0.05) 

(0.59) 

(1.29) 

[9] [00:57] [00:58] M

[01:00] [01:02]

ėto brat d'ėla taKOjė;  (1.38) 

[01:02] [01:03]

DA dymač'ka;  (1.79) 

[01:04] [01:05]

da moj chaROšyj;  (0.52) 

DA (.) moj SLATk (0.84) 

[10] ..

[01:06] [01:07]

[01:09] [01:09]

[01:11] [01:12]

M 'ij  ((lacht))  °hh ĖJ, (.) kraSAFč'ik; °h  (0.15)  krasafč'ik moj MAl'ėn'k'ij; 

krasafč'ik moj (1.25) 

193

194

Peter Kosta

[11] ..

[01:14] [01:14] [01:15] [01:16]

M SLATk'ij t; (0.23)  O:J; 

[01:17] [01:19]

O: kak my um'ėim; 

[01:20] [01:23]

a MAma n'ė uM'Ėit ana– 

O:

F

[:J;  [DA]

(1.29) 

(1.57) 

(2.16) 

[12] .. [01:23] [01:23] [01:24]

[01:25] [01:25]

M]

[01:27] [01:28]

my vot (.) O:T kak možėm; 

F

:– 

von v'id'iš dažė 

pač'ėMU on m'in'a tak–|  (0.22) 

(0.21) 

(0.62) 

[13] ..

[01:30] [01:30]

M žyVOt'ik patstavl'ait mn'ė;  F

[01:31]

[01:32]

[ty maja]

U:Mn'ica; 

[01:33] [01:33]

VO:T vot vot;  [ (dyma)–| ]

[01:34]

UMn'ica; 

(0.3) 

(0.11) 

(0.77) 

[14] [01:35] [01:35] [01:36] M (Od'a); 

[01:39] [01:40]

DYma dyma dyma ty n'ė s'uDA zal'ės; (.) nu ka daVAJ; (.) daVAJ;  (1.15) 

DYma– (.)  (1.48) 

Completions instead of repairs of turns hold for a conversational situation from ORD in a transcript which I shall present in example no. (7): (7) = ORD (35–19) [1] [00:00]

[00:07]

F1

[00:08] [00:09] [00:09] [00:10]

ana ušLA,  (7.43) ((Nebengeräusche und Gespräche im Hintergrund)) 

JA byla;  (0.74) 

DOč'in' (0.64) 

[2] ..

[00:11]

[00:15] [00:15]

F1ka; 

(zd'ės'/ėt'ich) vot JA pr'in'ėsla; (.) n'ėV' (3.8) ((Nebengeräusche im Hintergrund)) 

[3] ..

[00:17] [00:22]

[00:24] [00:27]

[00:28]

F1 ĖRna ta byla fs'o;  FA

aP'AT' ana vam vyp'isala n'ėprav'il'na;  (5.18) 

nap'iSAla žė;  (3.11) 

(0.76) 

195

Third Position Repair, Overlaps, and Code-Switching [4] [00:29]

[00:31] [00:31] [00:32] [00:32] [00:32]

F1 CHOSpad'i; (.) da ŠO žė ėta takoė t [a; ] F2

v

[a dur] DOM; 

[00:33] [00:33] [00:33]

[y ZNAėt'i] : što; 

uM'Ėjut 

[+++ +++ ] (0.42) 

(0.17) 

[5] ..

[00:34]

[00:34]

[00:36]

F1 kagda [ (chat'at); ] FA

[ė]

ta ispravL'Ėn'ijė n'ė dapuskaica;  (1.32) ((Nebengeräusche im Hintergrund)) 

[6] [00:37]

[00:40] [00:41]

F1 ėta vraČ'I (.) tam u nas tak'iė v pal'ikl'in'ik'i s'id'at;  FA

nap'iSAla žė dl'a panz'inorma (1.33) 

[7] ..

[00:45] [00:46]

FA draž– (.) n'ė v draŽĖ on; (.) a f tabL'ĖTkach; 

[00:48]

aboznaČ'Ėn'iė s'igna;  (0.91) 

(8.31) ((ab und zu

[8] ..

[00:56]

FA

[00:58] [00:58]

ŠTO ėta takoė vapšč'ė, (.) n'ė paN'ATna–   Stimmen im Hintergrund)) 

ana vam adNU vyp' (0.4) 

[9] ..

[01:00] [01:00]

[01:01] [01:02]

F1

[01:03]

[da JA ėta at]

FA isyvait?  (0.7) 

upaKOFku; 

[01:04] [01:04]

kuda znaju; 

[panz'iNORma; ] (0.99) 

vy panz'inorm SKOL'ka vy  (0.54) 

[10] ..

[01:06]

F1

[01:07] adNU l'i dv'ė– (.) ja n'ė ZNAju– (.) doč'

FA paluč'it' dalžny;  (1.26) ((Gespräche im Hintergrund)) 

196

Peter Kosta

[11] .. [01:10] [01:11] F1

[01:14] [01:15]

a; 

FA

nu jėsl'i adNU? (.) to TR'ITcat' tabl'ėtak ana vam dalžna vyp'isat';  (1.15) 

jėsl'i DV (0.59) 

[12] ..

[01:16]

FA 'Ė? (.) tak šėS'AT; (0.05) 

[01:16]

[01:19]

vapšč'ė n'ė paN'ATna; (.) a tut tr'icat' L'ITraf što l'i;  (1.0) 

[13] [01:20]

[01:23] [01:24]

F1

DOč'in'ka (.) ėta fs'o ras (.) ėta n'ič'ėVO n'ė daš pa ėt'im tabl'ėtka [m, ]

FA

[pa ]

Ėt’im n'ėt; (.) a 

[14] ..

[01:25] [01:25]

F1

[01:28]

a Ėta ja razarvu s'ič'as– (.) (ja/a) n'ė pajDU bol'šė n'ėt; 

FA vot Ėt'i ja vam–| F2

(.) schaD' (0.21) 

[15] .. F1

[01:29]

[01:29]

[01:31] [01:32]

[N'Ė:T n'ėt ] n'ėt; (.) ŠTO vy; (.) zaČ'ĖM ėta nužna; 

ėta bol'šė N'ĖRvaf; 

F2 It'ė; (.) scha [D'It'ė; ] (0.58)  [16] ..

[01:34]

[01:36] [01:36] [01:36]

F1 (.) ja isTRAt'ila; 

BOL'šė n'ėrvaf; (.) 

F2

DA:;  (1.43) ((Nebengeräusche im Hintergrund)) 

(0.11) 

In the transcript (35–19) of the ORD corpus, it is a talking situation in the pharmacy in which a senior woman with an incorrectly filled out recipe comes in and has been told by the pharmacist about the fact that the formula cannot be accepted in this form. The old woman is sick and dependent actively on the medicine, but apparently the overlaps are nevertheless not to be interpreted as interruptions or repairs, because they complement each other and show, despite the thematically given conflicting and emotional situation, in a formal sense instead features of completions and regular turn-taking contrary to the formal features of the first

Third Position Repair, Overlaps, and Code-Switching

197

five categories, which I have discussed in Sections 1–5. So I would again speak in most cases of completion and smooth turn-taking rather than about an intervening or abrupt change of interlocutors. It resembles more of a situation here, which changes between natural turn-taking and back channel behavior. This could be also motivated by the fact that we have a typically asymmetric case of conversation styles, where the pharmacist is the stronger and the patient the weaker interlocutor, in which the pharmacist has to make the decisions, and the patient depends on his/ her choice. The cooperative speech acts and conversational steps (turns) are therefore more likely dictated by the asymmetry between the speakers and by the topic of conversation rather than by repairs and interruptions or specific repair strategies.

7. Technical questions The challenge of transcription for conversation analysis is to acquire and to record data in an exact way. When analyzing overlaps, repairs and turn taking, one should pay particular attention to the pauses between the interlocutors’ segments, the duration and the content of overlapping passages. It is a rather easy task when the interlocutors are speaking in a polite and unworried manner. In such cases, conversation units are distinguishable and are alternating with short or long pauses as in the example 8-1. (8-1)

It is becoming more complicated when interlocutors speak in a very emotional way, interrupting each other. Milliseconds can be a decisive factor to distinguish between the self-initiated self-repair or self-repair initiated by other conversational partners. What we see in example 8-2 is an other-initiation of repair and ratification of other repairs rather than self-repair. If the speaker VA had begun to speak at least 0,63 milliseconds earlier, then we couldn’t have asserted the same. That is why it is so important to have one tier per speaker, to check the length of pauses between the conversation units.

198

Peter Kosta

(8-2)

At some points, it is also essential to know the exact time of the overlapping passages, for example, to investigate the matter of turn-taking, mainly when a conversation contains a lot of interruptions. Folker is an excellent tool, which enables to draw precise borders and to determine the exact spot of overlap on the syllable level as we can see in example 8–3. As we are investigating syntax, we are not trying to set up the exact moment of the interruption within an allophone or at the morphemic or morphosyntactic interfaces. (8-3)

Moreover, it might be an interesting matter to analyze the phases of conversation at which an overlap can happen. For these purposes, we are trying to determine intonation patterns, to be more precise, intonation movements at the end of the intonation phrase. The question mark (?) and comma (,) stand for rising intonation, dash (-) stands for unchanging tone, semicolon (;) and dot (.) stand for falling tone at the end of the phrase. This might help to determine, if the speaker completed the expression or if s/he wanted to continue the turn or to maintain the right for speech. As all these goals might be realized in the transcription editor Folker, we use this tool for our research purposes at the present moment.

8. Conclusion and outlook We have dealt – in the last two examples – with the two lengthy transcriptions from ORD s35 -1 and s35 -19 we were supposed to prepare for this meeting. The present contribution tries to demonstrate that not all formal overlaps (i.e., simultaneously expressed turns of two or more speakers) initiate “repairs” in a purely formal sense. The formal coincidence of utterances on a phonetic level sometimes has to do with speaker strategies: planning and organizing their turns or structuring the coordination of other turns. This is, for example, to be seen

Third Position Repair, Overlaps, and Code-Switching

199

w.r.t. the planning of “answers” to asked “questions” in adjacency question-answer repairs, as demonstrated under example 1 of our own material and data of the transcripts of UVP in the contribution, where it is evident that this is primarily the ‘Third Position Repair’ which often causes ‘source of trouble’ in communication. Other types of “repairs” are rather ‘content-related’, here one could also speak of metalanguage or “corrections proper”3 (of the content-related type): this can be seen in the example with the “Noski” (case 5). The first document from ORD s35 -1, namely example 6, demonstrates another discourse type – a ‘doing-gender’ like strategy (cf. also Thielemann 2010, but we would speak rather of otherness discourse. In s35 – 1/6 even though we are dealing with the male-female gender of speaking that does not lead to repairs – in (6), the woman plays a largely passive role and shows more often backchannel behavior, while the man rather speaks with the cat or better to the cat and otherwise the woman’s part can be seen more as “accessory” (backchannel behavior) – it is, in fact, the exclusion of the other against us-male. This conversational behavior of otherness could be demonstrated not only formally (backchannel behavior particles such as hmm, yes on woman’s part), but it also significantly appears at the lexical level of the turn-taking behavior (‘my muž’ik’i’ ‘we men’ etc.). In the second document from ORD, example 7, no repair mechanism is to be observed, but rather an asymmetrical discourse by the fact that – on the macro-level of conversation – we are dealing with an institutional “doctor/pharmacist – patient” discourse, where at the beginning of the talk, the “superiority” of the institution/ pharmacist is formally expressed by the utterance ‘ad oculos’, or better ‘ad aures’, as the pharmacist says to the interior woman, she could not hand over the recipe to her because of erroneously completed recipe, thus no medication. From this point, I speak of an ‘abruption of communication’ in the sense of cooperation proper, which can be demonstrated by the fact that the old, sick and almost powerless

3

It should be stressed that the term of ‘correction’ in Schegloff, Jefferson and Sacks (1977, p. 363) is quite general including not an ‘error’ or ‘mistake’ by what is ‘correct’. The phenomena they are addressing are neither contingent upon error, nor limited to replacement. For instance, some occurrences do not involve replacement of one item by another. Notably, it is clear that the presence and distribution of repair/correction are not well-ordered by reference to the occurrence of a grammatical or lexical or phonetic error. Thus, many barely comprehensible misspellings or hesitation marks of Sp1 do not indeed lead to a repair or replacement by Sp1 or Sp2, and vice versa, For the same reason, we consider repair a more general term in which comprehension or production between Sp1 and Sp2 is violated leading to a bright and distinguishable strategy of repair by self- or other correction.

200

Peter Kosta

female patient continues to talk to herself only and fear, distrust and anger mark her discourse. Therefore, here also no real overlap is visible. The asymmetry of knowledge and power over medication and the function of the medicine are very clear here also in the field of formal realization of turn-taking mechanism. Conversation analysis of such critical points of the discourse deserves – due to the numerous overlaps of turns – a more precise technical description and processing of the relevant loci and should, therefore, also build the subject of discussion of its functional design and implementation.

References Auer, Peter: “Delayed self-repairs as a structuring device.” In: Hakulibeb, Auli / Selting, Margret (eds.): Syntax and Lexis in Conversation. Studies on the Use of Linguistic Resources in Talk-in-interaction. John Benjamins: Amsterdam, Philadelphia 2005, pp. 75–102. Austin, John L.: How to Do Things with Words. Second Edition. Edited by J.O. Urmson and Marina Sbisà. The William James Lectures. Harvard University Press: Cambridge, Mass 1955. Grice, Paul H.: “Logic and Conversation”. In: Cole, Peter / Morgan, Jerry L. (eds.): Syntax and Semantics, Vol. 3, Speech Acts. Academic Press: New York 1975, pp. 41–58. Kosta, Peter: „Argumentation, Persuasion und der turn-taking-Mechanismus“. In: Čmejrková, Svetlana / Hoffmanová, Jana / Müllerová, Olga / Světlá, Jindra (eds.): Dialoganalyse VI: Referate der 6. Arbeitstagung, Prag 1996. Niemeyer: Tübingen 1998 (Beiträge zur Dialogforschung), pp. 115–131. Kosta, Peter: „Konversationelle Implikaturen und indirekte Sprechakte auf dem Prüfstein“. In: Kotin, Michail I. / Kotorova, Elizaveta G. (eds.): Die Sprache in Aktion. Pragmatik. Sprechakte. Diskurs. Language in Action. Pragmatics. Speech Acts. Discourse. Universitätsverlag WINTER: Heidelberg 2011, pp. 55–69. Kosta, Peter: “Code-switching and Code-mixing Revisited in Urban and Ethnic Styles: A Brief Sketch on Variation and Language Shift”. In: Vladislava, Warditz / Kress, Beatrix (eds.): Multilingualism and Translation. Studies on Slavonic and Non-Slavonic Languages in Contact. Peter Lang Edition: Frankfurt a.M. et al. 2015, pp. 111–129. Kosta, Peter: “Třetí faktor “relevance” mezi sémantikou, pragmatikou a syntaxí”. In: Bednaříková, Božena / Pittnerová, Monika (eds.): ČÍTANKA TEXTŮ Z KOGNITIVNÍ LINGVISTIKY II. Univerzita Palackého v Olomouci. Filozofická fakulta: Olomouc 2015.

Third Position Repair, Overlaps, and Code-Switching

201

Sacks, Harvey / Schegloff, Emanuel A. / Jefferson, Gail: “A Simplest Systematics for the Organization of Turn-Taking for Conversation”. Language 50(4) Part 1, 1974, pp. 696–735. Schegloff, Emanuel A. / Jefferson, Gail / Sacks, Harvey: “The Preference for SelfCorrection in the Organization of Repair in Conversation”. Language 53(2), 1977, pp. 361–382. Schegloff, Emanuel A: “Repair After Next Turn: The Last Structurally Provided Defense of Intersubjectivity in Conversation Author(s)”. The American Journal of Sociology 97(5), 1992, pp. 1295–1345. Schegloff, Emanuel A.: Sequence Organisation in Interaction. A Primer in Conversation Analysis. Cambridge University Press: Cambridge 2007. Selting, Margret: „Imagearbeit bei der Behandlung von Verständigungsproblemen in Gesprächen“. In: Abraham, Werner / Århammar, Ritva (Hrsg.) Linguistik in Deutschland : Akten des 21. Linguistischen Kolloquiums, Groningen 1986. Niemeyer: Tübingen 1987 (Linguistische Arbeiten, 182), pp. 325–337. Selting, Margret/ Couper-Kuhlen, Elizabeth: „Argumente für die Entwicklung einer interaktionalen Linguistik“. Gesprächsforschung – Online-Zeitschrift zur verbalen Interaktion 1, 2000, pp. 76–95. Tannen, Deborah: Conversational Style: Analyzing Talk Among Friends. Ablex: Norwood, NJ 1984. Tannen, Deborah: Conversational Style. Analyzing Talk among Friends. New Edition. Oxford University Press: Oxford 2005. Tannen, Deborah: Gender and Discourse. Oxford University Press: Oxford 1994. Tannen, Deborah: “Language and Culture”. In: Fasold, Ralph / Connor-Linton, Jeff (eds.): An Introduction to Language and Linguistics. Cambridge University Press: Cambridge 2000, pp. 343–372. Thielemann, Nadine: Untersuchungen zum weiblichen Diskussionsstil am Beispiel von Gesprächen russischer, ukrainischer und polnischer InteraktionspartnerInnen. Kubon & Sagner: München 2010. Thielemann, Nadine / Kosta, Peter (eds.): Approaches to Slavic Interaction. John Benjamins: Amsterdam (Dialogue Studies) 2013.

