1. Introduction
In Finnish and Danish interactions, people sometimes make use of utterances spoken on the inbreath, that is, ingressive speech. This device is predominantly used in responses. Excerpt (1) is an example of this from a telephone conversation in Finnish.Footnote 1 It is from a phone call between two dog breeders (mother and daughter) discussing a time for a possible coming together to buy a coat to Sini, the daughter (data not shown). In this sequence, both parties resort to the use of ingressive response particles at the same time (lines 15–16). A dot in front of a word or a sound indicates that it is spoken on the inbreath.Footnote 2
There are two obstacles before the two women may find a joint time to go buy the coat. Just prior to this excerpt, Irja (mother) has told Sini that she needs to go to a place to trim her puppies. This is what the inquiry in line 1 is about. In line 9, another possible obstacle for the coming together is mentioned: The plan may also depend on a common acquaintance (Marja) coming to Irja’s place to examine her puppies.
The sequence consisting of two question–answer pairs (Schegloff Reference Schegloff2007) is first responded to by Sini (line 13) with a change-of-state token ↑a:ha (see Koivisto Reference Koivisto2016). Kovisto shows that this specific change-of-state token receipts information that redirects the talk and has consequences for the participants’s projects. In this case, the information that Marja will come ‘in the morning’ deletes the potential obstacle that Sini raised in her negative question (line 9), bringing the situation back to where it was left in lines 2–5, that there is no ‘ready plan’. After Irja’s equivocal increment to her earlier turn, vissii ‘perhaps’, the parties jointly bring the sequence to an end using (a variant of) the ingressive particle.joo (lines 15–16). Sini has thus performed two jobs: Her ↑a:ha (line 13) receipts the answer to her question and displays that this deletes the potential obstacle raised by that question. Her ingressive t.joo (line 15) goes on to indicate that this means there is nothing more to add about this. Irja’s very similar ingressive .tjoo (line 16) shows the same orientation. And, complementing this, Sini utters a double joo with a ‘no news’ prosody (Ogden, Hakulinen & Tainio Reference Ogden, Hakulinen, Tainio, Elizabeth and Cecilia2004), implicating ‘there is nothing new in these arrangements’ (line 17). With these devices, the speakers indicate that they share the understanding of the situation: ‘this is the way things tend to go; there is nothing to add’. Here, the use of the ingressive response particles is almost iconic: both parties are pulling out from the topic, and having said enough, they pursue closing the topic.
Excerpt (2) shows a similar orientation. It is from a video-recorded interaction in an office that helps Danes find jobs abroad. Alexa is the job seeker and Bodil the clerk working at the office. Alexa has told Bodil that she wants to apply for a job in Switzerland, and that she has a Swiss citizenship, and Bodil has informed Alexa that in that case she can just go there and apply for a job.Footnote 3
In line 1, Alexa starts formulating her understanding of what Bodil’s information means for her situation. Such upshot formulations (Heritage & Watson Reference Heritage, Rod Watson and George1979) invite the recipient to confirm. Alexa pauses at a point where her turn is syntactically incomplete, but the use of the word bare ‘only; just’ conveys the implication that the application should be easy. Bodil enters in line 3 after 0.3 seconds with a suggestion for a revised completion of the turn, recasting it as ‘apply for the jobs like a Swiss’. Alexa repeats the suggested phrase in line 4, and Bodil confirms with an emphatic ‘yes’ in line 5, which is receipted by Alexa in line 6, while both are nodding. Alexa adds an ‘okay’ in line 8, which emphasizes that she has received the necessary information (Steensig & Sørensen Reference Steensig and Søren2019). In overlap with this, Bodil agrees with lige præcis ‘precisely’ (line 9), which in this case indicates that Alexa’s formulation is a correct version of something that the speaker of the ‘precisely’ turn has primary rights to know (Olesen Reference Olesen2019).
At this point, both parties have confirmed and agreed several times with the assertion that Alexa can ‘apply for the jobs like a Swiss’, and in line 10, Alexa receipts once again with an ingressive .ja ‘.yes1’.Footnote 4 The parties have thus displayed to each other in several ways that a common understanding has been reached and that the sequence can be closed. This jointly achieved and displayed closing of the sequence is further confirmed with an ingressive .ja.
Excerpts (1) and (2) evidence a usage of the ingressive response particles that is recurrent in our data. For now, we can gloss this usage in the following way: The ingressive response particles confirm or register a prior action and display that participants see the prior activity as potentially complete, treating the understanding that has been reached as sufficient. The registering and confirming actions are found equally with the ‘normal’ (or egressive) counterparts of our ingressive response particles, the Finnish joo and Danish ja.Footnote 5 However, whereas we find the egressive particles in a wide array of contexts, responding to all kinds of action, the ingressive ones have a more limited and specialized usage. In this article, we explore the usage of the ingressive response particles in Finnish and Danish talk-in-interaction and try to reveal the logic behind it. In excerpts (1) and (2), we saw the particles being used after sequences that were already potentially complete, displaying a shared understanding and (re-)confirming the relevance of closing the sequence. In Section 3.1, we will proceed to ingressive response particles used as receipts in third position after question–answer pairs. Then we will move to instances in which the particles come in second position after questions and other first pair part actions (Section 3.2). Next we will investigate instances where the particles occur after assessments that do not seek affiliation (Section 3.3), leading up to analyses of ingressive particles occurring after affiliation-seeking utterances (Section 3.4). In our concluding discussion, we will sketch the logic behind the use of these particles and discuss further perspectives.
