246 3 128MB
English Pages [450] Year 2021
OKAY across Languages Toward a comparative approach to its use in talk-in-interaction Studies in Language and Social Interaction
Edited by Emma Betz, Arnulf Deppermann, Lorenza Mondada and Marja-Leena Sorjonen
34
John Benjamins Publishing Company
OKAY across Languages
Studies in Language and Social Interaction (SLSI) issn 1879-3983
Studies in Language and Social Interaction is a series which continues the tradition of Studies in Discourse and Grammar, but with a new focus. It aims to provide a forum for research on grammar, understood broadly, in its natural home environment, spoken interaction. The assumption underlying the series is that the study of language as it is actually used in social interaction provides the foundation for understanding how the patterns and regularities we think of as grammar emerge from everyday communicative needs. The editors welcome language-related research from a range of different methodological traditions, including conversation analysis, interactional linguistics, and discourse-functional linguistics. For an overview of all books published in this series, please see benjamins.com/catalog/slsi
Editors Sandra A. Thompson
University of California, Santa Barbara, USA
Elizabeth Couper-Kuhlen University of Helsinki, Finland
Editorial Board Peter Auer
Barbara A. Fox
Galina Bolden
Makoto Hayashi
Arnulf Deppermann
Marja-Liisa Helasvuo
Paul Drew
K.K. Luke
University of Freiburg, Germany Rutgers University, USA Leibniz-Institut für Deutsche Sprache, Germany University of York, UK
University of Colorado, USA Nagoya University, Japan University of Turku, Finland Nanyang Technological University, Singapore
Volume 34 OKAY across Languages Toward a comparative approach to its use in talk-in-interaction Edited by Emma Betz, Arnulf Deppermann, Lorenza Mondada and Marja-Leena Sorjonen
OKAY across Languages Toward a comparative approach to its use in talk-in-interaction Edited by
Emma Betz University of Waterloo
Arnulf Deppermann Leibniz-Institut für Deutsche Sprache, Mannheim
Lorenza Mondada University of Basel
Marja-Leena Sorjonen University of Helsinki
John Benjamins Publishing Company Amsterdam / Philadelphia
8
TM
The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.
doi 10.1075/slsi.34 Cataloging-in-Publication Data available from Library of Congress: lccn 2020048726 (print) / 2020048727 (e-book) isbn 978 90 272 0815 6 (Hb) isbn 978 90 272 6028 4 (e-book)
© 2021 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Company · https://benjamins.com
Table of contents
Acknowledgements Chapter 1 Introduction: OKAY emerging as a cross-linguistic object of study in prior research Emma Betz and Marja-Leena Sorjonen Chapter 2 Data and methods used in the study of OKAY across languages Arnulf Deppermann and Lorenza Mondada
vii
1
29
Generic sequential uses of OKAY across languages Chapter 3 OKAY in responding and claiming understanding Emma Betz and Arnulf Deppermann
55
ith contributions from Elizabeth Couper-Kuhlen, Elwys De Stefani, Katariina W Harjunpää, Kaoru Hayano, Henrike Helmer, Leelo Keevallik, Mary Shin Kim, Stephanie Hyeri Kim, Aino Koivisto, Satomi Kuroshima, Seung-Hee Lee, Xiaoting Li, Anna Lindström, Lorenza Mondada, Ana Cristina Ostermann, Søren Sandager Sørensen, Marja-Leena Sorjonen, Jakob Steensig, and Matylda Weidner
Chapter 4 OKAY in closings and transitions Lorenza Mondada and Marja-Leena Sorjonen ith contributions from Emma Betz, Elizabeth Couper-Kuhlen, Elwys De Stefani, W Arnulf Deppermann, Katariina Harjunpää, Kaoru Hayano, Henrike Helmer, Leelo Keevallik, Mary Shin Kim, Stephanie Hyeri Kim, Aino Koivisto, Satomi Kuroshima, Seung-Hee Lee, Xiaoting Li, Anna Lindström, Ana Cristina Ostermann, Søren Sandager Sørensen, Jakob Steensig, and Matylda Weidner
93
vi
OKAY across Languages
OKAY in specific languages Chapter 5 The prosody and phonetics of OKAY in American English Elizabeth Couper-Kuhlen
131
Chapter 6 Rising OKAY in third position in Danish talk-in-interaction Søren Sandager Sørensen and Jakob Steensig
175
Chapter 7 OKAY as a response to informings in Finnish Aino Koivisto and Marja-Leena Sorjonen
205
Chapter 8 When OKAY is repeated: Closing the talk so far in Korean and Japanese conversations Satomi Kuroshima, Stephanie Hyeri Kim, Kaoru Hayano, Mary Shin Kim and Seung-Hee Lee
235
OKAY in specific activities and settings Chapter 9 OKAY in health helpline calls in Brazil: Managing alignment and progressivity Ana Cristina Ostermann and Katariina Harjunpää Chapter 10 A resource for action transition: OKAY and its embodied and material habitat Elwys De Stefani and Lorenza Mondada Chapter 11 OKAY projecting embodied compliance to directives Leelo Keevallik and Matylda Weidner Chapter 12 Coordination of OKAY, nods, and gaze in claiming understanding and closing topics Henrike Helmer, Emma Betz and Arnulf Deppermann
269
301
337
363
Appendix: Transcription conventions and glossing symbols
395
Bibliography
401
Name index
433
Subject index
437
Acknowledgements
This book took shape during two inspiring workshops held in Mannheim and Helsinki in 2017. We are grateful to the Leibniz-Institut für Deutsche Sprache in Mannheim and the University of Helsinki’s Centre of Excellence in Research on Intersubjectivity in Action for generously funding these workshops, and we thank the researchers who accepted our invitation and quickly shared in our enthusiasm for okay. Each contribution has been read and commented upon by several other authors, and we thank everybody for their extraordinarily collegial efforts in thus shaping the whole volume. Sandra Thompson carefully read and evaluated the project at different stages, and we thank her for her insight and support. We are most grateful to Kerttu Pihlajamaa for her efficient work and attention to detail in formatting chapters, generating the bibliography, and assisting with indexes. Our heartfelt thanks also goes to Isja Conen and Patricia Leplae, who guided the volume patiently and competently through the publication process at Benjamins. The research reported in this book rests on all of our participants’ availability and willingness to be recorded and scrutinized. Above all, we are indebted to them.
Basel, Helsinki, Mannheim, and Waterloo, September 2020
Chapter 1
Introduction OKAY emerging as a cross-linguistic object of study in prior research Emma Betz and Marja-Leena Sorjonen University of Waterloo / University of Helsinki
This chapter sets the context for the articles in the volume – explorations in the use of OKAY in a diverse set of languages, including American English, Brazilian Portuguese, Danish, Estonian, Finnish, French, German, Italian, Japanese, Korean, Mandarin, Polish, and Swedish. We first outline the origins of OKAY in American English and its spread to other languages as a loanword, motivating this study of OKAY. We then review the state of the art in research on OKAY in spoken interaction in a variety of settings. Since this volume makes a case for investigating OKAY empirically as it is actually used in particular occasions of spoken and embodied interaction, the review of existing work on OKAY will be connected to relevant developments in Conversation Analysis (Sidnell 2010; Clift 2016) and Interactional Linguistics (Couper-Kuhlen and Selting 2018). We close with a discussion of overarching themes and promising new research directions that emerge from existing work.
1. Origin and early spread of OKAY1 O.K.2 has been called a “spectacular expression” (Read 1963b, 83) and “America’s greatest invention” (Metcalf 2010, 26). This section provides a summary of what we know about its origins and spread. The origin of the word can be traced back to newspaper language use in Boston in 1830s, more specifically to page 2 of the 1. We use all-capitals spelling (OKAY) in general references to the particle. When referring to the use of the particle in a specific language or quoting from a particular study, we display the particle in italics (okay, okei, occhei, etc.). 2. The spelling as okay is a later development, so we use “O.K.” (as well as the spelling variants O. K., o.k., o. k., OK, and O K, depending on the quoted source) in describing earlier stages in the word’s development. Spellings with and without periods existed from the beginning, although O.K. was more common. https://doi.org/10.1075/slsi.34.01bet © 2021 John Benjamins Publishing Company
2
Emma Betz and Marja-Leena Sorjonen
Boston Morning Post of March 23, 1839 (Read 1963a, 1963b, 1964a, 1964b, 1964c; Metcalf 2010, 30). This first use was an instance of adjectival “o. k.” within a larger report by editor Charles Gordon Greene, and it included an inserted unpacking immediately following the abbreviation: “all correct”. The second use of “o.k.” in the same newspaper (“They were o.k.”), three days later, contained no such unpacking for readers.3 In the same year, it can be shown to have spread to other newspapers in Philadelphia, New York, and Baltimore (Metcalf 2010, 36). This is now considered the most credible origin of today’s OKAY. Read (1964b, 1964c) and Metcalf (2010, 80–95) also offer detailed accounts of some of myths of origin.4 What made the birth of O.K. in the 1830’s possible was in no small part its shape. It came into existence at a time when the use of witty abbreviations as joking devices had become popular and widespread (Read 1963a, 5).5 This coincided with a fondness for misspellings, especially of the use of the unusual and thus conspicuous letter K for C in English (Read 1963a; Metcalf 2010, 9).6 O.K.’s survival and spread during the 19th century was assisted by different historical events, including the presidential re-election campaign of Martin van Buren in 1840 and a political hoax connected to Buren’s predecessor Andrew Jackson and his purported proneness to misspellings (Read 1963b, 9; Metcalf 2010, 87–88). Practical, non-joking uses of O.K. moved to the foreground with the advent of regular telegraph transmissions in 1844 (facilitating the emergence of routine uses of O.K. for closing in this environment, see Section 2), and the word slowly entered as an affirmative into everyday use. O.K. soon spread beyond its U.S. American borders. Early popularity outside of the U.S. is credited to the British music hall performer Alfred Vance, who used O.K.
3. Using the abbreviation without an explanation so soon after its first mention suggests that it may have already existed in spoken interaction. Green’s use is the first documented use of O.K. in written language; we have no information regarding the particle’s use in spoken interaction before 1839. 4. These include an origin story about OKAY coming from the Native American Choctaw language, specifically from the word okéh, said to serve as a marker for assertions. This has been shown to be a false origin, but it explains the source of alternative spellings of OKAY as “okeh” (Metcalf 2010, 86–87). Greco and Degges (1975) point to Scottish och/ough aye ‘oh yes’ as an alternative pre-1830 origin. 5. O.K. was only one of many creative spellings at the time. Many other abbreviations that emerged in the 1830 and 1840 in the same context, such as o.w. and OW for all right, k.y. for no use, or n.s.m.g. for enough said among gentlemen (Metcalf 2010, 26), were, however, short-lived phenomena. 6. Since using the letter K makes an English word stand out and catch attention in writing, this device was particularly well suited for advertisement, as can still be seen today in product names such as Kodak or Kleenex.
Chapter 1. Introduction
around 1870 as a predicative adjective (“the O, K, thing to do”) in the song “Walking in the Zoo” (Metcalf 2010, 173–176). The connection to the U.S. is reflected in an entry in a British dictionary,7 in which O.K. is described as an “Americanism” and defined as describing something that is “perfectly in order” (Hotten 1874, 241). Different reasons have been suggested for the international spread of O.K. to, first, other parts of the English-speaking world and then beyond its boundaries. One is connected to what has been described as simplicity and possible universality of phonetic shape and a resulting ease in finding “native expressions in other languages that sound like OK” (Metcalf 2010, 79, 90–92; cf. Wait 1941). A second reason is connected to etymological origin stories: As O.K. spread widely, speakers of other languages began developing accounts that reconstructed O.K. as deriving from expressions in their own language,8 thus contributing to acceptance and integration of the word into new languages. At the same time, O.K. seems to have retained its connection to English/to its English origins into the present time, and attitudes toward the word show this.9 As the global influence of English has steadily increased, O.K. has spread to a vast number of the world’s languages,10 becoming the possibly most successful linguistic export from American English (Kessler 2010, 161). Our goal in this volume is to document the range of current uses of the
7. Hotten (1874), in the entry on “O. K.” on p. 241. See also an earlier entry in Hotten (1864, 191), which, however, does not include information on the word’s history: “O. K., a matter to be O. K. (OLL korrect, i.e., all correct,) must be on the “square,” and all things done in order.” 8. Examples include Aux Cayes, the French name of a port in Haiti; German ohne Korrektur ‘without correction’ or ‘no correction needed’, said to have been used on documents in the late nineteenth century; Finnish oikea for ‘correct’; Greek olla kalla ‘all good’ or ‘all right’ (Metcalf 2010, 90–92). See also Cassidy (1981), who refutes an African origin. 9. See for example a 2016 survey, which documents a critical position toward the use of loan words from English (such as Ok) in German: https://yougov.de/news/2016/08/22/auch-we r-englisch-kann-kritisiert-oft-menge-der-an/. See also the deconstruction and reframing of the traditional distinction between ‘necessary’ and ‘luxury loans’, including okay, by Onysko and Winter-Froemel (2011). Attitudes toward OKAY can also shed light on what is perceived as the particle’s core meaning: Beach (1995b, 266, with reference to Verby 1991) reports cases of medical trainers’ treating OKAY use by medical trainees as inappropriate, because it may carry connotations such as ‘good’ or ‘all right’, thus being hearable as “‘accepting or approving’ harmful behaviors” such as smoking (see Kovarsky 1989 for a similar discussion). 10. The number of entries on Wikipedia underscores this point: Entries exist in 50 languages, from Armenian to Vietnamese. This does not necessarily mean that the word is used in 50 languages, but it suggests that OKAY holds some interest and is recognizable to speakers of these languages. This is in addition to entries connecting the meaning of certain hand gestures (a circular thumb-index finger gesture or a thumbs up) to OK/O.K., as indicating ‘good’ or marking approval or agreement.
3
4
Emma Betz and Marja-Leena Sorjonen
particle that has resulted from this spread, but before surveying empirical research on its current use in interaction across languages, we ask what can be gleaned from sources on the history of O.K. about its early uses in English. 2. What can we glean from sources on the history of OK about its early uses? Much of the early work on O.K. concentrated on unraveling its origin and routes of spread. In that literature, certain types of uses are explicitly mentioned, or they can be inferred from the text. O.K. began as an abbreviation, or more precisely an initialism (Metcalf 2010, 12–13), and this is reflected in its common spelling with periods. With changes in how O.K. was used and perceived went along a change in how it could be spelled: O.K. became a word, and okay and kay appeared (cf. Metcalf 2010, 147–148).11 The likely first use of O.K. in 1839 was as an adjective, and work on the history of O.K. suggests that this is its most common use throughout its documented history. The following instance from 1866, used in a description of the transatlantic cable for telegraphing, provides a use of O.K. in a coordinate adjective phrase: “Simply through its entire length it was perfect, or, as it is technically called, O.K. (all correct)” (quoted in Metcalf 2010, 101). The O.K. is described as a technical expression in this context, and its meaning is unpacked with the paraphrase “all correct”. This unpacking, also delivered with its earliest documented use in 1839, connects O.K. and ALL RIGHT as overlapping in meaning, as Metcalf (2010) notes: Back in Boston in the late 1830s, the misspelling OW for all right was especially important in paving the way for a smooth launch and reception of OK. It’s not just that they both begin with A misspelled as O. They also have practically the same meaning. (Metcalf 2010, 39)
This is a common view still today. Not only is “all right”/“alright” commonly the, or part of the, definition of the adjectival use of OKAY in English dictionaries,12 it is also sometimes treated as an alternative or functional equivalent to OKAY in interaction research (see Beach 1993, 336, on pre-closings; Merritt 1978; Thompson et al. 2015, 224, 226, on responses to requests; Schegloff 2007, 120). However there is also work tracing differences between them in certain sequential contexts and 11. The first known spelling as “okay” is in 1869, in Louisa May Alcott’s Little Women; it is the more modern spelling, considered more refined, but it was still deleted in a revised edition 1880. Okay remained a marker of colloquial language and was more regularly spelled as okay only in/ after the 1920s (cf. Metcalf 2010, 126). 12. See, e.g, https://www.merriam-webster.com/dictionary/OK.
Chapter 1. Introduction
interaction types (Bangerter and Clark 2003; Filipi and Wales 2003; Schleef 2005, 2008; K. Turner 1999). Some of this work is discussed in Sections 3 and 4 below. In addition to its use an adjective, there is some evidence of other contexts of use early in O.K.’s emergence: Metcalf (2010, 99) quotes an example from a wire transfer, in which O.K. is used after a suggestion, understood as accepting this suggestion, and followed by a formulation of the agreed arrangement. He also shows an example of O.K. as acknowledging an answer to a request for information (ibid., 100), a use which in CA terminology would be described as occurring in sequence-closing third position.13 A closing function of O.K. is also visible in an 1876 beginner’s manual on telegraphing, specifically on marking the end of an exchange: An acknowledgment of the receipt of any kind of communication is made by returning O K, followed by the call of the office receiving the communication. … and no message is ever regarded as transmitted until the office receiving it gives O K, or commences to send back other dispatches. (Metcalf 2010, 100–101)
This example is particularly interesting from an interactional perspective, because it shows how O.K. was used in a routine way, that is, as a dedicated token in a specific sequential environment in telegraphic transmission. Work tracing the history of O.K. suggests an increase in variation of its form in English (oakie-doke, okey-dokey, okay, okey),14 along with a widening of the range of uses in the 20th century. This includes O.K. as agreeing and accepting response after a wider range of first actions: granting a request; giving a go-ahead, permission, or acquiescing response after an offer in early human-computer interaction. What also emerges in the written documents available are, however, transitional uses in literary dialogue and tag-like uses (cf. Metcalf 2010, 151–164). An example of the latter is the purported use of O.K. in gang language for territory marking in the UK, where the earliest use of phrases such as “(gang name) Rules, Ok?” may have been in the 1930’s in Glasgow.15
13. The Oxford English Dictionary entry contains a collection of examples of different uses documented in writing. See https://www.oed.com/viewdictionaryentry/Entry/130925. 14. For variation in other languages, see for example by Koivisto and Sorjonen (this volume) on Finnish and Sørensen and Steensig (this volume) on Danish. 15. Urbandictionary.com, accessed April 4, 2020, entry “Rules, OK?”: “A slang phrase appended to a word (usually a noun), to form a rhetorical question, which denotes superiority (ie. so-and-so Rules, OK?) It sometimes appears without a comma or question mark, but always appears with the informal ‘ok’ as opposed to ‘okay’. The phrase’s first recorded use was in 1975, but it is rumored to have originated as early as the 1930’s among the Glasgow “Razor Gangs”. Rival gangs were known to tag each other’s turf with “(gang name) Rules, Ok?” during disputes over territory as a part of gang warfare.”
5
6
Emma Betz and Marja-Leena Sorjonen
Research on the uses of OKAY beyond English and over the range of its current host languages has remained limited. OKAY, however, is not just an Anglicism which, like many others, has been borrowed into other languages. In addition to being phonetically integrated in its host language (e.g., Kubozono 2006 for Japanese; Martin Havlík, pers. comm., for Czech), OKAY is particularly interesting because of its high frequency in various languages (see, e.g., Sharp 2001, 81, for Swedish) and because of its emerging differentiation in interactional use – uses that may or may not overlap with what we know about American and British English (see, e.g., Adegbija and Bello 2001 on differing uses in Nigerian English; House 2013 on uses in English as lingua franca interactions in Germany; Bangerter and Clark 2003 on uses in similar contexts in American English and German, discussed in Section 4). We will offer an overview of existing work on the use of OKAY in interaction in different languages in the next two sections, broadly divided into research on everyday interaction and work on institutional discourse. This division will make visible some of the differences in research focus (and thus functions attributed to OKAY) emerging from investigating varied interaction contexts. 3. The emergence of OKAY as a research topic in conversation analysis This section presents some of the early research on OKAY in everyday interaction. We trace studies that started to develop a more nuanced understanding of what OKAY as a verbal element was used for. These studies were published from the late 1970’s onward, mostly in Anglo-American journals. We lay out a process through which some central meaning axes of OKAY started to emerge in different subject areas and research approaches, and how in the course of the process conversation analytic research, and particularly research on responding in interaction, started to play an increasingly important role. We then move to more recent studies on OKAY in everyday interactions. Research that focused on the usages of OKAY was preceded by studies that had other, larger phenomena as their main interest and where OKAY was shown to implement one important part of that phenomenon. Schegloff and Sacks discussed English okay as part of analyzing telephone conversation closings. They showed that entering the closing often included a pre-closing sequence, which was implemented with reciprocal okays forming turns of their own. An important argument presented by Schegloff and Sacks (1973) was that it is not the use of okays as such that displays that a pre-closing is on its way, but their sequential placement in a closing-relevant environment (see also Robinson 2013). Furthermore, they show that okays in closing-relevant environment do not need to lead to a closing but can be followed by a new phase in the conversation. Around the same time, Goffman
Chapter 1. Introduction
(1974, [1976] 1981), in discussing framings people use in making sense of experiences, situations and activities, mentioned okay as one of the conventional bracket markers (or boundary markers) that are used for separating different frames and for moving between them. Sinclair and Coulthard (1975) described okay as a boundary marker in their discourse analytic classroom interaction study, indicating that a task episode had ended or was about to start. Around the time when the first studies on OKAY appeared, conversation analytic research on response particles in English paved the way for a rigorous analysis of OKAY. We will briefly outline the methodological considerations developed in this work and some of its findings here. This early research highlights the importance of sequential context for understanding the interactional relevance of different responses, laying the groundwork for a positionally sensitive analysis of uses of OKAY.16 Several studies discussed usages that, in one way or another, dealt with the sequencing of actions and topical movement in interaction. Jefferson analyzed differences between different particles used in similar sequential slots (e.g., yeah and yes vs. mm hm and uh huh, Jefferson 1981b, [1983b] 1984, [1983a] 1993). Schegloff (1982) argued against assigning form-based functions to responses without considering their sequential context (e.g. signaling attention, interest or understanding, as in studies on ‘back-channels’, e.g., Yngve 1970). He argued that a relevant generic function of responses such as uh huh is crucially related to the sequential development of talk: uh huh is taken as a display that the primary speaker has an extended unit of talk in progress and should continue it. Thus, it is used, and oriented to by recipients, as a ‘continuer’.17 C. Goodwin (1986) subsequently showed a difference between continuers and assessments in the midst of extended talk. Other studies focused on displays of affective or epistemic stance to the specifics of the prior turn (e.g., Jefferson 1980, 1988; Heritage 1984a, 1985). 16. An early statement of key methdological issues in studying responses, including response particles, appeared in an influential volume edited by Atkinson and Heritage (1984, 297–298), in the introduction to its section on responses (particle oh, Heritage 1984a; organization of laughter during troubles talk, Jefferson 1984c; applause as audience response to public speaking, Atkinson 1984). The introduction argued for the potential analytic interest of responses, and thus for its potential as a conversation analytic research object. It also voiced a methodological caveat: it pointed to the dangers of presuming a priori that responses that look or sound the same whenever used will have the same interactional implications in all uses. The introduction can be seen as addressing researchers and research at the time both in conversation analysis and more generally. 17. Gardner (2001, 26), in discussing continuers, mentions that as uh huh and mm hm are used to pass up an opportunity to speak, and that they typically form a turn of their own. They are only rarely found with further talk in the turn, for example, in a brief utterances such as uh huh, okay. Gardner can be heard as implying that the turn with okay would achieve something else than merely pass up the opportunity to speak.
7
8
Emma Betz and Marja-Leena Sorjonen
In this methodological development of research on responses, and especially on response particles, more elaborated discussions on okay in English began to appear as well. Davidson (1984) drew attention to okay as a rejection finalizer, an element displaying that the speaker is going along with a rejection of an invitation or offer and will not be producing subsequent versions (cf. Davidson 1990 for a more extensive discussion). The use of okay in the combination oh okay in responding to a final successful repair in extended repair sequences was noted by Heritage (1984a). The more extended conversation analytic work focusing on OKAY was started in the early 1990’s. 1993 saw the appearance of a highly influential, now classical, study on American English okay by Wayne Beach. Beach called for the analysis of okay beyond its use as a free-standing and non-continuative particle (as when it is used in pre-closings). The paper lays out an argument for the transitory character of okay. Okay is treated as a pivotal (Beach 1995a, 127), “dual-character” element (Beach 1993, 339): It is both – and simultaneously – responsive to the prior talk and fundamentally projective, making expectable a move to some relevant next matters in the ongoing sequence. The responsive, backward-looking character of okay is prominent when okay occupies a second position in a sequence (providing, e.g., confirmation or agreement) or a third (e.g., going along with the rejection of an invitation), and when it forms all that its speaker produces in a turn. In other contexts in which okay forms a turn of its own, such as telephone call openings and pre-closing contexts, the transitional character of okay begins to be seen. Thus in call openings, okay projects a move to a reciprocal “how are you,” and in closings forms a step to a subsequent closure. The pivotal, transitional character of okay is seen in a compact form especially in cases when okay is located in a turn-initial position, where it responds to what was projected by the prior turn (e.g., a display of agreement), while projecting a move to the next matter made relevant by the ongoing sequence. Beach states that in a large number of cases, okay is designed to be neither backward- nor forward-looking; rather, these functions are intertwined and intimately tied to ongoing activities. The dual usage is unambiguous, the okay displaying a ‘state-of-readiness’ for moving on to the next relevant issues. Neither is okay to be treated as competitive or disruptive. Evidence for this lies in the fact that okay is not prepared (‘set up’) by other particles, as is the case for the particles mm hmm + uh huh + yeah, which form a movement from ‘passive recipiency’ (mm hmm, uh huh) to speaker readiness (yeah, Jefferson 1981b). In general, an attempt to move to speakership does not explain the kind of movement that okay is working toward in everyday interactions, that of next relevant tasks in the ongoing interaction. However, in institutional interactions, okay appears to be used by the persons responsible for managing the interaction as a means for moving to speakership and for topic and activity bounding (ibid., fn. 8; see also Barske 2006, 2009). With his article, Beach
Chapter 1. Introduction 9
laid a strong basis for subsequent research, which is shown especially in a wide range of research on the transitional character of okay in institutional interactions (see Section 4). Table 1, which is organized chronologically, provides an overview of studies that focus on, or include discussion of, OKAY in everyday interactions. Table 1. Studies on the use of OKAY in every-day interaction* Setting/interaction type
Language(s) investigated
Authors
Telephone calls, group therapy
American English
Schegloff and Sacks (1973)
Telephone calls
American English
Davidson (1984, 1990)
Telephone calls, face-to-face
American English, some British English
Beach (1993, 1995a)
Face-to-face, telephone (from the British English London-Lund Corpus 1953–1985) Spanish (Madrid) and Face-to-face adolescent talk (corpora British English COLAM 2002, COLT 1992) (London)
Stenström (1994) Stenström (2014)
Children in peer and family interaction, face-to-face (1989 corpus)
American English
Pak, Sprott, and Escalera (1996); Kyratzis and Ervin-Tripp (1999)
Telephone calls (closings)
Greek, German
Pavlidou (1998)
Telephone calls
Australian English
K. Turner (1999)
Examples from interactions picked up; newspapers
Nigerian English
Adegbija and Bello (2001)
Face-to-face, some telephone calls
Australian English; some American and British English
Gardner (2001)
Casual conversations between Israeli Hebrew friends, family (1994–1997, audio)
Maschler (2002)
Face-to-face (from the London-Lund Corpus 1961–1976)
British English
Bangerter and Clark (2003)
Mainly telephone calls
Danish
Knudsen (2003, 2015) Mortensen and Mortensen (2009)
Social events (video/audio), research interviews (audio), radio programs (1980’s–2000’s)
Swedish Estonian and Estonian in Estonia
Keevallik (2006, 2013)
Telephone calls
American English
Schegloff (2007)
Telephone calls (closings)
German
Harren and Raitaniemi (2008)
Face-to-face talk, in-car interactions
English
Rauniomaa and Keisanen (2012)
(continued)
10
Emma Betz and Marja-Leena Sorjonen
Table 1. (continued) Setting/interaction type
Language(s) investigated
Authors
Private conversations/dialogue (telephone and face-to-face)
South African English
Huddlestone and Fairhurst (2013)
Face-to-face, student union kitchen (2015)
Finnish
Pekkanen (2017)
Telephone, face-to-face conversations American and British English
Stivers (2018)
Face-to-face, multi-party (3 encounters, 3.5 hrs total)
Oloff (2019)
German
Telephone, face-to-face conversations Syrian Arabic (recorded in the UK)
Martini (2020)
Ordinary conversations
Beach (2020)
American English
Telephone, face-to-face conversations American, Canadian, and Singaporean English
DeSouza et al. (2021)
* Tables 1–4 provide an overview of studies that we were able to find and read. They are most likely not exhaustive.
Table 1 shows that after the paper by Schegloff and Sacks (1973), and after the paper by Beach (1993), some work was done in the 1990’s that included analysis of OKAY. From the beginning of the 2000’s, research on OKAY in everyday interactions increased, revisiting usages discussed in earlier research and focusing on usages that had remained unexplored. Transitional okays in different Englishes are discussed by Gardner (2001, 53–57; also 2007). He describes such usages as ‘change-of-activity’, classifying okay, as well as alright and right, in these usages as ‘change-of-activity’ tokens. One issue he takes up are earlier treatments of English okay and alright as similar in function. He presents evidence for an earlier suggestion by Beach (1995a) that okay operates on a more local level than alright by referring to a study by K. Turner (1999) on telephone calls in Australian English. Turner found that a majority of instances of okay occurred at local, on-topic shifts, whereas a majority of alrights were found at major topic boundaries. However, both okay and alright occurred at an ‘intermediate’ topic or activity shift level. The difference in scope of the response came out here, too: Some of the cases had to do with returning to a topical thread after a digression, and in these contexts, okay was used to return from minor digressions, whereas alright was associated with longer digressions. Prior research has also described several interactional functions of OKAY in responsive position, highlighting its essentially backward-looking character (cf. Schegloff 2007, 120–128): In responses to how-questions (e.g., ‘How are
Chapter 1. Introduction 11
you?’), English okay is used adjectivally, and its meaning amounts to “allright” (see Couper-Kuhlen, this volume; cf. Pillet-Shore 2003). When responding to deontic initiating actions (e.g., directives, offers, proposals, requests, or suggestions), OKAY signals compliance, willingness, agreement and/or acceptance (e.g., Thompson et al. 2015, 227). On the other hand, it can be used to accept an offer, especially after negotiations, where it can be heard as “giving in to the other’s agenda” (Keevallik 2017, 285). In investigating responses to requests for object transfers or services in ordinary face-to-face interactions in English, Rauniomaa and Keisanen (2012) show that okay projects or promises compliance before fulfilling the request, for example when the recipient is engaged in doing something else and cannot fulfill the request immediately. Stivers (2018), studying responses to information questions that serve as proposals or pre-requests in English, classifies okay, alright, sure, and fine as ‘acquiescent’ interjections. She states that a response form of this type carries a semantic layer of acceptance, foregrounding the questioner’s agency and thus a position of “going along with the other” (ibid., 16). Okay thus contrasts with other response formats such as repetitional answers, and it can be heard as “begrudgingly and [as] happily accommodating” (ibid.), that is, as an affiliative or disaffiliative response. The kind of stances that OKAY displays when it responds to epistemic initiating actions (e.g., informings, tellings) have also been subject to investigation from different viewpoints. In her study on the responsive use of okay in German interactions, for instance, Oloff (2019) concludes that okay displays a neutral epistemic stance instead of agreeing with the validity of the information provided and its function in the interaction. Evidence is provided by comparing okay as a turn of its own with cases in which okay forms a more elaborated response with a change-ofstate token that precedes it (ah okay, ach okay), and analyzing the kind of action the preceding turn accomplishes. Oloff notes that through its prosodic design, okay can hint at a possible disalignment. A number of other recent studies document how responsive OKAY turns can foreground stances other than acceptance and agreement through sequential placement and phonetic-prosodic realization. These range from a ‘neutral’ or ‘agnostic’ position, in which acceptance or endorsement of prior information is withheld (e.g., Lindström 2018; Oloff 2019), to a position where acceptance is cast as given reluctantly or under duress (Beach 2020; Stivers 2018), to incredulous or disagreeing positions (Beach 2020; see Chapter 3 this volume). These different uses could be said to share a family resemblance in that they convey, to different degrees, that the speaker has reservations about the prior action. One type of usage in this cluster is that of concession or assent. Such uses have been discussed, for example, by Keevallik (2017), who describes Estonian okei used
12
Emma Betz and Marja-Leena Sorjonen
to go along with a my-side offer. These okeis come after negotiations and are hearable as ‘assenting’ to the offerer’s agenda. Beach (2020) illustrates similar uses of okay in English: to ‘reluctantly comply’ with a request or commit to a future action, often after a pursuit of such a commitment. Depending on its sequential context, a stance of reluctance can amount to ‘being imposed upon’, ‘giving in to another’s wishes’ (Beach 2020), or conceding a point in an argument, that is, ‘standing corrected’ (Couper-Kuhlen this volume; cf. Heisler 1996, 300–301). Another type of usage in this cluster is OKAY projecting disagreement or opposition, or conveying surprise and incredulity. In these uses, speakers mark a prior piece of information, stance, or action as inconsistent or incompatible with prior knowledge (cf. Chapter 3, this volume), or as departing from what is acceptable reasoning or expectable behavior (Beach 2020). Beach (2020) finds that OKAYs displaying such ‘incongruous stances’ are prosodically marked in English, for example through increased syllable duration or final rising pitch. The above studies show that phonetic-prosodic features of OKAY can be crucial for the particle’s meaning, and experimental research has confirmed this. Van Zyl and Hanekom (2013) show for English that agreement can be particularized or qualified through cues such as word duration, making an okay-response interpretable as reluctantly, rather than unreservedly, agreeing (cf. Forsberg and Abelin 2018; see also Bangerter and Clark 2003, Hockey 1993, Stivers 2018 on ‘modulating’ interjections). In the present volume, several chapters deal with responsive uses. In addition to the chapter presenting an overview of OKAY in responding and claiming understanding in all the languages in the volume (Betz and Deppermann), the particle is discussed as a response to both deontic and epistemic initiating actions in English (Couper-Kuhlen). Its use in responding to deontic initiating actions is discussed in Estonian and Polish interactions (Keevallik and Weidner). OKAY as a claim to understanding in informings and tellings is analyzed together with a larger set of multimodal resources in German (Helmer, Betz, and Deppermann), and it is described as a response to informings in Finnish (Koivisto and Sorjonen). For Danish, prosodic format emerges as central in an analysis of okay used to manage understanding in third position following answers to questions (Sørensen and Steensig): The authors identify a use of Danish okay that systematically marks the prior answer as locally informative, yet at the same time unresolved or incongruous within the larger activity, and thus in need of expansion. Prosodically marked uses expressing different kinds of reservations about the prior action are also discussed for English (Couper-Kuhlen), Finnish (Koivisto and Sorjonen), German, and Polish (Chapter 3).
Chapter 1. Introduction 13
4. Research on OKAY in institutional discourse and elicited interaction Compared to responsive uses, transitional OKAY has received less attention in ordinary conversation. In institutional and task-based interaction, by contrast, a large body of work focuses on uses involved in activity and topic shifts. This likely reflects systematic differences in setting: As agenda- or script-based shifts between topics and between task or project phases is constitutive of many institutional interaction types, but not of mundane conversation, the relevant environments for using transitional OKAY exist to a greater degree in institutional interaction. In this section, we first summarize some representative studies and then trace common themes and issues emerging across the body of work we surveyed. We show how early interest in institutional OKAY, and particularly in OKAY as a device for managing transitions, reorientations, and closings, has shaped – and also possibly limited – what we know so far about OKAY. One of the first publications having OKAY as its focus investigates service encounters in American English. In two connected papers, Merritt (1976, 1978) studied okay used at the cash register counter in shops and stores. She shows first that, contrary to its description in some dictionaries, okay was not merely displaying acceptance, approval, or acknowledgement of the prior action, and it was not synonymous to yes. She gave evidence for this through sequential analysis of the encounter showing, first, that okay occurred in a particular position of a particular sequence, as a response to the customer’s request for action, for providing them a product. Okay is thus differed from yes-type responses (cf. Bolinger 1957), which were used as a response to customers’ requests for information (e.g., concerning the availability of a product).18 Merritt showed that simultaneously to displaying acceptance, okay contributes to managing a larger organizational task, that of building “a bridge, a linking device between two stages or phases of the encounter” (1978, 11), achieving a transition from the request for a product and its acceptance to the embodied fulfillment of the request (cf. Keevallik and Weidner, this volume, for an in-depth analysis of similar uses in Polish and Estonian). Merritt’s studies constitute methodologically exemplary early work on OKAY. In many ways, they prefigure usages of OKAY that were discussed and developed more systematically in later studies. One such study is Nevile (2004), who analyzes Australian English talk-in-interaction in the airline cockpit. In this ‘sociotechnical’ (Anderson, Hughes, and Sharrock 1989, 136) setting, in which humans interact with each other while also managing complex technical systems, OKAY emerges as a central lexical resource in 18. Merritt draws a distinction between “‘yes’ items (‘yes’, ‘yeah’, ‘yeh’, ‘yep’, ‘umhmm’) and ‘O.K.’ items (‘all right’, ‘O.K.’)” (1978, 3). She finds occasional examples of the former after requests for action of the Can I have x format, but no examples of the latter after Do you have x requests for information (ibid., 4–5).
14
Emma Betz and Marja-Leena Sorjonen
pilots’ coordination of talk with non-talk activities and in pilots’ integration of talk within and beyond the cockpit (e.g., with air traffic controllers). Nevile shows that responsive okay accepts the prior talk, for example an action-implicative informing, as “routine and unproblematic” (2004, 75). By accepting a prior informing, speakers also accept the concomitant responsibility for action and the institutional identities this implies (ibid., 41–42). In cockpit interactions, okay also regularly indexes “some talk as separate to prior talk, and the task which ‘okay’ initiates as separate to prior tasks” (ibid., 201). In this way, turn-initial okay can serve as a bridge between talk inside and outside the cockpit: A pilot’s okay marks talk with an air traffic controller as closed and the talk following okay in the same turn as initiating something new that emerges from the prior and is relevant for here-and-now in-cockpit tasks. Okay thus indicates that “[t]alk between the pilots was somehow occasioned by talk beyond the cockpit” (ibid., 181). Overall, Nevile shows through detailed sequential analysis how okay serves complex bridging functions and thereby reflects and reifies the specific institutional setting in which it occurs. Table 2, which is organized by setting, provides an overview of studies that focus on or include a substantial discussion of OKAY in different institutional contexts. Table 2. Studies on the use of OKAY in discourse in different institutional contexts* Setting/interaction type
Language(s) investigated
Authors
Service encounters, urban settings (indoor/outdoor, 1964–1973) Service request calls to local city council office Medicine dispensing interaction in a pharmacy department Calls to a student financial support office
American English
Merritt (1976, 1978)
Australian English
Kelly (2010)
South African English
Huddlestone and Fairhurst (2013) Lindström (2018)
Sociolinguistic interviews (1971 and 1984) Sociolinguistic interviews
Québécois French
Sociolinguistic interviews (1980’s-2000’s) Job interviews Clinical psychology interviews
Swedish
Laforest (1992), Heisler (1996) Liao (2009)
American English (L1 Chinese) Estonian, Swedish Estonian Keevallik (2013) German American English
Courtroom interrogations Police interview/interrogation
American English American English
Doctor-patient interaction
American English
Doctor-patient interaction Speech therapy sessions (video)
German American English
Birkner (2001) Antaki, Houtkoop-Steenstra, and Rapley (2000) Beach (1990b) Gaines (2011) Ainsworth-Vaughn (1992); Beach (1995b); Jones (2001) Kaiser (2011) Kovarsky (1989)
Chapter 1. Introduction 15
Table 2. (continued) Setting/interaction type
Language(s) investigated
Authors
Focus group meetings Academic meetings (different disciplines) Business meetings Business meetings
American English American English
Beach (1990a) Ford (2008)
English German
Boden (1994) Barske (2006, 2009); Meier (2002)
Daytime television talkshows Presidential primary debates
American English American English
Honeman (1995) Schubert (2019)
Academic advising sessions
American English
Master’s supervision meetings with international students
Scottish English (L1) and English (L2, various L1s) English as a lingua franca in Germany
Guthrie (1997); cf. Swales and Malczewski (2001) Bowker (2012)
Academic consultations Airline cockpit interaction, including flight instruction
Australian English American English
House (2013) Nevile (2004) DeSouza et al. (2021)
* This overview excludes work focusing on adjectival and adverbial uses of OKAY. For research on OKAY in assessments, and on the interactional functions of such OKAYs in parent-teacher conferences in particular, see Pillet-Shore (2003).
Beginning in the 1980’s, a second area of research emerged: OKAY was increasingly studied in relation to and as an index of general organizational levels of interaction in elicited talk and constrained discourse, for example to inform machine learning and improve human-machine interaction. Table 3 provides an overview of work on OKAY in constrained discourse and experimental settings. Table 3. Studies on the use of OKAY in elicited and constrained discourse and in experimental settings Setting/interaction type
Language(s) investigated Authors
Human-computer dialogues: problem-solving tasks; information retrieval
American English
Interaction tasks: families engaging American English in decision-making about vacations Student dyads make travel American English decisions: co-present (audio), or synchronously computer-mediated Interaction tasks: computer-mediated American English decision-making
Grosz (1977, 1982)
Condon (1986) Condon (2001) Condon and Cech (2007)
(continued)
16
Emma Betz and Marja-Leena Sorjonen
Table 3. (continued) Setting/interaction type
Language(s) investigated Authors
Info gap task: design reconstructions American English Info gap task: map tasks Australian English British English
Hockey (1993) Filipi and Wales (2003) Bangerter and Clark (2003)
Joint tasks: Tangram arrangement American English and task and Lego model building task Swiss German Joint task: identifying moving American English computer images Collaborative puzzle game British English
Bangerter and Clark (2003)
Conversations between work colleagues about pre-assigned topics (1992)
American English
Bangerter and Clark (2003); Bangerter, Clark, and Katz (2004)
Telephone directory enquiries (calls to operator, 1974)
British English
Bangerter and Clark (2003); Bangerter et al. (2004)
Appointment scheduling task with human interlocutor (1995), with simulated computer (2003)
American English
Fischer (2006)
Machine learning experiment (coding instances by discourse function)
American English
Gravano et al. (2007)
Elicited utterances rated for level of American English agreement (experimental studies) Swedish
Gravano, Hirschberg, and Beňuš (2012) Knutsen, Col, and Le Bigot (2018)
Van Zyl and Hanekom (2013) Forsberg and Abelin (2018)
Within this cluster of work, Condon (1986, 2001) investigated experimental decision-making interactions with the aim of showing how the distribution of ‘interjections’ can be related to more abstract levels of discourse organization, such as topical organization, adjacency relations, or interpretive frames (e.g., joking- serious). She focused on American English okay and found that it appears at “significant structural boundaries” (Condon 2001, 496; cf. Condon and Cech 2007), that is, where several organizing systems coincide. This parallels early findings on the role of okay in telephone opening and closing routines (Schegloff 1968, 1986; Schegloff and Sacks 1973). Condon also compared okay to other boundary markers, notably well and found that okay oriented to expected next business and instantiated routine transitions, while well signalled departures from routine transitions, conveying dispreference or unexpectedness. The work summarized in Table 3 generally treats OKAY in comparison with other discourse markers. A third area of research in which the study of OKAY figured prominently started growing in the late 1970s and early 1980s and focuses on pedagogical settings. Early in its emergence, this research highlighted discourse-structuring uses
Chapter 1. Introduction 17
of OKAY in academic discourse and specifically in lecturing activities, making one type of ‘structuring’ OKAY especially visible: the “lecturer’s OK” (Levin and Gray 1983). Table 4 provides an overview of work focusing on OKAY, or including OKAY among other discourse markers, in pedagogical contexts. Table 4. Studies on the use of OKAY in discourse in different pedagogical contexts Setting/interaction type
Language(s)/speakers investigated
Authors
American English
Levin and Gray (1983)
(a) Lecturing/monologic use: Graduate student seminar presentations
Reading-style university history American English (for ESL Chaudron and Richards lecture; effect of discourse students in the US) (1986) marker use on comprehensibility Research-in-progress seminars in computer science
Australian English
Rendle-Short (2000)
University lectures in social sciences American English and and humanities German American English
Schleef (2005)
Lectures (different disciplines)
Othman (2010)
British English
Schleef (2008)
Presentations at an international English, different varieties Querol-Julián and Bellésacademic conference (chemistry) Fortuño (2010) Lectures and seminars
South African English
Huddlestone and Fairhurst (2013)
Academic medical lectures, given in English (NNS, mostly L1 Saudi Arabia and England Arabic) and British English
Al Makoshi (2014)
Mathematics lectures by graduate American English students; focus: self-directed talk
Looney, Jia, and Kimura (2017)
Graduate math and undergraduate social science lectures; focus: closing phases
Tyagunova and Greiffenhagen (2017)
British English
(b) Different types of pedagogical activities, teachers’ use: Classroom interactions at the British English boundaries of different activities (teacher-led primary school)
Sinclair and Coulthard (1975)
Different academic speech events (MICASE corpus, audio only), focus: format shifts, redirection
Swales and Malczewski (2001)
American English
(continued)
18
Emma Betz and Marja-Leena Sorjonen
Table 4. (continued) Setting/interaction type
Language(s)/speakers investigated
Authors
Undergraduate seminars or discussion sections
American English (L1 Chinese) American English and German
Liao (2009)
Bilingual elementary school, US, Spanish and English instruction
Spanish and American English
Broderick and Broderick (2003)
Feedback in community adult ESL American English (ESL), US and TESOL methods classes interactions English, Indonesian EFL high school language class, (L1 Javanese) Indonesia EFL secondary school teacher-fronted English (L1 Korean) language class, Korea University seminars, Germany and Belarus; focus: closing phases
Schleef (2005)
Waring (2008); Fagan (2012); cf. Hellermann (2005) Karlina (2015) J.-M. Lee (2017)
German, Belarussian (data Tyagunova and in translation only) Greiffenhagen (2017)
(c) Different pedagogical activities: students’ use in language learning contexts Secondary-school group discussion/ English learners (L1 Fung and Carter (2007) negotiation task Chinese) in Hong Kong Pedagogical subcorpus of British English native Fung and Carter (2007) CANCODE (1998) speakers University classroom bookclub discussions
ESL students (L1 Japanese) Johnson (2016) in the US
Peer interactions: role play preparation outside of class
University learners of German in Canada
Reichert and Liebscher (2018)
Levin and Gray (1983) investigated American English okay used by graduate students in a lecturing role. They found okay at points of topic change, related to discourse boundaries of different strengths. Instead of treating the usages of okay they found as related to the setting (lecturing), Levin and Gray suggested that they were related to the type of ongoing task. The task in question, typical for a variety of settings, was to give a presentation built on a prepared agenda (manifested in, e.g., lecture notes) containing various items that needed to be covered one after another in a series. Within such a task, okay contributes to displaying that the speaker is taking stock of where they are in the agenda, whether everything is covered, and what the next step is. The interest in OKAY in pedagogical contexts, and particularly in lecturers’ use, continues today (e.g., Looney et al. 2017), underscoring the importance of teaching as a socially central, conventionalized activity. This body of work now increasingly documents students’ deployment of OKAY in learning
Chapter 1. Introduction 19
activities, an example of which can be found in Reichert and Liebscher (2018), who show that students use German/English okay (the particle cannot always be clearly assigned to a language, p. 133; cf. Flores-Ferrán 2014) to shift into and out of role-play practice, and thus also between the languages associated with different activities and between character roles. Across the research we surveyed, OKAY has been studied primarily as a structuring device, that is, as a resource for managing transitions, reorientations, and closings. These are uses that may be particularly salient/exposed in contexts that are pre-structured to a higher degree than is the case in every-day interaction: Transitions between topics, tasks, and phases of an activity (e.g., topical parts of a lecture, phases in a task-based teaching unit) are perceived as crucial work in managing these contexts through interaction and are thus more readily noticed by researchers interested in discourse structures. Some themes that emerge from the body of research on OKAY as a structuring device are the following. Backward/forward orientation: OKAY has been described in light of its relationship to preceding talk and to what follows or is projected to follow. In responsive uses, especially where OKAY forms its own turn, it is primarily backward-looking (see Section 3) but it “has been shown to possess fundamental ‘projective’ qualities” (Beach 1995a, 134). This forward-looking quality is particularly evident in OKAYs that preface fuller turns, and their work in institutional contexts has been described as changing footing in self-directed talk (Rendle-Short 2000; Schleef 2008; Looney et al. 2017), instantiating perspective shifts (Filipi and Wales 2003), or resuming a pending unit (Polanyi and Scha 1983, 265). Research on OKAY in institutional contexts thus particularly highlights how OKAY occupies sequential ‘opportunity spaces’ for transitioning to a next step or phase of an activity. This positional sensitivity is crucial for recognizing the role of OKAY in showing understanding of and shaping the unfolding interaction: OKAY is “positionally active and consequential for unfolding talk” (Beach 1995a, 122; see also Birkner 2001, 92f.).19 Topic organization: The dual-faced character of OKAY makes it particularly suitable for managing shifts at different levels of organization, thus reflecting participants’ perception of what the current talk amounts to. In work on lecturing and on agenda-setting in medical talk in particular, the role of OKAY in managing topic has been foregrounded (e.g., Ainsworth-Vaughn 1992; Broderick and Broderick 2003; Guthrie 1997; Kelly 2010; Rendle-Short 2000; Schleef 2008; see also Beach 19. This has been brought out for uses in responsive position as well: See Stevanovic and Peräkylä (2012) on acknowledging that a first action has consequences, and on Finnish okei and selvä ‘all right’ as “action-accepting compliance tokens” (p. 304; cf. Sorjonen 2001, 111); Stivers (2006) on okay as a display of commitment in response to treatment decisions; Couper-Kuhlen (2019a) on ‘consequential uses’ of okay in American English.
20 Emma Betz and Marja-Leena Sorjonen
1990a, 215, on regulating topics; Jensen 1987, as cited in Beach 1995b). In contexts such as lecturing, where turn-taking organization is constrained, ‘topic’ may be a stronger and more important source of coherence than in everyday interaction, and topical organization become more clearly traceable in, for example, the use of OKAY. Beach (1995b) acknowledges that topic is one layer of organization to which OKAY contributes, and he illustrates how okays are recruited in medical interviews in the service of the clinical agenda and goals: Okay can mark the sufficiency of a patient’s answer or elaboration of a topic but can also embody a unilateral shift that amounts to a refusal to address a response or topic raised (p. 284), thus controlling topical progression. However, Beach (1995b; cf. Schegloff 1990) also cautions against a narrow analytic focus on topic, because it may not adequately grasp the particle’s sensitivity to sequential context (see Fischer 2006) and limit our understanding of the functional range of OKAY and thus our grasp of the types of organizational structures and activities that are co-constructed by participants in teaching, medical interviews, etc. A consideration of topic development may, however, provide clues for distinguishing between tokens that appear in similar contexts doing very similar work, such as English okay and alright/allright. In a study on Australian English, Filipi and Wales (2003) connect uses of alright and okay to topic development. Analyzing map task interactions, they find that speakers used both okay and alright to signal perspective shifts when giving route descriptions, but differences could be seen when speakers resume the main route-giving activity after an aside: Here, okay treated the intervening talk as interruptive to the main activity, while alright indexed the next instruction as new rather than as a return. Thus, okay was “more strongly associated with topic continuance” (ibid., 453; see also K. Turner 1999, discussed in Section 3).20 Types of transitions: One recurrent aim in existing studies on transitional OKAY is to unpack the types of structural distinctions interactants make, and thus mark as relevant, in shifting. Typically, this is pursued by comparing OKAY with other discourse markers. Such work shows how OKAY manages or contributes to managing various types of shifts: from an aside or thematic digression back to the main task at hand (Beach 1995b; Filipi and Wales 2003; Looney et al. 2017, 55; see also K. Turner 1999), from topical talk to a game, from everyday interaction to medical talk in doctor-patient encounters (Beach 1995b), or between different keys and stances, such as from joking to serious talk in decision-making sequences (Condon 1986, 95–96). Some more specialized forms of transition where OKAY is used occur between tasks or type of activities which rely on different languages: 20. On the possible relevance of topic continuance for OKAY in everyday interaction in German, see Oloff (2019, 216).
Chapter 1. Introduction 21
Reichert and Liebscher (2018) show how students shift from discussing the task in English to implementing it in the form of a role play in German by using okay in the transition space. OKAY has also been identified in transitions from one section of a lecture to another (Rendle-Short 2000, ‘text bracketing device’; Schleef 2008, ‘structural marker’), where it makes the structure and the progression of the lecturer’s monological discourse intelligible (see also Tyagunova and Greiffenhagen 2017). An important finding emerging from this research is that in marking readiness for, or relevance of, transitioning, OKAY seems to be specialized for transition phases at which different levels of organization coincide and for facilitating movement to matters that are mutually understood or expected as next business (e.g., Condon 1986, 94, ‘at unmarked transitions’; Condon 2001, 492, ‘transitions to some expected sequence of talk’; Fischer 2006, 440, ‘previously agreed upon’ topic; Condon and Cech 2007, 21, ‘significant, yet expected, transition in the interaction’). The different types of transitions are conceptualized by Bangerter and Clark (2003) as ‘vertical’ vs. ‘horizontal’, the implementation of which they relate to different particles, termed ‘project markers’. Bangerter and Clark (ibid.) describe as ‘vertical’ transitions those that are made between levels in the hierarchy of some joint project (entering projects or subprojects), and these are implemented in their English data by okay and all right. ‘Horizontal’ transitions, by contrast, are ones where the participants move forward within the same level of hierarchy. These transitions are managed with uh-huh, m-hm and yeah (cf. Condon and Cech 2007; Guthrie 1997; Rendle-Short 2000). Bangerter and Clark (2003) show how both types of transitions are managed sequentially yet index larger project structures. They also offer some initial observations regarding cross-linguistic similarities and differences: Overall, German/Swiss German and American English speakers constructed the same general contrasts between horizontal and vertical transitions in the experimental game situations (Tangram). But the German speakers could be seen to use okay for transitions at different activity levels in comparison to the English speakers. In the present volume, several chapters expand on the themes outlined above: Kuroshima et al. analyze the specific role that duplicated okhe okhe/okke okke plays in closing ‘protracted’ sequences and thus in navigating boundaries between larger activities in Korean and Japanese. Ostermann and Harjunpää document how call-takers in health helpline calls in Brazilian Portuguese employ oquei in managing alignment and the progression of activities that are part of the institutional agenda. Focusing on transitional uses as well, De Stefani and Mondada point to the limitations inherent in understanding OKAY along the dimensions of backward v. forward orientation. They show that French oké and Italian occhei do not initiate or achieve transition on their own, but rather contribute to complex, multimodal arrangements, which are best conceptualized as extended ‘transition phases.’ These
22
Emma Betz and Marja-Leena Sorjonen
studies use data from settings such as helpline calls (Ostermann and Harjunpää), hospital emergency triage, sushi bar service, guitar tutoring sessions (Kuroshima et al.), participatory public meetings, and guided tourist tours (De Stefani and Mondada). Their analyses thus also broaden our understanding of the range of institutional structures and participants’ concerns to which OKAY is sensitive. The next section will outline further research directions emerging from existing work on OKAY, including those that, as in the case of Bangerter and Clark (2003), foreground cross-linguistic questions. 5. Questions and research directions emerging from existing work on OKAY Some prior research has suggested that ‘accepting’ may be a kind of basic function of OKAY (Beach 1993, 1995b, 2020; Schegloff 2007, 120–123; Stivers 2018 ‘acquiescence’; Oloff 2019 ‘neutral acceptance’) and some work treats ‘agreement’ as its core meaning (Pavlidou 1998; Maschler 2002; Schleef 2005; Gaines 2011). However, research on response particles in general and on OKAY in particular (e.g., Beach 1995b; Condon 1986, 98; Kovarsky 1989; Thompson et al. 2015) has repeatedly shown that particle meanings are heavily context-dependent and cannot be grasped without considering prosody (e.g., Beach 2020, Gravano et al. 2007, 2012; cf. Van Zyl and Hanekom 2013; Forsberg and Abelin 2018), sequential and turn placement, preceding action, and larger levels of organization, which may include roles and agendas in institutional settings (e.g., Barske 2006, 2009; see also De Stefani and Mondada, this volume; Ostermann and Harjunpää, this volume). A goal of interactional research thus has to be to specify, for example, what ‘acknowledgement’ or ‘acceptance’ or ‘agreement’ specifically ‘means’ (i.e., amounts to) in its local sequential and action environment: How is it particularized and consequential for the unfolding interaction and the participants’ next actions? Engagement with OKAY brings out this central issue in studying particles, and some recent work by Lindström shows what can be discovered about the complexities of OKAY when we take seriously its inherent indexicality: Lindström (2018) investigates okej-prefaced responses by administrators in calls to the Swedish Board for Study Support. She shows that they recurrently implement alignment shifts from a clients’ perspective to the institutional record and thus take an ‘agnostic’ position on the client’s prior narrative. In “bracketing off the prior turn as neither true or false” (ibid., 368), okej cannot be said to convey ‘agreement’ with prior talk, or ‘acceptance’ of prior information, but rather registers information in a neutral way, thus also “mitigating the potential opposition inherent in such shifts” (ibid.) to an institutional perspective.
