Unterbrechen in zielgerichteten Gesprächen. Eine gesprächsanalytische Untersuchung

Abstract

Keskeyttämisen ilmiö keskusteluissa on ollut akateemisen tutkimuksen kohteena useiden vuosikymmenten ajan, ja sitä on lähestytty niin laadullisista kuin määrällisistä lähtökohdista. Esimerkiksi määrällisissä tutkimuksissa on tyypillisesti haluttu laskea keskeyttämistapaukset sen selvittämiseksi, kuinka usein keskeytyksiä tapahtuu tarkastelun kohteena olevissa keskusteluissa tai kuka keskeyttää eniten (esimerkiksi miehet vai naiset, lääkärit vai potilaat, jne.). Yllättävän usein itse ilmiön problematisointi on kuitenkin jäänyt ikään kuin puolitiehen. Tutkimuksissa on ajateltu esimerkiksi että kaikki päättymättömät lauseet ovat merkkejä keskeyttämisestä. Mutta onko näin keskustelun osapuolten kannalta? Se, mikä näyttää ulkopuoliselle keskeytykseltä, ei välttämättä vastaa keskustelun osallistujan tulkintaa.

Keskeyttäminen on lisäksi arkikielessä voimakkaasti värittynyt käsite, johon sisällytetään tunteita ja arvolatauksia. Tämä tarkoittaa, että käsitettä ei voi käyttää neutraalisti, mikä tekee siitä tieteellisestä näkökulmasta katsottuna epäluotettavan. Keskeyttämisen ilmiö itsessään on kuitenkin kiinnostava tutkimuksen kohde silloin, kun keskustelun osapuolet itse ilmaisevat, että keskeyttämistä tapahtuu.

Tässä väitöskirjassa tarkastellaan keskeyttämisen luonnetta osallistujien osoittamasta näkökulmasta. Tutkimus tarkastelee keskeyttämisen ilmiötä keskustelunanalyysin menetelmällä. Tutkimuksen aineisto koostui ammatillisessa kontekstissa käydyistä pienryhmäkeskusteluista.

Analyysin kohteena olivat erityisesti sellaiset tapaukset, joissa keskustelun osallistujat esittävät selvästi, että keskeytys oli meneillään tai on meneillään. Keskustelunanalyysin tulokset osoittavat, että keskeyttäminen on hyvin monitahoinen ilmiö eikä sellaisena sovellu määrälliseen laskemiseen. Toisin kuin aikaisemmin on oletettu, ei lauseiden valmista rakennetta voida pitää varmana merkkinä siitä, että puheenvuoro on osallistujien näkökulmasta viety loppuun. Esimerkiksi uuden puheenvuoron alku voi olla osallistujille ongelmallisempi, jos se osuu kesken pitemmän puheenvuoron, huolimatta siitä että yksittäiset lauseet on saatettu loppuun. Osallistujat näyttävätkin monin eri tavoin, että puheenvuoro on vielä meneillään tai toisaalta jo valmis. On esimerkiksi mahdollista, että puheenvuoro päätetään nonverbaalisesti, sanattoman viestinnän keinoin. Tällöin osapuolet eivät käsittele uuden puheenvuoron alkua keskeytyksenä.

Lisäksi analyysi osoittaa, että keskeyttäminen ei ole vain yhden osallistujan teko. Aikaisemmissa tutkimuksissa on ollut vallalla käsitys siitä, että keskeytys on jotakin joka tehdään toisille. Tämä tutkimus osoittaa sen sijaan, että keskeyttämisen syntyyn tarvitaan ainakin kaksi osallistujaa. Keskeyttäminen tehdään yhteistoiminnassa.

Merkittävä osa uusista puheenvuoroista alkaa kesken vielä meneillään olevaa puheenvuoroa, mutta tutkimuksen aineistossa osallistujat käsittelevät näitä tapauksia vain harvoin ongelmina tai keskeytyksinä myös silloin kun ne tehdään aggressiivisesti tai kilpailuhenkisesti. On huomattava, että osapuolet korjaavat usein keskustelun kulkua näiden aloituksien avulla ja varmistavat näin keskustelun jatkumisen yhteisiä tavoitteita kohti niin nopeasti ja vaivattomasti kuin mahdollista. Tällainen keskeyttäminen on siis toiminnallinen teko. Jos keskustelussa on esimerkiksi tapahtunut väärinkäsitys tai siinä ilmenee jokin muu äkillinen ongelma, on paras vaihtoehto usein aloittaa korjaus heti, ennen kuin keskustelu kehittyy ongelmalliseen, turhaan tai ei-toivottuun suuntaan.

Over the last four decades, a considerable amount of studies has been dealing with interrupting in conversations. Different subjects and research areas, like psychology, linguistics, and sociology, have had an interest in the phenomenon for various reasons, and previous work on interruptions yielded numerous results. However, surprisingly little is known about the phenomenon itself. This lack can be explained by the problematic approaches of earlier investigations.