Before we do this, however, we will take a brief look at some general facts and myths about ingressive responses, and we will present our method and data.
2. Background
2.1 General facts and myths about ingressive responses
Ingressive speech occurs when people produce an utterance using the normal articulators in the vocal and nasal cavities while sucking air into their lungs, instead of using the airstream coming out from the lungs. An articulatorily accurate term for the former is, therefore, ‘ingressive pulmonic speech’ (Eklund Reference Eklund2008). According to Eklund, this kind of speech is found to occur in diverse functions in speech communities all over the planet. Ingressive speech as a general phenomenon (not limited to responses) has been interpreted as expressing things like conversational intimacy (Hill & Zepeda Reference Hill and Ofelia1999); sympathy, commiseration (Eklund Reference Eklund2008); reluctant compliance with exigent wishes and assertions (Laver Reference Laver1994:169); strong emotion and impatience (Léon Reference Léon1992; Finns running out of breath towards the end of a long utterance (Laver Reference Laver1994:170); stylistic variants of normal (egressive) speech (Ohala Reference Ohala and MacNeilage1983); and imparting bad news or news in confidence (Thom Reference Thom2005).
In this paper, we limit ourselves to considering ingressive responses of the type exemplified by excerpts (1) and (2) above. As we shall see, for these responses, the above descriptions do not seem to be relevant. This more restricted use of ingressive speech seems to be an areal phenomenon, documented in a North European area. Having gathered data sets from Finnish and Danish, we have seen and listened to recorded instances from interactions in Estonian, French, Icelandic, North German and Swedish, and we have heard it used in Faroese as well. Peters (Reference Peters1981, quoted in Stølen Reference Stølen1995) reports on its use in Scandinavian immigrant populations in Maine, USA, and Thom (Reference Thom2005) in Gaelic and its heritage across the Atlantic. John Heritage (p.c.) reports its use in English in Northumbria, where it is considered a Scandinavian influence (Hakulinen Reference Hakulinen, Anne-Marie, Hanna, Pirkko, Anne-Marie and Helena1993:49).
Some earlier studies of ingressive responses see them as a phenomenon used primarily by women and primarily in informal conversations (Peters Reference Peters1981, Stølen Reference Stølen1995). It is reported to mark femininity (Pitschmann Reference Pitschmann1987, Stølen Reference Stølen1995), non-involvement and passivity (Pitschmann Reference Pitschmann1987), but Stølen (Reference Stølen1995) also finds instances where a Danish ingressive response seems to mark emotional involvement. Hakulinen (Reference Hakulinen, Anne-Marie, Hanna, Pirkko, Anne-Marie and Helena1993) sees ingressive responses (and ingressive speech more generally) in Finnish interaction as markers of ‘intimacy’.
Hakulinen’s paper is the only one that takes up the sequential context in which the ingressive responses occur. It identifies two main environments: ingressive responses can signal a boundary at the end of a topic or as a sequence closer after a second pair part (Schegloff & Sacks Reference Schegloff and Harvey1973; Hakulinen Reference Hakulinen, Anne-Marie, Hanna, Pirkko, Anne-Marie and Helena1993:55–60), and they can be used as an ‘accompaniment’, with which a secondary recipient in a multi-party conversation can mark that she is a listener, not a contributor (Hakulinen Reference Hakulinen, Anne-Marie, Hanna, Pirkko, Anne-Marie and Helena1993:60–61).
Our data, from naturally occurring interactions in Finnish and Danish (see below), disconfirm that the use of ingressive responses should be primarily done by women. We have a substantial number of instances uttered by men. Also, we have found young, middle-aged and elderly speakers using ingressive responses. And, finally, our data show that the phenomenon occurs in both private interactions among people who know each other (which was the primary or sole environment reported in earlier studies) and in ‘institutional’ interactions (medical and social lay–professional encounters) between people who have had no prior contact with each other.
There may, of course, be sociolinguistic restrictions on the use of ingressive responses, and they are certainly worth investigating. Our approach, however, has been to discover the interactional functions that ingressive responses perform in Finnish and Danish interactions.
2.2 Delimiting and representing the phenomenon
In this section, we describe the particles we are investigating, and we compare them to the corresponding egressive particles.
The ingressive response particles we are investigating are variants of the corresponding egressive ones. They are, thus, clearly words that belong to the inventory of the languages even though they employ a phonation that is not used very much elsewhere in these languages. Keevallik & Ogden (Reference Keevallik and Richard2020) examine vocalizations ‘on the margins of language’ (such as sniffs, lip-smacks, grunts, moans, sighs, whistles, and clicks) and they operate with a ‘vocalization continuum’, between vocalizations on the ‘word’ end of that continuum and on the more somatic end on the other. The ingressive particles clearly belong to the ‘word’ end of such a continuum.
The Finnish response system is relatively complex. What can roughly be translated with ‘yes’ in English, can be expressed with a variety of particles, joo or juu, nii(n), mm and kyllä, or, primarily as responses to yes–no questions, by means of a repetition of the verb. The particles have different epistemic and sequential functions (Sorjonen Reference Sorjonen2001a, b), and in our analyses below we shall address those differences when necessary. Negation in Finnish is done with a negative auxiliary ei, conjugated in person, and, under certain conditions, it is used as a particle (for accounts of Finnish responses, see Sorjonen Reference Sorjonen, Elinor, Schegloff and Thompson1996, Hakulinen Reference Hakulinen2001, Hakulinen et al. Reference Hakulinen, Vilkuna, Korhonen, Koivisto, Heinonen and Alho2004). In our data and in our personal experience, the response particle .joo is by far the most frequent one used as an ingressive response. At times it is reduplicated, as .joo .joo, or it can have the form of a regional variant .juu. In addition, we find uses of .mm, .niin and .kyllä as well as of the negation verb/particle .ei. Occasionally we find the verb .on ‘is’, when repeated from the previous turn.