Chapter 1. Introduction 23
One way to aim for a holistic description of OKAY is to consider embodied aspects of its use, in addition to its phonetic-prosodic realization and precise placement in turns, sequences and activities. Some studies on OKAY in face-to-face interaction have begun this work. We show some of their findings below, as this is an area where we expect an increasing number of studies in the future.21 An early study that discussed complex aspects of the delivery of OKAY is Levin and Gray (1983). In monologic lecturing in English, Levin and Gray observe the recurrence of ‘unfilled pauses’ before okay, accompanied by decreased speaking volume and gaze aversion from the audience. They note that these features make visible a discursive break in a task- or agenda-based activity, “during which the speaker takes stock of where he is and what the next step is” (ibid., 199; cf. Looney et al. 2017).22 Rendle-Short (2000) also associates different uses of okay within sections of lecturing discourse with different kinds of patterns of pausing and phonetic-prosodic features. For example, as a ‘comprehension checking device’, okay may be preceded by a pause, delivered more quietly, and with rising intonation. A number of studies highlight the importance of gaze and orientation to objects for identifying the function of OKAY: Rendle-Short (2000) notes that falling-intoned okays at discourse boundaries are regularly accompanied by changing of projector slides after a closed-off topical section. This makes visible how this kind of okay “finally closes off the previous section to enable the next section of talk to begin” (ibid., 26). Querol-Julián and Bellés-Fortuño (2010) similarly draw on systematicities in body movement and gaze direction to delineate different uses of English okay in conference presentations. Okay as ‘pause-filler’ (i.e., transitional) co-occurs with changes in position, gaze direction at computer screens, and slide transitions. This contrasts with okay as a ‘confirmation check’, during which speakers gaze at the audience. Barske (2006, 2009) also traces a connection between eye gaze and the function of German okay in workplace meetings. As a continuer, okay is delivered with maintained eye-gaze, but as a transition marker, it is systematically connected to gaze withdrawal from current co-speakers, and with orientation toward relevant objects (e.g., the written meeting agenda) or other participants. 21. Note that the terminology we use in this section may vary, reflecting variation in the authors’ terms across time and research traditions. We mark authors’ own terms with single quotes. 22. Levin and Gray (1983) also offer anecdotal evidence for uses of similar ‘self-directed’ OKAY by students before answering a lecturer’s question. Pre-answer uses of OKAY, such as those in question-answer periods after academic lectures, are the focus of DeSouza et al. (2021), whose analyses consider gaze aversion, pauses, and posture adjustments before OKAY. They interpret these as working in tandem with OKAY to mark the transition to ‘beginning to respond’, while OKAY itself allows for projections about how the response will unfold (specifically: that it will not begin with the answer to the question).
24
Emma Betz and Marja-Leena Sorjonen
Lastly, Looney et al. (2017) show that gaze, gesture, and orientation to objects in addition to sequential position are crucial in distinguishing two kinds of okay for ‘doing thinking/planning’ in self-directed talk in mathematics lectures: Transitional okays at major boundaries differ from self-initiated self-repair uses in that they are produced with reduced volume, sometimes after pauses, and are characterized by thinking gestures and orientation to written notes. This indexes planning and makes projectable that the lecturer is returning to the main topic or beginning a new phase of the lecture (cf. Schleef 2008, 71, on paper shuffling accompanying okay during phases of reorientation). The research summarized above focused on transitional uses of OKAY. Some work on responsive uses also points to the importance of multimodal design. Merritt (1978) notes in her transcripts that OKAY that works as an affirming or granting response to a request for action “provides an expectation of something to follow” (p. 6) and is followed by bodily turns away (to retrieve the product). Overall, she concludes that “‘O.K.’ might have a special role in bridging the transition from a verbal to non-verbal mode of interaction” (ibid., 5). Nevile (2004, see Section 4) makes a similar point in his analysis of how pilots coordinate talk and non-talk activities in the airline cockpit and the role okay can play in this. Keevallik and Weidner (this volume) extend these analyses to directives in Estonian and Polish. Couper-Kuhlen (this volume), who highlights the importance of prosody in analyzing the breadth of uses of okay in American English, points to gaze aversion and lower volume as features of responsive okay that contribute to implementing ‘qualified acceptance’ of the prior turn, showing less than full commitment to a course of action. The detailed analysis of okay in its multimodal context is also pursued in this volume by De Stefani and Mondada as well as Helmer et al.; both chapters considerably expand the breadth of contexts in which multimodal realizations of OKAY are analyzed. A second cluster of research questions emerging from existing work concerns change in the use of OKAY across time: To which extent does work on English show systematic diachronic differences in uses and/or frequency? How did OKAY spread in the languages into which it has been imported, that is, which uses are observable early, which develop later in its integration? Are there differences between languages in this respect? One longitudinal perspective contributing to a fuller picture of OKAY relates to language socialization and the ontogenesis of pragmatic/interactional competence. In a child language development study on ‘little words’, Pak et al. (1996) report on systematic changes they observed in children’s use of English okay across time (1;5 to 7;5 years).23 Two main findings emerged: (1) Uses are initially more limited and then diversify, that is, children ‘decontextualize’ okay as a discourse device; (2) Social
23. 1;5 means one year and five months, and 7;5 seven years and five months.
Chapter 1. Introduction 25
functions related to face-to-face interaction emerge before discourse-structuring uses at levels higher than the adjacency pair. In other words, children begin using okay “in the highly constrained discourse context of adjacency pairs” (ibid., 295) at age 1;3 to accept offers and comply with requests, and at age 2;6 as tags making response relevant.24 They then, at age 3;9, add uses signalling agreement and verification, acknowledgment and approval, and around this age also begin expanding the range of okay to marking changes in activity or activity phases, topic shifts, and changes in speaker’s roles. Syntactically integrated adjectival, adverbial, and nominal uses are the last to appear as “the marker interacts with other aspects of syntactic knowledge and other syntactic frames which are more ideationally based than interactionally based” (ibid., 295). This kind of work not only adds to our knowledge of how children acquire language (forms) but also sheds light on how they orient to underlying norms of conversation and how children’s increased mastery of formal structures enables them to engage in social interaction in increasingly context-sensitive and fitted ways.25 Questions about change over generations have not yet been extensively explored in existing work on OKAY. In line with what we can observe in early accounts on English (see Section 2 above), some researchers hypothesize a systematic functional development from acknowledging and agreeing (as well as complying) to transitioning uses (Condon 2001, 510). Changes in observable functions over time are suggested for American English by Couper-Kuhlen (this volume), and she also offers some initial observations regarding frequency of use across time. For languages other than English, we only have first indications for changes in frequency or functions over time: Heisler (1996) observes “a nine-fold increase in the frequency of use” (p. 294) over the span of his data collection (interviews recorded in 1971 and 1984) in Montréal French. He analyzes this change in okay use by function (as agreement, interactive, and structural marker) and different sociolinguistic variables (sex, age, social class, level of education) and sees the clearest increase in almost all uses in younger male speakers (15–33 yrs). He concludes that “the era of OK’s proliferation in Montréal French was sometime in the seventies” (ibid., 309), rooted in the influence of Canadian and American English. Koivisto and Sorjonen (this volume) also show differences in frequency in Finnish conversational data between the late 1980’s and the 2010’s, in okei’s frequency in general, and in its use after informings in particular. Future longitudinal studies in different languages could pursue this direction and provide answers to questions of frequency and spread over time. 24. Some work in the present volume also discusses uses of OKAY to elicit a response (Couper-Kuhlen; Ostermann and Harjunpää). 25. For examples and an overview of the emerging body of conversation-analytic inquiry into early childhood development, see Wootton 1981, 1997; M. H. Goodwin 1990; Cekaite 2007; Sidnell 2016; Stivers, Sidnell, and Bergen 2018.
26 Emma Betz and Marja-Leena Sorjonen
Lastly, promising avenues for future cross-linguistic research lie in exploring the relation of OKAY to other particles in a language. OKAY entered different languages at different times, with sources in many pointing to the early(er) 20th century. According to Danish and Swedish lexicographic sources (Ordbog over det danske Sprog, n.d., and Svensk ordbok, n.d.), for example, the word appeared in the two languages in the early 1920’s and the 1930’s respectively. Similarly, written records in Korean and Japanese point to its appearance as early as the 1920’s in popular Korean magazines (M. Y. Lee 1932) and 1930 in a popular Japanese song (Kato 2001). A Korean loanword dictionary includes the word in 1937 (J. K. Lee 1937). First attested uses of okay/occhei in Italian also date back to the 1930’s (Panzini 1931), as does okei in Finnish (Paunonen and Paunonen 2000), which is included in the main dictionary of modern written Finnish (Kielitoimiston sanakirja 2012, s.v. okei). Okay is said to have entered German in the 19th century already (Pfeifer 1993), but a noticeable spread in written sources begins in the 1940’s, and it made its first standard dictionary appearance in 1954 as O.K. (Dudenredaktion 1954). In Estonian okei appears in written sources only in the 1990’s, but its use in spoken interaction in informal settings seemed already well established at that time. This contrast may reflect changes in language editing after the Soviet era (Leelo Keevallik pers. comm.). The above is a sample of languages and some indications of when OKAY first appeared in them; in the remainder of this section, we want to outline questions regarding OKAY’s spread once it has entered a new language. Longitudinal studies on the use of OKAY in different languages can show us how OKAY progressively integrates into existing systems of discourse particles in a given language and how this integration affects (shifts, limits) the range of uses of other particles. On Hebrew, for example, Metcalf (2010) offers the following anecdotal observation: “The Hebrew word that OK has replaced is b’seder. Seder is “order,” and b’seder is “in order.” So b’seder could be translated as “alright.” But OK is now more common than b’seder” (ibid., 172). It would be worth exploring whether interaction research supports this. The talk-in-interaction data in Maschler (2002) documents okey for agreement, but the numbers of okey and b’seder/beseder are too low to allow for conclusions about relative frequency. Maschler (ibid.) allows for a tentative conclusion about the functional spectrum of okey in Israeli Hebrew, however: While okey has taken its place among discourse markers used for agreement (along with, e.g., ken ‘yeah’, naxon ‘right’, bídyuk ‘exactly’, beseder ‘alright’), it does not appear as a structural marker. The range of uses in Israeli Hebrew thus differs from that in English and other languages – this, at least, was the case in the mid-1990’s, and it may have changed since then. How the integration of OKAY has precisely worked and works in different languages is something that is not known at the moment. It remains to be seen whether a usage of OKAY in particular sequential environments (e.g., transitioning
Chapter 1. Introduction 27
but not compliance) may become salient when there is, for one reason or another, space for a ‘newcomer’ (cf. Sorjonen 2001, 89–91 on joo, loaned from Swedish into Finnish). Work attempting to trace the progressive integration of OKAY into a language and the resulting restructuring (e.g., a possible displacement of b’seder by okey in Hebrew) requires extensive mapping of the interactional uses of a broad range of particles at different points in time, which is a daunting task. Incomplete knowledge of the system of discourse markers and response particles in different languages is currently a significant limitation for describing the relation of OKAY to other particles from both a diachronic and a synchronic perspective. However, some initial observations regarding the place of OKAY in the systems of particles in specific languages – based on existing studies and from research that underlies the present volume – may serve as a starting point. Kaiser (2011) observes that affirmative okay and gut ‘good’ appear in similar sequential environments in German, but okay affirms and evaluates more strongly (p. 74; cf. Birkner 2001, 93). Transitional okay and gut, however, do not show differences in use (closing and projecting) in Kaiser’s data, leading her to suggests that individual speakers’ preference may guide their selection; further work comparing okay and gut in German could shed light on this matter (cf. Werlen 1984 on okay and gut for summarizing). Koivisto and Sorjonen (this volume) compare Finnish okei with the particles joo, joojoo and just in similar sequential contexts and find differences at the level of topic and sequence development: okei in third position orients to the just-closed informing sequence as part of a larger activity. They also differentiate okei from the change-of-state particle proper, ai jaa. Their research suggests that the particle system for receiving informings in Finnish may be in a process of reorganization involving okei, moving toward a more differentiated treatment of informings by recipients. In work on languages other than English, authors’ translations can provide hints at functional equivalents of OKAY and thus point at expressions with which OKAY may compete once it has entered a language. In our discussion of Hebrew above, we noted that OKAY does not appear in transitional uses, as a “structural marker introducing an action” (Maschler 2002, 22). Maschler lists tov as the most common lexical item used for this function and translates tov as ‘okay’. If okey expands its range of uses in Hebrew, then it may affect the range of uses of tov. Similar questions could be asked about hao ‘good’ in Mandarin, which has a wide range of interactional uses, including approval in responsive uses, topic closing, and transitioning (cf. Xian 2007). Hao is regularly translated as ‘okay’ into English (cf. Wang et al. 2010; Yeh and Huang 2016), and as ‘sure’ when granting a request (Kendrick 2010). Okay exists in Mandarin (see Chapters 3 and 4 in this volume for examples), and its functions overlap with those of hao (e.g., receipt in second and third position, transition, closing). Similarly, we can observe that Turkish tamam
28
Emma Betz and Marja-Leena Sorjonen
for claiming understanding and acceptance, for example of a piece of instruction done through an embodied demonstration, is translated as ‘okay’ in existing conversation analytic research (Tekin 2019), but that speakers of Turkish also use the word okay (Burak Tekin, pers. comm.).26 The present volume features work on okej in Polish (Keevallik and Weidner), but we know little about OKAY in other Slavic languages. In a study on closings in Russian, Bolden (2017) analyzes (nu) ladno, ladnen’ko, davaj as pre-closings and, reflecting existing work on OKAY in English closings (Schegloff and Sacks 1973), translates them as ‘okay’ and ‘alright’. Okej exists as a separate lexical item in Russian and appears in Bolden’s data in pre-closing and other sequential contexts. For Czech, Oloff (2018) notes that oukej is a recent addition and infrequent in its use in conversation. Her data suggests, however, that a comparison, and possibly emerging differentiation, between responsive oukej and dobrý/dobře ‘good’ in Czech may be worth exploring. Cross-linguistic data sessions that were part of the work toward the present volume suggest further avenues for exploring the place of OKAY in different languages, including the distinctions drawn in Brazilian Portuguese between ta ‘sure, all right’ and tá bom ‘good’ and oquei; in French between d’accord (‘agreed’) and oké; in Italian between allora/dunque ‘so, then’ and okay in transitions, or bene ‘good’ and okay in sequence closings; in Korean between al-ass-e ‘got it’ (based on the verb ‘to know’), ung ‘yes’, e ‘yes’, kwaynchan-a ‘fine’, and okhe, or in Japanese between hai ‘yes’, yoshi ‘good, alright’, sequence-closing nn ‘yes’, and okke. Many of the tokens listed above can combine with OKAY in their respective language, either in one or under different prosodic contours. Work describing the place of OKAY in a specific language thus also needs to consider which co-occurrences with OKAY are possible and recurrent (cf. Heritage 1984a on oh okay in English; Schegloff 2007; Helmer et al. this volume on change-of-state tokens preceding okay in German) and explore the kind of systematic division of labor this makes visible in the particle systems of different languages.
Funding Emma Betz’s work was supported by Canada Foundation for Innovation funding for the creation of a Social Interaction, Language, and Culture Lab at the University of Waterloo (CFI-JELF project #37510). 26. Other languages in which OKAY is used and which await investigation include Farsi (okay, Carmen Taleghani-Nikazm and Reihaneh Afshari Saleh, pers. comm.) and Hungarian (oké, Csilla Dér, pers. comm.). There are studies on Arabic particles that in some of their usages come close to the use of OKAY, including Syrian Arabic (see Al-Khalil 2005; Martini 2020 on aywa) and Lebanese Arabic (Ayash 2016). Some varieties also seem to have OKAY borrowed from English, for instance Palestinian and Lebanese Arabic (Irina Piippo and Véronique Traverso, pers. comm.).
Chapter 2
Data and methods used in the study of OKAY across languages Arnulf Deppermann and Lorenza Mondada
Leibniz-Institut für Deutsche Sprache, Mannheim / University of Basel
In this chapter, we overview the specificity of comparisons made within the perspective of Conversation Analysis (CA), and we position them in relation to other fields. We introduce the analytical mentality, methodology, and procedures of CA, and we show how we used it for the analysis of OKAY in this volume. Keywords: comparison, data, transcription, recording, multimodality, sequentiality, context
1. Comparing languages and language use across cultures Comparison has been at the core of the linguistic enterprise since the very beginning of its history: in diachronic linguistics, one state of a language is compared to another state; in anthropological linguistics, the language of one culture is compared to the language of another. Even within a definition of language as “un ensemble où tout se tient” (Meillet 1903, x), i.e., a system where every element is related to all the others, structural linguistics identifies and combines minimal units on the basis of similarities and differences. Comparison is omnipresent in linguistic analysis. Yet it is implemented through very different analytical procedures and concerns very different objects (from phonemes to words and syntactic constructions, from speech acts to interactional practices). Comparative studies have classically focused on formal resources (phonology, morphology, lexicon, syntax), and much less on actions and interactional practices, although it is possible to find some precursors for these fields as well. Comparisons have been at the core of linguistic anthropology from Franz Boas (1889) onward, grounding relativity theories and controversies about the diversity versus universality of language structures. Building on Humboldt’s thoughts about the close connection between world-view and the structure of languages (Humboldt 1836) and on writings of his teacher Edward Sapir, himself a student of Boas, Benjamin Lee Whorf ’s studies of the Hopi language (Whorf 1956) led him https://doi.org/10.1075/slsi.34.02mon © 2021 John Benjamins Publishing Company
30
Arnulf Deppermann and Lorenza Mondada
to the conclusion that the Hopi have a concept of time that is completely different from that of Western cultures. He based this conclusion on the fact that he did not find linguistic forms in Hopi devoted to the expression of temporal properties. Thus, so-called “linguistic relativism” – claiming that cognitive conceptualization of actions, events, and objects essentially rests on the affordances yielded by the lexical and grammatical structure of the individual language of the thinker – ultimately rests on comparison between languages. While this claim exists in different versions and is far from being undisputed (see, e.g., Lucy 1997), comparative approaches have also led into the opposite direction. Starting with Greenberg’s Universals of Language (1963), linguistic typology has looked for universal properties of languages whose origins are unrelated and whose surface systems are organized in very different ways (e.g., synthetic vs. analytic coding of meaning in grammar). Earlier typological research was mainly interested in universal rules of word order and morphological properties. More recent approaches adopt a functional and cognitive perspective. They try to show how the syntactic, morphological, and lexical options of different languages are used to express universal cognitive categories like animacy, causation, and thematic roles, and how linguistic structure can be explained by reference to general human cognitive capacities (e.g., Comrie 1989). While typological approaches are interested in the comparison of linguistic systems, pragmatic and interactional approaches compare language use. This ties back to linguistic anthropology, to the interest of Boas in both language and cultural practices, and to the later developments of linguistics inspired by ethnography, sociology, and discourse analysis in the 1970s. In the 1970s, an interest in action and practices characterized the linguistic approaches focusing on discourse and contexts of language use, using qualitative methods and fieldwork. Ethnography of speaking (Hymes 1972) has proposed an explicit comparative perspective with the SPEAKING model, offering a general list of categories to compare speech events. While much of this research tradition has been interested in ritualistic, ceremonial speech events that are unique to a certain culture (Bauman 1975; Duranti 1997), others have turned to everyday practices, like Frake’s study of how to ask for a drink in Subanun (1964). John Gumperz’ Interactional Sociolinguistics has focused on intercultural encounters in ordinary and institutional settings. At the heart of his theory of contextualization lies the claim that misunderstandings between members of different cultures arise from different contextualization conventions. Gumperz showed how prosodic contours, formulaic speech, and code-switching give rise to different inferences concerning, e.g., the type of action, speaker’s intentions, or emotional stance towards the interlocutor depending on cultural conventions (Gumperz 1982). This interest in culture-specific interpretations of action was also developed using speech-act theory. For instance, launching cross-cultural politeness research,
Chapter 2. Data and methods used in the study of OKAY across languages 31
Brown and Levinson (1978) compared English with Tzeltal and Tamil, and searched for universal face-concerns and linguistic strategies to manage them in different languages. Within cross-cultural pragmatics, Blum-Kulka and Olshtain (1984) and Blum-Kulka, House, and Kasper (1989) compared the realizations of speech acts like requests and apologies across different languages. At the same time, the usability of speech act theory for comparative research was criticized by linguistic anthropologists like Rosaldo (1982; see also Duranti 1988) questioning the applicability of Searle’s categories and typologies of speech acts to other cultures. Moreover, the focus on specific speech acts, rather than cultural practices in actual settings, led Blum-Kulka et al. (1989) to use survey and experimental methodologies, involving questionnaires, interviews, and discourse completion tasks, rather than fieldwork and participant observation. In anthropology, there have also been early comparative studies on embodied conduct. Examples include David Efron (1941), who compared the gestural behavior of eastern Jews and Italians in New York City – using sketches but also slow motion films – as well as Margaret Mead and Gregory Bateson, who engaged in cross-cultural studies of bodily practices, comparing ordinary parenting practices in New Guinea, Bali, and the US (such as bathing babies, in Bateson and Mead 1951; see Jacknis 1988). Later on, Edward Hall (1966) researched cultural differences in the ways people position their bodies and manage physical distance, and human ethology researchers sought to identify universal strategies of behavior through cross-cultural comparison (Eibl-Eibesfeldt 1979). Comparative studies within CA are both inspired by these traditions and propose a distinctive approach. Like studies in linguistic anthropology, CA research is interested in practices observed in naturalistic settings, but unlike many ethnographic studies, it makes systematic use of audio- and video-recordings in order to focus on the details of vocal and bodily conduct. Practices are studied by focusing on practical problems participants have: managing turn-taking, securing and repairing understanding, producing meaningful responses, etc. Rather than focusing on pre-defined singular actions, the focus is on sequential environments in which specific actions can be identified as locally designed for and interpreted by the co-participants. In this framework, the emic, rather than the etic, definition of action is fundamental, enabling a focus on problems, categories, and meanings that are intelligible and oriented to by the participants themselves. Thus, in contrast to other approaches in the social sciences and sociolinguistics, comparison in CA does not relate language use to external variables (like gender, age, social class, or ethnicity) and therefore does not formulate its analyses in terms of correlations. The conversation analytic approach, instead, focuses on those pragmatic and contextual properties that participants themselves index to be relevant within their discursive practices – that is, in an emic way. Instances are compared in
32
Arnulf Deppermann and Lorenza Mondada
order to discover what participants themselves treat as relevant differences, which matter for their conduct and their understanding – for instance asking which contextual features and which resources make a difference for the ways in which an item like OKAY is employed and understood. This comparative procedure leads to a refined and structured collection (Mondada 2005). It includes the identification of generic structures, subtypes, contextual variants, boundary cases (which shade into other practices), and deviant cases (Deppermann 2001, Chapter 6.3). In contrast to statistical procedures, deviant cases are not just outliers which can be discarded. On the contrary, they can be particularly valuable for refining the analysis (ten Have 1998; Schegloff 1968, 2009). They may falsify an analysis and be key for its revision; they may point to conditions for the use of a practice which remain unnoticed in the more routine cases; or they may be particularly strong warrants for the normative character of the practice, if participants sanction, explain, or otherwise account for what appear to be deviant cases. 2. Comparative approaches in conversation analysis One consequence of CA’s focus on how participants engage in local methods to solve fundamental organizational problems is an interest both in formats of actions characterizing specific languages and in important similarities across very different cultures and languages. For instance, in his study of repair-practices in Tai languages (Thailand), Michael Moerman (1977) was the first to test whether interactive practices which had been identified on the basis of American English data were also used in culturally and linguistically distant communities. Moerman’s results highlighted the relationship between basic interactional problems (like the organization of turn-taking and repair) and culture- and language-specific practices to deal with them: “although the problems are generic and the abilities apparently universal, the actual forms that interaction takes are shaped by and adapted to the particular resources that are locally available for their expression” (Sidnell 2009, 4). Further studies of repair have highlighted similar issues: repair of troubles of hearing as well as troubles of understanding are practices present in all languages, but the way they are implemented is specifically tied to the grammatical constraints and resources of specific languages. For instance, comparing self-repair practices in English and Japanese, Fox, Hayashi, and Jasperson (1996) observe that while Japanese verb inflections have the form of a whole syllable, have a well identified meaning, and are not subject to grammatical agreement, English verb inflections are more closely bound to the verb, do not have a unique meaning, and are subject to agreement. This results in Japanese verb-endings being more readily available for repair compared to those in English (Fox et al. 1996, 203).
Chapter 2. Data and methods used in the study of OKAY across languages 33
More recently, building on CA and anthropological linguistics, the Pragmatic Typology approach developed a focus on “the typology of systems of language use and the principles that shape them” (Dingemanse, Blythe, and Dirksmeyer 2018, 322). Studies in Pragmatic Typology seek to identify the use of different linguistic formats in different languages: they aim at explaining commonalities in terms of universal pragmatic concerns and constraints, whereas differences are accounted for through linguistic and cultural specificities associated with individual languages. Studies include typologically different and geographically remote language communities, and they aim for a variable-based comparison of conversational practices involving formal coding procedures (e.g., concerning gaze behavior, Rossano, Brown, and Levinson 2009; turn-taking, Stivers et al. 2009; question-answer systems, Enfield, Stivers, and Levinson 2010; other-initiated repair, Dingemanse, Torreira, and Enfield 2013; Dingemanse et al. 2015; answers to polar questions, Enfield et al. 2018; recruitments, Floyd, Rossi, and Enfield 2020). This approach requires researchers to use interactional data which conform to uniform sampling criteria, using only instances of “maximally informal social interaction in familiar settings between people who knew each other well” (Enfield et al. 2010, 2617). Projects of this kind use CA for the inductive development of coding categories and add linguistic (morphological) description. Results rest both on qualitative analysis and statistical evidence. In sum, since the early 2000s, there has been a growing body of work that has studied how basic interactional concerns and structures are managed and implemented in different languages and cultures. Concerning their methodological approach, we can distinguish several types of studies: a. Studies that test if and how interactional structures and practices that have been identified with respect to (American) English data in seminal studies are implemented in other languages and cultures (e.g., Moerman 1977, 1988). b. Edited volumes or issues of journals that collect a number of independent studies dealing with the realization of comparable interactional structures (Steensig and Drew 2008 on questions; Enfield and Stivers 2007 on person reference; Margutti et al. 2018 on invitations) and linguistic formats (Heinemann and Koivisto 2016b on change-of-state tokens; Auer and Maschler 2016 on nu in different Germanic languages; Sorjonen, Raevaara, and Couper-Kuhlen 2017 on imperatives). In these collective books, each study uses its own specific set of data, and has particular research questions and specific analytic angles; comparisons between the individual studies, and discussions about commonalities and differences in terms of their results, are elaborated by the editors of the volumes.
34
Arnulf Deppermann and Lorenza Mondada
c. Jointly designed projects with a common research focus, e.g., on self-repair (Fox et al. 1996, 2009), repair initiation with huh (Dingemanse et al. 2013), and second assessments (Sidnell and Enfield 2012). These projects have taken a qualitative approach with the aim of comparing how linguistic resources in different languages are used to deal with the same interactional concerns. As shown by these examples, comparative studies in CA deal with very different objects: some studies start from a linguistic form (like change-of-state tokens, a specific particle, or the repair-initiator huh), whereas others start from an interactional practice within a specific sequential environment (like different types of repair, responses to polar questions, etc.). Moreover, some studies try to restrict and control the data set used (like informal conversations), although strictly speaking corpora set up specifically for comparative analyses are still very rare, and references to “informal conversations” remain abstract and cover a variety of activities, participation frameworks, and contexts. Other studies do not constrain the data examined but rather the sequential environments and types of actions. Finally, most studies proceed on the basis of an analysis of “collections” in Schegloff ’s (2009) sense. Some have introduced coding systems as a basis to secure comparability and statistically valid results, although this remains controversial from an emic perspective. 3. The conversation analytic methodology used in this book on OKAY CA is an approach to the study of social interaction with a distinct focus on situated action, and it has developed a distinctive research methodology to analyze situated action (see overviews in ten Have 1998; Hutchby and Wooffitt 1998; Deppermann 2001; Sidnell 2013). The main interest of CA is to identify practices that members use to perform social actions through which knowledge is conveyed and negotiated and through which interaction is organized and relationships are managed. “A ‘practice’ is any feature of the design of a turn in a sequence that (i) has a distinctive character, (ii) has specific locations within a turn or sequence, and (iii) is distinctive in its consequences for the nature or the meaning of the action that the turn implements” (Heritage 2010, 214; Selting 2016). We will exemplify the notion of practice with respect to the particle OH in English, which has been studied in depth by John Heritage (1984a, 1998, 2002, 2006, 2010, 2018a) and which is quite similar to OKAY in many ways. A practice is an observable, routine way of dealing with a certain pragmatic concern (e.g., claiming understanding). Practices are implemented by using linguistic formats (e.g., OH, with different prosodic contours, Local 1996; Barth-Weingarten, Couper-Kuhlen, and Deppermann 2020) and various other co-occurring resources, such as gaze
Chapter 2. Data and methods used in the study of OKAY across languages 35
(e.g., at prior speaker), gesture (e.g., upward nod), body posture, facial expression (e.g., eyebrow raise), and object manipulation, whose temporally organized configuration can constitute complex multimodal gestalts (Mondada 2014a). Practices are tied to a certain sequential context (e.g., elicited or volunteered informing, other-repair) and are used to perform certain actions in that context (e.g., claiming a change of the speaker’s epistemic state). Practices are generic structures that have context-free properties, which hold across the different local occasions at which they are deployed. At the same time, they are irremediably indexical and flexible, which means that they are always implemented with respect to the particulars of their specific, “unique” local context (Schegloff 2009). Building on these general, formal properties of practices, and using them as methodological guidelines, the methodology of CA allows us to identify new practices inductively. It is designed to be a methodology of discovery. The starting point for the identification and analysis of a practice can be a certain form (e.g., a token like OKAY), a certain sequential environment (e.g., how to respond to an informing), or a certain interactional concern (e.g., claiming a revised understanding). However, it is important to note that the sheer identification of instances, e.g., all instances of OKAY that can be found by searching a corpus, does not yet represent a “collection” of instances of a practice in a given sequential environment, but only a list of candidate items. Rather, a collection (in the technical sense of Schegloff 2009) is based on an analysis of all candidate items for the specific actions they do in specific sequential environments. A “collection” is the result of both an in-depth sequential and linguistic analysis of each candidate instance of the practice, and a comparison of instances, identifying a common formal, sequential, and action pattern (Deppermann 2001, Chapter 6.2; Mondada 2005). Comparing can be seen as a fundamental operation within all conversation analytic studies (Schegloff 2009; Arminen 2010): in order for a researcher to build a collection, and to decide whether an instance belongs to it, a comparative analysis between tokens, their sequential environments, and the action(s) they achieve has to be made. Moreover, comparisons have characterized CA from its early stages, not between languages and cultures, but between contexts. For instance, a fundamental comparison between the large, rather unrestricted, options available in informal conversation and the constrained and specialized options characterizing institutional settings has prompted a reflection on the theoretical importance of “conversation” as distinct from other forms of “talk-in-interaction, ‘namely in institutional settings (Drew and Heritage 1992; Drew 2003). Specificities of the organization of turn-taking and the organization of sequences have also been identified by distinguishing telephone (and other technologically mediated) vs. face-to-face interactions, dyadic vs. multi-party interactions (and other types of participation frameworks), as well as different contexts of activity for the same
36
Arnulf Deppermann and Lorenza Mondada
action (e.g., giving advice, delivering bad news, asking questions, requesting). Moreover, comparisons between the same type of action at different moments in history (Clayman et al. 2006) and along longitudinal processes (Pekarek Doehler, Wagner, and Gonzalez-Martinez 2018) have been proposed. The study of OKAY across 13 languages (Danish, English, German, Swedish, Estonian, Finnish, Polish, French, Italian, Brazilian Portuguese, Mandarin, Japanese, and Korean) enables both an exemplary implementation of CA analysis and some distinct challenges. Although referred to here with the generic form OKAY (in capitals), the particle is realized in many different ways (e.g., okei, occhei, oké, okra) in specific languages. Whereas most often these forms still show phonetic similarities with English okay, they cannot be automatically treated as identical. This shows the importance of detailed transcripts of the forms studied. Furthermore, OKAY is used in highly diverse contexts, and this creates the need to specify in detail the sequential environments and actions concerned. The following properties of occurrences of OKAY have been attended to in the analyses gathered in this volume: – type of interaction, activity, and participation framework: the social setting and type of interaction, the participation framework including the number of parties and participants to an encounter, their situated roles (including official status, epistemic and deontic rights, being acquainted vs. unacquainted); – sequential environment: the position of OKAY within interactional sequences, e.g., in responses to stories, informings, and repairs (Chapter 3, 7, 12), closings of sequences and encounters, and transitions between activities (Chapter 4, 8, 9, 10); – actions performed (or co-constituted) by OKAY; – position within a turn: stand-alone uses of OKAY, constituting the whole turn; OKAY as a TCU of its own within a larger turn; OKAY in TCU- and turn-initial, turn-medial, and turn-final positions (see Chapter 5, 9); – phonetic and prosodic realization of OKAY, including the following aspects: realization as monophtong vs. diphtong, morphophonetic variation (like mkay, kay, okidoki); stress, contour, loudness, onset and pitch register, creak, syllable duration, aspiration/inserted laugh particles; in addition to auditory analysis, instrumental analyses using PRAAT were performed for some studies (e.g., Chapter 5, 6); – part of speech: OKAY as a particle vs. as noun, adverb, or adjective. For this book, only particle uses have been studied in detail (see below); – co-occurrence with other particles (like ah okay, well okay) and reduplication (okay okay, see Chapter 8);
Chapter 2. Data and methods used in the study of OKAY across languages 37
– multimodal realization: OKAY as integrated in a multimodal gestalt, together with various bodily resources, like gaze and nodding (see Chapter 12), as well as body movements, postural changes, and object manipulations (see Chapter 10, 11). As a consequence of the emphasis of CA on the situated details in which actions, turns, or forms are produced, and as a consequence of the challenges of OKAY across very different languages, transcription is a fundamental methodological practice. It lies at the core of not only the preparation of the data, but also the analysis itself. This book adopts the conventions of Jefferson (2004) for transcribing talk, elements of the GAT 2 transcription system for prosody (Selting et al. 2011), the Leipzig glossing system (Lehmann 1982; Croft 2003), and the conventions of Mondada (2018a) for transcribing embodied conduct. The Jeffersonian conventions (Jefferson 2004) insist on the temporality of talk in its diverse manifestations (Hepburn and Bolden 2017). This enables us to capture the fundamental organizational aspects of turn-taking (gaps and overlaps) and of the formatting of actions within sequences and wider sequential environments, revealing how they relate to each other. Not only temporal but other detailed formatting elements of the production of tokens are considered in these conventions, such as intonation patterns. In the transcripts of this volume, these prosodic notations have been expanded by taking into consideration the GAT 2 system (Selting et al. 2011) when suitable for the language transcribed. This enables us to specify the intonational contours of OKAY, which might be fundamental for the identification and differentiation of the actions accomplished by the particle. Moreover, since the project involves 13 different languages, glosses and translations of the original data constitute an essential part of the transcripts. This is essential for readability but also has immediate analytical consequences. Using glossing categories for different languages is a practical necessity and at the same time presupposes a problematic assumption, as discussed by Haspelmath (2010), who points at the tendency of cross-linguistic glossing categories towards “categorial universalism.” In the transcripts, the glosses are a pragmatic instrument for making data in their original language understandable (rather than an analytical basis for comparison). In this sense, the original, the glosses, and the translation in English together build the intelligibility of the data. They are meant to support the reading of the original transcript, and not to autonomize the reading of the translation. In this sense, the translation provided is not an equivalent of the original, but rather a further gloss that helps readers understand the original. This is why in many cases the translation is quite close to the original, providing for a rather literal version of it, instead of seeking an idiomatic rendering of the original. Moreover, in
38
Arnulf Deppermann and Lorenza Mondada
order not to impose any translation or equivalent on the specific forms of OKAY in different languages, the particle has been glossed but not translated. An important contribution of the volume is the fact that the comparative analysis includes elements not only of the talk but also of the embodied conduct of the participants, when face-to-face situations are involved. This invites a multimodal comparative analysis that considers talk and the body together. The conventions adopted here are those of Mondada (2018a), which focus on the specific temporality of multiple embodied conducts, ordered both sequentially and simultaneously. This mode of transcription can provide hints about OKAY as belonging to possible multimodal gestalts (Mondada 2014a), that is, configurations in which embodied resources such as gesture, head movements, body postures, body movements, etc. are integral parts of the intelligibility of OKAY, in the same way as other specific phonetic, prosodic, and turn-constructional features. Considering these aspects together opens up new perspectives for the ways in which comparative analysis in a sequential-interactional perspective can be done and enriched – never considering a formal token alone, but instead its sequentially meaningful combination with other multimodal properties. This shows both the locally specific situatedness of the uses and the conditions of their cross-contextual comparability. 4. Data used in the present studies of OKAY This book gathers CA studies of OKAY in 13 languages. Following the CA analytic mentality and procedures, all languages have been documented using naturalistic data, audio/video-recorded in their social context, without any other intervention of the researchers. Two kinds of studies have been generated on this basis. First and foremost, transversal studies (Chapter 3, 4) using data from all the corpora, illustrate the sequential environments in which OKAY can feature in systematic ways throughout the languages studied. Claiming understanding (Chapter 3) and managing closings and transitions (Chapter 4) have been identified as recurring functions in which OKAY is pervasively used in all languages studied. In these chapters, a diverse range of contexts is illustrated, both ordinary and institutional, including phone and face-to-face encounters. Whereas the type of data – the social contexts in which they have been recorded – has not been constrained, the sequential environments in which OKAY features have been rigorously controlled in the analyses. This approach enables us to demonstrate the variety of uses of OKAY as well as the methodic ways in which it is used in precisely defined sequential environments across the languages.
Chapter 2. Data and methods used in the study of OKAY across languages 39
Second, in-depth studies (Chapter 5–12) have focused on specific languages (American English, Chapter 5; Danish, Chapter 6; Finnish, Chapter 7; Korean and Japanese, Chapter 8; Brazilian Portuguese, Chapter 9; French and Italian, Chapter 10; Estonian and Polish, Chapter 11; German, Chapter 12) and on specific formal aspects (such as prosody, Chapter 5, 6; repetitions, Chapter 8; multimodality, Chapter 10, 12) or specific sequential phenomena (such as responses to informings, Chapter 7; alignment in responses, Chapter 9; action transition, Chapter 10; directive sequences, Chapter 11; displays of understanding, Chapter 12). Some chapters focus on a certain activity type, others use data from a wider range of contexts. Table 1 at the end of the chapter presents an overview of the corpora used in the book, both in the initial cross-linguistic chapters and in the following in-depth analyses. It shows the diversity of the corpora as well as recurrent settings across the languages (such as telephone calls, informal conversations, and institutional encounters). Although the number of occurrences of OKAY indicated in Table 1 does not differentiate between specific uses, and must therefore be read with caution, the numbers shows the variability of OKAY not only across languages (in some languages only a limited number of occurrences have been found, in striking contrast to other languages) but also across different and similar settings (e.g., a phone conversation in a language can contain an abundant number of OKAYs, whereas another phone call in the same language can exhibit only few occurrences). These indicative numbers not only suggest that OKAY is not used in the same range and frequency across languages, but also that its use is sensitive to variations in types of activities, as well as types of speakers (as far as age, social class, and education are concerned – although these variables have not been collected in the meta-data of most corpora and are not used in CA analyses). The way this book has been conceived is both as a collective project on OKAY in social interaction and as a collection of specific studies on OKAY in interactions in different languages and socio-cultural contexts. In a first step, a survey was made by all authors about the occurrences of OKAY found in corpora representing a diversity of languages. Each language team prepared an overview of OKAY in their data and sampled recurrent uses. These first analyses gave an overview of the diversity of sequential environments in which OKAY is used in different languages. At two meetings in Mannheim and Helsinki in 2017, prototypical examples for each language were presented and analyzed by all contributors. In order to enhance the comparability of analyses and findings, it was decided to focus on OKAYs that are used as particles, and on the interactional work they achieve (for turn-taking, sequence and topic organization, managing understanding, closing, etc.). The resulting focus is thus on stand-alone OKAYs that constitute a whole turn and on OKAYs that occur as a prosodic, mostly initial, unit
40 Arnulf Deppermann and Lorenza Mondada
of their own, within a longer turn. Additionally, two chapters include a discussion of OKAYs used as tags in turn-final position (Chapter 5, 9). Syntactically integrated uses as predicative adjective (“that’s okay with me.”), adverb (“it looks okay.”), or noun (“they gave their okay.”) have not been considered. On this basis, each language team contributed extracts for the two most pervasive sets of uses, receipts and claims of understanding (Chapter 3) and closings and transitions (Chapter 4), which were identified as recurrent sequential contexts across languages. After convening on shared transcription conventions (especially for orthographic and prosodic representations of the OKAY tokens), terminology, and analytic foci, researchers for each language also developed collections (Chapter 5–12) concerning at least one language and a specific use of OKAY. 5. Some challenges and limitations In this book, we present studies that cover a maximally wide range of the uses of OKAY in different languages in all sorts of naturally occurring interactions. We have aimed at an exploration of the phenomenological richness of the uses of OKAY across languages, covering a wide range of cultural contexts, interaction types, and participation frameworks. This approach has been preferred over an attempt to pin down differences across languages, as is often the case with comparative projects. Our choice has been informed both by methodological and empirical reasons. A comparison of instances of OKAY whose results would unequivocally speak for differences between languages and cultures would require that all other interactional dimensions that could account for differences in the use of OKAY were kept constant. In other words, we would need to have corpora of data that were uniform in terms of the distribution of interaction types, participation frameworks, and sequential environments, within comparable contexts and activities. Such uniformity can only be achieved if a highly restrictive sampling procedure is used. On the one hand, this would require collecting new data for most of the languages to be studied. On the other hand, this would necessarily rule out much variation of interactional practices in the languages studied. Most importantly, institutional interaction could be covered only to a very small degree, and would run the risk of adopting a very general (not at all emic) typification of institutional contexts (this critical argument applies to “informal conversation,” too, a description often adopted for comparative samples as implying an homogeneous set of data, yet actually referring to a diversity of contexts, activities, number of participants, and participation frameworks). As this book shows, OKAY is routinely used in many institutional contexts, in which it accomplishes specific interactional work, which depends on
Chapter 2. Data and methods used in the study of OKAY across languages 41
the specificity of the activity, the institutionality, and the material ecology of these contexts. Furthermore, a control of the types of contexts documented in the data across languages would run the risk that only a restricted set of languages and cultures would conform to them, or that only some activities in particular languages or cultures would be considered – probably the more standardized and attuned to a global socio-economy – with the loss of many other more specific contexts. This could lead to data that are rather artificial, marginal, and not representative of the discursive practices in a given language. Thus, while comparability of data could be maximized and controlled, the ways in which the data could be said to cover the varieties of naturally occurring interaction in the individual languages and cultures would be severely restricted, with the effect that only small segments of interactive practices in an individual language would be studied. In order to cover phenomenological variation most comprehensively, the contributors to this book have drawn from a rich variety of corpora that exist for the languages under study. Yet since these corpora are of different sizes and comprise different types of interactions, activities, and participations frameworks, a direct comparison concerning differences in the distribution of forms and uses in the different languages is not possible. The finding that a certain use does not occur in the corpora available for one language could suggest that this use is indeed rare in that language. However, it does not provide robust negative evidence: we cannot infer that a use does not exist in a language just because it does not occur in the corpora available for that language. The same applies for relative differences in distribution. Because of the incommensurability of the composition of corpora, but also because we do not know how interaction types, sequence types, and other pragmatic factors are distributed in social interaction within the individual languages and cultures, any differences that we could find in the use of OKAY would be very hard to assess. The heterogeneity of the available corpora thus limits their comparability (cf. Couper-Kuhlen 2019b). For future research that aims for comparison across languages and cultures, there is a need for large, diversified, and open-access corpora of naturally occurring multimodal interaction. Such corpora are currently only available for a small number of languages, notably Danish (Corpus Samtalebank, https://samtalebank.talkbank.org), French (Corpus CLAPI, http://clapi.ish-lyon. cnrs.fr/), and German (corpus FOLK, https://dgd.ids-mannheim.de/dgd/pragdb. dgd_extern.welcome). The use of a wide variety of audio- and video-recordings from different languages has led to a robust description of the spectrum of current uses of OKAY in this book. In our analyses, we have found that the core interactional jobs done by OKAY and the range of its uses are strikingly similar across languages. This is represented in Chapters 3 and 4, in which we describe the most pervasive uses
42
Arnulf Deppermann and Lorenza Mondada
of OKAY, which are recurrent in the languages considered. Thus, our inductive approach, which involves a methodology that strongly emphasizes and allows for the representation of contextual and linguistic particularities of the use of OKAY in the individual languages, has yielded the conclusion that there is a large area of commonality between the different languages concerning the use of OKAY. This is not to be understood as a claim to universalism – the number of languages is much too small for such a claim. However, it shows that the appropriation of OKAY in very different languages and contexts has led to a range of largely similar practices to an astonishingly high degree. Our findings highlight the importance of a holistic analysis of OKAY turns that includes prosody and the larger sequential and multimodal context (see Chapters 5–12). This has allowed us to delimit individual uses of OKAY in a more systematic way than has been done before, showing how the realization of OKAY and its context are constitutive for different uses. The inductive approach has also made it possible to identify uses that were not documented previously, for example the use of OKAY as a continuer or for “qualified acceptance.” Another area in which new observations have begun to emerge concerns the relationship of OKAY to other particles in a given language (see Chapters 1, 7). OKAY coexists with other particles that do very similar or even equivalent interactional work in the specific languages. The way OKAY progressively integrates into the system of particles of each language, and eventually changes their distribution over history, is a topic that could be fruitfully expanded by future research. This would require systematic knowledge of the system of response particles in every language – something that does not exist for most of the languages studied in this book. Another related line of future investigation, which is not addressed in this volume, concerns the diachronic development of different uses of OKAY and their spread in different languages. Chapter 5 suggests avenues for doing this for English, and it would be exciting to see such work pursued for OKAY in other languages as well (see also Chapters 1, 7).
Chapter 2. Data and methods used in the study of OKAY across languages 43 Table 1. Inventory of the corpora used in this book and their characteristics Number of OKAYs in that transcribed recording
Reference to the examples in the book (chapter/ excerpt number)
Name of the corpus
Language
Place of the recording
Date of the recording
Type of interaction
Type of recording: Audio or video
Length of the recording that has been transcribed
NB 028
American English
California, USA
1960s
Audio
44 min 31 s
8
Chapter 4/ Excerpt 6
Joyce & Stan
American English
California, USA
1960s
Audio
8 min 28 s
11
Chapter 5/ Excerpt 4, 13
HGII
American English
USA
1960s
Audio
20 min
17
Chapter 5/ Excerpt 1, 5, 12, 17
Call Home 4544
American English
USA
1994–1997
Audio
25 min 34 s
1
Chapter 5/ Excerpt 9
Call Friend 4984 American English
USA
1996–1997
Audio
30 min
26
Chapter 5/ Excerpt 3, 6
Call Friend 6239 American English
USA
1996–1997
Informal telephone conversation Informal telephone conversation Informal telephone conversation Informal telephone conversation Informal telephone conversation Informal telephone conversation
Audio
30 min
9
Call Friend 6899 American English
USA
1996–1997
Informal telephone conversation
Audio
30 min
25
Chapter 3/ Excerpt 11 Chapter 5/ Excerpt 8 Chapter 5/ Excerpt 10
(continued)
44 Arnulf Deppermann and Lorenza Mondada
Number of OKAYs in that transcribed recording
Reference to the examples in the book (chapter/ excerpt number)
Name of the corpus
Language
Place of the recording
Date of the recording
Type of interaction
Type of recording: Audio or video
Length of the recording that has been transcribed
Kamunsky 3
American English
USA
1960s
Informal telephone conversation
Audio
13 min 27 s
42
Two girls
American English
USA
1960s
Audio
16 min 25 s
5
Debbie & Shelley American English
USA
1960s
Audio
7 min 37 s
8
Chapter 5/ Excerpt 15, 16
Farmhouse
American English
Colorado, USA
1998
Video
38 min 36 s
20
Chapter 5/ Excerpt 2, 7, 18
Samtalebank
Danish
Denmark
2009
Audio
6 min 48 s
1
HB
Danish
Denmark
2015
Video
6 min
6
ADK
Danish
Denmark
2002
Audio
unknown
unknown
ADK
Danish
Denmark
2002
Audio
unknown
unknown
AULing
Danish
Denmark
2003
Video
23 min
10
Samtalebank
Danish
Denmark
2009
Informal telephone conversation Informal telephone conversation Informal face-to-face conversation Private phone call Dietician-client interaction Private phone call Private phone call Everyday conversation Everyday conversation
Video
44 min
21
Chapter 3/ Excerpt 16 Chapter 4/ Excerpt 11 Chapter 6/ Excerpt 1, 2 Chapter 6/ Excerpt 2 Chapter 6/ Excerpt 3 Chapter 6/ Excerpt 4
Chapter 4/ Excerpt 4 Chapter 5/ Excerpt 11 Chapter 5/ Excerpt 14
(continued)
Chapter 2. Data and methods used in the study of OKAY across languages 45
Name of the corpus
Language
Place of the recording
Date of the recording
Type of interaction
Samtalebank
Danish
Denmark
2009
AULing
Danish
Denmark
1990s
Samtalebank
Danish
Denmark
2009
Samtalebank
Danish
Denmark
2009
Samtalebank
Danish
Denmark
2009
AULing
Danish
Denmark
2015
AULing
Danish
Denmark
2008
Everyday Video conversation Private phone Audio call Private phone Audio call Everyday Video conversation Everyday Video conversation Everyday Video conversation Shop encounter Video
KTA
Finnish
Helsinki, Finland 1988–1989
KTA2
Finnish
Western Finland
KTA3
Finnish
Helsinki, Finland 2007–2008
1997
Everyday telephone conversation Everyday telephone conversation Everyday telephone conversation
Type of recording: Audio or video
Length of the recording that has been transcribed
Number of OKAYs in that transcribed recording
50 min
9
7 min 22 s
2
6 min 44 s
1
43 min
10
30 min 35 s
12
Reference to the examples in the book (chapter/ excerpt number) Chapter 6/ Excerpt 5 Chapter 6/ Excerpt 6 Chapter 6/ Excerpt 7 Chapter 6/ Excerpt 8 Chapter 6/ Excerpt 9 Chapter 6/ Excerpt 10 Chapter 6/ Excerpt 11 –
unknown
unknown
2 min 15 s
1
Audio
1 h 30 min
25
Audio
1 h 55 min
47
Chapter 7/ Excerpt 1
Audio
1 h 49 min
73
Chapter 7/ Excerpt 2, 4, 6, 10, 11, 15 Chapter 4/Excerpt 23
(continued)
46 Arnulf Deppermann and Lorenza Mondada
Number of OKAYs in that transcribed recording
Type of interaction
Type of recording: Audio or video
Length of the recording that has been transcribed
Everyday telephone conversation
Video
1 h 10 min
99
Helsinki, Finland 2015
Student union
Video
6 h 45 min
114
Helsinki, Finland 2017
Radio interview Audio
13 min
3
Japanese
USA; Japan
1996
Audio
3h
2
Japanese
Tokyo, Japan
2017
Video
50 h
6
Architect office
Japanese
Tokyo, Japan
Sep 2006
Telephone conversation Ordinary conversation; Various institutional settings Meeting
Video
1h
2
Sushi bar encounters
Japanese
Osaka, Japan
Sep 2005
Restaurant encounter
Video
7h
2
Guitar lesson
Japanese
Tokyo, Japan
Jan 2018
Guitar tutoring Video
2h
21
Name of the corpus
Language
Place of the recording
Date of the recording
Repair
Finnish
Helsinki, Finland Mar 2011
Pekkanen
Finnish
Radio
Finnish
Call Home Corpus Corpus of Everyday Japanese Conversation
Reference to the examples in the book (chapter/ excerpt number) Chapter 7/ Excerpt 3 Chapter 7/ Excerpt 14 Chapter 3/ Excerpt 3 Chapter 7/ Excerpt 5, 13 Chapter 7/ Excerpt 12 Chapter 3/ Excerpt 4 Chapter 8/ Excerpt 9, 10
Chapter 8/ Excerpt 8 Chapter 4/ Excerpt 8 Chapter 8/ Excerpt 2 Chapter 8/ Excerpt 5
(continued)
Chapter 2. Data and methods used in the study of OKAY across languages 47
Length of the recording that has been transcribed
Number of OKAYs in that transcribed recording
Name of the corpus
Language
Place of the recording
Date of the recording
Type of interaction
Type of recording: Audio or video
MC3
Japanese
Tokyo, Japan
May 2016
Audio
1
Linguistic Data Corsortium CallFriend Korean KGSA
Korean
USA; Canada
1996
Telephone conversation Ordinary telephone conversation
Audio
19
Korean
USA
4
Korean
USA; Korea
ER
Korean
Korea
Oct 2011
ECOCLIN
Brazilian Portuguese
City in Southern Brazil
Jan 2013
HMF ACONGEN
Brazilian Portuguese
City in Southern Brazil
Feb 2014
LAVA JATO
Brazilian Portuguese
Brasília, Brazil
Mar 2016
Student group meeting Ordinary telephone conversation Medical encounter Ultrasound exams (doctor-patient) Genetic counseling (doctor-patient) Tapped telephone interactions among politicians
Video
Korean Telephone Calls
Apr 2005, Aug 2006 Jan 2003– Dec 2004
Audio
5h
Reference to the examples in the book (chapter/ excerpt number)
9
Video
1
Audio
2
Audio
53
Audio
12
(continued)
48 Arnulf Deppermann and Lorenza Mondada Number of OKAYs in that transcribed recording
Reference to the examples in the book (chapter/ excerpt number)
41 h 30 min
688
Audio
10 min
12
Chapter 9/ Excerpt 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 Chapter 3/Excerpt 14
Video
2 h 50 min
68
Chapter 4/ Excerpt 3, 15
Video
37 min
13
Chapter 10/ Excerpt 3, 4
Video
24 min
6
Chapter 10/ Excerpt 8
Video
2h
85
Chapter 10/ Excerpt 1
Type of recording: Audio or video
Length of the recording that has been transcribed
Jan 2007–Dec Health helpline 2009 telephone interactions
Audio
Lyon, France
2003
French
Paris, France
2016
AGRO_Inra
French
Nancy, France
2001
CAB4
French
Lyon, France
2008
CAB7
French
Lyon, France
2008
Name of the corpus
Language
Place of the recording
Date of the recording
DISK
Brazilian Portuguese
Brasília, Brazil
CALL_Jean2
French
FRO_Par1007
Type of interaction
Calls for help to service providers Cheese shop encounter, 1–3 customers with 1 salesperson (27 encounters, 3 salespersons) Work meeting between 2 agronomists and 1 computer scientist Grassroots political meetings, ca. 100 part. Grassroots political meetings, ca. 100 part.
(continued)
Chapter 2. Data and methods used in the study of OKAY across languages 49 Number of OKAYs in that transcribed recording
Reference to the examples in the book (chapter/ excerpt number)
Name of the corpus
Language
Place of the recording
Date of the recording
Type of interaction
Type of recording: Audio or video
Length of the recording that has been transcribed
CAB12
French
Lyon, France
2008
Video
1 h 45 min
19
Business meeting Italian
Milan, Italy
2013
Video
4 h 20 min
more than 50
Guided Tour Adults Guided Tour School Cold calls
Italian
Naples, Italy
Feb 2009
Video
3 h 20 min
14
Italian
Naples, Italy
Feb 2009
Video
2 h 30 min
unknown
Italian
2011
Audio
37 min 20 s
24
Driving Lessons
Italian
Caller: Olten, Switzerland; Called: Whole Switzerland Ticino, Switzerland
Video
7h
882
Chapter 4/ Excerpt 10
Grocery shopping
Italian
Ticino, Switzerland
2004
Video
1h
16
Chapter 4/ Excerpt 16
Merilin
Estonian
Tallinn, Estonia
1998–1999
Grassroots political meetings, ca. 100 part. Business meeting, 5 part. Guided tour, 8 participants Guided tour, 24 participants Outbound call from bank employee to existing clients Driving lesson 7 lessons, 2 in-car participants (2 instructors, 7 trainees) Grocery shopping, 2 customers Phone call
Audio
30 s (all calls 2 h 10 min 50 s)
1 (53)
Chapter 3/ Excerpt 6
2009–2010
Chapter 10/ Excerpt 7 Chapter 10/ Excerpt 2, 5, 6 Chapter 10/ Excerpt 10 Chapter 10/ Excerpt 9 Chapter 3/ Excerpt 10
(continued)
50
Arnulf Deppermann and Lorenza Mondada
Reference to the examples in the book (chapter/ excerpt number)
Name of the corpus
Language
Place of the recording
Date of the recording
Type of interaction
Type of recording: Audio or video
Length of the recording that has been transcribed
Number of OKAYs in that transcribed recording
Sõnnik
Estonian
Läänemaa, Estonia
2014
Physical labor
Video
3 (ca. 20)
Chapter 11/ Excerpt 3, 5
Kelder
Estonian
Tallinn, Estonia
1998–1999
Phone call
Audio
6 min transcr. (5 h 30 min recorded) 5 min (all calls 2 h 15 min 22 s)
4 (25)
Pille
Estonian
Tallinn, Estonia
1998–1999
Informal phone Audio call
Gądki
Polish
Gądki, Poland
2009
Phone call
Audio
3 min (all calls 3 h 13 min 23 s) 9 min 28 s
Chapter 4/ Excerpt 2 Chapter 11/ Excerpt 1 Chapter 4/ Excerpt 22
Gądki
Polish
Gądki, Poland
2009
Phone call
Audio
2 min 30 s
6
Gądki
Polish
Gądki, Poland
2009
Phone call
Audio
4 min 6 s
3
Gądki
Polish
Gądki, Poland
2009
Phone call
Audio
1 min 15 s
1
Komorze
Polish
Komorze, Poland 2011
Professional talk Video
7 h rec.
unknown
Komorze
Polish
Komorze, Poland 2011
Professional talk Video
4 h 30 min rec.
unknown
I. Studio
Polish
Poznań, Poland 2016
Talk at work
3 min transcr. (35 min rec.)