Н. В. БОГДАНОВА-БЕГЛАРЯН

О возможных «выходах» из хезитационного поиска в процессе речепорождения1 Natalia V. Bogdanova-Beglarian

On Options for the Resolution of Hesitation Phenomena in the Process of Speech Production Summary: The article covers some possible ways of breaking out the hesitation impasse, which is very common in colloquial speech. Spontaneous speech implies some time deficit: a speaker speaks and thinks simultaneously. So some specific items appear inevitably, such as hobbles, slips, repetitions, self-correction, hesitation and so on. All these features enlighten the very mechanism of speech production. Hesitation pauses, being an integral part of colloquial speech, are very often not pauses at all: a speaker fills them with some sounds, either not communicative (e-e, m-m, ts (tut-tutting), mp (lipsmacking), sl (air intake), laughter, sigh, cough etc.), or partly communicative (verbal but meaningless). The latter ones give us reason to talk about a new class of conventionally verbal speech items – verbal hesitatives. The research shows that all these items have variative structure and are multi-functional. The most common ones are searching hesitatives, and it is their analysis that allows us to see some details of speech production mechanism. Apparently, there are several possible ‘outcomes’ of hesitative search. This article is based on analysis of “One Day of Speech” corpus material. The corpus is being created and processed in Saint-Petersburg State University. At the moment there are more than 1000 hours of recorded colloquial speech in it, mostly of dialogs and polylogs, got from more than 120 volunteer speakers (native Russian speakers, men and women, age 17–70) and more than 800 interlocutors (3–85 years old). Several different ways of hobble overcoming were given consideration to: • the speaker finds the word he/she needs (on the spot or within some time): o vot est’ / vešči takie / vot / nu / u ljudej chobbi naprimer / da ? *P *V nu(:) / tam skažem / *P nu / ne znaju / poet čto-to (S39, man., 53 # F1);

1

Исследование выполнено при поддержке гранта РНФ № 14-18-02070 «Русский язык повседневного общения: особенности функционирования в разных социальных группах».

204

Natalia V. Bogdanova-Beglarian

• the speaker turns the word sought into its descriptive substitute: o izvini menja / chachal’ iz dvuchkomnatoj chaty / @ uhu // *P *N *P @ kak ego ? *P tam evroremontik u nego / plastikovye okna / kuchon’ka novaja (S13, wom., 22 @ F1); • the speaker requests his interlocutor’s help: o a ja… ne ot menja zavisit / ja by podsadila na *N # nu dak podsadila ljudej na ėto samoe / ty znaeš’ / da ? ėto narkozavisimost’ pojavilas’ (F4 @ S4, wom., 34 @ F1 # F3); • the speaker does not find any item and just leaves the phrase unfinished: o Petr_Naumych% delaet // u nego est’ zamečatel’nye spektakli // vot vidite tut daže idet / ėto samoe / kak ego / potom on igraet / čto li ? da (S45, man., 65). The interlocutors demonstrate various reactions: • prompt the speaker with the word he/she needs: o ėto… ėto @ ne pirožnoe @ ėto ne pirožnoe / ėto… ėto samoe… @ miralgin // na (F3 @ F1 – in talking to: S4, wom., 34); • continue the description (demonstrating understanding) – backchannel: o u nego že posle meningita / u nego(:) ėto samoe / *P on medlenno govorit / on (e-e) / dolgo emu nado du… obdumyvat’ vse ėti… # dlinnaja reakcyja vidimo (S6, wom., 27 # F1); • answer the incomplete question (demonstrate that communication is complete): o ladno / uspokojsja // u tebja est’ / ėto samoe … # net (M1 # S11, wom., 28); • ease the speaker saying that everything is clear and no words are needed (communication is complete): o von tuda postav’ poka // *P eščo možet byt’ budem s sokom nu… na… kak ėto ? # ja znaju / da (F1 # S4, wom., 34). It is interesting to turn attention to these phenomena, particularly in terms of cognitive analysis because it clears up some speech processes and specific features of colloquial discourse. Аннотация: Спонтанная речь порождается в условиях временнoго дефицита: человек думает и говорит одновременно. В таких условиях неизбежны заминки, оговорки, паузы хезитации, повторы, самокоррекция и т. п. явления, проливающие свет на механизмы порождения речи. Паузы хезитации, являющиеся неотъемлемым свойством любой устной речи, часто бывают не физическими (полное отсутствие звучания), а так или иначе заполненными. Заполнять такие паузы могут либо неречевые звуки (э-э, м-м, тс (цоканье), мп (причмокивание), сл (хлюпанье, или шумное втягивание воздуха) и под.), либо паралингвистические элементы (смех, вздох, кашель, стон), либо какие-то вербальные, но практически незначимые единицы, что дает основания говорить о целом классе условно-речевых функциональных единиц устной речи – вербальных хезитативов. Все эти единицы, как показали наблюдения, часто структурно вариативны и полифункциональны. Наиболее распространены при этом поисковые хезитативы, и именно анализ контекстов с такими единицами позволяет увидеть некоторые детали механизма порождения речи.

On Options for the Resolution of Hesitation Phenomena

205

Можно, по-видимому, говорить о разных возможных «выходах» из ситуации хезитационного поиска. Анализ, результаты которого представлены в настоящей статье, проведен на материале Звукового корпуса русского языка «Один речевой день» (ОРД), создаваемого в Санкт-Петербургском государственном университете. К настоящему времени корпус содержит более 1250 часов звучания, представляющего собой записи естественной повседневной речи (преимущественно диалогической и полилогической), полученной от более чем 120 волонтеров-информантов, носителей русского языка, мужчин и женщин в возрасте от 17 до 70 лет, а также около 1000 их коммуникантов (от 3 до 85 лет). На материале ОРД прослеживаются различные варианты преодоления возникшей коммуникативной заминки: • говорящий сам находит искомое слово (сразу или дистантно): o вот есть / вещи такие / вот / ну / у людей хобби например / да ? *П *В ну(:) / там скажем / *П ну / не знаю / паяет что-то (И39, муж., 53 года # Ж1); • говорящий не находит подходящего слова и заменяет его описанием: o извини меня / хахаль из двухкомнатной хаты / @ угу // *П *Н *П @ как его ? *П там евроремонтик у него / пластиковые окна / кухонька новая (И13, жен., 22 года @ Ж1)2; • искомая единица не найдена, и говорящий апеллирует к собеседнику, полагая, что тот и так понял, о чем речь: o а я… не от меня зависит / я бы подсадила на *Н # ну дак подсадила людей на это самое / ты знаешь / да ? это наркозависимость появилась (Ж4 @ И4, жен., 34 года @ Ж1 # Ж3); • говорящий, ничего не найдя, просто бросает начатую фразу: o Пётр_Наумыч% делает // у него есть замечательные спектакли // вот видите тут даже идёт / это самое / как его / потом он играет / что ли ? да (И45, муж., 65 лет). Собеседник при этом по-разному реагирует на производимый говорящим поиск (знак @ в расшифровках означает наложение речи обоих собеседников, знак # – мену говорящих): • подсказывает говорящему искомое слово: o это… это @ не пирожное @ это не пирожное / это… это самое… @ миралгин // на (Ж3 @ Ж1 — в разговоре с И4, жен., 34 года); • продолжает описание искомого предмета/явления (backchannel), демонстрируя, что понимает, о чем речь: 2

Here, @ means speech overlaps, and # means the change of speakers/interlocutors.

206

Natalia V. Bogdanova-Beglarian

o у него же после менингита / у него(:) это самое / *П он медленно говорит / он (э-э) / долго ему надо ду… обдумывать все эти … # длинная реакция видимо (И6, жен., 27 лет # Ж1); • искомая единица в ходе поиска не найдена, но собеседник не только все понимает, но и отвечает на недоформулированный говорящим вопрос (коммуникация определенно состоялась): o ладно / успокойся // у тебя есть / это самое … # нет (М1 # И11, жен., 28 лет); • успокаивает говорящего, что он все понимает и слово искать не надо (коммуникация состоялась): o вон туда поставь пока // *П ещё может быть будем с соком ну… на… как это ? # я знаю / да (Ж1 # И4, жен., 34 года). Анализ подобных явлений на материале естественной речи интересен, в частности, в когнитивном аспекте, поскольку проливает свет на некоторые механизмы порождения речи и в целом на специфику построения устного дискурса.

Введение Спонтанная речь порождается, как известно, в условиях временнОго дефицита: человек думает и говорит одновременно. В таких условиях неизбежны заминки, оговорки, паузы хезитации, повторы, самокоррекция и т. п. явления, проливающие свет на механизмы порождения речи. Паузы хезитации (от англ. Hesitation – ‘колебание’), являющиеся неотъемлемым свойством любой устной речи, «характернейшей особенностью» такой речи, «отражающей ее сиюминутность, творческий характер» (Фонетика спонтанной речи 1988: 144), часто бывают не физическими (полное отсутствие звучания), а так или иначе заполненными: либо неречевыми звуками (э-э, м-м, тс (цоканье), мп (причмокивание), сл (хлюпанье, или шумное втягивание воздуха)3 и под.), либо паралингвистическими элементами (смех, вздох, кашель, стон), либо вербально, что дает основания говорить о целом классе условно-речевых функциональных единиц – вербальных хезитативов (ВХ) (см. о них подробнее: Богданова-Бегларян 2013а; Звуковой корпус… 2015). Все эти единицы, как показали наблюдения, часто структурно вариативны и, как правило, полифункциональны. Наиболее распространены при этом поисковые ВХ (см. о них, например: Богданова-Бегларян 2013б) и именно

3

О таких мало описанных типах хезитационных явлений, как цоканье языком, причмокивание и хлюпанье, или шумное втягивание воздуха, см., например: Казак 2015, Чэн 2018.

On Options for the Resolution of Hesitation Phenomena

207

анализ контекстов с такими единицами позволяет увидеть некоторые детали механизма порождения речи. В целом исследование хезитаций стало «основой целого направления, весьма интересного и перспективного, возникшего на рубеже лингвистики, психологии, медицины и теории вероятностных процессов» (Николаева 1970: 117). Сегодня это направление называют коллоквиалистикой, или теорией разговорной речи (см.: Девкин 1979; Скребнев 1985). Настоящее исследование выполнено в рамках этого направления.

Материал исследования Анализ проведен на материале Звукового корпуса русского языка «Один речевой день» (ОРД), создаваемого в Санкт-Петербургском государственном университете. К настоящему времени корпус содержит более 1250 часов звучания, представляющего собой записи естественной повседневной речи (преимущественно диалогической и полилогической), полученной от более чем 120 волонтеров-информантов, носителей русского языка, мужчин и женщин в возрасте от 17 до 70 лет, а также около 1000 их коммуникантов (от 3 до 85 лет) (см. подробнее об ОРД: Богданова-Бегларян и др. 2015; Sherstinova et al. 2015; Русский язык повседневного общения 2016). Основой дискурсивного анализа стали транскрипты (расшифровки) этих звукозаписей объемом около 1 млн. словоформ.

Результаты анализа Хезитационный поиск является, как уже было отмечено, одной из наиболее распространенных функций пауз хезитации, как физических, так и заполненных тем или иным способом. Анализ материала в этом отношении позволяет говорить как об отдельных вербальных хезитативах (это, вот, это самое, скажем, как его, знаешь, (я) не знаю и под.), так и о цепочках (контаминациях) ВХ и/или иных хезитационных явлений: • были (э) как-то вот / (э) (…) вот эти / как их ? лямблии ? или как это ? (ОРД: И44, муж., 41 год # Ж1)4.

4

Об особенностях орфографического представления материала в корпусе ОРД см.: Шерстинова и др. 2009; Русский язык повседневного общения 2016. Все примеры в статье атрибутируются с указанием номера информанта, его пола и возраста (например, И44, муж., 41 год) или его коммуникантов (например, Ж1). В обозначении последних учитывается только их гендерная принадлежность (М/Ж).

208

Natalia V. Bogdanova-Beglarian

Преодоление хезитационной заминки в процессе речепорождения происходит по-разному, что позволяет говорить о разных способах «выхода» из хезитационного поиска. В большинстве случаев говорящий сам тем или иным способом преодолевает возникшее затруднение, но часто в процесс включается и его собеседник. Рассмотрим все выявленные в ходе анализа ситуации. 1. Говорящий сам находит искомую единицу (сразу или дистантно): 1) а / всё … # а у меня на да… д… дача на этом / как его / на Дунае$ // # а ну / на Дунае$ хорошо (ОРД; И39, муж., 53 года # М1); 2) вот есть / вещи такие / вот / ну / у людей хобби например / да ? *П *В ну(:) / там скажем / *П ну / не знаю / паяет что-то (ОРД; И39, муж., 53 года # Ж1); 3) а я Санычу% и сказал / я говорю / нам (…) своих некуда сажать // *П а у них и (э-э) и зачёт и(:) (…) этот (м) / как его (…) ну и у Мельченко% и у… и(:) у Токарева(:)% зачётное занятие (ОРД; И7, муж., 45 лет). В примерах (1)-(2) «выход» из хезитационного поиска происходит довольно быстро (on-line), сразу после использования говорящим более или менее длинных хезитационных конструкций. В контексте (3) между достаточно протяженной цепочкой хезитативов (и(:) (…) этот (м) / как его (…) ну) и найденным словосочетанием (зачётное занятие) говорящий, словно помогая себе в предпринятом поиске, вставляет ряд элементов, зависимых от искомой единицы, продолжая при этом хезитировать: и у Мельченко% и у… и(:) у Токарева(:)%. Здесь можно говорить об off-line коррекции речевого фрагмента. В любом случае говорящий самостоятельно справляется с хезитационным поиском, который оказывается в данном случае вполне успешным (найденная единица во всех примерах подчеркнута). Примеры подобного рода позволяют сделать ряд грамматических наблюдений над формированием устного дискурса. Так, в случае поиска имени вербальный хезитатив может находиться в грамматической гармонии или дисгармонии с искомой формой. Гармонию (в роде, числе и падеже) можно видеть, например, в контексте (1): на этом / как его / на Дунае$. Здесь можно, по-видимому, говорить даже о двойной гармонии, которую демонстрирует найденная форма (на Дунае$) по отношению к обоим ВХ (на этом / как его). См. также:

On Options for the Resolution of Hesitation Phenomena

209

4) ну на фотографиях я посмотрел / он там //-// как бы вылетел он из этой самой ∫ из машины и где-то вот ∫ метра-ах наверно ∫ в пяти от машины / лежит тело (САТ; муж., 43 года, юрист, высокий УРК; рассказ)5. В примере (5) налицо грамматическая дисгармония (по числу) ВХ и искомого (и найденного) имени: 5) а потом нас еще и повезли / как бы на эти самые ∫ на день рождения (САТ; муж., 43 года, юрист, высокий УРК; рассказ). В следующем примере (6) наблюдается полная гармония с найденным именем первого из цепочки ВХ (вот этих […] Мегах$) и дисгармония – второго (как его / Мегах$). Такая же дисгармония видна и в следующей поисковой ситуации в том же контексте: у этих бл**ь (…) в(:) … в Океях$ стоят / в Лентах$: 6) я понял / пацанам короче / ну / подрабатывал (…) на заводе // пацанам помогал / короче / знаешь / какие короче / стеллажи / которые вот в Ике…$ / вот этих / как его / Мегах$ ст… () у этих бл**ь (…) в(:) … в Океях$ стоят / в Лентах$ (ОРД; И105, муж., 37 лет # М2). 2. Искомая единица в результате произведенного говорящим поиска так и не найдена, но информант заменяет ее описанием слова или ситуации: 1) а сейчас с флэшкой всё таки / ага / сейчас // он ушёл в этот самый @ он ушёл @ в (э-э) … бельё понёс / я ему сказала / он сейчас бельё отнесёт и придёт (ОРД; И6, жен., 40 лет – Ж5 @ Ж2 @ Ж5); 2) вот этот вот дом / *П он … вот это всё / можно тоже таким // это как его / ну как брев… бре… брёвна / как () что ? доски // *В а вот эту сторону … она будет немножко что / потемнее / в тени потому что (ОРД; И37, жен., 59 лет);

5

Часть примеров в статье – из модуля САТ (сбалансированная аннотированная текстотека) в составе Звукового корпуса русского языка, представляющего монологическую речь носителей русского языка разных социальных групп (см. о нем подробнее: Звуковой корпус… 2013). Особенности расшифровок этого материала несколько отличаются от ОРД, см. о них подробнее: Русская спонтанная речь 2008. В атрибуции этих контекстов указан тип монолога (рассказ, пересказ или описание изображения), а также метаданные информанта: его пол, возраст, профессия и уровень речевой компетенции (УРК) – высокий, средний или низкий.

