4.4 Taking turns at talk
4.4.2 Parsing the turn
Turn design and its interactional relevance have been described by Schegloff (1996) and a number of other scholars who base their discussion on Schegloff’s model (e.g.
Steensig 2001; J. Lindström 2002; Karlsson 2006). These studies have shown that the production of turns goes through six different stages: pre-beginning, beginning, post-beginning, pre-possible completion, possible completion and post-possible completion.48
Schegloff (1996: 92) characterizes pre-beginnings as “elements which project the onset of talk, or the beginning of a (next) TCU or a turn, but are yet not recognizable beginnings.” In other words, they initiate the turn without initiating a TCU. Similar elements may also be found in the post-beginning, just after the beginning of a turn.
Pre-beginnings may be both visual such as lip parting or gaze, or audial such as in-breath, coughing or throat clearing. In some studies, however, the pre-beginnings are interpreted loosely and include even lexical items, such as address terms and particles (cf. Steensig 2001; J. Lindström 2002). I, however, treat these elements as the actual turn beginning, because they are certainly meaningful entities which either constitute a TCU in their own right, or are prosodically and pragmatically integrated into larger TCUs. Examples of both pre- and post-beginning may be located in the following stretch of talk. This conversation is drawn from a phone-in program in which a car owner is getting advice about his broken car from a mechanic:
(4.4) BAD TIRES: Soul 23.05.96 (→8.7)
(M = Mechanic; H = Hrafnkell, a caller)
1 H Ég ver(ið)að pælí: (.) Ég var að kaupa dekk á bílinn hjá mér¿
I be.PP to think+in (.) I be.1.PT to buy.INF tires on car.DEF with I.DAT I’ve been wondering (.) I just bought tires for my car
2 M j↑á:
PRT yeah 3 (1.2)
4 H Og. (0.8) þa er eitthvað’ (.) Hann rá:sar svo eftir að ég fékk and (0.8) it be.3 something (.) he slide.3 so after that I get.1.PT and (0.8) there is something (.) it slides so back and forth that I got
48 As mentioned above, Schegloff is mostly focusing on turn-shifts and, thus, he does not address what happens between the post-beginning and pre-possible completion. This will be discussed in more detail in section 4.4.3 in which I address Auer’s (1996a; 1996b) ideas of syntax as a projection potential.
5 ný d↑ekk \og ég fór og talaði viððá þarna¿
new tires and I go.1.PT and speak.1.PT with+they.DAT there new tires and I went to talk with them
6 (0.2) 7 M j↑á:↓:
PRT yes 8 (0.3)
→ H .hhhhh (.)Og þeir #e# vildu ekkert fyrir mig gEra ég var .hhhhh (.) and they eh want.3.PT nothing for I.ACC do.INF I be.1.PT .hhhhh (.) and they eh didn’t want to help me I was
10 sona pælí hvað ég- Hvernig ég ætti að snúa mér í þessu.
PRT think+in what I- how I can.1. PT.SUBJ to turn.INF I.DAT in this like thinking what I should do about this
→ M .hhhhhh .mt #N↓ú hv-e e# hérna voru þau ekki ballanseruð .hhhhhh .mt NÚ wh- e eh PRT be.3.PT they not balance.PP.N.PL .hhhhh .mt NÚ wh- e eh ehm weren’t they balanced
12 og annað sl/íkt,=
and other such and such things?
The caller, Hrafnkell, signals very clearly that he is planning to enter the floor by producing audible in-breath before he starts producing the actual TCU (line 9). After he has produced the connective og ‘and’ and the subject þeir ‘they,’ he produces a planning marker. The planning marker occurs at the point of “maximum grammatical control” (cf. Schegloff 1996: 93), which is before the finite verb has been produced.
Thus, the mechanic can expect a continuation. After the post-beginning, Hrafnkell continues his turn by posing a question to the mechanic. The mechanic also prepares his answer with a pre-beginning, an audible in-breath followed by a click caused by parting of the lips (line 11). He begins his turn with a non-temporal nú but hesitates and produces a post-beginning. After the post-beginning, the mechanic marks the continuation of his turn by producing the particle hérna ‘here’ followed by a syntactic construction designed as a question.
Possible completion (see beginning of overlap in excerpt (4.5), line 4) is, according to Schegloff (1996: 83–92), the first point at which the TCU is viewed as complete. Speakers may also withhold the anticipated completion and yield the floor before it is ever delivered (cf. also Steensig 2001: 255–262). These endings are known as trail-offs (see excerpt (7.2), line 7) (cf. Jefferson 1983; Schegloff 1996: 87). The completion point is preceded by a point which is known as the pre-possible
completion (see the beginning of overlap in excerpt (4.5), line 2), the point at which the interlocutor is able to project the timing and the content of the closure and can thus enter the floor without waiting any further. Research has shown that interlocutors clearly orient to these points in the conversation, for example, by initiating self-repairs (Schegloff 1996: 85) or collaborative completions (Lerner 1991). Although this phenomenon has not been examined in Icelandic, I would suggest that it is often, yet not exclusively, marked with a pitch peak similar to what Schegloff (1996: 84–5) has noted on pre-possible completions in English (cf. also Steensig 2001: 255 on prosodic cues in Danish). Consider the turn-shifts in the following excerpt:
(4.5)TRADITIONS AND THE EUROVISION SONG CONTEST: Soul 17.05.96 (→7.5)
(M = Moderator; F = Fjóla, a caller)
1 M >Jú maður verður eiginlega að fara í keppni< með yes man.IMP must.3 actually to go.INF in competition with yes, you really must enter a competition with
→ því hugarf[ari að maður ætli að-]
that mind [set that man.IMP intend.3.SUBJ to ] that in min[d that you are going to- ]
3 F [ N e i e k k ] e r t endilega, [ PRT noth]ing especially [ no no]t particularly
→ M gera sitt bes:ta: [og helst að vinna. ] do.INF his.REF best.ACC [ and preferably to win.INF ] do one’s best [ and preferably win ] 5 F [J↓áj↑á h ú n h ]ú n
[ PRT she s]he [ yes she s]he
6 þa- [ég efast ekki um að hún geri sitt besta.