In the comprehensive grammar of Finnish (Hakulinen et al. Reference Hakulinen, Vilkuna, Korhonen, Koivisto, Heinonen and Alho2004), the above-mentioned egressive particles belong to a subgroup of ‘dialogue particles’ that can display confirmation, acceptance, agreement, etc. (§798), and it is mentioned that joo and niin can also be spoken on the inbreath (§800).
As pointed out in Hakulinen (Reference Hakulinen, Anne-Marie, Hanna, Pirkko, Anne-Marie and Helena1993, Reference Hakulinen2010), some speakers have been found to use ingressive speech in other utterance types as well, e.g. in exclamations, such as .voi .kauhee ‘oh how awful’. In our paper, however, we are only investigating the single word responses, mainly because these are found in both the Finnish and the Danish data.
The Danish response system is less complex than the Finnish one. The three particles that we find as ingressive responses are ja ‘yes1’, jo ‘yes2’ and nej ‘no’. After utterances with positive polarity, ‘yes1’ and ‘no’ are used as in English, but after utterances with negative polarity, nej ‘no’ confirms, and jo ‘yes2’ disconfirms (Heinemann Reference Heinemann, Auli and Margret2005, Reference Heinemann2010, Reference Heinemann2015; Hansen & Heltoft Reference Hansen and Lars2011:1116–1121). The response word jo ‘yes2’ can also confirm assumptions, if it occurs after tags and tag questions with negative polarity, such as gør han ikk ‘doesn’t he’ (Heinemann Reference Heinemann, Auli and Margret2005, Reference Heinemann2015). This is the only place we have found the ingressive .jo. In our data, ingressive .nej and .jo are only found doing confirmation.
Hansen & Heltoft (Reference Hansen and Lars2011:1113) categorize the egressive particles as ‘response words’, a subgroup of ‘neutral interjections’. The grammar of Danish talk-in-interaction (samtalegrammatik 2020) and the dictionary of Danish spoken language (Ordbog over dansk talesprog n.d.) contain descriptions of various uses of the egressive particles. The latter mentions that ja can be ingressive, but none of the grammatical or lexical descriptions deal with the uses or meanings of the ingressive particles.
As mentioned above, we represent the ingressive versions of the response particles as we would write their non-ingressive counterparts, but with a dot in front of them. An ingressive word sounds ‘breathy’, mainly because the ingressive airstream is not sufficient to generate regular vibrations in the vocal cords (Eklund Reference Eklund2008).
2.3 Methodology and data
Our methodology relies on conversation analysis (see Heritage Reference Heritage and David2010, Sidnell & Stivers Reference Sidnell and Stivers2013) and interactional linguistics (Couper-Kuhlen & Selting Reference Couper-Kuhlen and Margret2018). This involves using data from naturally occurring interactions (Mondada Reference Mondada2013) and studying extracts from these data in a way that describes and uncovers the logic in how interactants use and understand each other’s actions. For this purpose, transcripts with a high level of detail are produced and used as a window onto the way people do things in interaction. The transcripts emphasize the precision concerning the timing and production of sounds, and, if necessary, also give representations of physical actions (Hepburn & Bolden Reference Hepburn and Bolden2013). In analyzing extracts, the analyst pays attention to the exact way actions are performed and uses the co-interactants’ reactions as a key to investigating what a conversational object is doing on each occasion. Building on the detailed descriptions and analyses of particular cases, generalizations and conclusions are made in a fashion that attempts to stay true to the insights gained from the analyses of the individual cases.
More concretely, the process of collecting and analyzing the data has been as follows: The second author had a collection of instances from transcribed audio and video recordings in Finnish, used for her first publication about this phenomenon (Hakulinen Reference Hakulinen, Anne-Marie, Hanna, Pirkko, Anne-Marie and Helena1993). The first author made a collection of similar instances from transcribed audio and video recordings in Danish. When we started working together comparing our data, we supplemented our collections with more instances, so that we had a few hundred cases from each language. Our first observation was that the uses we found in the two languages were very similar. However, some uses were poorly represented in the Danish collection even though the first author had a strong intuition that they could be found. When the third author joined the project, she brought a large collection of instances from video recordings of institutional data in Danish, and in this collection, we found Danish instances of the uses that were earlier poorly represented in the Danish corpus.
The collections of instances that form the basis of our investigation are thus a so called ‘opportunity sample’. We simply took all the instances of ingressive responses that we found in our corpora. The corpora, however, had not been collected with the purpose of investigating this phenomenon. This means, of course, that our collections of instances make no claim of representativity, other than the important one that these cases exist and in them, the ingressive responses are used in the ways that we have found. We, therefore, refrain from making distributional claims, except for the obvious ones, for instance, that some uses were recurrent in our data.
The data come from both everyday conversation with relatives and friends, and from institutional settings, among them, patient education sessions in chronic disease self-management, counselling data, emergency calls, conversations at the hairdresser’s, interactions between health care professionals and between such professionals and patients.
All data were anonymized to secure that participants cannot be recognized. Data have been collected, stored and treated in accordance with data protection legislation.