210

Natalia V. Bogdanova-Beglarian

3) извини меня / хахаль из двухкомнатной хаты / @ угу // *П *Н *П @ как его ? *П там евроремонтик у него / пластиковые окна / кухонька новая (ОРД; И13, жен., 22 года @ Ж1). В данных примерах поиск тоже можно считать удачным: говорящий в примере (1) заменил искомую номинацию описанием действия (бельё понёс), которое помогает собеседнику понять, о чем идет речь (например, о прачечной). В примере (2) хезитационная заминка возникла при попытке говорящего описать цвет, в который, по его мнению, стоит покрасить дом. В результате произведенного поиска найденным оказался цветовой аналог: как брёв… брё… брёвна / как () что ? доски. В контексте (3) искомое имя заменено описанием наиболее важных характерных черт того человека, о котором идет речь: там евроремонтик у него / пластиковые окна / кухонька новая. Приведенные иллюстрации позволяют увидеть и еще некоторые особенности устной спонтанной речи, вполне типичные для повседневного общения. Так, в контексте (1) хезитационный поиск начинается после предлога, который повторяется во фрагменте дважды, вместе с различными поисковыми маркерами: он ушёл в этот самый @ он ушёл @ в (э-э) … бельё понёс. Подобные ситуации, когда в процессе речепорождения грамматические элементы дискурса воспроизводятся без задержки, а содержательные, семантически нагруженные, производятся, требуют от говорящего определенных усилий и вынуждают его использовать маркеры хезитационного поиска, оказываются весьма типичными для устной спонтанной речи. Поиск может начаться не только после уже произнесенного предлога, но и после грамматического элемента составного глагольного сказуемого (люблю это самое отдыхать), после отрицательной или иной частицы (они говорят / даже ∫ это самое ∫ даже паркет отрывали; а оно не это самое / не долбанёт ?), после сравнительного союза (подают как это самое / как будто пре-парте какое-то) и даже после грамматически значимого префикса (люди уже(:) по несколько раз при… это самое / приходят). Иными словами, при порождении речи грамматика (форма) опережает семантику (значение), ср.: «Во многих языках, в том числе и в русском, в этой ситуации (хезитационной заминки – Н. Б.-Б.) используются согласуемые маркеры-заместители, типа этот (самый)/эта (самая), такой/такая, как его/её, которые демонстрируют, что говорящий уже выбрал грамматическую форму планируемой группы и колеблется лишь в выборе конкретной номинации» (Подлесская 2013: 634–635) (подробнее о грамматических наблюдениях такого рода на материале ОРД см.: Богданова-Бегларян 2015).

On Options for the Resolution of Hesitation Phenomena

211

Любопытен с точки зрения структурирования устного дискурса и пример (2): хезитируя в поисках нужного слова, говорящий словно бы вступает в минидиалог с самим собой: как () что ? доски. Сам себе задает вопрос и сам же на него отвечает. Происходит своеобразная диалогизация монологических реплик, что вполне типично для нашей повседневной устной речи. Ср. еще ряд подобных примеров: 4) он говорит какое отчество она старше вас ненамного / оона получается так что она действительно старше нас где-то там ∫ буквально лет на пять / почему? потому что она Колю родила ∫ когда наверно когда ей восемнадцать ей было / всего лишь (САТ; муж., 43 года, юрист, высокий УРК; рассказ); 5) итак / как провел лето? лето я провел хорошо (САТ; муж., 30 лет, юрист, высокий УРК; рассказ); 6) что это такое ? ну не знаю / @ ну понятно @ то есть / а () она тоже зарегистрирована / и че… () связывается с сервером / (э) я бы (э-эм) не стал тут ничего портить / я поставлю ноут / а он возьмёт завтра и перестанет обновляться / и что тогда ? (ОРД; И38, муж., 58 лет @ М1); 7) значит это что? это диктофон (ОРД; И43, муж., 60 лет); 8) что же он увидел? он увидел десятки / сотни рыбаков / в маленьком пруду (САТ; муж., 37 лет, юрист, высокий УРК; описание изображения); 9) что что здесь / мы видим? мы видим что человек / поймал рыбу в пруду (САТ; муж., 40 лет, юрист, высокий УРК; описание изображения). Видно, что это наблюдается в равной степени и в диалогической, и в монологической речи. 3. Искомая единица не найдена, и говорящий апеллирует к собеседнику, полагая, что тот и так понял, о чем речь: • а я… не от меня зависит / я бы подсадила на *Н # ну дак подсадила людей на это самое / ты знаешь / да ? это наркозависимость появилась (ОРД; Ж4 @ И4, жен., 34 года @ Ж1 # Ж3). Обратившись за поддержкой к собеседнику, говорящий, не дожидаясь ответа, все же сам дает описание искомой единицы: это наркозависимость появилась. Важно отметить, что и в данном случае предлог говорящим произнесен еще до вербального хезитатива (подсадила людей на это самое), и только после него начинается собственно поиск нужной единицы. Иными словами, и здесь грамматика несколько опережает семантику порождаемой единицы, которая в ходе произведенного поиска так и не нашлась.

212

Natalia V. Bogdanova-Beglarian

4. Искомая единица не найдена, фраза говорящим просто брошена: 1) сейчас // *П тоже цветные / да ? # цветные / да // @ да // @ цветные матовые // цветные матовые / хорошо // *П надеюсь(?) этот самый / как его / *Н *П *В шестнадцать штук три на четыре ещё надо (ОРД; И39, муж., 53 года @ М1); 2) так а у меня(:) чуть чего-то тоже(:) (э-э) как его ? # после этого и панкреатит может развиться // # *П ур… уролог мне там тоже что-то прописывал после чего я в госп… в госпиталь и попал (ОРД; И7, муж., 45 лет # Ж1); 3) Пётр_Наумыч% делает // у него есть замечательные спектакли // вот видите тут даже идёт / это самое / как его / потом он играет / что ли ? да (ОРД; И45, муж., 65 лет). По контекстам видно, что неудачный поиск не разрушает, тем не менее, коммуникации: в большинстве случаев она продолжается так, словно говорящие прекрасно поняли друг друга. 5. Вербализованная хезитация часто вынуждает собеседника включиться в построение высказывания и помочь говорящему выйти из хезитационного поиска. Это тоже происходит по-разному. Так, например, собеседник может просто подсказать говорящему искомое слово или выражение: 1) ой / а ягодка опять этого / от фы… фы… как он ? # физалис # физалис // я когда-то выращивала / *В один куст достаточно / ты знаешь / он … (ОРД; Ж1 # И4, жен., 34 года); 2) это… это @ не пирожное @ это не пирожное / это… это самое… @ миралгин // на (ОРД; Ж3 @ Ж1 – в разговоре с И4, жен., 34 года); 3) и потом у нас было очень сильное вот (э) / логопед говорил / грассирующее эр / *В которое сейчас даже уже / логопеды / они относят / *В уже хотят это / столько нарушений / хотят оставить как это / @ к норме отнести @ да / к норме (ОРД; Ж2 @ Ж1 @Ж2 – в разговоре с И22, жен., 35 лет). Видно, что в примерах (1) и (3) говорящий не просто принимает подсказку собеседника, но и вербализует свое согласие с ней, повторяя найденную единицу (физалис; да / к норме). 6. Искомая единица в ходе поиска не найдена, но собеседник не только все понимает, но и отвечает на недоформулированный говорящим вопрос (то есть коммуникация определенно состоялась):

On Options for the Resolution of Hesitation Phenomena

213

1) ладно / успокойся // у тебя есть / это самое … # нет (ОРД; М1 # И11, жен., 28 лет); 2) нормально ? *П ничего там тебе / не это самое (…)*П # да нет (ОРД; М1 # И28, муж., 19 лет); 7. Искомая единица не найдена, говорящий пытается заменить ее описанием, собеседник также не находит искомого слова, но поддерживает описание (коммуникация снова состоялась, несмотря на неудачный поиск): • у него же после менингита / у него(:) это самое / *П он медленно говорит / он (э-э) / долго ему надо ду… обдумывать все эти … # длинная реакция видимо (ОРД; И6, жен., 27 лет # Ж1). 8. Искомая единица в ходе поиска также не найдена, но собеседник успокаивает говорящего, что он и так все понял (коммуникация состоялась): 1) вот / просто вы когда запишетесь позвоните я тогда это самое… # да / хорошо (ОРД; М1 # И1, жен., 33 года); 2) вон туда поставь пока // *П ещё может быть будем с соком ну… на… как это ? # я знаю / да (ОРД; Ж1 # И4, жен., 34 года). В примере (1) здесь собеседник прерывает хезитационный поиск, осуществляемый говорящим, и демонстрирует своей репликой, что он не только понял, о чем речь, но и согласен с говорящим (да / хорошо). В примере (2) собеседник также останавливает довольно мучительный хезитационный поиск (ну… на… как это ?), в ходе которого говорящий фактически обращается за помощью к своему коммуниканту, который и успокаивает его, что искать дальше не стоит, все и так понятно (я знаю / да).

Некоторые выводы 1. Анализ подобных явлений – различных «выходов» говорящего из хезитационного поиска – на материале повседневной естественной речи интересен, в частности, в когнитивном аспекте, поскольку проливает свет на некоторые механизмы порождения речи и в целом на специфику построения устного дискурса. 2. В диалогической речи вербальный хезитатив как маркер поиска часто стимулирует собеседника включиться в построение дискурса. В ряде случаев это приводит к ситуации, которая в лингвистике получила название Backchannel – продолжение конструкции, начатой одним говорящим, в речи его собеседника, фактически – построение одного дискурса

214

3.

4.

5. 6.

Natalia V. Bogdanova-Beglarian

усилиями двоих говорящих (см., например: Cathcart, Carletta, Klein 2003; Гренобль 2008; Grenoble 2013). Вербальный хезитатив как маркер поиска может способствовать свертыванию диалога (еще одно проявление закона экономии, столь свойственного устной неподготовленной речи): собеседник дает говорящему понять, что уже все понял и можно не продолжать поиска нужной единицы. Неудачный поиск, производимый говорящим, отнюдь не обязательно приводит к коммуникативной неудаче: практически при любом «раскладе» коммуникация может считаться состоявшейся, то есть собеседники прекрасно понимают друг друга, чему в немалой степени способствует, по-видимому, общность ситуации и некоторого предыдущего опыта, что в принципе свойственно диалогу в отличие от монолога. Монологической речи хезитационный поиск свойствен практически в той же степени, что и диалогической, разве что на помощь собеседника говорящему в этом случае рассчитывать не приходится. Поскольку круг вербальных хезитативов русской речи достаточно узок, а их поисковая функция является хоть и наиболее распространенной, но все же не единственной, целесообразно поставить задачу создания специального Словаря прагматем русской разговорной речи, в который войдут и вербальные хезитативы, упоминаемые в настоящей статье, и, например, дискурсивы (значит, ну вот), метакоммуникативы (знаешь (там), скажи, понимаешь), рефлексивы (или как это ?, скажем так, или как его?), междометные прагматемы (драсьте пожалста!, Щас прям!, будет тебе, давай), аппроксимативы разного типа (все дела, и всё такое (прочее), бла-бла-бла, то-сё, туда-сюда, пятое-десятое), дейктические маркеры (вот здесь вот, вот это вот), маркерыксенопоказатели (типа (того что), такой, вроде того (что), грит), ряд ритмообразующих маркеров (там, короче, блин) и т. п. (подробнее о проекте такого словаря и о структуре его словарной статьи см.: Богданова-Бегларян 2014). «Думается, что такой словарь может найти своего читателя и пользователя, среди которых могут быть специалисты-лингвисты, исследователи повседневной русской речи, создатели грамматики русской речи (антропоцентрической грамматики), переводчики спонтанных текстов на другие языки, хотя бы в рамках художественного произведения, при передаче речи персонажей, преподаватели русского языка как иностранного, поскольку носители других языков вынуждены учиться воспринимать и правильно понимать

On Options for the Resolution of Hesitation Phenomena

215

русскую спонтанную речь как устно, так и письменно, при чтении русскоязычных текстов, а также все те, кому интересна наша повседневная речь» (Богданова-Бегларян 2014: 16).

Список литературы Богданова-Бегларян, Наталья В.: О словаре «не-слов»: возможности лексикографического описания вербальных хезитативов русской речи. Слово. Словарь. Словесность: Коммуникация. Текст. Синтаксис (к 90-летию со дня рождения С. Г. Ильенко). Материалы Всероссийской научной конференции, Санкт-Петербург, РГПУ им. А. И. Герцена, 13–15 ноября 2013 г. Отв. ред. Черняк, Валентина Д. Сага: Санкт-Петербург, 2013а, сс. 359–364. Богданова-Бегларян, Наталья В.: Кто ищет – всегда ли найдет? (о поисковой функции вербальных хезитативов русской спонтанной речи). Компьютерная лингвистика и интеллектуальные технологии. По материалам ежегодной Международной конференции «Диалог» (2013) (Бекасово, 29 мая – 2 июня 2013 г.). Вып. 12 (19). В двух томах. Том 1. Основная программа конференции. Гл. ред. Селегей, Владимир П. РГГУ: Москва, 2013б. сс. 125–136. Богданова-Бегларян, Наталья В.: Прагматемы в устной повседневной речи: определение понятия и общая типология. Вестник Пермского университета. Российская и зарубежная филология. Вып. 3 (27), 2014, сс. 720. Богданова-Бегларян, Наталья В. Из наблюдений над спонтанной речью: грамматический аспект. Труды международной конференции «Корпусная лингвистика-2015». 2226 июня 2015 г., Санкт-Петербург. Филологический ф-т СПбГУ: Санкт-Петербург, 2015, сс. 129–136. Богданова-Бегларян, Наталья В., Асиновский, Александр С., Блинова, Ольга В., Маркасова, Елена В., Рыко, Анастасия И., Шерстинова, Татьяна Ю.: Звуковой корпус русского языка: новая методология анализа устной речи. Язык и метод: Русский язык в лингвистических исследованиях XXI века. Вып. 2 / Ред. Шумска, Дорота, Озга, Кшиштоф. Wydawnictwo Uniwersytetu Jagiellońskiego: Kraków, 2015, сс. 357–372. Гренобль, Ленора А.: Синтаксис и совместное построение в устном русском диалоге. Вопросы языкознания, № 1, 2008, сс. 25–36. Девкин, Валентин Д.: Немецкая разговорная речь: Синтаксис и лексика. Москва: Международные отношения, 1979. Звуковой корпус как материал для анализа русской речи. Коллективная монография. Часть 1. Чтение. Пересказ. Описание / Отв. ред. БогдановаБегларян, Наталья В. Филологический ф-т СПбГУ: Санкт-Петербург, 2013.

216

Natalia V. Bogdanova-Beglarian

Звуковой корпус как материал для анализа русской речи. Коллективная монография. Часть 2. Теоретические и практические аспекты анализа. Том 2. Звуковой корпус как материал для новых лексикографических проектов / Отв. ред. Богданова-Бегларян, Наталья В. Филологический факультет СПбГУ: Санкт-Петербург, 2015. Казак, Мария В.: Паузы хезитации в спонтанной речи на родном и неродном языках (на материале речи франкофонов). Lambert Academic Publishing: Saarbrücken, 2015. Николаева, Татьяна М.: Новое направление в изучении спонтанной речи (О так называемых речевых колебаниях). Вопросы языкознания, № 3, 1970, сс. 117–123. Подлесская, Вера И.: Нечеткая номинация в русской разговорной речи: опыт корпусного исследования. Компьютерная лингвистика и интеллектуальные технологии. По материалам ежегодной Международной конференции «Диалог». Вып. 12 (19). В двух томах. Том 1. Основная программа конференции. Гл. ред. Селегей, Владимир П. – М.: РГГУ: Москва, 2013, сс. 631–643. Богданова, Наталья В. (Отв. ред. и автор предисловия): Русская спонтанная речь. Свободные монологи-рассказы на заданную тему. Тексты. Лексические материалы. Сост. Куканова, Виктория В. Факултет филологии и искусств СПбГУ: Санкт-Петербург, 2008.  Русский язык повседневного общения: особенности функционирования в разных социальных группах. Коллективная монография. Отв. ред. Богданова-Бегларян, Наталья В. ЛАЙКА: Санкт-Петербург, 2016. Скребнев, Юрий М. Введение в коллоквиалистику. Изд-во Сарат. ун-та: Саратов, 1985. Фонетика спонтанной речи. Отв. ред. Светозарова, Наталья Д. Изд-во ЛГУ: Ленинград, 1988. Чэн, Чэнь.: Хезитации в русской устной речи носителей китайского языка. Дис. … канд. филол. наук. Санкт-Петербург, 2018. (машинопись). Шерстинова, Татьяна Ю., Рыко, Анастасия И., Степанова, Светлана Б.: Система аннотирования в звуковом корпусе русского языка «Один речевой день». Формальные методы анализа речи. Материалы XXXVIII Международной филологической конференции. Факультет филологии и искусств СПбГУ: Санкт-Петербург, 2009, сс. 66–75. Cathcart, Nicola, Carletta, Jean, Klein, Ewan.: A Shallow Model of Backchannel Continuers in Spoken Dialogue. 10th Conference of the European Chapter of the ACL. Vol. 1, 2003, рp. 51–58.

On Options for the Resolution of Hesitation Phenomena

217

Grenoble, Lenore A.: Talking out of turn: (co)-constructing Russian conversation. Approaches to Slavic Interaction. Thielemann, Nadine, Kosta Peter (eds.) – John Benjamins, Amsterdam, 2013. рp. 17–33. Sherstinova, Tatiana, Bogdanova-Beglarian, Natalia, Martynenko, Grigory.: The «One Day of Speech» Corpus: Phonetic and Syntactic Studies of Everyday Spoken Russian. SPECOM 2015, Lecture Notes in Artificial Intelligence, LNAI, vol. 9319. Springer International Publishing, Switzerland, 2015, рp. 429–437.

Lenore A. Grenoble

Multimodality and Interactional Grammar in Russian Conversation Abstract: Conversation, in Russian as in other languages, is inherently interactive. Following an analytic stance that treats all details at all levels as potentially relevant to interlocutors, it is shown that the scope of devices used in interactional grammar includes both linguistic and nonlinguistic phenomena. Interlocutors collaboratively build a conversation, relying on a complex interplay of linguistic elements and non-linguistic cues, together with shared knowledge. One key part in this co-construction of conversational structure is projection, an individual action or part of it foreshadows another. The paper presents a preliminary account of several non-verbal projection devices in Russian interactional grammar, arguing for an expanded category of gesture that encompasses kinesic gestures, non-verbal cues such as eye gaze and body position, kinesic gestures, along with vocal cues such as inhalations and clicks. These are shown to play a role creating conversational structure and providing cues to interlocutors about that structure.