2
Video
4 (112) 9
Chapter 3/ Excerpt 9 Chapter 3/ Excerpt 15 Chapter 11/ Excerpt 2 Chapter 4/ Excerpt 21 Chapter 11/ Excerpt 4 Chapter 11/ Excerpt 6 (6–6.4) Chapter 4/ Excerpt 18
(continued)
Chapter 2. Data and methods used in the study of OKAY across languages 51
Name of the corpus
Language
Place of the recording
Date of the recording
German FOLK subcorpus (specific videos + telephone)
German-speaking 2008–2014 countries
Betz’s corpus
German
2004–2016
PECII
German
Mandarin Multimodal Corpus
Mandarin Chinese
Different regions in Germany, incl. transatlantic calls from USA to Germany Palatine region, Germany Beijing, China; Shanghai, China; Edmonton, Canada
Bertils ARUU
Swedish
Sweden
2016
JohanssonSyUU
Swedish
Sweden
2015
2016 May 2008– Oct 2016
Length of the recording that has been transcribed
Number of OKAYs in that transcribed recording
Reference to the examples in the book (chapter/ excerpt number)
Type of interaction
Type of recording: Audio or video
Informal conversation and institutional interaction, 2–6 part. Informal conversation (e.g., game playing, meals), 2–5 part. Driving
Audio and 26 h 7 min video
697
Chapter 3/ Excerpt 13, 17 Chapter 4/ Excerpt 13
Audio and 4 h 35 min video
97
Chapter 3/ Excerpt 1
Video
1
Tea/coffee conversation, dinner-table conversation, playing card games Breast feeding support help line, 2 part. Handicraft circle, ca. 8 part.
Audio and 19 h video
Chapter 4/ Excerpt 19 Chapter 3/ Excerpt 2, 8; Chapter 4/ Excerpt 17
1 min 27 s
57
Audio
22 min 24 s (1 call)
unknown
Chapter 3/ Excerpt 12
Video
3 h 10 min
unknown
Chapter 4/ Excerpt 12
Generic sequential uses of OKAY across languages
Chapter 3
OKAY in responding and claiming understanding Emma Betz and Arnulf Deppermann
University of Waterloo / Leibniz-Institut für Deutsche Sprache, Mannheim With contributions from Elizabeth Couper-Kuhlen, Elwys De Stefani, Katariina Harjunpää, Kaoru Hayano, Henrike Helmer, Leelo Keevallik, Mary Shin Kim, Stephanie Hyeri Kim, Aino Koivisto, Satomi Kuroshima, Seung-Hee Lee, Xiaoting Li, Anna Lindström, Lorenza Mondada, Ana Cristina Ostermann, Søren Sandager Sørensen, Marja-Leena Sorjonen, Jakob Steensig, and Matylda Weidner OKAY originates from English, but it is increasingly used across languages. This chapter presents data from 13 languages, illustrating the spectrum of possible uses of OKAY in responding and claiming understanding in contexts of informings. Drawing on a wide range of interaction types from both informal and institutional contexts, including those crucially involving embodied practices, we show how OKAY can be used to (i) claim sufficient understanding, (ii) mark understanding of the prior informing as preliminary or not complete, and (iii) index discrepancy of expectation. Keywords: informing, second position, third position, closing, understanding, intersubjectivity, newsmark, continuer, (discrepancy of) expectation, prosody
1. Introduction The uses of OKAY1 discussed in this chapter form a family of uses whose members are related in various ways to the management of understanding in interaction, specifically in contexts of informings.2 OKAY is indexical by nature: Its use for indexing understanding of a prior turn is tied to specific sequential and turn positions (cf. Sidnell 2014; Deppermann 2015). When OKAY is used in response to an 1. We use the all-capitals spelling of the particle to refer to the item in general, not to a particular prosodic realization or language-specific variant. When quoting a particular instance of OKAY, we display it in italics in its specific realization. 2. Chapter 4 also examines OKAY in other contexts, such as request and proposal sequences. https://doi.org/10.1075/slsi.34.03bet © 2021 John Benjamins Publishing Company
56
Emma Betz and Arnulf Deppermann
interlocutor’s turn,3 it can embody some sort of understanding of that turn. Sacks (1992, vol. 2, 141) distinguishes claiming from demonstrating understanding (cf. Schegloff 1979). The latter involves giving proof or evidence of an understanding gained, for example by formulating an inference or by executing a fitting or mandated next action. The former can be performed by repetitions, nodding, and various response particles like OKAY. Being merely a claim to understanding, response practices do not give any evidence of the precise kind of understanding gained, and they therefore do not allow the producer of the first action to assess whether the claimed understanding is correct or not. OKAY, like other verbal resources for claiming understanding, can also be used for other actions involved in managing understanding and intersubjectivity in interaction, such as initiating repair (Rossi 2020), projecting the continuation of the interlocutor’s turn (Schegloff 1982) or merely registering a piece of information, while not yet having understood its relevance to the larger topical, argumentative, or pragmatic context (Betz et al. 2013). Prosody plays a major role in distinguishing between these uses (cf. Thompson, Fox, and Couper-Kuhlen 2015 on the role of prosody in responses). Responsive OKAY can receipt information as understood and sufficient (Section 2), thus allowing for sequence closing and a move to pending or next matters (see Chapter 4). OKAY, however, can also mark prior information as preliminary or not yet complete (Section 3) and thus invite continuation. It can furthermore communicate that a prior turn was counter to expectation, thus engendering elaborations and accounts (Section 4). There are still other uses of OKAY which contribute to managing understanding in interaction, for example its use as a tag which pursues a response or as a turn-final particle (see Couper-Kuhlen on English and Ostermann and Harjunpää on Portuguese, both in this volume) and its use at the beginning of an extended turn (see Lindström 2018 on Swedish and DeSouza et al. 2021 on English). In this chapter, however, we will restrict ourselves to stand-alone cases, that is, OKAYs which have their own prosodic contour and perform a single action. These tend to be OKAYs that form their own turn. We begin, however, with an OKAY in German which combines with the change-of-state particle aAH ‘oh’ and is followed by talk that explicates the claim of understanding which it embodies. This suggests that ‘claiming understanding’ is indeed a participant category for the responsive uses in which we are interested. Excerpt 1 is from the four-person card game Phase 10, in which players advance
3. We use a broad understanding of response/responsive in this chapter and include next actions that are positionally sensitive but not sequence-specific, such as continuer uses of OKAY (see Section 3). See Enfield (2011, 286–287) for a broader and Thompson, Fox, and Couper-Kuhlen (2015, 2–10) for a narrower definition of what constitutes a ‘response’.
Chapter 3. OKAY in responding and claiming understanding 57
through a set series of successively more difficult ‘phases’.4 In lines 1–2, Vincenzo (VIN, whose first language is Italian), who is learning the game, is asking for an explanation of the rules guiding Anna’s just-completed move. Excerpt 1. versteh ich ‘I understand’ (Phase10.ms_03.2016_23:12–24:06), 2016, German, face-to-face, informal game-playing interaction 01
VIN:
02 03
ANN:
04
BRI:
05 06 07 08
09
bri
ANN: ann
BRI: bri
VIN:
BRI:
aber JEtzt wir sind auf den dritte, u:nd (.) sie but now we’re on(to) the third ((phase)) and she
spielt auf den zweite. wie[=s geh]t das jetz; is playing the second how does that work now [>ja< ] yeah JA WEI[%L- es ist] ein%facher für sie rauszulegen;%= well cause it’s easier for her to lay down cards %RH fwd--------%RH palm up gest twd VIN----% [+↑jaJA.+ ] yeayeah +moves hands apart+
=%als für uns.% than for us %retracts RH-%
natürlich hat sie bessere chancen ↑aufzuholen. of course she has better chances to catchup (m)okee;
aber es is: geht ganz nor↑mal. but it’s proceeds as usual
10
du kannst bei ihr anlegen= you can add to her cards
11
=so[bald du ]deine phase ausgelegt hast. as soon as you have displayed your phase [genau. ] exactly
12
ANN:
13
ANN:
[>ja> das ganze.° the whole thing (0.2)
eine phase w(h)eit(h)er.=hähähä[hä one more phase down [HÄ:=chähä. huh huhuh (0.1) ich komm glEICH_ I’ll be there soon
Brigitte, the most experienced player, provides a response to Vincenzo’s question in lines 4–11. Vincenzo receipts her explanation regarding Anna’s play with (m)okee; (line 8). Brigitte then adds a prospectively-oriented reminder that the remaining game rules still apply (line 9) and formulates one of those rules (lines 10–11). This is endorsed by Anna (lines 12, 13) and receipted by Vincenzo in line 14: Vincenzo marks the prior as new information with aAH (Imo 2009; Golato 2010; see also Koivisto 2015b on now-understanding in Finnish), and the following OKAYs signal that the new information is to be taken as understood.5 That OKAY is proposing understanding is supported by the continuation of Vincenzo’s turn in lines 17–18: He does not move to next or new matters but instead elaborates the claim adumbrated by the particle. Additionally, in his next turn a few seconds later (not shown), Vincenzo demonstrates his understanding by offering a correct application of a related rule. Examples in which OKAY is followed by self-reformulation suggests that OKAY is directly implicated in claiming understanding. In the remainder of this chapter, we illustrate different ways in which speakers across languages use free-standing OKAY as a resource in managing understanding in responsive turns. 2. OKAY claims (sufficient) understanding What OKAY does crucially depends on the action that the prior turn has been implementing. While after assessments, proposals and suggestions, a free-standing OKAY conveys agreement, after informings, it works as a claim of (sufficient) understanding (Beach 1993, 329–331). An informing can be an announcement, an 5. For a more detailed analysis of change-of-state + OKAY combinations, see Helmer et al. on German and Koivisto and Sorjonen on Finnish, both in this volume.
Chapter 3. OKAY in responding and claiming understanding 59
explanation, a piece of news, a reporting, or part of a telling, for which particles are a common format for responding (for English, see Thompson et al. 2015, 64–84). Informings can be volunteered, elicited, or offered as part of a larger activity (a story or other telling), and particle responses thus appear in different sequential positions (second or third). We first show how OKAY is used as a second- or third-position receipt after informings and explanations, both solicited and unsolicited. Excerpt 2 illustrates the use of OKAY in Mandarin in second position to an announcement. Four student friends – Lim, Lien, Ruiling, and Peng – are playing cards around a table (see Figures 1, 2). Lim verbally (line 1) and gesturally (line 3) marks the beginning of the card game. Before the excerpt below, participants had decided to play – and be recorded – for an hour.
Figure 1. Participants from left to right, clockwise: Lim, Lien, Ruiling, and Peng. Ruiling (back right) is pointing and Lim (front left) is gazing at the timer (line 4)
Figure 2. Mutual gaze between Ruiling (back right) and Lim (front left) during okay; (line 6)
c3-fig2
Excerpt 2. jishi qi ‘timer’ (card game_Video_02:50 (V5) Corpus 3), 2014, Mandarin, face-to-face, informal game-playing interaction 01 LIM:
(okay) na women kaishi ba. then let’s begin
02
(0.3) ((LIM: brief gaze up))
03 LIM:
((knocks table with cards twice))
60 Emma Betz and Arnulf Deppermann
04 RUI:
05
lim fig
06 LIM: rui fig
07 LIM:
08
+zhe bianr &you #shi+jian& you this side have time have there’s time a timer on this side +....points to timer-+gaze to LIM-------> &gaze to timer&gaze forw to RUI--> #fig.1 jishi q[i ke]yi kandao. timer can see ((we)) can see the timer [&°o#kay;°&+] &nod---->& ,,,,,,>+gaze away #fig.2
&.hh &wo- women *shi ba:& we be BA6 ,,,>&gaze to left to LIE& &circling gest w/r/index& *head tilt to right tow LIM
suoyou de pai dou fawan dui bu dui. all ASSC cards all hand out right NEG right we will take all the cards, right?
In line 4, Ruiling (who is sitting diagonally across from Lim) announces that the recording device on her side of the table includes a timer. Lim directs his gaze to the device to which Ruiling is pointing (Figure 1) and receipts her announcement with OKAY (Figure 2) and a simultaneous down-up head nod (on the coordination of nodding and OKAY, see Helmer et al. in this volume). Formal features of the particle include low pitch onset, low volume and falling pitch. Ruiling orients to Lim’s OKAY-turn as a receipt of information and claim of sufficient understanding: She retracts her pointing finger and shifts her gaze away from Lim (line 6). Lim’s bodily conduct also indicates that no further expansion of the subject matter is needed: He shifts his gaze to another participant (Lien, sitting to his left) and resumes game preparations by addressing to Lien a candidate understanding of a game rule (lines 7–8). OKAY-turns such as the one shown here mark information as sufficient, thus claiming understanding and, as a corollary, promoting sequence closing. After OKAY, speakers or recipients return to pending matters or move to next matters. OKAYs regularly constitute third-position responses in question-elicited informings. In such sequences, elicited information is treated as sufficient with OKAY produced by the speaker who explicitly requested the informing. A claim of understanding systematically coincides with the proposal that sequence closing is appropriate (cf. Grosz 1982, 153); such OKAYs thus function as sequence-closing thirds (Schegloff 2007, 115–168; see also Chapter 4 in this volume). Excerpt 3 comes
6. Glosses used: BA = pretransitive marker in the BA-construction; ASSC = associative marker; NEG = negator.
Chapter 3. OKAY in responding and claiming understanding 61
from a Finnish interaction between friends and family on a Sunday afternoon. Paavo, Susa and Maija usually eat Sunday lunch with Susa’s and Maija’s father, who lives upstairs in the same building. On this particular Sunday, Maija was not present at the family lunch, and this motivates the question in line 1. Previously, the participants have been talking about the rings Maija is wearing, and during a 1.8 second silence just prior to line 1, Susa has been inspecting Maija’s rings, trying to read the rings’ inscriptions (Figure 3). Maija’s question in line 1 is thus unrelated to the prior topic.
Figure 3. PAA (Paavo), IIN (Iina), SUS (Susa) and MAI (Maija), clockwise, seated in Susa and Paavo’s home. Susa is trying to read the inscription on Maija’s ring (line 12)
Excerpt 3. ruoka ‘food’ (Aineisto1_30_40), 2011, Finnish, face-to-face, informal conversation 01 MAI:
ettekste ollu yläkerrassa syömäs; didn’t you.PL have lunch upstairs
02 PAA:
olti[in. ] ((we)) did [°↑(aa)], ‘oo°, (aa) yeah
03 MAI: 04 SUS: 05 MAI: 06 SUS: 07 PAA: 08
09 SUS: mai 10 PAA: 11
mai
kaisa °j[aok↓ke.< sorede sa::, then PRT and you know u:n, yeah
11 WIF:
HUS moves toward sequence closing in line 6, but with WIF’s addition of specific information in line 7, a possible understanding problem seems to emerge for him. HUS offers a candidate understanding (line 8), which amounts to a question making relevant a yes- or no-response. HUS’s understanding is confirmed by WIF. In next position, HUS receipts this with a falling-intoned >ok↓ke.< and then shifts to next matters. HUS’s >ok↓ke.< proposes sufficient understanding in the sense that it marks that the answer aligns with what HUS had assumed and is sufficiently informative to enable HUS to carry out his planned course of action (sending a letter to arrive in Okinawa while his wife is still there). The function of OKAY to indicate that an information-focused question has been sufficiently dealt with can also be used to communicate a more specific stance toward the prior turn in third position: An OKAY-turn can propose that the recipient has provided too much information, thus curtailing the (projected) further development of a sequence, for example an elaborate correction of a candidate understanding. Proposing that the prior is more than sufficient or “overdone” (Stivers 2004, 271), that is, that a speaker has “persisted unnecessarily in the current course of action” (Stivers 2004, 260), can be done with multiple OKAYs, which then target not only the just-provided information but rather the larger course of action.7 Third-position OKAY in informing sequences can also receipt an answer as sufficiently informative within a larger activity organized by sequences of questions 7. For a particularly clear example of this use with okhe=okhe in Korean, see Excerpt 4 “Unable to make it” in Kuroshima et al. (this volume); for an example of multiple OKAYs in English, see Stivers (2004, 270–271). This effect of reduplication, however, is not specific to OKAY; see for instance Barth-Weingarten (2011b) on German jaja, Kunnari (2011) on Finnish joojoo, and Keevallik (2010c) on reduplicated imperatives in Estonian.
64 Emma Betz and Arnulf Deppermann
and answers, for instance the history-taking phase in a doctor-patient encounter (see Beach 1995b, 266–267 for English). Excerpt 5 is from a Brazilian Portuguese interaction during an obstetric ultrasound exam performed by a physician. It is taken from the anamnesis phase, at a point at which the physician has already asked several questions. His query in line 1 concerns the medical history of the patient’s mother. Excerpt 5. pressão alta ‘high blood pressure’ (ECOCLIN_GABRIEL_Agata_ RM_07_01_2013 00.38), 2013, Brazilian Portuguese, face-to-face, doctor-patient interaction 01 ECO:
>↑sabe< se a tua mãe teve do you know if your mother had
02
pressão alta na gestação?= high blood pressure during pregnancy
03 PAT:
=nã:o. no
04
(0.1)
05 ECO: 06 PAT:
.hh[h
[não ↓teve; ((she)) didn’t
07 ECO:
°oq(h)ue:i?° e a< gestação and the and the pregnancy
08
(.) é:: foi espon↑tânea uh was ((it)) spontaneous
09
>ou tu< precisou fazê algum tratamento or did you need to have some treatment
10
pra [(ele::)] in order to [foi ] espontânea. ((it)) was spontaneous
11 PAT:
In lines 1–2, the physician asks a ‘do you know’-framed next question, to which the patient responds with nã:o ‘no’ (line 3). After a pause in line 4, the patient redoes her answer with a verb repeat response (nã:o teve ‘((she)) didn’t’). She thus removes a possible ambiguity in her initial answer (in line 3) between responding to the ‘do you know’-frame (‘no’ = I don’t know the answer) and responding to the proposition of the question (‘no’ = she didn’t have high blood pressure). The physician’s third-position °oq(h)ue:i?° receipts the information given in line 6, marking it as now sufficient. In the larger activity of history-taking, the understanding claim with OKAY enables progression to the next point, in which the physician now topicalizes the patient’s own pregnancy (lines 7–10). This OKAY thus is a boundary case, combining the use of indexing sufficient understanding with the use as a transition marker as discussed in Chapter 4 of this volume. The occurrence of °oq(h)ue:i?° at the juncture between two question-elicited informing sequences within a larger activity may also account for the prosodic delivery of the particle
Chapter 3. OKAY in responding and claiming understanding 65
with rising intonation (rather than with the falling intonation we observe as typical for the uses illustrated in this section; on the interaction between the intonation of OKAY and topic shift, see also Heritage and Clayman 2010, 110–115). OKAYs with final rising intonation and marked prosody systematically index not-yet-complete understanding and counter-to-expectation informings and thus make more talk from the prior speaker relevant (see Sections 3 and 4; cf. Oloff 2019).8 This is not the case in the present example: The particle is quiet and aspirated, and it is not oriented to by the patient as encouraging more talk. Instead, rising intonation here seems to contribute to projecting that the physician will continue with a next item within the larger activity (cf. Barth-Weingarten 2002, 2009), and the and-prefacing of the next item supports this (Heritage and Sorjonen 1994; Nevile 2007). The larger activity context is also relevant in the next excerpt from an Estonian phone conversation, in which the OKAY-speaker explicitly formulates the action that the just-received information makes relevant. Merilin (MER) is at work in an office building, when Hannes (HAN) calls her from his car. Merilin responds to the summons in line 1 and asks, immediately after Hannes’ self-identification (line 3), for his location (line 4). This points to a prior arrangement between the two: Hannes seems to be picking up Merilin from work. Excerpt 6. Balti jaamast ‘from Balti station’ (CASE 2. M1A8, Tallinn), 1998/99, Estonian, telephone, informal conversation 01 MER: 02 03 HAN:
-lo; (hel)lo (0.3) tere hannes on siinpool. hi it’s Hannes ((speaking)) here
04 MER:
jah. kuspool. yeah where
05 HAN:
ole-n liikvel. (.) [sinu] poole. I am moving toward you [jah;] yeah; [balti jaama-st.] from Balti ((name)) station.
06 MER: 07 MER: 08 HAN:
[ ole-n TEE-L ] su poole. I am on my way to you
09
(.) balti jaama-st jah; from Balti station, yeah
10 MER: 11
oKEI.
(0.2)
8. For a similar argument regarding English oh in informing sequences (that is, a systematic connection between oh indexing “now K+” vs. “not yet fully K+” and falling vs. rising intonation), see Thompson et al. 2015, 56–57.
66 Emma Betz and Arnulf Deppermann
12 MER:
akka-n kokku pane-ma asju ja alla tule-n. I’ll begin to collect my stuff and come down
13 HAN:
no: akka jah, do that
14 MER:
jah; yeah
Hannes’ location description in line 5 as ‘moving toward you’ leads to an understanding check by way of a syntactic extension (line 7). In line 9, Hannes provides a repeat-formatted confirmation, and this confirmation is receipted by Merilin with oKEI.. That this OKAY constitutes an understanding claim, that is, a claim to having been sufficiently informed for present purposes, is supported by Merilin’s announcement after oKEI.: With akka-n kokku pane-ma asju ja alla tule-n. ‘I’ll begin to collect my stuff and come down’, she formulates the next actions following from Hannes’ informing concerning his current location, thus essentially offering an understanding demonstration (cf. Grosz 1982, 153). Since the understanding which is embodied by oKEI: here concerns an expected next action,9 it is also related to the use as a device indexing transition to a next activity (cf. Bangerter and Clark 2003; Bangerter, Clark, and Katz 2004; Merritt 1978; see Chapter 4, this volume). This promise of action reflects Merilin’s understanding of agreed-upon plans and is endorsed by Hannes in line 13 with no: akka jah, ‘do that’, a responsive imperative that provides a go-ahead for the action to which Merilin is already committed (Zinken and Deppermann 2017; Heinemann and Steensig 2017; Keevallik 2017; Sorjonen 2017). In Excerpt 7, similar to Excerpt 6, a third-position OKAY follows a repeat-formatted confirmation of an understanding. In this phone conversation, the caller makes a request for help with a malfunctioning elevator (lines 1–2), which is rejected by the operator in line 3. The caller then formulates an understanding for confirmation in line 5. Excerpt 7. ascenseur ‘elevator’ (Jean2/rec2/2.05), 2003, French, telephone, service encounter 01
APP:
02 03 04
.h euh: ce soir ça serait possible .h uhm this evening would it be possible de me dépanner ou [pas? to assist me or not?
OPE:
[>non.< no (0.4)
9. In responding to an informing, this OKAY thus not only marks sufficient understanding of the informing but also conveys that the informing has implications for the respondent’s own future behavior. Couper-Kuhlen (2019a) terms this a “consequential” use of OKAY.
Chapter 3. OKAY in responding and claiming understanding 67
05
APP:
06
07 OPE: 08 APP:
ah ça fait demain? oh this means tomorrow? (0.2)
demain oui. tomorrow yes o:ké.
09 OPE:
au r’voir, bye
10 APP:
très bien, à demain very good, see you tomorrow
After having completed a description of the problem that he is currently experiencing (not shown), the caller launches a request that targets the time at which help could be delivered. Ce soir ‘tonight’ (line 1) is positioned early in the caller’s turn, which is formatted as a polar question and closed by ou pas ‘or not’, indicating that the speaker now expects a negative answer. This final ou pas is overlapped by the operator’s non ‘no’. The caller registers the straightforward rejection of the time with ah in line 5 (in this context similar to oh in English; Heritage 1984a) and then offers the new temporal reference demain ‘tomorrow’. This turn is designed as a conclusion from the prior sequence (with ça fait ‘this/that means’) and as dependent on the prior talk.10 The conclusion is confirmed in line 7 by the operator and followed by the caller’s o:ké. The claim to having been sufficiently informed for present purposes in this example allows for sequence closing and a transition to the next relevant step, in this case conversation closing. Note that the caller reuses the lexical item demain in the final greetings (line 10), which is another way of ratifying the new, agreed-upon time. The specialized use of OKAY as a claim to sufficient understanding of information emerges particularly clearly in contexts in which the particle is followed by a separate element dedicated to sequence closing or transition (e.g., Polish do:bra: ‘good’; Portuguese tudo bem. ‘alright’, see line 21 of Excerpt 10 “That type of information” in Ostermann and Harjunpää, this volume). Prosodic shape seems to be an important feature of OKAY: As the excerpts above illustrate, OKAY in this use tends to be realized with falling pitch (cf. Gravano et al. 2012).11 OKAYs also serve as receipts after other-corrections and disconfirmations, thus marking revised understanding. In these contexts, OKAY regularly co-occurs with 10. Compare uses of German das heißt ‘that means’ and du meinst ‘you mean’, as described by Helmer and Zinken (2019). 11. See Golato and Betz (2008) and Golato (2010) for similar observations regarding understanding claims in third position and the prosodic realization of the German particle combination achso ‘oh I see’. Thompson et al. (2015, 55) note that falling intonation is typical of sequence-curtailing responses to question-elicited informings in English in general.
68 Emma Betz and Arnulf Deppermann
tokens registering a change of epistemic state (e.g., a: in Korean, ah in Chinese; cf. Oloff 2019; Helmer et al. this volume). Yet, more rarely, there are also cases in which a free-standing OKAY occurs by itself as change-of-state token. Such OKAYs thus appear to index a transition from a state of not knowing to now knowing (Heritage 1984a, 2012a; cf. Heinemann and Koivisto 2016a). The use of OKAY as a stand-alone change-of-state token can be observed in the next example from Mandarin Chinese. Here, however, it combines with an additional note of concession. Don, Yua, Hon, and Min are friends who are talking about children’s reaction to divorce. Don has maintained that divorcing parents should talk with their teenage children about divorce reasons, rather than merely informing their children. This, he argues, would make it easier to understand and accept the situation. Yua, to the contrary, has argued that no matter how much parents talk to their children, they would not be able to understand zhezhong qingkuang ‘this situation’. Don now objects that Yua does not have any experience of such a ‘type of situation’ (lines 1–2). Excerpt 8. zhezhong qingkuang ‘this type of situation’ (Tea talk_Audio 01:11:21, Corpus 3, Edmonton), 2014, Mandarin, face-to-face, informal interaction 01
DON:
02 03
shouxian shi zheyang de; first of all it’s like this
yi:: ni meiyou yudao zhezhong qingkuang. first you didn’t encounter this type of situation YUA:
04
wo yudao zhezhong qingkuang; I encountered this type of situation [ni ping] shenme shuo wo mei= you depend on what say I NEG what makes you say I didn’t
05
DON:
[ah yu ] AH en(counter)
06
YUA:
=yudao [zhezhong qingkuang. encounter this type of situation
07
DON:
08 09
fig
MIN: DON:
10 YUA: 11 DON:
]2
[+O:#°kay.°+=[ni yudao zhe]1zhong]2= you encountered this type of +head/gaze up+ #fig.4 [&kHHH!& ]1 &gaze to Don, then Yua&
=qingkuang.=[na +jiu%shiʔ] (0.2)+(0.2)% situation that just be situation then (it’s) just +gaze to Yua----+ %head tilt to left% [*°dui.°* ] right. *gaze down, away from Don*
.h danshi wo ye wo queshi but I also I indeed
Chapter 3. OKAY in responding and claiming understanding 69
Figure 4. Participants clockwise from left: Min, Don, Yuan, and Hon. Don moves his head upward on O:°kay.° 500
300
Pitch (Hz)
200
100 70 50 O: 0
kay
ni 0.3
yu 0.6
dao
zhe
zhong 0.9208
Time (s)
Figure 5. Pitch trace for O:°kay.°=ni yudao zhezhong, in line 7
Yuan corrects Don’s assumption (line 3) and then asks Don for a justification (lines 4, 6), a move which amounts to an accusation. In overlap, Don receipts Yuan’s correction as news with a change-of-state token (ah, line 5), and, still in overlap with Yuan, produces an OKAY and repeats Yuan’s correction (lines 7, 9). The OKAY here is produced in second position in a repair sequence: correction – OKAY. The O:°kay.° follows a change-of-state claim (line 5) and is realized with a higher-pitch, lengthened first syllable and a softer, falling second syllable (see Figure 5).12 There is a slight hiatus between the first and the second syllable, whose prosodic deliveries 12. Compare Koivisto (2015b) on indexing revised understanding with Finnish aa after corrections, and her discussion of how the prosodic realization of the particle is systematically connected to upgrading or downplaying the proposed cognitive shift.
70 Emma Betz and Arnulf Deppermann
contrast sharply, sounding like two quite distinct parts: O: + kay. This may suggest that both parts are designed to express the two different actions in play here: indexing a change of state with the first part (which has more phonetic prominence) and indexing concession with the second part (which is delivered like an agreeing OKAY). This analysis is also supported by the fact that on the first syllable of O:°kay.°, Don nods up (Figure 4), as is common for change-of-state tokens (cf. Helmer et al. in this volume), shifting his gaze away from Yuan, and then lowers his head again on the second syllable. In this sequential position, after an other-correction, OKAY conveys a change of epistemic state, but also concession, because the OKAY-speaker backs down from his prior assumption (that his interlocutor is not knowledgeable about the matters talked about) in response to the interlocutor’s correction and accusation.13 This double function of OKAY as indicating a change in epistemic state and indicating a concession is reinforced by the ensuing repetition of the concession by the OKAY-speaker. The subsequent interaction (not shown here) gives further evidence of Don’s revised understanding after OKAY. We have seen that OKAY can be used to display sufficient understanding of a volunteered first-position or elicited second-position informing. In the response to multi-unit narratives or informings, OKAY is contrastively used with other displays of recipiency: When items such as English mh, mhm, and yes have been used as continuers supporting the production of a lengthy multi-unit turn under way, OKAY is used only later to convey that the recipient considers the story or informing to be now complete and that they have understood it sufficiently. Excerpt 9, taken from a telephone call in Polish, is a case in point. A client has called a company selling shelves. In the excerpt, the salesperson instructs the client on how to find a specific type of shelf on their website. He tells the client that he needs to use the search window (line 1) and enter the product code (line 2). He then starts dictating the code number by number (lines 2–15). Excerpt 9. siedemdziesiąt sześć ‘seventy-six’ (VN680313_numbers okej), 2009, Polish, telephone, service encounter
01
SAL:
w lewym górnym rogu::: jest pole szukaj i tam= in left top corner is (a) field search and there
CLI:
=proszę wp[isać ] (.) .hh twadzieścia trzy:, please write .hh twenty three [tak, ] yes (2.4) ((C is typing the numbers))
02 03 04 05
CLI:
yes
13. A concessive use in American English can be found in Couper-Kuhlen (this volume; Excerpt 15 “Blow off your girlfriends”). See also Heisler (1996, 300–301) for a description of a concessive use of OKAY in Montréal French.
Chapter 3. OKAY in responding and claiming understanding 71
06 07
SAL:
08
CLI:
09
SAL:
10 11
CLI:
12
SAL:
13 14
CLI:
15
SAL:
16 17 18 19
CLI:
CLI:
sześćdziesiąt sześć, sixty six (0.8) ((C is typing the numbers)) ta:k_ yes
pięćdziesiąt cztery, fifty four (1.0) ((C is typing the numbers)) °ta:k_° yes
osiemna:ście, eighteen (1.0) ((C is typing the numbers)) ta::k_ yes
sie:demdziesiąt sześć. seventy six (0.5) ((C is typing the numbers))
sześć. ↑o↓ke::j. seventy six (1.0) ((audible key stroke)) o:::: y:::: jest. PRT PRT I got it oh uh I got it.
When dictating the code, the salesperson delivers it in two-digit chunk installments (cf. Clark and Brennan 1991; Svennevig 2018; see also Sorjonen 2001, 251–253 on receiving three-digit chunks of phone numbers with joo in Finnish), each indexed through a final rise as being not yet complete. Each of the subsequent installments is received by the client with the acknowledgment token tak ‘yes’ (lines 5, 8, 11, 14), after he has typed the respective two-digit chunk into the search window. Tak at the same time is used as a continuer, giving the salesperson a go-ahead to produce the next digits. In line 15, the salesperson delivers the last two digits, bringing his multi-unit turn to a possible completion (note the final falling intonation). Having typed the numbers in the search window (line 16), the client repeats the last two digits and then, in his second TCU in the same turn (line 17), produces an accentuated ↑o↓ke:j.. This ↑o↓ke:j. exhibits a pitch peak on the first syllable and a falling contour on the second syllable, which indexes the speaker’s understanding that the activity of dictating is now complete and has sufficiently been understood. OKAY does not itself close the sequence but projects closure, which is then realized with the client’s key stroke (line 18), through which he sends his web-query. This action, which builds on the informing, provides further evidence that ↑o↓ke:j. indexes the sufficiency of the informing. In his subsequent turn, the client declares that he has
72
Emma Betz and Arnulf Deppermann
found the merchandise he was looking for. The first item in the client’s turn is the noticing particle o::::, followed by the stretched hesitation marker y::::, which may indicate the client’s attempt at locating the item, and by the verb jest, which refers to the item as ‘just found’. Excerpt 9 shows that in responding to a lengthy multi-unit turn in fieri, OKAY has greater ‘weight’ than response particles which precede OKAY as acknowledgment tokens and continuers. OKAY indexes that the main information has been conveyed and that an understanding which is sufficient for the current practical purposes has been gained. However, using OKAY to index understanding of a major point made in an interlocutor’s lengthy turn does not necessarily imply treatment of that turn as closed. Rather, the use of OKAY as an understanding claim may combine with continuer functions. This is the case in Excerpt 10, taken from a phone-call in Italian. A bank employee (PAD) has called a client (BOR) to ‘invite him to a meeting at the bank’ (not shown), where he would be presented with new products and services (see De Stefani 2018). BOR works towards declining the invitation, mentioning that his second son was just born and that he is therefore unable to make an appointment at this point in time. The excerpt starts with the employee’s response to the client’s account. Excerpt 10. alla banca ‘at the bank’ (obfc5, 06:47–07:08), 2008, Italian, telephone, sales call
01
PAD:
02
BOR:
04
BOR:
03
PAD:
05 PAD:
07
BOR:
08
10 11 12
[va-
[f-
=questo momento ((chuckles))[(che è:) this moment which is [sì: adesso::: se yes now if non è questa settimana [sarà nella prossima= it’s not this week it will be the next
06
09
niente [passi[:: adesso (.) con calma si:: goda= fine come now without hurry enjoy
[no::: ( ). no =però boh intanto io:: son sempre di passaggio but I dunno in the meantime I am always passing alla [banca perché poiat the bank because then-
PAD:
BOR:
[sì (.) esatto. yes right. (0.4)
eh::: lavoro: lavorando nella ristorazione uh I work ((since I’m)) working in catering ho: abbastanza:: I have enough
Chapter 3. OKAY in responding and claiming understanding 73
13 14
PAD: BOR:
15 16
↑occhei;=
=contatto colla banca e poi al massimo posso contact with the bank and then at the latest I can chiedere anche lì al:^eh ask also there at uh
17
PAD:
18
BOR:
19
PAD:
20
BOR:
(0.5)
[ad ((place name)).] in ((place name)) [allo sportel]lo at the counter sì sì anche allo [sportello. yes yes also at the counter [ad ((place name)). in ((place name))
PAD’s response (lines 1, 3) leaves the initiative to visit the bank to the client and thus displays her orientation that the phone call will be closed without an appointment. BOR accepts the invitation without a definite commitment by affirming that he will visit the bank within the next two weeks (lines 4–5, 7). He explains that he regularly passes by the bank (lines 7–8) because of his work in catering (line 8–14). During BOR’s lengthy TCU in line 11–15, PAD produces an ↑okay; with high onset and falling final pitch (line 13). It is produced at a point at which BOR’s talk is incomplete syntactically, prosodically, and pragmatically; the particle therefore clearly is not at a TRP, yet it is produced in the clear, without overlap. With ↑okay;, PAD acknowledges BOR’s working in the catering business as new information (it was not mentioned in previous talk), thus displaying a change of epistemic state. In addition, the precise placement of ↑okay; after ho: abbastanza:: ‘I have enough’ (line 12) may also convey that BOR already anticipates the argumentative point BOR’s turn is designed to make: that BOR’s work is the reason why he has regular contact with the bank. In this sense, ↑okay; would not only claim understanding of new information but simultaneously index an understanding of what the turn-inprogress projects and is to accomplish as an action. Yet, the precise positioning of ↑okay; also indexes that PAD expects BOR to keep the floor. In continuing his turn in line 14 without any delay and by producing a coherent next syntactic constituent (a direct object NP) which fits the clausal frame established by ho: abbastanza::, BOR also treats ↑okay; as a continuer,14 through which PAD aligns herself with his ongoing turn-production. 14. We use the term ‘continuer’ here in a rather broad sense, as conveying an understanding that the prior/ongoing turn is not complete and that the prior speaker is expected to continue. We will show and discuss examples of continuer uses as understood more narrowly in Schegloff (1982, 2007) and Heritage (1984a), that is, of OKAY used to simply or primarily pass a turn at talk (and thus distinct from, e.g., newsmarks), in Section 3.
74
Emma Betz and Arnulf Deppermann
Excerpt 10 has shown that, depending on the sequential position of its production, OKAY can carry out multiple local functions. In the specific case from our Italian data, proposing an epistemic change of state combines with the use of OKAY as a continuer. In addition to this, a claim to anticipating a projectable course of a turn in progress and to understanding its pragmatic import are possibly in play as well. Prior research has described ‘accepting’ as a core function of responsive OKAY (Beach 1993, 1995b; Schegloff 2007, 128). Section 2 has illustrated a range of uses of OKAY after informings, where ‘accepting’ can be specified as ‘marking a prior turn as sufficiently informative for present purposes’. In this specific context, accepting then amounts to claiming understanding. As a corollary to claiming understanding, these OKAYs propose that sequence closing is appropriate and thus enable a move to pending or next matters.15 The OKAY uses illustrated here are thus sequentially non-continuative, and this seems to be reflected in the prosodic realization of the particle as well. In the uses illustrated in Section 2, OKAY tends to be realized with falling pitch, which has been suggested as typical of sequence-curtailing responses in some languages (see Golato 2010; Thompson et al. 2015, 55, 135). As highlighted in Chapter 2 of this volume (Betz and Sorjonen), the functional spectrum of OKAY emerges more clearly if OKAY is compared to alternative response types available to speakers in similar sequential environments, such as minimal clausal responses (repeats) or other particles. In third-position uses, especially in informing sequences that involve corrections, OKAY contrasts with (partial) repeats: Repeat responses register information without necessarily showing commitment to it or claiming understanding of its relevance. They are thus not sequence-closing (Svennevig 2004; Betz et al. 2013) or at least equivocal for participants (Schegloff 2007, 126–127; Koivisto 2019). In foregrounding ‘acceptance’ and understanding, non-continuative OKAY also contrasts with particle responses that anchor (now-)understanding in claims to have just retrieved or reactivated existing knowledge, and with particles that foreground the receipt of new or newsworthy information (see Heinemann and Koivisto 2016a). In the next section, we will shows that responsive OKAY can also be sequence-continuative. Sequential placement (e.g., at the beginning or in the middle of extended informings) and prosodic design are crucial in distinguishing such continuative uses from those seen in Excerpts 1–9.
15. Transitional uses of OKAY, that is, OKAYs that project and prefigure next moves, topics, and activities, are described in depth in Chapter 4.
Chapter 3. OKAY in responding and claiming understanding 75
3. OKAY marks understanding of prior informing as preliminary or not complete OKAY can, as we saw in Excerpt 10, be used (and taken) to propose an understanding of the prior action as preliminary to further talk and thus as projecting more. Such OKAYs are followed by more talk by the prior speaker. In this section, we look more closely at uses of the particle that convey an understanding of the prior talk as not (yet) complete. Incompleteness may already have been indexed by the prior speaker themselves (e.g., by continuative prosody), but OKAY can also suggest that the prior turn is in need of expansion when the prior speaker did not project turn-continuation. A primary case is the use of OKAY as a continuer during extended tellings (Schegloff 1982, 2007), which we illustrate with the next two excerpts. In Excerpt 11, taken from an American English phone conversation, Debbie is responding to her friend Sara’s query (line 1) by explaining what she does on her part-time job. The explanation proper is launched in line 3 with a well-prefaced unit but then discontinued for a parenthetical insert sequence (lines 4–10). It is resumed with we:ll in line 11, a preface which contributes to projecting an expanded response (Heritage 2015). Excerpt 11. Scheduling (Call Friend Engn 6239, 1669.808), 1996/97, American English, telephone, informal conversation 01
SAR:
03
DEB:
02 04 05 06
SAR:
09
DEB:
10 11
16 17 18 19
>I mean< it has nothing to do with fi:lm; [°you kno:w°] [↑I know
=or [what.
] is it like secretarial stuff?=
]
[yea:h; ]
it’s (0.3) it’s secretarial;= i:s u:m (0.5) I: (1.1)
13 15
!p well I- I’m working in (0.8)
=↑we:ll_ .hhh I- ↑what I do ↓m:ost of the ti::me;
12 14
(0.9)
(.) °at a:ll.°
07
08
what are you doing.
↑each (.) ↑they bring in like these people
SAR: DEB:
to interview, for high level positions, [mhm:,]
[.hhh ] and (0.2) when they come in
to interview;=the:y (.) have to interview with like eight peopl:e, (0.6)
76
Emma Betz and Arnulf Deppermann
20
21 22
SAR:
DEB:
23
[°m:↓kay_°] [!t they
] have to do like
eight hours of interviewing.
↑and the(y) these people come from
24
all over the country;=
26
of the panel members, a:nd the (.) candidate?
25
=so I have to arra:nge schedules so that a:ll
27
.hhh are free: on the sa:me da:y,=
28
=and then I have to arra:nge the schedules
29
with the panel members,
30
.hhhh and then I have to li:ke (.)
31
se:nd documentation to the candidate
32
and °to the panels.°
33
(0.2)
34
35 36
37
38
so ↑that’s what I do most of the time
SAR:
DEB:
is sche:duli:ng? [↑O:↓kay;]
[.hh
] but- (0.3) but I also do::-
like whatever else °they need me to do:.°
In line 18 Debbie has not yet reached a point of possible pragmatic completion. The slightly rising pitch, and possibly the lengthening on people, project continuation of the telling (Ford and Thompson 1996; Wennerstrom and Siegel 2003); however, she leaves a little room for a response (line 19). After a brief pause, Sara uses a soft, level-intoned m:kay_ in line 20.16 This particle does not mark the informing as complete and does not propose that the sequence can be closed; rather, it conveys that Sara is following and Debbie should continue, and it is understood as such. In her continuation, Debbie provides further background information necessary for appreciating the scope of her work in arranging schedules (lines 25–35). Debbie’s continuation comes in overlap with Sara’s continuer (line 21). In line 34–35, Debbie reinvokes the frame with which she started her explanation (cf. line 11), indicating that this explanation has now come to a point of possible completion. Sara receipts this with a falling-intoned ↑O:↓ kay. in line 36 (see Section 2 above), thereby proposing closure of the sequence she initiated in line 1 with what are you doing.
16. See Couper-Kuhlen (this volume) on recurrent prosodic patterns for sequence-internal continuer uses of OKAY in American English spontaneous interactions. Gravano, Hirschberg, and Beňuš (2012), using data from American English task-oriented dialogues, also find distinct patterns in word-final intonation for OKAY with “backchannel function” (rising) v. “agreement” OKAY (plateau or falling) (p. 17).
Chapter 3. OKAY in responding and claiming understanding 77
Excerpt 12 is also drawn from an extended telling/reporting on the phone, but it is taken from an institutional context: a volunteer breast-feeding support helpline in Sweden (Bertils 2016). We see below the mother’s (MOT) problem presentation to the call taker (CTA) at the beginning of the interaction. In lines 5–6, the mother reports taking her child to the children’s health center (Bevece ‘BVC’, line 5), where, as a matter of course, the boy was weighed. The mother’s men: ‘but’ in line 6 projects a problem formulation; dom/rom ‘they’ (lines 8, 10) and sa ‘she’ (line 15) refer to the staff at the health center. Excerpt 12. hundratjuge ‘one hundred and twenty’ (Bertils ARUU6, 00.38–1.21), 2016, Swedish, breastfeeding support helpline interaction 01
MOT:
ja födde barn för ungefär tre vecker sen, I gave birth about three weeks ago
02
eller a tre- jaa den sextonde så att de e or well three yeah the sixteenth so that it is
03
#ju:#- han e tre veck(h)er å en d(h)a här nu då, well he is three weeks and one day here now then
04 05
CTA:
mm,
CTA:
[å så hade han- han hade: gått upp men:= and so had he he had gone up ((in weight)) but [mm,
MOT:
06 07 08
MOT:
ä:h och=ä::h så va ja på: bevece i torsdas, and I was at BVC on Thursday and he had
=äh (0.3) #ä::h:# hhh dom räkna me äh uh uh hhh they counted with uh
09
hundrafemti ti tvåhundr#a# #ä::h# per vecka a hundred and fifty to two hundred uh per week
10
eller va re va rom sa; or what it was they said
11
CTA:
mm,
14
CTA:
.h[hh så då tyckte- #äh#= .hhh so then ((she/they/I)) thought- uh [°okej,°
16
MOT:
12
MOT:
13 15
MOT:
17
äh äh och han hade gått upp hundratjuge, uh uh and he had gained one hundred and twenty
=så då=äh då sa hon att= so then uh then she said that
=a:men då kan du ge han en flaska well then you can give him a bottle t:vå gånger per dygn. twice every twenty-four hours
18
MOT:
.h [ä::hm:
20
MOT:
ä:h .hh Å:=äh >så sa ja< A:, oke:j men de- kommer and .hh so uh I said yeah okay but it- will
19
CTA:
[°oke:j,°
78
Emma Betz and Arnulf Deppermann
21 22
de störa min amning nu då eller >liksom såhär< it disturb my breastfeeding now then or like this MOT:
23 24
ne:j nej nej, de behöver du inte va orolig för. no no no you don’t need to worry about that .hh ä:h men nu så: hah har de gjort de. .hh uh but now it has
CTA:
mhm:.
Within the mother’s report, the call taker produces two OKAYs (lines 14 and 19). Both are rather quiet and produced after the completion of the prior turn but not of the whole telling. Specifically, they appear in overlap with or shortly after an in-breath of the mother, who continues or projects continuation of her telling in overlap with the particle. These features indicate that both OKAYs are indeed heard as continuers, that is, as supporting the progressivity of the current telling. However, the call taker produces different vocal continuers before okej, and oke:j,: in lines 4, 7, and 11, mm, is produced. One may therefore ask how continuer OKAY differs from other continuers available to speakers. As already noted, the excerpt is a telling in which the mother introduces a problem, which is that the baby refuses to take the breast after having been bottle-fed for a few days (not shown in excerpt). One can note that both instances of continuer OKAY are produced after turns in which the mother reveals something that can be heard as problematic and as foreshadowing further problems: In line 12, the mother presents the first possible key issue, which is that the baby has been found to have gained less weight than expected. In lines 16–17, the information/problem solution that the mother reports as having received (namely to introduce the bottle) might be understood by the call-taker, based on her professional knowledge, as projecting a troubles telling. Thus, continuer OKAY seems to index a speaker’s understanding of certain pieces of information within a larger telling as particularly salient; in contrast, scene setting and background information (lines 1–3, 5–6, 8–10) are followed by more minimal acknowledgement tokens. We suggest that the choice between continuers enables recipients to distinguish between foregrounded and backgrounded information. Excerpt 13 illustrates a similar OKAY in a different sequential environment. In contrast to Excerpts 11 and 12, the particle is not used mid-telling but is in this German conversation employed as a response to an elicited informing, possibly facilitating the expansion of this informing into a telling. It thus shows a different type of orientation to non-completeness. Anna (ASC), a student interested in a room-for-rent in a shared apartment, is meeting with prospective flatmates. Three other participants have shared that they are studying to be teachers. With du auch? ‘you too?’ (line 1), Anna is inquiring whether Marvin (MLA) is also a teacher in training.
Chapter 3. OKAY in responding and claiming understanding 79
Excerpt 13. technikerschule ‘technical school’ (FOLK_WGCA_01_A01, 02:36–02:55), 2012, German, face-to-face, informal conversation
01
ASC:
02
MLA:
03
SLU:
05
MLA:
04
ASC:
06
SLU:
08
ASC:
07
09 10 11
asc
MLA: asc asc
+du auch? [((starts to laugh))] you too +gaze to MLA ----------------->> [ ich?=n]ee nee. me no no [thh HH HÄ] HÄ HÄ [hä hä hä ]
>ch ma[:ch die technikerschule g]r↑ad_ I’m studying at a technical school right now [.hhh
(0.22)
]
o*kee,* *head nod*
*(.) *slight head nod-->
für da*ten und informati↑onstechnik_ for data and information technology --->*
*u:nd* bin jetzt im abschlussjahr; and am now in my final year *slight nod*
Marvin expands his response to Anna’s polar question in line 5. His turn is ambiguous in status: It may be pragmatically complete (its syntax and the subsequent pause would suggest this) or it may be a first item in a longer informing, projecting more (the level intonation maintains this possibility). Anna’s OKAY-response to Marvin’s turn (line 8) carries slightly rising intonation and is accompanied by a head nod during and another slight nod after OKAY (line 9). It treats the prior turn as a first bit of information rather than as a complete informing. Speakers maintain eye-contact throughout the sequence, and after OKAY, Marvin indeed expands his turn with a syntactic increment: He specifies the referent technikerschule ‘technical school’, thus offering more detail about his field of studies, and then adds information about his degree progress. Similar to the OKAY uses described in Section 2, these kinds of OKAY responses to actions that inform seem initially to function as news receipts (Maynard 2003; Couper-Kuhlen 2012b; Koivisto 2015a) in that they are minimal and acknowledge the prior as news or new information. However, in orienting to the prior as not complete, OKAYs such as the one in Excerpt 13 invite more, and in this, they contrast with news receipts, which mark the potential completion of a sequence. They instead advance an informing sequence, a function they share with newsmarks and with what has been termed continuers in existing research (Schegloff
80 Emma Betz and Arnulf Deppermann
1982; Heritage 1984a, 339–344; C. Goodwin 1986; Gardner 1997, 1998; Sorjonen 2001, 2002; Maynard 2003). Using a response particle that, through lexical choice and placement, can be heard as “inviting” continuation or elaboration (rather than as either marking the informing sequence as potentially closed or as simply passing the floor) certainly fits a context in which, as is the case for Excerpt 13, a larger purpose for the speaker is to show interest in her co-participants and thus potentially increase her chances of being selected as the new roommate.17 The specific selection of OKAY in this context may be motivated by an additional consideration: the structuring of information.18 In our discussion of Excerpt 12, we suggested that OKAY may be used to foreground information within a larger telling.19 This may also be the case here. Research on responses to informings shows that lexical choice and prosodic design in particle responses matters for its function in a particular sequential context: For example, Heritage, summarizing Jefferson, notes that “all newsmarks project further talk by the news deliverer/newsmark recipient by reference to the news, but […] different newsmarks project different trajectories for such talk” (1984a, 340). Similarly, Thompson et al. (2015, 68–69) 17. We describe Marvin’s turn in line 05 as ambiguous with respect to whether it projects more. It is the following OKAY that treats it as not complete (cf. Sorjonen 2001, Chapter 8). Here is an example from English in which the turn which is receipted with OKAY itself clearly projects more. OKAY again follows a question-elicited informing, and the data (presented and analyzed in Heritage and Raymond 2012) come from a pediatric visit: [Pediatric Visit], Extract (17) in Heritage and Raymond (2012, 191), original transcription retained
01 02 03 04 05 06 07 08 09 10 11
DOC: MOM: DOC: MOM: MOM: DOC: MOM:
Has he been coughing uh lot? (0.2) .hh Not uh lot.=h[h [Mkay:?, But it- it deep. (1.0) An’ with everything we (heard) on tee v(h)ee=hhhh £we got sca:re.£ Kay. (And fer i-) It sounds deep? (.) Mm mm.
Here, the clinician’s Mkay:? (line 4) is produced in third position after an answer to a question that by design (through a negated repetition response) projects expansion, specifically the introduction of information that was not directly asked about (ibid., 190–192; cf. Stivers 2007). The particle carries rising intonation; it shows an understanding of the prior turn as projecting, indeed as launching, an expansion. 18. This echoes calls for further specification of the term ‘continuer’ (see, e.g., Sorjonen 2001), that is, a description of the varied context-specific work continuers do in addition to passing on an opportunity to take a full turn. 19. See also Helmer et al. and Koivisto and Sorjonen, both in this volume, and Oh and Park’s (2017) analysis of the differential interactional import of acknowledgement tokens ung and e in tellings in Korean.
Chapter 3. OKAY in responding and claiming understanding 81
show that particle responses in English can be produced with “affective lamination, cued […] via prosodic upgrading or downgrading” and can additionally convey an assessment or appreciation of the information responded to. We suggest that selecting OKAY among other resources that speakers have to respond to informings in the language of the conversation is a device for contextualizing a piece of information as important or substantial within the current activity. Excerpt 14, taken from a phone interaction in Québécois French, shows another example in which a speaker uses OKAY to invite elaboration. SIM is updating MAN on his (professional) activities since leaving the Navy. In line 3, SIM begins an announcement of news with j’suis passe- ‘I have gon-’, which he reissues in the clear in lines 8–9. Excerpt 14. responsable du niveau deux ‘teaching level two’, (CLAPI_Fraq_appel_ 5136_01:49–02:10), 2004, Québécois French, telephone conversation 01
MAN:
02 03
SIM:
04
MAN:
05
o:h mon dieu. oh my god
ouais, c’est [l’fun.] yes that’s fun [ ouais] !p j’suis [passe-] (well/yes) I have gon[il ] est he’s a
commandant lui? commander right?
06
SIM:
ouais.= yes
07
MAN:
=[ah ouais.] oh I see
08
SIM:
[j’suis I have
] pa:ssé de p’tits entraîneme::nts gone from little trainings
09
à: responsable du niveau deux. to responsible for level two
10
(0.3)
11
MAN:
13
MAN:
15
SIM:
12 14
16 17
MAN:
18
SIM:
okay,
(0.5) hUH.
(0.4)
c’fait que j’ferais rien fuck all but actually I’d do (nothing) fuck all (0.3)
c’est pas trop=h↓ein,= that’s not too much, right? =j’avais plus de responsabilité I had more responsibility
82
Emma Betz and Arnulf Deppermann
19
quand j’étais (elop) à baie comeau. when I was (a student) at baie comeau ((=place name))
20
MAN:
((laughs))[((laughs))
22
MAN:
U[gh HU:GH.
24
MAN:
facque ((=fait que)) là tu fais moins que rien_ so now you do less than nothing
25
SIM:
là j’fais crissement rien_ now I do absolutely nothing
21
23
SIM:
SIM:
[pis j’faisais rien. then I was doing nothing
[pis là j’vais avoir mai capitaines ben vite; and now I’ll have the rank of captain really soon
Through the formulations j’suis passé ‘I have gone/moved’ and niveau deux ‘level two’, SIM’s announcement is recognizable as good news, specifically of his recent professional advancement. Relevant responses to this include a positive assessment by MAN or, alternatively, a newsmark validating the announcement’s status as news and encouraging its elaboration (Jefferson 1980; Heritage 1984a; Maynard 1997, 2003). The latter would also advance the sequence toward a fitted (that is, more particularized) appreciation of the news, for example in the form of an assessment (cf. Maynard 2003, 108). In line 11, MAN produces an OKAY with prominence on the second syllable and slightly rising intonation. As in Excerpt 13, the OKAY allows its speaker to treat the prior as informative but the sequence as not complete. In this case, MAN’s OKAY conveys that she expects (and needs) more to provide a fitted appreciation of the announcement. SIM, however, does not immediately treat the okay, as showing MAN’s understanding that a news telling is underway, that is, as a kind of newsmark designed to advance a fuller telling. A gap emerges in line 12. MAN then produces what is hearable as a small laugh token in line 13. In its specific sequential context, the laugh token comes off as a generic assessment of SIM’s announcement. In being ambiguous as to the stance it conveys, the laugh token is an apt solution to the issue MAN seems to be facing in this moment: The stance conveyed by SIM in lines 8–9 is not unambiguous. By describing his path as de p’tits entraîneme::nts à: responsable du niveau deux. ‘from little trainings to responsible for level two’, SIM highlights a contrast between what one would expect as the normal course of advancement and what he himself experienced. It is thus not yet clear what exactly should be treated as newsworthy here: that he was promoted or how this happened. As SIM has not indicated a readiness to continue or elaborate (lines 10, 12), MAN finds herself in a position to appreciate the news without having enough information to provide a particularized uptake. She does this with an equally ambiguous hUH (line 13).
Chapter 3. OKAY in responding and claiming understanding 83
In line 15, SIM formulates the import of his announcement: What makes the news newsworthy is the contrast between SIM’s advancement to a more demanding level and the fact that he is actually doing less work. An extreme-case formulation (Pomerantz 1986) foregrounds this contrast, and this allows MAN to provide a fitting uptake, appreciating the irony conveyed (line 17, 20, 22, 24). In the present example, OKAY treats the prior as news but does not convey a particular stance (e.g., surprise) toward the news. In its placement and through its prosodic shape (slightly rising intonation), it encourages more talk from the prior speaker, specifically talk that advances a fuller delivery of the news and provides the basis for a fitting appreciation. 4. OKAY marks discrepancy of expectation Whereas Section 2 dealt with cases of sufficient understanding and Section 3 dealt with cases in which some new information or action was registered but indexed as not yet complete for local practical concerns, this section will show that OKAY can also index a lack of understanding by marking a prior informing as counter to expectation (cf. Heritage 1984a; Robinson 2009; Persson 2015; Thompson et al. 2015, 75–85; Seuren, Huiskes, and Koole 2018). The aspect of epistemic discrepancy between the OKAY-speaker’s expectation and the information gleaned from the interlocutor’s prior turn can be layered with additional affective and/or evaluative aspects, such as surprise, skepticism or perplexity. These OKAYs are thus often delivered with marked prosody. Excerpt 15 from a Polish business phone-call illustrates this. In line 1, the client (CLI) announces her intent to place an order and inquires as to whether this is possible on the phone. The salesperson (SAL) then informs her that the company needs to have a confirmation of every order via email (lines 3–4). The long silence in line 5 points to a problem on the part of the client. Excerpt 15. złożyć zamówienie ‘place an order’ (PhoneVN680234_QA okeeeeeej), 2009, Polish, telephone, service encounter 01
CLI:
02 03 04
ja: chciałam złożyć zamówienie. I wanted to place (an) order
czy mogę to zrobić telefonicznie:? can (I) do it (on the) phone SAL:
.hh ym potrzebujemy m:ieć (.) potwierdzenie .hh uhm we need to have (a) confirmation takiego zamówienia na mejla:. of such (an) order in (our) email.