3. The functions of ingressive response particles
In most cases, ingressive responses are used by one party at a time, in the way shown in excerpt (2) above, but, occasionally even with both speakers as in excerpt (1). In what follows, we are going to discuss four interactional environments in which the ingressive responses occur: after answers to questions, as responses to questions, after assessments that do not seek affiliation, and after affiliation-seeking utterances.
3.1 Ingressive responses as receipts after answers to questions
After a question–answer sequence, or other adjacency pairs (Schegloff Reference Schegloff2007), it is relevant for the initiator of the sequence to display how s/he understands the response. It can be done with a receipt in third position. This is one of the positions where we find ingressive response particles.
The Finnish excerpt (3) below comes from a ‘pre-party’ gathering: two middle-aged couples have gathered to drink mulled wine before moving on to a pre-Christmas celebration in the vicinity. The talk drifts from one domestic topic to the next. Previously, a recent acquisition, the family dog, which is lying under the table, has been discussed, when the hostess Jaana is reminded of something, and turns her head towards her husband Jaska.
Jaana marks her turn with the word muute ‘by the way; incidentally’, thereby showing that her question is not continuing the dog topic. Furthermore, she specifically marks the object of the clause, ne, pussit ‘them, bags’ as something that is known to the recipient. Finnish does not have a definite article, but the use of the pronoun ne ‘them’ here underlines the fact that this ought to be known to Jaska. Thereby, Jaana seems to assume that the recipient knows the issue, and, by choosing the positive interrogative, she also builds in an expectation that Jaska has carried out the action. Responding to her query with a juu (line 5), Jaska confirms having performed the activity expected of him. This is receipted by Jaana with an ingressive .juu (line 6). By using the ingressive particle, Jaana indicates, as she did with her muute in line 1, that this was not meant to be the next topic but that it is a parenthetical sequence that need not be expanded. Consequently, one of the visitors, Mirja, brings up a next topic, the imminent pre-Christmas celebration (line 9).
Excerpt (4) below comes from the same interaction as excerpt (2) above, but later in the interaction. At this point, Bodil has recommended that Alexa gets in touch with the Swiss embassy in Denmark. The reference to ‘Aalborg’ in line 1 is (an anonymized version of) the city where the interaction takes place.
In lines 1–3 Alexa requests confirmation of her assumption that the Swiss embassy does not have an office in the city where they are. Bodil confirms with a nej ‘no’ in line 4, which is immediately receipted by Alexa in line 5 with an ingressive .nej ‘no’. The fact that the negative particle is used both to confirm and (as an ingressive) to receipt, is in line with the general rule in Danish talk-in-interaction that negative utterances are confirmed with the negative particle (Heinemann Reference Heinemann, Auli and Margret2005, Reference Heinemann2015), and the nod that Alexa makes at the same time as her ingressive receipt, seems to express that her assumption has been confirmed. After the receipt, which proposes a closing of the sequence, Bodil adds a slight modification of her epistemic stance in line 6, while Alexa, in overlap, adds something that could be a consequence of the fact that there is no office in the hometown. As was the case in excerpt (3), a questioner receives the expected answer to a question with a matching ingressive response.
In general, the ingressive response particles occurring after question–answer sequences indicate that the answer confirmed the assumption displayed in the question. In that way, it contrasts with the change-of-state responses that a questioner can also use after an answer to a question, indicating that the answer informed her and changed her state of knowledge (Heritage Reference Heritage1984, Heinemann & Koivisto Reference Heinemann and Aino2016). Ingressive responses in this same position index that the answer was to be expected, that the assumptions built into the answer were confirmed, and that there is no need for expansion.
3.2 Ingressive particles as responses to questions
In excerpts (3) and (4), the ingressive .juu and .nej occurred as receipts in third position after a question–answer adjacency pair. This position has been claimed to be the prototypical one for the use of the ingressive response particles (Hakulinen Reference Hakulinen2010), and it is indeed a position in which ingressive responses are frequently found in the present data. However, we also find ingressive responses as second pair parts, as answers to questions (Schegloff Reference Schegloff2007). Next, we will consider instances of this usage.
Excerpt (5) is taken from an encounter at a hairdresser’s. In line 1, the talk is resumed after a long silence. The hairdresser (KA) and his client (AS) are old acquaintances from years back.
The question posed by the hairdresser is a polar interrogative, but with a biased implication, which could be paraphrased as ‘You just go on living in Vinhu, don’t you’. The default polar interrogative in Finnish is verb initial, which in this case would be Asut-ko sä vielä Vinhussa ‘Live-Q you still in Vinhu’. Such a question could receive either of the two default positive answers, a verb repeat asun ‘live.1sg’, or a particle answer joo (Hakulinen et al. Reference Hakulinen, Vilkuna, Korhonen, Koivisto, Heinonen and Alho2004:993, 1147–1148). Instead, the question begins with the temporal adverb vielä ‘still’, at the end of which the question clitic -kO is attached. The unmarked positive answers to this type of interrogatives would be a repeat of the questioned item, vielä, or one of the particles niin or joo (Sorjonen Reference Sorjonen2001a:Chapter 4). However, the client answers the question with an ingressive .joo. On the basis of what we have said above, the response is hearable as registering or confirming information that is already shared. The choice of this response variant is accounted for in the next turn of the client (line 4) – an account explicating her reasons, which can be paraphrased as ‘there is no other place for me to go to’. Her line of thinking is subsequently supported by the hairdresser in line 5, who in fact had assumed this state of affairs to begin with, choosing a marked form of the polar question. The ingressive answer thus treats the information it confirms as obvious and sufficient.