1. Introduction This paper argues for a complex, multi-faceted model to account for the interaction of linguistic, multi-modal and paralinguistic resources in creating and maintaining conversational structure in Russian (Thielemann and Kosta 2013, p. 4). Face-to-face interaction is multi-modal and multi-channel; the study of this complex system is still in its infancy with regard to Russian. Conversation is inherently interactive, and interlocutors collectively contribute to the organization of its structure; the turn-taking system is central to this structure. Turn-taking has been shown to be rapid and to have universal parallels in both spoken and signed languages (see e.g. Levinson 2015). The question of how the turn-taking system is created and how interlocutors know when to start, stop or continue talking. An integral part of the organizational model is projection, defined as an individual action or part of it that foreshadows another (Auer 2005, p. 9). Auer’s work on projection focuses on syntax and its role in interactional grammar, but projection can be understood more broadly to encompass a range of linguistic phenomena (prosodic, phonological, semantic) and non-verbal cues (e.g. manual gesture, head movements, eye gaze) that anticipate and co-construct upcoming discourse. Here I consider this broad range and discuss not only linguistic projection, but also the role that manual gestures play and their function as projecting

220

Lenore A. Grenoble

devices. Projection occurs at various levels in discourse, including: (1) the topical level, across larger discourse units; (2) locally, at the level of the clause; (3) across clauses; and (4) across turns. In the present paper I address the interaction between projection and turn structure, focusing on the interactive use of projection in turn-taking, turn-yielding and turn-holding and the interplay with topical structure. Interlocutors collaboratively build, or co-construct, a conversation, relying on syntax and a complex interplay between other linguistic elements (semantics, prosody and other non-linguistic vocal cues) and nonverbal kinesic devices (such as manual gesture, eye gaze, and body position and movement) as well as shared knowledge. Nonverbal gestures have multiple functions, which include serving to project a turn change, to project turn holding, or as backchanneling devices. In Russian frequent verbal gestures include clicks, hesitation markers and inhalations; they are used to project turn construction units (TCUs) and/ or new topic units for a continuing speaker. (See Paschen 2014 and this volume for clicks in Russian conversation; Local and Walker 2012 for English.) Both verbal and nonverbal gestures in the database can be classified in three of the four categories of Kendon’s Continuum of Gestures (McNeill 2005): gesticulations, emblems, and pantomimes (Section 3). Non-verbal gestures may occur on their own, but in spoken language co-speech gestures are pervasive; research into how they combine with speech and verbal cues in projection in Russian interaction is still at its early stages. Projections may or may not align with one another, and sometimes there are mismatches as, for example, misalignments between syntactic and semantic completion (that signals the possible end of a turn and prosodic cues (such as a rise in pitch, lack of a pause) that project continuation (see Section 2). The paper presents a preliminary model for weighting these various factors in an account of Russian interactional grammar, showing that different kinds of projection devices are used for different kinds of projection, and the differences in distribution in conjunction with the interlocutor’s role in the conversation. Syntax, defined here as a combination shared syntax (including routinized patterns and constructions) and online syntax (collaboratively constructed by the interlocutors, Auer 2009), plays a fundamental organizational role in the organization of turn-taking system of Russian.

1.1. Methodology Data are taken from audio and video recordings of spontaneous conversation, drawn from the ORD corpus, the Russian National Corpus, television broadcasts

Multimodality and Interactional Grammar in Russian Conversation

221

and my own field data. The prime sources of gestural data are two video recordings: (1) an interview between Vladimir Pozner and Aleksandr Zaldanov (known as Chirurg ‘Biker’) that aired 02 June 2014, Pervyj kanal and is available online (Pozner 2014), and (2) an online television program with Mixail Labkovskij and two interviewers on the topic “What is love?” (Labkovskij 2014). Both broadcasts were selected because of the relatively free format of the interviews and the camera angles which show the interlocutors, including their hands, although there are segments where only one or the other is visible. Still, it is relatively natural discourse, although it is important to keep in mind that the interview setting has an influence on the structure of the text. The present paper provides a close analysis of some segments of Pozner (2014). The format of the program clearly defines the roles of each participant as interviewer (Pozner) and interviewee (Zaldanov). Pozner thus has a privileged role in asking questions (and thus in controlling discourse topic), and Zaldanov responds to them. The conversation between them was transcribed and processed in ELAN. The entire interview runs approximately 52 minutes.

2. Projection One goal of this paper is to analyze how projection interacts with the turn-taking structure. Projection refers to the fact that in conversation “an individual action or part of it foreshadows another” (Auer 2005, p. 8). Projection is thus linked to temporality in discourse: an action projects onto the next time slot and makes some other action expected. Trajectory is the time course over which projection develops and comes to closure. Interlocutors have knowledge of this sequencing and how one action predicts, or projects, another. This kind of knowledge is an integral part of what fluent speakers know and is fundamental to what they do in conversation. Projection occurs at various levels in discourse, both at the topical level, across larger discourse units and more locally, at the level of the clause. With conversation, we find projection occurring across clauses, across turns, and at turn boundaries. In tracking the role of projection in Russian in terms of linguistic levels: semantic, syntactic, prosodic, phonetic and non-verbal. The very first line of the longer interview excerpt presented as example (3) is extracted here in (1) to demonstrate semantic, syntactic and prosodic projection: (1) A gost’ v programme ‘And the guest in the program’

There is semantic projection in that this first line clearly sets up a frame to name the program’s guest; syntactic projection in that this is a syntactically incomplete

222

Lenore A. Grenoble

clause, one which projects a predicate; and prosodic projection in that the rising intonation on v programme ‘in the program’ signals that speech will continue, it signals a lack of finality. Phonetic projection and non-verbal projection are discussed in detail in Sections 2.1 and 2.2. Projections may or may not align with one another. In (1), syntactic, semantic and prosodic projection align and indicate an incomplete turn. Sometimes, however, there are mismatches, e.g. between syntactic completion and prosody that projects continuation, as in line 2 of (2). Here the rising intonation (indicated ¿ to signal a rise in pitch; see Appendix for transcription conventions) strongly signals a continuing turn, but the subsequent pause of 0.8 seconds, relatively long in a conversation, coupled with syntactic completion, indicates turn completion. Zaldanov ends the pause by starting a turn, and is interrupted by Pozner, who continues his turn in line 4, overlapping with Zaldanov’s speech and gaining the floor: (2) Pozner (VP) and Chirurg (Zaldanov, AZ), (00:0023.615-00:00:44.700) 1

VP

Chirurg, ėto vse-taki prosvišče ¿ (0.8)

2 3

AZ

4

VP

1

VP

2

nu

d[a XXX [no svjazano s vašej byvšej professiej. “Surgeon,” that is, after all, a nickname ¿ (0.8)

3

AZ

4

VP

well

y[es XXX [but connected with your former profession.

These different kinds of projection combine in various ways in Russian interaction to signal turn continuation with ongoing topics or (2) new topics.

2.1. Phonetic (vocal) projection Phonetic projection refers to an array of vocal devices, including clicks, glottalizations, hesitations, modulations in voice quality, and inhalations, that project continuing turn, turn end, or turn beginning. Prosodic projection (changes in pitch) is itself a type of phonetic projection; its role in interaction is so important and salient, that it has been comparatively well-studied and warrants being treated as a separate category. Other kinds of phonetic projection have been studied somewhat for a handful of languages; Wright (2011) discusses clicks in English; Local and Walker (2012) analyze English phone conversations for both talk-projecting and turn-projecting devices; Milroy and Milroy (1997, pp. 85–7) the lack of glottalization at the ends of speaker turns in Tyneside English. Certain

Multimodality and Interactional Grammar in Russian Conversation

223

hesitation devices—such as uh [ǝ:] or mm [m::], with the final vowel or consonant lengthened—are used to “extend” the current turn; often this kind of hesitation device appears to be used when the current speaker plans to continue speaking, i.e. does not want to relinquish his/her turn, but is searching for a word or something to say. Hesitation devices such as these are often found in Russian conversation and have been noted elsewhere (see e.g. Grenoble 2013). Here I focus on only two phonetic projection devices: clicks and (audible inhalations). Their use and distribution in Russian discourse is understudied and warrants more detailed investigation on a broader corpus; here I present what at best should be considered a preliminary analysis with suggestions for how to focus future investigation.

2.1.1. Clicks as projection devices Clicks are used in Russian in two different ways. One is as a verbal gesture, indicating some sort of negative evaluation, which Gil (2013) identifies as an affective paralinguistic click. Beyond this function, they are used in regular ways in Russian interaction by some (but not all) speakers. They are not verbal gestures in the sense of the paralinguistic clicks, but rather project information about the turn-taking structure. In a preliminary study of the ORD (Odin rečevoj den’) corpus, Paschen (2014) notes that clicks occur in three different places: (1) at domain/subdomain boundaries, (2) as turn-holding devices, and (3) in strongly emotional contexts. Putting aside the third category (as belonging to the distinct category of verbal gestures), work with video recordings of Russian interaction confirms Paschen’s analysis of clicks as projecting boundaries or as projecting a continuing turn. They function in conjunction with other projection devices, including other phonetic cues. This is illustrated in example (3): the clicks in lines 8 and 10 have somewhat different functions at the local level—in line 8 it projects turn continuation, and in line 10 projects topic completion—but at the macro level these clicks indicate that the current speaker will continue talking. It is a preliminary move to turn completion. The pitch contour works in sync: pitch rises at the end of line 6 on the word kluba ‘club-gen’ and stays high throughout line 7, and then falls at the end of line 9. (3) Pozner (VP) and Chirurg (Zaldanov, AZ), (00:00:23.615-00:00:44.700) 1

VP

A gost’ v programme,

2

Aleksandr Zaldostanov,

3

on že, Chirurg ¿

4

kotorogo v Vikipedii predstavljajut uh sledujuščim obrazom. (.)

5

sovetskij i rossijskij bajker,

224

Lenore A. Grenoble

6

osnovatel’ i lider starejšego v Rossii bajkerskogo kluba,

7

Nočnye volki, (0.05)

8



click prezident Rossijskoj associacii bajkerov.

9 10



click Zdravstvujte, Aleksandr Sergeevič

11 12

AZ

13

VP

1

VP

Zdravstvujte,

[Vladimir] Vladimirič [head nod] [inhale]



Da, skažite mne, požalujsta And our guest in the program is,

2

Aleksandr Zaldostanov,

3

he is, the Surgeon]

4

who is presented in Wikipedia uh this way. (.)

5

Soviet and Russian biker,

6

the founder and leader of the oldest biker club in Russia,

7

the “Night wolves”,

8



9

President of the Russian Association of Bikers.

10



11 12

click click Hello, Aleksandr Sergeevich.

AZ

Hello,

[Vladimir] Vladimirich.

[head nod] 13 14

VP

[inhales]

 So, tell me, please,

Zaladanov utters zdravjsvujte ‘hello’ with low, relatively level intonation (line 12), starting from a low point of 95.36 Hz and rising to a peak of only 110 Hz on the stressed vowel in the first syllable, then dropping down to 78.58 Hz at the end. This utterance could be (syntactically, semantically and prosodically) a complete turn, but continues to utter his interlocutor’s name and patronymic. The transcript indicates that Pozner interprets the first word as the end of the utterance as the end of the turn. His (very) audible inhalation (line 13) overlaps with Zaldanov’s utterance Vladimir. The inhalation indicates the beginning of a new TCU, as Pozner attempts to take the floor, which he achieves in line 14.

Multimodality and Interactional Grammar in Russian Conversation

225

Pozner also often uses a click at a topic transition point, where one topic ends and a new one is picked up. This is seen in (4): (4) Pozner (VP) and Chirurg (Zaldanov, AZ), (00:11:39.720-00:11:44.670) 1

Ponjal, chorošo.

VP

2



click Ob”jasnite mne, kogda vy upotrebljaete slovo zapad (00.860)

3 4

AZ

da

1

VP

Understood, good.

2



4

click Tell me, when you use the word the West (00.860)

3 AZ

yes

In Paschen’s (2014) study of clicks in ORD, he notes that the use of clicks and other phonetic devices show considerable idiosyncratic variation. This variation is also found in my corpus. Pozner regularly uses both clicks and inhalations in predictable ways, but Zaldanov does not use them at all. Rather, he uses a considerable number of gestures, while speaking and listening. It might be tempting to conclude that this distribution is because of their different roles in the interview context: Pozner is unambiguously the interviewer: he asks questions and controls both the topical and temporal flow of the conversation, while Zaldanov is the interviewee. However, recordings of other examples of Russian interaction show that the use of phonetic devices like clicks is individual and idiosyncratic: some speakers use clicks, and some do not. Note that this is in sharp contrast to clicks in Wolof or Basaá, where they can be used by the current listener as a back-channel to signal that he or she is following what the speaker is saying (see Grenoble et al. 2015; Pillion et al. (2019)). For Russian, clicks are used by the current interlocutor to project the onset of a new sequence or as a turn-holding device. The first function is in keeping with Wright’s (2011) analysis of clicks in English interaction, where they are used to signal the beginning of new and disjunctive sequences. In both languages, these clicks function in conjunction with other phonetic projecting devices, such as pitch, pauses and (in English at least) voice quality.

2.1.2. Inhalations as projection devices One potentially surprising projection device is audible inhalations. A review of the Pozner interview shows that their use is systematic and salient: they are highly audible and thus readily noticeable. (Other non-audible intakes of breath are

226

Lenore A. Grenoble

omitted from the analysis, as are any ambiguous cases.) Pozner uses inhalations quite frequently during the interview; the distribution is similar to his use of clicks. They occur in three places: (1) turn-initially, when Pozner begins a TCU; (2) turn-finally, when Pozner ends a TCU; and (3) turn-medially, when Pozner continues an ongoing TCU in which he is current speaker. This would seem to suggest that these audible inhalations can occur anywhere in the interaction, but actually Pozner consistently inhales at turn boundaries or to hold a turn, and only when he is current speaker, not when he is the current listener. For the first eight minutes of the interview, Zaldanov does not use audible inhalations; the first occurrence is at a clear boundary, after Pozner asks a pointed question, given in English in line 0 in (5), as background for Zaldanov’s response: (5) Pozner (VP) and Chirurg (Zaldanov, AZ), (00:08:12.967-00:08:18.100) 0

VP

1

AZ

And you can’t say that you have given many interviews. Moreover, you have said, that you [inhale] avoid journalists. I don’t know why, and I don’t even know why you came here today, because after all, here with me you have, in my person as it were, you also have a journalist. 

[inhale] [RH, PD  PD] {OpenB} Ja vam otveču.

2

[RH, PDPLdiTC] {OpenB} 3

VP

nu,

4

AZ

inhale znaete, mne uh:: ja po- ispytivaju k vam interes.

5 6 1

AZ



inhale [inhale] [RH, PD  PD] {OpenB}

2

I will answer you. [RH, PDPLdiTC] {OpenB}

3

VP

Well?

4

AZ

inhale

5

You know, me uh:: I po- am interested in you.

6

inhale

Zaldanov’s turn in its entirely is given in lines 1–6. He both opens and closes the turn with an audible inhalation (lines 1 and 6); this excerpt functions as the second pair part of an adjacency pair; it is the response to Pozner’s question/statement. I now turn to a discussion of the manual gestures in (5) in Section 3.