84
Emma Betz and Arnulf Deppermann
05 06
CLI:
07
SAL:
(1.8)
a::: (2.4) hy:::::::: (2.4) .hh (1.0) and PRT .hh and huh .hh
a proszę mi powiedzieć co pani dokładnie and please tell me what exactly you
08
chce zamówićto od razu może (.) moglibyśmy want to order then maybe at once (we) could
09
moglibyśmy ustalić koszty transportu. determine transport costs
The salesperson’s informing in lines 3–4 could be taken to imply that placing an order on the phone is not possible at all, or that it is possible but requires an additional record in the form of an email sent by the client after placing the order. The salesperson’s information may be contrary to what the client could have expected: that it would be possible to place an order over the phone. In any case, from the client’s point of view, there is trouble with “the sequential import of the utterance or turn as a whole […], and accordingly what type of talk/action is relevant or appropriate next” (Schegloff 1987, 206). After a pause of 1.8 seconds, the client produces an OKAY (line 6). Its formal realization is highly marked: Both syllables are stretched and accentuated, and it is delivered with slightly rising intonation. The client continues her turn with a::: ‘and’, which may project a follow-up question designed to clarify the procedural consequentiality (Schegloff 1991) of the salesperson’s informing for her future action, namely the ordering procedure; yet, this projection is not fulfilled. Instead, further pauses and the hesitation marker hy point to trouble with the action implications emerging from the salesperson’s response. The caller clearly has trouble proceeding with the reason for the call. In the next turn (lines 7–9), the salesperson offers steps that can be taken on the phone now. Excerpt 16 provides another example of OKAY making visible a mismatch between prior expectations and just-received information. It comes from the beginning of a phone call in Danish between MOR (mother) and BO (daughter). An earlier call had been interrupted by BO’s phone turning off. The call below resumes this earlier conversation, and with hva ↑sker der. ‘what’s going on’ (line 2), MOR is eliciting an explanation for the interruption. Excerpt 16. tænder å slukker ‘turns on and off ’ (samtalebank:bilen:00), 2009, Danish, telephone, informal conversation
01 02
MOR:
03
BO:
(1.0)
hva ↑sker der. what’s going on jamen min telefon den slukkede;=hhh heh °heh° well my telephone it turned off hhh ((laughs))
Chapter 3. OKAY in responding and claiming understanding 85
04
(0.2)
05
BO:
07
BO:
08
MOR:
09
BO:
=her en af dagene_= one of these days
10
BO:
=fordi: dn den tænder å slukker sån i ↑utide_ because it it turns on and off like out of control
06
.hhh
(0.2)
ja[m (ve’ du hv]a:) jeg ska ned me’ den= well (you know what) I am going down with it [nå::? ] oh
11
#å:# den sætter alarmen sådn .hhh (.) til: her; an’ it sets the alarm like .hhh here
12
å den slår den fra:_ å:=hhh an’ it turns it off an’ hhh
13
(0.5)
14
MOR:
15
BO:
16
BO:
når jeg ska op om morgenen_= when I need to get up in the morning
17 18
BO: MOR:
=hhh hnh [h(n)] [hnh ] heh heh heh he-=
MOR:
£_o[:↑`´kay,£
19 20
BO:
21
BO:
23
BO:
24
MOR:
22
25 26
MOR:
27
BO:
28
MOR:
29
BO:
30
MOR:
31
BO:
[ja: men det da ] yeah but that’s certainly [(å jeg har jo) ( ] ) alarmen sat te’ (an’ I have) ( ) the alarm on
=.hhnh
[(hm:./hnh.)
(0.8)
°så_°= so =ja: ↓ja:, yeah yeah (0.3)
#a:rhm’# der må da være noget garanti på (.) well there must surely be some guarantee on ja[: ]men det er der oss_= yeah but there is [den_] it =faktisk h[am jeg] har købt den af nede i butikken.= actually the guy that I bought it from in the shop [ja:_ ] yeah =ham har jeg på et hold nede i °håndboldklubben°_ I have him on a team in the handball club
Frequency (kHz)
86 Emma Betz and Arnulf Deppermann
5 4 3 2 1
Pitch (Hz)
80
300 200 105
60
ↄ
kh
Intensity (dB)
400
i
hm: 19 Time (s)
Figure 6. Pitch, intensity, spectrogram and soundwave of £_o:↑`´kay£, in line 20
BO’s answer in line 3 provides a first explanation, followed by laugh tokens and a short silence. MOR’s delayed nå::? ‘oh::?’ (line 8) receipts BO’s explanation as news. In overlap, BO expands her answer by describing plans to solve the problem (going to a repair place, lines 7, 9) and characterizing her phone issues as recurrent (lines 10–12) and potentially significant (lines 15–16). When troubles are relayed, participants tend to indicate how seriously these should be taken (Jefferson 1984c). BO produces laughter tokens in line 17 and can thus be seen to downplay the trouble (ibid.). This opens up for MOR the option of either joining the troubles-resistant line, taking a “time-out” from the trouble (ibid., 351) or treating the trouble as serious, being “troubles-receptive.” MOR chooses the former option and joins the laughter (line 18). She then produces an OKAY in line 20. The particle is produced with smile voice, and its prosody is ‘marked’ for Danish: It has a pitch upstep from the first to the second syllable and a slight fall-rise contour on the second, stressed, syllable (see Figure 6). In overlap with OKAY, BO produces what could be a further laugh token, and after a silence (line 22), she expands her prior turn with a stand-alone °så_° ‘so’ (line 23), possibly projecting an upshot that will not be produced (cf. Raymond 2004 for such uses of so in English). MOR then formulates an assumption (lines 26, 28) which can be heard as a suggestion for solving BO’s problem. It is also hearable as a sort of objection to the implicit claim that BO’s trouble is just something she has to live with, an impression MOR could have gotten from the way in which BO has presented her troubles as ‘no big deal.’ Additional evidence for MOR’s at least slightly disaffiliative stance on how the problem should be treated can be found in the repeated ja ‘yes’ (line 24; see
Chapter 3. OKAY in responding and claiming understanding 87
Heinemann 2009) and the turn-beginning item a:rhm’ (line 26), a variant of ja men ‘yes but’ or nej men ‘no but’ (Steensig and Asmuß 2005; Pedersen 2015). MOR’s smile voice on OKAY resonates with a ‘no big deal’ interpretation of BO’s stance. The particle is produced with rising intonation, however, which seems to indicate some kind of counter-expectedness and thus non-affiliation with BO’s trouble-resistant stance. This is corroborated by the fact that MOR shifts to a serious mode after OKAY by treating BO’s trouble as in need of a solution. OKAY can be said to index a discrepancy of expectation in the sense that the OKAY-speaker has, on the basis of her interlocutor’s account and her own assumptions, arrived at an assessment of the seriousness of the trouble that is different from the troubles-teller’s. While MOR’s OKAY indexes this discrepancy, it does not explicate its precise nature or its basis. Therefore, in order to enter into a negotiation of how to deal with and how to assess the trouble, a more explicit account is in order, as it is given by MOR’s turn-continuation. In Excerpt 15, an informing which is discrepant with the OKAY-speaker’s expectations leads to a cooperation problem. In Excerpt 16, OKAY indexes that its speaker does not share a co-participant’s assessment of a reported state of affairs. An unexpected informing can, however, also affect the intelligibility of a story. This is the case in the next excerpt from a phone-call between two female friends in German. EG has just told FR that her cat had to be saved from falling off the roof by EG’s boyfriend. FR’s responds to this in lines 1–3. Excerpt 17. gekipptes fenster ‘window ajar’ (FOLK_E_00084_SE_01_T_01_DF_01_ c130), 2014, German, telephone, informal conversation 01
FR:
.h jet jetzt weißte ja dann kannste des .h no- now you know well then you can’t
02
fenster ja nichmal me:hr hier auf ↑kipp machen even partly open the window any more
03
wenn [der ( )] if the [ ja=aber das] wollt ich (well/yeah) but I wanted that
04
EG:
05
ich wʔ äh=hä(n) °wollt ich ja auch nie;° (.) I wʔ uh=huh(n) never wanted that (y’know)
06
also grade in der küche. weil das=äh I mean especially in the kitchen. cause that=uh
07
[weil da eben] das dach is; cause there’s obviously the roof ] [ okay,
08
09 10
FR:
EG:
un dann: (.) die £checken des ja nich_£ and then (you know) they don’t get that dass es da runtergeht. (.) that it goes down there
.h .h
88
Emma Betz and Arnulf Deppermann
11 12
EG:
13
FR:
14 15 16 17 18
EG:
(0.4)
weißt[e, you know
]
[(unʔ)](0.5) wie: ist des dann: passiert?=hh (and) how did that happen then hh (0.2)
ja: keine ahnung. (.) das wʔ äh=also s war well ((I’ve)) no idea. it w uh=(so) it was ja (0.2) des war halt äh (0.5) °en stückchen (y’know) it just was uh slightly offen_° open (1.3)
In response to EG’s story, FR remarks that EG can now, as a result of the reported event, no longer keep the kitchen window open (line 1–3). She thus shows that she presupposes that an open window must have been the reason why the cat managed to escape onto the roof. In lines 4–7, EG responds that she never wanted to open the window, especially not in the kitchen (from where the cat escaped). Still in overlap with EG’s explanation about her intentions, FR produces okay, (line 8) with rising intonation, realized as a diphthong, with a stress on the first vowel of the diphthong. It indexes that FR claims to have understood EG’s statement about her intentions on a propositional level, but not how it matches the context of the story so far.20 From EG’s statement in lines 4–7, FR seems to have drawn the inference that, because of EG’s stated intentions, the window was in fact closed. This inference, however, is discrepant with FR’s assumption that the cat must have escaped through the window. As Seuren, Huiskes, and Koole (2016, 183) note, the okay, only “indexes but does not identify a problem”, that is, it does not indicate why the new information is at odds with FR’s understanding of EG’s story so far. Yet, after EG has finished the account of intentions concerning keeping the windows closed (line 12), FR topicalizes the problem (that which okay, had only indexed) with wie: ist des dann: passiert ‘how did that happen then’ (line 13). Because EG had continued the account of her intentions after FR’s okay, without attending to the epistemic discrepancy that the particle had marked, FR’s explicit query now asks EG to resolve the epistemic puzzle. In response to this query, EG explains that the window was open after all (lines 15–17), which confirms FR’s earlier assumption. This explanation reveals that in saying ‘I never wanted that [the window is open]’ (line 5), EG had implied the opposite implicature
20. See Betz et al. (2013) concerning the difference between registering new information and claiming to understand its pragmatic relevance.
Chapter 3. OKAY in responding and claiming understanding 89
(+>21 the window was open) from the one which FR had inferred (+> the window was not open). The misunderstanding and consequently the discrepancy of expectation in this case thus rest on competing inferences from a narrative. The prosodic shape (a rising pitch contour) seems to be a crucial feature of OKAYs marking a discrepancy of expectation (cf. Selting 1988, 1996, on repair-initiation in German). It is a stable feature across the different languages in which we could find OKAY in this function.22 In many cases, there is additional prosodic marking by heightened volume, strong accentuation of one or both syllables of OKAY, high onset and/or an especially high rising contour.23 These kinds of OKAY responses mark an epistemic mismatch between already communicated, known, or expected information and just-conveyed information. They are not sequence-closing but engender elaborations and accounts in next turn. They are thus similar to next-turn “open class” repair-initiators (Drew 1997) such as huh? (Dingemanse, Torreira, and Enfield 2013) or English what? (Robinson 2014). However, in contrast to these items, OKAY does not index ‘not understanding’ but rather conveys that the prior information has been registered (i.e., understood on a propositional level), but is not sufficient for present purposes, because it is at odds with what was expected. Non-sufficiency can operate on a pragmatic, interactional level (as in our Polish example), if the OKAY-speaker cannot glean a sufficiently clear projection for their own next action from the prior turn or if the prior turn interferes with projected next actions or an interactional goal of the OKAY-speaker. It can also operate on a topical level (as in our German example), if new information cannot be accommodated within the OKAY-speaker’s prior assumptions about the state of affairs talked about, thus leading to contradictions, incoherence, or lack of motivation for reported events. The epistemic mismatch is both temporal and interpersonal: Temporally, the
21. ‘+>’ denotes ‘implicates’ (see Levinson 1983, 104–108). 22. We did not find this use of OKAY in all the languages that we investigated. Existing research (see, e.g., the discussion in Keevallik 2003 on Estonian) suggests that the prosodic form – interactional function patterning we observe here may be less clearly binding in other languages and/ or limited to certain language families. 23. See also Selting (1987, 130–140) on prosodic marking of TCUs indexing a discrepancy of expectation and Thompson et al. (2015, 75 –83) on certain rising-intoned particle responses in English and counter-expectedness. In an experimental study on the prosody of American English OKAY in particular, Van Zyl and Hanekom (2013) identified word duration as an important cue for interpreting a response as reluctantly (v. unreservedly) agreeing. Similarly, Beach (2020) identifies increased syllable length, accentuation, final rising pitch movement, and high pitch onset as recurrent features of OKAYs that “accomplish a range of incongruous actions” such as projecting disagreement or indexing that some statement or state of affairs is odd, bizarre or incredible.
90 Emma Betz and Arnulf Deppermann
new information is in contrast to given information or prior existing assumptions on the part of the OKAY-speaker; interpersonally, there is a mismatch between the prior speaker’s and the OKAY-speaker’s epistemic states. While OKAY indexes this mismatch (without, however, explicating its nature and its sources), it strongly makes relevant a resolution of the mismatch by the prior speaker as the next turn. By assigning the responsibility to solve the mismatch to the prior speaker, it can carry a more or less acute moral connotation of blaming the prior speaker for producing a turn that was not intelligible, credible and/or acceptable and for having failed to prevent the discrepancy of expectation from developing. In contrast to the use of OKAY as a continuer or newsmark, the prior speaker is not expected to proceed as they planned independently from the OKAY, but the OKAY acts as a kind of repair-initiator (Schegloff 1997), calling for an account or additional information which provides for the intelligibility, credibility and/ or acceptability of the prior turn, thus restoring intersubjectivity in terms of its action-projecting potential and/or its informative value. 5. Summary and conclusion Uses of OKAY within and across languages can show different orientations to a prior informing action and thus have different sequential implications. This overview chapter, and the in-depth individual studies that follow in this volume, illustrate that the sequential implications of OKAY responses to informings rest on a combination of contextual, prosodic, and sequential factors, as well as its positioning in the larger activity (Thompson et al. 2015, 52; cf. Gardner 1998, 2007; McCarthy 2003). We have shown that OKAY is directly implicated in managing matters of understanding and intersubjectivity in interaction. There is a continuum of uses in terms of the degree to which an understanding which is locally sufficient with regard to the practical concerns of the interaction at that moment has been achieved. OKAY in second position or third position, closing a question-answer or some other adjacency-pair sequence, produced with falling intonation, can index sufficient understanding. “Sufficiency” means that the understanding gained allows for interactional progression, that is, for the closing of a topic and/or the move to next actions or topics. In this sense, sufficient understanding paves the way for closing (see Chapter 4). These next actions may be scheduled on an agenda or emerge from the prior sequence, pragmatically or logically building on it. OKAY with level or rising intonation, however, indexes that some action or new information has been registered (i.e., understood on a propositional level) but is not yet sufficient for local pragmatic concerns. “Insufficiency” means that the prior action is incomplete, because it does not fulfill the informational needs
Chapter 3. OKAY in responding and claiming understanding 91
of the OKAY-speaker or the conditional relevance which has been set up by the informing speaker comprehensively enough, or because its relevance to the larger project, argumentation or discourse topic is not yet evident. This epistemic expression of insufficiency combines with the interactional projection/expectation that the prior speaker continue their turn in order to remedy this insufficiency. OKAY thus serves to invite, encourage, or even push for continuation. Thus, in contrast to what have been described as continuers (C. Goodwin 1986; Gardner 2001; Schegloff 1982; Sorjonen 2001), these OKAYs are not just a floor-pass to the prior speaker to continue what they have projected or planned, but more specifically require them to provide for the means to engender a pragmatically sufficient understanding of what they have already said. The other extreme of this continuum ranging from sufficient over yet partial, incomplete understanding to non-understanding in a pragmatic sense is inhabited by OKAYs which display a discrepancy of expectation. These OKAYs are also produced with rising intonation, but tend to have a more expressive design, which means that they are louder, cover a greater pitch range, or are stretched (cf. Beach 2020). Like the other two variants, they index propositional understanding as well. Yet, they do not only indicate that something is missing but also convey that the interlocutor’s prior turn runs counter to an expectation which the OKAY speaker has had and maintained, given the sequence so far. This expectation of the OKAY-speaker may have existed already before the interaction, or it may have been touched off by more remote prior turns of the interlocutor or even by the prior turn of the interlocutor, seeming self-contradictory to the OKAY-speaker. Relevant expectations can concern different cognitive and interactional orders: Assumptions about events and states of affairs, about the partner’s assessments and emotions tied to reported events, or about expectations about possible, appropriate, or required next actions. OKAY indexes that expectations which the OKAY-speaker claims to hold accountably are not fulfilled. Again, this cognitive, expressive component combines with pragmatic and interactional properties: OKAY here projects disaffiliation or disalignment with the prior speaker, which again can concern different levels: propositional credibility, interactional cooperation, emotional and/or evaluative stances. OKAY creates an interactional projection that the interlocutor remedy this discrepancy, for instance by giving an explanation or offering a justification, or by backing down from their position. Yet, as OKAY does not by itself specify the nature and the source of the problem which prevents the accomplishment of intersubjectivity, it is often only an initial index of an intersubjective mismatch. It requires subsequent clarifying explications by the OKAY-speaker in order to resolve (or to accentuate) the rupture of intersubjectivity.
92
Emma Betz and Arnulf Deppermann
Acknowledgements Thanks to Betty Couper-Kuhlen for her close reading and insightful comments and to Sam Schirm and Johanne Léveillé-Schirm for additional help with transcription.
Funding Emma Betz’s work was supported by Canada Foundation for Innovation funding for the creation of a Social Interaction, Language, and Culture Lab at the University of Waterloo (CFI-JELF project #37510).
Chapter 4
OKAY in closings and transitions Lorenza Mondada and Marja-Leena Sorjonen University of Basel / University of Helsinki
With contributions from Emma Betz, Elizabeth Couper-Kuhlen, Elwys De Stefani, Arnulf Deppermann, Katariina Harjunpää, Kaoru Hayano, Henrike Helmer, Leelo Keevallik, Mary Shin Kim, Stephanie Hyeri Kim, Aino Koivisto, Satomi Kuroshima, Seung-Hee Lee, Xiaoting Li, Anna Lindström, Ana Cristina Ostermann, Søren Sandager Sørensen, Jakob Steensig, and Matylda Weidner This chapter discusses OKAY in transitions and closings, based on findings in the data from 13 languages that are analyzed in the project on OKAY. The chapter shows that OKAY features in a variety of languages in closing environments at different levels of organization, where the particle might occupy diverse positions and do similar but also complementary work across the languages. The sequential and activity environments to be discussed include (i) sequence closings, (ii) activity closings, (iii) transitions between activities or topics, and (iv) the closing of an entire conversation. Keywords: closing, pre-closing, transition, second position, third position, completion
1. Introduction Completing or closing an interactional Gestalt, which is often multimodally organized (Mondada 2015), is a pervasive practical problem in the emergent organization of social interaction. Speakers project and interlocutors anticipate completion of turn constructional units, as well as turns (Sacks, Schegloff, and Jefferson 1974). They make use of positional and compositional grammatical resources (Schegloff 1996), as well as, in face-to-face interactions, embodied resources for doing so (Sacks and Schegloff 2002; Schegloff 1984). Sequences – minimally constituted by an adjacency pair, or shaped in more complex and expanded turns (Schegloff 2007) – are also brought to closing, in second, third or further position, in ways that are intelligible for the participants. Particles feature importantly in this environment, such as voilà in French (Mondada 2018b), nå in Danish (Heinemann 2017b), and joo in Finnish (Sorjonen 2001). Series of sequences (Schegloff 2007), stories (Jefferson 1978), complaints, trouble tellings (Jefferson and Lee 1981), and other more complex sequential arrangements https://doi.org/10.1075/slsi.34.04mon © 2021 John Benjamins Publishing Company
94 Lorenza Mondada and Marja-Leena Sorjonen
are also closed by specific resources, among them particles. In institutional settings, “phases” of the activity are opened and closed in orderly and recognizable ways, for example, when moving from the openings to the presentation of the problem in consultations in health and therapeutic settings (see R. Turner 1972; Modaff 2003; Robinson and Heritage 2005). Finally, the conversation as a whole is brought to closing, after the pre-closing sequence has been completed and no further mentionable has been introduced (Schegloff and Sacks 1973). In all these sequential environments at different levels of organization, transitioning and closing is a situated and emergent accomplishment. For doing so, participants mobilize a series of linguistic (including phonetic-prosodic) and embodied resources that might be specific for the kind of sequential environment, of activity, of setting and of interactional issues going on. Among the lexical resources contributing to closing, particles play an important role, in particular OKAY (Beach 1993, 1995a, 1995b). This chapter illustrates the use of OKAY in different languages at different levels of organization, from sequence and activity transitions and conversation closings. 2. Sequence closing In our dataset of 13 different languages, OKAY participates in closing the smallest sequence of actions, the adjacency pair, occurring either in second or third position of the sequence. In these cases, due to the tight organization of the sequence, the closing work of the particle is perhaps among the clearest ones. OKAY can form either a turn of its own, or be located in turn-beginning, to our knowledge mostly as a prosodic unit of its own. 2.1
Second position
As most response particles, OKAY participates in completing an adjacency pair by providing a verbal second pair part to the preceding first pair part. In these cases it can respond to a request that makes it relevant by the recipient to give permission or grant the request verbally. Beach (1993) gives the following example of okay granting a request to borrow the car. Excerpt 1. Car (Beach 1993, 330; Sacks 4/1/72, 16), American English 01 A: can I borrow your car? 02 B: when?
03 A: this afternoon. 04 B: okay.
Okay does a double task here: it provides a preferred verbal response to the prior turn, and in so doing treats the sequence as closed.
Chapter 4. OKAY in closings and transitions 95
A central sequential context of a second-position OKAY is one where it provides a response to a directive turn, to one that, for example, requested the OKAY speaker to do some action or instructed them to do some action, or proposed a joint action (see Keevallik and Weidner, this volume; Couper-Kuhlen, this volume; Chapter 3, this volume). In the cases we will show next, the action nominated in the directive turn is something that will be executed after the OKAY response, either at some point later in the on-going interaction, or after the current interaction. When a request for embodied action has been made in face-to-face interaction, the immediate response can be implemented with OKAY, while the requested action will be done after OKAY, or at some point after the current encounter (see about such uses of OKAY in shop encounters, Merritt 1978; in car conversations, Rauniomaa and Keisanen 2012, 838; Couper-Kuhlen, this volume; Keevallik and Weidner, this volume).1 When the action nominated is to be done some time after the current encounter, OKAY participates in assuring that compliance is under way or will follow. The following excerpt contains OKAY as a response to a request for immediate action in a phone call in Estonian. Ene is calling to a congregation to arrange for someone to bring a tape recorder to the aerobics class she is running. She first talks with the person who answered the phone. That person however cannot help her (lines 1–3). Excerpt 2. anna mulle Reet ‘give me Reet’ (KELDER_K3B1), 1998/99, Estonian, telephone, informal conversation
01 HEL:
=ta ei sa- ta ei saa makki tuua. she NEG ca- she NEG can tape.recorder.PAR bring.INF she canno- she cannot bring the tape recorder.
02
ta n- ta on siin; she i- she is here she is here;
03
me ole-me siin; varju-s präegult. we be-1PL here NAME-INE now we are here; at Varju ((church name)) at the moment.
04 ENE:-> aa; anna mu-lle reet sis. PRT give.IMP I-ALL NAME then oh; give me Reet then. 05 HEL:=> okei; 06
07 REE:
(9.5)
eaa_ PRT yeah_
1. See Mondada and Sorjonen (2016, 737–739) on cases in which the seller, who is processing the products that the customer has picked up in the shop, registers the customer’s verbal request for a product with a response token, implying that the request will be fulfilled later.
96 Lorenza Mondada and Marja-Leena Sorjonen
08 ENE:
.h tere reet; ENe si[in]. hi NAME F NAME F here .h hi Reet; this is Ene.
Ene then requests to talk with another person (line 4). The okei (line 5) is a straightforward display of compliance, after which the other person is summoned to the phone. The following excerpt from an encounter in French in a cheese shop also contains a request for immediate action. The seller is slicing cheese and while doing that, she asks the customer to tell her when she has cut enough slices (line 2). Excerpt 3. sandwich (PA1007_CLI21 2–20–42), 2016, Paris, face-to-face, shop encounter
1 2
sel
SEL:
3
CUS:
4
SEL:
5
sel
(2.9) >>slicing cheese->
donc vous m’arrêtez quand vous l` souhaitez so you stop me when you want euh: bah j’vous laisse en faire deux autres ehm well I let you make two more ↑oké.
(2.0) ->cuts two more slices->>
The seller presents her request when she is just preparing to cut the next slice (line 2); this timing gives the customer enough time to tell when enough slices have been cut. In response, the customer asks the seller to cut two more slices (line 3). This is received by the seller with an OKAY produced with a high-onset and a final fall (↑oké.) that treats the request as having provided understandable and sufficient information for the embodied action to be produced. In so doing ↑oké announces the imminent compliance with the request, as the complying action is already ongoig. It also works to close the sequence of dealing with the size of the purchase. Within an activity progressing as a series of instructions, it is relevant for the recipient to display at a possible completion of each instruction that the instruction has been understood and will be followed. In the following excerpt from an American English phone call, Alan is providing Myra instructions on how to drive to his place later (remote action, Lindström 1997, 2017, 2018). These instructions are responded with okay by Myra (lines 3 and 5). Excerpt 4. Directions to Alan’s place (Kamunsky 3, 5–6), 1960s, American English, telephone, informal interaction 01 02 03 04
ALA: MYR:
=’kay< then turn right to
vineland [en then] turn left on vineland. [okay, ] (0.4)
Chapter 4. OKAY in closings and transitions 97
05
MYR:
le:ft o::n vineland. o[ka:y,
07
ALA:
uhh-hh-huh-˙hhe:_h uhh-heh-heh-˙hhi:_::h ahh[hee-hee]
06
ALA:
[an’then just go:.
Alan’s first instruction in lines 1–2 is received by Myra with okay, carrying a slighty rising pitch at the end (line 3). It is produced after the possible completion of the instruction (the name of a street to be turned to). Simultaneously with this response, Alan proceeds without a break to the next item, thus treating the okay as an appropriate and sufficient response. The second OKAY (line 5), on the other hand, is produced by Myra as a turn-final turn-constructionl unit (TCU) in her response to Alan’s subsequent instruction. The oka:y follows a repetition of the prior instruction, produced slowly by lengthening the vowels, which possibly manifests that Myra is taking notes. In overlap with the beginning of the second syllable of o[ka:y, Alan proceeds (line 6) to a new instruction that indicates that the specific instructions have now been given, treating Myra’s o[ka:y, as an unproblematic and sufficient response to them. Similar to its use as a response in other types if larger activities consisting of a series actions, such as when a series of pieces of information are given (see Chapter 4, this volume), OKAY here is a resource used for claiming and projecting compliance with directives that step-by-step instruct the recipient to act in a specific future situation. In sum, OKAY can be used as a claim of compliance in the second pair in both telephone calls and face-to-face encounters, and as a response to a directive concerning an immediate and a deferred action. 2.2
Third position
OKAY is found at a later stage in a sequence, as a minimal post-expansion of an adjacency pair in third position. OKAY treats (as a TCU of its own) the preceding response as accepted and thereby closing relevant. As Schegloff (2007, 118) points out, “minimal” here does not have to do with the size of turn but with the fact that the turn forms a potential closing of the sequence and does not project any further talk. OKAY can follow either a preferred or a dispreferred second pair part, claiming to accept it, and in so doing treat the sequence as potentially closing relevant (see e.g., ibid., 120–123; Beach 1993). The types of sequences encompass a variety of types (directive, as well as question – answer sequences). Davidson (1984, 127, fn. 6), for example, reports a use of OKAY after an invitation or an offer has been made and rejected (the transcript is from the original publication, with slight adaptations to our conventions):
98 Lorenza Mondada and Marja-Leena Sorjonen
Excerpt 5. Trip (Davidson 1984, 127, fn. 6), American English 1
A:
you wan’ me bring you anything?
3
B:
no: no: nothing.
2 4
A:
(0.4)
AW:kay.
The particle in the excerpt above displays that its speaker – who is the one who made the offer or the invitation – is aligning with the rejection, choosing not to expand the sequence with a subsequent version of the initial action. The following example is from a phone call in American English. Excerpt 6. Richard’s for lunch (NB 028, 29_39–43), 1960s, American English, informal interaction 01
LOT:
↑w’l I jus thought mayb we c’d go over to richard’s
03
EMM:
awri:ght?
05
EMM:
02 04
LOT:
for lunch then after I get my hair ↓fixed. oka:y,
.hh I: C’D WA:LK DOW:N MEETchu:,
In this excerpt, a proposal for having lunch together (lines 1–2) is responded with a display of acceptance by Emma (line 3). Following that, Lottie receives Emma’s turn with OKAY (line 4). By producing OKAY, she reconfirms the agreement arrived at in the proposal sequence, and by producing OKAY only in her turn instead of continuing, she transfers the turn to Emma. Emma now moves to the arrangements of getting to the lunch venue. The fact that oka:y, is produced with a slightly raising intonation shows that the particle closes the local sequence sequence but through its slightly rising final contour, a space for a possible pursuit of the larger sequence with further arrangements is opened. In the example above, OKAY as a third position element formed a turn of its own. In the next turn, the co-partipant treated OKAY as a closure of the sequence by proceeding to start a new sequence of actions that were related to the completed sequence. OKAY can also form the first element and action in a multi-unit turn, as in the following example from a telephone call from the study on English interactions by Beach (1995a; the transcript from the original publication, with slight adaptations to our conventions). Excerpt 7. (UTCL: J10.1, Beach 1995a, 136), English
01
ALL:
allan
03
ALL:
hi
02 04 05
TUP: TUP: ALL:
hi: this is tuppel. you r(h)eady for today’s go rou:nd? sure h
Chapter 4. OKAY in closings and transitions 99
06
07
TUP:
oka(h)y hih hih hhhh well- I just had a call from
joe and he says ((continues))
After Tuppel has received Allan’s response to his checking question as acknowledged, he moves through a pivotal laughter and breathing to a new activity and topic, starting to report a telephone call he has just received, thereby showing his treatment of the preceding sequence as closed (see Section 4 below on pivotal usages of OKAY). OKAY is also deployed in third position in a sequence that its speaker initiated by requesting information or confirmation that they needed for doing some main activity. In such a case, OKAY displays that the information provided was sufficient for proceeding to the main activity, and it proposes to close down the preparatory sequence and activity. This happens in the example below where a chef in a Japanese sushi restaurant (Chef 2, line 1) checks his understanding of the number of soups that the customers ordered from his recipient (Chef 1, line 2). Excerpt 8. Miso soup (UH_09_21_05 00:11:48), 2005, Japanese, face-to-face, institutional interaction 01 CH2: ch2
02 CH1: ch1
*hitotsu:, one one (soup). *hand gesture of “one” +hitotsu de::, one by one, +hand gesture of “one”
03 CH2:=> ’k↓ke. 04
((CH2 goes to the kitchen to prepare the soup))
Here a display of the sufficiency of the information for fulfilling the order, and a display of compliance with the order are intertwined in the OKAY response, produced with a fall on the final syllable. The evidence of the closing character of OKAY is provided by the subsequent embodied action done by its speaker (Chef 2 going to the kitchen). Sequence closing has been reported in the literature on interactions in English as being done by a variety of actions and resources in third position, which can be combined with OKAY. These include assessments and change-of-state tokens. For example, OKAY was part of the high-grade assessment sequences that were frequent in the quality-of-life interviews data of Antaki, Houtkoop-Steenstra, and Rapley (2000). In this case, the sequences took the form [answer receipt] + [right / ok] + [high-grade assessment] + [move to next item]. OKAY and assessments are also found in classroom interactions where the three-part sequence Inquiry-ResponseFeedback/Evaluation (IRF/E) has been recognized as an established pedagogical
100 Lorenza Mondada and Marja-Leena Sorjonen
practice for long time (Mehan 1979; Sinclair and Coulthard 1975). In classroom, OKAY with falling intonation can constitute an acceptance of an answer (Waring 2008). It can also be combined with assessments and used in third position to acknowledge the student’s answer and close the sequence, also disengaging bodily from interaction with that student (Fagan 2012; Waring 2008). This is however not always the case – sometimes the sequence is expanded by the teacher asking for further specifications (Waring 2008). In both types of institutional situations in which the use of OKAY was just discussed, quality of life interviews and classroom interactions, OKAY occurred before the assessment. This position of OKAY may make one wonder about its force as a closing implicative element. These studies however mention that assessments are conventionalized in these interactions as closing implicative, and they state that sequences containing assessments “sound institutional” (Antaki et al. 2000) and that assessments are treated as closing implicative (Fagan 2012). However when OKAY combines with a change-of-state token such as oh in English, or aa in a number of languages (see, e.g., Koivisto 2015b; Helmer, Betz, and Deppermann, this volume; Chapter 3, this volume), the order is fixed, so that OKAY is the last element in the sequence. This final position indicates the order in which the preceding turn and sequence is treated: dealing first with the particular contribution of the prior turn (e.g., epistemic), and thereafter the relevance of the sequence for the progression of the interaction. In the previous three cases, OKAY occurred in third position responsive to a prior verbal action in a sequence. It is however also used in sequences where the preceding action is produced in an embodied fashion. In the following excerpt, the Okay in line 9 occurs in an instruction and request sequence. It is produced as a response to an embodied compliance by a child patient to an instruction by his uncle how to sit to place his arm properly with respect to a blood pressure monitor in a Korean emergency room. Excerpt 9. Stay still (ER 10.23–10_0:45), 2011, Korean, face-to-face, institutional interaction 01
UNC:
tway-ss-e tway-ss-e ( become-PST-IE become-PST-IE okay okay you can do ( )
02
NUR:
03
UNC:
uum [ike (.) son ( ) uhm this hand uhm this (.) hand ( ) [kotaylo ham ca (.) ccwu:k ccwu:k ca as:is once well straight straight well just like that well (.) strai:ght strai:ght well
04
) hayto-tway do-become:IE
(ike) tasi ttokpalo anca pwa, this again straight sit try:IE (this) seat yourself straight again
Chapter 4. OKAY in closings and transitions 101
05 06
UNC:
07 08 09 10 11 12 13
(0.8)
ttokpalo anc-a ttok:palo kulay kuleh-ci ttok:palo straight sit-IE straight do:so:IE do:so-COMM straight sit straight. strai:ght. that’s right sit strai:ght. anc-a twa:y-ss-e. sit-IE become-PST-IE goo:d. ((Child places his arm in the monitor as instructed.))
UNC:=> okay kulay iss-e: like:that exist-IE like that:, (0.8)
UNC:-> kotay:lo iss-e: as:is exist-IE stay just li:ke that: NUR:
(3.0)
myech khilokulaym naka-yo? how:many kilogram weigh-POL how many kilograms does he weigh?
When the child has inserted his arm into a blood pressure monitor, his uncle instructs him first to correct the position of the arm, and then requests him multiple times to sit straight, helping him to sit in the right position for the cuff (lines 3–4 and 6). He then indicates that the child is seated in the right position (line 7, tway-ss-e ‘good’). When the child has placed his arm in the monitor (line 8), the uncle responds with okay (line 9), displaying that the embodied execution of the action was as requested. In so doing he displays the potential closure of the sequence of getting the child to have the right position. He then proceeds within the same prosodic unit, and again slightly later, to specify the importance of sustaining the position (lines 9 and 11). A similar usage of OKAY to receive an embodied response occurs in the next excerpt from a driving lesson in Italian. In line 11, the instructor uses occhei to display that the student’s embodied action to his instruction was appropriate. The car is in a queue behind several other cars, and the traffic proceeds very slowly, so that the student has to stop and restart several times (lines 1–9 until the end of °porca miseria°, commenting on the queue). Excerpt 10. frizione ‘clutch’ (15sg2BM1, 52:45–52:57), 2010, Italian, face-to-face, driving lesson 01
STU:
02 03
INS:
04
vanno in vacanza i signori? are they going on holiday these people? (0.4) ((click)) (.) °que°sti qua? °the°se here? (0.2)
102 Lorenza Mondada and Marja-Leena Sorjonen
05
INS:
vanno tutti a san provino. they’re all going to San Provino. 06 (0.2) 07 STU: hh 08 (1.5) 09 INS: °porca +miseria°+ vieni su colla# *frizione eh?*# °holy crap° come up with the clutch huh? ins +looks down+ stu *left leg up-* fig #fig.1 #fig.2 10 (0.4) 11 INS:=> +occhei.+ Ins +looks down+ 12 (0.8) 13 INS: lascia rotolare. sempre col motore inseri°to° eh? let (it) roll. always with the motor install°ed° huh?
1
2
Chapter 4. OKAY in closings and transitions 103
When making the soft-spoken comment °porca miseria° on the queue (line 9), the instructor can be seen to look downwards during miseria, likely towards the set of pedals he has on his side of the car. He then moves to give the student an instruction concerning the handling of the car, instructing her to release the clutch pedal. The instruction is produced with a louder voice contributing to the production of transition to another activity. The student releases the clutch pedal towards the end of the instruction (compare Figure 1 and Figure 2, note the positioning of the student’s left leg), complying with the teacher’s instruction. Slightly later, the instructor produces occhei and at the same time looks downwards again. That allows him to see that the student accomplished the action, and with occhei he both displays verbally that the action was executed correctly and closes down the sequence. In this case, the closing implicative use of OKAY also serves to pave a way to the next instructional action (line 13). In the examples so far, OKAY was used as a third position response in relatively short sequences, many of them directive sequences. In the last excerpt of this section, OKAY (line 15) occurs as a third position turn in a lengthy question-answer sequence in dietician’s consultation in Danish. The dietician is filling in a form about the patient’s situation in computer. The dietician’s question at line 1 is a follow-up question after the patient answered to the prior question that he does not walk even little as he is not able to. Excerpt 11. vælter jeg bare rundt ’I’m just tumbling around’ (HB_011015_ny_BF_OQ), 2015, Danish, face-to-face, dietician-client interaction 01
DIE:
02
03
PAT:
04
DIE:
05
pat
PAT:
die die
hv%a er det for nog&en begrænsninger what be.PRS it.N for some limitation-PL what kinds of limitations is it: %looks at PAT---> &gestures---> du sån&’n €ø:h_ you.SG like uh that you uh: -->&rests hands---> €leans forward--->
jamen altså +det [mindste jeg r]ører- altså (nem-),+ PRT PRT DEF.ART least I move PRT well you know the tiniest I move/touch you know (eas-), [°du har° ] you.SG have.PRS °you have° +................hand in front of face----+
+(det %mindste) jeg an&strenger+ mig (DEF.ART.N least) I strain.PRS me if I strain myself +rubs eyes---------------------------+ --->%looks at computer screen---> --->&puts hands on keyboard--->>
104 Lorenza Mondada and Marja-Leena Sorjonen
06
+en lille smule_+ a-C little bit(C) just a little bit_ +hand on forehead, bends head forward+
07
08 09
10
11
12
+så har jeg ( ) bøvl me’ så_ then have.PRS I trouble with then then I get ( ) trouble with then_ +rubs forehead--->
Pat die
PAT:
pat pat die
PAT:
%(0.7)+% -->+ %......%looks at PAT--->>
+så (#bu os#) PRT ( ) then ( ) +leans head on hand, shakes head---> (0.4)+∆(0.6)$+(0.2)$ --->+.......+waves hand---> ∆looks up at DIE--->> $nods--$
så er +jeg bare væk, PRT be.PRS I just gone then I’m just gone ,,,,,+rests chin on hand--->> (0.2)
13
DIE:
14
PAT:
15
DIE:=> $okay.$ $nods-$
16 17
DIE:
°·pt°=ja= PRT °·pt° yes, =(vælter) jeg bare rundt, (tumble.PRS) I just around I’m just (tumbling) around,
(0.7)
så helst ikk op å: å PRT preferably not up and and so rather not up and: and
18
få pulsen op_ get.INF pulse-DEF.C up getting your pulse up_
19
[(eller) s]ån nogen ting_ (or ) such some thing or things like that_ [nej_ ] PRT no_
20
PAT:
21
DIE:
nej, PRT no_
Chapter 4. OKAY in closings and transitions 105
22 23
DIE:
24
(2.2)
hva me’ sån:=ø:h gåtur for eksempel what with like uh walking-trip for example what about like=u:h the walk for example me’ den hastighed du havde med her, with DEF.ART-C pace(C) you.SG have-PST with here with the pace you had here
25
PAT:
’a:,= yes yeah:,
26
DIE:
=°ned° ad gangen, down.DYN by corridor-DEF.C down the corridor,
27
ka (.) ku du ka du gøre det_ can.PRS can.PST you.SG can.PRS you.SG do.INF it.N can (.) could you can you do that_
The dietician produces his okay response (line 15) at a point when the patient has proceeded in his description to a completion-implicative description of a complete loss of strength after having strained himself (lines 11 and 14). Earlier in the course of the patient’s description of his difficulties, the dietician has implied a move to type in the patient’s answer (line 5), and slightly later responded to the patient with nodding (10). While the first description by the patient about loosing the strength gets a ja response (line 13), the second, stronger description is received with okay (line 15). Okay carries a stress on its first syllable and a fall to low on the second syllable, accompanied with a nod. With this response, the dietician displays his treatment of the answer as sufficient and completion-implicative. A silence follows, during which the participants orient visually to each other. The silence is broken by the dietician who presents a formulation of the upshot of the patient’s answer, working supposedly to formulate information to be recorded. After a confirmation by the patient, the next question on the agenda follows (lines 23–24, 26–27). We have shown in this section cases where OKAY is used to receive a response by the recipient to the OKAY speaker’s first-position action, that is, where OKAY implements a third position action. In our data from several languages, OKAY functions as a third-position response both to verbal and to embodied responses. Several of the examples that we have shown come from directive sequences, where OKAY displays that the prior response was appropriate, providing sufficient information for accomplishing the action at issue and in need of no elaboration. OKAY displayed its speaker’s stance to the sequence as closing relevant. We also showed that OKAY could be followed by a specification of the action that the instruction sequence had focused on, thereby starting a new short sequence (Excerpt 7).
106 Lorenza Mondada and Marja-Leena Sorjonen
3. Closing larger sequences In the previous section, we have seen how OKAY might be used to close simple sequences in second or in third position. However, OKAY is also quite often used to close larger sequences. Its scope backwards might vary. The scope can also be quite ambiguous in some cases: the exact sequential span that is closed by OKAY is not always clearly bounded within the previous stretch of interaction. Moreover, it is not always clear what OKAY per se does, as it is combined with other resources, preceding or following it. This section discusses these issues by examining some larger sequences that are brought to close by the use of OKAY as an element (among others) in manifesting the completion of a previous extended sequence. The segment closed by OKAY might be an extended sequence, or a series of sequences, that is, a sequential development expanding the sequences examined in the previous section. The following case, in Swedish, is an instance of these extended sequences: okej closes a sequence of instructions, which was extended by the instruction giver who corrects the responsive action to her instruction. The fragment was video recorded in a knitting club (Johansson 2016). Prior to the excerpt, Felicia has asked Cecilia how to do a purled stitch and has received some instructions in response, which she has then followed, by engaging in knitting. We join the action as Cecilia gives a further instruction to Felicia, orienting to her knitting as wrong (line 1). Excerpt 12. purled stitches (Swedish 8; purled stitches 1; JohanssonSyUU: Aviga maskor), 2015, Swedish, face-to-face, handicraft circle 01
CEC:
02
03 04
fel
FEL: CEC:
05
06 07
fel
FEL: CEC: fel
felicia .hh du ska ha den runt ditt=ä:h ((name)).hh you will have it round your uh Felicia you should have it around your
pekfinger (kan) *man säja* indexfinger can one say indexfinger so to speak *puts thread around indexfinger* okej, men [så?, but so but like this then [a fast på andra sidan. yeah except on the other side *ä:h f:ramför. s::å. uh in front so uh in front like that *changes position of thread-> ja[ja yeah yeah oh yeah [yez:.* yes ->*
Chapter 4. OKAY in closings and transitions 107
08 09
FEL: CEC:
okej. Mm,
å sen så tar du (.) liksom (.) här, and then so take you like here and then you take like here
Seeing what Felicia is doing, Cecilia engages in a corrective instruction (lines 1–2). She uses Felicia’s name in turn-initial position as an attention getting device as well as a way to possibly project a criticism or a reproach (see Günthner 2016). The instruction is complied with by the corresponding embodied action (line 2), as well as okej in second position (line 3, cf. supra, Section 2). But Felicia herself proceeds from okej to a new TCU in which she requests a confirmation of what she is doing. This occasions in response a correction by Cecilia (line 4), which is responded to by Felicia with a modification of the position of her fingers, in overlap with the continuation of a new instruction (line 5). Felicia also responds verbally, with a duplicated jaja after the completion of the corrective multi-unit turn (line 6), and Cecilia provides a confirmation with the particle yez: (line 7). At this point, Felicia has been able to perform the adequate movement for achieving the purled stitch. The extended sequence of instruction is closed by her with okej. Mm, (line 8). Next, Cecilia moves to the instruction of the next move to do (line 9). This example shows how sequences of paired actions, such as instructions, might be expanded (e.g., with corrections) before the action targeted by the instruction is actually achieved in an adequate and accepted manner. While other particles might be used to close single sequences (like jaja line 6; see also Koivisto and Sorjonen, this volume), OKAY is particularly used to close a series of sequences, which present the characteristics of featuring manual actions in the first place. In this case, the particle can be produced with a falling intonation, like in the previous Swedish example, or with a lengthening, like in the next German example, adjusting the delivery of the particle to the actual closing of the previous package. In the following excerpts, extended sequences like a story or an explanation are closed with OKAY by one of the participants (more precisely, in both of cases, by the recipient). In the next excerpt, from a job application training in Germany, the trainee (DOS) is telling to the trainer (TNE) a success story about his former employee who had a reputation to be a difficult person. The trainer’s OKAY (line 9) displays the closing-relevance of the story. Excerpt 13. mitarbeiter ‚employee‘ (FOLK_00174_SE_01_T_02_DF_01_cf 1215–1265), German, face-to-face, institutional interaction 01
DOS:
so hat sich das verhältnis gedreht,=weil er (0.6) this is how the relationship changed because he (0.6)
02
äh (.) von anfang an gemerkt hat dass ich ihn uh (.) he realized from the beginning that I
03
(0.3) fair behandle, (.) un nich (0.8) ähm (0.3) (0.3) treat him fairly and don’t (0.8) uhm (0.3)
108 Lorenza Mondada and Marja-Leena Sorjonen
04
er hat auch durchaus den ruf nich unbedingt he also has the reputation to not
05
ganz einfach zu sei(h)n;=hm mh ((laughs)) be the easiest ((colleague))
06
TNE:
08
DOS:
07
mhm? (.) mhm, (0.4)
09
aber (1.1) wunderbar °geklapp[t.° ] but (1.1) ((it)) worked beautifully TNE:=> [ o]↓khee;
11
TNE:
10
fig
(0.2)
#.h* in welchen regio#nen ham sie gearbeitet. .h which regions have you worked in *turns to his notes, manipulating them->> #fig.3 #fig.4
3
4
Chapter 4. OKAY in closings and transitions 109
The trainee DOS tells about a difficult person he had the opportunity to employ and work with (lines 1–5). The trainer TNE responds minimally (line 6), and invites DOS to continue, which leads to a proper happy end of the story (line 8). On the possible completion of the trainee’s turn (in overlap with its last sound, line 8), the trainer produces a closing o↓khee; (line 9), produced with a low onset and low volume, a slightly aspirated /k/ and an intonation falling to mid. After okhee; and a short gap (line 10) TNE produces an inbreath (line 11). He looks down to his writing pad and changes his body posture, displaying disengagement. Manipulating the writing pad and documents lying on the table before him, TNE prepares the next action and initiates a new topic with a high onset (line 11; Couper-Kuhlen 2001). Thus, a change of posture accompanies this closing (Figures 3 and 4). While in the previous excerpt a story was told, in the next excerpt an extended explanation is provided. In a genetic counseling consultation recorded in Brazil, a pregnant woman (PAT), whose fetus has been diagnosed with gastroschisis (a congenital malformation involving the intestines and the abdominal wall), is accompanied by her mother (COM), so that the geneticist can explain to both of them, what the malformation consists of and what are the available treatments. Excerpt 14. melhor forma possível ‘best possible way’ (HMF_ACONGEN_tarsila_ JEFERSON_04_02_14 14:22), 2014, Brazilian Portuguese, face-to-face, doctor-patient interaction in genetic counseling 01
GEN:
°°então°° .h ia ser importante colocá essas so .h it would be important to discuss these
02
coisas ↑pra vê↑ >que de repente issues to check because perhaps
03
vai ter que< fazê um exame de uri::na; it will be needed to make a urine examination
04
.hh [ou a ] lguma outra coisa [pra entendê]= .hh or something else in order to understand [°mhm°]
05
COM:
07
GEN:
06
08
PAT:
09
GEN:
10
COM:
11
GEN:
12
[ãrrã. uhuhm
]=
=um pouco melhor que aí tá aconte↑↓cendo °né°. a little better what is happening there you see (1.1)
tá::_= PRT
=°então tá°? then PRT
.h °tá° >mas< a ideia >como eu falei< .h PRT but the idea like I said
a gente vai tentá planejᣠtu:↑do pra que >tudo< we will try to plan everything so that everything
110 Lorenza Mondada and Marja-Leena Sorjonen
13
trans↓cor↑ra .h da melhor fo↑rma pos↑sível tá:; goes in the best possible way PRT
14
COM:=> °o↑↓que:i. [>tá bom< então.°] be.3SG good PRT all right then
15
GEN:
16
COM:
[essa é: that is °tá ↑↓bom°. PRT all right
] £essa é a ideia tá:_£ that is the idea PRT
The geneticist engages in a detailed explanation (lines 1–4), which is further elaborated on (line 7), after both recipients produced continuers (lines 5–6). In absence of a response at a possible completion of the explanation (line 8), the geneticist pursues a response with the particle tá2 (line 9). The mother responds by displaying a need for further information (line 10), which generates further explanations by the geneticist (lines 11–13), again brought to completion with the same particle tà. At this point, the mother receives this expanded explanation with an oque:i. (line 14), followed by another token (tá bom então) that displays that she has now reached sufficient understanding and is agreeing about the upcoming procedures, and therefore the possible closure of the larger sequence. Subsequently, both participants ratify this closing (lines 15–16), displaying their alignment in treating the explanation as clear and elaborated enough. In sum, in this section we have focused on expanded sequences that constitute complex actions, such as detailed instructions, stories, and explanations. OKAY is used at a possible completion of the activity, at a moment treated as closing relevant by the OKAY speaker. It should be noticed that in the cases of closing an extended sequence examined here, the producer of OKAY is the recipient, who in institutional situations is either the professional (Excerpt 11) or the client (Excerpt 12): the completeness of the ongoing action is treated as relevantly defined by the recipient of the action (the person who receives an instruction, story, or explanation), for all practical purposes. These cases can be seen as parallel to the ones in Section 2.2 in which the professional in an institutional encounter used OKAY in third position to close the sequence (e.g., a directive or question-answer sequence). Also in these cases, the OKAY speaker was the relevant recipient (e.g., the teacher evaluating the sufficiency and hence the closing-relevance of the sequence). 2. The discourse particle tá has been left untranslated, as it is almost impossible to find corresponding expressions in English, given the variety of actions it might perform. Tá is the third singular form of the verb estar (‘to be’), but in addition to the copula verb usage it is also used as a stand-alone token to display agreement or acceptance, and in turn-final position as a taglike element. c4-fn2
Chapter 4. OKAY in closings and transitions 111
4. Transition between activities OKAY occurs not only in the closing of the previous sequence but, like some other lexical elements, also in initiating a move to a next or to the main activity (see Beach 1993 for an early study). In prior research, diverse types of connectives, termed as “transition markers” (Keevallik 2010a, 2010b), or “action markers” (De Stefani 2013), have been explored as lexical resources in transitions from one activity to another. The role of OKAY in transitions has been observed early on by several authors, starting from, for example, Goffman ([1976] 1981, 49]) who considers expressions like okay, well, now and then as “bracketing markes” (Goffman 1974, 251–269) that delimit the boudaries of a frame, for example in “the management of some phase transition of the business at hand”. Beach (1995a, 142–146) characterizes okay as both prospectively and retrospectively oriented simultaneously, used as a pivotal element. The “dual-character use” of OKAY (Beach 1995a, 143) has been explored subsequently, for example, in interaction in different kinds of meetings, regarding for example topical transitions, transitions within the agenda, and with respect to the physical material of presentations (e.g., Birkner 2001; Meier 2002; Ford 2008, 64; Deppermann, Schmitt, and Mondada 2010). Another central field of research focuses on the use of OKAY to achieve transitions in instructional encounters, for example, in transitions from discussion sections to practicing object of learning (e.g., some aspect of foreign language) and moving to the next phase in more monological teaching (e.g., Rendle-Short 2000; Schleef 2008; Tyagunova and Greiffenhagen 2017; Reichert and Liebscher 2018). We will show that in some cases, the need to move to another activity is explicated with a meta-communicative turn. In most cases, however, the transition is achieved in a more subtle way, by using OKAY while transiting to the new activity, together with other resources, which are mainly multimodally organized in face-toface interactions. The kinds of transitions we will be showing occur in contexts in which several changes are observable at different levels: (i) they have both a retrospective and prospective character; (ii) there is a tight association of closing the previous actions or activities, and moving to a different type of action or coming back to a main action that was suspended (e.g., after a joke); (iii) they contain a topic closure, a shift in the topic or a change of topic (within the same activity, within a series of similar actions, or while initiating a new line of action); and (iv) there is a change in the participation framework. We begin the analyses of this section with two instances of transition in face-toface shop encounters. In these cases, the transition is achieved in a multimodal way,
112 Lorenza Mondada and Marja-Leena Sorjonen
deploying not only OKAY but also a range of embodied resources making the shift of activities visible. We join the first instance in a cheese shop in France. The excerpt shows how the participants reorient their bodies while moving from the phase in which a series of products are requested (not shown, but hinted at by the question, line 2) to the moment of paying. Excerpt 15. autre chose ‘something else’ (FRO_paris1007_ CLI3 33–47), 2016, French, face-to-face, shop encounter
01 02
VEN: fig
(6.1) vous avez envie d’autre chose?# do you want something else? >>looking at the customer-> #fig.5a+5b
5a=5b 03
CUS:
04
ven VEN: fig
c’est tout +bon I’m good +turns to till, away from CUS-> o:ké.# #fig.6a+6b
Chapter 4. OKAY in closings and transitions 113
6a=6b 05 06 07
ven cus VEN:
(1.0) +* (2.0) * (5.8) ->+weights the cheeses and makes the final bill->> *takes wallet out of the pocket* donc ça va vous faire un total de+ treize euros, so it’s a total of thirteen euros -->+ trente-neuf s’il vous plait thirty-nine please
By asking whether the customer wants something more (line 2), the seller orients to a possible closing of a series of purchases. She maintains her posture, facing the customer, looking at him (Figures 5a and 5b). The response (line 3) by the customer is formatted as a positive form of contentment. Even before this response, the seller orients to it in an embodied way, by shifting her posture, as a response that nothing more is needed: She turns towards the till (Figures 6a and 6b) while uttering o:ké (4). With the particle, she both orients retrospectively to the completion of a series of purchases and prospectively to the next step in the encounter, the paying. Turned to the till and after typing some prices there, she reads aloud the total price for the customer (lines 6–7). In this interaction OKAY thus does not just complete the previous series of request sequences. It is uttered as the seller has already turned towards the next activity, and by having done that she responds to what she has inferred from the emerging response. The transition is achieved with a set of multimodal resources and practices articulated together. The next fragment offers another instance of an embodied shift from one shopping activity to another, this time in a supermarket in Italian-speaking Switzerland. The couple PIE (Piero) and CAR (Carmine) is shopping for the weekend. In the excerpt, they negotiate moving from one type of product (bread) to another (cold cuts) (see De Stefani 2013 for other cases). We join the action as they are still focused on the bread:
114 Lorenza Mondada and Marja-Leena Sorjonen
Excerpt 16. pane ‘bread’ (cons 42271V, 16:20–16:44), 2004, Italian, face-to-face, grocery shopping 01 02 03 04 05
CAR:
PIE:
CAR:
06 07
10 11 12
13 14 15
(1.1)
>cioè< proviamo a prendere questo io non lo conosco. >that is< let’s try to get this one I don’t know it.
(0.2)
neanch’i:o. no va beh: andiamo sul classico me neither. no alright: let’s go for classic [prendiamo quello: let’s get the one:
PIE:
08 09
va beh. alright.