Another context for the use of ingressive responses is after requests for clarification. Excerpt (6) comes from a chronic disease self-management training session in a Danish university hospital. A patient (P) with a cardio-vascular condition is instructed by a nurse (N) in how to use certain medical devices that will allow him to monitor and treat his blood coagulation (INR) values at home. Prior to this excerpt, N has been instructing P in how to set up and use a handheld INR-monitor, which includes how to check, enter and adjust information in the settings of the device:
In lines 1–3, N addresses special circumstances that arise from the need to occasionally change batteries on the INR-monitor. Specifically, she informs P that he will need to reset the device. She does not explain precisely how the reset operation is accomplished, but points to the manual (line 3) and informs P that the procedure is described therein.
Gazing towards the manual, P makes a series of minimal nods (line 4) but does not produce a verbal response to either claim or demonstrate an understanding of N’s instructions. In lines 5–6, N treats P’s minimal nods as an insufficient response to her instructions and she pursues a new response from P by picking up on her instructions (from lines 1–3) and producing an account for simply referring P to the manual. She explains in particular that if she did provide more details about how the reset operation is performed, P would likely have forgotten them by the time a battery change next became relevant.
During N’s account, P has put out the tip of his tongue and positions it against his upper lip in a manner conveying deep concentration (line 6). This action can be seen as supporting the suspicion that he did not indeed quite follow N’s first version of the instructions (from lines 1–3). Upon completion of the account, P releases the tongue gesture with a smack (first sound in line 7) and produces a request for clarification, ‘you mean how one makes changes in that one’ and points to the INR-monitor. This displays an understanding of what the nurse refers to with her de:t ‘it’ in line 6, and with her three first lines, that is, what N’s instructions amount to and, as such, calls for a confirmation. The nurse does provide a confirming answer in line 8. The confirmation is presented in the form of an up-initial head-nod and an ingressive ‘yes1’.Footnote 6 The ingressive response particle seems to convey a sense of obviousness (similar to the one noted in excerpt (5) above).
Though N now treats the matter as established, P continues to display hesitation. After a 1.4-second silence (line 9), during which P continues to gaze silently at the INR-monitor, P nods and produces a particle m:m¿ (line 10). While these actions jointly serve as sequence closing (Schegloff Reference Schegloff2007), receipting N’s ingressive confirmation (from line 8), P seems to convey – through the delay – that N’s instructions and their possible future implications are (still) not entirely clear to him, although they were treated as in no need of further explaining by N.
3.3 Ingressive particles after alignment-seeking assessments
Ingressive particles also occur after assessments in both Finnish and Danish. According to Pomerantz (Reference Pomerantz1984), first assessments of items that both parties have epistemic access to, call for agreeing responses. The question for our examination is, thus, whether an ingressive response particle can display agreement in such situations. In our analyses of these cases, we have found it useful to distinguish between affiliation and alignment. We use the term affiliation in the sense introduced by Stivers (Reference Stivers2008) to indicate that a recipient ‘displays support of and endorses the teller’s conveyed stance’ (Stivers Reference Stivers2008:35). After assessments, an affiliating response is, thus, an agreement. Affiliation is to be distinguished from alignment, defined by Stivers for recipients of storytelling as ‘support[ing] the structural asymmetry of the storytelling activity’ (Stivers Reference Stivers2008:34). In more general terms, alignment implies the recipient’s going along with an activity without necessarily affiliating, that is, going along with the stance or perspective of the co-participant (see also Lindström & Sorjonen Reference Lindström and Marja-Leena2013, Steensig Reference Steensig and Chapelle2019). In this section, we examine cases in which the speaker of the ingressive token is aligning but not necessarily affiliating with an assessment.
In excerpt (7), we see an ingressive token after a first assessment. In Finnish, second assessments that agree with a first assessment, typically consist of the verb repeated from the first assessment, either preceded by a particle (e.g. niin +V) or followed by one (V+ joo) (Sorjonen & Hakulinen Reference Sorjonen, Auli and Jack2009). Here, however, an assessment is responded to with an ingressive .joo.
The excerpt comes from a conversation between two young men, Mikko and Tero. Mikko is visiting his home town, and would like to take part in playing basketball in the street with his old pals. However, playing will prove to be difficult without proper foot gear. A local friend Tero promised (before this excerpt) to rummage his brother’s stock for suitable shoes.
At lines 1–2, Mikko is explaining that he has only brought walking shoes with him and begins an assertion about the quality of the walking shoes. Before he manages to complete this, Tero produces an affiliating response nii at line 3, which also claims knowledge (Sorjonen Reference Sorjonen2001a:181). Mikko’s attempt at a further explanation (line 6) is cut short by Tero’s assessment (line 7). This assessment exhibits a shared understanding of the situation by virtue of Tero taking Mikko’s perspective and judging the quality of the shoes. An agreeing, and affiliative, response could, thus, be nii on or on joo ‘it is/they are’. Instead, Mikko responds with an extended ingressive particle .joo-o. In responding in this way, he is not as much showing agreement with the stance as indicating that the preceding assessment revealed the speakers’ shared understanding and that the stance conveyed in the assessment was self-evident. So, Mikko aligns with the epistemic stance (the shared knowledge), but he does not go on to affiliate with the assessment. This does not mean that Mikko disaffiliates with Tero; his use of an ingressive particle indicates that they share the knowledge but it also indicates that there is no need to assess this more. In this way, Mikko retrospectively suggests that Tero’s assessment was not to be understood as the beginning of an assessment activity (Goodwin & Goodwin Reference Goodwin and Marjorie1987), in which affiliation is relevant, but rather as an expression of shared understanding, with which he aligns.
