Although the original conversations are always regarded as the actual data, in this case by audio and video recordings, they have to be reproduced in visual form as transcriptions. As I mentioned earlier, there were no CA transcriptions available in Icelandic when I started the transcription process. Therefore, my first task was to study how the transcription process has been approached in other languages, and then to formulate my own guidelines for transcribing Icelandic conversation. The transcription conventions used in this study are first and foremost based on a system
which was developed by Gail Jefferson for American-English (cf. e.g. Ochs, Schegloff, and Thompson 1996: 461–5, Hutchby and Wooffitt 1998: 73–92).
According to Sacks, Schegloff, and Jefferson (1974: 734), the primary purpose of a CA transcription is “to get as much of the actual sound as possible into the transcripts, while still making them accessible to linguistically unsophisticated readers.” By this, Sacks, Schegloff, and Jefferson are probably referring to the fact that CA transcriptions are not based on IPA or other phonological transcription systems. This popularization of the transcription conventions and the attempt to make them available to the general reader have been criticized by Couper-Kuhlen and Selting (1996: 40). However, the transcription is always a selective representation of the primary data, the conversation itself. Hutchby and Wooffitt (1998: 88) point out that a good CA transcription is designed to “highlight analytically relevant features of talk-in-interaction.” My goal, therefore, has not been to do phonetically accurate transcriptions. Instead, I have considered each extract separately and aimed to transcribe it in such a way that it is comprehensible without leaving out anything that could be of importance for the analysis. Hence, I have focused on aspects which I consider especially important for the interaction. I notate pauses, overlapping talk, laughter, voice quality, hesitation, restarts, and terminal contours. Stress is marked when I perceive the stress pattern to be particularly prominant and potentially meaningful. Similarly, I note the soft pronunciation particular to speakers from Reykjavik (e.g., strætó ‘bus’ pronounced as strædó) only when I perceive it to be particularly prominent in the context.
When transforming spoken interaction into written characters, it can be difficult to assess the degree to which traditional spelling may be changed. In talk-in-interaction, speakers often use verbal contractions (maður > mar ‘man’) or utter two or more words without any clear prosodic boundaries. It is not always clear if these units should be considered as one or two words (cf. Ottesjö 2005: 21–2 on transcription of Swedish). This is particularly striking with respect to pronouns as objects, for example, in phrases such as ég sáann ‘I saw him,’ which in proper writing would be ég sá hann.
To show the layout of the examples in the following chapters, I will revisit section (1.1), here referred to as (5.1). The first line provides the Icelandic original, the second
a gloss in English, and the third a free translation into English. A key for transcription and glossing61 conventions is provided in Appendices A and B:
(5.1) I’M GETTING FED UP: Soul 07.05.96 (←1.1)
(M = Moderator; A = Arngrímur, a caller)
1 M Og Arngrímur Guðmundsson fyrstur í dag komdu sæll.
and 1nameM PatrM first.M in day GREET And Arngrímur Guðmundsson is first today hello
2 A Já komdu blessuð.
PRT GREET Yes hello
→ A >Vi (h)öfum nú talað saman áður?<
we have.3.PL NÚ talk.PP together before we have NÚ talked before
4 M J↓á:j↑áj↓á: Hvort við ekki höfum, PRT whether we not have.3.PL Yes yes yes we certainly have
5 A Ég skal segja þér eitt eh:: e- Eva- Eva Maríe.
I shall.1 tell.INF you.DAT one eh- e- 1nameF 1nameF 1nameF Let me tell you one thing eh- e- Eva- Eva Maríe.
6 A <Nú: fer mér að leiðast svolítið hvernig að mYnd e::::::>
NÚ go.3 I.DAT to bore.MV little bit how that shape eh- NÚ I’m getting a bit fed up with that shape eh
7 forsetafram- forsetakosningarnar eru að taka á sig ºnúnaº. president- presidential elections.DEF be.3.PL to take.INF on it.REF NÚNA the president- the presidential elections are taking NÚNA
→ M N↑ú:::
9 A J↓á::↑: mér finnst þetta nú orðið ansi .hhh ansi PRT I.DAT think.IMP.MV this NÚ become.PP.N quite .hhh quite Yeah I think this has NÚ become quite like
10 á:þekkt e:: (.) bara sona: *eh .h* hörku: alþingiskosningum?
similar eh- (.) PRT PRT ((Chuckle)) tough parliament elections.DAT eh (.) just like eh .h real parliamentary elections
The extract above has ten numbered lines. I use these numbers in my analysis when I refer to something in the conversation. The instances of nú and núna which will be discussed are boldfaced and marked with an arrow in the margin (lines 3 and 8). All instances of nú and núna are left untranslated, cited in uppercase. Each numbered line has three lines of text. When an excerpt occurs more than once, the numbers of the
61 Due to the richness of Icelandic grammar, the gloss is not grammatically complete. Only those grammatical features that are potentially important for the interaction are noted.
Time (s)
0 0.479862
100 150 200 300
50 500
70
other examples are given after an arrow, and after the identification line (for example
←1.1).
Of particular importance is the prosodic realization of nú and other particles.
Thus, N↑ú::: in line 8 represents an instance which is produced with: a) beginning intonation symbolized by a capital N, b) a prolonged vowel symbolized by four colons (one for each 0.1 seconds), and c) a rising intonation contour symbolized with the arrow pointing upwards. Nú in line 3, by contrast, is prosodically non-prominent.
As I will show in Chapter 9, the intonation contour is particularly important in the case of the syntactically non-integrated nú, such as N↑ú::: in line 8. The speakers treat the different contours as meaningful cues, and, therefore, they are important factors in my analysis.
To further support my analysis, I will use figures generated by the computer program Praat. These figures illustrate the different kinds of intonation contours and intensity patterns which the transcription symbols mentioned above represent. Figure 5.1 represents N↑ú::: in line 8 given with frequency in Hertz (Hz) displayed logarithmically on the vertical axis. Time in seconds is displayed on the horizontal axis:
N↑ú:::
Figure 5.1: Fundamental frequency (f0) in line 8, excerpt (5.1)
As I mentioned earlier, nú in line 3 is prosodically non-prominent, and it sounds as though it is cliticized to the preceding verb. In such cases, no arrows are used to show the intonation contour of nú. However, a question mark at the end of the utterance represents a rising terminal contour. Figure 5.2 shows the fundamental frequency for the TCU in line 3:
Time (s)
0 0.94001
100 150 200 300
50 500
70
Time (s)
0 0.94001
47.35 73.06
vi (h)öfum nú talað saman áður?
Figure 5.2: Fundamental frequency (f0) of Line 3, excerpt (5.1)
As Figure 5.2 illustrates, the particle nú is produced as a prosodically integrated part of við (h)öfum nú ‘we have NÚ,’ while the rest of the TCU, talað saman áður ‘talked (together) before,’ has vague boundaries between each word.
In addition to using prolonged vowels and recognizable intonation contours, speakers may also use intensity. Figure 5.3 shows the intensity of the utterance in line 3 given with frequency in decibels (dB) on the vertical axis and time on the horizontal axis:
vi (h)öfum nú talað saman áður?
Figure 5.3: Intensity of the 2nd TCU in Line 3, excerpt (5.1)
The discussion above shows that transcribing data is by no means an easy task. The transcriptions which appear here are a product of my own work, done with the study
of particles in mind. Other conversation analysts may choose other solutions, in particular, if they are focusing on other phenomena.