[tanto (.) alla fine non ce lo mangiamo mai. anyway (.) in the end we never eat it. perché mangiamo altre cose. because we eat other things.
car
(1.8)+(0.1)+ +withdraws gaze from bread+
car
%(2.7) %turns around 180°-->
CAR: fig
CAR: car fig
PIE:
occh#e:i. #fig.7
e qua# %è tutto. (.) spostiamoci, (1.5) verso, (1.2) and here that’s it, (.) let’s move, (1.5) towards, (1.2) -->%walks ahead-->> #fig.8
questa parte. this area. (1.9)
tu vuoi dei salumi? do you want cold cuts? CARmine
7
PIEro
8
Chapter 4. OKAY in closings and transitions 115
The process of deciding which bread to buy is brought to an end with PIE’s final comment (lines 7–8). Once the decision has been made, they can move to the next type of product and decision-making. That is what happens, with ocche:i uttered by CAR as the first indicator of a transition (line 10), followed by a change in his body orientation and gaze. With these resources he orients to another area of the supermarket (Figures 7 and 8). He also formulates explicitly the activity he is carrying out (spostiamoci verso questa parte ‘let’s move towards this area’, lines 12–13). Here ocche:i is thus used not only as a transition marker between two activities, but as a resource that allows its speaker to organize the overall shopping activity. The position of OKAY is here before its speaker (as customer) turns to the next product, area in the shop and activity. In the previous excerpt, OKAY was positioned after the speaker (as seller) turned towards the till, that is, to the place where the performing of the next actions needed to be made. This shows that the association of different types of interactional resources in the transition can vary in their temporal distribution. OKAY is also found at the juncture between qualitatively different frames of interaction, for example, between a moment of joking and serious activities. This is the case in the next fragment, from an everyday leisure activity among four Mandarin speaking friends (Lim, Lien, Ruiling, and Peng) playing cards. Lien and Peng form one team playing against the team of Lim and Ruiling. Ruiling is new to this card game, whereas the other three participants are more experienced players. We join the action as Ruiling is about to play and seeks confirmation from her partner in the game, Lim, of a rule (line 1). Excerpt 17. daydreaming (Mandarin 8; V5 A6, audio 20:57, Video 02_06:45), Chinese Mandarin
01
RUI:
shi tamen na fen shiba. be they take point right it’s them who get points, right?
02
LIM:
[women shi zhu. [womenshi zhuangjia we be master we be maker we are the masters. we are the game makers,
03
PEN:
[ni hai ni hai [zai mengzhong; you still you still at dream you’re still you’re still dreaming.
04
LIM:
womenshi womenshi[taofen we be we be escape point (so) we prevent them from getting points.
05
PEN:
06
RUI:
[ni hai zaimengzhongne. you’re still dreaming. e::he he [he he
116 Lorenza Mondada and Marja-Leena Sorjonen
07 08
LIE: RUI: fig
[((breathy laughter))
oK*A:#Y?* *puts one card on the table* #fig.9
9
10 09
(0.4)
Lim
Lien
Ruiling
10
RUI:
>oKAkuai kuai kuai.< quick quick quick.
12
LIE:
=ni meiyou fangkuair le.= you NEG have diamond PRT you don’t have diamond cards anymore?
13
RUI:
=>meiyou le.< NEG have PRT no.
Peng
Chapter 4. OKAY in closings and transitions 117
In response to Ruiling’s query, Lim provides a confirmation and explains their role in lines 2 and 4. In overlap, Peng produces a joking turn about Ruiling’s lack of understanding of the basic rules of the game (lines 3 and 5). This occasions some laughing together by Ruiling and Lien (lines 6–7), both displaying their orientation to Peng’s prior negative comment as joking and teasing. At the end of the laughter, the joke comes to a possible closure. At this juncture, Ruiling produces two OKAYs (lines 8 and 10). Both have a low pitch on the initial vowel o, moving from low to a markedly high rise on KA:Y. KA:Y is also significantly longer than o.3 While producing the first oKA:Y (line 8), Ruiling puts a card on the table (Figure 9), and the lengthening of the particle seems to be adjusted to her embodied action. She further rearranges the cards in her hand while adding a new TCU after the second o>KA:ja echt
21
cor AND:
22
JES:
23 24
*ja< yeah *gazes away, out the window--> (0.2)
okee;=ja=*°thema[wechs]el.° PRT topic.change yeah chance of topic -->*gaze forward towards JES-> [°(ja)°] (yeah)
JA und der +pfarrer+ der war=HH (0.8) yeah and the pastor he was +gazes to left to AND or into rearview mirror+
der war witzich? oder_ hähähähä ((laughs)) he was funny or COR:
@ja(h) [d(h)er war witzich.@ yeah he was funny
When Corinna launches the story about the pastor, she treats the funeral in question as either known or unproblematic (see the definite article in auf der beerdigung ‘at the funeral’, line 1). But the late repair initiator by Jessica (line 11) shows that it is not straightforward, since it occasions the announcement of the death of Corinna’s close relative (lines 14–15) and subsequently an affective response and commiserations by Jessica (lines 13 and 16). Corinna’s minimal response to the commiserations (ja< ‘yes’, line 18) displays a stance to curtailing a further expansion of the topic. She also withdraws from the interactional space by looking away, through the window (and this contrasts with her having been previously engaged in the conversation by looking forward, oriented towards Jessica). After a pause, Jessica’s okee-turn displays a closure of the prior sequence and achieves a transition away from a delicate issue. Okee is followed by the particle ja ‘yes’ and an explicit proposal to change the topic in the same turn, with the metapragmatic formulation themawechsel ‘change of topic’. The particle okee carries a stress on its second syllable. At the end of the syllable the pitch falls to mid, and the following ja is latched onto okee. The formulation themawechsel is produced with a lower voice and at a faster pace. During this formulation, Corinna reorients again towards the other participants, looking forward. Jessica then invites a resumption of the prior activity, Corinna’s story, which she puts on hold in line 10 (lines 22–23). The turn-initial ja ‘yes’ and und ‘and’ mark the turn as skip-connecting back to, and resuming the previous story. Corinna aligns with this invitation and continues her story. In this last case, OKAY works to manage the transition to and a resumption of an activity that was previously initiated and suspended after a side sequence.
Chapter 4. OKAY in closings and transitions 121
In the other cases examined in this section, OKAY operated a transition to a new activity. As the extracts from different languages clearly show, OKAY managed centrally this shift, although in slightly different positions relative to (a) other verbal elements contributing to the transition (such as particles, agreement tokens and other expressions), and (b) practices making the transition visible and implementing it in an embodied way in face-to-face interactions. Its variability shows that in these moments of transition, participants orient both retrospectively to the previous activity about to be completed and prospectively to the one about to start. These orientations might co-exist and combine in different ways, depending on the local constraints of the participation framework, and in face-to-face interactions, the embodied organization of the ongoing actions and the spatial and material context. 5. Moving into closing the interaction OKAY is used not only for moving from one activity on the next, but also to initiate the closure of the entire interaction. It is thus used for shaping the overall organization of the interaction. Such OKAYs can be used in a variety of positions that are all sensitive to the projected imminent closing, even before the proper pre-closing and the actual closing of the exchange (Schegloff and Sacks 1973; see also Button 1987, 1990). In this section, we will first briefly explore early indications of an orientation towards closing, and then move on to discuss the use of OKAY in a pre-closing and closing context. In the temporal sequential development of interaction, the participants move towards the closing-relevant actions step-by-step. We find OKAY as a display of closing relevance for example at places where the participants have produced similar types turns that imply closing relevance of the sequence and topic, and possibly of the interaction (e.g., yeah in English, see Jefferson 1981b, also on closing relevance of assessments). In these cases, OKAY implements a strong display of closing relevance, stronger for example than an acknowledgement token (such as yeah), and it can be followed by its speaker’s moving to closing the interaction. An opening to a systematic research on the contribution of the phonetic-prosodic resources in transition to closing the interaction is Wright (2011) who discusses the relevance of phonetic parameters, such as, pitch, loudness, voice quality and ‘articulatory segmental’ features in segments of talk that include OKAY initiating closings in everyday English telephone calls. In the following segment from a phone call between the former Brazilian President Lula and his lawyer Roberto Teixeira, the speaker moves from OKAY in his subsequent talk to a next phase of the closings (line 4).
122 Lorenza Mondada and Marja-Leena Sorjonen
Excerpt 20. tá bom querido ‘all right dear’ (Lava Jato LIS X ROBERTO TEIXEIRA March 16 2016 02.30), 2016, Brazilian Portuguese, telephone, tapped calls among policians 01 02 03 04 05 06
LUL:
tá ↑bom querido. PRT dear ROB: t(h)á bom amigo. hheh .hh= PRT friend hheh LUL: =tá bom. PRT ROB:=> #e# o:quei então. e am[an↑hã então↑ vou vendo; #e# then and tomorrow then I’ll see LUL: [o: o: o:: (.) othe the the the-5 já vol↑tô o fabi↑↓ano. has (the) Fabiano already returned
After a series of exchanges of tá bom ‘good; fine’ (lines 1–3), Teixeira moves to display closing relevance of the activity so far with a turn-initial expression o:quei então ‘OKAY then’, which forms a prosodic unit of its own with a falling final contour. He then moves to evoke earlier made arrangements (‘and tomorrow I’ll see’, line 4), which orients to moving to closing down that sequence. However, instead of aligning with the closing implication by Teixeira, Lula initiates a move out of the closings, by starting a new topic and activity: he asks a question soon after Texeira has proceeded to the arrangements (lines 5–6). Especially in institutional phone calls, another place for OKAY to occur is before a participant thanks the other. The following segment comes from a phone call in Polish to a business. Prior to the fragment, the customer has inquired about the practicalities concerning an order collection, such as the address of the business, its opening hours and the method of payment. It turns out that the company is closed on the weekends and open only till 4 pm on weekdays, which turns out to be problematic for the customer. At the beginning of the segment, the customer ensures that his colleague will come pick the order up before 4 p.m. (line 1). Excerpt 21. okej to dziękuję ‘okej thank you’ (GĄDKI_VN680287), 2009, Polish, telephone, service encounter 01
CUS:
do szesna:stej je odbierze. to sixteen them collect.3SG (he’ll) collect them till four
02
SAL:
dobrze=dobrze:, alright alright
03 04
(0.2)
CUS:=> ↑oke:j.=to dzięku:ję,=do u[słysze:nia, it thank.1SG to hear.N thank (you),=till later,
5. In Brazilian Portuguese, as in many other languages, the definite article (o [masc], a [fem]) is used with proper name references, such as in o Fabiano ‘the Fabiano’.
Chapter 4. OKAY in closings and transitions 123
05
SAL:
06
[dziękuję ró:wnież,= thank.1SG also also thank (you), =do usłysze:nia, to hear.N till later,
The salesperson receives the customer’s turn with a repeated dobrze dobrze (line 2) that treats the customer’s arrangement as acceptable and sufficient. Following a brief pause, the customer’s next turn moves into closing the interaction. The turn (line 4) is composed of three turn constructional units, each accomplishing a different task. The turn starts with ↑oke:j, with a relatively high onset on the first syllable and a slight lengthening on the second one. The TCU is brought into a possible completion with a falling final contour, being part of the orientation to the larger activity of the order collection arrangement-making as now complete. The other two TCUs, which contain units used systematically in conversation closings (thanking and leave taking with an orientation to the next contact), provide further evidence that the interaction is now moving into closing. This understanding is reciprocated by the salesperson’s subsequent turn (lines 5 and 6). In the close preceding proximity of a closing relevant OKAY, we can find also other occurrences of OKAY used slightly differently. The segment below from the end of an Estonian phone call contains okei on lines 5 and 6, as well as on line 9. Pille and Ragne are talking on the phone about meeting the same night, which Pille who needs a listener has suggested. Ragne has just arrived home and has been slightly resistant to the idea of going out. Excerpt 22. elistame ‘let’s call’ (PILLE_P1A3), 1998/99, Estonian, telephone, informal conversation
01
PIL:
02
RAG:
noh; Elista-me ühesõ[naga]. PRT call-1PL in.one.word so we’ll call. [ma::] helista-n su-lle veel I call-1SG you-ALL again I’ll call you back
03
ku ma olen ära söö-nd; when I be-1SG PRT eat-PPT once I’ve had (something) to eat;
04 05
[ja siis ma üt]le-n mis on. and then I say-1SG what be.3SG and then I’ll say what (I) have. ] PIL:=> [ o: k e i:
07
RAG:
06
PIL:=> o:[kei,]
[eks,] PRT right,
124 Lorenza Mondada and Marja-Leena Sorjonen
08
PIL:
no:h_ PRT well/so
09 10
RAG:=> o:[kei. ] PIL: [tšau_] bye_
11
RAG:
tšau. bye.
The first two responses o:kei: and o:kei, (lines 5 and 6), produced with an accented and lengthened first syllable and completed with a slightly rising contour (line 6), respond to a promise by Ragne to call back soon. They simultaneously display acceptance of the promise and, considering Ragne’s situation (just back at home, needing to eat), they can also understood to work toward the closing of the call. The third o:kei. in line 9, now produced with a falling final contour, is already moving into a closing proper, and it is followed by the closing goodbyes. The last example from a Finnish phone call contains two sets of okei by both participants, the first one (lines 12 and 14) moving towards closing after consecutive joo turns displaying problems in turn-taking, and the second set (lines 22–23) following wishing each other happy new year. Sani has returned Vallu’s call and found that he wanted to ask for some eggs for a dish he was preparing. They report briefly each other’s Christmas activities and we join them as they move then to talk about seeing each other, from line 1 onwards. Excerpt 23. Tekstit ‘texts’ (KTA3 Sg401_lauri), 2007, Finnish, mobile phone call, informal conversation 01
VAL:
02 03
SAN:
04 05
VAL:
06
SAN:
07 08 09 10 11
SAN:
VAL:
voiaham me (0.2) >e ehkä kannattaa tekstitt< we could y’know (0.2) maybe it would be worth to text (0.7) .hh t[ teks]tit, (0.7) .hh texts, [joo, ] yeah (.) lähettää [(ku mä en send (‘cause I [lähetetää yes let’s
(0.6)
joo, yeah, (0.8)
jooh. yeah. (2.6)
osaa) can’t) vaa. do that.
Chapter 4. OKAY in closings and transitions 125
12
VAL:
14
SAN:
16
VAL:
17
SAN:
18
VAL:
13 15
19 20 21
VAL:
22 23
VAL: SAN:
25
SAN:
26
VAL:
24
27 28
VAL:
↑okei;
(0.3)
°okei:?°
(0.8)
↑joo? yeah, joo hauskaa [uutta vu]ott(h)a. yeah happy new year. [↑no mut,] well but (0.4)
↑joo? h samoin? yeah? you too? (0.8)
>ok[eiok[ei 06 07=> 08 09 10
HYL: NAN: NAN: HYL: NAN: HYL: NAN:
=.hh ↑MAYbe we can go out for a drink tonight. (.) ye::ah. that soun- yeah I owe you a dri:nk. (.) I wanna buy you a dri:n[k. [AOW. alri[:ght, [o↑KA:Y, so we will for sure;= =alri[ght.] [ af ]ter; (.) the pl[ay.]
Nancy’s proposal in line 5 makes acceptance or rejection from Hyla relevant next, but Hyla at first only registers the proposal with a variant of oh (Heritage 1984a). Nancy now follows up with OKAY (line 7), which ends up in partial overlap with Hyla’s subsequent acceptance token alright (line 6). After Nancy’s renewed commitment to buy Hyla a drink (line 8), Hyla reiterates her acceptance of the proposal in line 9. The prosodic design of the OKAY token in line 7 of Excerpt 1 is typical when the particle accompanies requests and proposals: the first syllable of OKAY is lower, shorter, and softer than the second. There is a step up of approximately three semi tones to the next syllable, which remains high, ending here with a very slight final rise. Figure 1 presents a pitch trace of OKAY in Excerpt 1 for illustrative purposes, with the caveat that other, often just as crucial prosodic and phonetic features (e.g., loudness, accentuation, timing, vowel quality) are not captured in this diagram.9 It is thus only an imperfect representation. When OKAY is produced with the prosodic design documented in Excerpt 1 as a follow-up to a request or a proposal, it solicits commitment to comply or acceptance. The interlocutor typically provides this in next turn, often with a responsive OKAY (see Excerpt 18 below) or alright, as in line 9 above. While this freestanding variant of OKAY is found in the newer collection as well, it also occurs there as a tag latched immediately on to the end of a TCU. 9. Because the division between the two syllables of OKAY is invariably marked through the voicelessness of /k/, phonemic segmentation has been omitted from the pitch trace figures.
136 Elizabeth Couper-Kuhlen
Pitch (Hz)
500 300 200 150 100 0
0.4397 Time (s)
Figure 1. Pitch trace of OKAY in line 7 in Excerpt 1 A drink
Moreover, it attaches to TCUs that implement not only directive, or deontic, actions (dealing with obligation: see Stevanovic and Peräkylä 2012) as in Excerpt 1, but also informing, or epistemic, ones (dealing with knowledge; see Heritage 2012a, 2012b). Here is a case in point: Excerpt 2. Tax and tipping (Farmhouse, 34.25), 1998, American English, face-to-face, informal interaction [Michelle, who works part-time as a waitress, is explaining to her friends Donna and Laura how to reckon a tip based on a bill that includes sales tax.] 01
DON:
it comes to twenty percent with tip and tax?
02
MIC:
[no no no
03
LAU:
[so you do twenty percent and] then
no
no
]
04
↑say your twenty percent was five bucks and your
05
tax was another dollar,=then it’d be six bucks?
06
(0.4)
07
LAU:
is that what you’re saying?
08
MIC:
*I’m saying okay let- let’s just say the bill is *gazing at Laura------------------------------->
09 10
twenty dollars total,* -------------------->* LAU:
11 12
(0.2) MIC:
13 14 15
and then ta:xes >would be like what< eighty ceno. how’s that work. ( ) a dollar sixty
LAU: MIC:
16 17
oka:y,
whatever. yeah [a dollar sixty ] [°I don’t know° ] (0.5)
LAU:
18-> MIC:
sounds good. *£well (percentages) (.) *gazing at Donna------->
Chapter 5. Prosody and phonetics of OKAY in American English 137
19=>
no(h)t my spe(h)cia(h)lty,= o↑KA:Y_£* ---------------------------------->*
20
DON:
mine either.
21
MIC:
↑so anyway let’s just say you have this amount
22 23
for your tax. LAU:
right.
When Donna proffers a candidate understanding of what Michelle is saying (line 1), Michelle rejects it (line 2), while Laura proposes another interpretation (lines 3–5) and asks for confirmation (line 7). But Michelle now advances her own case example, using the sum of twenty dollars as an imaginary bill for food and drink (lines 8–9). She then begins to reckon what the sales tax (8% at the time) would be on twenty dollars, trying eighty cents first (line 12) and then, after correcting herself, repairing to one dollar sixty (line 13). Laura indulgingly acknowledges this repair with whatever (line 14) and sounds good (line 17), whereupon Michelle – with an overlay of laughter particles – produces a self-deprecatory account for her uncertainty in arithmetic (lines 18–19), gazing now at Donna. To this explanation she appends an OKAY, and Donna responds with a congruent self-deprecation (line 20). Michelle now proceeds with her case example. The OKAY token in line 19 has roughly the same prosodic-phonetic design as that in Excerpt 1: the first syllable is shorter, lower, and softer than the second. The second syllable is produced with a pitch step up, here encompassing approximately seven semitones, and remains on a high pitch level until the end (the ‘flutter’ in the pitch trace of Figure 2 is due to the presence of ‘speech-laugh’ or ‘wobble’: see Ford and Fox 2010 for more on this phenomenon):
Pitch (Hz)
500 300 200 150 100
0
0.5158 Time (s)
Figure 2. Pitch trace of OKAY in line 19 of Excerpt 2 Tax and tipping
What is interesting about the OKAY tag in Excerpt 2 is that it attaches, not to a request or proposal for which it is soliciting compliance or acceptance, but to an account, in this case for a lack of skill in arithmetic: not my speciality (line 19). Note that in the next turn Donna responds, not by saying okay or alright, but
138 Elizabeth Couper-Kuhlen
by orienting to Michelle’s self-deprecation and making a similar self-deprecatory statement about herself (line 20).10 With this newer use of OKAY as a tag, the speaker does not solicit acceptance, nor even acknowledgment, of the action it accompanies. In contrast to freestanding OKAYs accompanying requests and proposals, there is no pursuit of such a response, should one not be forthcoming. That is, the OKAY tag has become part of the prior TCU and serves now as a turn-final particle, retaining its characteristic up-stepped pitch contour. To summarize, OKAY accompanies first-position actions in both the older and the newer collections, where it is produced with a lower, shorter, and softer first syllable and up-stepped or rising pitch on the second. While it attaches to directive actions such as requests and proposals in both data sets, soliciting commitment to comply or acceptance, in the newer data set it also occurs latched onto informing actions such as accounts, promises, and informings. In these cases, however, it lacks strong conditional relevance. 4.2
OKAY in the second position of a sequence
Let us turn now to the use of OKAY as a responsive particle. Here distinguishing different responsive actions based on the nature of the initiating action has proven to be useful (Sorjonen 2001; Thompson, Fox, and Couper-Kuhlen 2015). In the present case this means distinguishing OKAY as a response to a directive action, e.g., a request, proposal, offer, or invitation, from OKAY as a response to an informative action, e.g., an informing or a telling. Whereas the former type of sequence could be said to be deontically driven, the latter is epistemically driven (Drew 2012). In deontically driven sequences, an OKAY response indexes acceptance of and/or commitment to the obligation conveyed by the prior turn (Thompson et al. 2015). In the case of epistemically driven sequences, an OKAY response conveys acknowledgment or registering of the information just imparted (see also Beach 1993). In the following we consider these two types of OKAY responses separately.
Responsive OKAY in deontically driven sequences When the particle OKAY is used to respond to a directive action, it signals acceptance and willingness to comply. This is what we find happening in the following excerpt:
10. In another case in the newer data we find a similar OKAY tag being added on to a predictive-like promise: it’s just not gonna happen.=OKAY?.
Chapter 5. Prosody and phonetics of OKAY in American English 139
Excerpt 3. University address (CF 4984_17), 1996–1997, American English, telephone, informal interaction [Roberta has just announced that she will be sending Liz a CD by the folk singer Susan Werner.] 01
ROB:
will that fit in the mailbo:x?
02
LIZ:
hell if ↑I: kno:w,
03
.hhh [you wanna make su:re]
04
ROB:
05
LIZ:
06
[hhh hhh hhh hhh
]
if you wanna make su:re; the best thing to do is just send it to the
07
university;=otherwise I gotta go li:ke
08
hhh [all the way to
09
ROB:
10
[°okay (.) why don’t] I do tha:t° °just [say°
11
LIZ:
12
ROB:
]
]
[fourtee:nth] street to [pick] it u:p. [o:h ]
13
okay.
14
↑so
15
(0.4)
16->
[give ] me the address at the univers(ity).
17
LIZ:
18
[(
)]
(0.6)
19=> LIZ:
O `ka:y;
20
(.)
21
it’s um-
22
(0.5)
23
long island university?
24
(0.3)
25
ROB:
m hm:
26
LIZ:
o:n:e university pla:za:
In lines 1–13, Liz informs Roberta that the best way to send the tape is through her university, as she would have to go the post office to pick it up if it does not fit into her mailbox at home. Roberta agrees to do so (lines 9–10, 13). She now asks Liz for her university address (line 16), whereupon Liz responds first with the particle OKAY (line 19) and then proceeds to call out the address (lines 21, 23, 26). This OKAY is produced with stress on both syllables: The first is louder, the second begins only a bit higher and is lengthened with a falling glide to mid: See Figure 3. The same pattern can be heard on Roberta’s (overlapped) OKAY in line 9, with which she agrees to Liz’s suggestion that she (Roberta) should send the tape to her
140 Elizabeth Couper-Kuhlen
Pitch (Hz)
500 300 200 150 100
0
0.4407 Time (s)
Figure 3. Pitch trace for OKAY in line 19 of Excerpt 3 University address
university address. This prosodic-phonetic design for a second-position deontic OKAY is unmarked, by which we mean that the response is straightforward: There is no affective overlay on the agreement or commitment to comply. We return to prosodically-phonetically marked variants of responsive OKAY in Section 5 below. However, the data reveal a second recurrent pattern for unmarked responsive OKAY in deontic sequences. Here is a case in point: Excerpt 4. Some suggestions (Joyce & Stan_2), 1960s, American English, telephone, informal interaction 01
STA:
02
to as:k your uh:: advice on two little matters: uh.
03 04
.hhhh well the ↑main reason I called you up joyce was (0.4)
STA:
I ↑might be going shopping either tomorrow or saturday
05
and I’m- ↑what I’m looking for is a couple of things.=
06->
>I thought maybe you might have some< suggestions
07->
where I could find it.
08=> JOY:
O ↓´ka:y?
09
↑first of all: I’m looking for: a: pair a sa:ndles:?
STA:
10 11
(0.7) STA:
and a hat.
In lines 1–2 Stan announces the reason for his call to Joyce as wanting to ask […] advice on two little matters. After a brief pause he then elaborates that he will be going shopping the next day and is hoping for some suggestions from Joyce as to where to go (lines 4–7). Joyce signals her willingness with an OKAY token in next turn (line 8), whereupon Stan launches an enumeration of what exactly he is looking for (lines 9–11). Joyce’s OKAY in line 8 sounds quite different from that documented in Excerpt 3 above: although it also has stress on both syllables, the first syllable is lower, shorter,
Chapter 5. Prosody and phonetics of OKAY in American English 141
Pitch (Hz)
500 300 200 150 100
0
0.7379 Time (s)
Figure 4. Pitch trace for OKAY in line 8 of Excerpt 4 Some suggestions
and softer than the second. The second syllable has a step down of approximately four semitones with some lengthening and a rising glide to high: See Figure 4. With this prosodic-phonetic design on OKAY, Joyce gives Stan a go-ahead, signaling that she is willing to comply with his request and make some shopping suggestions. But at the same time, her go-ahead is somewhat cautious: It implies ‘yes but tell me more’. In fact, this is what transpires next: Stan now proceeds to detail at great length (not shown here) what exactly he is looking for. In sum, there are two recurrent patterns for unmarked OKAY responses to directive actions. In both cases, the OKAY signals willingness to comply with the directive. In one, the second syllable of OKAY is somewhat lengthened and has a falling glide (to mid); in the other, the second syllable has a rising glide. A final rising OKAY marks provisional acceptance of the directive, intimating that more details are needed to ‘seal the deal’.
Responsive OKAY in epistemically driven sequences As an example of OKAY in second position of an informative sequence, consider the continuation of Excerpt 1: Excerpt 5. A drink, cont’d (HGII 17), 1960s, American English, telephone, informal interaction [Hyla and her friend Nancy are planning to go to the theater together in the evening.] 01
HYL:
02 03
=.hh ↑MAYbe we can go out for a drink tonight. (.)
NAN:
04
ye::ah. that soun- yeah I owe you a dri:nk. (.)
05 06
NAN: HYL:
07
NAN:
I wanna buy you a dri:n[k. [AOW. alri[:ght, [o↑KA:Y,
142 Elizabeth Couper-Kuhlen
08
so we will for sure;=
09
HYL:
10
NAN:
11
HYL:
=alri[ght.] [ af ]ter; (.) the pl[ay.] [.hh]
12
(.)
13-> HYL:
I can’t drink too much cuz I’m dri-i-vhh[i(h)i(h)ng,]=
14=> NAN:
[↑O↓k a: y; ]=
15
NAN:
16
HYL:
17
NAN:
18
=(↑well) (0.4) I said one dri[nk. [hhheeh .heh .eh [.hh [you think I’m made of money or something-hhn=
19
HYL:
=.e.e=
20 21
NAN: HYL:
=.hhi::[::hh] [.t.k]°h-h°
Once the two friends have agreed that Nancy will buy Hyla a drink after the play (lines 5–11), Hyla now announces that she will not be able to drink much because she will be driving (line 13). Nancy acknowledges this information with OKAY (line 14) and adds that she only promised a single drink anyway (line 15). Nancy’s OKAY in line 14 is delivered in terminal overlap with stress on both syllables, the first, however, being higher and louder than the second. On the second syllable there is a pitch step down of five semitones, with some lengthening. The final rise at the end is barely audible.
Pitch (Hz)
500 300 200 150 100 0
0.4843 Time (s)
Figure 5. Pitch trace for OKAY in line 14 of Excerpt 5 A drink, cont’d
The prosodic-phonetic design of OKAY in line 14 of Excerpt 5 is characteristic of unmarked OKAY responses to informings: There is no affective overlay on the acknowledgment or registering of the information. We return to prosodically-phonetically marked variants of responsive OKAY in Section 5 below. There is, however, a second recurrent pattern of delivery for responsive OKAY in epistemically driven sequences. This is the one exemplified in the following excerpt:
Chapter 5. Prosody and phonetics of OKAY in American English 143
Excerpt 6. Native speaker (CF 4984_2), 1996–1997, American English, telephone, informal interaction [Liz is explaining the Call Friend set-up to her interlocutor Roberta.] 01
LIZ:
they said ↑you kno:w
02
(0.2)
03
if you wanna: make a free: thirty minute phone call;
04->
to anyone=it has to be a native speaker of Engli:sh,
05
(0.7)
06=> ROB:
°o↓ka:y,°
07
[well th ]at I am:,
08
LIZ:
[you_know]
09
LIZ:
.hhh so we’re being recorded.
10
↓so you know you can like- tell all your secre:ts↓
11 12
[n ROB:
hhh hhh hhh hhh]
[okay. hhh hhh °( )]
13
(just) don’t use any four letter wor:ds
14
[o:r ]
15
LIZ:
16 17
[.hhh] £↑A:H_I think you can say anything you want.£
ROB:
°°#okay good#°°
In line 4 Liz reports that the instructions for the Call Friend telephone call include the requirement that the person being called must be a native speaker of English. Roberta acknowledges this information in next turn with OKAY (line 6) and then applies it to herself in declaring that she satisfies the requirement (line 7). In lines 9–10 Liz goes on to detail the implications of her informing. Prosodically speaking, Roberta’s OKAY in line 6 does not set in immediately after Liz’s informing, but is instead delayed by 0.7 sec. Its volume is softer than that of prior talk. The pitch begins mid high and then drops by approximately four semitones to a lower level, from where it rises to mid high.
Pitch (Hz)
500 300 200 150 100 0
0.6994 Time (s)
Figure 6. Pitch trace for OKAY in line 6 of Excerpt 6 Native speaker
144 Elizabeth Couper-Kuhlen
This prosodic design treats the prior informing as preliminary, implying the expectation that more will be said (see also Heritage and Clayman 2010, 113).11 In Excerpt 4 Roberta may be anticipating that Liz will go on to explain why she is imparting this information at this point in time. In fact, Liz does go on to explain as of line 9. In epistemically driven sequences then, prosodically unmarked OKAY responses to informings have a first syllable that is somewhat higher and often louder than the second, with the second displaying a pitch step-down and some lengthening. These responses acknowledge or register the information just imparted without affective overlay. If the second syllable is produced with a rising pitch glide, there is an implication that more talk is expected from the other. 4.3
OKAY in the third position of a sequence
According to Schegloff (2007, 120), OKAY can be used in minimal sequence expansion as a ‘sequence-closing third’, in which case it “mark(s) or claim(s) acceptance of a second pair part and the stance which it has adopted”. Interestingly, Schegloff describes OKAY as the “appropriate” sequence-closing third in directive sequences such as those involving requests, offers, and invitations, while he treats the particle OH as appropriate for claiming receipt of information in informative sequences (ibid., 120). Yet our data collections show OKAY occurring in third position of both directive and informative sequences; its epistemic use has increased dramatically in frequency in the newer data (see Table 1). In deontically driven sequences, the OKAY speaker has issued a directive and the interlocutor has indicated either acceptance or rejection in next turn. An OKAY particle produced in third position either finalizes the deal in the case of acceptance, or finalizes the rejection (Davidson 1984). In epistemically driven sequences, on the other hand, the OKAY speaker has typically asked a question, to which the interlocutor provides an answer in next turn. A freestanding particle OKAY in third position accepts the information provided in prior turn and proposes to close the sequence. In the following we will examine these two types of third-position OKAY separately.
11. Heritage and Clayman find that in doctor-patient interaction, specifically during problem presentation, doctors employ OKAY with rising intonation to counteract the implication that the patient has reached a unit boundary in their talk: “ ‘Okay,’ (with comma intonation) permits doctors to hedge their bets as to whether a patient is complete” (2010, 113).
Chapter 5. Prosody and phonetics of OKAY in American English 145
Table 1. Frequencies for OKAY in different turn and sequence positions OKAY Position
OKAY (Older data)
OKAY (Newer data)
OH OKAY* (Newer data)
Total
2 26 5 8 5 – 27 4 19 96
9 22 14 2 34 4 15 2 – 102**
– – 7 – 16 – – – – 23
11 48 26 10 55 4 42 6 19 221
1-Follow-up or Tag 2-Deontic 2-Epistemic 3-Deontic 3-Epistemic Continuer Transitional_Preface Transitional_Close Conversational_Preclose Total
* Not dealt with here. ** Not included in table: OKAY in Reported speech (# = 11), Concessive OKAY (# = 4)
Third-position OKAY in deontically driven sequences In a deontic sequence, third-position OKAY marks acknowledgment of the interlocutor’s acceptance of the suggestion, invitation, proposal, etc., that the speaker has just advanced. Here is a case in point; this excerpt is from the same conversation as Excerpt 2: Excerpt 7. Water and coffee (Farmhouse_50, 31.53), 1998, American English, face-toface, informal interaction [Mom is offering her guest Donna something to drink.] 01-> MOM:
+would you like some m:ore water, or some hot tea, +standing, gazing at Donna------------------------>
02
[or coffee,
03
DON:
04-> MOM:
[hhhh
05-> DON:
o[r+ -->+ [↑I’ll have some water and coffee.=
06=> MOM: don
=%>↑O°kay;° SAR: 10 11
[what ↑are you doing. (0.9)
DEB:
!t (0.6) well I- I’m working in (0.8)
12
>↑I mean it has nothing to do with< fi:lm;
13
°#at a:ll#.°
14
[°you kno:w°]
15
SAR:
16
DEB:
[↑I know
] is it like secretarial stuff? or [what.] [yea:h]
17
it’s (0.3) it’s secretarial;=
18->
=↑we:ll_ .hhh I- ↑what I do ↓m:ost of the ti::me;
19
i:s u:m (0.5) I: (1.1)
((26 sec. of talk omitted in which Debby details her work)) 36->
=so ↑that’s what I do most of the time;=
37->
=is sche:duli:ng,
38=> SAR:
[↑O:↓kay;]
39
[.hh
40
DEB:
] but (0.3) but I also do::–
like whatever e:ls:e °#they need me to do:#°.
In lines 1–8 Debby is explaining to Sarah why she only works part-time: It would be too much for her to work full-time at something unrelated to her screenwriting. When Sarah now asks what are you doing (line 9), Debby launches an extended telling detailing her part-time work, introduced after some background talk with the preface what I do most of the time (line 18). When she now returns to this phrase
148 Elizabeth Couper-Kuhlen
so that’s what I do most of the time (line 36), she has hearably completed her telling, at least provisionally, and Sarah acknowledges the information imparted with a third-position OKAY (line 38). Sarah’s OKAY has a louder and somewhat lengthened first syllable (both syllables are stressed), which is higher in pitch than the second. The second syllable has a step down of approximately five semitones to a pitch which remains low until the end:
Pitch (Hz)
500 300 200 150 100 0
0.566 Time (s)
Figure 8. Pitch trace for OKAY in line 38 of Excerpt 8 Scheduling
Other third-position epistemic OKAYs sound roughly the same, with some variation encountered on the height of the first syllable and the extent of the step-down on the second syllable. Yet the new data contain several instances of an alternative prosodic design for third-position epistemic OKAY: This variant occurs exclusively in sequences where the prior informing runs counter to what the OKAY-speaker ostensibly thought was the case. Consider, for instance, the following excerpt: Excerpt 9. College tuition (Call Home En 4544_1246), 1994–1997, American English, telephone, informal interaction [Beth and Ann, two middle-aged friends, are talking about what tuition was like when they went to college.] 01
BET:
02 03
(0.5) ANN:
04 05
.hh I wanted to go to the university of rochester. at that time, and it [was ( )-
BET:
06-> ANN:
[and (
)
AND AT THAT TIME WHAT WAS IT.
07->
FIVE THOUSAND DOLLARS?
08-> BET:
no it was much more=
09->
=it was like (0.5) twelve thousand dollars.
Chapter 5. Prosody and phonetics of OKAY in American English 149
10=> ANN:
11
it was like unbelievable.
BET:
12
and I told my parents (.)
13
you know (.)
14
forget it (.)
15
I’m not going=
16
=I’m not gonna ask (.)
17
you know I’m >NOT gonna get in debt for the rest of
18
my lifeand I don’t want YOU to get in debt for the
19
re(h)st [of m(h)y li(h)fe.
SAL:
like I was waking up so early out in arizona;
20
(0.9)
21=> MOM:
°Oka:y,°
22
(0.6)
23
SAL:
24 25
(3.0) SAL:
26 27
becau::s:e_ ↑I don’t know↑= =cause I wanted to sit outsi::de_=↑you kno:w;
MOM:
m hm,
Sally confesses to Mom that she has not been able to get up in the morning at her new home in Atlanta, information that Mom accepts with stoicism (lines 1–9). Sally’s account for this is that she does not have any reason to do so (line 15). After a continuer by Mom (line 18), Sally now moves to contrast her present situation with that in Arizona, where she woke up early (line 19). This line, however, is both prosodically and pragmatically incomplete: It only sets the stage for the explanation that Sally has projected. But Sally proceeds to make a significant pause, whereupon Mom produces OKAY (line 21). Mom’s OKAY is not responding to the prior informing as sufficient and complete, but is instead signaling that Sally should go on. Sally now provides an account for why she woke up early in Arizona (lines 23, 25–26). When used in this position as a continuer, OKAY has a characteristic design: It is typically low and soft in volume; both syllables are stressed, with the second being at roughly the same height as the first. There may be a slight upwards glide at the end:
Pitch (Hz)
500 300 200 150 100 0
0.8207 Time (s)
Figure 10. Pitch trace for OKAY in line 21 of Excerpt 10 Waking up early
152 Elizabeth Couper-Kuhlen
The OKAY in line 7 is another instantiation of this pattern but is only barely audible. Although the standard continuers for English are considered to be uh-huh and mm-hm (Schegloff 1982), the newer data set examined here suggests that OKAY is joining the group. When used as a continuer, OKAY is delivered prosodically with low pitch and soft volume. It may have an optional low rise at the end, but overall it displays little pitch excursion. 4.5
OKAY in transitional positions
Early work by Beach (1993, 1995b) has pointed to so-called “continuative” uses of OKAY, in addition to the “non-continuative” ones which we have described as second-position and third-position OKAYs. “Continuative” OKAYs in Beach’s understanding are forward-looking; they prefigure a fuller turn by that speaker, often one that will entail a shift of topic or activity. As transitions, OKAY particles both close off what has preceded and foreshadow the initiation of a new topic or sequence. Despite this dual task, however, it has proven useful for the present investigation to distinguish OKAYs that primarily preface a new topic or sequence by the same speaker (often occurring after a break or significant pause), from OKAYs that primarily close a prior topic or sequence and are followed by other-speaker talk (often occurring before a significant break or pause). As we will see, one retrospective justification for making this distinction is that these two types of transition with OKAY have different prosodic-phonetic designs.
Preface to a new topic/sequence The following excerpt illustrates an OKAY that transitions to a new sequence, in this case to an announcement of the reason for the call: Excerpt 11. Reason I’m calling (Kamunsky-3_2), 1960s, American English, telephone, informal interaction [Alan is the caller, but Maryanne preempts the first topic to tell him about recently talking to a mutual friend who is caught in a triangular love relationship. Alan now announces that Bruce, the other man, has decided to back out.] 01
MRY:
02
ALA:
well that’s good.[at least it’s (o:[:ff?)
03
ALA:
finally.=
04
ALA:
=that[s what I-
05
MRY:
06
ALA:
[so he
[eeyeah.
]
[(close) the]su:bject, that’s what hhIhh told him I go it’s ab(h)out t(h)i:me.
Chapter 5. Prosody and phonetics of OKAY in American English 153
07 08 09
MRY: ALA:
you kno[w. [Go::::::[:::::d] [.hhh ]
10=> ALA:
↑O↓kay;
well the reason I’m calling=
12
ALA:
13
MRY:
=there[is a reason behind my madness.
14
MRY:
uh-huh,
15
ALA:
uh next saturday night’s a surprise party here for p-
[°( ).°
16
kevin.
17
(0.2)
18
ALA:
!p and if you can make it.
19
MRY:
OH REALLY::::, =
20
ALA:
=yeah.
Alan brings his telling to a close by reporting what he said to Bruce after the break-up (lines 6–7), whereupon Maryanne produces an appreciative response cry (line 8). The way is now clear for a change of topic, which Alan introduces with a transitional OKAY (line 10). He then proceeds to an announcement of the reason for his call, which is to invite Maryanne to a surprise party for their mutual friend Kevin (lines 11, 15–18). The transitional preface OKAY in Excerpt 11 is delivered with stress on both syllables, the first syllable being higher in pitch than the second. The second syllable has a step down of three semitones and is somewhat curtailed in length.
Pitch (Hz)
500 300 200 150 100 0
0.3609 Time (s)
Figure 11. Pitch trace for OKAY in line 10 of Excerpt 11
Transitional OKAYs that function as prefaces display a certain amount of variation in the size of the pitch interval between the first and the second syllable but they typically begin higher than same-speaker prior talk. They can be curtailed and/or have fast speech rate.
154 Elizabeth Couper-Kuhlen
Closure of prior topic/sequence OKAYs which bring a prior topic or sequence to a potential close in the service of transitioning are often characterized by a form of prosodic stylization. In the following extract from the same conversation as Excerpts 1 and 5, OKAY is not only stylized but also delivered in a high pitch register: Excerpt 12. What are you gonna wear (HGII_20), 1960s, American English, telephone, informal interaction [Hyla and her friend Nancy are planning to go to the theater together in the evening.] 01
NAN:
↑what’re you gonna wea::r;
03
NAN:
just nice pa:nts,=or some[thing,]
04
HYL:
02
(0.9)
05 06
NAN:
07
HYL:
08
NAN:
09
]nna get dressed,=
[°°okay;°°] =cause it’s supposed to> rai:n tonight;=t[oo:; ] [OH
HYL:
11
NAN:
12
[ least] there’s a cha:nce of it; ↑Okay::_ ==
HYL:
=
14=> NAN:
15
(.)
16
HYL:
17 18
]
that’s r[i:ght.]
10
13
[yeah. ] BEE:
=>O-KAY SHE:
=that’s no:t >I mean< that’s not thee reason I’m not
10->
going.
11=> DEB:
↑´mm↓kay_↑
12
(1.0)
13
DEB:
.hh °okay° .hhh alright well I’ll call jay tee:,
14
and I’ll just tell him tha:t youknow we gotta’n extra
15
ticket or whatever.
When Debbie accuses Shelley of prioritizing guys over girls (line 1), Shelley initially denies that this is case (line 3). But she then goes on to concede that it would indeed be easier for her if her boyfriend came along (lines 5–6). This admission is immediately followed by a contrastive move insisting that the boyfriend’s not coming is not the main reason for Shelley cancelling the trip (lines 9–10).13 The latter, however, is precisely what Debbie’s original injunction just don’t blow off your girlfriends for guys (lines 1–2) implied. Debbie thus now ‘stands corrected’. It is in this context that her subsequent prosodically and phonetically marked OKAY (line 11) must be interpreted. Debbie’s OKAY is produced as mm-kay, delivered with high pitch register (245–325 HZ) and an upwards glide on the first syllable. With the high pitch register Debbie comes off as innocently distancing herself from any derogatory imputation concerning Shelley’s intentions.14 The OKAY token itself signals acceptance of what Shelley has claimed, but at the same time, because the first syllable here is articulated with closed lips, its delivery might be heard, on purely iconic grounds, as embodying some resistance. The subsequent 1.0 second pause may be providing space for Shelley to counter. However, when no further talk is forthcoming, Debbie ‘gives in’ (lines 13–15) – at least provisionally, as the next excerpt will show.
13. For more on concession as a rhetorical device in conversation see Couper-Kuhlen and Thompson (2000). 14. Debbie’s OKAY in line 4, produced in response to Shelley’s initial denial, has the same marked prosody (including high pitch register) and is arguably doing something similar.
160 Elizabeth Couper-Kuhlen
Pitch (Hz)
500 300 200 150 100
0
0.4974
Time (s)
Figure 15. Pitch trace for OKAY in line 11 of Excerpt 15
Several seconds later, Shelley continues to proclaim that her boyfriend is not the reason for her decision to back out and Debbie produces a similarly designed OKAY: Excerpt 16. Funding (Debbie and Shelley_3), 1960s, American English, telephone, informal interaction15 01
SHE:
now that he’s not going I have to pay for the whole
02
thing and: that’s fi:ne, except for: .hh you know I
03
have my sister coming in and stuff and I’m like well
04
do I really want to do this? well yeah I wanna do it,
05
but do I have the money to do: [all the]se things.=
06
DEB:
07
DEB:
[ri:ght.] =↑okay↑
08-> SHE:
so: I mean it’s not becuz he’s- he’s- I mean it’s not
09->
becuz he:’s not going it’s becuz (0.5) his money’s not
10->
(0.5) funding me.
11=> DEB:
↑ ´o↓kay ↑
12
SHE:
so and ↑when other time have I ever [done that.]
13
DEB:
14
[.hhh well ] I’m jus say:in it jus seems you-
15
you base a lot of things on-on guy:s.(.) I do’know:,
16
it just- a couple times I don- I don- .hh it’s not a
17
big deal.
This OKAY is also delivered with a strong upwards glide on the first syllable and high pitch register (270–360 Hz); it is again hearable as proclaiming innocence and, despite the lexical content of OKAY, mild disbelief at Shelley’s disclaimer. As can be seen from what transpires next, Debbie indeed believes there is evidence that speaks against what Shelley is asserting (lines 13–17). 15. See also Heritage (2012b, 23), who examines this extract without, however, discussing the prosody of Debbie’s OKAY.
Chapter 5. Prosody and phonetics of OKAY in American English 161
Pitch (Hz)
500 300 200 150 100
0
Time (s)
0.9112
Figure 16. Pitch trace for OKAY in line 11 of Excerpt 16 Funding
In epistemically driven sequences, prosodic marking on a responsive OKAY will be interpretable in relation to the main action of the turn, namely accepting and/ or acknowledging the information that has been imparted. The examples we have seen here have involved high pitch register, although other types of prosodic marking such as widened pitch span and increased volume are also documented in the collection.
Deontically driven sequences In deontic sequences, especially ones in which a request has been made, a prosodically-phonetically unmarked OKAY produced in next position will be hearable as the recipient straightforwardly agreeing to or committing to comply. The use of prosodic-phonetic marking on responsive deontic OKAYs is interpretable with respect to the action of compliance. Here is a case where, again, high pitch register is involved. This exchange takes place later in the conversation in which Hyla and Nancy are making arrangements for the theater (see Extracts 1 and 5 above). Excerpt 17. Book back (HGII_17), 1960s, American English, telephone, informal interaction 01-> HYL:
.hhh oh and you know what I want my book ba:::ck.=
02
=your book.
NAN:
03=>
↑ ´O↓`kay:; ↑
04
I’ll have to look for it,=
05
HYL:
=dUhhhhh=
06
???:
=(k-k-k)=
07
HYL:
=.eh-.uh .hhh
08
(0.2)
09
NAN:
10
HYL:
11
I don’t know where it [is but I’ll fin[d it.
]
[°.hhhhhhhhhhh° [#u.-oh#]:: alright,=
162 Elizabeth Couper-Kuhlen
Hyla’s request that Nancy return the book she has lent her (line 1) comes completely out of the blue: Note the oh preface, suggesting it is something she has just thought of (Heritage 1984a; Bolden 2006). In next turn, after partially repeating its key element (your book), Nancy signals her agreement to comply with OKAY (line 3), but she goes on to claim that she will have to search for it (line 4). Nancy’s OKAY is prosodically-phonetically marked not only through the strong upwards glide on its first syllable and wide falling pitch on the second (with an interval range of nine semitones) but also through its high pitch register (228–383 Hz). The effect achieved is that of utmost willingness, if not eagerness, to fulfil Hyla’s request – an affective display possibly designed to offset the upcoming announcement that she does not know where the book is (with its undesirable implication that she may have lost it).
Pitch (Hz)
500 300 200 150 100
0
Time (s)
0.5586
Figure 17. Pitch trace for OKAY in line 3 of Excerpt 17 Book back
Yet prosodic-phonetic marking can achieve something quite different in responsive deontic OKAYs. Here is a case where a request, accompanied by an OKAY tag, is responded to with a marked OKAY. This exchange takes place in the Farmhouse conversation, before Excerpts 2 and Excerpt 7 shown above: Excerpt 18. Chunky chocolate ice cream (Farmhouse_30), 1998, American English, telephone, informal interaction [Michelle, who is a weight watcher, is complaining to her friend Laura that she (Laura) left some chunky chocolate ice cream in Michelle’s freezer.] 01
MIC:
[and I’m >sittin’ there< staring ] at it
02
saying ↑I’m not gonna eat this, I’monna-
03
°(h)I(h) was gonna° bring it over this afternoon
04
but (0.2) forgot.
05->
*so you have to get that.=o↑kay? *gazes at LAU->
06
(0.5)
07=> LAU:
+°nOkay_° +gazes down at her hands on table
Chapter 5. Prosody and phonetics of OKAY in American English 163
08
I’ll try [to [remember.
09
MIC:
10
DON:
11
MIC:
*[cuz otherwise, *gazes at others [hee hah hah hah I’m throwin it out:(h)! £hhh£=
In lines 1–2 Michelle complains that having the ice cream in her freezer requires her to exercise vigilance in order not to eat it and in lines 3–4 she claims that she intended to return it to Laura herself but forgot. She now pleads with Laura to come pick it up (line 5), whereupon Laura, with some delay, produces a token of compliance with OKAY (line 7). Yet although Laura’s OKAY is positioned to signal compliance, there are a number of indications, including prosodic-phonetic marking, to suggest that her willingness to commit is only half-hearted.16 First, on a visual level we can note that Laura breaks away from Michelle’s gaze at the end of line 5 to gaze down at her hands. Second, she delays her response by 0.5 seconds and only then produces an OKAY which is low-pitched and has soft volume (line 7). The first vowel of OKAY begins with a nasal sound, resulting in a curious hybrid of ‘no’ and ‘okay’. This OKAY is also marked because it departs from local prosodic norms. As Figure 18 shows, the pitch is roughly the same on both syllables:
Pitch (Hz)
500 300 200 150 100
0
0.7757 Time (s)
Figure 18. Pitch trace for OKAY in line 7 of Excerpt 18
Laura’s OKAY in Excerpt 18 thus has all the trappings of an OKAY that does not mean ‘okay’ (Beach 2020). Rather than suggesting full commitment to comply, it remains ambivalent about whether Laura will pick up the ice cream at all. As she herself goes on to remark, she will try to remember, implying that she may actually ‘forget’ to do so (line 8).
16. Kendrick and Torreira (2015, 274) discuss further examples of responsive OKAY whose prosodic-phonetic marking indexes that the speaker’s commitment to comply is qualified.
164 Elizabeth Couper-Kuhlen
In sum, in deontically driven sequences, prosodically-phonetically marked variants of OKAY become interpretable in relation to the compliance or acceptance that such a response is signaling. In one case examined here, a wide rising-falling glide as well as high pitch register index strong willingness to comply, while in the other case low pitch, soft volume and level intonation are indicative of the opposite kind of stance: reluctance and lack of enthusiasm. 6. Summary and provisional conclusions At the conclusion of our survey, readers may feel bewildered at the number of different prosodic patterns for OKAY particles in the older and newer collections. And it is true that there is a good deal of variation. Yet recurrent patterns of delivery can be identified: The overview of prosodic-phonetic patterns at the end of the chapter shows the ones we have singled out here for discussion. Several of the patterns shown in this overview are binding, in the sense that this is the only way OKAY is done in such and such a type of turn or sequential position. If a different pattern were to be used, sequential position permitting, OKAY would be doing a different job. For instance, if OKAY accompanying a first-position action were to be delivered not with a low, short, soft first syllable and up-stepped or rising pitch on the second syllable but, say, with high pitch onset on the first syllable and a step down or fall on the second, this would be hearable as the speaker abruptly transitioning to a different matter. If a sequence-internal (continuer) OKAY were produced with a pattern other than low pitch, soft volume and minimal pitch excursion, the speaker would be heard to be taking a full turn at talk. And in both epistemic and deontic sequences, final rising pitch on second-position responsive OKAYs is binding if the prior turn is to be treated as preliminary. Without this prosodic-phonetic design, there is no such implication. Other patterns are not binding but instead constitute particularly clear instances of the sequence-specific work accomplished by OKAY. This is, for example, the case of the stylized patterns found when OKAY is proposing to close a sequence/topic or to pre-close a conversation. Not all sequence/topic closing OKAYs nor all conversational pre-closing OKAYs are prosodically stylized, but when they are, they identify OKAY as doing this job. The same might be said of the stylized pattern used to acknowledge a counter-informing (Excerpt 9): This design is not binding but when present, it identifies the responsive OKAY as doing this particular job. Finally, there are a variety of patterns found in second and third position in epistemic and deontic sequences that are neither binding nor identificatory. We
Chapter 5. Prosody and phonetics of OKAY in American English 165
have attempted to single out ‘neutral’ designs for actions implemented straightforwardly by second- and third-position OKAYs, and have contrasted these with similar cases in which the prosodic-phonetic marking of OKAY leads to emotive or affective overtones. What exactly the overtones are depends heavily on the nature of the prosodic-phonetic marking and the type of sequence in which the OKAY is embedded. OKAY produced with glides to high and with high pitch register can imply disbelief in epistemic sequences but keenness to comply in deontic sequences. Low pitch and volume as well as level pitch on OKAY in similar deontic sequences can imply lack of full commitment. We conclude that the picture for OKAY and its prosodic-phonetic design in American English is complex, but not without its regularities. 7. Some observations on changes in the use of OKAY over time Has American English OKAY changed over time in its use as a particle? If so, how? In the following, we note changes in frequency, positional use, and prosodic-phonetic design of OKAY, based on observations that have emerged through a comparison of the older and newer collections.
Frequency OKAY occurs less frequently in the newer data, where there is on the average one token every 2 min. 22 sec., than in the older data, where one token occurs on the average every 1 min. 26 sec.: see Table 2.17 The greater frequency of OKAY in the older collection may have to do with the recent rise of new competitors for OKAY, e.g., right or alright (Gardner 2007). But it may simply be an artifact of the data: The newer collection includes data from face-to-face conversations and, in the case of the Call Friend/Call Home telephone calls, 30-minute excerpts prior to conversational closings. Consequently, there are no conversational pre-closing OKAYs in the newer collection. Furthermore, there is one conversation in the older data (Kamunksy-3) where the frequency of OKAY is well above average (one token every 19 sec.) due to the presence of driving instructions.
17. These frequency counts include not only freestanding OKAY but also OKAY in combination with other particles, e.g., oh okay, oh okay right, yeah okay, yeah right okay.
166 Elizabeth Couper-Kuhlen
Table 2. Frequencies for OKAY in the older and newer data (including oh okay, oh okay right, yeah okay, yeah right okay) Older data Conversation Debby & Shelley Hyla II Joyce & Stan Kamunsky-3 NB 27 NB 28 SBL 10 SBL 11 SBL 15 SF2 Two girls Total
Duration
Tokens
Frequency (1 token every…)
457 s 1200 s 508 s 807 s 107 s 2671 s 738 s 232 s 208 s 880 s 985 s 8793 s = 146 min 55 s
8 17 11 42 3 8 3 3 8 8 5 116
57 s 71 s 46 s 19 s 36 s 334 s 246 s 77 s 26 s 110 s 197 s 76 s = 1 min 16 s
Duration
Tokens
Frequency (1 token every…)
1085 s 1800 s 1800 s 1800 s 300 s 524 s 1800 s 1800 s 1534 s 1140 s 1140 s 2316 s 1320 s 1517 s 19,876 s = 331 m 27 s
4 26 11 9 4 3 25 14 1 8 10 20 7 7 149
271 s 69 s 164 s 200 s 75 s 175 s 72 s 129 s 1534 s 143 s 114 s 116 s 189 s 217 s 133 s = 2 min 22 s
Newer data Conversation Before Bed Call Friend 4984 Call Friend 5926 Call Friend 6239 Call Friend 6255 Call Friend 6278 Call Friend 6899 Call Friend 6938 Call Home 4544 Camp Reunion-1 Camp Reunion-2 Farmhouse Game Night Hey Cutie Pie Total
Chapter 5. Prosody and phonetics of OKAY in American English 167
Positional use There are some interesting developments in the way OKAY is used now (data from roughly 1994 onward) as compared to its earlier use (1960s). A number of these developments are captured in Table 1: i. OKAY is found more often as a follow-up or tag accompanying a first-position action in the newer data. Although not evident from Table 1, OKAY appears more frequently as a tag (no intervening silence or talk) than as a follow-up (with intervening silence or talk) in the newer data. Moreover, the nature of the first action that OKAY accompanies is changing. In the older data, the action was always one that mandated acceptance or rejection, e.g., an invitation or proposal (I wanna buy you a drink, I’ll see you about eight). But in the newer data the action to which OKAY is attached is on occasion a self-deprecating assessment (not my speciality) or a prediction/promise (it’s just not gonna happen). These actions do not mandate acceptance or rejection; at the most they invite acknowledgment, although the speaker does not pursue this if absent. ii. The use of OKAY and OH OKAY in third position in epistemically driven sequences (Question-Answer-OKAY/OH OKAY) has risen dramatically since the 1960s. iii. OKAY is found as a continuer (floor pass) in the newer data; this use was lacking completely in the older data. iv. OH OKAY (not dealt with here) is found in second and third position of epistemic sequences only, and exclusively in the newer data.