Tero, however, does go on with his assessment, in line 9, in a way that again appeals to shared knowledge (with varmaa ‘surely’), now specifying what the problem with the shoes could be ‘slippery’. This time, the assessment is treated by Mikko as the beginning of an assessment activity. He makes an agreeing second assessment with one of the formats typical for this action (o:n joo ‘they are’, line 11), but then goes on with a modification (lines 11–12) indicating that he does not see the slippery shoes as a big problem, it is one that can be overcome with some humor. By doing that, he recasts his initial agreement (line 11) as a ‘token agreement’ (Pomerantz Reference Pomerantz1984) and ends up not totally agreeing with Tero’s assessment.
We have seen here two successive, but different, treatments of an assessment. The first time around, the assessment was treated as a display of shared understanding with the ingressive particle, avoiding taking a stance. When the assessor insisted on making affiliation relevant, the recipient presented a second assessment that ended up not fully agreeing with the assessment.
Excerpt (8) is another instance of an ingressive response particle used after an assessment. It comes from the end of a telephone call to the Danish emergency line (for an account of the data, see Larsen Reference Larsen2013). CT is the call-taker and C the caller. The caller has called on behalf of a person who has an overheated car, the ‘him’ of line 7:
The call taker makes explicit that he has delivered the caller’s request to the fire assistance (lines 1–2). This is assessed positively by the caller (line 3) with a ‘low grade assessment’ (Lindström & Heinemann Reference Lindström and Trine2009), det var godt (literally: ‘that was good’). Lindström & Heinemann describe how low grade assessments are used in interactions between senior citizens and their caregivers to close care-giving activities. The only existing general account of assessments sequences in Danish talk-in-interaction (Garly Reference Garly2019) also describes how mutual assessments can be used to close sequences and activities, but it does not account for the formats of such closing assessment sequences. However, it has been established that phone calls often have assessments just before the actual closing of the call (Schegloff & Sacks Reference Schegloff and Harvey1973). In our data, such assessments are reciprocated, responded to with the particles ja ‘yes1’ or okay ‘okay’, or not responded to, after which participants proceed to the terminal sequence.
In excerpt (8), the assessment in line 3 is receipted by CT with a short ja in line 5. Here, participants might open up closings (Schegloff & Sacks Reference Schegloff and Harvey1973), but instead, C expresses his appreciation (line 6) and goes on to report that he will pass the positive outcome on to the person with the overheated car (line 7). In line 8, the call taker receipts this report with another positive assessment, a slightly contracted version of the caller’s assessment in line 3. Again, this opens the possibility of entering into closings. The caller receipts this assesment with an ingressive .jahh, after which he starts the terminal exchange with a ‘g’dbye’ (line 11), reciprocated by the call taker in line 12, upon which the call is closed.
In excerpt (8), the ingressive response aligns with the activity of opening up the closing while not affiliating with the evaluative stance in the utterance it responds to. This exhange is in no way problematic as the participants here collaborate in closing the activity.
In excerpts (7) and (8), assessments are not treated as calls for agreement and affiliation, but rather as displays of mutual understanding of the activity. In excerpt (7), line 7, Tero merely voices his understanding of the problem with the shoes, in a situation where the problematic nature of the shoes is already established and action is taken to solve the problem. The ingressive response to Tero’s assessment registers and accepts the assessment, but it also indicates that the assessment sequence need not be expanded. In excerpt (8), this non-expandability is even clearer, the assessments are part of opening up closings, the first one (line 3) gets an egressive ‘yes’ response, after which there is a minimal expansion, the next one (line 8) receives an ingressive particle response, after which the participants effectuate the closing.
The instances we have looked into so far occur in situations where the participants agree, at least momentarily, on the status of the utterances that they respond to with ingressive response particles. But the device can also be used in a less harmonious way – as interrupting or leading away from something. We will discuss instances of such, more problematic, uses of the ingressive responses in order to see what they reveal about the interactional functions of the device.
3.4 Ingressive particles after affiliation-seeking utterances
Utterances can be designed to seek affiliation from the recipient. This is true not only for (specific kinds of) assessments, but also for other actions that invite recipients to go along with an expressed perspective or stance. Ingressive particles that occur after such utterances are not treated as affiliative, as we will see from the next three excerpts.
Excerpt (9) comes from a phone conversation. Two women are catching up with latest news. They have come to talk about their mutual friend Soili, who has problems with sub-tenants in her rental apartment. Arja is informed of the recent developments, and she is here reporting on the unpleasant events to her friend Jaana.
The lengthy report has turned out to become a complaint story on behalf of the mutual friend Soili. The change between two sub-tenants while Soili herself is going to be abroad, has not been running smoothly. The latest of her problems is a missing key, the disappearing of which has been attributed to Soili, who is also going to lose money. The story has reached a culmination point at line 4, a place where the narrator pursues affiliation: the recipient is expected to evaluate, to commiserate on the hard luck of the protagonist (Jefferson Reference Jefferson2015). But instead of commiserating, Jaana, in overlap with the narration, utters a reduplicated joo joo, which merely indexes that she has followed the story so far and understands Soili’s situation. After the teller rephrases the culmination at line 6, Jaana’s response is a mere ingressive .joo. An ingressive particle suggests, as we have seen, that the speaker has nothing to add; the prior talk presented already established knowledge. In this situation, where an engaged response is called for, it becomes an instance of what Gail Jefferson termed a ‘perverse passive’, that is, a response that functions as a way of ‘preserving her status as recipient, and thereby preserving her co-participant’s status as teller’ (Jefferson Reference Jefferson1983:15) in a situation where ‘it is exceptable and appropriate that a recipient assume speakership’ (ibid.:14).