Multimodality and Interactional Grammar in Russian Conversation

227

3. Non-verbal projection: Gestures Gesture is defined here as “visible action when it is used as an utterance or as part of an utterance” (Kendon 2004, p. 7). This description encompasses a very large range of kinesic behavior, including head and body movements, posture, eye gaze, and facial expressions. Here the discussion is limited to the to the ways movements of the hand and arm are used in interaction, and presents only a very preliminary sketch of some of the possibilities. Kendon’s seminal work on gestures (1982) is the first to classify four different kinds of gestures, which McNeill (1992) then placed on a continuum. The four points on the gesture continuum are gesticulation, pantomime, emblem, and sign language. Each of the four types of gesture, or four different points on the continuum can be roughly defined as follows (see McNeill 2005, pp. 1–12): gesticulation: motion that embodies a meaning relatable to the accompanying speech; gesticulation accompanies speech and is not conventionalized; emblems: conventionalized signs, such as thumb’s up or first finger and thumb touching to signal OK; emblems are produced with accompanying speech; pantomime: a gesture or sequence of gestures conveying narrative line; pantomime imitates functional motor activities and does not occur with speech; and sign language: signs in a sign language are conventionalized and have the properties of full linguistic signs; where such signs are iconic, their iconicity is highly constrained. (Note that Russian Sign Language is a distinct linguistic system and not relevant to the analysis of spoken Russian interaction.) The present corpus has no examples of pantomime, which is found instead of speech rather than co-occurring with it. For example, in an utterance such as: (6) And then he went [hand with extended index finger points upward] the gesture is a pantomime. Such gestures provide rich information about the relationship between language and cognition (McNeill 2005) and, To the best of my knowledge, they are unstudied in Russian. Lacking data, however, it is unclear that they function as projection devices in the same way as co-speech gestures: the upward movement of the hand in (6) is synonymous with up; it provides semantic, not interactional, content. It is interactional in the broader sense that is part of interaction and must be visible in order to be interpreted. In understanding the role of gesture in conversation, it is useful to distinguish between gestures that indicate something about the content of the utterance (representative gestures) and those that discourse devices having to do with the turn-taking

228

Lenore A. Grenoble

mechanisms and not involved in content (interactive gestures; see Abner et al. 2015, p. 438 for more discussion; Richter 2014 for discussion of the use of co-speech gestures and prosody in Russian, as well as an overview discussion of the relevance of gesture in Russian interaction). Both occur in the present corpus. Example (5), Section 2.1.2, provides a clear example of an interactive gesture.1 In this excerpt Zaldanov responds to a direct question from Pozner. He is seated facing Pozner, his right arm is extended on the table in front of him, touching from just below the elbow. The hand lies palm down on the table in rest position. As he inhales (line 1), he raises the hand off the table, maintaining the palm facing down so that it is parallel to the table at midchest height by the end of the audible inhalation. As he utters line 2, he simultaneously turns his wrist so that the palm is lateral, bending in four fingers while keeping the index finger extended, pointing in the air, in the general direction of the interviewer, Pozner. The hand then returns to the rest position. This second gesture in line 2 is more iconic, symbolically extending a hand as an offering to the interlocutor. The first gesture together with the inhalation project the onset of a new turn, which from a structural point is the second-pair part of an adjacency pair, the onset of a response to Pozner’s question. In both instances, the gestures do not add propositional content to the utterance but provide both interactional and discourse information. The first gesture, in conjunction with the inhalation in line 1, projects the onset of a new turn, and the inhalation in line 3 projects a closing. Here they project information about the conversational structure, the beginning and ending of the TCU. The co-speech gesture in line 2 points to Pozner, who is also referenced as vam (2.pl.dative) and is interactional in this sense. In the present corpus, there are many examples of gesticulations and emblems are co-speech gestures and are of particular relevance to the study of multimodality in Russian interaction. Finally, there are a number of deictic gestures in this recording, used to indicate points in space as if the object of pointing were located there (Section 3.3). The available video recordings of naturalistic conversations do not make it possible to quantify the number of gestures of any type, since neither interlocutor is filmed continuously for the duration of the interview. However, of the visible gestures, gesticulations are considerably more frequent than any other type.

1

Transcription of gestures loosely follows the conventions outlined in Bressem (2013). The specific conventions used here are in the Appendix, modified and simplified as a full transcription would require many separate tiers to capture all features. Because the modified transcription only partially describes the gestures, and because they can be cumbersome to interpret, I also describe each in prose.

Multimodality and Interactional Grammar in Russian Conversation

229

3.1. Gestures in Russian interaction: emblems Emblems are found in such examples as (7), where the speaker (Pozner) indicates quotation marks by raising both hands and flexing his index and middle fingers twice, a standard gesture to represent quotes (called air quotes or finger quotes in American English). This is a co-speech gesture and coincides with his utterance of the words v kavyčkach ‘in quotes’ in line 2: (7) Pozner (VP) and Chirurg (Zaldanov, AZ), (00:01:55.750-00:02:02.370) 1

I sejčas požaluj samoe proischodjaščee

VP

[v kavyčkach ]

2  4

AZ

mmm

5

VP

inhales

takoe vyraženie

inhales

click

6 VP

And currently, I would think, the most “occurring event,”

2

[in quotes] 

3

[air quotes]

if you can use

in

such an expression

inhales

is Ukraine

4

AZ

mmm

5

VP

inhales

6

in

ėto Ukraina.

3

1

esli možno upotrebit’

[air quotes]

click

The use of air quotes is not common in Russian conversation but rather a borrowing from English, although from the standpoint of American English it is unusual to utter the words ‘in quotes’ while gesturing the air quotes. In line 6 the inhalation is another example of its use as a phonetic demarcation of a boundary, projecting the onset of a new unit. Note that Pozner inhales before uttering the potentially problematic completion of the phrase esli možno upotrebit’ in line 2. This entire TCU (lines 1–3) invokes the politically sensitive topic of Ukraine. The literal meaning of these lines is given in the English version of (7) but the actual pragmatic nuances are more accurately captured in a looser translation as ‘And currently, I would think, the most “occurring event,” in quotes, because I’m not sure you can use [inhales] such an expression, is Ukraine.’ The inhalation both holds the turn and sets off the NP takoe vyraženie ‘such an expression’.

230

Lenore A. Grenoble

3.2. Gestures in Russian interaction: gesticulations Gesticulations frequently occur in Russian interaction. In (8) Pozner uses an iconic gesture of a rolling hand, which he produces simultaneously while uttering xoteli by vnesti ‘would like to bring in’: (8) Pozner (VP) and Chirurg (Zaldanov, AZ), (00:00:44.700-00:00:48.565) 1

Skažite mne požalujsta

VP

ničego ne

2

[xoteli by vnesti]

3

AZ

nu::: možno bylo by dobavit’,

1

VP

Tell me, please

[…]

isn’t there anything you [would like to bring] into this representation?

2

[RH, PLdiTB, circular motion 2X]

 3

v ėto predstavlenie

[RH, PldiTB, circular motion 2X] {OpenB}



AZ

well::: you could add,

[…]

Pozner begins the gesture with his right hand facing his body, and gestures with a spiral motion twice. This is a prime example of a gesture invoking the propositional content of the utterance, with the rolling spiral motion iconically representing the concept of bringing something in.

3.3. Deictic gestures Finally, a number of iconically deictic gestures are found as co-speech gestures in the Pozner interview. They are used specifically when the speaker talks about pointing at something, reinforcing the content in a manner similar to the use of gesticulations seen in Section 3.2, but the use of an index finger to point at something is unambiguously deictic. This indexical pointing motion is a basic referential gesture, and is most probably universal, although the specific motion involved in pointing varies across languages and cultures. (Some use lip or chin pointing, for example.) It is acquired by children around the age of 11 or 12 months (Tomasello et al. 2007, p. 705). Pointing is highly interactive; it is used to direct the interlocutor’s attention to some place. Deictic gestures are not necessarily purely spatial (Richter 2014, p. 1394), but can also be used in temporal or abstract contexts. In the two examples here, the speaker uses the deictic gesture to point at a space as if the referent were there. In Russian interaction, pointing typically accompanies the use of linguistic deixis. This is illustrated in (9) and (10):

Multimodality and Interactional Grammar in Russian Conversation

231

(9) Pozner (VP) and Chirurg (Zaldanov, AZ), (00:13:09.584-00:13:12.104) 1

VP

Vy ne možete emu [pal’cem pokazat’] [RH, PLdiTC, jab forward/back 2x, moves to side] {G}

2

1

VP

vot on

vot

vot

[RH, jab 1x] {G}

[RH, jab 1x]

[RH, jab 1x]

You cannot

[show him with your finger] [RH, PLdiTC, jab forward/back 2x, moves to side] {G}

2

here he is

here

here

[RH, jab 1x] {G}

[RH, jab 1x]

[RH, jab 1x]

Here as Pozner utters the end of line 1 (pal’cem pokazat’), he moves his right arm with the index finger pointed (G handshape) in front of him, elbow bent, pointing toward Zaldanov and points his finger two times, then moves his arm to his right and extends the arm more fully, pointing each time that he utters vot ‘here’. The use of the gesture underscores the act of pointing. The word vot is itself a presentational deictic, uttered when presenting something to someone or pointing something out; in this usage, it prototypically occurs accompanied by an indexical gesture as seen here. Zaldanov responds in kind, also using the same hand shape and pointing each time he utters ėtot ‘this one.masc’. His right arm is bent at the elbow and he jabs quickly, once each time as he utters ėtot, moving his hand to his right each time: (10) Pozner (VP) and Chirurg (Zaldanov, AZ), (00:13:26.141-00:13:28.021) 1

AZ

1 2

Esli by ja mog konkretno skazat’ ėtot

2 AZ

ėtot

ėtot

[RH, jab 1x] {G} [RH, jab 1x] If I could concretely say:

[RH, jab 1x]

this one

this one

this one

[RH, jab 1x] {G}

[RH, jab 1x]

[RH, jab 1x]

In both (9) and (10) the same handshape is used to point to the space where the imagined referent is, but the position of the arm and the motion is different in (9) lines 1 and 2 and in (10). It is unclear that the differences are contrastive, as opposed to being representative of different speaking styles. Both interlocutors are speaking rapidly; this is a rather heated moment in the discussion. The accompanying words are also different: Pozner uses the presentational deictic vot ‘here’ and Zaldanov the proximal demonstrative ėtot ‘this one’, with differences in meaning. As Kibrik (2010, p. 140), the use of this pointing gesture is nearly required in Russian interaction in deictic contexts where the speaker indicates a referent.

232

Lenore A. Grenoble

The overwhelming majority of gesticulations in this interview are unique to each speaker: both Pozner and Zaldanov gesture while speaking and Zaldanov, at least, also gestures while Pozner is speaking. (The camera does not always show both participants simultaneously; at times it shows only the current speaker. So it may be that Pozner gestures while he is off-camera and Zaldanov speaks; it is impossible to tell in this clip.)

4. Conclusion This paper provides evidence of the role of multi-modal devices in the organization of Russian interaction, underscoring the need for more focused research on the wide range of devices used in discourse, with particular attention to different modalities. Here I have briefly illustrated the importance of a subset of hand gestures and vocalizations. This preliminary research demonstrates the need for study of the interaction between the vocal channel (including linguistic and non-language vocal devices) and the manual channel (including not only manual gesture, but also facial expressions and body movements). Based on an analysis of the present corpus, we can make the following conclusions: 1. Clicks and inhalations have two functions as projection devices in Russian interaction. They are used to demarcate boundaries, including change of topic and the onset of a new TCU. They also serve as turn-holding devices. These two functions appear to be contradictory: on the one hand they signal the beginning on new units (topical, turns), while on the other they project continuation. Their interpretation is dependent on their position in the discourse and on the co-occurrence of other projection devices. Clicks and inhalations that occur within a unit (topical or a turn) signal that the speaker intends to continue, while those co-occurrence with other projection devices, including not only other phonetic projections (such as declination and F0 reset), but also syntactic and semantic projection, as well as gestures. These devices mutually reinforce one another. An important area of future research is the study of mismatch between projection devices, to determine if there is any ordering or hierarchical preference, in terms of both processing and production, as to which device outweighs which. For example, where there is a prosodic device projecting continuation and also syntactic completion, what is the outcome, continuation or completion? Is the outcome predictable? 2. Gestures have multiple functions as projection devices, including backchannels, turn-holding devices and as signals for turn-taking. In the present corpus

Multimodality and Interactional Grammar in Russian Conversation

233

they occur both at the edges of units, projecting openings and closings, and internally, projecting continuation. The preliminary study presented here documents their uses in all of these environments. While emblems are conventionalized, gesticulations are not, and there appears to be no upper limit on the kinds of gesticulations that can co-occur with speech. At the very least, that is an open question for future research. Furthermore, there is considerable variation in the usage of all three of these, clicks, inhalations and gestures. There has been very little research on the use of verbal and non-verbal gestures in Russian interaction. Until we have a better understanding of the systems, we cannot say if, for example, the differences in the pointing gestures that Pozner and Zaldanov use in (9) are phonetic or phonemic. The data presented here do, however, make several solid conclusions. One is that all the devices studied here are involved in projection, and are an important part of the collaborative structure of interaction. Second is that the use of clicks and inhalations is very idiosyncratic: many speakers do not use them at all. This conclusion is drawn from the corpus as a whole, and supports Paschen’s (2014; this volume) findings. Third, co-speech gesture is widespread and pervasive. Speakers in the two videos used for this study (Labovksij 2014; Pozner 2014) gesture while speaking; this is ratified by observation of any Russian interaction. (Speakers gesture even when speaking on the phone, when they are not visible to the interlocutor.) This provides fruitful ground for future work on multimodality in Russian interaction.

Appendix: Transcription Conventions CLICK = dental click IN = audible inhalation, produced without co-speech XXX = incomprehensible speech Conventions for prosody (from Ochs et al. 1996); punctuation marks are used to indicate intonation, not grammar. . , ¿ ↓ =

Falling, or final, intonation, not necessarily at the end of a sentence. Continuing intonation, not necessarily at a clause boundary. A rise that is stronger than that of the continuing intonation indicated by a comma, but not the same rise as question intonation. Sharp fall in pitch. No perceptible pause between lines. Equal signs are used in pairs, one at the end of a line and the other at the beginning of the next line,

234

Lenore A. Grenoble

(.)

A “micropause,” one that is perceptible but not easily measurable, generally lasting less than 0.2 seconds (0.9) Numbers in parentheses indicate the duration of a pause or silence, measured in tenths of a second : Lengthening of the sound preceding the colon. The more colons, the longer the sound. [ Overlap. […] Ellipsis; text omitted by author (LAG) Conventions for transcribing gestures used here (adapted from McNeill 1992, 2005 and greatly simplified; see Bressem 2013 for a more complete system, including discussion of how to code gesture in ELAN). The notation system used here is a very abbreviated shorthand for what is actually a very complex phenomenon: RH PD PL di TB TC

= right hand = palm down = palm lateral = directed = toward body = toward center

 

= movement up = movement to the side

Handshape is provided in curly brackets following the gesture notation. The ASL alphabet is used to refer to the handshape, so that {G} indicates the handshape created by the ASL manual letter G.

References Abner, Natasha / Cooperrider, Kensy / Goldin-Meadow, Susan: “Gesture for linguists: A handy primer”. Language and Linguistics Compass 9(11), 2015, pp. 437–449. Auer, Peter: “Projection in interaction and projection in grammar”. Text 25(1), 2005, pp. 7–36. Auer, Peter: “On-line syntax: Thoughts on the temporality of spoken language”. Language Sciences 31, 2009, pp. 1–13. Bressem, Jana: “A linguistic perspective on the notation of form features in gestures”. In: Müller, Cornelia / Cienki, Alan / Fricke, Ellen / Ladewig, Silva H. / McNeill, David / Teßendorf, Sedinha (eds.): Body – Language – Communication / Körper – Sprache – Kommunikation. Handbücher zur Sprach- und Kommunikationswis-

Multimodality and Interactional Grammar in Russian Conversation

235

senschaft / Handbooks of Linguistics and Communication Science. (HSK 38.1). Mouton de Gruyter: Berlin, New York 2013, pp. 1079–1098. Gil, David: “Para-linguistic usages of clicks”. In Dryer, Matthew S. / Haspelmath, Martin (eds.): The World Atlas of Language Structures Online Max Planck Institute for Evolutionary Anthropology: Leipzig 2013, retrieved 5.2.2016, from http://wals.info/chapter/142. Grenoble, Lenore A.: “Verbal gestures: toward a field-based approach to language description”. In: Plungian, Vladimir / Daniel, Michael / Fedorova, Olga / Lyutikova, Ekaterina / Tatevoso, Sergei (eds.): Jazyk. Konstanty. Peremennye: pamjati Aleksandra Evgen’eviča Kibrika [Language. Constants. Variables: In memory of A.E. Kibrik]. Аleteija: Saint Petersburg 2014, pp. 105–118. Grenoble, Lenore A.: “Talking out of turn. (Co)-constructing Russian interaction”. In: Thielemann, Nadine / Kosta, Peter (eds.): Approaches to Slavic Interaction. Amsterdam: John Benjamins 2013, pp. 17–33. Grenoble, Lenore A. / Martinović, Martina / Baglini, Rebekah: “Verbal gestures in Wolof ”. In: Kramer, Ruth / Zsiga, Elizabeth C. / Boyer, One Tlale (eds.): Selected Proceedings of the 44th Annual Conference on African Linguistics, 2015, pp. 110–121. http://www.lingref.com/cpp/acal/44/paper3130.pdf. Kendon, Adam: Gesture: Visible Action as Utterance. Cambridge University Press: Cambridge 2004. Kendon, Adam: “The study of gesture: some observations on its history”. Recherches Semiotique/Semiotic Inquiry 2(1), 1982, pp. 25–62. Kibrik, Andrej A.: “Mul’timodal’naja lingvistika”. In: Alexandrov, Yuri I./ Solovyev, Valery D. (eds.): Kognitivnye issledovanija IV. RAN: Moskva, 2010, pp. 134–152. Labkovskij, Mixail: Čto takoe l’jubov’? Uploaded 15 August 2014 by OnlineTV.ru, retrieved 2.11.2014, from https://www.onlinetv.ru/video/1712/?autostart=1. Levinson, Stephen C.: “Turn-taking in human communication – Origins and implications for language processing”. Trends in Cognitive Sciences 20(1), 2016, pp. 6–14. Local, John / Walker, Gareth: “How phonetic features project more talk”. Journal of the International Phonetic Association 42(3), 2012, pp. 255–80. Local, John / Walker, Gareth: “Methodological imperatives for investigating the phonetic organization and phonological structures of spontaneous speech”. Phonetica 62, 2005, pp. 120–30. McNeill, David: Gesture and thought. University of Chicago Press: Chicago 2005. Milroy, James / Milroy, Leslie: “Exploring the social constraints on language change”. In: Eliasson, Stig / Jahr, Ernst Håkon (eds.): Language and its Ecology. Essays in Memory of Einar Haugen. Mouton de Gruyter: Berlin 1997, pp. 75–101.