Prosodic design For a number of the uses of OKAY, the same prosodic patterns are documented in both the older and newer collections. This is the case of OKAY as a follow-up to an initiating action (e.g., Excerpt 1), of unmarked responsive OKAY in deontic and epistemic sequences (e.g., Excerpt 3 and Excerpt 5), and of unmarked third-position OKAY in deontic and epistemic sequences (e.g., Excerpt 7 and Excerpt 8). It also holds for final rising second-position OKAY in deontic and epistemic sequences (e.g., Excerpt 4 and Excerpt 6) and it applies to transitional OKAYs (e.g., Excerpt 11/ Excerpt 12). In all these cases, although the illustrative example may have come from one of the collections, the same pattern is attested in the other collection. And yet there are prosodic patterns for OKAY that are documented only in the newer collection. This is trivially the case for the continuer OKAY, which is not found in the older data set. However, the tag use of a latched OKAY to accompany
168 Elizabeth Couper-Kuhlen
first-position epistemic actions (e.g., Excerpt 2) is not found in the older collection, and the stylized form of OKAY for responding to counter-informings (e.g., Excerpt 9) is attested only in the newer collection. This may be a genuine innovation. Given the developments so far, we can undoubtedly expect both constancy and change in the frequency, positional use, and prosodic design of OKAY over the years to come.
Overview of prosodic-phonetic patterns in different turn and sequence positions OKAY accompanying a first-position action
(1) “A drink” (Older collection)
o ↑KA:Y,
Pitch (Hz)
500 300 200 150 100 0
0.4397 Time (s)
(2) “Tax and tipping” (Newer collection)
o ↑KA:Y_
500 Pitch (Hz)
300 200 150 100
0
0.5158 Time (s)
Chapter 5. Prosody and phonetics of OKAY in American English 169
OKAY in the second position of a sequence Deontic
(3) “University address” (Newer collection)
o `ka:y;
Pitch (Hz)
500 300 200 150 100
0
0.4407 Time (s)
(4) “Some suggestions” (Older collection) o ↓´ka:y?
Pitch (Hz)
500 300 200 150 100
0
0.7379 Time (s)
Epistemic (5) “A drink”, cont’d (Older collection)
↑o ↓kay;
500 Pitch (Hz)
300 200 150 100 0
0.4843 Time (s)
170 Elizabeth Couper-Kuhlen
(6) “Native speaker” (Newer collection)
°o ↓ka:y,°
Pitch (Hz)
500 300 200 150 100 0
0.6994 Time (s)
OKAY in the third position of a sequence Deontic
(7) “Water and coffee” (Newer collection) >↑o °kay;°
well< when does he get home? (0.6)
jam:: ø:h hva’ er klokken nu?= well u:h what time is it now?= =den e::::r o:tte =it i::::s eight (0.4)
↑ja::; (0.2) >han er ikk’< hjem:: fø:e:r mm ↑well::; (0.2) >he isn’t< hom::e befo::re mm
07
(0.2) °halv ti kvart i ti° (0.2) °half past nine quarter to ten°
08
(0.3)
09
SIM:
11
LIV:
10
↑ohkay.
(.)
>↑jeg syns< du ska’ prø:ve å ringe °til ham° >↑I think< you should try: to call °him°
After an insertion sequence, through which Liv gets a clarification of what time it is, she provides an answer in lines 6–7. Simon receipts this answer with an OKAY in line 9. After a brief silence, Liv adds a recommendation, which builds on her having heard that the OKAY treated her answer as sufficient and informative. The pitch of OKAY begins at a high point in Simon’s register, the first syllable is stressed and the second syllable has a falling pitch. This is an instance of what we call okay1. Knudsen claims that “through the use of it [okay1], the producer indicates that the information received is complete, and, therefore, that the user is ready to move on to a new topic, a new activity, or to resume an activity that was put on hold” (2015, 41).1 Thus, the sequential environment of okay1 is after an answer to a question, where it receipts that answer as sufficient in both informational and sequential terms. Excerpt 2 presents the use of an okay2, again from Knudsen (ibid., 24). We enter at a point when Elin (ELI) is telling Lina (LIN) about the boyfriend of a mutual acquaintance:
1. Our translation. The original says: “med brugen af den indikerer producenten, at den modtagne information er komplet og at brugeren derfor er klar til at bevæge sig videre til et nyt emne, en ny aktivitet eller til at genoptage en aktivitet, der står i venteposition.”
Chapter 6. Rising OKAY in third position 177
Excerpt 2. gammel elevfest ‘former student party’ (ADK:02:L&E:7.0:gammelelevfest), 2002, Danish, telephone, everyday 01
ELI:
altså han hun er kærester med well he she is dating
02
en eller anden som sådan er helt vildt somebody who is like totally max
03
økofreak og de har lige været på sådan en eco-freak and they have just been on such a
04
KULhydrat kur å::: CARBONhydrate cure an':::
05
(0.2)
06
07
LIN:
ELI:
o£ka[y?£
[£han laver alt muligt sådan mærkelige ting£. £he makes all kinds of strange things£.
Elin depicts the boyfriend as a ‘totally max eco-freak’ (lines 2–3), and starts giving instances of the freaky nature of the relation by telling about a strange cure. This ends with a conjunction (å::: ‘an'::’) in line 4, which indicates that the telling may continue. There is a brief silence in line 5, after which the recipient of the telling utters an OKAY (line 6). Immediately after this, Elin continues her telling (line 7). Thus, Lina’s OKAY indicates that she understands the telling, and that it may be continued, which is also what Elin orients to when she does continue. In this case, the OKAY is stressed on the last syllable, on which the pitch also rises to a high point of the speaker’s register, that is, a typical instance of okay2. It comes at points when new information is given, and where the speaker indicated that there is more to tell, and, according to Knudsen, this version of OKAY “elicits a continuation from the informing party” (2015, 30).2 Following Knudsen (2015), we use the terms okay1 and okay2 to refer to the prosodic formats of the tokens we analyze: okay1 is stressed on the first syllable and has a falling contour, okay2 is stressed on the last syllable, and has a rising contour. Knudsen’s examples, cited above, demonstrate what, at the outset, we see as the typical sequential environments and interactional functions of the two OKAY types: okay1 occurs in third position after questions and answers, and treats the answer as informative and sufficient; okay2 occurs after utterances that are designed to be parts of larger packages, and it accepts and elicits the projected continuation of that package. However, the data supporting Knudsen’s findings (2015) and similar observations by Mortensen and Mortensen (2009) come from small data sets, and their conclusions are tentative. Therefore, our first research question is, “Can we find the 2. Translated from “eliciterer en fortsættelse fra informantens side.”
178 Søren Sandager Sørensen and Jakob Steensig
same distinction between okay1 and okay2 in a larger data set?” It turns out that we can, but also that a number of instances of the rising okay2 occur in the same position as okay1, that is, as a receipt to answers to questions, where continuation is not projected. This leads us to our main research question: What is the interactional logic behind the use of okay2 tokens in seemingly okay1 environments? This chapter begins by providing some background on the Danish OKAY types in the available literature. Then we describe our data and the procedures behind our extraction of instances. The analysis section begins with a quantitative overview of the two types of OKAY in their relevant sequential positions. This takes us to an analysis of okay1 and okay2 in their most frequently occurring positions, the ones exemplified above. After this, we analyze the okay2 tokens that occur in third position, according to their main jobs: indicating that the answer should be expanded, and dealing with unresolved matters. Finally, we discuss the logic behind the use of okay2 in what could be termed okay1 environments, and whether the okay2 retains some of the features found in its “continuer” positions. We also raise questions about the role of prosody, and make some suggestions about how this may be further investigated. 2. Background: Danish OKAY The word OKAY (often OK in Danish written language) was first documented in dictionaries from the 1930s (Ordbog over det danske Sprog n.d.). It has become a very widely used word. In our 20 hours of various types of interaction, we registered 483 OKAY tokens of the types we were investigating (see Section 3 below). In a subset of three hours, we systematically counted OKAY tokens of all types. There were 95 tokens in total, which means that this word was uttered every 1.9 minute on average. Only 6 of the 95 OKAY tokens are adjectives or adverbs, so we conclude that OKAY functions mainly as an interjection in talk-in-interaction. When OKAY was borrowed from American English, its phonetic shape needed to be adapted only slightly to Danish phonology. Danish already has diphthongs that come close to the American English versions, viz, [ɔʊ̯] as in bog [b̥ɔʊ̯ˀ] ‘book’ and [ɛɪ̯ ] as in mægle [ˈmɛɪ̯ lə] ‘mediate’.3 Plosives (such as the /k/ in OKAY ) are aspirated [kʰ] in the beginning of stressed syllables and in the beginning of unstressed syllables in words of foreign origin, which includes OKAY, so the standard pronunciation of OKAY in Danish is [ˈɔʊ̯kʰɛɪ̯ ] or [ɔʊ̯ˈkʰɛɪ̯ ] depending on the stress (Den 3. This rather crude IPA representation of the vowels will be used consistently in this paper. The more precise IPA cardinal vowel representation of the Danish [ɛ] is [e] or [e̞], and of [ɔ] probably [ɔ̝̟ ̟], that is, a raised and advanced version of [ɔ].
Chapter 6. Rising OKAY in third position 179
Danske Ordbog [DDO] n.d.; Ordbog over Dansk Talesprog [ODT] n.d.). According to ODT, OKAY also exists with monophthongs instead of diphthongs, which corresponds to what we have found. In our data, the pronunciations [ɔʊ̯khɛ], [ɔkhɛɪ̯ ] and [ɔkhɛ] are found, and so are versions with prolonged vowels and without aspiration of the /k/. In our data, we also encounter OKAY tokens without any audible first syllable, which may sometimes be due to the recording or noise. It is also possible to utter an OKAY token wherein the Danish pronunciation of the two letters, “O” and “K,” is used, which in Danish would be [ˈoːˀ ˈkʰɔːˀ]. There are only two instances of this in our data, and they do not appear in this paper. The difference between the forms that we call okay1 and okay2 is attested in the dictionaries (DDO; ODT), where the pronunciations with stress on first and stress on second syllables are mentioned and exemplified, but the dictionaries do not systematically address the functional differences that we exemplified in the introduction. 3. Data and method The data for this chapter come from 20 hours of recordings of naturally occurring interaction from two corpora: the publicly accessible Samtalebank (MacWhinney and Wagner 2010; Talkbank 2018) and AULing, a collection of recordings from Aarhus University (Samtalegrammatik.dk 2018). Most of the recordings are videotaped face-to-face interactions, but we have also included a smaller subset of audio-recorded phone conversations (around 1 hour, 42 minutes). Most of the recordings come from private, everyday settings, in which people are chatting, cooking, eating, and playing together. A smaller part is “institutional,” either layprofessional interactions (interactions in shops, advice-giving) or task-oriented interaction between peers (mainly study groups). For the purpose of this chapter, we have not made any a priori distinction between types of data (e.g., in terms of institutional roles, task-orientation, turn-taking system, modality), but we do address such factors in our analyses, when relevant. All participants gave their informed consent to being recorded prior to the recordings, and transcripts have been anonymized so that participants and people they discuss cannot be recognized.4 The transcripts of the recordings have been searched for all occurrences of OKAY. As the transcripts have different levels of quality and detail, we have also 4. The AULing corpus is registered with the Aarhus University data protection register with the ID no. 923, and registered with the Danish Data Protection Agency (“Datatilsynet”), and kept in a safe location, in accordance with the general data protection regulations of the European Union (EU 2016).
180 Søren Sandager Sørensen and Jakob Steensig
listened to, and looked through, most of the recordings to be sure that we located all the OKAYs, and that the OKAYs that were transcribed were actual instances of OKAY. We retranscribed all excerpts that we investigated in depth, in accordance with the transcription conventions of this book (see Appendix). For the present investigation, we focused only on “non-integrated” OKAY tokens. This means that of the 483 registered OKAY tokens, we have excluded instances in which OKAY occurs as a constituent in a clause or phrase (i.e., OKAY as an attributive modifier or as a predicative complement) and instances in which OKAY is prosodically integrated with other tokens (as in ja=okay ‘yes=okay’, nå=okay ‘oh=okay’) or repeated (okay=okay) or inside a quote. Thus, all the remaining OKAY tokens (356 tokens) potentially perform actions of their own. We made an initial overview of these OKAY tokens, in which we categorized the tokens according to sequential position, possible action, and prosodic and phonetic characteristics. This overview formed the basis of our qualitative investigations of the regularities of okay1 and okay2, and of prosodic and phonetic features. 4. Analysis In this section we take a closer look at the distribution of the OKAY tokens, and we analyze okay1 tokens in third position and okay2 tokens used as continuers. This will form the backdrop for our main analysis sections, in which we try to find the logic of okay2 occurring in third position, and how this differs from what okay1 most often does in this context. 4.1
Distribution of tokens across positions
The configurations that we exemplified in Excerpts 1 and (2) are rather stable in our data. Of the 356 non-integrated OKAY tokens, 217 (61%) are clearly distinguishable okay1 tokens, and of those, 153 deal with managing understanding, that is, they primarily deal with the local sufficiency of the content of the utterances to which they respond (see Chapter 3, this volume), and not, for instance, with accepting or complying with an action or indicating an activity transition (see Chapter 4, this volume). 98 tokens (28% of all the non-integrated tokens) are distinguishable as okay2, of which 82 deal with understanding.5 Table 1 shows the number of distinguishable okay1 and okay2 tokens that occur in the environments illustrated by 5. The rest of the tokens (42 tokens or 12%) are ones we have not been able to distinguish clearly, either because they are difficult to hear or because their prosodic shape is unclear.
Chapter 6. Rising OKAY in third position 181
Table 1. Numbers of understanding-managing, freestanding okay1, and okay2 tokens in the data corpus Receipt in third position Receipt for unsolicited information “Continuers” Total
Okay1
Okay2
108 (70.6%) 28 (18.3%) 17 (10.8%) 153 (100%)
36 (43.9%) 10 (12.2%) 36 (43.9%) 82 (100%)
Excerpts 1–2, but with two types of receipts: either as sequence closing thirds in question-answer sequences (marked as “receipts in third position”) or in response to unsolicited information (marked as “receipts for unsolicited information”). Those marked as “continuers” occur at potentially incomplete places in others’ talk.6 Table 1 shows a strong tendency for okay1 tokens to occur as sequence closing thirds after answers to questions, as we saw in Excerpt 1 above. However, the pattern for okay2 tokens is less clear. Only slightly less than half of the clearly distinguishable okay2 tokens occur in the environment described by Knudsen (2015), and illustrated in Excerpt 2, that is, at points of potentially incomplete talk by a co-participant, whereas the rest are receipts. In Sections 4.2 and 4.3 below, we exemplify and extract relevant observations of what earlier research established as the typical pattern: okay1 as receipts in third position following answers to questions, and okay2 as continuers. In Sections 4.4 and 4.5 we focus on okay2 tokens occurring in apparent third positions, to determine the logic of okay2 occurring in seemingly okay1 environments. We will not analyze the cases of okay1 occurring as apparent “continuers,” as we have relatively few of these cases, and we have not been able to find any recurrent patterns. 4.2
Okay1 as a receipt in third position
Excerpt 3 shows an instance of okay1. It comes from an interaction among four women who are sitting in a lunchroom at their school, chatting and eating cake. We enter the conversation at a point when Ani has asked Crystal (CRY) about an apartment that Crystal, who is searching for a new place to live, was supposed to 6. As noted in Chapter 3, there may be a difference between OKAY used at points of projected continuation and typical “continuers,” in Schegloff ’s (1982) or C. Goodwin’s (1986) sense. We know too little about the use of words such as ja ‘yes’, nej ‘no’, and m ‘m’, ‘mhm’) in these contexts of Danish talk-in-interaction to describe their (possible) differences. Therefore, we have used the term “continuer” (in quotation marks) for the OKAY tokens used in these contexts, pending further research on the matter, to determine whether these OKAYs are used differently from the more prototypical continuers.
182 Søren Sandager Sørensen and Jakob Steensig
view. Crystal has just said that she has seen the apartment, that it was in the basement of a building, and that “it was really not very big” (not included, immediately precedes the excerpt). She goes on to specify the size of the apartment in line 1. Our focus is on the inserted question–answer sequence launched by Ani in lines 5 and 7. Excerpt 3. etværelses ‘one-bedroom’ (AULing:KC), around 2003, Danish, face-to-face, everyday
01 02
CRY:
·hnh >de havde jo< sagt den var fyrre kvadratmeter; ·hnh >they had PRT< said it was forty square meters; (0.3)
03
CRY:
04
CRY:
05
ANI:
[+å så var det +en] to[er; ] ikko]ss,+= and then it was a “two”; right,=
06
BIA: ani
[khrm:, ] +.............+points down----------+
07 08
09 10 11 12
ANI: CRY: ani cry cry cry
CRY: BIA: ani
men det var faktisk hele but that was actually all the underetag[en ( lower floor (
)(særlig] meget) ] )(particularly much)
=&+en to+værelses€lejl[ighe+] =a two-bedroom apartment
[m-& ] ((negative mm)) +.....+full hand gesture-+ &chews------------------& €shakes head-->
(0.2)€ -->€
=etværel[+ses.= =one-bedroom.= [+°khrm° +nods once-->
ANI: ani
= okay;+ -->+
14
BIA:
°nå:,°= °oh:,°
15
DIA:
=mm↓hm_ =mm↓hm_
16
CRY:
å det vil sige det sted hvor stuen der var and that is to say the place where the living room was
13
(0.2)
Ani asks her question in line 5 at a point of possible terminal onset (Jefferson 1984a) of Crystal’s turn in lines 1–4. Here, Crystal makes it clear that the apartment was less than the forty square meters ‘they’ said it was, and Ani now makes a statement
Chapter 6. Rising OKAY in third position 183
about Crystal’s situation, and presents it as a next part in Crystal’s account: ‘and then it was a “two” (colloquial for a two-bedroom apartment); right,’. In line 7, she adds a more elaborate version: ‘a two-bedroom apartment’, with an accompanying gesture. This amounts to a request for confirmation, as it is an A statement about a B event, and designed as an upshot formulation (Heinemann 2010; Heritage and Roth 1995; Heritage and Watson 1979; Labov and Fanshel 1977). Ani does not finish the word lejlighed ‘apartment’, because Crystal starts shaking her head once Ani has completed toværelses ‘two-bedroom’, and then Crystal answers Ani’s request for confirmation with a negative ‘mm’ sound, that is, with a cut-off m- (in line 8), and follows up with the correct term, ‘one-bedroom’, in line 10. Thus, Ani made an incorrect assumption, which was rejected, and in line 12, she receipts the rejection and indicates her revised understanding with an OKAY. We hear the OKAY in line 12 as an okay1; the pitch is lower on the second syllable than on the first, and the last syllable seems to have a falling contour. This is confirmed when we analyze the sound in the Praat software (Boersma and Weenink 2018.7 Figure 1 shows the soundwave, pitch (fundamental frequency in Hz, displayed logarithmically with a dotted line), intensity (in dB, thin line) of Ani’s OKAY in line 12.8 The height of the box showing the pitch represents the speaker’s pitch range based on one minute of talk by this speaker. Thus, it may be seen that the first syllable lies in the upper part of the speaker’s mid-register, it has a slightly rising contour, and then there is a drop to the second syllable, which then falls to a low part of the mid-register. The difference between the highest and lowest pitch across the whole word is 5.9 semitones, the downstep from the last part of the first syllable to the beginning of the second is 1.5 semitones, and the difference between the average pitch of the first and second syllables is 2.7 semitones. We hear the first syllable as stressed, although in terms of duration and intensity, the difference between the syllables is not very large. In accordance with the conventions for marking pitch on OKAY tokens in this book (see Appendix), we have indicated that the first syllable has level pitch (with an _ before the syllable) and that the second one is falling to mid (with ; after the word). This OKAY is a typical representative of an okay1: The falling tendency is clearly audible (and visible in Praat), but the fall is not dramatic, and neither syllable is significantly longer or louder than the other, though the first leans toward being slightly louder and longer than the second.
7. Accessed May 5, 2018. 8. We used a script devised by Gareth Walker (2017) to display our measurements in Praat.
184 Søren Sandager Sørensen and Jakob Steensig
350 75
250 70
200
Intensity (dB)
Pitch (Hz)
300
65
145 0
kh
εə
Time (s)
Figure 1. Acoustic properties of Ani’s OKAY in Excerpt 3, line 12
After Ani’s OKAY, the two other participants, who also showed an interest in Crystal’s apartment project, but did not participate actively in this sequence, also display that they have been informed, Bia with the change-of-state token °nå:° (approximately ‘oh’, Heinemann 2016a; Samtalegrammatik.dk 2019a), and Diana (DIA) with an acknowledging mm↓hm_ (lines 14 and 15). After this, Crystal continues her telling about the apartment, which she abandoned in line 4 because of Ani’s inserted question. Sequentially, the questioner’s okay1 functions as a sequence closing third (Schegloff 2007), marking that the inserted question-answer sequence may be completed, and the interactional trajectory may continue along its projected course. Most of the 108 okay1 tokens that occur in third position after questions and answers (see Table 1 above) update a questioner’s knowledge; the OKAY token accepts the answer as sufficient and, thus, orients to the sequence as potentially complete. In Excerpt 1, the answer did not give the precise information that the question sought, but it was still treated as informative and sufficient, and in Excerpt 3, the answer was contrary to expectations: It updated the recipient’s understanding, and it was treated as sufficient. Such answers may be receipted with the change-of-state token nå as described by Heinemann (2017a), and in Excerpt 3, we also saw one of the other participants use exactly that token. We also have cases of okay1 in third position, where the answer updates the questioner’s understanding, without it going counter to expectations. Excerpt 4 is an example of this. Four women are chatting together. In this excerpt, three of them are sitting around a table, and Dorte (DOR) has told of a Christmas event that she attended with her daughter. In line 1 she describes the candy bag that her daughter got at this event, using a gesture to display the enormousness of the bag.
Chapter 6. Rising OKAY in third position 185
Prior to this, Sus has told about candy at another event provided by a company, “Candinavia,” which, according to her, was nothing but “chemical waste.” Excerpt 4. oss fra Candinavia ‘also from Candinavia’ (samtalebank:sam4:moedregruppe1), around 2009, Danish, face-to-face, everyday 01
02
03 04 05
DOR:
dor sus
SUS: sus dor
MIA?: dor sus
DOR:
06
SUS:
07
DOR: dor sus
08 09 10 11 12 13 14
15 16
DOR: sus
SUS: sus
DOR: dor
SUS: sus dor sus
de fik (0.2) &sådde%n (.) en slikpose &her. they got (0.2) such (.) a candy bag like this. &looks at SUS------------&looks at MIA--> %looks tow DOR-->
€‘a_ €& yes_ €nods€ -->&
°(.hnej)°
&%(1.1) &looks at SUS--> %looks down at butter and knife in her hand-->
+.mt å det [var helt+ vildt_] .mt and it was completely crazy [%det var oss ]§fra [candinavia;§ that was also from Candinavia [å: and: +shakes head 1x-----+ -->%looks at DOR, does not move--> §nods 2x---------§ (1.1)
var det d%et_= was it_= -->%looks down, begins using knife--> §=ja_§ =yes §nods 1x§
det var det det så jeg i avisen; it was I saw it in the newspaper; (0.3)
_ok&ay;& -->&looks down, retracts head, smiles&
men-%& candinavia har vel nok andet end but- Candinavia surely has other things than -->%looks at DOR, uses knife--> &looks at SUS-->> kemisk affald de har oss: %andre ting; chemical waste they also have: other things -->%looks down-->> vil jeg tro; I assume
186 Søren Sandager Sørensen and Jakob Steensig
Dorte’s assessment in line 5 potentially completes her story, but she has just initiated it when Sus interrupts with a piece of occasioned information: ‘that was also from Candinavia’ (line 6). First, Dorte indicates that she will continue (line 7), then she breaks off, and during a 1.1 second silence in line 8, Dorte and Sus look steadily at each other, all movement momentarily suspended. In line 9, Dorte requests a reconfirmation with ‘was it_’. This is not designed as especially incredulous or surprised, but given that it asks about something that may be counter to what Dorte expected, it seeks more than just a confirmation, that is, it is an expansion eliciting question (Steensig and Heinemann 2013). Sus answers with a short ‘yes’, with an even pitch contour (line 10), accompanied with a brief nod, and then in line 11 goes on to substantiate her piece of information by identifying her source. Dorte receipts this with an OKAY in line 13. At the same time, she stops looking at Sus, she retracts her head and smiles briefly. The answer is not in any way counter to Dorte’s expectations, as she already had the main information before she asked for reconfirmation, but she does get an account that updates her knowledge and may revise her understanding, and this seems to be sufficient for her. At least, she withdraws from her engagement, and she does not continue enquiring into the matter. However, Sus continues with possibly apologetic speculation that addresses the potential offense that may have been caused by her possibly implying that Dorte lets her daughter eat chemical waste. The first syllable of the OKAY token is stressed, realized mainly by higher intensity on that syllable, and by the pitch being higher on the first than on the second syllable. Its pitch lies in the upper part of the speaker’s register, and there is a fall of around 4.8 semitones from the highest point of the first syllable to the lowest of the second. Figure 2 shows the acoustic details. We hear this as a clear instance of an okay1. It is a rather unmarked case, in that the fall is clearly audible but not very steep, and the difference in intensity is also minor, compared to the more dramatic cases of the okay1 type. Excerpts 1, 3, and 4 are representative of our okay1 tokens in third position after question–answer sequences. They are used to indicate both substantially revised understanding and change-of-state, as in Excerpt 3, and an updated, more expected, current understanding, as in Excerpts 1 and 4 (see Chapter 3 for a treatment of the difference of these orientations).9 Thus, they display some degree of change-of-state and, importantly, they express acceptance of the information as sufficient, so that the question-answer sequence can be completed.
9. We have not been able to establish a clear difference in the prosody of tokens that “claim sufficient understanding” and tokens that “mark a discrepancy in expectation” (Chapter 3).
Chapter 6. Rising OKAY in third position 187
70
250 200
60
129
ε
k
0
Intensity (dB)
Pitch (Hz)
300
Time (s)
Figure 2. Acoustic properties of Dorte’s OKAY in Excerpt 4, line 13
4.3
Okay2 as a “continuer”
Excerpt 5 comes from a recording of three persons who are cooking a large meal in a big kitchen. Previously, Mette (MET) announced that her son has stopped smoking, and is now telling the others about how he began to smoke and came to realize that he had to stop. Thus, this is part of an unelicited telling. Excerpt 5. for nogen måneder siden ‘some months ago’ (samtalebank:sam3:225_deller: 429), around 2009, Danish, face-to-face, everyday 01
MET:
så STARTEDE han jo på amerikansk fodbold then he STARTED you know playing American football
02
ø::h (.) n:ede i >odense nu< her fo::r (0.9) uh (.) d:own in >Odense now< he::re (0.9)
03
nogen måneder siden, some months ago,
04
TOM:
oka:y,
06
MET:
å så tror jeg han kom hjem fra træning, and then I think he came home from practice,
05
07 08 09 10
MET:
MET:
(0.9)
(0.2)
å s:: så::=ø:h måtte han sande at hvis han ville det, and th:: then::=u:h he had to realize that if he wanted that, (0.6)
så måtte han holde op me’ å ryge. then he had to stop smoking.
188 Søren Sandager Sørensen and Jakob Steensig
From the outset, this story was introduced as being about how the son stopped smoking (not included), and at the beginning of this excerpt, this crucial event has not yet been told. So, when Mette says, in lines 1–3, that her son started playing American football, it is clear that the story cannot yet be over; a continuation is projected. At this point, the main recipient of the story, Tom, says OKAY (line 4). After this, Mette continues her story, and in lines 8–10, reaches the announced climax of how the son realized that he had to stop. Thus, this OKAY is produced at a point when the story is projectedly incomplete, and it is treated by the storyteller as a “go-ahead” or a “continuer” (Couper-Kuhlen and Selting 2018, 511f.; C. Goodwin 1986; Schegloff 1982), by which the recipient indicates that the storyteller should continue their story. Figure 3 shows the pitch, intensity and sound of the OKAY token in Excerpt 5. The first syllable begins slightly above the middle of the speaker’s pitch register, it is level, and then there is an upstep (of 1.8 semitones) to the second syllable, which rises to a mid-high level. The difference between the lowest and highest points is 4.7 semitones, and both the upstep and the pitch rise at the end are clearly audible. The second syllable is stressed: It has higher intensity, considerably longer duration, and more pitch movement. This is typical of the tokens we hear as okay2: They have an audible upstep and/or rise on the second syllable, the second syllable tends to be longer than the first, and longer than average second syllables on okay1. This seems to be a rather stable configuration. The OKAY tokens that come at points when an activity is potentially incomplete, and which are understood as proposing a continuation of that activity, are okay2 tokens. These OKAY tokens also address understanding, in the sense that they indicate that the OKAY speaker
80
300
75
200
70
150
65 95 u
εi
k Time (s)
Figure 3. Acoustic properties of Tom’s OKAY in Excerpt 5, line 4
Intensity (dB)
Pitch (Hz)
250
Chapter 6. Rising OKAY in third position 189
has understood what has been said so far, so that the speaker of the informing or telling may continue. We have now outlined the interactional and prosodic features of okay1 and okay2 in the positions that fit the results from earlier research. We will now analyze instances that seem to differ from those earlier descriptions: rising okay2 tokens in third position after answers to questions. 4.4 Okay2 indicating that an answer was the beginning of a telling In some cases of okay2 in third position, the answer to the question that the OKAY receipts is treated as being only the beginning of a telling or an account, which should be expanded. Excerpt 6 is an instance of this. It comes right after the opening of a phone call. Ann has called a landline phone and Maria (MAR) has answered the phone. Maria has said “welcome home” in a high tone just before we enter the conversation: Excerpt 6. lige kommet hjem ‘just come home’ (AULing:EMRI:ligekomm), 1990s, Danish, phone, everyday 01
MAR:
02
ANN:
04
ANN:
03
05
MAR:
I: ↑li[ge k]ommet hjem?= you.PL just come.PPT home you’ve ↑just come home?= [(hh-)] =er I ikk be.PRS you.PL not =haven’t you_
(.)
[det_ it.N
[jo:= onsdag; yes2 Wednesday yes:=Wednesday;
06
MAR:
08
ANN:
09
MAR:
=↑pfu:h_= =↑woo:h_=
10
MAR:
=hvodden ↑har I det; how have you.PL it =how ↑are you doing;
07
11 12
ANN:
okayh?
(0.3)
så: øh det er vi £lige ↓kommet£,= so: uh we have £just arrived£,=
(0.3)
vi har det så godt,h we are doing so well,h
In line 1, Maria asks a declarative question. Apparently, she has inferred that Ann has just come home, and now she seeks confirmation of this (‘you’ve ↑just come
190 Søren Sandager Sørensen and Jakob Steensig
home?’). She adds a tag question in line 3, and at the projected completion of this, Ann answers with a type-conforming jo: ‘yes2’ (line 4; Heinemann 2005), and adds the information, ‘Wednesday’. This is receipted by Maria with an OKAY in line 6. This OKAY is in third position and clearly does the job of receipting the answer as informative. In line 8, Ann adds a concluding utterance, summing up that her arrival on ‘Wednesday’ amounts to having ‘just arrived’. This is receipted with an enthusiastic exclamation by Maria in line 9, and a follow-up question about how they (probably Ann and her fellow traveler(s)) are doing (line 10). Ann answers this positively in line 12, and then goes on to assess her trip (not included). The question in line 1 is formatted as a request for confirmation, but we see the question as an expansion-eliciting question (Steensig and Heinemann 2013), which seeks more than just the confirmation and factual information that it initially receives (in line 4). This claim is based on the question requesting confirmation of something that is obvious from the context, as before the beginning of this excerpt, Maria had already extended her welcomes, which were appreciated (not included), and the topic (the homecoming) is clearly expandable. Therefore, the rising OKAY may be seen as displaying that Maria expects more. Ann seems to initially resist doing this, going by the silence after OKAY and the expansion in line 8, ‘so: uh we have £just arrived£’, which only states the obvious, and does not give more information. However, it has a light tone (the smiley voice at the end), which may still signal some sort of openness. Maria’s exclamation in line 9 maintains the “light tone” (this time through a high pitch and lexical choice), and her question in line 10 also has a high-pitched and enthusiastic feeling, which adds to its potential as a “telling question” (Fox and Thompson 2010, 136), which we also see in the way it is understood by Ann, who, in line 12 and in her subsequent talk, expands on the trip, while Maria keeps enquiring into it (not shown). The OKAY in Excerpt 6 is a quite dramatic instance of an okay2. Figure 4 shows its acoustic properties, which conform to what we hear, which is that the second syllable is much higher than the first, it has rising intonation, and it is stressed. In acoustic terms, the second syllable is longer, has a higher pitch (there is an upstep of 3.6 semitones from the end of the first to the beginning of the second syllable) and a clearly rising pitch (the difference from the lowest to the highest pitch is 12.7 semitones, an entire octave). The first syllable lies in the low part of the middle of the speaker’s register, and the second syllable rises to the high part of the register. There is no visible difference in intensity. The two vowels are monophthongs in this case, and the consonant is clearly aspirated. In particular, the wide pitch span and the rise in pitch to a high part of the speaker’s register make the OKAY token read as “enthusiastic,” “engaged,” which is consistent with the stance taken by the OKAY speaker in her assessments in the ensuing talk (in lines 9 and following, and not included).
Chapter 6. Rising OKAY in third position 191
70
300 60
Intensity (dB)
Pitch (Hz)
400
50
155
ε
kh Time (s)
23
Figure 4. Acoustic properties of Maria’s OKAY in Excerpt 6, line 6
In the discussion of Excerpt 6, we argued that although the OKAY receipted the answer as informative, the original question sought more, and we could see an orientation to this in the subsequent behavior of both parties. In Excerpt 7, below, we see a similar orientation, this time displayed most clearly by the behavior of the OKAY speaker. Excerpt 7 is from a phone call between Bo and her mother (MOR). Bo has been talking about problems with a sporting event that she is involved in planning: Some necessary documents have disappeared, and in line 2, Mor asks about what will happen then: Excerpt 7. ser det ikk rigtig ud til vi kan ‘doesn’t really look like we can’ (samtalebank: telefon:bilen), around 2009, Danish, phone, everyday 01 02 03 04
BO:
∙hh så:[:, ·hh so::, MOR: [men hva så, but what then, BO:
(0.7)
jam ib han prøvet å ringe ti’ arvid i >går= well NAME he tried to call NAME >yesterday=
05
=om vi ku< få noget dispensation. =if we could< get some dispensation.
06
(0.9)
07
BO:
08
MOR:
09
me[n ø]::h_ but u::h_ [(m:,)] (m:,) (0.6)
192 Søren Sandager Sørensen and Jakob Steensig
10 11
BO:
det ser det ikk rigtig ud ti’ vi kan, it doesn’t really look like we can, (1.0)
12
MOR: okay,
14
MOR: ellers (så) det jo ikk sikkert otherwise (then) it’s not certain
13
(0.5)
15 16 17 18 19
(at) det blir ti’ noget,= (that) anything will become of it,= BO:
BO: BO:
=nhej_h =nhoh:h (0.2)
°hh° ↑∙hhh
det e:::r øh nærmest katastrofal; it i:::s uh almost catastrophic;
Mor’s question in line 2 enquires into the consequences. It is a “telling question” (Fox and Thompson 2010, 136), in that it invites an expanded account. Bo delivers the beginning of such an account in lines 4–10, ending with a negative assessment of the situation, in line 10. This is receipted by Mor with an OKAY in line 12. Here, Bo provided an account that could be the full, expanded clausal answer to Mor’s question in line 2, but after a half-second silence, in lines 14–15 Mor requests further elaboration by providing a candidate understanding, formulating (Heritage and Watson 1979) what she sees as the possible consequence, that nothing ‘will become of it’. This indicates that at least in retrospect, Mor did not consider the account complete, and we believe that the OKAY already indicated that. The formulation is confirmed, and Bo provides a further negative assessment in lines 16–19. The OKAY in line 12 is a non-dramatic instance of okay2. Figure 5 shows that the two syllables have approximately the same length, the consonant is aspirated, the vowels are (slightly) diphthongized, and the intensity of the second syllable is marginally higher than that of the first syllable. The pitch of the first syllable lies in the lower part of the speaker’s mid-register, and the second is just a little higher, with a shallow rise on the second syllable. The upstep in pitch from the first to the second syllable is small (1.3 semitones), and the difference between the lowest and highest part of the pitch trace is 6.8 semitones. The OKAY in Excerpt 7 may have functioned as sequence closing third. It is only slightly rising, but still sounds unfinished to our ears. This interpretation is supported by the fact that the OKAY speaker asks a follow-up question after the OKAY, and thus expands the sequence. So, at least in retrospect, the OKAY is treated more as a continuer than as a sequence closing third. The next instance, Excerpt 8, is similar to Excerpt 7 in that the OKAY speaker expands a question-answer-receipt sequence. This comes from the same situation
Chapter 6. Rising OKAY in third position 193
400
80
Pitch (Hz)
200
60
Intensity (dB)
300
40
100
kh
εi
339 Time (s)
Figure 5. Acoustic properties of Mor’s OKAY in Excerpt 7, line 12
as Excerpt 5, but from a different recording of the same three people, cooking in a large kitchen. Tine (TIN) has been talking about what her daughter is doing, and in line 1 Tom enquires about her daughter’s future plans. Excerpt 8. på seminariet ‘at the teachers’ college’ (samtalebank:sam3:kartofler_og_broccoli), around 2009, Danish, face-to-face, everyday 01
TOM:
02
hva er det hun regner me’ å: (0.4) ville what is it she counts on: (0.4) wanting på længere sigt, in the long term,
03
TIN:
ja men så vi’ hun jo gerne ø::h (0.5) ind å:: well then she would like to u::h (0.5) get into::
04
TIN:
(0.2) >°altså°< (.) bli’e ↑skolelærer. (0.2) >°PRT°< (.) become a ↑school teacher.
05 06
TIN:
07
TOM:
09
TIN:
10
TOM:
08
11 12
(0.4)
på semi↑nariet. at the ↑teachers’ college. °o°kay,
(0.4)
ø::h= u::h= =så hun ø:h søger ind på seminariet her ti’ =so she u:h is applying to the teachers’ college this >sommer eller [hva,summer or what,
summer or what,> ......€takes cards€ .......§takes card-§ [NÅ_ ] ((WELL_)) [°eller°] °or° (0.1)
mar
MAR: tor
MAR: tor
TOR:
€((TAP SOUND))€ €puts cards on table€
>kun hvis jeg< skal.§= >only if I< have to.= ....................§throws card
=°eller° €må:_ §↓eller°_ =°or am allo:wed to_ ↓or°_ ........€takes cards-->> §sits up, looks tow. SVE-->> nu: >nu der lige en< øh spisepause. no:w now there’s a small meal break. (0.4)
Svend’s yes/no interrogative in line 1 may be understood as an accusation that Margrete may have cheated. This seems to be the way Margrete treats it, as she utters a very loud and insistent ‘n=NO↑::_’ in line 3, adds another assurance in line 4, ‘I certainly hadn’t_’, and then goes on to accounting via a moral statement in line 5, ‘I don’t look at other people’s cards’, from which she then backs down, in line 6 with, ‘(although) I wave my own ((cards))_’. Both the question/accusation in line 1 and Margrete’s defensive answer are said in a light tone, not directly laughing or smiling, but certainly with a sense of humor. This is receipted by Svend first with a very soft (possible) °(mm.)° (line 8), followed by an OKAY in line 10. Svend then disengages from this interaction, bends forward (line 12), and takes some cards. Torben’s NÅ_ ‘well’ in line 13, and his line 19 orient to changing the activity to eating, but Margrete decides to expand her explanation of her card-viewing behavior in lines 14–18, listing the exceptions (required by the game) to the rule that she does not look at other people’s cards. Svend’s OKAY in line 10 does close the question-answer sequence, at least as far as Svend is concerned. Still, it is an okay2, at least the way we hear it. This impression is partially borne out by acoustic evidence. From Figure 8 it is apparent that the first syllable cannot be measured. Still, it is clear that the first syllable (which is audible) is very soft, and that the second syllable has an even pitch, is monophthongized, and lies in the middle of the speaker’s register. So, to the extent that this is an okay2, and it certainly does not have the falling features characteristic of our okay1 tokens, this is a non-dramatic one.
200 Søren Sandager Sørensen and Jakob Steensig
56 54
200
52 50 48 50 0
k
ε
Intensity (dB)
Pitch (Hz)
300
46
Time (s)
Figure 8. Acoustic properties of Svend’s OKAY in Excerpt 10, line 10
Svend’s accusation in line 1 was not a very serious one, and he may understand Margrete’s long defensive turn in lines 3–6 as overdone. In any case, he disengages, first by uttering an indecisive sound (line 8), then with his unenthusiastic OKAY in line 10, before totally disattending what Margrete is doing. On the other hand, Margrete may have heard Svend’s lack of engagement (lines 8–10) as a (mock) display of skepticism, which may explain why she goes on with her accounts in lines 14–18. So, in this case it is not the OKAY speaker who deals with unresolved matters. It is, instead, the OKAY recipient, Margrete, who hears the OKAY as leaving something open, marking skepticism or mocking her, which is, arguably, why she adds to her account after the OKAY. Our last instance of an okay2 used to deal with unresolved matters comes from a service encounter. In Excerpt 11, Karl (KAR) is the customer in a kiosk, and Erling (ERL) is the salesperson. At the point when we enter, Karl has already purchased some goods and paid, and Erling is collecting the change (which he puts on the counter later, during lines 8–12), when Karl notices something below the counter where the magazines are displayed (line 1). Excerpt 11. helligdagen her på fredag ‘holiday this Friday’ (AULing:Kiosk2B20), 2008, Danish, face-to-face, institutional
01
KAR:
02
KAR:
+nå, oh, +looks down at magazines-->
er det nye +billedbladet allerede +udkommen; is the new Billedbladet already out; -->+......................+looks tow Erl-->
Chapter 6. Rising OKAY in third position 201
03 04 05
06 07 08 09
10 11 12 13 14 15
ERL: kar kar
+(1.7) +looks down at magazines-->
kar
(1.9)+ +(1.2)+ -->+ +scratches ear+
KAR:
kar
ERL: kar
ERL:
ERL: erl kar
KAR: ERL:
16
KAR:
17
ERL:
18
‘a:: det på grund af helligdagen her på fredag.+ yeah:: it’s because of the holiday this Friday. -->+
kar
kay,
+(4.3) + +takes up magazine, looks at it+ +sådden der. here you are. +moves magazine downwards--> fyrre kroner. forty kroner.
(1.0)
her_ here_ &+(3.0)& &puts money on counter and looks at Kar& +puts magazine in bag--> hva er ‘t den koster; what is it that it costs; syvogtyve; twenty-seven; (2.3)
å det gør se å hør ka jeg s[e oss.+ and that is the case I can see for Se og Hør too. -->+ [ja_ yes
+(5.3) +takes up another magazine and looks at it-->>
Karl’s nå (‘oh’ in line 1), registers something he has discovered in front of him, and this leads him to ask ‘is the new Billedbladet already out;’ (line 2).10 Erling confirms this, and explains why in line 3. Karl looks at the magazines for 1.7 seconds (line 4) before he receipts Erling’s answer with an OKAY in line 5. After this he first looks a bit more at the magazines, scratches his ear, then picks up a magazine (presumably Billedbladet) and puts it into his bag. All this happens without his paying any attention to Erling putting change on the counter and announcing it (lines 9–12). 10. Billedbladet “Picture Magazine” is the name of the magazine, while Se og Hør “See and Hear” is another magazine.
202 Søren Sandager Sørensen and Jakob Steensig
In line 13, Karl follows his physical actions with a question about the price of the magazine he has put into his bag, gets an answer to this (line 14), and proceeds to check the price of another magazine, and look at, it in lines 16–18. This time, the OKAY speaker’s physical actions show that he does not see the activity as over when he utters his otherwise possibly sequence closing OKAY in line 5. When considering what he is doing, Karl’s question in line 2, apparently requesting confirmation of something that he can already see (that the magazine is out), is a preparatory move to his considering buying the magazine. The OKAY just receipts the answer but keeps the transaction open, in that it occurs during activity that establishes a trajectory to buying the magazine. Only the second syllable of the OKAY is audible. Its measurable acoustic properties may be seen in Figure 9.11 We hear it (and see it) as an okay2 because the second syllable is stronger than the inaudible (and possibly non-existent) first syllable, and because the second syllable has a rise (of 1.9 semitones).
250
80
75 150 70
95 k 1510
Intensity (dB)
Pitch (Hz)
200
ε
Time (s)
Figure 9. Acoustic properties of Karl’s OKAY in Excerpt 11, line 5
Thus, this is a “low key” okay2, which matches the very mundane and casual style of this service encounter generally. But we still see its rising okay2 quality as oriented toward the activity of buying the magazine, that is, here, what is unresolved is that he is not yet ready to move on to the next phase of the transaction.
11. We do not have enough talk by Karl to determine his pitch register, so we made an estimate based on the lines he speaks.
Chapter 6. Rising OKAY in third position 203
5. Rising OKAY tokens as a way of keeping the interaction open Our initial investigations confirmed results from earlier research that the falling okay1 is used in sequence closing third position following answers to questions, to indicate a change-of-state and sufficient understanding, and the rising okay2 is used as a “continuer” or continuation elicitor, to mark that the OKAY speaker sees a projectedly or potentially incomplete telling or account as something that should be continued. What differs from findings of earlier research is the large number of cases of rising okay2 tokens that occur in what seem to be okay1 environments. They follow answers to questions, and seem to treat them as satisfactory, at least in terms of information. The cases of this analyzed in Section 4.4 seemed to function much the same as the “continuers” we exemplified in Excerpts 2 and 5, except that they followed answers to questions. They treated what came before as informative, but were treated as inviting or encouraging the answerer to expand his or her answer. The cases we analyzed in Section 4.5 share the feature that an activity was treated as not complete, as still being in some way unresolved, despite the OKAY having treated a question as informative and sufficient. This was more of a mixed bag, as the unresolved matters range from an “epistemic” battle to an expression of mocking skepticism and to a transaction activity. Our observations are supported by acoustic measurements. However, we have been unable to find any systematic distinctions in the size of the pitch movements or the contributions of other prosodic and phonetic factors. We saw one instance of what we termed a more “dramatic” okay2 (in Excerpt 6), but we can only speculate that this may be part of marking affectivity (see Couper-Kuhlen 2009; Koivisto 2015b for more thorough discussions of prosody and affect). The implications of our findings are that certain prosodic formats of specific particles may carry important interactional functions from one sequential position to other positions. In other words, different prosodic formats may have stable and systematic practices associated with them, as is the case with different lexical and syntactic formats. Some evidence for this has been found in several languages (Benjamin 2013; Selting 1996 on different prosodic patterns of English and German question words and repair initiators; Local 1996 and Couper-Kuhlen 2009 on prosodic and phonetic versions of English oh; Samtalegrammatik.dk 2019a; 2019b on versions of Danish nå ‘oh’), and this points to the need to always include prosody when describing turn design (Drew 2013), especially when they consist of one word or short phrases.
204 Søren Sandager Sørensen and Jakob Steensig
We make no claim that these prosodic formats go beyond the specific particles, for instance, that there is such a thing as “sequence closing” or “continuer” intonation, but it is possible that some prosodic formats manifest across different particles that may occur in the same positions.12
Acknowledgements Trine Heinemann connected us to the OKAY project and helped us get started, Birte Asmuß participated at the beginning of the Danish project and provided data for our collection. All the contributors to the OKAY project made useful comments during workshops, the editors of the book helped and supported us all the way, and the reviewers, Elizabeth Couper-Kuhlen and Arnulf Deppermann, made very precise and insightful comments to two earlier versions. Michaela Scioscia corrected our English. We are very grateful for this, and we also wish to thank data-session participants, who discussed excerpts and helped us understand them better, and especially, all the people who agreed to have their everyday lives documented by the recordings that form the basis of our research. All shortcomings are our responsibility. The authors contributed equally to this chapter.
12. To the best of our knowledge, the proposal that there may be similar pitch patterns for different Danish particles was first made by Caroline Grønkjær and Christel Tarber during an informal discussion that followed a presentation by Trine Heinemann of what later became the main points in Heinemann (2017a).
Chapter 7
OKAY as a response to informings in Finnish Aino Koivisto and Marja-Leena Sorjonen University of Helsinki
This chapter examines the use of the particle OKAY in Finnish as a response to turns that are designed to provide new information. The study focuses on the ways in which the speakers of OKAY orient toward the epistemic character of the informing, and on its sufficiency for the purposes of the ongoing interaction. These aspects are analyzed by paying attention to the sequential and activity context of the particle, its phonetic-prosodic design, and the possible subsequent talk in the turn. OKAY is also compared with some other Finnish response particles in similar sequential and activity contexts. The data come from telephone and face-to-face interactions between family and friends spanning a period from the late 1980s to the present. Keywords: change over time, change-of-state, comparison of response particles, doubt, Finnish, instrumental information, newsworthiness, prosody, sufficient information, understanding
1. Introduction The most well-known uses of the particle OKAY in different languages are for achieving transitions and closings, and for providing or projecting acceptance of the nominated action of a directive (cf. Schegloff and Sacks 1973; Beach 1993; Barske 2006, Chapter 2; 2009; Schegloff 2007, 120–123; Seuren 2018; Chapter 4, this volume; Keevallik and Weidner, this volume). These are also its most mentioned uses in Finnish so far (e.g., Leskelä 1999; Malaska 2017; Pekkanen 2017; see also the references in footnote 1),1 and represent our initial understanding of the core uses of okei. 1. The use of okei in Finnish dates back to the 1930s. The dictionary of slang (Paunonen and Paunonen 2000) contains entries for different forms of OKAY and provides information, with examples of the time when the form can be attested. The use of okei as a response to a directive can also be traced back to the 1930s (s.v. okei; also the variants ogei, okey and okay are mentioned), ok to the 1950s, okeido to the 1950–1960s, okra to the 1990s, and okka to the 2000s. The longest https://doi.org/10.1075/slsi.34.07koi © 2021 John Benjamins Publishing Company
206 Aino Koivisto and Marja-Leena Sorjonen
Quite recently, however, we encountered usages that were “new” to us, as well as a noticeable increase in the frequency of okei in the data from the 2010s onward (Koivisto 2016, fn. 4). In these “new” usages, okei responds to a piece of telling, or to an answer by the questioner – to an informing of sorts (see also Pekkanen 2017 for an analysis of okei in a present-day dataset, including its use as a display of change-of-state). Okei appears to function similarly to change-of-state tokens, proposing “that its producer has undergone some kind of change in his or her locally current state of knowledge, information, orientation or awareness” (Heritage 1984a, 299). Its work thus seems to resemble the use of response particles that are conventional means for showing such displays (e.g., ai jaa or aha; articles in Heinemann and Koivisto 2016b on different languages). In this article we explore the use of okei as a response to different kinds of informings in a range of sequential contexts. In investigating occurrences of okay as a response to informings, we use the term “informing” as defined by Thompson, Fox, and Couper-Kuhlen (2015, 51). Thompson et al. use the term for actions that are designed “to provide information to a non-knowing recipient such that they become (more) knowing” (ibid.). “Informing” for them is a cover term for presenting news, announcements, informings and reportings. Sequentially, information may be provided by the speaker on their own initiative (volunteered informing in first position), or it may be requested by the interlocutor with a question (question-elicited informing in the second position) (ibid.). Accordingly, a response to an informing may be produced in the second or third position. To acknowledge and make public their orientation toward the transfer of information, the recipient may produce a response that displays that they are now informed, and in so doing indicate that they did not have that information before (Heritage 1984a). In contrast, claims of already knowing the piece of information would make the sequence an unsuccessful informing sequence (ibid., 303–304). For instance, a preferred response to a piece of news would be a response that entry okei also mentions the use of the word as a tag, as a predicate nominal (miten kundit? onks ne tiutau vai onks ne okei? ‘How (are) the guys? are they tiutau or are they okay?’), and as an adverbial (soitettiin se keikka, se meni ihan okei ‘We played the gig, it went okay.’). In translations from English into Finnish, okei is traced back as far as a translation by Alex Matson of John Steinbeck’s 1930 novel The Grapes of Wrath in 1944 (Pulkkinen 1984 s.v. O.K.). The wikibased Urban dictionary (Urbaani Sanakirja), which was started in late 2007, contains eleven entries of OKAY, eight of them approximately from the first year of the dictionary, and many of them the same as in the slang dictionary. Okei has also been accepted in the main dictionary of written language (current continuously updated Kielitoimiston sanakirja 2012 (s.v. okei; pronounced as okei or oke´i).
Chapter 7. OKAY as a response to informings in Finnish 207
acknowledges the prior turn’s newsworthiness, which enables more talk on the topic (Maynard 1997, 104). Besides actual news tellings, the category of informings also includes deliveries of new information that are not “news” to be talked about as such but something that is relevant for the ongoing interaction and that can be understood and appreciated from that perspective. Examples of the latter type are, for instance, responses to simple information-seeking questions (see e.g., Thompson et al. 2015, 16–49), such as inquiries into the current whereabouts of the recipient at the beginning of a mobile phone call, or an inquiry about the time (see examples below). For cases like this, it may be relevant to receive the piece of information as understood and sufficient for the current purposes instead of foregrounding its newsworthiness and starting to elaborate on it. The key point in both cases is that the information is transferred from a knowing participant to an unknowing participant (Heritage 1984a, 2012b). Languages have conventionalized practices for displaying that the recipient has been informed and undergone a change in their epistemic state (see Heinemann and Koivisto 2016a). These include particle responses as well as phrasal and minimal or extended clausal responses. In English, for example, the central resources are the particle oh, and the minimal clausal responses such as you did and did you (see Heritage 1984a; Thompson et al. 2015, 53). Finnish, on the other hand, has several particles that are specialized in treating the prior talk as having provided new information (e.g., aijaa, aha(a), jaa, vai nii, ja(a)ha). While the exact functions of most of the particles are yet to be investigated, two common particles have been studied. Aijaa is the most frequent particle of this group; it treats the prior turn as having provided new and newsworthy information, which means that the prior turn may be topicalized. It is thus not in itself closing-implicative like the English oh. (Koivisto 2015a, 2016; see also Kastari 2006.) Another frequent particle aha(a), on the other hand, is used to indicate that the prior turn involved a redirection from the projected course of action and was thus unanticipated (Koivisto 2016). Other response forms, depending mostly on their prosodic shape, may also be used as a response to turns that offer new and even newsworthy information. These include partial repetitions of the prior turn (+ question particle, e.g., on-ko (be-PRT) or on vai (be PRT)) and other particles such as joojoo (a reduplication of the central acknowledging particle joo, Kunnari 2011). Another particle that does not “look like” a news receipt but can be used as such, is okei. In English, the terminology of response types conventionalized for displaying an epistemic change of state varies. For example, Heritage (1984a, 307–309) deploys the term information receipt to describe the use of the particle oh to display that the preceding talk to which it responds has been informative to the oh speaker.
208 Aino Koivisto and Marja-Leena Sorjonen
Oh is differentiated from response types such as yeah and mm hm, which do not treat the prior talk as informative. Furthermore, Heritage (ibid. fn. 13, by reference to Jefferson 1981a, 62–66), differentiates information receipts from news receipts/ newsmarks, forms that treat the prior talk as news and solicit more talk on it (e.g., did you; oh did you; oh really). However, Thompson et al. abandon the distinction between news receipts and newsmarks. They show that prosodic formats play a central role in distinguishing the work that the different syntactic forms do, for example, in terms of emotive involvement and whether they project sequence expansion (Thompson et al. 2015, 113–114). Thompson et al. provide a more detailed discussion on the terminology (ibid., 52–53). In this chapter, we will develop the description of okay as a response to informings through analyzing its use in its sequential and activity contexts together with its prosodic design, instead of treating it a priori as a member of any of the categories mentioned above. We will show that the use of okay as a response to an informing ranges from acknowledging (accepting) the relevance and sufficiency of the informing for the ongoing activity, to orienting to the prior turn as having provided something newsworthy or of (special) interest. In the latter cases, okei is produced with a marked, salient prosody. 2. Data We collected the data for the study from different telephone and face-to-face conversations held at different points of time. The data come from the Conversational Data Archive on Finnish at the University of Helsinki. They include telephone conversations between friends and family, recorded over four decades (from 1980 onward), and videotaped face-to-face interactions between friends and family. We went through sets of corpora from different points of time to obtain an initial understanding of the frequency of okei between the late 1980s and the 2010s. After this, we counted all the cases that could be considered informings. The striking observation was that while the use of okei has become more frequent in general, its uses as receipts of informings had also become commonplace by 2010. We also analyze the use of other response particles in the contexts in which we examine the use of okay, in order to sharpen the analysis of okay. The analyses of these other particles rest on existing prior research, mainly our own.
Chapter 7. OKAY as a response to informings in Finnish 209
Table 1. Number of okeis in data in different decades. Year = year of recording(s), Age = age(s) of speakers Year
Age
Total length
Number of okeis
Number of okeis in informing sequences
1988–1989
young adults
90 min
25
0
1997
18–20
115 min
47
6
telephone, face-to-face
2007–2008
17–30, 50
109 min
73
28
face-to-face, dyadic and multiparty Radio interview
2011, 2015
20–30
476 min
230
161
2017
~50
13 min
3 378
3 201
Decade Medium
Late 1980s
telephone
Late 1990s
telephone
Late 2000s 2010s
Total
3. Contrasting the use of okei to change-of-state tokens proper An essential question pertaining to the use of okei in response to an informing is its relation to response types that are specialized in expressing that the prior turn provided new information. In essence, these particles can be characterized as change-of-state tokens, proposing that “its producer has undergone some kind of change in his or her locally current state of knowledge, information, orientation or awareness” (Heritage 1984a, 299).2 Consider the following two excerpts from mobile phone call openings. In the first excerpt, the change-of-state token aijaa functions as a third position turn in an informing sequence, and the second excerpt contains an instance of okei in a similar environment.3 2. The basic findings on Finnish change-of-state tokens presented here are mainly based on previous work (Koivisto 2015a, 2016; Kastari 2006). However, we will illustrate their use with examples from the current database. 3. It should be noted that Excerpt 1 is taken from a series of phone calls recorded in the late 1990s, which included no instances of okei as a receipt of new information (its use was restricted to closing and transition contexts). Excerpt 2, on the other hand, was recorded twenty years later, in 2008. At that time, okei was used in a wider range of sequential contexts, also as a response to informings.
