transcription

introduction

At the start of the tutorial, I made a bet with you. I bet that you would say how are you differently if you were 'greeting' someone, rather than 'genuinely inquiring' into their well being.

That poses an obvious problem for the way we present and share data. If we just write down the words, there would be no way of distinguishing one 'how are you' from the other. The work you did, to produce and distinguish the two actions, would be lost and that is what we want to analyse.

So, a number of people - most importantly Gail Jefferson - developed transcription notation to help capture not only what is said, but also how it is said. Some CA transcripts look quite formal, even off-putting. What people are trying to do is give you, the reader, as much access to the data as possible. This section talks a little about why such notation is used and the business of  producing a transcript.

why so much detail?

In transcripts, pauses are calculated, attempts are made to represent intonational patterns, parts of talk where people speed up and slow down, stretched words, especially loud words and phrases, etc. Each has a distinctive notion. As a hint, most people transcribe in new courier font, because of the way items on adjacent lines get neatly lined-up. 

When there is no real gap between two speakers turns, we use = to denote latching

A micro-pause is represented by (.) and gaps are measured likewise in brackets (0.6)

The beginnings of changes in pitch are denoted by arrows ↑ ↓ 

Where words are cut-off you get a dash, hel-

Overlapping talk is represented by square brackets across adjacent lines

Talk that speeds up >looks like this<, that slows down <looks like this> 

Breathing is also captured .hh for in-breath hh for out breath

Noticeably quiet talk ºhas these signs at its boundaries, whilst loud talk is underlined and sometimes capitalised.

Bits of laughter are captured using (h) where they occur

Why do this? Isn’t it overly technical and pseudo-scientific? Measuring pauses seems especially vulnerable to criticism. So what if the pause was (.7) and not (.6), does it really matter? So let’s start here. In some instances it might not matter one bit that the pause was (.7) rather than (.6). But, we also know that details, such as pauses, as well as many other details, can be significant for the management of social affairs. Temporality is a big sociological topic. It shouldn't be surprising then that it is a phenomenon for people themselves. Imagine, for example, extended pauses opening up after the following utterances. In each case, there would be consequences.

 ‘Will you marry me’?

‘Did you kill him?

 ‘Do you like my new haircut'?

Because people orient to pauses as having social significance, in many different ways, it seems a bit arbitrary just to leave them out. This basic point applies generally. People produce and recognise actions, not just based on what people say, but how they say it. We might put the matter this way - either you risk sounding overly technical and note lots of details, or ignore them, which might mean your analysis is well-off the mark.

In practice, each researcher has to find a transcription style they are happy with. Some people prefer lots of detail, others less so. Perhaps transcription is not about getting the transcript right in some objective sense, but a matter of producing something that is good enough to work with. It has to be good enough to enable you to explore those aspects of the interaction you are most interested in.

how to link talk and embodied activity?

People do this in different ways. If you look at Charles Goodwin's work there is a huge amount of innovation there in the way talk, materiality and embodiment are combined so the reader can access the unfolding nature of interaction. For the most part, people include still images that tally pertinently with the talk. In recent years Lorenza Mondada has further developed the approach to transcription.

Continuing with the Big Issue example, the figure below give some sense of the key moments in this interaction. The images capture the trajectory of bodily movements as the passer-by walks by the vendor, then stops and turns to engage with him. Some gestural aspects of the interaction are captured too.

In terms of the details of the talk. The pitch is a little stretched (denoted by the semi colon) and there is an intonation shift on 'issue' (arrow), followed by a gap of a little over a second. The passerby's response isn't completely clear, so there are brackets around '(no)'. The vendor hurries through the initial part of his response, so >'I get'< is marked up to show that. And so on. 

The Schegloff transcription tutorial is amazing and well worth a look. My aim here is just to introduce the basics.

exercise 2

As a quick transcription exercise. Play the clip, look at the transcript below and try to identify 2 or 3 issues that are variously incomplete, under-transcribed, debatable or poor! Click the link below to see some 'suggestions'.