236

Lenore A. Grenoble

Ochs, Elinor / Schegloff, Emmanuel A. / Thomason, Sandra A. (eds.): Appendix. Interaction and Grammar. Cambridge University Press: Cambridge 1996, pp. 461–465. Paschen, Ludger: Click sounds in Russian: Worth investigating, worth transcribing? Paper presented at the 3rd Urban Voices Workshop, Pragmatics. Frankfurt/ Oder, 28.03.2014. Paschen, Ludger, this volume: On clicks in Russian everyday communication. In: Thielemann, Nadine / Richter, Nicole (eds.): Urban Voices – The Sociolinguistics, Grammar and Pragmatics of Spoken Russian. Peter Lang: Frankfurt am Main, Berlin, Bern, Bruxelles, New York, Oxford, Warszawa, Wien 2019, pp. 237–257. Pillion, Betsy / Grenoble, Lenore A. / Nsué Um, Emmanuel / Kopper, Sarah A.: “Verbal gestures in Cameroon”. In: Clem, Emily / Jenks, Peter / Sande, Hannah (eds.): Theory and description in African Linguistics: Annual Conference on African Linguistics. Language Science Press: Berlin 2019. http://langsci-press. org/catalog/book/192. Pozner, Vladimir: “Pozner vs. Chirurg. O Putine, motoprobege, velikoj otečesтvennoj i patriotizme”. Pervyj kanal. Aired 02 June 2014, retrieved 24.04.2015, from www.1tv.ru/sprojects_edition/si5756/fi31741. Richter, Nicole: “Structuring discourse: Observations on prosody and gesture in Russian TV-discourse”. In: Müller, Cornelia / Cienki, Alan / Fricke, Ellen / Ladewig, Silva H. / McNeill, David / Bressem, Jana (eds.): Body – Language – Communication / Körper – Sprache – Kommunikation. Handbücher zur Sprachund Kommunikationswissenschaft / Handbooks of Linguistics and Communication Science. (HSK 38.2). Mouton de Gruyter: Berlin, New York 2014, pp. 1392–1400. Thielemann, Nadine / Kosta, Peter: “Introduction and overview”. In: Thielemann, Nadine / Kosta, Peter (eds.): Approaches to Slavic Interaction. Peter Lang: Amsterdam: 2013, pp. 1–13. Tomasello, Michael / Carpenter, Malinda / Liszkowski, Ulf: “A new look at infant pointing”. Child Development 78(3), 2007, pp. 705–722. Wright, Melissa: “On clicks in English talk-in-interaction”. Journal of the International Phonetic Association 41(2), 2011, pp. 207–29. Wright, Melissa: “Clicks as markers of new sequences in English conversation”. ICPhS XVI, 2007, pp. 1069–72.

Ludger Paschen

On Clicks in Russian Everyday Communication Abstract: This paper discusses clicks produced in spontaneous Russian speech. On the basis of several case studies from the Russian speech corpus ORD, I show that clicks do not occur at arbitrary points in Russian conversations; rather, they are used in a systematic way as a means to accomplish interactional goals such as sequence management, turn-taking, and stance marking. Turn-initial clicks support self-selection at competitive turn transitions, whereas sequence-initial clicks often co-occur with one or more discursive breakpoints. The ORD data suggest a high inter-speaker variability concerning form, function and frequency of clicks. By identifying clicks as part of a meaningful practice in Russian, this paper contributes to a better understanding of how speakers of Russian employ phonetic resources in talk-in-interaction.

1. Introduction One of the core assumptions of analytical frameworks for spoken interaction such as Conversation Analysis (CA) is that speakers make systematic use of verbal resources to help accomplish interactional tasks such as the organisation of turn-taking. Subtle phonetic cues, including rhythmic structure, voice quality, and spectral characteristics (Couper-Kuhlen 1993; Ogden 2012) have been shown to play an important role in this process. Despite the growing interest in these phenomena, many of those resources are still obscure, and little is known about their cross-linguistic distribution. Click sounds in particular have only recently been recognised as a relevant device by which speakers indicate meanings relevant in discourse (Wright 2005, 2007, 2011ab; Ogden 2013; Kendrick & Torreira 2015; Trouvain 2014, 2015; Trouvain & Malisz 2016). The bulk of research on clicks has been confined to data from English (and, to a lesser extent, German), and much of our knowledge about clicks to date has been informed by the study of telephone calls, a special genre which differs from face-to-face interaction in crucial ways (Schegloff & Sacks 1973; Schegloff 1979; Biber & Conrad 2009). This paper is set out to contribute to the ongoing discussion on clicks by examining conversational data from Russian covering a wider range of spontaneous interactions in everyday situations. The main finding is that speakers of Russian make systematic use of clicks to achieve interactional goals such as turn-taking and stance marking. The paper is organised as follows. In section 2, I will give an overview of previous research on clicks in interaction. In section 3, I will present

238

Ludger Paschen

six case studies, each of them discussing a conversation fragment from Russian everyday speech in which click sounds are used. Lastly, section 4 highlights the importance of accessible and reliable acoustic data.

2. Background: Clicks in oral communication Click sounds or clicks are commonly defined as stops produced with velaric ingressive airstream and a loud burst (Ladefoged & Maddieson 1996). Click sounds are regular members of the consonant phoneme inventories in Khoisan and in some neighbouring Bantu languages (Gil 2013). In these “click languages“, up to five primary places of articulation for click sounds are distinguished (bilabial [ʘ], dental [ǀ], lateral [ǁ], alveolar [!], and palatal [ǂ]), and combinations of plain clicks with different phonation types, secondary articulations and other (posterior) sounds give rise to unusually large sound inventories. One example is the Northern Khoisan language Juǀ’hoan, which is reported to have 46 non-click and 48 click consonants, giving a total of 94 consonants (Miller 2013). Occurrences of clicks are not restricted to click languages, however, as they are frequently produced by speakers of non-click languages as well. A first distinction that has to be made is that between voluntarily and involuntarily produced clicks, the latter referring to clicks that are produced as epiphenomenal articulatory artefacts. Using recordings from a text reading task with German speakers, Simpson (2007) finds consistent occurrences of velar airstream mechanisms at word junctures with a final coronal voiceless stop and an initial velar stop that closely resemble articulatory movements found in the production of „true“ clicks. The study further suggests that glottal airstream mechanisms may arise at word junctures with a final stop and an initial preglottalised vowel, though no indication is given of how consistent this pattern is across speakers. Voluntary usage of clicks is sometimes labelled para(-)linguistic, though a clear definition of what is meant by this term is missing in many cases1. Gil (2013) includes logical and affective meanings („yes“, „no“; „positive“, „negative“) as instances of „para-linguistic usages“ of clicks in his typological survey. He mentions other contexts in which clicks may also be used, including turn-taking and addressing animals or babies, but it is not explicitly stated if he considers such clicks „para-linguistic“, too. Wright (2007) acknowledges „paralinguistic“ usage of clicks, defined as referring to the „emotional and attitudinal state of the speaker“

1

Notable exceptions are the careful discussions of the terms paralinguistic and paralanguage in Laver (1994: 22–25) and Ladd (1996: 34–42).

On Clicks in Russian Everyday Communication

239

(p. 1069), but argues clicks in English are not restricted to this particular function. Instead, clicks are used systematically to signal the onset of new and disjunctive sequences. It thus appears that the label para(-)linguistic is most often used as a cover term for pragmatic and discourse-related functions that are not readily compatible with pre-defined linguistic categories, and a debate that centres on whether or not clicks are used paralinguistically seems not to be particularly helpful in informing the enquiry as to how exactly they contribute to the interactional meaning of an utterance. Instead, it appears more reasonable to focus on the concrete interactional settings that clicks occur in, taking into account all potentially relevant discursive parameters, and examine their form and function within a solid theoretical model of verbal interaction. I will now briefly review some relevant contexts that have been identified as attracting click sounds in the growing body of literature on that topic. Wright (2005) in her seminal study finds that clicks in English telephone conversations occur between the first and second units of multi-unit first-closing turns where they signal the beginning of the second unit (this context is discussed further in Wright 2011a), at the beginning of new and disjunctive sequences (see also Wright 2007), and as a turn-holding device in word searching environments. Wright (2011b) shows that initiation sequences of closings in English telephone calls have specific phonetic designs, including click sounds and prosodic resources such as pitch and voice quality. Ogden (2013) confirms the findings in Wright (2011ab) and argues that one should distinguish three main functions of clicks in English: a) clicks marking incipient speakership, b) clicks serving the goal of sequence management, including word search and indexing new sequences of talk, and c) clicks used as part of a construction that displays a stance. Trouvain (2015) reports clicks occuring in sequence-initial position and in word-search environments in German. The former context is the more frequent one and includes clicks before feedback tokens such as ja ‘yes’ or ach so ‘I see’, clicks at the onset of prosodic units, and clicks indicating incipient speakership. Curiously, the functional environment of about one fourth of all click occurrences in Trouvain’s corpus could not be determined with certainty. It should be noted that the line between voluntarily and involuntarily produced clicks is not always easy to draw. A peculiar example is the case of percussives discussed in Ogden (2013). Percussives are „sounds made as the articulators separate and the speaker prepares to speak“ (p. 302), and they are „close relatives“ (p. 301) to „true” clicks produced independently of word-initial articulatory movements (see also Schaeffler et al. 2015). He observes that some English speakers produce percussives in the same interactional contexts in which voluntary clicks occur,

240

Ludger Paschen

too. This strongly suggest that all click events are potential candidates for being meaningful pieces in a conversation.2 While there exist a number of studies on how speakers of Russian convey relevant discursive meanings in talk-in-interaction (a.o. Graf 2011; Bolden 2016; Thielemann this volume), so far no attention has been paid to the role of click sounds to my knowledge. Against this background, the next chapter will show that Russian speakers do make systematic use of click sounds as a phonetic resource in similar contexts as speakers of English and German.

3. Clicks in Russian conversation My analysis is grounded in the theoretical framework of Conversation Analysis (Sacks et al. 1974 and much subsequent work), extended by a discourse analysis component based on knowledge domains (Sappok 2010). The latter approach, originally developed for annotating Russian dialect speech in the RuReg database (Sappok et al. 2016), highlights the interaction of five distinct knowledge domains (thematic, textual, linguistic, situational, voice) in oral communication. Each domain contains several sub-domains (e.g. reaction and initiative in the situational domain), which in turn provide a fine-grained system of descriptive labels. The model predicts co-occurrences of discourse-related events and (sub-)domain breakpoints to reveal the key mechanisms of how speakers signal important places in monologic and dialogic speech. Sappok’s domain-based appraoch is by and large compatible with the basic assumptions of Conversation Analysis because most of the crucial concepts of the latter have correspondences in one or more (sub-)domains; sequence management, for instance, belongs to the situational domain, while prosodic features are encompassed by the linguistic domain.

3.1. The ORD speech corpus For the purpose of the current study, recordings of four speakers from the ORD corpus of Russian everyday speech (Šerstinova 2009) were analysed. The ORD corpus contains a vast number of long-time recordings of Russian speakers in St Petersburg who carried the recording device with them over the course of a whole day. The ORD corpus includes male and female speakers from various age groups and social backgrounds. ORD is particularly well-suited for exploring

2

In a similar vein, Simpson (2014) hypothesises that ejectives in German and English are not always epiphemonenal but may present a sociophonetic and/or interactional resource that speakers make conscious use of.

On Clicks in Russian Everyday Communication

241

the nature of spontaneous speech in real-life interactions across a representative cross-section of contemporary urban speakers of Russian. In what follows, I will discuss six sequences from the ORD corpus illustrating the three main functional contexts in which I found clicks to occur in Russian: domain boundaries, turn transitions, and stance marking. Transcripts follow the basic GAT2 guidelines (Selting et al. 2009) and only contain special characters that are relevant to the current discussion. Click events were identified on the basis of their distinct auditory impression and acoustic fingerprints (see figures 1–3 and discussion in section 4).

3.2. Clicks at domain boundaries The first example is an excerpt from the first recording of speaker S5, a 27 year old female lecturer in psychology, from a scene at her home in the morning. It is situated within a larger narrative sequence of about 3 minutes in which S5 tells her neighbours, M11, M13 and F12, about her first driving lesson, which she took earlier that morning. The small fragment presented in example (1) begins one minute into this sequence and lasts 19 seconds; only one of her interlocutors, F12, can be heard speaking in it. Example (1): ((ordS05-01.183-202)) 01 02 03 04 05 06 07 08 09 10 11 12 13

S5: F12: S5: F12: S5: →

znaeš' po imeni srazu takoj tak sprosil kak po imeni °h vot Lenočka davaj tak [otžima:j tak tichonečko] [((laughs)) ] da da da da da [vot tak vot] znaeš' °h [((laughs)) ] (-) (.)nu (.) normal'no tak vyechali po trasse tam poechali

01 02 03 04 05 06 07 08 09 10 11 12 13

S5: F12: S5: F12: S5: →

you know by name right from the start he asked for my name °h ok Lenochka go that way [relea:se the clutch slowly] [((laughs)) ] yeah yeah yeah yeah yeah [that's it ] you know °h [((laughs))] (-) (.)well (.) okay we drove out on the road we drove there

242

Ludger Paschen

S5’s interlocutors learn that the instructor addressed her by her first name instead of the more formal combination of first name and father’s name3, right from the beginning (01–02). S5 is then giving her interlocutors a vivid image of her driving instructor in a reported speech sequence that is characterised by an overall lower pitch register, with which S5 imitates the instructor’s voice (03–06). S5 concludes that her driving instructor is obviously used to working with beginners such as herself (08–10). Lines 08–10 form the descriptive peak of the whole fragment. The use of smile voice, which turns into laughter in line 10, reflect the general humorous setting of this scene, flanked by the choice of the slightly derogative self-designation lochuška ‘idiot’ in l. 09, with which S5 describes herself in an ironic way to underline the discrepancy between her knowledgable instructor and her own apparent lack of driving skills. The humorous setting is echoed by F12 several times during the conversation by shorts bursts of laughter (05, 07). The descriptive sequence is ended by the closing particle vo:t ‘ri:ght’ in l. 11, before S5 takes up the narrative portion from lines 01–06 up again (12–13). It is at the beginning of l. 12, i.e. at the beginning of the new (renewed) narrative sequence, that S5 produces a clearly audible dental click (see fig. 1). Several discursive dimensions are involved in the environment of this click. First, the click is relevant for turn-taking, as it marks incipient speakership at a TRP lacking overt speaker selection. F12 could have taken the floor after the closing particle vo:t to make a remark or to ask a clarification question, for instance. Instead, it is S5 who self-selects. Second, the click serves as a disjunction marker, as it occurs at the boundary between a descriptive and a narrative sequence on the meso-level, i.e. at a boundary between two textual subdomains. Note that no clicks are produced at minor domain breaks, e.g. at the transition to the reported speech sequence (03) which is embedded in the first narrative portion (01–06). Third, another discourse dimension that can be identified as having a major breakpoint at the time of the click is voice quality. For an extended period of time (7 seconds), S5 speaks in smile voice, as indicated by spectral characteristics and an overall high F0 (Drahota et al. 2008). The new click-initial sequence, however, is accompanied by a neutral voice quality. Moreover, it is worth noting that the click is not accompanied by an inbreath. Clicks frequently co-occur with

3

As it is highly marked to address someone by his official first name alone in Russian, it is not surprising to hear S5 have her driving instructor pronounce a diminutive form, Lenočka ‘Lenochka’, when re-enacting the driving lesson scene. The use of several diminutive suffixes on top of each other in Lenočka ‘Lenochka’ underlines the image of an intimate relation with the instructor that S5 wants to create.

On Clicks in Russian Everyday Communication

243

audible inhalation, either preceding or following the burst (Wright 2007, 2011a; Trouvain 2015). S5 does produce inbreaths at other sequence boundaries (02, 06, 10), however. Lastly, it should be pointed out that it is not the click alone that fulfils a disjunctive function here, but rather the combination of the click, the particle nu ‘well’ projecting more talk to come (see discussion of ex. 2 below), a change in voice quality (from smile to normal) and a pitch reset. Clicks followed by a particle or an interjection have been observed to be a frequent collocation used for disjunction marking in English, too (Ogden 2013). Figure 1: A dental click prefacing the beginning of a new TCU and a new sequence (ex. 1). A secondary burst following the primary burst after about 5 msec is a typical acoustic diagnostic of clicks and can be observed in this example as well. Also note the creaky portion in the micro-pause following the particle nu by which the particle is connected to the rest of the TCU.