210 Aino Koivisto and Marja-Leena Sorjonen
Excerpt 1. Lahdessa ‘in Lahti’ (KTA2 Sg112 A01), 1997, Finnish, mobile phone call, informal conversation 01
VIK:
halojaa_
03
VIK:
lahde-ssa. NAME.CITY-INE in Lahti
04
MIS:=> ai jaa. mi-tä-s sie-llä.h what-PAR-CLI DEM3.LOC-ADE what’s ((happening)) there
05
VIK:
02
MIS:
no moi:_ hh mi-ssä sä oo-t. Hh PRT PRT where-INE you.SG be-2SG hi hh where are you
ol-la-an salla-lle ol-tu osta-ma-s >kaikke-e 06
=>
öö miko-lla? hh 1nameM-ADE er at Mikko’s hh (0.6)
okei mä oo-n just .hhh tii:na-n kanssa tä-ssä? hh .hh PRT be-1SG PRT 1nameF-GEN with DEM1-INE I’m here right now .hhh with Tiina hh .hh
kävele-mä-ssä kohti ↑kir:jasto-a; walk-INF-INE toward library-PAR walking toward the library
In both cases, the sequence-initial question is part of the opening of the phone call: the caller inquires about the location of the recipient, which is a customary part of mobile phone call openings (see e.g., Laurier 2001; Weilenmann 2003; Arminen 2006). In both cases, the recipient also provides a mere phrasal answer to indicate their location, and in so doing transfers the turn back to the questioner (Vepsäläinen 2019, 84–91). In Excerpt 1, the answer is received with aijaa (one of the equivalents of the English oh, see Koivisto 2015a, 2016), produced as a prosodic unit of its own with a final falling contour. In Excerpt 2 (line 5), the response is okei, followed by a reciprocal informing without a prosodic break. We can safely say that in both cases the previous turn provided new information as an answer to an information-seeking question, which was then received with aijaa/okei. What is different, though, is that
Chapter 7. OKAY as a response to informings in Finnish 211
in Excerpt 1, the answer merely provides the name of the town in which the speaker is, whereas in Excerpt 2, the answer providing the location (‘at Mikko’s’, i.e., visiting somebody) implies that the speaker may be unable to engage in a longer conversation. This difference is reflected in the ways in which the questioners treat the answers. In Excerpt 1, the questioner produces a follow-up question (‘what’s happening there’) after the particle aijaa, thus treating the piece of information as topicalizable. In general, ai jaa differs from the use of English oh in that it is not closing-relevant (Koivisto 2015a). In Excerpt 2, okei is followed by a reciprocal informing of the current whereabouts of the questioner, verifying the fact that the prior description was received as understood and sufficient, and needed no elaboration. Another piece of evidence of aijaa and okei operating differently when responding to a piece of new information is that they can occur together, in the same turn. Consider the following Excerpt 3, in which aijaa is followed by okei twice within the same sequence (lines 9 and 12). This is an excerpt from a face-to-face conversation in which Iina is visiting her friend Susa. Later, Susa’s sister Maija arrives at Susa’s from her boyfriend’s (Tomi), but apparently earlier than Iina had expected. Excerpt 3. Nopeesti ‘quickly’ (Repair Aineisto1_30_40), 2011, Finnish, face-to-face, informal conversation 01 02
IIN:
03
MAI:
04
IIN:
05 06
MAI:
↑mi-st sä tul-i-t siis. pa:sila-sta vai. where-ELA you.SG come-PST-2SG PRT NAME.SUBURB-ELA or where did you come ((here)) from then. Pasila or (.) .mthh #joo#. .tch yeah
>mite sä tul-i-t< näi nopeesti. how you.SG come-PST-2SG DEM1.MAN quickly how did you get here so quickly (0.6)
↑bu:ssi #tulee suoraan# °toho°. ((the)) bus comes straight there
07 IIN:-> ↑ai↓jaa. 08
(0.4)
11
(.)
09 IIN:=> [°okei°. ] 10 SUS: [yhek↑sän]↓viis menee siit tomin edestä ja; ((the)) nine five leaves from Tomi’s place and 12 IIN:=> aija::a. o[kei.] 13 SUS: 14
15 SUS:
[ tul]ee toho_ comes there
(1.0) ((SUS makes a pointing gesture)) toho mannerheimintielle, (.) there to Mannerheim Street
212 Aino Koivisto and Marja-Leena Sorjonen
Iina’s question in line 1 is a simple information-seeking question followed by a candidate answer (‘from Pasila or’). After receiving a confirming response Iina, instead of receiving the answer as understood, sufficient or new information, asks for an explanation for Susa’s quick arrival, thus revealing the motivation behind her initial question (line 4). As a response, Maija provides a characterization of the bus route from the point of view of Iina’s question (‘comes straight’). Iina responds with aijaa in line 7, produced with a movement from slight rise in pitch at the beginning of the response relative to Iina’s prior talk (marked with ↑) to a falling final contour. With aijaa Iina treats the answer as a piece of new and newsworthy information (Koivisto 2015a), which is emphasized by the wide prosodic movement in the particle. In other words, by saying aijaa, Iina claims that she did not previously know of the existence of such a bus connection, thereby also confirming her status as a now-informed party (Heritage 1984a, 310). By not continuing she may be understood as making an elaboration of the information by Maija relevant. That not coming, she produces okei, uttered sotto voce and with a falling final contour (line 9). With okei she displays acceptance of the relevance of the information and its sufficiency for explaining Maija’s unexpectedly quick arrival. This is followed by an additional, more detailed piece of information (and thus explanation) by Maija’s sister Susa, to which Iina again responds with aijaa, with a narrow range modulation and a final fall, followed by okei as a separate prosodic unit (line 12). Excerpt 3 shows that the two particles can occur in combination, as a composite sequence-closing third (Schegloff 2007, 127). The order of the particles is always aijaa (or some other change-of-state token), followed by okei (cf. Schegloff 2007 on oh okay; see also Helmer, Betz, and Depperman, this volume). In this case, we can say that aijaa is used for acknowledging the newsworthiness of the answer (epistemic orientation), and okei targets its success in providing a satisfactory explanation, thereby proposing a sequence closure. Thus, in this context, okei is used as a device for indicating action-acceptance, especially from the point of view of its sufficiency (similarly to English okay, Schegloff 2007, 128). This section has shown some basic similarities and differences between the Finnish “change-of-state token proper” aijaa and okei. We now take a closer look at the properties of okei as the receipt of an informing. We first discuss cases in which okei receives informings as relevant and sufficient for the current purposes (Section 4). We then move on to cases in the most recent data in which okei also has usages that come close to the use of aijaa. In these cases, it provides the first response to the informing, marking it as newsworthy or of special interest (Section 5). We suggest that in these cases, the characteristics of okei implying the sufficiency of the informing merge into and exploit the characteristics of the change-of-state tokens. This means that okei can treat the informing as having provided something newsworthy or of special interest but also as a relevant and sufficient whole from
Chapter 7. OKAY as a response to informings in Finnish 213
the perspective of the ongoing activity. It turns out that prosody plays a central role in displaying which properties of the okei response are more emphasized, in addition to the sequential placement of the informing. Finally, we briefly consider okei produced with stylized prosody, which seems to add an affective coloring of doubt or perplexity to the response, whereas the meaning of indicating sufficiency is present to a lesser extent (Section 6). 4. Receiving information as understood and sufficient for current purposes In this section we analyze the instances of the okei response that treat the prior informing as understood and sufficient for the current purposes, thus orienting to the instrumental nature of the informing in the context of a larger ongoing activity. In so doing, okei closes or suggests closure of the local sequence. These informings include, for example, sequences of talk that deal with checking and setting the conditions for the interaction or the activity proper to start. In these cases, okei is produced as prosodically unmarked. We start with the mobile phone call from Excerpt 2. We consider a larger excerpt that contains two instances of okei as a response to an informing, both volunteered and question-elicited. Excerpt 4. Testipuhelu ‘test call’ (KTA3 Sg404 testipuhelu), 2008, Finnish, mobile phone call, informal conversation; analyzed partially in Excerpt 2 01
MIR:
moro? tää on testi-puhelu siit nauhotta-va-sta? PRT DEM1 be-3SG test-call DEM3.ELA record-PTCP-ELA hi this is a test call from the recording ((machine))
02 TAR:=> £aha ok(h)ei oo-t-sä mi-s↑sä£ PRT be-2SG-you.SG where-INE where are you 03 04
MIR:
05 TAR:=> 06
öö miko-lla? hh 1nameM-ADE er at Mikko’s hh (0.6)
okei mä oo-n just .hhh tii:na-n kanssa tä-ssä? hh .hh I be-1SG PRT 1nameF-GEN with DEM1-INE I’m here right now .hhh with Tiina hh .hh
=> kävele-mä-ssä kohti ↑kir:jasto-a; walk-INF-INE toward library-PAR walking toward ((the)) library
In line 2, okei, following the change-of-state token aha, is a response to a volunteered informing regarding the fact that the phone call will be recorded, and in line 5, okei initiates a turn with which the speaker responds to an answer that she received to the question she asked her co-participant. In both cases, the speaker
214 Aino Koivisto and Marja-Leena Sorjonen
moves from okei to the following turn-component without a prosodic break: in line 2, to a request for information (‘where are you’) and in line 5, to a reciprocated informing. This is an indication that the speaker treats the information provided by the co-participant as understood and sufficient for the current purposes, closing the local sequence. Okei in line 2 also functions as an indication of its speaker’s commitment to talk despite the call being recorded (these usages are outside the scope of our analysis); producing the response aha with a smile, and the laugh token in ok(h)ei may index the delicacy of the situation. The following turn-components in this turn in line 2 begin a new adjacency pair, which makes it clear that the okei speaker expects no elaboration from the informer.4Instead, okei is used to acknowledge the current status and whereabouts of the recipient before the call can move on to other matters. Another type of example of a minimal question-answer-okei sequence is provided below. The excerpt is taken from a situation in which a group of students is preparing food for a get-together of a bigger group. In this excerpt, okei receives an answer to a question about time during the preparation of the meal. Okei is used to receive an answer to a question that Irja, who is baking bread, has just asked. Excerpt 5. Kello ‘clock’ (Pekkanen), 2015, face-to-face, conversation in student union kitchen
01 02 03
IRJ:
(2.3)
>karri mi-tä< o(h). 1nameM what-PAR clock be.3SG Karri what time is it (1.4)
04
KAR:
05
IRJ:=> okei. hyvä? good
06
HIL:
07 08
kello o-n tasan kuusi. clock be-3SG exactly six. it’s exactly six o’clock
°nonni,° (0.4) paitsi et mi↑tenkäs alright (0.4) except how did meiän uuni-po#litiik#ka nyt meni. our oven-politics go
IRJ:
en tii£jä(hh)£. I don’t know
4. It is noteworthy that the precise position of okei in its turn and hence its temporal relation to the prior informing by the co-participant is different in these two cases. Whereas the latter okei (line 5) occupies the initial position in the turn, the okei in line 2 is preceded by the change-ofstate token aha, which implies that the information received requires a reorientation from the aha speaker (Koivisto 2016).
Chapter 7. OKAY as a response to informings in Finnish 215
With her okei (line 5), produced as a prosodic unit of its own with a falling final contour, Irja displays understanding and the sufficiency of the information Karri provided. She then produces an assessment (hyvä ‘good’) in a separate prosodic unit. This assessment implies the relevance of her question and the relevance of its answer for her ongoing activity of baking – there is no time pressure to prepare the bread. Again, then, while functioning as a receipt of new (here: question-elicited) information, okei invites neither elaboration of the prior turn nor the topic initiated by the question but rather closes the sequence. As in Excerpt 4, the function of okei is related to establishing conditions for the ongoing larger activity (see also Pekkanen 2017). Okei can also be part of a conventional pre-sequence, that is, function as the receipt of an answer to a question that is designed to project another action. In the excerpt below, a child (TAN) is making a mobile phone call to his mother (MOM). Excerpt 6. Ääni ‘sound’ (KTA3 Sg400_07), 2008, Finnish, mobile phone call, informal conversation
01 02
MOM:
03
TAN:
04
MOM:
05 06
TAN:
07
MOM:
08 09
TAN:
10
MOM:
11 12 13 14 15
TAN:
MOM:
TAN:
>mer↑ja< 1nameF (1.0)
moi mamma, hi mommy @no moi@, PRT hi (0.8)
mi-tä kuul-u-u? what-PAR hear-REFL-3SG how are you? @no hyvä-ä:@_h PRT good-PAR PRT fine (0.8)
mm: o-n-k-s [su-l huome-n ilta PRT be-3SG-Q-CLI you.SG-ADE tomorrow-ESS evening mm do you have an evening [(mi °joo°. no miltä se näyttää_ well how does it look
218 Aino Koivisto and Marja-Leena Sorjonen
Excerpt 8. Smirgeli ‘smirgeli’ (KTA2 Sg094_06), 1991, Finnish, landline call, informal conversation
01
PEK:
hh ö Pekka Virtanen_hh= 1nameM surname Pekka Virtanen
02
MAT:
=no joo: Matti Makkonen morjens_= PRT yes 1nameM surname hi yes Matti Makkonen hi
03
PEK:
04
=no moi moi.= PRT hi hi
MAT:-> =ö ryntäs-i-t sä pi:tkä-ltä.hh= rush-PST-2SG you.SG long-ABL =er did you ((have to)) rush from far away=
05
PEK:
06 07
MAT:=> =jo[o joo:. PEK: [.nhh
08
MAT:
=#ei:: tuo-lt yläkerra-st tänne alas vaan#_hh= NEG.3SG DEM2-ABL upstairs-ELA DEM1.LOC.ALL down just =no just from upstairs down here
tota: mä tein ne (.) kii:lat nyt:te just [äskön.] well I made the (.) wedges just a moment ago
Excerpt 9. Tanssimaan ‘dancing’ (KTA2 Sg056 A02), 1991, Finnish, landline call, informal conversation 01
KAI:
leppäne? surname Leppäne?
02
SIR:
°no sirpa moi.° PRT 1nameF hi Sirpa moi
03
KAI:
no ↑moi PRT hi
moi. hi
04 SIR:-> e-t-hän nukku-nu. NEG-2SG-CLI sleep-PPT you weren’t sleeping were you 05
KAI:
↑e:-n nukku-nu (.) mä (.) vaa ol-i-n tuo-l NEG-1SG sleep-PPT I just be-PST-1SG DEM2-ADE ((I)) didn’t (.) I (.) just was there
06
toise-s huonee-s .mu .men-i .vähä other-INE room-INE I.ADE go-PST.3SG little.ADV in the other room it took me
07
.aika-a .ku .mä dyykkas-i-n sie-ltä time-PAR when I dive(SLANG)-PST-1SG DEM3.LOC-ABL a moment to jump here from
08
#tännehh# m hhh there
09 SIR:=> just.=no mä aattel-i-n siitä perjantai-sta= PRT I think-PST-1SG DEM3.ELA Friday-ELA =well I was thinking about the Friday=
Chapter 7. OKAY as a response to informings in Finnish 219
10
KAI:
=.hh joohh (0.5) mitä_ ootsä soitellu ihmi[silleh =.hh yeah (0.5) what_ have you been calling people
In the segments above, the caller poses an inquiry about the recipient’s current situation, displaying an orientation toward a delay in responding to the call (in Excerpt 7 line 5, in Excerpt 8 line 4, and in Excerpt 9 line 4). After receiving the answer with the response particle, the callers proceed to the reason for their call (lines 14, 8 and 9). This demarcation between sequences and separate activities is enforced by producing the response particle with a falling final contour as a prosodic unit of its own. These particles are also used in the more recent data. In the following excerpt, joojoo is used in a similar sequential environment in a telephone call recorded in 2007. In the same set of data we also have a case of okei in a similar kind of sequential context (Excerpt 11). Excerpt 10. Lasku ‘bill’ (KTA3 Sg399l), 2007, Finnish, mobile phone call, informal conversation 01
RAI:
02
TEE:
03
RAI:
04
TEE:
05
.hh noni moikka_ alright hi no morjesta hh_ PRT hi
soita-n-ko paha-an aika-aihan to-s joojoo.hh eiku hh mä vahingo-ssa ava-si-n ku PRT I accident-INE open-PST-1SG PRT no hh I accidentally opened ((‘it’)) 07
tänne tul-i äsgee-n lasku_ DEM1.LOC.ILL come-PST.3SG NAME-GEN bill a bill from ÄsG that came here
Excerpt 11. Kotona ‘home’ (KTA3 Sg399i), 2007, Finnish, mobile phone call, informal conversation
01
TEE:
moi_ hi
02
MIR:
03
TEE:
n_moi oo-t-sä (.) nyt koto-na-s vai_ PRT hi be-2SG-you.SG now home-ESS-POSS1 or PRT hi are you (.) at home now or joo:_ oo-n_hh just tul-i-n_ PRT be-1SG PRT come-PST-1SG yes I am hh ((‘I’)) just got here
220 Aino Koivisto and Marja-Leena Sorjonen
04 MIR:=> ok#ei:#; (.) joo no e-m-mä< (.) viel tiädä PRT PRT NEG-1SG-I yet know (.) yeah well I don’t (.) yet know 05 06
siis että (.) tota ei me nyt viel olla PRT PRT PRT NEG.3SG we now yet be PRT whether (.) um we aren’t at least now ainakaa tulo-ssa [mutta hh] at.least coming-INE but coming but hh
Compared to the use of okei in similar sequential and activity contexts, the particles joo, joojoo and just suggest that the prior topic has been dealt with. In contrast, okei indicates that the conversation can move on to projected next matters. We suggest that, in comparison to okei, joo, joojoo and just provide a slightly different kind of response, and each of them a response type of their own. Joo (Excerpt 7), which can be described as a general acknowledgement token,5 provides a weaker display of the sufficiency and possible closing relevance of the sequence, topic and activity than okei (on joo, see Sorjonen 2001). The reduplication joojoo (Excerpts 8 and 10), on the other hand, can take a longer segment of talk in its scope, receiving it as understood and sufficient, and it may imply a suggestion to the co-participant to finish their activity (see Kunnari 2011). It can thus be used as a stronger closingand transition-implying device than the simple joo. The third particle just implies, as suggested by Waris (2006), stronger understanding of the prior talk than joo, and in so doing, orients slightly more toward the prior turn as having provided new information.6 Okei as a response that treats the informing as understood and sufficient, and further talk on the issue as unnecessary for the ongoing activity, thus resembles the usages of some other particles. However, while there are clear similarities, differences also emerge: in the examples above, joo, joojoo and just are followed by an initiation of a new sequence, topic and activity. Thus the particles close the previous trajectory of talk. In contrast, okei – while also treating the prior talk as sufficient for the current purposes and thus closing the local sequence – also orients toward the informing being relevant from the perspective of a larger, still ongoing activity. A more comprehensive understanding of the use of the different resources relative to each other requires further extensive study.
5. Joo also has uses other than an acknowledgment token: it is used as a confirming response to polar questions, and as a claim of compliance to imperative formatted directives (see Sorjonen 2001). 6. Just is also used as a temporal adverb and a focus particle (see Vilkuna 1993).
Chapter 7. OKAY as a response to informings in Finnish 221
5. Receiving information as newsworthy: okei produced with marked prosody In the preceding section we saw instances of okei as the first response to informing, treating it as understood and sufficient for the current purposes. In all the cases, the okei speaker proceeded after okei to further talk that oriented toward the informing as being sufficient and not relevant to elaboration. In these cases, okei was produced as prosodically unmarked, i.e., flat and with a final fall (cf. Couper-Kuhlen, this volume). We now turn to the cases in which okei treats the prior informing as more than just understood and sufficient, orienting toward the epistemic salience of the informing. This means that the information provided is treated as newsworthy or of special interest. These cases cluster in sequences of telling. In this context, the use of okei thus comes close to that of the dedicated change-of-state tokens (Heritage 1984a). In the following, we present two cases produced with marked prosody. We will start with an example in which the use of okei as treating a piece of information as epistemically salient emerges very clearly because of its marked prosody. The segment is from a radio interview in which the interviewee is the famous rock musician, Ismo Alanko. Previously (data not shown) the interviewer (INT) has asked Alanko (ALA) whether he thinks about the genre of songs when starting to work on new material. When the excerpt begins, Alanko is describing his long-standing dream of making a “beautiful peaceful album”. Excerpt 12. Haave ‘dream’ (Radio Ykkösvieras Ismo Alanko), 2017, Finnish, radio interview 01
ALA:
mul on ollu monta- pitkän aikaa semmonen haave et I’ve had many- for a long time a dream that
02
mä haluisin j:oskus >sellasen< I would like at some point ((‘to make’)) this kind of
03
kau:niin rauhallisen levyn. .nff ja se on aina beautiful peaceful album. .nff and it’s always
04
jääny ja sit on tullu joku muu idea ((just)) got left and then some other idea has arisen
05
joka onki toteutet[tu sitte siinä välissä.= that gets carried out in between.=
06
INT:
08
INT:=>
10
INT:=>
07
09
ALA:
[mm;
=↑nyt mä oon tekemässä sitä [mikä pitä-s =↑now I’m doing it it’s supposed ((to come out)) [.t o↑ke:i.
ens syksynä sit[te et se on nyt; .hhh niinku next fall then so it’s now .hhh like [okei.
222 Aino Koivisto and Marja-Leena Sorjonen
11 12
ALA:
13
ALA:
15
ALA:
16
INT:
14
21 22
[.nhhh
[joo. [kyllä. PRT PRT yeah. it is. [no ni. kaunis alright. a beautiful rauhallinen le[vy. nyt ku enpeaceful album now when I don’t-
ALA:
19 20
(0.2)
INT:-> [ai se on jo ↑niin pitkäl[lä. oh it’s already ↑that far along.
17 18
miksauksessa menossa. being mixed.
INT: ALA:
[niin. mut on siin- (tai) sanotaan niin et yes. but there’sor let’s say that siinä pyritään [k(h)auneuteen mutta; .hh mä en pysty it strives for beauty but .hh I can’t [hh heh heh välttämään sitä että avoid the fact that
maailman; .hhhh nf jyly kythe rumble of the world
kuitenkin kuuluu [sieltä niinku että ((can)) still be heard in it like that
In the course of Alanko’s extended turn, the interviewer first merely produced continuers such as mm at syntactic and prosodic boundaries. Another mm is produced in line 6. In line 7, after having described his dream on a general level, Alanko begins a new TCU. It starts with the temporal adverb nyt ‘now’, produced prosodically with a high onset and stress, which projects the beginning of something new. That is, the temporal adverb is anchored to the current moment, contrasting with what he has talked about previously. Alanko announces that he is now in the process of making the album that he has been dreaming of. Right after the possible last element of the first clause, the anaphoric pronoun (sitä ‘it’, ‘now I’m doing it’), the interviewer responds in overlap with a prosodically marked okei response (line 8). It is produced with a step up, stress and vowel lengthening in the second syllable. Okei with this type of prosodic design in this particular point of talk is implemented for highlighting the immediately-prior informing and marking it as worthy of special interest and attention (cf. Heritage 1984a, 305; see also Thompson et al. 2015, 55–56). In doing so it also aligns with the designed-as-newsworthy characteristics of the just-prior announcement. The first okei is followed by another, prosodically less marked instance of okei, receiving the second part of the informing (‘it is supposed ((to come out)) this fall’). Another, more elaborated and explicit type of news response occurs in line 14 after Alanko reports that the album is already being mixed. The
Chapter 7. OKAY as a response to informings in Finnish 223
interviewer produces a paraphrase-type candidate understanding (‘oh it’s already that far along’), targeting a specific element of the prior talk and retrieving it as news (see Thompson et al. 2015, 128–134). After the candidate understanding has been confirmed, the interviewer produces a summary assessment (e.g., Jefferson 1984b, 211), now indicating the sufficiency of the information provided about the present state of the production of the album (lines 16–17). The interviewee then moves gradually to talk about whether his music can indeed be considered ‘beautiful’ (from line 18 onward). However, even if okei responding to an informing may be produced with a marked prosody, which highlights the newsworthiness of the informing, it does not need to make an elaboration by the informer relevant. This is the case in the following excerpt from the kitchen of a student union. The excerpt begins when a newcomer to the student union, Tanja, enters the kitchen. Excerpt 13. Multia-edustus ‘Multia-representation’ (Pekkanen), 2015, Finnish, face-toface, conversation in student union kitchen 01
ANN:
02
HAN: 1
03
TAN: 2 öö farmasia-a alot-i-n nytte< pharmacy-PAR start-PST-1SG now er I just started ((to study)) pharmacy
04
terve[tuloo? ] welcome?
[mitä sä] opiske↑let_ what are you studying
HAN:=> o[↑kei, ]
05
TAN:
06
HAN:-> niij justii.= that’s right.
07
ANN:
08
HAN: 1 mistä sä oot koto:sin_ where are you from,
09
TAN: 2 Multialta? Multia?
10
HAN:=> mt o↑°kei:°. mul[tia-edustusta.] loistavaa? Multia representation. splendid? TAN: [joo. kyllähh. ] yeah. that’s right.
11
[ihan_ ] (.) ihan tänä syksynä että. just. (.) just this fall so.
=kiva. nice.
12
TAN:
13
HAN:-> niij justii. mites sä [oot right. how do you (like it ANN: [NOO well
14
vähän pienemmäst paikasta. from a bit of a smaller place. (viihtyny).] here) HAnna ] on Hanna is from
224 Aino Koivisto and Marja-Leena Sorjonen
15 16 17
[ka:nnonkoskelta ettäh? place.name so Kannonkoskelta so? HAN: TAN:
[£ehhhh (ni(h)ij -).£ £joo. oke:i.£ joo. yeah. okay. yeah.
In the segment above, successive Q-A-okei sequences are a part of a larger activity of ‘getting to know a new person’, featuring two “old hands” (Hanna and Anna) and a new-comer (Tanja). In the segment, the same person, Hanna, presents two requests for information to the newcomer, Tanja, and receives the answers with okei (lines 2 and 4, and 8 and 10). In both cases, okei is produced with a rise in pitch at the beginning of the second syllable and a glide to a final fall (with a vowel lengthening in line 10). The prosodic pattern is thus similar to the first okei in the preceding example, Excerpt 12. The rise in pitch makes the two okeis prosodically marked. In both cases, the okei is followed by an assessment by one of the two old hands. In the first case, okei (line 4) responds to information about the newcomer’s study field. Overlapping with the second syllable of okei, the informer continues her turn by saying that she has just started her studies (continuation possibly projected with a glottal stop at the end of the prior word nyttehuhuu £ok↑:e:i£. [hi
24
TOM:
25
VIL:
(.hhh)
27
VIL:
28
TOM:
. yeah
26
29 30
VIL:
31
TOM:
32
VIL:
[niih. yes
(0.6)
nii nii semmosta. so that sort of thing (1.0)
. yeah ni[i. yes
[pittää yrittää kek#ssii jotakii#. have to try to come up with something.
Tommi begins to prepare for the reason for his call in line 12 by apologizing that he is letting Vilma know so late and then producing a pre-announcement (‘you know what’, line 13). After Vilma’s go-ahead no (line 15, Sorjonen 2002) he proceeds to
230 Aino Koivisto and Marja-Leena Sorjonen
his projected turn that is formulated as a plan he has made with some other people (‘we were you know planning’) regarding small gifts that everybody is supposed to buy for a Christmas party (lines 16–21). Reporting this to Vilma implies that the plan is relevant to her. However, Vilma does not immediately respond to the turn but a 0.6 gap develops. This can be heard as foreshadowing a dispreferred response, or as an indication that extra time is needed for processing the information. Finally, Vilma responds with okei (line 23). The prosodic delivery of the particle bears similarities to the instances that orient toward the informing as newsworthy. That is, there is a pitch step-up and lengthening on the second syllable. However, there are also features that can be seen as adding an additional, affective layer. The stop k remains unreleased for slightly longer than is expected. In fact, the whole word is produced considerably slowly. This gives the impression that Vilma is processing the information she received (possibly because Tommi had woken her up), while the rise in pitch in the second syllable foregrounds the treating-as-news character of the particle. Together, these features contribute to a doubtful stance, that is, the speaker is not yet ready to respond to the turn as a proposal or to display acceptance or rejection of it. Treating the announced plan as a piece of news that is still in need of extra processing is obviously not enough for the sequence to proceed. That is, the announcement of a plan that also concerns Vilma would require a more explicit display of acceptance of and thereby commitment to the plan and/or an evaluation of it. What is noteworthy is that if produced as prosodically unmarked, the particle okei could constitute a sufficient response to the report as a proposal (directive action), indicating acceptance of and commitment to the plan. However, the prosodically heightened okei is insufficient, which is reflected in the next turns by Tommi. He produces turns that do not introduce anything new (‘so that sort of things’, line 28; nii ‘yes’, re-asserting the plan, and making Vilma’s response relevant, lines 24, 31, cf. Sorjonen 2001, 195–199), prompting a stronger commitment from Vilma. Vilma finally gives a response (line 32, ‘have to try to come up with something’), which then closes the sequence. In summary, although okei can be used to accept proposals, in this case the prosodic delivery conveys a stance towards the proposed activity. That is, the rise in pitch on the second syllable brings in orientation toward the news aspect of the informing of the plan (cf. Example (10)–(11)), which suggests that the content of the prior turn still requires processing before the plan can be accepted. Although the prosodic pattern differs from the previous example, it works similarly: it can also be understood as a provisional acceptance of the proposed plan.
Chapter 7. OKAY as a response to informings in Finnish 231
7. Conclusion In this chapter, we have focused on the usages of okei that we found to be relatively recent in Finnish interactions, those in which okei occurs in epistemically-driven sequences (cf. Couper-Kuhlen, this volume), responsive to the co-participant’s turn that implements an informing. In these usages, okei seemed to have entered the area of interactional activities that form the home base for response particles described by the cover term “change-of-state tokens” in English, such as oh in English and aijaa, aha(a), jaa, vai nii, ja(a)ha in Finnish. This raised the question of the kind of interactional work okei does compared to the dedicated change-of-state tokens, as well as the association between the design, activity and sequential contexts of the informings, and the action implemented by okei. Instead of situating okei in a category such as change-of-state tokens a priori, we wanted to tease out its work by exploring the different kinds of contexts of use within the domain of “informings”. As a first step in exploring the use of okay as a response to informings, we briefly compared its use with the use of response particles that have become conventionalized as responses that orient toward the epistemic aspects of talk. We found evidence for differences in the use between okei and the conventional change-of-state tokens from two sources: from the character of the subsequent interaction after the response, and from the order in which they occurred together. Thus the central change-of-state particle aijaa was followed by its speaker’s topicalizing question of the informing, whereas the okei speaker treated the informing as understood and sufficient, and not needing elaboration. Secondly, the order of the response particles was invariant: The response type that is lexicalized as a resource for epistemic work (aijaa) was directly adjacent to the informing. Okei, in turn, treated the informing as having formed a sufficient action in the current sequential context and thus proposed a sequence and activity closure (see also Koivisto 2013, on ainii as a similarly structured composite response). For understanding the work of okei as a response to informings, we further investigated the design and placement of the informing with respect to its placement in the interaction. The relevant aspect here was, in addition to the placement of okei in its home sequence (i.e., in third position in a question-answer sequence), the placement and design of the informing in a larger sequence and activity. Thus it matters whether the informing to which okay responds forms an independent informing (e.g., telling a piece of news), or whether it is understood as preliminary and instrumental to another main action or activity. In the latter cases, okei is produced as prosodically flat (without any marked internal modulation) and with a final fall, and the okei speaker moves after okei to an action that treats the informing as understood and not needing elaboration.
232 Aino Koivisto and Marja-Leena Sorjonen
As part of this exploration, we briefly discussed a group of particles other than the change-of-state tokens (like aijaa) which are used for treating the prior turn as understood and sufficient, the particles joo, joojoo and just. We explored these particles and okei in similar sequential contexts (in the third position of a question – answer sequence at the beginning of telephone calls), finding that okei provided a slightly different type of response in comparison to joo, joojoo and just. What appears to be relevant is the relationship between the sequence that the particle was part of and the subsequent talk by the particle speaker in the turn. More specifically, we found, subject to further work with more data, that joo, joojoo and just were followed by a sequence and topic that was unrelated to the prior sequence and the informing to which these particles responded. The demarcation between sequences and separate activities was enforced by producing the response particle with a falling final contour, and hence the subsequent talk formed a prosodic unit of its own. In contrast, the talk subsequent to okei in the turn oriented toward the informing being relevant from the perspective of a larger, still ongoing activity. Here the work of okei as receiving the informing as understood and locally sufficient, and its work in transitions and closings intertwine (cf. De Stefani and Mondada, this volume; Chapter 4, this volume). Our observations point to a possible ongoing reorganization of how informings are received, in which okei plays a role. To find out more about the use of okei in relation to change-of-state tokens (e.g., aijaa) and as a resource for orienting to the newsworthiness of the information, we studied contexts in which telling news about one’s life and catching up with someone was made relevant, that is, where the point of informings was not instrumental. In contrast to the okei responses produced with flat prosody and a final fall, we found here that okei was often produced with marked prosody, more specifically with a pitch step up and lengthening of the second syllable, ending with a final fall. This kind of prosodically marked okei response is the first response to the informing, and it receives the information as newsworthy. The okei was subsequently followed by an elaboration that specified the stance by the okei speaker, for example, by a positively valenced assessment. To what extent the topic will be talked about, however, is another matter, which depends on the specifics of the type of larger ongoing activity and the social roles that the speakers orient toward in the situation. We finished by considering instances in which okei was produced with stylized prosody. In these cases, okei appears to display doubt, puzzlement or bafflement with respect to the information received, making the acceptance of the informing provisional. We showed examples of two prosodic patterns. In the first case, the expression of doubt was displayed through second-syllable vowel lengthening, with a gradually rising final intonation. In the second, the prosodic delivery contained similarities with okeis that oriented toward the informing being newsworthy, having
Chapter 7. OKAY as a response to informings in Finnish 233
a pitch step up and lengthening on the second syllable. However, the additional feature that contributed to the stance of puzzlement in the segment we analyzed was the slow production of the response. In this chapter we have provided empirical information on the use of okay in some of the sequential contexts in which it is used as a response to an informing in Finnish interactions. We have discussed okei from some of the perspectives that we see as important in discovering what kind of interactional resource a particular response form, in this case okei, is. These include its sequential context and the possible larger activity context, the prosodic design of the response, and the design of the rest of the turn it occupies, as well as the development of interaction after okei. On a more general level, we have sought a more holistic analysis of the target phenomenon, including its larger context and the larger indexical field of response particles in the given language.
Acknowledgements We are grateful to Leelo Keevallik and Arnulf Deppermann for their apt and thoughtful comments on earlier version of this paper. We would also like to thank the participants of the OKAY meetings, as well as all the other authors of this volume for inspiring and wise discussions. Our thanks to Emma Betz and Lorenza Mondada for their helpful editorial work.
Chapter 8
When OKAY is repeated Closing the talk so far in Korean and Japanese conversations Satomi Kuroshima, Stephanie Hyeri Kim, Kaoru Hayano, Mary Shin Kim and Seung-Hee Lee
Tamagawa University / California State University, Northridge / Japan Women’s University / University of Hawaiʻi at Mānoa / Yonsei University
This chapter explores the use of the repetition of OKAY in a third position and in a transitional phase between activities in Korean and Japanese conversations. We have identified that (1) the duplicated OKAY, by being placed either in the middle of the other speaker’s turn or after a gap, is employed to accept the second position turn as sufficient and thereby propose to close the protracted sequence, and (2) it is deployed to serve as a boundary marker between two activities. These actions are commonly observed as an achievement of multiple actors displaying their orientation toward the activity through various resources, including the duplicated OKAY. Keywords: duplicated OKAY, curtailing, sequence-closing third, expanded sequence, transitional phase, proposal to close, gap, lapse, sufficient, perseverance
1. Introduction This chapter explores the ways in which the repetition of OKAY in duplication accomplishes various social actions when used in a response position and in a transitional phase between activities in Korean and Japanese action sequences of various types (e.g., telling, question-answer, making a suggestion, etc.). In this paper, we call the repetition of OKAY duplication or double sayings of OKAY (cf. Keevallik 2010c; Stivers 2004). Our main focus is on the instances of OKAY used as a reactive particle (Stivers 2004). It is possible that this practice is not exclusively observed in Japanese and Korean conversations. Our interest here is not to demonstrate the use
https://doi.org/10.1075/slsi.34.08kur © 2021 John Benjamins Publishing Company
236 Satomi Kuroshima et al.
of the duplicated OKAY unique to Japanese and Korean but rather to explore the workings of the duplicated OKAY using data from Japanese and Korean. Historically, OKAY was introduced to both languages more than half a century ago. In Japanese, its first appearance in public records dates back to 1930, when it was used in a popular song (Kato 2001), and 1951, when it was used in a popular novel (F. Hayashi 1954). In Korean, an early recorded instance of OKAY can be traced to a loanword dictionary published in 1937 (J. K. Lee 1937). Records show the use of OKAY in Korean popular magazines and newspapers, as well as in one novel, published in the 1920s and 1930s (M. Y. Lee 1932). Perhaps due to its foreign nature and its inclusion of phonemes that are indistinguishable in the target language, there are several phonological deviations from English in both Japanese and Korean OKAYs. In Japanese, neither syllable within the word is pronounced as in the original English okay. Moreover, the Japanese word has a lexical accent, realized as high-low (H-L), that differs from that of the English okay. In the case of duplicated OKAY, the first OKAY is typically produced with a higher pitch and a broader pitch range, and the second OKAY is produced with a lower pitch and a much narrower pitch range, equivalent to a downstep in intonation (Kubozono 1996). These phonological adjustments indicate that the word has been adapted into the Japanese language system (Kubozono 2006). In Korean, the first syllable is always pronounced as /o/, a short vowel, in contrast to the English /oʊ/. Although there is no documentation of the pronunciation of OKAY, it is typically pronounced as /okʰei/. In terms of the duplicated OKAY in Korean, both OKAYs are produced in more or less the same manner. However, the pitch movement within each OKAY token is a reversal of that in Japanese; that is, the first syllable of OKAY has a lower pitch than that of the second syllable. What is consistent across both languages is that the two OKAYs in the duplicated OKAY are produced and heard as a single unit, without a pause, rather than being produced as two separate units in succession, thus constituting a single turn. This suggests that the repeated tokens are performing a single action as a whole, rather than performing two separate actions (cf. Stivers 2004; Golato and Fagyal 2008). In both languages, the use of the duplicated OKAY is empirically differentiated from a single OKAY as a response token. As in other languages, a single OKAY can appear in the second position and third position after a sequence without any expansion (see Chapter 4, this volume); however, the duplicated OKAY usually occurs after a sequence expansion in the third position and between activities. The following excerpt, taken from a student group staff meeting, shows a single OKAY’s usage in the second position. VPR is the vice president, and PRE is the president.
Chapter 8. Duplicated OKAY in Korean and Japanese 237
Excerpt 1. Let’s take it out (KGSA 0607(1) 00:14:31), 2004, Korean, face-to-face, institutional interaction 01
VPR:
02
03 04 05
pre
umsik-un: mos ppay-telato, food-TOP NEG exclude-even.if even though (we) can’t take any food out,
*ike ppay-p-si-ta¿ this take.out-HON-RQ-PR let’s take this out¿ *looks up at VPR (1.5)
PRE: => .hhh *okhay.* *nodding* all
%(0.5)% %look at their budget sheet and make the change%
In the context of budgeting and planning for their next event, VPR’s proposal to remove china plates from the budget to reduce costs is accepted by PRE with okhay in line 4. PRE’s acceptance is taken as a final decision, and everyone looks down at their papers to remove china plates from their budget sheets. A single OKAY also occurs in the third position. For instance, in the following excerpt, taken from an interaction at a sushi restaurant, Chef 2, who has sought confirmation of the number of miso soups being ordered in line 1, produces OKAY in line 3 (’k↓ke.) to accept Chef 1’s confirmation, thereby proposing to close the sequence (Schegloff 2007). Subsequently, Chef 2 goes to the kitchen to prepare the order. Excerpt 2. Miso soup (UH_09_21_05 00:11:48), 2005, Japanese, face-to-face, institutional interaction 01
02
03 04
CH2:
*hitotsu:,* one one (soup), ch2 *hand gesture for ‘one’* CH1: +hitotsu de::,+ one COP one, ch1 +hand gesture on ‘one’+ CH2: => ’k↓ke. ((CH2 goes to the kitchen to prepare the soup))
Here is a similarly positioned OKAY from a Korean conversation. Two friends are having a phone conversation as research participants, and the excerpt occurs at the very beginning of the recorded conversation.
238 Satomi Kuroshima et al.
Excerpt 3. Research participants (LDC 6648 00:00:01), 1996, Korean, telephone conversation, informal interaction 01
A:
02
B:
kwaynchan-keyss-ci?= alright-should-COMM it’s all right, right?
03
=e, kulem ku[lem. ] yes sure sure yes, sure sure. A: => [okhe.] B: A:
hhh [hh hh [camkkanman, one:moment wait a second,
06
A:
nay-ka tto il pen-ul nwulle-ya I-NOM again one number-ACC punch-CONN
04 05
07
toy-nun kes kath-ta. must-ATTR thing seem-DC I think I need to push “one”.
In line 1, A asks for B’s permission to have a conversation for the purposes of research (i.e., asks if B feels comfortable participating in the research and being recorded). B confirms this in line 2. Then A closes this permission-granting sequence with okhe in line 3. After B’s laughter, A moves on to a different topic and sequence (lines 5–7). In our data, the duplicated OKAY was not observed in sequential environments in which a single OKAY was observed (i.e., in response to a first or second action of various sequence types without expansion). Instead, all of the duplicated OKAYs we found in our data occurred in two sequential environments: (1) after some type of sequence expansion and (2) at a “transitional phase” (see De Stefani and Mondada, this volume) between two activities. This distributional pattern suggests that the duplicated OKAY is employed to accomplish a different social action than that of the single OKAY. Closer examination of this phenomenon has yielded three related yet distinct uses of the duplicated OKAY in Korean and Japanese conversations. In the first type, speakers deploy a duplicated OKAY to curtail a protracted sequence by claiming the matter has been sufficiently dealt with and proposing the closure of the sequence. This practice is the closest to what Stivers (2004) has demonstrated for multiple sayings. In the second type, speakers use a duplicated OKAY to propose the closing of an expanded sequence that none of the participants show an interest in continuing (typically demonstrated through an ensuing gap). This is achieved by both the speaker of the duplicated OKAY, who treats the second action as appropriate and fitting for their project and now calls for a closure, and the recipient, who has produced a relevant response and is asked to align with the
Chapter 8. Duplicated OKAY in Korean and Japanese 239
proposed closure. Finally, the participants employ a duplicated OKAY to propose a transition to the next activity at a transitional phase by showing their readiness to shift to a new activity. In our analyses, we will investigate how different semiotic resources, such as language, prosody, bodies, and tools, are utilized so that the action performed by a duplicated OKAY is made recognizable as such. Rather than arguing that the duplicated OKAY is the sole resource for achieving a given action, we view it as part of the “contextual configuration” (C. Goodwin 2000; De Stefani and Mondada, this volume). Throughout the analyses, we hope to show how the duplicated OKAY is deployed to multimodally accomplish the speakers’ various practical purposes and how it is used in sequence closure. We also hope to demonstrate that transitions are not spontaneous but rather temporally organized emerging phenomena resulting from the achievement of action. 2. Background 2.1
OKAY in English, Korean, and Japanese
The usage of a single okay in English conversation has been previously described (e.g., Schegloff and Sacks 1973; Beach 1993; Schegloff 2007). The classic paper on the closing of telephone conversations (Schegloff and Sacks 1973) documents how okay is used to initiate and close pre-closing sequences. Schegloff (2007) also draws attention to how okay brings about closure by examining cases in which okay is produced in the “sequence-closing third” position. However, Beach (1993), who examines a wider variety of sequential environments in which okay appears, sheds light on its “pivotal” character, arguing that okay is not only responsive to the prior turn, serving to bring the sequence so far to closure; it can also be “preparatory in movements to what is offered as relevant for ensuing talk” (ibid., 338; italics in the original). On the other hand, notwithstanding OKAY’s long history as a loanword in Korean and Japanese, there has been no research on the use of the single OKAY in these two languages. Thus, the point of departure for our investigation was our own initial observations based on our collections of OKAY, which suggested that the basic positions in which OKAY appears in the two languages and the functions they serve are compatible with those reported for OKAY in English conversation. As in English, the single OKAY in Korean and Japanese is used to mark the completion of an activity, both in response to requests, suggestions, or informings and in third position as a sequence-closing third; thus, it is primarily responsive to prior turn(s), but it can also be a preparatory response to what is about to come.
240 Satomi Kuroshima et al.
Describing how the single OKAY in Korean and Japanese compares with that in English is an interesting line of investigation in its own right, part of which has been reported in Chapters 3 and 4 in this volume. In this chapter, however, we turn our attention to one particular way in which OKAY is deployed (i.e., the duplicated OKAY). 2.2
Multiple sayings in interaction
There is a growing body of literature on duplication in interaction (Barth-Weingarten 2011a, 2011b; Golato and Fagyal 2008; Heinemann 2016b; Keevallik 2010c; Stivers 2004). Although duplication is generally glossed as a means to intensify the semantics of the token being repeated (Haiman 1980), conversation analytic studies have revealed that when an item is repeated, the token performs “an action that is discrete from the item being repeated” (Stivers 2004, 268). The literature has examined the duplication of various linguistic items (Heinemann 2016b on the Danish change-of-state token nå; Keevallik 2010c on verbs, nouns, and tokens of other word classes), but the studies most relevant to the current study are those that focus on response particles. Among them, Stivers (2004) was the first to adopt a conversation analytic approach to multiple sayings (i.e., more than one). Examining multiple sayings of a response token (e.g., no no no) in English conversation, she shows that they convey that the prior speaker has persisted unnecessarily in the prior course of action and should properly halt it. Furthermore, Stivers (ibid., 280–285) has also claimed that multiple sayings in Japanese and Korean, as well as in some other languages, work in a similar way, suggesting that multiple sayings are a pervasive linguistic practice across languages. Golato and Fagyal’s (2008) study on double sayings of the German response token ja provides further support for Stivers’ claim: jaja is produced to indicate that the prior speaker has uttered something that is obvious and/or already known. They also show that the specific stance expressed by jaja varies depending on the placement of the pitch peak. Whereas jaja with the pitch peak on the first syllable (i.e., ^jaja) merely indicates that the prior utterance contains already known information, jaja with the pitch peak on the second syllable (i.e., ja^ja) treats the prior speaker’s utterance as containing something either unwarranted or self-evident. Barth-Weingarten, who presents a more detailed analysis of prosodic features of jaja (Barth-Weingarten 2011a, 2011b), further suggests that the stance that jaja indicates depends on accompanying bodily behavior (e.g., nodding) (Barth-Weingarten 2011b). These studies, therefore, suggest that the uses of multiple or duplicated sayings have cross-linguistic commonalities. This does not exclude the possibility that there are variations in their uses depending on the nature of the linguistic item that is multiplied/duplicated or the number of times the item is multiplied. This chapter
Chapter 8. Duplicated OKAY in Korean and Japanese 241
attempts to contribute to this line of research by focusing on duplicated sayings of a response token of foreign origin and the interactional functions they perform in Korean and Japanese. 3. Data Our paper draws data from a range of naturally occurring interactions between friends, family members, and colleagues, including audio-recorded ordinary telephone conversations and videotaped recordings of conversations in various settings. The Korean data come from the authors’ personal collections, as well as a publicly available corpus, the Linguistic Data Consortium Korean Corpus of Telephone Speech (available at http://www.ldc.upenn.edu/). Among the personal collections are institutional interactions, such as triage interactions at an emergency department, airline calls, graduate student group meetings, and ordinary conversations over dinner. In total, we have examined approximately 100 hours of data, from which we found 38 instances of a single OKAY and five instances of a duplicated OKAY. The institutional data (24 hours) were nearly devoid of OKAY, with only two single OKAY incidents found, and the rest were found in ordinary or semi-structured conversations between acquaintances, friends, and family. The Japanese data also come from the researchers’ own collections from both ordinary and institutional settings, such as a meeting at an architect’s office, sushi bar service encounters, a group meeting for an event, and guitar tutoring sessions. The Japanese dataset also includes the public CallFriend Corpus, distributed by Talkbank (http://talkbank.org/, see MacWhinney 2007), and 50 hours of the Corpus of Everyday Japanese Conversation (Koiso et al. 2018). In total, 95 hours of Japanese conversations have been examined: 55 hours of ordinary conversations and 40 hours of institutional interactions. Among them, we have found 40 cases of OKAY used as a response particle, the vast majority of which, in contrast to the Korean data, are from institutional settings, whereas only seven instances were found in ordinary interactions. Duplicated OKAY cases, however, only account for about one-fifth of the total occurrences (i.e., 10 cases) – eight in institutional settings and two in ordinary interactions. Although the total duration of these corpora exceeds 190 hours, astonishingly few instances of duplicated OKAY were found in each language. However, as noted above, these instances share largely similar practices. It should be noted that the study has drawn on the two languages equally and was not designed to be comparative. The main purpose of the study was to illustrate the use of the duplicated OKAY using data from the two languages and to describe its interactional relevance.
242 Satomi Kuroshima et al.
4. Analysis The duplicated OKAY was found in broadly similar sequential environments across Korean and Japanese. One common place of appearance was after an expansion of sequences, such as telling, question and answer, and suggestion and acceptance/ rejection. Whereas the single OKAY in the third position is used to simply register and accept the responsive action without expansion and propose to close it (as shown in Excerpts 2 and 3), the duplicated OKAY is employed to propose the closure of an expanded sequence as a sequence-closing third (cf. the similar use of both the single and the double OKAY in Italian reported by De Stefani and Mondada in this volume) or to propose a transition to a new course of action. In what follows, we first demonstrate two different ways of closing an expanded sequence with a duplicated OKAY: by curtailing an expanded sequence (4.1) and by proposing to close an expanded sequence after a gap (4.2). Finally, we will show how the duplicated OKAY is also employed to propose a transition (4.3). 4.1
“That’s more than sufficient”: Curtailing an expanded sequence with a duplicated OKAY
The following case from a Korean conversation is a prime example of how the duplicated OKAY is used to curtail the further development of a sequence by the recipient. Specifically, the OKAY speaker Myounghoon (MYO) interdicts the recipient’s turn in progress, proposing that the recipient has provided more than sufficient information. Here, four members of the Korean graduate student association staff are deciding the date of an upcoming event. After some collective discussion, Hongwon (HON) seeks confirmation from the president of the group, Myounghoon, that the date of the event will be May 13th at lunchtime (line 1), which Myounghoon confirms in line 2. Then Jian (JIA) informs them that she will likely not make it then (line 3). Excerpt 4. Unable to make it (KGSA0506(4) 00:03:10), 2004, Korean, face-to-face, institutional interaction 01
HON:
02
MYO: myo hon
+sipsam-il cemsim. thirteen-day lunch thirteenth, lunch. +nods at MYO-->
*yey, *cem*sim-ey=+ yes lunch-at yes, at lunch. *looks down at his phone*looks up at HON* -->+
Chapter 8. Duplicated OKAY in Korean and Japanese 243
03 04 05
JIA: myo
06
MYO:
07
JIA
=na *an-toy-l hwakyul kwoyngchanghi I NEG-become-ATTR chance very *looks at JIA-->
nopha-yo, high-POL it’s very likely I won’t be able to make it. (0.1)
a, mwe qual iss-ta kulay-ss-ci, PRT DM Qual have-QT say-PST-COMM oh, ((you)) said ((you)) have a qualifying exam, right, qua- qual-un yuwel mal-eyqual-TOP June end-at the qual- qualifying exam is at the end of June,
08
yuwel tal-ey iss-nuntey June month-at exist-but it’s in June, but
09
sipten
10 myo fig
11 12 13 14
owel isip-il-kkaci (.) nonmwun May twenty-day-until paper
nay-l ke #*iss-ketun-yo? submit-ATTR thing exist-CORREL-POL ((I)) have a paper to submit by May 20th -->*looks down at his cellphone-->> #fig.1
Figure 1: MYO (on the far left), looking down at his cell phone; JIA is on the right with a laptop. HON:
(0.2)
[a:: PRT Oh JIA: [deadline-i iss-e[se, deadline-NOM exist-because because ((I)) have a deadline, MYO: => [*okhe=okhe* *nods-----*
244 Satomi Kuroshima et al.
15
JIA:
16
HON:
kukka [na-nun mos ka-l ci-to moll-a. I:mean I-TOP NEG go-ATTR whether-also don’t:know-IE in other words, I may not be able to make it. [kulay (( right right ((
)) +kulay kulay kulay.+ right right right )) right right right. +nods------------------+
Myounghoon’s confirmation-seeking question (line 5) in response to Jian referring to their earlier conversation is formed with the committal suffix -ci, which embodies Myounghoon’s knowledge of the proposed understanding and evokes shared knowledge between the two (Kim and Suh 2004). The question is designed for, and presupposes, a simple confirmation. However, Jian’s answer is a transformative one (Stivers and M. Hayashi 2010) that is expanded to repair the misunderstanding displayed in Myounghoon’s question, as well as to account for her own limited availability for the event. Jian first corrects the date of her qualifying exam (lines 7–8) and provides information about another commitment (lines 9–10), which is the main reason for her predicted unavailability. Toward the end of Jian’s turn in line 10, Myounghoon’s gaze moves from Jian to his cell phone, likely looking at the dates on the calendar but also displaying his understanding that Jian’s turn is nearing a possible completion point. After a slight gap (line 11), however, Jian further starts to expand on her talk to more explicitly state the reason for her unavailability (line 13). Immediately upon hearing the expansion of Jian’s response, Myounghoon produces a duplicated OKAY accompanied by nodding (line 14). This proposes to close the sequence so far, suggesting that sufficient information has already been given for the question, so the discussion of Jian’s availability need not be continued. Myounghoon’s gaze, which has been down since the end of line 10, continues to be on the phone throughout the duplicated OKAY turn (see Figure 1), contributing to Myounghoon’s stance that the expansion is unwarranted. Jian’s line 15 also shows that she has understood Myounghoon’s duplicated OKAY as such; she accepts the proposal to close the sequence by quickly summarizing her expanded response. More notably, Jian’s summary is begun with kukka, a discourse marker that is used to initiate the rephrasing of a prior turn constructional unit (TCU) when encountering some trouble (Kim and Suh 1996). Hongwon responds to Jian with kulay ‘right’ multiple times throughout Jian’s turn, intensifying and strengthening the work Myounghoon’s duplicated OKAY has done (i.e., proposing that more than sufficient information has been provided and that the sequence should be closed). In summary, what is being proposed by the duplicated OKAY is to halt the in-progress expansion of the account of the problem and to close the sequence. Producing the duplicated OKAY in overlap with the
Chapter 8. Duplicated OKAY in Korean and Japanese 245
recipient’s mid-TCU displays the speaker’s stance that the expansion is unwarranted and is “overdone” (Stivers 2004, 271). The duplicated OKAY can also be used to call for an imminent closing of a sequence by coming in near its possible completion. In the next excerpt from a one-on-one guitar lesson, the instructor (INS) asks the student (STU) to play a chord sequence as an important part of demonstrating the basic chord progression. During the student’s demonstration, the instructor gives several instructions, the last of which is given in line 2. While the student plays two measures for each chord, the instructor counts the rhythm of eight with his right hand and two with his left, tapping both on his lap. Excerpt 5. Guitar lesson (GL_00010 2:51), 2018, Japanese, instructional setting
01
02
03 04 05
stu ins INS:
*%(2.0) *>>playing the sequence of C and Am--> %>>counting eight w/ his R hand and two w/ L hand by tapping his lap-->
stu
jaa tsugi, (0.8) kawaru yatsu.* then next change thing then next, (0.8) the alternating one. -->*
stu
*(4.0)* *playing the sequence of C, Bm7–5, E7, and Am*
stu
*(0.5)* *playing the last chord of Am*
INS: ins ins fig stu
=> soo iu %koto desu ne, +u:n,%#˚okke ok*ke.˚+#*% soo say thing COP FP ITJ that’s it, right. yeah. -->%puts 2 hands on lap%raises L palm up-down% +nods-------------->+ #fig.2 #fig.3 -->*makes last stroke* *with her guitar--*
Figures 2 and 3: Instructor raising his left palm up and down at the duplicated okke (line 05)
246 Satomi Kuroshima et al.
06 07
ins stu
INS: Ins
08
09 10
INS: ins
INS: STU:
%(1.0)%*(1.0)*%(0.5)% %nods % %nods-% *nods*
%on’naji %nakama tte iu:: same member QT say (they) belong to the same %brings up and down%grabs %his left hand-----%right
koto o%, thing ACC group, his---% arm---%
>%onaji nakama tte iu ka, e okhe okhe. yes
22
MIN:
ung. yes
23
WOO:
kulay. right:IE right
24
MIN:
kulay. right:IE right
25
WOO:
akka cip-ulo palo tuleka-ss-e? moment.ago home-LOC directly go.in-PST-IE you went back home right away?