The narrator Arja does not use a reciprocal .joo nor any other similar closing implicative device. She clicks her tongue, as if in waiting for an appropriate response (line 8). As Jaana still shows no sign of affiliation, the teller has to do the work of the recipient. The awkwardness in her having to present the evaluating turn herself (line 10) is hearable in the way she produces it with laughter (Jefferson Reference Jefferson2015). Finally, the recipient joins in the evaluation with a second assessment in line 11. In this instance, the use of the ingressive particle is disaffiliating: it did not suffice for the narrator, and therefore, it leads to further work by the narrator in order to get jointly out of the sequence.
In excerpt (10), an ingressive particle is used as a response to a pursuit of affiliation. The excerpt is from a telephone conversation between P, an employee in an office providing au-pair services for Danish clients, and K, a client who has called to complain that her au-pair has not arrived yet. We enter when P has explained that it is not the bureau but the immigration authorities (de ‘they’ in line 1) who are at fault:
Line 1 is the conclusion and the climax of P’s telling about the incompetent authorities. It invites affiliation, which could take the form of a claim of agreement (for instance, ‘That’s right’). Instead, there is a short silence and then (line 3) a prolonged ja: ‘yes’, indicating a certain degree of reservation or resistance (Lindström Reference Lindström1999, Steensig Reference Steensig2015). This is clearly not sufficient for P, which is evidenced by the fact that she quickly adds an emphasizing repetition of her point in line 4 (‘That’s the way it is’). This is not presented as a personal opinion but as a generalization that no one can dispute. The wording has an idiomatic ring to it. As pointed out by Drew & Holt (Reference Drew and Elizabeth1988:398), ‘a complaint is formulated idiomatically at a point where there is some conflict or lack of alignment between complainant and recipient’. In this case, where P is seeking an acceptance from K that there are powers beyond her control who are at fault, an affiliative response might be something like ja det er rigtigt ‘yes that’s true’ or ja det ka jeg godt se ‘yes I can see that’. K’s response is, however, merely an ingressive .ja, (line 6), which does claim acknowledgement, but also, as we have seen above, that what was said is sufficient and that the sequence can be closed. This is followed by a lengthy silence (line 7), and then K proceeds with a men ‘but’ to making the point that she believes that it is the bureau’s job to ‘do something about it’. This is clearly a non-affiliating move, which in retrospect supports the understanding that the ingressive ‘yes’ in line 6 was anything but affiliative and that it was premonitoring disaffiliation.
A final example of how ingressive particles can be used in disaffiliative situations, in (11) below, comes from a recording of an experimental situation carried out in the 1950s. A group of four men, selected for a sociological experiment, are solving problems related to work. The experiment is carried out in a restaurant cabinet, and at intervals, the men are being served alcohol by a waitress (TA).
After solving one of the tasks, the men turn to the waitress who has just reappeared. While she is filling up their glasses, a preceding topic is dropped, and another topic is picked by Veka at random from the situation (Bergmann Reference Bergmann, Ivana and Klaus1990), commenting on the shape of the table they are seated at. However, the randomness develops on the spot into a flirtatious mood: From the heart shape of the table, to the room being labeled as a bridal house in an utterance (lines 6–7), directed as a question to the waitress. What makes this interpretable as flirtatious is the fact that it oversteps the institutional roles and claims the right to say something that has romantic implications (Speer Reference Speer2017). Of course, as Speer (Reference Speer2017) shows, flirtatious utterances are designedly ambiguous. In this case, the waitress (TA) has the opportunity of orienting or not to the flirt in her response. Instead of explicitly answering the polar question affirmatively ‘Yes, this is a wedding room’), the waitress receipts it with the ingressive .joo, as if going through the motions of confirming. This use amounts to implicating, ‘I have nothing to add here’. The ingressive device is something one cannot be made accountable for, so there is no way for the men to pursue the flirting, and so the topic is killed in the bud. In the second TCU of her turn in line 8, the waitress utters a phrase that belongs to her profession, accordingly responded to with a ‘thanks’ by the men, and the situation is returned to the professional mode.
Ingressive responses in situations where affiliation is the relevant next move (Lindström & Sorjonen Reference Lindström and Marja-Leena2013), cannot be heard as affiliative. Even though they may express a degree of going along, by virtue of belonging to ‘positive’ (affirmative) response particles, they are not understood as delivering the necessary affiliation. They indicate that there is nothing more to say and that the sequence can be closed; in certain situations, such an action is going against the affiliation that was called for.
4. Concluding discussion
The ingressive response particles in Finnish and Danish match the polarity of the utterances they respond to, and their basic meaning potential seems to be that of their corresponding egressive versions: They express confirmation or acknowledgement. In addition, however, they indicate that what is being responded to is already shared and, consequently, that the sequence they occur in can be closed, suggesting that there is nothing more to add.
Therefore, it makes sense that we find the ingressive response particles in third position as a receipt of an answer to a question. The examples we presented of ingressive particles in this position in Section 3.1 show that the particles occur at points where all relevant issues have been clarified, and that the answers to the questions were expected and found sufficient. Furthermore, ingressive particles may occur together with other indications of sequence closure.