i- [ I doubt.1.MV not about that she do.3.SUBJ her.REF best.ACC i- [ I’m sure she will do her best
7 M [hún gerir sitt besta.
[ she do.3 her.REF best.ACC [ she will do her best
Fjóla and the moderator are having an argument in the excerpt above, and this is evidenced by the large number of early on-sets which result in overlaps. In the first turn in (4.5), the moderator produces an utterance which is an objection to Fjóla’s previous turn. The moderator’s turn is potentially completed after the noun hugarfar
‘mindset.’ Fjóla starts her response after the first two syllables of that word, treating the moderator’s turn as if it is coming to an end. Thus, Fjóla’s turn starts at the point
of a potential pre-possible completion. It turns out, however, that the moderator has not completed her turn, and, after Fjóla completes her own turn, the moderator continues with the turn she paused during the overlap. There is also a possible completion after besta ‘best,’ during which Fjóla takes the floor and overlaps the moderator again (lines 4 and 5).
As I mentioned above, talk-in-interaction is incrementally produced in real time.
Consequently, speakers rarely plan every word they are going to utter, and sometimes they choose to modify or add items to their turns. When speakers seem to have finished a turn, for example, after a possible completion, they may choose to add other elements to these turns. These additional elements are known as post-possible completions (Schegloff 1996: 90). According to Schegloff (1996: 90), these post-possible completions can be mainly of two sorts: post-completion stance markers (cf.
Karlsson 2006) or grammatically structured extensions to the preceding turn. The latter type is closely related to the notion of syntactic expansions (Auer 1996b).
Expansions are syntactically symbiotic with other constructions and can be understood only in that context. Consider Magga’s turn in line 6, which is designed as a syntactic expansion to her turn in lines 1 and 2:
(4.6)COMPOSING CLASSICAL MUSIC: Reunion (→6.7)
(M = Magga; B = Brynhildur; V = Vala; E = Erna)
1 M .mt þa er náttlega líka hægt að sem:ja eitthvað °í°
.mt it be.3 naturally also possible to compose.INF something in .mt it’s of course also possible to compose something in
2 klassískum stíl [til dæmis,]
classical style [ for example ] classical style [ for example ] 3 (E) [j↑á: ]
[PRT ] [ yes ] 4 B j↓áj↑á=
PRT yes 5 V =j↓á[:
PRT[
yea[h
→ M [n↑ún↓a.
7 (0.4)
Magga’s turn reaches a possible completion with the noun stíl ‘style’ (line 2). Note how the co-particiants start responding to her utterance immediately after that point.
Following this turn completion, however, Magga adds the phrase til dæmis ‘for example,’ and, after backchannels from three co-participants, she adds another post-completion, the temporal marker núna (line 6). This last word is uttered as a non-beginning, and so it has a continuing intonation (cf. Schegloff 1996: 73–7).
Continuations, by contrast, are produced as new units, and, consequently, they are often prosodically marked as turn beginnings. They are often initiated with connectives such as og ‘and’ or en ‘but,’ but their dependency on a previous TCU is usually merely pro-forma. In (4.7), the caller, Bjargey, produces a continuation:
(4.7)IT’S NICE THAT IT IS YOU: Soul 03.06.96 (→9.10)
(M = Moderator; B = Bjargey, a caller)
1 B Gaman að það skulir vera þú (.) #á# (.) #ö-# (.) ##
fun that it shall.3.SUBJ be.INF you (.) on (.) eh (.) eh Great that it’s you (.) on (.) eh (.) eh
2 þessum tíma núna, this time now at this time now 3 (0.2) 4 M N↓ú
→ B En maðurinn þarna me:ð+#e# (.) með matið hann ætti að minnka but man.DEF there with eh (.) with “matið” he should.3.SUBJ to reduce.INF the man with eh (.) with “matið” he should reduce
6 nota þetta sko líka.
use.INF this “sko” also using this “sko” also
→ B En þa var ekki þa sem ég ætlaði að tala um. .hhh [...]
but that be.3.PT not that which I intend.1.PT to talk.INF about .hhh [...]
but that wasn’t what I was going to talk about .hhh [...]
Bjargey begins the conversation by giving the moderator a compliment (lines 1–2) and then comments on a previous speaker who complained about the frequent use of the noun matið ‘the evaluation’ in the Icelandic media. This new topic is introduced with the connective en ‘but.’ According to Bjargey, that same speaker should reduce his use of the particle sko ‘y’see’ (line 5–6). Here, Bjargey’s turn seems to have come to a completion. However, Bjargey takes the floor again and produces a continuation, which she also introduces with the connective en ‘but,’ followed by another
topic-shift. Thus, Bjargey has in fact initiated two new turns which she designs as continuations of the previous turns.
Schegloff’s model for parsing the turn is an excellent tool for understanding the interactional mechanism taking place during turn production. The model explains how speakers use breathing and gaze when they wish to take the floor, and how their co-participants may orient to these more or less subtle cues. Hence, Schegloff shows that elements which in traditional linguistics have not been considered worth mentioning actually play a crucial role in interaction.