Let us now consider another example of clicks co-occurring with domain boundaries from the same driving lesson episode. The fragment presented in example (2) starts 16 seconds after the end of the fragment presented in ex. (1) above. S5 produces a dental click inside a larger narrative sequence in which she re-enacts several events happening in rapid succession: from regular driving exercises (01–02) to a near miss (03–04) that was fortunately resolved by the driving instructor’s intervention (05). The driving lesson then proceeded as usual (06–07), and next, the instructor pointed at an object in his field of vision (08–09). At the beginning of the new TCU in l. 05, S5 produces a dental click. The click occurs

244

Ludger Paschen

in between two events that are part of a list of events within an extensive narrative sequence. At the same time, it is also placed in between two thematic sequences, one about the driving instructor’s actions (05–09) and another about S5’s own actions (01–04). Thus, the click again occurs not only at the beginning of a TCU but also at discursive breakpoints in the textual and the thematic domains. The click could potentially also be interpreted as projecting the upcoming reported speech sequence in 09, i.e. the relevant boundary in the situational domain. However, since the interval between the projected sequence and the projector would be rather large (ca. 5–6 seconds), the projection function does not feature as prominently as the other functions. While the fragment in ex. (2) is monologic, it is taken from a larger polylogic episode. One could therefore speculate that the click is also used to establish S5’s self-selection after the end of her previous TCU. In section 3.3, I will present more straighforward examples of clicks signalling self-selection at turn transitions. Here, the click seems to be primarily used to structure the speaker’s own monologue (see Trouvain & Malisz 2016 for a study of clicks as hesitation markers in monologic speech). Note, however, that the distinction between monologic and dialogic speech is not a trivial one (Sappok this volume). Example (2): ((ordS05-01.218-233)) 01 02 03 04 05 06 07 08 09

S5: →

a ja tam nu ne uverena mogu tam chop tam v odin moment tam čut' v jamu ne v''echala (-) °hnu on tam vyrulil konečno potom poechali °h poechali na druguju ploščadku zaechali i on takoj mne govorit vot vidiš' tot (.) kust šipovnika

01 02 03 04 05 06 07 08 09

S5: →

and I am well I'm not sure if I can whoops suddenly I'm almost driving into a hole (-) °h well he got us out of course and then we drove on °h onto another square and he is like do you see that (.) rose bush over there

In contrast to the first example, the click in ex. 2 is directly followed by an inbreath. While there are several more instances of inbreath-click combinations attested in the whole ORD corpus, the one in ex. 2 is the only such case out of the seven conversation fragments discussed in this paper. Similar to ex. 1 above, the click is also followed by a discourse particle. The polyfunctional particle nu ‘well’ is omnipresent in Russian spoken interaction and has been

On Clicks in Russian Everyday Communication

245

described as a resource for signalling sequence transitions and for prefacing talk, among many other things (Graf 2011; Bolden 2016). Here, a combination of click, inbreath, and particle is used to simultaneously mark several discursive breakpoints at once.

3.3. Clicks at competitive turn transitions I will now turn to cases of clicks occurring at the beginning of new turns. In example (3) below, the setting is similar to the ones discussed in the previous section, as the conversation takes place at S5’s home the same day later that morning and the topics of the conversation are again matters of everyday life. S5’s interlocutors are some of her neighbours, among them F1, who can be heard in example (3). The fragment starts with S5 talking about a specific piece of clothing, a coat (01–04). She suggests to F1 that the current weather allows for wearing this type of garment. When S5 completes her suggesting sequence at the end of l. 04, she reaches the end of a syntactic and a prosodic unit (in the sense of Szczepek Reed 2010) and thus a TRP, at which point F1 takes the floor to comment on S5’s statement. However, until the end of l. 05, it is not quite obvious what exactly F1 wants to say: her token of acceptance (konečno ‘of course’) is immediately followed by the negative particle ne ‘no’, and no clear indication is given as to what exactly the negation relates to. The final particle vot ‘well’, which signals the end of a non-final sequence, is accompanied by an accelerando, a prosodic feature that is frequently used as a turn-holding device. However, the pitch contour in the passage in l. 05 does not signal incompleteness and the overall pitch level is low, so that the short passage seems to both project and not project more talk to come. When F1 continues her turn in l. 06, she gets interrupted by S5, who has obviously interpreted the ambiguous signals in l. 05 in terms of a TRP with no strong self-selectional preferences from F1. This results in a considerable amout of overlap of about 500 msec (06–07). At this competitive turn transition, S5 produces a dental click. She then continues her turn in which she talks about another piece of clothing, a scarf (07–08); this turn has a total duration of 14 seconds, of which the first 3.3 seconds are shown in the transcript. Example (3): ((ordS05-03.247-258)) 01 S5: 02 03 04 05 F1: 06 07 → S5: 08

ja prosto sejčas dumaju vsë ravno pogoda ėto cholodnaja °h možno uže vot ė-ė odevat' vot ėto sinee pal'to tvoë (.) ka- konečno ne vot budetpo[teplee ] [jaznaeš'] ja choču takoj birjuzovogo cveta šarfik k nemu kupit'

246

Ludger Paschen

01 S5: 02 03 04 05 F1: 06 07 → S5: 08

I simply think now the weather it's cold anyway °h you can already uh wear this blue coat of yours (.) of course no well itwillbewar[mer ] [youknowI] I want to buy a turquoise scarf to wear with it

S5’s incoming is competitive for at least two reasons. First, F1 does not clearly indicate that there is a potential TRP at the end of line 5, let alone explicitly give the floor to S5, and second, F1 has already been speaking for 1.7 seconds at the point when S5 comes in with her turn. The outcome of this competition is in an overlap of about 500 msec and, eventually, the end of F1’s turn. It is intriguing that the main phonetic resource for establishing the new turn in this situation is the click, as S5’s voice is not particularly loud, high-pitched, or fast-going, and no other deviations from her usual prosody can be detected, either. F1 does not show any resistance to S5’s incoming. Thus, with the help of the turn-initial click, S5 achieves her interactional goal of reclaiming the floor to take up her previous topic. It should be mentioned that the click also co-occurs with a micro-thematic breakpoint at which the focus of attention is temporarily shifted from the coat to the scarf, but it is obvious that signalling a breakpoint within the thematic domain only plays a subordinate role here, the primary function of the click being to mark incipient speakership at a competitive turn transition. Another case of a click occuring at a competitive turn transition is presented in example (4) below. The fragment is taken from a longer episode in which S5 is preparing her classes which she is going to teach later that day. S5 is talking to several of her colleagues at the department, among them F2. The fragment in ex. (4) begins 25 seconds into a sequence in which F2 and S5 discuss one of F2’s seminars. F2 makes a somewhat worried impression, and S5, though busy preparing her own courses and talking to various other people as well, offers her a helping hand by going through some of the relevant literature with her. After the end of ex. (4), the conversation proceeds for another 42 seconds; at a later point in the recording, F2 can be heard thanking S5 for her assistance. Example (4): ((ordS05-07.175-195)) 01 02 03 04 05 06 07 08 09

S5: F2: S5: F2:

vtoroj ėto Abramova no Abramova idët čisto tol'ko po vozrastnoj psichologii °h i to nu ty predstavljaeš' kak Abramova pišet čitala net net [( )] [nu u neë znaeš'] (-) ė-ė-[ė ] [chotja net ] kakuju-to ja čto-to ja

On Clicks in Russian Everyday Communication 10 11 → S5: 12 13 14

čita[l ]a [] vobščem (.) razol'ëmsja (.) ė-ė smyslom po drevu nazyvaetsja da (-) °hh vot ė-ė to est' u neë po suti dela ėto kak ne učebnik da

01 S5: 02 03 04 05 06 F2: 07 S5: 08 09 F2: 10 11 → S5: 12 13 14

the second is Abramova but Abramova is only about developmental psychology °h and that's well you imagine how Abramova writes you read it didn't you no [( )] [well she you know] (-) u:[h ] [but no ] there is one that I didre[a ]d [] anyway (.) she's getting (.) uh completely lost in details right (-) °hh ok uh so her book is in principle like not a textbook right

247

In lines 01-03, S5 introduces the second book that she wants to suggest to F2. In order to assess the common ground, S5 asks whether F2 has read something from the literature in question (04-05), and after F2 negates that question (06), S5 initiates a new turn in which she tries to give more details on the book in question (07)4. The second TCU of that turn begins with the hesitation marker ė-ė-ė ‘u:h’, the final portion of which is accompanied by strong creak (08). F2 interprets this as a TRP at which she self-selects to conduct late self-repair, as she has suddenly realised that she did in fact read one of the books by the author that S5 asked her about. F2’s incoming in l. 09 is clearly non-cooperative, as it is in no way directed towards assisting S5 in her search for words (08). F2 thus ranks the protection of her own face above the face-threatening impolite act of interrupting her interlocutor. S5 does to not react to the update that F2 has just given her. Instead, she is more concerned about keeping up her (temporary) dominant role as an advisor, and pushes forward to close the question-answer insert sequence (05–06; 09–10). To that end, she initiates a new turn starting with a dental click in l. 11., where she eventually proceeds with her lecture from ll. 01–04. The click overlaps with the second to last segment of F2’s turn and is followed by the particle v obščem ‘anyway’, which serves as a closing bracket to the insert sequence.

4

Unfortunately, a small portion of what F2 says after the net ‘no’ in l. 06 cannot be deciphered due to S5 speaking at the same time.

248

Ludger Paschen

Since the overlap at the beginning of S5’s turn in l. 11 is only about 125 msec long, one could raise the question if the turn transition is genuinely competitive. In fact, it is not trivial to decide whether F2’s short turn in ll. 09–10 projects more talk to come. The turn-final TCU какую-что я что-то читала ‘there is one that I did read’ (09–10) is syntactically and prosodically complete, and it is safe to assume that S5, anticipating the end of F2’s turn, interprets the temporal dimensions of the transition space at the end of line 10 rather generously. However, the unresolved reference of kakuju-to […] čto-to, literally ‘some [book] […] something’ in F2’s turn (09) might project more talk to come, and one easily imagine an alternate scenario in which F2, being eager to further elaborate on the one book that she recalls having read, challenges S5’s incoming in l. 11. Assuming that S5 at least entertains the possibility that F2 wants to continue her turn after ll. 09–10, and considering the small but not insignificant portion of overlap, the click can tentatively be described as occurring at a “mildly competitive” turn transition. Lastly, S5’s click-turn in ll. 11 thus can also be ascribed a sequence-managing function. In the episode that ex. (4) is taken from, S5 is conducting a series of descriptive and mostly monologic sequences. Her question in l. 05 is built as a short dialogical insert sequence that is virtually finished in l. 06 but is then taken up again by F2 shortly afterwards (09–10). The click in l. 11 thus signals both a return to the descriptive monologic sequence on the textual and to the previous topic from ll. 01–04 on the thematic5 level.

3.4. Affective clicks and stance marking Clicks have been reported to convey meanings associated to emotionally loaded stances in a large group of languages including Russian (Gil 2013), and the ORD corpus indeed contains several cases in which a click can be interpreted 5

The English translation of razol’ëmsja […] smyslom po drevu (11–12) as ‘she’s getting […] completely lost in details’ in the transcript may be misleading in the sense that the deictic element she referring to Abramova in l. 04 is not present in the Russian original and can thus not directly co-constitute S5's return to the previous topic. The quote is an idiosyncratic deviation from the idiom rastekat’sja mysl’ju po drevu, which literally means ‘to get one’s thought spilled on a piece of wood’; the literal translation of what S5 says is ‘let us get [our] sense spilled on a piece of wood’. Since it is clearly the book by Abramova that S5 is criticising for being too unstructured, there is still a (covert) aboutness link to that topic. The packaging of her criticism within an ironically used first person plural hortative construction serves to ease up the potentially tensed atmosphere between the two colleagues after the competitive turn transition.

On Clicks in Russian Everyday Communication

249

as expressing a negative stance. One such case is presented in example (5), in which S24, a 63-year-old female lecturer in history, is experiencing technical issues with her computer while she is working at home alone. At some point after turning on her PC (01)6, a typical MS Windows error sound can be heard (03). The dental click produced by S24 shortly afterwards (04) is a direct reaction to the puzzling noise that is now calling for her attention and is immediately followed by a short verbal exclamation (čto takoe ‘what’s that’). The click thus suggests a negative stance towards the unpleasant situation, though it is nontrivial to pinpoint the exact pragmatic content of the click compared to the čto takoe – perhaps a more adequate way to characterise the click would be to say that the click and the exclamation are part of a larger construction which in its totality marks a negative stance. Example (5): ((ordS24-23.0-19)) 01 02 S24: 03 04 → S24:

((Windows startup sound)) (3.5) (tak) ((Windows error sound)) (1.1) čtotakoe(0.9) bam bam

01 02 S24: 03 04 → S24:

((Windows startup sound)) (3.5) (ok) ((Windows error sound)) (1.1) what'sthat(0.9) bam bam

Another case of a click used in a stance-marking construction is presented in example (6). The fragment is a part of a telephone conversation in which S35, a 70-year-old male construction engineer, tries to clarify some accountancy-related questions as part of his daily working routine. The parts spoken by his interlocutor are muffled and cannot be deciphered, but from him addressing his interlocutor shortly before the beginning of the fragment in ex. (6) it is certain that he is talking to a female. Example (6): ((ordS35-06.698-719)) 01 S35: 02 03 04 → 05 06 07 08

6

vy znaete u menja °h ne chotel ja est' tut sto ė-ė printery kupili sto (--) tridcat' s čem-to tut tysjač (--) (.)ojnutakneochota pisat' eë ėtu(1.9) m::: da no (1.5) nu ladno chorošo oj (-) oj ženščiny ėti večno (1.3) večno ugovorjat aga (-) nu vsë

There is a small portion of silence in the recording between lines 02 and 03 hinting at a cut-out portion of unknown length, which makes it impossible to say how much time exactly elapsed between the two acoustic events in lines 01 and 03.

250 01 S35: 02 03 04 → 05 06 07 08

Ludger Paschen you know I have °h I didn't want here there are a hundred uh bought printers a hundred (--) and thirty or more thousand (--) (.)ohwellIdon'tfeellike writing it this (1.9) hm::: yeah but (1.5) well alright oh (-) oh these women always (1.3) always persuade you (-) alright then

The click produced in line 04 is the only bilabial click in the small sample of conversation fragments containing click sounds in this study. The click is directly connected to the following sequence (04–05) in which S35 expresses his lack of motivation (oj nu tak neochota ‘oh well I don’t feel like’) to perform a certain kind of work (pisat’ её ėtu ‘writing it this’) that was discussed before (01–03). These first three lines clearly contain business-related talk and thus belong to the other mode or the informational discourse level (Yokoyama 1993). When S35 decides that his negotiation is to no avail and that he will have to conform to the requests of his business partner, he switches into the svoj mode or the metinformational level, lamenting his unwillingness to fulfil the task at hand (04–05), only to announce indirectly shortly afterwards that he will eventually comply with what he has been told to do (07–08). S35 instantiates the change into svoj mode by pronouncing a bilabial click (04) and drawing on general socially accepted themes (lack of motivation to work in ll. 04–05) before he goes on to playing on gender clichés (oj ženščiny ėti večno (1.3) večno ugovorjat ‘oh these women always (1.3) always persuade you’ in lines 07–08). The frequent use of the interjection ой ‘oh’ emphasises the inter-personal (svoj) level which S35 has steered the conversation to. The click also initiates a new thematic sequence in which S35 shifts the topic from accounting-related issues to personal and emotional matters. It is important to note, however, that the click is not at all used affectively, i.e. as an unintentional and uncontrolled verbal reaction to an emotional trigger. Rather, it is used support a deliberate change in conversation mode. The ways speakers indicate such a shift can be very subtle, down to dropping of isolated features, and the click in ex. 6 may well be considered such a modal feature. Note also that the transition into svoj mode is accomplished by use of a construction containing a click, and not by the click on its own. The construction also foreshadows the end of the telephone conversation, cf. the closing formula nu vsё ‘alright then’ in l. 08 after which the conversation comes to an end. The conversation fragment in ex. (6) thus reveals interesting parallels but also subtle differences to the opening up closings-function identified for clicks in English telephone conversation (Wright 2005).

On Clicks in Russian Everyday Communication

251

Figure 2: A bilabial click in a construction marking stance produced by speaker S35. Note the rising-falling intonation contour on the unusually long vowel segment in the interjection oj and the short rise on the particle nu.

4. Further discussion In spite of the sometimes noisy nature of the recordings in the ORD corpus, which is due to speakers carrying the recording devices with them during their daily routines, all of the clicks discussed in the previous paragraph could be clearly recognised as click consonants articulated by one particular speaker. However, the ORD corpus also contains cases in which it is difficult to ascertain whether an acoustic event that might be a click is indeed a click, or which speaker a click sound has to be attributed to. This can be related to recording artefacts sharing some of the acoustic properties of clicks: a short burst with energy concentration at a specific frequency range, possibly followed by a secondary lower-energy burst. It is therefore important to carefully evaluate whether a potential click is indeed a speech sound in order to avoid misinterpretation of other acoustic events that are in fact are irrelevant for a specific interactional sequence. Consider example (7) below. The fragment is taken from a longer episode in which S5 is teaching a class of psychology students at St Petersburg State University. The students returned from a trip abroad only the day before, and S5 has to deal with a low level of student motivation and preparation for the class. At the beginning of line 06, a clicking sound can be heard, and it would be tempting to assume that this is a dental click uttered by S5. In the particular dynamics that

252

Ludger Paschen

are characteristic of classroom settings (Mehan 1979; Becker-Mrotzek & Vogt 2009), each individual can be an interlocutor by its own right, but at the same time, both the teacher and the students may also negotiate a group identity for the latter. When that is the case, the number of interlocutors is effectively reduced to two. Although her role as teacher would normally give S5 certain priorities with respect to turn-taking, the circumstances on that particular day lead to a situation in which S5 is in constant struggle to execute her right to speak and to reduce the noise level of the group of students to a minimum. The idea of her producing a click to support her efforts fits well into this scenario, also considering that the click in line 06 could be said to project a longer monologic sequence (05–10 and beyond: in the recording, S5’s address to the class goes on for several minutes after the end of l. 10), although it does not coincide with any of the previously discussed domain boundaries. Example (7): ((ordS05-08.235-252)) 01 S5: 02 03 M1: 04 M2: 05 S5: 06 → 07 08 09 10

°hh davajte tak my da my projavim (-) čërt voz'mi ((nrb.)) my projavim drug drugu uvaženie (1.0) (?) da i ponimanie (1.0) ja projavljaju so svoej storony ponimanie k vašej situacii k tomu čto vy ne možete mne segodnja čto-to tam otvečat' i rasskazyvat'

01 S5: 02 03 M1: 04 M2: 05 S5: 06 → 07 08 09 10

°hh let's do it this way we yes we will (-) damn it ((unclear)) we will treat each other with respect (1.0) (?) and with understanding (1.0) I am going to acknowledge your situation that today you cannot give me answers and tell me

The audible impression of the alleged click, however, is inconclusive, and a closer look at the oscillogram and the spectrogram (figure 3) suggests that the source of this sound is indeed far from evident. The clicking noise displays the characteristic dual burst, but it lacks in frequency regions below 3 kHz, which is fairly atypical of dental clicks (cf. figure 1 as well as the acoustic data in Wright 2011a and Ogden 2013). In addition, it can be seen that the overall noise level even in the “silent” intervals is fairly high, with the clicking sound barely standing out distinctly against the background noise. The status of the click is thus dubious, and the analysis of its potential functions must consequently be treated with due caution.