In line 1, Woo’s wh-question asks about Min’s claim that something has changed about Woo’s response to the answer to the ringing telephone. Min, the CA student, attempts to answer in lines 2–20. Min starts responding by explaining the basis of the noted change (lines 2–3). However, before Min’s turn is complete prosodically and grammatically (specifying what the change is), Woo tries to resist such an understanding by reference to his “feeling.” He, however, does not fully articulate the utterance (lines 4–5). This leads to Min’s initiation of repair in the next turn (line 6). After line 6, Woo repairs by indicating that the basis of the difference is his knowledge of being recorded. However, Min directly rejects this (eyi kukey anila
Chapter 8. Duplicated OKAY in Korean and Japanese 249
‘nah, it’s not that’ line 9) and explains the difference she noticed as one between the opening of the first call (lines 9, 11, 13, 14) and that of the second call (lines 19–20); it was ‘hello’ in the first call but ‘yes’ in the second call. At line 18, Woo again rejects Min’s understanding, conveying that the difference is not because it is the second call. At lines 19–20, Min likewise rejects Woo’s understanding with ani ‘no’ and maintains her argument by re-explaining the basis of the difference. After these unsuccessful attempts to convince each other and to resolve the issue, Woo produces e ‘yes’ (line 21), which merely acknowledges Min’s argument. Immediately following this acknowledgement token, Woo proposes to close the sequence and the discussion as a whole with a duplicated OKAY. His proposal for closure treats the discussion as sufficiently dealt with. However, given that the information provided and negotiated in lines 2–20 was initiated by Woo’s question ‘what is changed?’ in line 1, and given the extent to which Min had to repair Woo’s misunderstanding, Woo’s mere acknowledgement e ‘yes’ is not fully appreciative of the information. In response to Woo’s proposal to close the sequence, Min passes the turn with ung ‘yes’ (line 22), leaving it as is and accepting the proposal. In other words, Woo proposes to close the sequence and the attempt to resolve the issue, despite the failure to come to an agreement, and Min accepts this by responding with ung. The closure of the sequence is further coordinated and mutually agreed upon with another sequence of kulay (roughly translatable as ‘all right,’ ‘right,’ or ‘okay’) in lines 23–24, which serves as an “exit device” (M. Hayashi and Yoon 2009). Then, at line 25, Woo moves to a different sequence and activity. In the examples thus far, the recipients of the duplicated OKAY accepted the proposal for a closure without issue. However, they can also persist with their course of action. Here is such an example from Korean. Sue and Hee are friends having a phone conversation. Hee lives in New York City (NYC). Sue is visiting the city and telling Hee about how beautiful their drive was to NYC from Buffalo. The excerpt begins with Sue’s story preface in lines 1–2. Excerpt 7. Beautiful scenery (LDC_4012 00:12:55), 1996, Korean, telephone conversation, informal interaction 01
SUE:
02
03
kuntey yeki new york o-nun kil-i by:the:way here New York come-ATTR road-NOM
elmana yeyppe-ss-nunci molu-n-ta? how pretty-PST-whether not:know-PRS-DECL by the way, ((you)) have no idea how beautiful the road was to New York. HEE:
way: why why? ((=what happened?))
250 Satomi Kuroshima et al.
04
SUE:
05 06
kuttay wuli-ka: han sey si ccum that:time we-NOM about three o’clock about chwulpal-ul hay-ss-ta¿ departure-ACC do-PST-DC at the time we left around three o’clock
HEE:
ung. yes uh huh.
((Lines 7–15 omitted, in which Sue describes with whom she drove and what time they were driving on the road)) 16
SUE:
17
kulaykaciko:, so
[awu ] ne:mwu nemwu yeyppun ke-ya. wow so so pretty thing-be:IE so, wow, ((the road)) was so, so beautiful.
18
HEE:
[ung.] yes uh huh
19
SUE:
20
HEE:
[kuntey] yeki o-ta-New York o-nun tey-nun? DM here come-TRANS New York come-ATTR place-TOP on the way here, the road to New York? [ung. ] yes uh huh
21
HEE:
ung. yes uh huh
22
SUE:
23
HEE:
mak ccwa::k mak- wancenhi [san-i-ya. DM spread DM totally mountain-be-IE it’s like, ((spread far and wide)) like- mountainous [al-a na-to enni. know-IE I-too sister I know it too, Sue.
24 25
SUE:
26 27
MWE-L AL-E↓ hh what-ACC know-IE ((YOU)) KNOW NOTHING hh
HEE:
28
29
ku ccok ka-nun ki:l yeysnaley [na-that way go-ATTR road long:time:ago I that road to ((NYC)). a long time ago I[↑AL-KI-N know-NML-TOP
↑CAKNYEN-EY: CONNECTICUT KA-NUN KIL-EY KEKI last:year-at Connecticut go-ATTR road-on there TULLE-SS-TA↓ stopped:by-PST-DECL last year ((I)) stopped by there on the way to Connecticut
SUE:
↑O:KA:Y [O:KA:Y↓
Chapter 8. Duplicated OKAY in Korean and Japanese 251
30
HEE:
31
[kulentey ku ccok-- ku ccok-ulo ka-nikka DM that way that way-to go-when and over there- when ((you)) go over there phwungkyeng-i tallaci-ko kuleh-canh-a:. landscape-NOM change-and be:so-you:know-IE the landscape changes, right.
32
SUE:
33
HEE:
kuntey mak nemwu nemwu yeyppun ke ya[:: but DM so so pretty thing be:IE but the view is like so, so beautiful. [ung. yes uh huh.
Upon Hee’s go-ahead to continue with the proposed telling (line 3), Sue begins the telling from the beginning of the trip (line 4). Then, in lines 16–22, Sue describes the scenery, which is the climax of the story. Sue’s recounting of the scenery treats Hee as an unknowing recipient. However, when Sue searches for a proper word to describe the scenery (evidenced by the discourse marker mak and several attempts to continue the turn in line 22), Hee intervenes, stating in overlap that she has independent knowledge of this scenery (line 23). Hee further explains how she knows, starting with yeysnaley na- ‘a long time ago, I-’ (line 24), which is aborted due to Sue’s outright rejection of this account but is possibly on the way to describing her direct experience with the view. Sue’s intervening turn jokingly (displayed with laughter and loud prosody) challenges Hee’s claimed epistemic status. This somewhat disaffiliative move is produced noticeably loudly throughout the turn. Hee responds to the challenge with the same prosodic intensity (i.e., loud throughout) by providing an account of how she has independent knowledge of the scenery (lines 27–28). Hee first specifies the time she saw the scenery by repairing the time formulation in her previous turn from ‘a long time ago’ to ‘last year.’ Hee also formulates the occasion as having occurred more intentionally (possibly a detour) than being on the way, with the verb tulle-ss-ta ‘stopped by’. These resources work to upgrade Hee’s knowledge of the scenery to counter Sue’s argument that Hee “knows nothing.” It is at this point – when Sue is faced with having to respond to Hee’s rebuttal – that Sue produces the duplicated OKAY. It proposes that the discussion (in lines 25–28) be halted, at the risk of displaying a dismissive stance. However, Sue also downgrades this stance and makes the proposal for a closure playful by prosodically emphasizing each syllable, so as to match the loudness of prior turns in lines 25–28. What unfolds after the duplicated OKAY distinguishes this case from the previous excerpts. Hee does not acquiesce to Sue’s proposal to close the course of action; instead, she continues to support her claim by providing her own description of the scenery and solicits Sue’s agreement (lines 30–31). However, neither does Sue budge. At line 32, Sue simply resumes her telling that was halted earlier in
252 Satomi Kuroshima et al.
line 22 without responding to Hee’s immediately prior turn, thereby deleting Hee’s lines 30–31 as if Hee has accepted the proposal to close the sequence. Note that both Sue’s and Hee’s turns in lines 30 and 32 are initiated with ku(len)tey (often translated as ‘but’), a discourse-organizing connective used when dealing with a problematic situation (Kim and Suh 1996). In the end, Hee acquiesces to Sue’s agenda by simply acknowledging and allowing Sue to continue with ung ‘yeah’ (line 33). Although it is not included in the excerpt, Sue carries on the telling by continuing to talk about the animals she saw on the road. In summary, we have shown two places in which the duplicated OKAY occurs: in overlap with the previous speaker’s TCU curtailing the sequence (Excerpts 4–5) and after the turn is possibly complete but often not the course of action (Excerpts 6–7). In both positions, the duplicated OKAY is used to propose the closure of a larger course of action (rather than responding to an immediately prior turn) that has already been, in the OKAY speaker’s view, sufficiently dealt with, and to prevent a likely further expansion (cf. Stivers 2004). 4.2
Confirming no more expansion with a duplicated OKAY
Unlike the duplicated OKAY in previous excerpts, which is positioned in overlap with or immediately after the recipient’s turn, the duplicated OKAY in the following examples is produced after a “gap” (Sacks, Schegloff, and Jefferson 1974), as the sequences could potentially be more expanded. Thus, the gap indicates that neither participant has more to say, and both are ready to close the sequence. The analysis will also demonstrate that the participants’ embodied orientation to their readiness to close a sequence is already observable through various resources in their last turns. The next excerpt, recorded in Tokyo, is from an architect design firm. The four participants – Ken, Tet, Hiro, and Yuto – are discussing a plan in front of a computer monitor that displays the design that the lead designer of this project, Ken, developed. This was taken from a moment in which the discussion was nearly ended. Excerpt 8. Square plan (Architects), 2006, Japanese, face-to-face, professional meeting
01
02 03
KEN:
ken
KEN: ken ken
*(d’)ii jan. sore de. toriaezu.* fine TAG that PRT for.now that’s good. that way. for now. *looking at the computer monitor*
*(0.5) *looks at Tet-->
+°dou.* dou.° see? see? -->*looks down--> +starts to walk away from the group-->
Chapter 8. Duplicated OKAY in Korean and Japanese 253
04 05
06 07
08
09 10
11
12 13
14 15 16
TET: ken ken ken
TET:
KEN:
tet tet tet
KEN: tet
(2.3)
nn.+ ITJ yeah. -->+
+* (0.2) +* (0.2) * +stops-------+ *looks at Tet*at the computer-->*
↑puran:: >honto< mashikaku de ii %no plan really square PRT fine FP can (it) be perfectly square in the plan? %looks at Ken-> ≠a↓:a≠ (.) >toriaezu% %±(0.5)± ±nods± %diverts gaze from Ken--> ano saigo,% uhm last uhm at the end, -->%
tet ken ken
+%*(1.0)+ %looks at Ken--> *looks at TET--> +moves hands as if modifying a square object+
tet
%(0.5) %middle distance-->
TET:
saichoo[≠see, readjustment readjustment, KEN: [≠chousei≠%sure ba ii to *omou adjust do if fine QT think if (we) adjust, (it) will be fine, (I) think. tet ≠nods---≠ tet -->%looks down to %an eraser in his hands--> ken -->*looks at *the monitor--> TET: => #%˚’kke ok↓ke˚% tet %nods--------% fig #fig.4
KEN: ken tet
%mo- (.) *ma↓a toriaezu ↑ima no %↓dankai de wa well for.now now PRT stage PRT TOP wewell, for now, at this stage, -->*looks at Tet----> -->%looks at Ken-->
254 Satomi Kuroshima et al.
17
18 19
20
21 22
23 24 25 26 27
KEN: tet tet
KEN: ken tet
ken tet ken
KEN:
TET:
%(0.2)
˚˚shikaku˚˚
square (it should be) a regular square. %nods--> %(0.5)% %turns to his right while nodding%
*%’to nanka koo maru ga do- nan shurui atte:,* and like this circle NOM what kinds be and how many kinds of circles are there, *averts gaze from Tet and looks at middle dist--> %looks at the screen---> *dou iu takasa de how say height COP how tall it is, *looks at Tet--> %*(0.2)%* %nods--% *......*
[*doo hashira ga miete kuru *ka tte iu *no ˚ga.˚ how pillar NOM look come Q QT say NML PRT how the pillars will come to look (should be adjusted). *hand gesture of a pillar--*,,,,,,,,,,*looks at monitor-->
[nn: nn ITJ ITJ yeah yeah (0.5)
TET: => #okke ↓okke% tet --->% fig #fig.5 tet
TET:
%(0.8)% %grabs an eraser once%
%jaa dou shi yo kka. then how do AUX Q now how shall we do this? %turns away from Ken and starts to move with an assistant--->>
TET
TET KEN
Figure 4: Tet is looking down at his hand
KEN
Figure 5: Tet is looking toward his front
Chapter 8. Duplicated OKAY in Korean and Japanese 255
When the previous discussion of the plan reaches its completion, Ken almost disperses the team (line 3). However, Tet, who is responsible for making a 3D model, looks at Ken and asks him to confirm whether the plan of the room can be ‘perfectly square’ (line 7). Because this question picks up an issue that has already been agreed upon, it can be heard as potentially challenging to the recipient. Ken’s simple confirmation in line 8, repeating the ‘for the time being’ first spoken in line 1, receives only nodding and a shift in Tet’s gaze in line 9, rather than a verbal uptake. Tet’s response occasions Ken to further elaborate that they will make an adjustment afterwards, voluntarily expanding his own response (Stivers and Heritage 2001). With Tet’s repair solution to Ken’s possible search in lines 10 and 11, Tet’s understanding of ‘for the time being’ is made available. Upon this seeming agreement on the issue, Ken diverts his gaze from Tet to the computer screen, and Tet looks down at an eraser held in his hand (line 13), observably making a gradual exit from the interaction. It is at this point that Tet produces a duplicated OKAY in a small voice (˚’kke ok↓ke ˚) with concurrent nodding (line 14; Figure 4), treating the provided account for the confirmation as acceptable and sufficient and proposing to close the expanded sequence. However, Ken immediately reconfirms the plan of the square shape in lines 16 and 17, which indicates that the duplicated OKAY only suggests to propose the closure of the sequence, and that the closure could be open for further negotiation by the other party (cf. Excerpt 7). Ken displays his orientation toward revisiting the matter, with an emphasis on the tentative nature (‘at this stage’) of the original plan. He shifts his gaze from the monitor to Tet, who proposes to close the topic, further demonstrating his persistence on the explanation. Because Ken is revisiting a matter that was once accepted by the recipient, the provision of a further account is made relevant (Stivers and Heritage 2001). Tet’s nodding and quick turning away from Ken at line 18 after the recognizable completion of Ken’s turn embody his understanding of it as such. Thus, Tet allows Ken to take a further turn. In fact, Ken elaborates by revealing what kind of chousei ‘adjustment’ they can make for the current project (lines 19 and 20) (i.e., the number of circles, the height, and the appearance of the pillars). The slot after the duplicated OKAY is utilized by Ken to volunteer a justification of the plan. Tet’s recipiency and alignment with Ken’s course of action via nods and a continuer is induced by Ken’s halting of his turn (line 21). When Ken possibly completes his account in line 22 with a falling intonation – which was also realized by a subsequent gap in line 24, indicating no more expansion to be made by Ken or Tet – Tet produces another duplicated OKAY (okke ↓okke) while looking forward (line 25; Figure 5). Positioned in this environment, the duplicated OKAY, produced more loudly than the previous one, along with the concurrent nods, accepts the volunteered justification and aligns with Ken’s displayed orientation to closure. Because this justification consists of Tet’s possibly challenging question in line 7, through his
256 Satomi Kuroshima et al.
embodied alignment with Ken’s action and the duplicated OKAY, he can claim that he is treating Ken’s explanation as a sufficient, reasonable basis for the plan and thereby proposes a concrete closure of the extended sequence. Indeed, the sequence closes as proposed; Tet starts to leave the meeting and initiates a dyadic talk with an assistant in line 27. The two duplicated OKAYs we have examined in Excerpt 8 suggest that the token alone cannot achieve the action of closure proposal. Rather, the sequential environment, the participation framework (i.e., Tet initiates a possibly challenging question to Ken; Tet is entitled to judge the justification of the plan), and their embodied orientation through gaze and interaction with various objects (e.g., an eraser and a computer monitor), all of which constitute a “complex multimodal gestalt” (Mondada 2014a), configure the activity organization and enable the speaker of the duplicated OKAYs to propose to close an expanded sequence, and this proposal might either be aligned with by co-participants in closing or resisted by continuing the previous action (De Stefani and Mondada, this volume; C. Goodwin 2000, 2017). Excerpt 9 is another case in Japanese in which a duplicated OKAY is produced to close an expanded sequence, during which participants handle (or negotiate) alignment and affiliation. In this excerpt, three family members (two parents and their adult son) are dining at a sushi bar counter in Japan. Prior to this segment, the son was deciding on his next drink with his mom. The mom said she was ordering wine next, and the son replied that he might as well have another beer, but with the chef ’s recommendation, he decided to have Japanese vodka (not shown). Excerpt 9. Wine ordering (CEJC T004–003 12:43), 2017, Japanese, face-to-face, informal interaction 01 02 03 04 05 06
mom son
MOM:
SON: mom
SON: mom son
%(0.4) >>looks at her left where SON is--> %looking at a middle distance--> ↑wain non’ demo ii: n da yo:? wine drink if find SE CP FP ((you)) can have wine, you know. (0.2)
˚u:n.˚ yeah.
*(2.0) *turns to her right-->
wain wa::: *%ii ya% hhh wine TOP fine FP ((I)) don’t want wine. -->*brings back gaze to her left--> %nods---%
Chapter 8. Duplicated OKAY in Korean and Japanese 257
07
08 09
10
11
12 13 14
15
MOM:
mom son
KAT: mom
MOM: mom son
SON:
*i::i?* fine ((you)) don’t? *nods*
*%(0.8) *,,,,,grabs a sake glass--> %looks down-->
↑hai doozo* ((serving for the other guest)) ITJ please here you are. -->*
*pia%dooru da yo?, piat.d’or COP FP (it)’s Piat d’or. *slowly brings a sake glass to her mouth--> ->%looks towards mom-->
son
%a %piadooru ↓ka= ITJ piat.d’or Q oh ((it))’s Piat d’or. %,,,,,%looks up-->
mom son
*(1.0)*% *drinks sake* -->%
MOM:
SON: mom son mom mom son fig
=nn yeah
*%nomu n dat tara hito kuchi ˚˚nomi tai na.˚˚* drink SE COP then one sip drink want FP if (you) are drinking (it), (I) want a sip. *puts down a sake glass-----------------------* %looks towards mom--> *(0.2)+(0.4)#% *turns head/gaze towards SON--> +nods--> ->% #fig.6
Figure 6: Mom is nodding 16
MOM: mom son
=>o%kke okke.+ -->+ %looks down-->>
258 Satomi Kuroshima et al.
17
18
MOM: mom mom
+aa jaa atashi no age mashoo: ITJ then I GEN give POL:VOL oh then (I) will give you some of mine. +puts her both hands on her lap-->> (1.5)* -->*gaze to her empty glass-->>
After the son orders, the mom suggests in line 2 that he can have wine, implying that it is also a good choice for him. However, her suggestion is rejected by him in line 6. Although the mom once registers the rejection (line 7), she further “reworks” (Schegloff 2007, 162) the suggestion by referring to the brand of wine in line 10, which seems to be similar to what Schegloff (ibid.) calls a “follow-up inducement.” Her provision of a piece of information, presumably intended to appeal, conveys the persistent stance of her suggestion. The son indeed shows his interest by directing his gaze toward his mom after hearing the brand name (line 10) and receives the information as worthy of consideration; he embodies a “thinking face” looking upward (line 11). After giving thought to the selection, the son displays provisional acceptance of his mom’s proposal in line 14 by conveying his interest in taking a sip but not ordering a glass. When the son concedes in this way, the mom displays her orientation to the son through her head direction and starts nodding as she wordlessly puts her sake glass down (line 15; Figure 6). Thus, realized as a gap, the moment is occupied with her bodily conduct, embodying her acceptance of his concession and indicating that no more negotiation needs to ensue. It is at this point that she produces a duplicated OKAY in line 16 (okke okke.). With it, the mom not only demonstrates that she is not pursuing full acceptance of her suggestion any further but also announces a closing of the persistent negotiation. Hence, the mom also concedes by treating her son’s response as satisfactory for her project, as she is entitled to do so. Indeed, her subsequent expression of her intent to offer her wine in line 23 is prefaced by a change-of-state token aa ‘oh’, the turn design of which arguably shows that she makes this offer independently, rather than as a continuation of her previous suggestion, to accommodate her son’s wish. Similarly to Excerpt 8 – with its duplicated OKAY, its projected acceptance of the recipient’s alignment through nodding during a gap, and the sequential positioning – the speaker conveys that she is ready to close the course of action she originally began. The last two excerpts above have shown how duplicate sayings of OKAY after a gap can close a prolonged sequence by treating the matter as sufficiently dealt with. This practice of proposing to close a course of action that the speaker themselves has initiated is compatible with that of the earlier cases we examined in 4.1. However, what systematically sets them apart is that the speakers in the latter examples produce the duplicated OKAY during a gap, after there has been an embodied indication (in both cases through gaze direction and nodding) of the speaker treating
Chapter 8. Duplicated OKAY in Korean and Japanese 259
the response as sufficient and appropriate for the current project, as well as the recipient’s alignment toward the speaker’s action. As such, the duplicated OKAY is designed to be both “responsive” and “preparatory” for what immediately ensues (cf. Beach 1993 on the single okay in English). It should be noted that participants are showing their orientation to move toward sequence closure prior to the production of a duplicated OKAY (cf. De Stefani and Mondada, this volume). The provision of a sequence closure via a duplicated OKAY confirms that the issue has been sufficiently dealt with, and therefore no more expansion is made. 4.3
Proposing to move on to a next relevant course of action: Transitional work
The duplicated OKAY is also employed during a “transition phase” (De Stefani and Mondada, this volume), in which proposing the completion of a previous activity and moving on to the next activity are both relevant. This dual work is clearly shown in the following excerpts, in which, rather than being produced as a closing third of the extended sequence, it is produced after the prior sequence is closed, which is underscored by a “lapse” (Sacks et al. 1974), so that the transition to a new course of action is now relevant. What can be commonly observed throughout these excerpts is that the participants’ orientation to the transition is observable during the lapse, thus demonstrating that the transition is done not solely via a duplicated OKAY but as a joint achievement by the mutual orientation of the co-participants’ alignment. In Excerpt 10, three young mothers – Yumi (YUM), Haru (HAR), and Nozomi (NOZ) – are planning on baking prior to their children’s event at a restaurant. Each of them has a list of items (e.g., a shifter, etc.) that they need to prepare for the baking, and they are going through the list to see who can supply these items. At the beginning of this excerpt, all of the participants are engaged in looking at the list in front of them, and Nozomi maintains her posture throughout, until she initiates a telling (line 12). Haru’s question regarding the time when they must leave the venue in line 1 was initiated after Nozomi’s offer to supply items was accepted (not shown in the transcript). Because of Haru’s question, the activity of going through the list has been put on hold. Excerpt 10. Baking (CEJC: T003–004 5:45), 2017, Japanese, face-to-face, group meeting
01
02 03
HAR: yum
HAR: YUM:
(place name) tte, juuni ji QT 12 o’clock as for (place), (we) need >>checking off on the list--> tes[shuu: na kanji da yone.= withdraw COP like COP FP to leave at noon, right? [ right.
260 Satomi Kuroshima et al.
04
YUM:
05
HAR:
06
YUM: yum
07
HAR:
08
YUM: yum
09 10
HAR: har yum yum yum fig
11
12
YUM: yum har yum har
13 NOZ:
14
noz noz
15 NOZ:
=>soo soo [soo< right right right. [un un un.= yeah yeah yeah. =juuni ji ni owaru *kara *ne::. 12 o’clock at end so FP (it) will finish at 12 o’clock. -->*,,,,,*places her hand at *bottom of sheet--> un, [un. yeah yeah. [are *(.) ano asoko heya ga ne::= that well that room SP FP that (.) that room, you know. -->*starts erasing with an eraser *attached to the pencil------> =un, %un.% yeah yeah. %...%picks up her cup-->
(2.0)*(1.0)#*(1.0)#*(0.5) -->*changes the direction of pencil *corrects something-> *put both hands on the table and reviews the sheet-----> #fig.7 #fig.8
Figure 7: Making a correction after erasing
Figure 8: Reviewing the list
=> ↑okkee ↓okkee. %*.hh jaa ato go. then what’s.left five then (we) need 5 more items. -->*grabs a pencil and starts writing--> ...%puts down her cup------->
(1.0) *(1.8)% *%(1.8) -->*stops----*writes again->> -->% %looking at the list-->> &ah,& ITJ oh &looks up&
&(1.0)& &puts her pencil down&
jitsu wa sa:::, ((starts telling about someone)) truth TOP PRT truth be told,
Chapter 8. Duplicated OKAY in Korean and Japanese 261
When Haru receives a response to her question in lines 3–4, Haru proposes to close the sequence in line 5. However, Yumi further adds a specification regarding the ending time of the reservation (lines 6 and 8), which is received by Haru in lines 7 and 9, and thus this expanded sequence is recognizably completed (further confirmed by Haru’s sipping her tea at this point). While this exchange is taking place, Yumi works to update the list; she writes something down on the list with her pencil (lines 1–6) and keeps her gaze on the sheet. While she offers the specification (lines 6 and 8), Yumi places her hand toward the bottom of the page and starts erasing something there (line 8). After the verbal exchange is complete, 2.0 seconds of erasing pass; then she corrects something for 1.0 seconds (Figure 7), puts both hands on the table, and reviews the sheet (Figure 8), embodying that she has finished updating the list. All of this writing and erasing activity is completed as Yumi’s sole activity; she does not look at the other co-present participants, who are also looking down at the list. This is when Yumi produces the duplicated OKAY in line 11 (↑okkee ↓okkee.), which serves to organize a transition to a next activity (cf. “transition phase” in De Stefani and Mondada, this volume). This duplicated OKAY is not addressed to anyone (it is rather heard by the entire group), because it is produced during Yumi’s overview of the list. Also, it should be noted that Yumi is the one who created the list and is thus entitled to take the lead in this particular activity. With the duplicated OKAY produced after Yumi’s solitary writing, following her overview of it (perhaps to check her work), the speaker proposes transitioning to the next relevant activity. Indeed, Yumi announces a resumption of the previously halted activity next in line 11. Because the duplicated OKAY is produced between the lapse after the sequence completion and the announcement to resume their remaining business, it may serve as a marker of the boundary between activities. The prosodic property of the duplicated OKAY – high onset on the first one followed by a contrastively low pitch on the second one – might further suggest this dual aspect of the action, i.e., the closing implicativeness and projection of beginning a different activity. Evidently, in her effort to resume the previous activity, Yumi clearly marks the sequential relationship of this turn with what happened before in line 11. First, the ordering marker jaa ‘then’ projects the incipient next action. Second, ato ‘what is left’ clearly formulates her action of returning to the unfinished list. Finally, she grabs her pencil and starts writing while producing an announcement, displaying that she is engaged in the next activity. Correspondingly, Haru, who was taking a sip of her drink, upon hearing the duplicated OKAY and the announcement, puts down her cup (lines 11–12) and starts re-engaging in looking at the list (line 12). In this way, Haru also displays her orientation toward transitioning to another activity by disengaging from her solitary action of drinking.
262 Satomi Kuroshima et al.
Although somewhat differently organized, the following telephone conversation offers strong evidence of this dual function of a duplicated OKAY in a transitional phase, as the recipient of the duplicated OKAY indeed proposes to shift to a new activity. In Excerpt 11, Ken, who is a close friend of Ryo, calls Ryo to invite him for a drink during the upcoming holiday. After the initial invitation, which is the main business of the call, is accepted, they talk about Ryo’s partner, whom Ken has met before. Excerpt 11. Let’s have a drink (MC3), 2016, Japanese, telephone, informal interaction 01
KEN:
aa:: soh na n’ da:. ITJ so COP SE COP oh, I see. ((receipting information about A’s partner))
02
RYO:
°°↓mm°° yeah.
03
KEN:
.hhh
04 05 06
RYO:
KEN:
07
08 09
a::n hmm. (0.5)
ja’, >toriae-< t- ma- ma’ hakidashi then well for.now spill tai koto toka mo aru daro(h)h sh(h)i, .hhh want thing etc also exist maybe because then, for- well, for now, you might have something to spill out so,
RYO: KEN:
10
heh[ehehe [ko- ko(h)ndo no- ko(h)ndo no next GEN next GEN re(h)n(h)kyuu ni.hh hh holidays on on the next holiday,
11
RYO:
£aa.£= yeah.
12
KEN:
=shippori- .hh shippori nomimash[oh. quietly quietly drink-POL:VOL let’s quietly have a drink.
13
RYO:
14
[soh da na. sounds good. (0.7)
15
RYO:=> okke okke::.
17
(0.3)
16
KEN:
mm, gomen ne shigoto chuu ni. ITJ sorry FP work during at yeah, sorry (to call you) during work
Chapter 8. Duplicated OKAY in Korean and Japanese 263
18 19
RYO:
20
KEN:
21
RYO:
aa, daijohbu. ITJ fine oh, no problem. (0.3)
hoijaa [mwell, t[>a jaa sore made-< sore made wa modotte ITJ then that till that till TOP return
22
ko nai no? mada. kocchi wa. come NEG Q yet here TOP oh then, (you) won’t be back until then? not yet? to here?
23
(0.6)
When the small talk comes to a possible completion in line 4, Ken returns to the arrangement they have just made for getting together for a drink in line 6. When the invitation is reissued and accepted by the recipient in line 13, 0.7 seconds of silence ensues. When such silence manifests itself as a lapse, as it becomes obvious that neither of them will take a further turn, a transition to a new activity is expected to occur. Then, Ryo produces a duplicated OKAY in line 15 (okke okke::.). Because it is produced following the acceptance of the invitation by the same speaker, it can be understood that the speaker has no issue with it; therefore, they can move on to another activity. Indeed, upon hearing this, the caller, B, acknowledges it (‘yeah’) and re-refers to his calling the other at work (he has mentioned this at the beginning), confirming no “unmentioned mentionables” (Schegloff and Sacks 1973). This move shows the speaker’s understanding of what is proposed by the duplicated OKAY. Thus, because the transition has an emerging organization – with which co-participants align themselves and display their orientation as such to each other – the sequential environment of the duplicated OKAY in relation to the overall organization of the phone call facilitates the achievement of the transition. Whereas Excerpts 9 and 10 show that a duplicated OKAY is produced to confirm a closing of a sequence, which has been embodied by the speaker and the recipient, in Excerpts 11 and 12, the duplicated OKAY is produced to propose proceeding to a next relevant activity after the prior course of action has come to a recognizable ending, and the subsequent lapse further underscores the realization of a sequence completion.
264 Satomi Kuroshima et al.
5. Discussion By drawing on both informal and institutional interactions captured in video and audio recordings in Korean and Japanese, we have described the usages of the duplicated OKAY. Depending on how the duplicated OKAY is positioned with respect to the prior talk, the speakers can display different stances and activity orientations to propose the closing of a course of action, after which they can shift to something else. The distributional pattern common to both languages suggests that the use of a duplicated OKAY constitutes a distinctive interactional practice. The features of OKAY described in the literature as backward-looking (i.e., responsive to prior turn(s)) and shift-implicative forward-looking (cf. Beach 1993) are also prominent in our examples. Our analyses have further shown that (1) the duplicated OKAY, by being placed either in the middle of the other speaker’s turn or after a gap, is employed to accept the prior turn as sufficient and thereby propose to close the protracted sequence, and (2) it is deployed to serve as a boundary marker between two activities. These actions are commonly observed as an achievement of multiple actors displaying their orientation toward the activity through various resources, including the duplicated OKAY. Furthermore, we have also noted that the duplicated OKAY shares some features with “multiple sayings” (Stivers 2004). According to Stivers, multiple sayings “convey that the speaker has found the prior speaker’s course of action to have perseverated needlessly and proposes that the course of action be halted” (ibid., 280). Her excerpts include multiple sayings in both the second and third positions (i.e., sequence-closing third and third position repair initiation). In fact, the instance cited as exemplifying a sequence-closing third was the multiple saying of okay in English (Excerpt 10 in Stivers 2004), which is akin to our excerpts in 4.1, in which the duplicated OKAY is used to close the sequence originally begun by the speaker of the duplicated OKAY. The commonality does not seem to be a mere coincidence; rather, it points to a general feature of the practice of using OKAY in this position: differentiating the development of a course of action they initiated without any negotiation (i.e., a single OKAY) or with some negotiation (i.e., double and multiple sayings of OKAY). On the other hand, in our corpus, we did not find any occurrences of multiple sayings of OKAY (i.e., more than duplication, such as OKAY OKAY OKAY). Given that Stivers (2004) has in fact demonstrated the presence of multiple sayings in Korean and Japanese (p. 280), the duplicated OKAY in Korean and Japanese could be a distinctive practice beyond the range of actions that multiple sayings can accomplish. However, as mentioned above, some properties are shared between these two practices in both languages, particularly that of stance display. Therefore, it
Chapter 8. Duplicated OKAY in Korean and Japanese 265
can be suggested that these practices lie along a continuum of ways to propose the closure of a protracted sequence, one end being more token-specific (i.e., as in a duplicated OKAY) without displaying any particular stance, and the other more specific to the practice of multiple saying (i.e., in various formats), with the stance against the other person’s perseverance in various response positions. In addition, as our final two examples show, a duplicated OKAY can appear at the boundary between courses of action, which differs from what multiple sayings can do. Furthermore, given that speakers duplicate response tokens in other languages (cf. Barth-Weingarten 2011a, 2011b; Golato and Fagyal 2008; Heinemann 2016b; Keevallik 2010c), what is distinctive about the duplicated OKAY awaits further empirical demonstration through comparison with the duplication of other items in Japanese and Korean. More broadly, this study also contributes to our understanding of how a loanword inhabiting a typologically different language can be added to speakers’ practices for achieving various purposes.
Acknowledgements The Japanese portion of the project was led by Satomi Kuroshima and the Korean portion by Stephanie H. Kim. The work conducted by the second author was supported by the California State University Northridge College of Humanities’ Faculty Fellowship and Grant Program. The authors are grateful for the helpful comments from two of the anonymous reviewers and the editors of this volume – Emma Betz, Arnulf Depperman, Lorenza Mondada, and Marja-Leena Sorjonen – and for the participants in the two workshops held in Mannheim and Helsinki. We would also like to thank Yusuke Arano for allowing us to cite his data.
OKAY in specific activities and settings
Chapter 9
OKAY in health helpline calls in Brazil Managing alignment and progressivity Ana Cristina Ostermann and Katariina Harjunpää
Universidade do Vale do Rio dos Sinos / University of Helsinki
This chapter investigates phone calls to a health helpline in Brazilian Portuguese, focusing on how call-takers employ OKAY as a resource in managing the participants’ mutual alignment and the progression of activities during the calls. Call-takers make use of OKAY in transitioning between main and subsidiary actions and activities and in eliciting uptake from the recipient. In both contexts, the use of OKAY relates to the work of managing the institutional agenda and technical resources in relation to the situated concerns of the callers. The analysis thus contributes to the study of particles and institutional interaction by showing how OKAY is used in organizing the routinized, agenda-driven conversational structures of helpline interactions. Keywords: Brazilian Portuguese, eliciting uptake, transitions, helpline, information delivery
1. Introduction This paper investigates the use of the Brazilian Portuguese variant of OKAY in phone calls to a government-run toll-free health helpline in Brazil, here referred to with the pseudonym Centro de Informações sobre Saúde (CIS). Callers to the helpline make inquiries concerning health matters. The call-takers respond to the inquiries by searching for the right information in a health information database, consisting of written texts that they then read aloud to the callers. We analyze the use of oquei1 mainly in the call-takers’ turns, as related to organizing different activities in the calls and balancing between the institutional agenda, technical resources and the situated concerns of the callers.
1. In Brazilian Portuguese dictionaries and textual sources, the abbreviation OK is typically used. We adopt an orthographic representation that is in line with the pronunciation of the token. https://doi.org/10.1075/slsi.34.09ost © 2021 John Benjamins Publishing Company
270 Ana Cristina Ostermann and Katariina Harjunpää
The CIS has two overarching aims: (i) informing citizens about health promotion, disease prevention and treatment, and (ii) referring them, when necessary, to local health-units for diagnosis and treatment. The call-takers are trained and expected to follow certain guidelines, some quite standardized, such as passing the information to the callers in the same order set by the computer-database. These features create constraints on how the call-takers can adjust their responses to an individual caller’s local concerns. The callers to the helpline, by contrast, are not necessarily familiar with these routines and constraints. The potential gap between organizational policy and call-specific contingencies has to be worked out in situ in the interaction between the institutional representatives who take the calls and the laypersons (Zimmerman 1992). The call-taker needs to get the caller “on board” in making the caller’s inquiry one that fits what is available in the database, in checking whether the caller is actually receiving the requested information from the partially prefabricated answers, as well as in locally implementing the larger institutional goals. We argue that the call-takers employ oquei in routine ways for the local management of alignment and progressivity of the reception and handling of the caller’s inquiry and the provision of the information as its response. By ‘progressivity’, we refer to ways in which the parties continue to the relevant next phases in the call, be it within a multi-unit turn, sequence or larger activities (S.-H. Lee 2011; Stivers and Robinson 2006). To proceed from one phase to the next, the parties need to negotiate their alignment with regard to sufficiently and appropriately handling the current task, which then allows them to move on to the next steps. We show how oquei is used to transition between actions and activities at various points of the calls and to elicit recipients’ uptake during the delivery of information, manifesting the call-takers’ work of managing the institutional agenda and technical resources in relation to the situated concerns of the callers. 1.1
Interaction in helplines
Helplines, a worldwide growing business, consist of a specialized phone-in service that provides support in a number of areas, such as legal and emotional counseling, technical and financial assistance, and health information (Baker, Emmison, and Firth 2005). Health helplines, in particular, have become an important means of increasing and expediting access to health care and health information. They are able to serve geographical areas and populations hitherto unreached, with the additional benefit of lowering costs (Lopriore et al. 2017). The degree of specialization in the services provided by health helplines varies from attending to general health concerns (ibid.) to specialized ones, such as pregnancy (Shaw and Kitzinger 2007),
Chapter 9. OKAY in health helpline calls in Brazil 271
cancer (Ekberg et al. 2014; Leydon et al. 2013; Woods, Drew, and Leydon 2015), and poisoning (Landqvist 2005). The nature of the help, whether specialized or not, also varies between the provision of information and support (for example, how to quit smoking) and of actual medical consultations over the phone (Lopriore et al. 2017). Health helpline services are of interest for investigating not only macro social issues (such as social inequalities) but also the microinteractional organization of the service. It is by means of talk that help is sought, given, resisted, and negotiated (Baker et al. 2005, xv). In fact, helpline calls awoke the interest of Harvey Sacks, one of the founders of conversation analysis, who conducted his very first study on calls to a suicide prevention center (Sacks 1967), data that became pivotal to his first set of UCLA lectures.2 With the advances of technology, call-centers have become increasingly assisted by computer software. Despite offering many benefits, the use of software may involve complexities and lack situated context sensitivity, thus posing “hidden labor” to call-takers. In particular, call-takers have to find ways to reconcile local, situated demands from callers with institutional agendas and software-designs that do not necessarily fit individual needs. These challenges may even jeopardize the goals of the service (Del Corona and Ostermann 2012; Murdoch et al. 2015; Nattrass et al. 2017). Operating a computer during the calls also entails attending to other, simultaneous tasks that can be autonomous or concurring with the objectives of the call (for example, searching for information), and yet involve a temporal organization different from talking (Mondada 2008). Call-takers and callers both need to cooperate in the local organization of the talk-external resources in order to successfully manage the call and the response to the inquiry posed by the caller. Investigating calls to the Brazilian health helpline is of interest not only due to the questions concerning their local organization but also because of the social importance that such a helpline assumes in a country where inequality of access to health care and information is still a major issue (Landmann-Szwarcwald and Macinko 2016; World Health Organization 2006). The choice of the helpline as the site of investigation is also supported by preliminary observations of the recurrency of oquei in data from different settings.3 We find it much more often used in our data from institutional (6.3 oqueis per hour) as compared to ordinary conversations (0.3 per hour), and over the phone (16.9 per hour) as compared to 2. See the Special Issue of Human Studies 12 (3/4), “Harvey Sacks Lectures 1964–1965” (Dec., 1989). 3. The settings analyzed consisted of the helpline (41h 29min), face-to-face medical consultations (75h 50min), phone-calls among acquaintances (3h 24min), and face-to-face ordinary conversations (25h 45min).
272 Ana Cristina Ostermann and Katariina Harjunpää
face-to-face interactions (1.3 per hour). Thus, the helpline calls allow us to examine the Brazilian Portuguese oquei in a context where it is typical and where its uses relate to the fundamental issues handled in the interactions, namely, how access to health information is locally organized in the unfolding of the calls. 1.2
Data and setting
The data consist of 308 phone calls to CIS, of which 171 include the use of oquei. The corpus consists of calls directed to women’s health (see Ostermann 2005) under the following topics: contraception, cancer, AIDS, and (other) sexually transmitted diseases.4 All callers in the dataset, with only one exception, are women, largely with low income and low educational background, with difficult access to health care and information (including access to the internet).5 The recordings were conducted between 2007 and 2009, after the research project had been reviewed and approved by the Ethics Committee.6 In order to protect the participants’ identity, all proper names that might identify the call-center or the participants (including names of cities, streets, and institutions) were replaced by fictitious ones. The overall structural organization of CIS calls is represented in Figure 1. It outlines the two main activities structured around an inquiry by a caller and a responsive information delivery by the call-taker. These can comprise several sequences and actions that sustain the larger activities (Heritage and Sorjonen 1994), such as establishing and adjusting the details of the inquiry as searchable in the database, and searching, adjusting and expanding the information provided as the response. In addition to the main activities, surveys at the end of the calls become part of the CIS agenda in the latest data. Call-takers are trained by means of oral instruction (instructional sessions and monitoring of the calls with immediate feedback) and fairly short written guidelines containing some scripted sentences. When providing information, call-takers are expected to rely on the support of written texts from a database, rather than their own knowledge. The guidelines prescribe formulaic sentences to be used in the opening and closing of the calls, and when initiating a search in the computer. In addition, they provide a list of expressions, named by the institution as “feedback of understanding”, for the call-takers to use while reading the texts to “solicit 4. Callers indicate the reason for their call at the beginning of the call, by means of selecting a topic in an automated menu. 5. Information provided by the call center management during field work. 6. The research project has been reviewed by the IRB at Universidade do Vale dos Sinos (Unisinos), Brazil, which has issued the approval certificate “Resolução 045/2005.”
Chapter 9. OKAY in health helpline calls in Brazil 273
Opening of the call Request for information Inquiry
Confirming and specifying the request Database search
Response to the inquiry
Information delivery Offer of further information on a related topic Referral Inviting new inquiries (Survey)
Closing of the call
Figure 1. Overall structural organization of CIS calls
confirmation regarding [the caller’s] understanding of the information.” Among the listed expressions are Está entendendo? ‘Do you understand?’, Certo? ‘All right?’, and Correto? ‘Correct/Right?’ (Ministério da Saúde 2014, 109). Oquei, however, is not noted on the list, perhaps due to its relatively recent appearance in Brazilian Portuguese. Including the ones listed, other expressions and particles are used in the calls for tasks that overlap with those of oquei. Next we analyze two recurrent environments of okay in the calls: (1) in transitioning between actions and activities, and (2) in eliciting recipient uptake. 2. Transitioning between actions and activities Transitional oqueis, similar to okay in English (Beach 1993, 1995a, 1995b), respond to prior turns, and project a move to next-positioned matters. The transitional oqueis analyzed here close an inserted sequence in third position and are followed by further talk by the same speaker (see also Betz and Deppermann, this volume; Mondada and Sorjonen, this volume). Thus they occur in turn-initial position, either prosodically integrated with the following talk or produced in a separate intonation contour with an intervening pause. In the continuation of the turn, the speaker launches another action, achieving the postponement of an answer to the inquiry (Section 2.1), the resumption of the main action after a search (Section 2.2), or moving to further activities (Section 2.3), such as the final survey. Oquei does not achieve transition by itself but is used in routine ways in this institutional context to manage transitions between sequences of action and activities (see also Mondada and Sorjonen, this volume; De Stefani and Mondada, this volume).
274 Ana Cristina Ostermann and Katariina Harjunpää
2.1
Postponement of the answer to the inquiry
After a caller (hereafter, CAL) has made an inquiry, call-takers (C-Ts) typically first request confirmation of CAL’s inquiry by repeating it partially or otherwise presenting their understanding of it. Answers to the inquiries, therefore, are not provided right away but involve some inserted, preparatory action. In addition, C-Ts may engage in further preparing the grounds for answering by asking for some background information before performing a search on the computer. Oquei is used in these contexts by C-Ts to receive CAL’s (dis)confirmation of the inquiry or their response to the background question, thus occurring as a sequence-closing third. By responding with oquei, C-T implies the closure of the sequence and projects a transition to another action, which then turns out to be another inserted sequence of action. This seems to have become a conventionalized practice: the turn involving oquei achieves postponement of what would be the next action in this sequential environment – to answer CAL’s inquiry. Excerpt 1 shows two instances in which C-T moves into another action before answering the inquiry. Excerpt 1. O que que a senhora já sabe ‘What do you already know’ (DISK240707Paula2 0:10), 2007, Brazilian Portuguese, telephone, helpline interaction 01
CAL:
eu queria sabê quais são I’d like to know what are
02
os métodos anticon:cepcio↓nais. the contraceptive methods.
03
(0.5)
04 05 06 07 08
C-T:
CAL:
C-T:
09 10
(0.2) sim. yes
(0.5)
>o↓queio ↑que que a↑ senhora< já sabe (.) uh what what do you already know
sobre esse assu:nto. about this topic
11
CAL:
12
C-T:
13
CAL:
14
quais são os ↑métodos ↑anticoncepcionais; what are the contraceptive methods
(1.2)
°e:u?° me i:[sso. ] right ((“that”)) [°(-)°] .h >eu só conheço a camipor favor< a↑gua:rde enquanto then please wait while
eu realizo uma pes↑quisa [>>vô podêo↓queipor favor< agua:rde. please wait
278 Ana Cristina Ostermann and Katariina Harjunpää
04 05 06
C-T:
07
CAL:
08
C-T:
(14.6)
>cêstá< ou↑vin↓do are you listening (0.5)
tô. ((I)) am >>°°o↓quei°°>°°o↓quei°°assim< é:: (2.0) homens também ↑dá (.) câncer de ma:ma? like this u::h (2.0) men too get (.) breast cancer (1.2)
por favor a↑gua::rde. please wait (48.5)
tá me ouvi:ndo? can ((you)) hear me (0.7) si:m yes
então o câncer de mama não é exclusi:vo de so breast câncer is not exclusive to mulhe:↓res (.) podendo mesmo sendo ra:ro apresentar women (.) being able even though rare to occur
280 Ana Cristina Ostermann and Katariina Harjunpää
14
em ↑homens. in men
15
(2.6)
16
C-T:
17
CAL:
18
C-T:
19
CAL:
não só i:sso. no just that
20
C-T:
o↓que:i. >gostaria< de ouvir as fo:rmas would ((you)) like to hear about ways
21 22
ce:rto? right ce:rto; right
teria alguma outra dú:vida? would ((you)) have any other question
de prevenção dos cânceres em geral? of preventing cancers in general
23
CAL:
24
C-T:
(0.4) sim yes
por favor agua:rde enquanto eu realizo uma ↑pesqui:sa. please wait until I perform a search
CAL’s new question as to whether men can also have breast cancer opens up another cycle of a database search, a response to the inquiry, and an invitation to make further inquiries (lines 7–18), which is declined by CAL (line 19). C-T makes use of the closing-implicative environment to initiate a further activity. She receives CAL’s disconfirmation of having further questions with o↓que:i., followed by an offer that specifically orients to one of the institution’s missions: the prevention of illnesses. In the next example, oquei precedes an offer of further service. Excerpt 5. Posso ajudar em algo mais ‘Can I help with anything else’ (DISK160707Luana 3:11), 2007, Brazilian Portuguese, telephone, helpline 01
C-T:
02
↑mais pró:ximo à sua resi↑dência. closest to your residence
03
CAL:
04
C-T:
05
.h você já ↑sabe onde tem um posto de saúde .h do you already know where is the health center
06
CAL:
07
C-T:
sim. yes
↑o↓quei >posso ajudar< em algo ↑mais; can ((I)) help with anything else
(1.0)
não ob↑rigada (filha). no thank you (dear) >tudo bem< o ministério da saúde que agra↑dece a (sua) all right the ministry of health thanks for (your)
Chapter 9. OKAY in health helpline calls in Brazil 281
08
ligação ligue sempre que necessário tenha uma ↑boa tarde. call call always when necessary have a good afternoon
09
CAL:
10
C-T:
igualme:n[te. likewise [obri↑ga:da. thank you
After C-T has provided the information requested by CAL on menstruation and directed her to visit a public health unit for more specific information (not shown here), she checks whether CAL knows where the nearest health unit is (lines 1–2) – aligning with the institution’s mission “to refer”. C-T receives CAL’s confirmation that she already has that information (thus preempting the need to provide it) with a turn-initial, rise-fall intonation ↑o↓quei (line 4), and then moves to offer further help, within the same prosodic unit. Oquei precedes the offers in the sequential context of receiving (dis)confirmation (Excerpt 4, lines 18–20; Excerpt 5, lines 1–4), whereas we do not find it preceding offers that come after understanding-checking sequences, such as line 3 in Excerpt 4 (after ficou claro – sim), or line 18 (after certo – certo). Thus oquei is responsive to a more substantial display of the sufficiency of prior information in a transitional environment. These oquei turns open up the possibility of either moving into closing or to further activities. Even though oquei is here used with an action typically associated with movements towards closing (i.e., an offer of further help in a service-call), it does not by itself work as a pre-closing token (Schegloff and Sacks 1973). Instead, it is followed by C-T’s own talk that calls for at least one more turn by CAL, in which she either confirms or (dis)confirms the need of a particular service, before an actual closing becomes relevant. The cases in this section show how C-Ts use oquei as a device in implementing the institutional agenda, organizing related activities and managing the use of technical and textual resources in the unfolding of the calls. 3. Eliciting recipient’s uptake We now move to examining the call-takers’ delivery of information to the callers and their use of turn-final oquei to elicit uptake from the caller. In these sequences, the participants negotiate the unfolding of an activity, for instance, in terms of whether (a segment of) the response to the inquiry is understood, accepted and sufficient for the current purposes. C-T’s elicitation of uptake from CAL can concern CAL’s general following and understanding of an ongoing explanation (Section 3.1), the acknowledgement of emphasized details in the explanation (Section 3.2), acquiescence with events in the immediate or distant future (Section 3.3), and joint closing
282 Ana Cristina Ostermann and Katariina Harjunpää
of the inquiries (Section 3.4). We also consider a set of deviant cases in which an apparent eliciting oquei is, instead, followed by C-T’s own elaboration of their prior talk (Section 3.5). 3.1
Following and understanding an ongoing explanation
It is the call-taker’s task to search for and provide information from the database in a way that responds to the caller’s inquiry. However, according to the guidelines, and as attested in the data, C-Ts maintain a pre-set order in the delivery of information: first they give a general description of the issue (e.g., AIDS and menstruation), then a list of symptoms or characteristics, followed by forms of health promotion or disease prevention, and finally treatment. Due to this routine, C-Ts may, instead of directly providing the information requested, give lengthy explanations on related matters. During these extended actions, which consist of reading aloud texts from the computer, C-Ts use oquei to invite displays from CAL that she follows and understands the information so far, after which they can continue. In Excerpt 6, CAL has asked whether breast cancer affects bones or only the breast tissue. C-T has responded that she will first finish the ongoing explanation of the forms of prevention, and only then answer this question. The excerpt starts when C-T re-engages with CAL after a new search in the database and launches the explanation. Excerpt 6. Câncer de mama ‘Breast cancer’ (DISK170707Celia 6:25), 2007, Brazilian Portuguese, telephone, helpline 01
C-T:
02
CAL:
03
C-T:
>a senhora está< me ou↑vin[↓do. are you hearing me
[ouvin::ndo; yes (‘hearing’) então senhora o câncer de mama (.) que é provavelmente so mam breast cancer (.) that is probably
04
o mais temido pelas mulhe:res devido à sua alta the most feared by women because of its high
05
frequê:n↑cia .hh e sobretudo pelos seus frequency .hh and especially for its
06
efeitos↑ psicológicos que afetam a- a percepção da psychological effects that affect the- the perception of
07
sexualidade? .hh .tsk ↑e da própria imagem corporal. sexuality .hh .tsk and of one’s own body image
08
corre:to.= right
09
CAL:
=corre::to; right
Chapter 9. OKAY in health helpline calls in Brazil 283
10
C-T:
ele é relativamente ra:ro (.) antes dos trinta e cinco it is relatively rare (.) before thirty five
11
anos de ida:de .hh ↑mas acima desta faixa etária years of age .hh but above that age range
12
>a sua incidência< cresce ↑rápido e progressiva↓mente. its occurrence increases rapidly and progressively
13
(0.3)
14
15
16
C-T:
CAL:
C-T:
!t oque::i?=
=oquei::;
no brasil o câncer de mama é o que ↑mais causa morte in Brazil breast cancer is what causes most deaths
17
entre as mu↓lheres. (0.6) .th de acordo com a estima>tiva among women (0.6) .th according to an estimate
18
da incidência do câncer no brasil para dois mil e< seis of cancer occurrences in Brazil from two thousand and six
19
.hh o câncer de mama será o segundo mais incide:nte .hh breast cancer is the second most recurrent
20
com quarenta e nove mil novecentos e tri:n↑ta ca::sos. with forty nine thousand nine hundread and three cases
21 22
23
24
C-T:
CAL:
C-T:
(0.3)
o↓↑que:i:?=
=oque:i:;
.hh os sintomas do câncer de mama são o nódulo ou .hh the symptoms of breast cancer are a nodule or
25
o tumor no se:io. .hh acompanhado ou não >de dor< tumor in the breast .hh accompanied or not by pain
26
ou seja senhora .hhh o câncer >ele vai acontecê< (.) ↑na that is, mam .hhh the cancer it will occur (.) in the
27
ma:ma; (.) >pra senhora percebê< na ma:ma não ↑tem breast (.) for you to understand in the breast there’s no
28
↓o:sso. (.) >então< é a ma:↑ma que vai sê afe↓tada. bone (.) so it is the breast that is going to be affected
29
CAL:
ce::rto; right
C-T provides a general description of breast cancer (lines 1–25) and only then the specific piece of information requested by CAL (lines 26–28). During the general description that she reads aloud from the computer, C-T stops at three points to elicit uptake by the recipient: corre:to. (line 8), and oquei (lines 14 and 22). CAL responds to all of them by immediately echoing the tokens (lines 9, 15, 23). C-T’s tokens are prosodically different, with corre:to. produced with falling intonation and oque::i? and o↓↑que:i:? with a final rise, whereas the prosodic design of CAL’s responses is rather uniform, with a stretch on the stressed syllable and falling
284 Ana Cristina Ostermann and Katariina Harjunpää
intonation. The echo responses give the impression of being almost automated, suggesting that CAL is confirming being present and listening, but perhaps actually not claiming understanding. On the one hand, C-T’s elicitation of CAL’s uptake between read-aloud segments of information might be interpreted in light of them following the institutional guidelines in using understanding-checks. On the other hand, C-Ts may coordinate reading aloud by pausing every now and then to re-establish some “conversational contact”. In fact, we suspect that the positioning of the elicitation of uptake within the C-T’s talk may reflect the division of the source text into paragraphs. In the current example, the units of explanation that are marked off by the elicitation of uptake constitute topical units: the effects of breast cancer (lines 3–7), age range (lines 10–12), statistical frequency (lines 16–20), and symptoms (lines 24–25). In any case, the elicitation of uptake during reading aloud succeeds in segmenting the delivery of information and moving between its parts, thus allowing joint progress in the activity that is underway. 3.2
Uptake of emphasized information
Instead of segmenting an explanation in a way that resembles textual units, the call-taker’s elicitation of uptake can also emerge in a more situated manner during the unfolding of a larger explanation. In such cases, C-T can establish prominence or contrast within the explanation in order to respond to or anticipate the fact that CAL might have trouble understanding the talk at hand. One way of building the emphasis is through an insertion that expresses a contrast with prior items, often including negation, followed by an oquei that elicits uptake. The analysis indicates that variation in the prosodic integration of oquei with the preceding talk may contribute to differentiating the actions that the turns implement. The first extract includes two instances of this use of oquei, the first one as part of the information delivery itself (line 20), and the second one in a clarification requested by CAL (line 40). Excerpt 7. Febre ‘Fever’ (DISK020909FaniAIDS 00:51), 2009, Brazilian Portuguese, telephone, helpline 01 02
C-T:
03
CAL:
04
C-T:
consegue me ouvir? can you hear me (0.6)
con↑si:go, yes ((‘I can’)) ce:rto; >senhora< en↑tão os sinto:mas da infecção da right mam so the symptoms of the infection of
Chapter 9. OKAY in health helpline calls in Brazil 285
05
(0.6) o- >assim< o vírus do hiv o vírus da da AIDS (0.6) t- or the HIV virus the virus of of
06
i↑munodeficiência humana é o vírus que cau↑sa a a:ids. human immunodefiency is the virus that causes AIDS
07
(0.6)
08 09 10 11
certo< [.hh] right CAL:
C-T:
12 13 14 15
[oi?] huh (0.7)
>a senhora< (.) o vírus do hiv senhora (.) é o ví:rus you- (.) the HIV virus mam (.) is the virus que cau↑sa a↑ aids>. (0.2) ↑c[erto;] that causes AIDS (0.2) right ((‘sure’))
CAL:
C-T:
[°tá°.] all right (0.3)
>então< .h a pessoa que apresenta a aids .hhh so .h the person who has AIDS .hhh
16
o:u a infecção ↑pelo vírus do hiv !m ela p- apresenta or the infection by the HIV virus mh she h- has
17
(.) suor inte↑nso durante a noi↑te .h certo po↑de:ndo intense sweating during the night .h right that can
18
estar acompanhado ↑ou não de feb↑re, .thh (1.0) e essa be accompanied by fever or not .thh (1.0) and this
19
fe:bre senhora ↑não é uma febre muito↑ a:lta .h mas fever mam is not a very high fever .h but
20
é uma febre que aparece ↑todos os dias_ (.) >oquei?< ((it)) is a fever that appears every day (.)
21 22
CAL:
23
C-T:
(0.4) ã. ahn
.h é comum uma sensação ↑constante do cansaço .h .h ((it)) is common to have a feeling of fatigue .h
24
mesmo que a pessoa esteja em repouso, (0.8) a pessoa even if the person is resting (0.8) the person
25
pode apresentá .hh diarréia .h por mais ↑de um mês (.) can have .hh diarrhea .h for over a month (.)
26
e a presença de caroços embaixo do bra:ço no pescoço and the presence of lumps below the arm in the neck
27
na virilha_ (.) c↑erto; in the groin (.) right
28
(0.4)
286 Ana Cristina Ostermann and Katariina Harjunpää
29
CAL:
30 31
nas costa (.) tam↑bé:m ou só: ou- tem o::: nas costa on the back (.) too or only or is it uhm on the back >tam↑bém pode aparecê