Ingressive response particles were also found to occur as answers to questions (Section 3.2). In these instances, however, the questions were asked from a position of shared knowledge, conveying a mutual understanding that had already been created. By responding with ingressive particles, the answerers acknowledged not only the accuracy of the demonstrated understanding but also that this understanding was already shared and that there was nothing more to add.
In Section 3.3, ingressive particles were used after assessments. Our analyses showed that ingressive responses did not treat the assessments they responded to as the beginning of an assessment sequence, but, rather, as indexing sufficiency and opening up a closing of the sequence (or the interaction). Thus, the ingressive responses did not treat the assessments as affiliation-seeking, instead they aligned with the activity of closing.
In the excerpts in Section 3.4, we argued that the utterances that the ingressive particles responded to were indeed seeking affiliation. The affiliation could have been in the form of agreeing with an assessment and its perspectives or in the form of going along with an activity (e.g. a flirtatious one). In these cases, the sufficiency-indexing and closure-implicative nature of the ingressive response particles was mobilized to disaffiliate with the move suggested by the prior speaker, which became evident in the ensuing interaction.
Speaking on the inbreath, or ingressive pulmonic speech, is a phenomenon that has been attested in many languages. This article focuses on one particular use of ingressive speech, that of ingressive response particles. We have used data from an opportunity sample of naturally occurring interactions in two languages that are known to have this feature. As an interactional linguistic resource, ingressive response particles add to the functions of their non-ingressive counterparts that they indicate that what is responded to was already shared and that there is no need to go any further. This makes the particles useful for closing sequences, but also for indicating sufficiency in other contexts, of which we have examined the ones that stood out clearly from out data corpus.
By way of conclusion, we would like to add some remarks about the implications of our study and spell out limitations and perspectives. We will begin with some study-internal methodological caveats and then briefly address other phonations and actions in other modalities, sociolinguistic implications, cross-linguistic comparison, and the grammatical status of ingressive response particles.
All the instances of ingressive response particles in our study are responsive and designed to index a possible closing of a sequence or activity. This means that their interactional import is never addressed explicitly. We do not get instances of next speakers commenting on, or explicitly disaligning with, this kind of particles. The lack of explicit accountability means, of course, that strict ‘next turn proof procedure’ (Sacks, Schegloff & Jefferson Reference Sacks, Schegloff and Jefferson1974:728–729) is not very useful as an analytic tool (for participants and analysts alike). Instead, we have relied on spelling out relevant alternative actions at the point where the ingressive particles occur and used ‘circumstantial evidence’ as support for our analyses.
This lack of explicit orientation to the ingressive particles also means that we cannot claim as yet that ingressivity has a special status. The inventory of response tokens contains other minimally engaging ways of pronouncing response particles – for instance, with low volume, creaky voice, final closure, etc. It is possible that some of these phonations do the same jobs as the ingressive phonation does in our study. The use of such phonations could be investigated in languages that do not make use of ingressive phonation (for this type of utterances) and, in languages that do have ingressive response particles, the use of other phonations when uttering response particles might shed more light on both the ingressivity and the other phonations. These might again be compared to vocalizations further down the continuum towards ‘non-verbal’ or ‘somatic’ phonations (Keevallik & Ogden Reference Keevallik and Richard2020). We also saw that ingressive response particles could cooccur with specific nodding behavior. The interplay between these modalities could be studied in more depth, including the possibility that there may be multimodal ‘response packages’ (Kärkkäinen & Thompson Reference Kärkkäinen and Thompson2018) involving ingressive phonation and specific bodily behavior.
We explained above that our corpus is an opportunity sample and was not designed to be representative of particular types of speakers or activities. Therefore, we are not able to say anything about the frequency or sociolinguistic distribution of the practices that we describe, other than the facts we accounted for above, that the use of ingressive speech is not restricted to particular gender identities, private situations or intimacy. With reference to our data, it appears that ingressive response particles are a resource available to all users of the language, although not all speakers may use it. We do have recordings of long interactions in which nobody uses these particles, so that it is quite possible that there may be factors related to activities, relations, identities or other social circumstances that inhibit their use. Such factors remain to be investigated.
For the interactional functions that we have examined, we found similar patterns in interactions in the two languages, Finnish and Danish, that are typologically unrelated but may share areal features. This fact does not amount to saying that ingressive particles, and even less ingressive speech more generally, have the same status and interactional usages in the two languages overall. Further studies might shed light on differences between the two languages and on the situation in other languages that share the feature of having ingressive response particles. Judging from our limited data sets, and anecdotal evidence, it seems that Finns use ingressive speech more often and on more extended utterances than Danes, and the same might be true for other languages sharing this feature. What this means for possible generalizations about the role of ingressive speech in the different languages is something that further studies might investigate.
We believe that a grammar of talk-in-interaction should include the features that contribute to determining not only which particles can be used in specific environments but also what the different phonations of the particles may mean in interaction. As mentioned above, the ingressive response particles are not, or only cursorily, mentioned in existing grammars. It is our hope that the present description may contribute to including descriptions of these particles, both in grammars and in analyses of interaction.
Acknowledgements
We have presented earlier versions of this paper at numerous conferences and workshops, and we have discussed data extracts used in the paper at data sessions on many occasions. The paper has benefitted from reviews by anonymous reviewers from Nordic Journal of Linguistics. The authors, of course, still take the full responsibility for this result. Finally, we wish to thank all the people who have agreed to have their interactions recorded and, thereby, made this research possible.