On Clicks in Russian Everyday Communication

253

Figure 3: Oscillogram and spectrogram of a potential click (ex. 7). Two small bursts occurring in rapid succession (3 msec) are visible ca. 10 msec before the beginning of the next TCU. Note the general noisiness of the fragment, especially during the silent intervals before and after the acoustic event in question.

Another interesting observation from the ORD data is that there are strong indivual preferences when it comes to usage of clicks. For the purpose of the current study, four full one-day recordings from the ORD corpus were systematically checked for click sounds. Speaker S5 (female) regularly produced a number of clicks at domain boundaries and at competitive turn transitions. Speakers S24 (female) and S35 (male) produced only a few clicks, almost all of them in emotional contexts. S15 (male) did not produce any clicks at all. Previous studies suggest that the degree of inter-speaker variation in the production of clicks may well be different across languages. Trouvain (2015) reports huge inter-speaker variation among a group of 6 male and 6 female speakers of German, from those who produced no clicks at all and those who produced only one or two clicks to those who produced up to 19 clicks within 10 minutes of dialogue. On the other hand, Gold et al. (2013) found that 100 speakers of Standard Southern British English did not vary substantially from one another in their clicking behavior, and that there was an intricate within-conversation instability for extensive click users. The data from the ORD corpus suggest that Russian is similar to German in that usage of clicks is highly individualized, making it a strong candidate for being a sociophonetic and not only an interactional resource. However, more

254

Ludger Paschen

extensive research is required to study if and how speakers of Russian make use of clicks to construct social meaning. Finally, it should not be overlooked that clicks are still sorely neglected in contemporary speech corpora despite the fact that they have been recognised as an important phonetic resource for talk-in-interaction for more than a decade. This neglegence manifests itself in the fact that clicks are often not annotated at all or subsumed under a “misc/non-linguistic/other/…” label but not marked separately. For instance, most sub-corpora of the German spoken language database DGD2 (IDS 2017) lack annotations of click sounds altogether. In the few sub-corpora that do include clicks (FOLK, Berliner Wende), they are annotated according to the guidelines for „nichtsprachliche Ereignisse“ (‘non-linguistic events’), which means that they are mentioned in the transcript (e.g. „((schmatzt))“ ‘munches’, „((schnalzt))“ ‘clicks’) but cannot be searched for via the web interface. Future research on talk-in-interaction would without doubt benefit from adding clicks to the inventory of common annotation symbols.

5. Conclusion In this paper, I have argued that speakers of Russians make systematic use of click sounds in everyday conversations to convey interactional meanings. Clicks occur in at least three functional contexts: at sequence (domain) boundaries, at competitive turn transitions, and in emotionally non-neutral situations. More specifically, Russian speakers employ clicks for marking the beginning of a new thematic, textual, or actional sequence, as a device to help conduct self-selection at competitive turn transitions, and for prefacing stance-taking. While clicks can stand on their own, e.g. as an affective spontaneous reaction to an unexpected event, most clicks discussed in this paper were parts of larger constructions linked to specific discursive meanings comprising both verbal and non-verbal elements. Turn-initial clicks after other-selection were not attested in the ORD data investigated in this study, which stands in stark contrast to observations on clicks in other languages. The domain-based approach overall proved fruitful, as it revealed the interplay of clicks and domain breakpoints. Future research will hopefully shed more light on the cross-linguistic distribution and functional load of clicks in oral communication.

Acknowledgements Many people contributed to this research by giving helpful comments on previous versions of the paper, among them the audiences of Urban Voices: linguistic

On Clicks in Russian Everyday Communication

255

variation and communicative diversity (St. Petersburg, 2015) and 12. Deutscher Slavistentag (Gießen, 2015), as well as all members of the network „Urban Voices“. I am particularly grateful to Lenore Grenoble and to Richard Ogden for their valuable feedback and ideas.

References Becker-Mrotzek, Michael / Vogt, Rüdiger: Unterrichtskommunikation: Linguistische Analysemethoden und Forschungsergebnisse. Niemeyer: Tübingen 2009. Biber, Douglas / Conrad, Susan: Register, Genre, and Style. Cambridge University Press: Cambridge 2009. Bolden, Galina: “The discourse marker nu in Russian conversation”. In: Auer, Peter / Maschler, Yael (eds.): Nu/Nå: A family of discourse markers across the languages of Europe and beyond. DeGruyter: Berlin 2016, pp. 48–80. Couper-Kuhlen, Elizabeth: English speech rhythm : Form and function in everyday verbal interaction (Pragmatics & Beyond New Series, vol. 25). J. Benjamins: Amsterdam 1993. Drahota, Amy / Costall, Alan / Reddy, Vasudevi: “The vocal communication of different kinds of smile”. Speech Communication 50, 2008, pp. 278–287. Gil, David: “Para-linguistic usages of clicks”. In: Dryer, Matthew S. / Haspelmath, Martin (eds.): The World Atlas of Language Structures Online. Max Planck Institute for Evolutionary Anthropology: Leipzig 2013, retrieved 31.08.2017, from: http://wals.info/chapter/142. Gold, Erica / French, Peter / Harrison, Philip: “Clicking behavior as a possible speaker discriminant in English”. Journal of the International Phonetic Association 43(3), 2013, pp. 339–349. Graf, Elena: Interjektionen als interaktive Einheiten. Lang. IDS [Institut für Deutsche Sprache]: Frankfurt a.M. 2011. Datenbank für Gesprochenes Deutsch, 2017, retrieved 31.08.2017, from: http://dgd.ids-mannheim.de. Kendrick, Kobin H. / Torreira, Francisco: “The timing and construction of preference: a quantitative study”. Discourse Processes 52(4), 2015, pp. 255–289. Ladefoged, Peter / Maddieson, Ian. The sounds of the world’s languages. Oxford: Blackwell 1996. Mehan, Hugh: Learning lessons: Social organization in the classroom. UP: Harvard 1979. Miller, Amanda L: “Northern Khoesan”. In: Rainer Vossen (ed.): The Khoesan Languages. Routledge: Abingdon 2013, pp. 139–149. Ogden, Richard: “Making sense of outliers”. Phonetica 69, 2012, pp. 48–67.

256

Ludger Paschen

Ogden, Richard: “Clicks and percussives in English conversation”. Journal of the International Phonetic Association 43(3), 2013, pp. 299–320. Sacks, Harvey / Schegloff, Emanuel A. / Jefferson, Gail: “A simplest systematics for the organization of turn-taking for conversation”. Language 50(4), 1974, pp. 696–735. Sappok, Christian: „Russische regionale Varietäten und Dialekte – eine akustische Datenbank mit diskursiver Annotation“. Wiener Slawistischer Almanach 65, 2010, pp. 163–190. Sappok, Christian, this volume: “Addressing by voice: deixis and design” In: Thielemann, Nadine / Richter, Nicole (eds.): Urban Voices – The Sociolinguistics, Grammar and Pragmatics of Spoken Russian. Lang, Berlin et al. 2019, pp. 37–64. Sappok, Christian / Krasovitsky, Alexandr / Paschen, Ludger / Brabender, Katrin / Koch, Andreas / Kühl, Nadja: RuReg: Russische Regionen. Akustische Datenbank 2016. URL: www.rureg.de. Schaeffler, Sonja / Scobbie, James M. / Schaeffler, Felix: “Complex patterns in silent speech preparation: preparing for fast response might be different to preparing for fast speech in a reaction time experiment”. In: Proceedings of ICPhS 18 2015, Glasgow. Schegloff, Emanuel / Sacks, Harvey: “Opening up closings”. Semiotica 8, 1973, pp. 289–327. Schegloff, Emanuel: “Identification and Recognition in Telephone Conversation Openings”. In: Psathas, George (ed.): Everyday Language: Studies in Ethnomethodology. New York 1979, pp. 23–78. Selting, Margret et al.: „Gesprächsanalytisches Transkriptionssystem 2 (GAT 2)“. Gesprächsforschung 10, 2009, pp. 353–402. Simpson, Adrian: “Acoustic and auditory correlates of non-pulmonic sound production in German”. Journal of the International Phonetic Association, 37(02), 2007, pp. 173–182. Simpson, Adrian: “Ejectives in English and German – linguistic, sociophonetic, interactional, epiphenomenal?” In: Celata, Chiara / Calamai, Silvia (eds.): Advances in sociophonetics. Benjamins: Amsterdam 2014, pp. 187–202. Szczepek Reed, Beatrice: “Intonation phrases in natural conversation: A participants’ category?” In: Barth-Weingarten, Dagmar / Reber, Elisabeth / Selting, Margret (eds.): Prosody in interaction. Benjamins: Amsterdam 2010, pp. 191–212. Šerstinova, Tatjana: “The structure of the ORD Speech Corpus of Russian Everyday Communication”. In: Matoušek, Vaclav / Mautner, Pavel (eds.): TSD 2009. Berlin, Heidelberg: Springer-Verlag 2009, pp. 258–265. Thielemann, Nadine, this volume: “Constructional patterns emerging in talkin-interaction – Some preliminary observations on the grammar of govorit’

On Clicks in Russian Everyday Communication

257

in colloquial Russian” In: Thielemann, Nadine / Richter, Nicole (eds.): Urban Voices – The Sociolinguistics, Grammar and Pragmatics of Spoken Russian. Lang, Berlin et al. 2019, pp. 67–94. Trouvain, Jürgen: “Laughing, breathing, clicking – the prosody of nonverbal vocalisations”. In: Campbell, Nick / Gibbon, Dafydd / Hirst, Daniel (eds.): Proceedings of Speech Prosody 7, Dublin 2014, pp. 598–602. Trouvain, Jürgen: “On clicks in German”. In: Leemann, Adrian / Kolly, Marie-José / Schmid, Stephan / Dellwo, Volker (eds.): Trends in phonetics and phonology. Studies from German-speaking Europe. Peter Lang: Bern et al. 2015, pp. 21–33. Trouvain, Jürgen / Malisz, Zofia: “Inter-speech clicks in an Interspeech keynote”. In: Proceedings of Interspeech 2016, San Francisco 2016, pp. 1397–1401. Wright, Melissa: Studies of the phonetics-interaction interface: Clicks and interactional structures in English conversations. (PhD thesis) 2005. Wright, Melissa: “Clicks as markers of new sequences in English conversation”. In: Proceedings of the ICPhS XVI, Saarbrücken 2007, pp. 1069–1072. Wright, Melissa: “On clicks in English talk-in-interaction”. Journal of the International Phonetic Association 41(2), 2011a, pp. 207–229. Wright, Melissa: “The phonetics-interaction interface in the initiation of closings in everyday English telephone calls”. Journal of Pragmatics 43(4), 2011b, pp. 1080–1099. Yokoyama, Olga T.: “Oppozicija svoj-čužoj v russkom jazyke”. In: Maguire, Robert A. / Timberlake, Alan: American Contributions to the Eleventh International Congress of Slavists, Bratislava 1993, pp. 452–459.

Potsdam Linguistic Investigations Potsdamer Linguistische Untersuchungen Recherches Linguistique à Potsdam Edited by / Herausgegeben von / Edité par Peter Kosta, Gerda Haßler, Teodora Radeva-Bork, Lilia Schürcks, Nadine Thielemann and / und / et Vladislava Maria Warditz Band 1

Peter Kosta / Lilia Schürcks (eds.): Linguistics Investigations into Formal Description of Slavic Languages. Contributions of the Sixth European Conference held at Potsdam University, November 30–December 02, 2005. 2007.

Band 2

Lilia Schürcks: Binding and Discourse. Where Syntax and Pragmatics Meet. 2008.

Band 3

Christiane Hümmer: Synonymie bei phraseologischen Einheiten. Eine korpusbasierte Untersuchung. 2009.

Band 4

Svetlana Friedrich: Definitheit im Russischen. 2009.

Band 5

Matthias Guttke: Strategien der Persuasion in der schriftkonstituierten politischen Kommunikation. Dargestellt an Parteiprogrammen der Neuen Rechten in Polen. 2010.

Band 6

Peter Kosta / Lilia Schürcks (eds.): Formalization of Grammar in Slavic Languages. Contributions of the Eighth International Conference on Formal Description of Slavic Languages – FDSL VIII 2009. University of Potsdam, December 2–5, 2009. 2011.

Band 7

Roman Sukač (ed.): From Present to Past and Back. Papers on Baltic and Slavic Accentology. 2011.

Band 8

Diego Gabriel Krivochen: The Syntax and Semantics of Nominal Construction. A Radically Minimalist Perspective. 2012.

Band 9

Teodora Radeva-Bork: Single and Double Clitics in Adult and Child Grammar. 2012.

Band 10

Anja Hennemann: A Context-sensitive and Functional Approach to Evidentiality in Spanish or Why Evidentiality needs a Superordinate Category. 2013.

Band 11

Diego Gabriel Krivochen / Peter Kosta: Eliminating Empty Categories. A Radically Minimalist View on Their Ontology and Justification. 2013.

Band 12

Christina Behme: Evaluating Cartesian Linguistics. From Historical Antecedents to Computational Modeling. 2014.

Band 13

Kathleen Plötner: Raum und Zeit im Kontext der Metapher. Korpuslinguistische Studien zu französischen und spanischen Raum-Zeit-Lexemen und Raum-Zeit-Lokutionen. 2014.

Band 14

Marion Eva Ernst: Produktnamen der Lebensmittelindustrie. Eine empirisch-strukturelle Untersuchung. 2014.

Band 15

Stefanie Wagner: Eine „unbekannte“ Sprache lesen oder Von der Entdeckung des Nissart durch Interkomprehension. 2015.

Band 16

Nataša Todorović: The Indicative and Subjunctive da-complements in Serbian: A SyntacticSemantic Approach. 2015.

Band 17

Vladislava Warditz / Beatrix Kreß (eds.): Multilingualism and Translation. Studies on Slavonic and Non-Slavonic Languages in Contact. 2015.

Band 18

Nadia Varley: Optionality and overgeneralization patterns in second language acquisition: Where has the expletive ensconced itself? 2015.

Band 19

Verónica Böhm: La imperfectividad en la prensa española y su relación con las categorías semánticas de modalidad y evidencialidad. 2016.

Band 20

Roland Wagner: Reflexivität im tschechisch-deutschen Sprachvergleich. Möglichkeiten und Grenzen einer Prognose. 2016.

Band 21

Ray C. H. Leung: Institutional Construction of Gamblers’ Identities. A Critical Multi-method Discourse Study. 2017.

Band 22

Markéta Ziková: Licensing of Vowel Length in Czech. The Syntax-Phonology Interface. 2018.

Band 23

Ludmila Veselovská: Noun Phrases in Czech. Their Structure and Agreements. 2018.

Band 24

Carmen Conti Jiménez: Complejidad lingüística. Orígenes y revisión crítica del concepto de lengua compleja. 2018.

Band 25

Nadine Thielemann / Nicole Richter (eds.): Urban Voices: The Sociolinguistics, Grammar and Pragmatics of Spoken Russian. 2019.

www.peterlang.com

Urban Voices: The Sociolinguistics, Grammar and Pragmatics of Spoken Russian The volume Urban Voices presents studies, analysing how speakers of Russian convey social meanings across a variety of speech situations. Rooted in quantitative and qualitative methodological frameworks, the contributions show how various linguistic, paralinguistic and pragmatic means relate to sociolinguistic dimensions (e.g. display the social, ethnic, local identity of a speaker, the institutional character of a communicative situation). The analyses are the results of the research network Urban Voices, which focuses on the sociolinguistics, grammar and pragmatics of spoken Russian and investigates its linguistic and communicative diversity. The contributors are Olga Blinova, Nina Bodganova-Beglarian, Beatrix Kreß, Lenore Grenoble, Peter Kosta, Elena Markasova, Ludger Paschen, Nicole Richter, Christian Sappok, Tatiana Sherstinova and Nadine Thielemann.

The Editors

Nadine Thielemann / Nicole Richter (eds.)

Nadine Thielemann / Nicole Richter (eds.)

25

Urban Voices

Potsdam Linguistic Investigations Potsdamer Linguistische Untersuchungen Recherches Linguistiques à Potsdam 25

Potsdam Linguistic Investigations Potsdamer Linguistische Untersuchungen Recherches Linguistiques à Potsdam 25

Nadine Thielemann / Nicole Richter (eds.)

Urban Voices: The Sociolinguistics, Grammar and Pragmatics of Spoken Russian

Nadine Thielemann is a professor of Slavic linguistics at Vienna University of Economics and Business. She is interested in the linguistics and pragmatics of talk-in-interaction, political discourse and multilingualism at the workplace. Nicole Richter is a professor of Multicultural Communication (Slavonic and English Linguistics) at the European University Viadrina, Frankfurt (Oder) and at Collegium Polonicum, mainly interested in contrastive phonetics and prosody.

PLI_025 266462_Thielemann_JA_A5HC 151x214 fusion.indd 1

10.12.18 17:21