Previous work on interrupting can be sorted into two main groups: Firstly, a large number of studies have been conducted based on quantifications. By formulating operational definitions accordingly, certain formal features (mostly simultaneous and syntactically incomplete talk) were (and are) assumed to be interruptions and counted as such often without exception. This research tradition neglects that interruption is a mundane concept: With this term we not only name trouble that has occurred in talk, but we also accuse someone else, or even ourselves, of causing the disturbance. Further, being disturbed or possibly disturbing a conversation partner are very personal matters, individual perceptions, which may vary considerably among participants and across situations. This means that research based on predefinitions not only ignores the term s moral implications (and paradoxically uses them at the same time, by determining people s personalities and status on grounds of the number of their interruptions ), it also neglects the fact that interruption is a participant s interpretation. It is impossible to determine interruptions from an outsider s perspective, namely on grounds of cases appearances, and stay neutral at the same time. Analysts should rely instead on participants mutually displayed understandings. Simply counting occurrences of unfinished talk as interruptions, or trying to prove predefined categories on the material, results in a distorted picture of a very complex phenomenon and bears the danger of groundless judgements of participants or their behaviour.

Secondly, some (significantly fewer) studies follow a qualitative approach, highlighting the importance of fine details when dealing with incidents of start-ups while someone else is already speaking. To avoid the negative connotation related to the term interruption , most researchers using this approach propose the use of the neutral and descriptive term overlap . One main outcome of this kind of work is that most cases of overlapping talk are systematic, a result of participants orienting to places in talk, which mark possible completion (see Jefferson 2004). While studies in this area are of great importance, since they actually show that looks do not automatically allow analysts to treat incidents as interruption, they do not provide an account of interruption per se which is after all an existing everyday category and as such scientifically interesting. They point out only what interruption is in fact not. The problems evident in the previous approaches lead to a very basic question: What is interrupting?

This dissertation strives to answer the posed question. Its main objective is to pinpoint the phenomenon as such, to describe how interaction partners make interruptions, and how they handle interrupting. The study follows a qualitative approach. It utilizes (ethnomethodological) conversation analysis, taking the perspective of the interactants into account. Conversation analysis considers talk-in-interaction as mutually achieved orderliness, as a result of participants organized doings. Conversation analysts are interested in the methods interlocutors employ to accomplish and organize everyday conversations. Analysis focuses on details and pursues to detect regularities and order. Accordingly, conversation analytic studies rely on recordings of naturally occurring talks. To be true to the perspective of the participants means that the analysis only (or at least mostly) takes into account what the data shows itself as interactively relevant. (see for example Heritage 1984)

The data utilized in this study consisted of videotapes of five authentic teacher conferences at schools in Finland. Recordings took place between the years 2003 and 2004. Two conversations were held in Finnish by native speakers, three talks were carried out in German two of which were accomplished by native speakers, and one by native speakers as well as non-native speakers of German (their mother tongue being Finnish). The situations lasted from one to two hours. Participants number ranged from three to six teachers.

In conversation analytic research it is assumed that participants constantly display to each other their mutual understandings. This way it is possible to achieve intersubjectivity. In other words, if an interactively relevant interruption occurred from the perspective of the participants, at least traces of such an interpretation should be observable. This work concentrates on clear cases, incidents that are explicitly treated as interruptions by the participants themselves. This seems necessary as a first step : Since the field is short of an actual account of interrupting in conversation, analysis of clear cases could provide secure grounds for working out a basic, participant orientated, and profound definition of the phenomenon to start with. The results of this analysis are complex. It appears that each of the analyzed clear cases is also a single case with its own, specific features. For example, incompleteness of an ongoing turn can be marked in different and sometimes even opposing ways. This heterogeneity is proof that interruption cannot be simply predefined for quantifications. Apart from that it seems that the pragmatic state of a turn is of greater importance than its syntactic structure, when it comes to its (in-)completeness. This outcome is especially relevant because the classic literature on the subject usually takes syntactic incompletion of a previous turn as a certain sign that an interruption has occurred. Another important observation is that interruption contrary to the established notion in the field does not seem to constitute a one-sided activity. As the analysis points out, at least two participants are needed for an interruption to occur. Interruptions are mutually accomplished.

The main outcome of this study is a participant-centred definition of interrupting in goal-orientated talks, which combines the aforementioned results: the phenomenon appears to be a turn-internal event which is mutually put forth as interruption by the participants themselves through the use of different resources. The conversation partners orient noticeably to syntactic, semantic, prosodic, nonverbal and/or pragmatic (in-)completion.

Looking at the material, it can also be noted that even though new turns are very often placed in the midst of an already ongoing turn, participants, differing from the majority of the literature, rarely treat these turn-internal activities as actual problems, or even interruptions. That is why this work pursues yet another research question: What are participants doing with turn-internal starts, what is their function within goal-oriented talk-in-interaction? Contrary to the assumptions of earlier, especially quantitative studies, the results of this conversation analytic work indicate that starting to talk in the midst of someone else s turn is most often functional, even when done aggressively or competitively. By turn-internal starts the participants usually begin to deal with some sudden trouble or other acute issue in the conversation. These starts are placed systematically namely closely to the trouble or issue to which they refer, before the talk can develop further (e.g., in a possibly problematic or unnecessary direction), and before the new activity needs some explanation to be related. These findings point out that participants orientate to some kind of conversational economy, and that by starting in the midst of an ongoing turn they resort to a sensible resource to adjust the (e.g., problematic) course of the talk as quickly and as effortlessly as possible. In doing so, participants also show their orientation towards their mutual goals.