Politics of Transcription seminar, Roskilde, 4 June 2010
CALPIU will host a seminar on a range of issues relating to transcription, in the Lecture Room 3.2.5, at the CALPIU Research Center, Roskilde University
Roskilde University
The politics of transcription
Seminar with Celia Roberts (King’s College, London), Johannes Wagner (SDU, Kolding), Martha Sif Karrebæk (Copenhagen), Hartmut Haberland (RUC) and Brian MacWhinney (Carnegie Mellon University, Pittsburgh)
Programme
Celia Roberts: Issues of representation
Panel and general discussion
Johannes Wagner and Brian MacWhinney: Difficult issues in transcriptions and some of the ways they have been solved
Panel and general discussion
This seminar takes its point of departure in some issues that have been discussed within the CALPIU research network and research center and elsewhere. It is easy to agree that interactional data – even in telephone interviews and focus groups – are the full communicative and interactional event, not the ‘visible marks’. Bloomfield (Language, p. 21) wrote ‘Writing is not language, but merely a way of recording language by means of visible marks’. In the same way a transcription is not our data, but merely a way of recording our data by means of visible marks. But apparently the transformation of linear data (as we can record them in video or audio recordings) need to be transformed into a
two-dimensional transcript before we can work with them in a serious manner. But what happens on the way from the data to the transcription? Some reduction and reconstruction, not to talk about decontextualization and recontextualization. This implies decisions both on the side of the transcriber and on the side of the software developer.
These decisions are neither trivial nor mechanical, one can say: they are political.
Within CALPIU we have decided to stick to using the CHAT format for transcription and its associated software, CLAN. One of the reasons is that CLAN makes the relationship between data and transcription more transparent than other software packages. CLAN indeed helps the transcriber to produce a text (the transcript), but not only that. CLAN does not encourage the analysts to forget the video or audio recording of the data but reminds them constantly of the source of the transcript in the data by linking it back to the recording (by synchronizing data recordings and data transcriptions). Hence CLAN emphasizes that the transcript is not the data and what is being analyzed is not visible marks (on paper or on a screen), but the interactive event.
The panel participants present their views on issues of representation and transcription of data and invite the participants to discuss them.
Relevant literature that is recommended (but not required) as preparatory reading:
Celia Roberts n.d. Issues in transcribing spoken discourse.
http://www.kcl.ac.uk/content/1/c6/01/81/04/part1.pdf
Cécile Vigouroux 2009. The making of a scription: a case study on authority and authorship. Text and Talk 29(5):615-637
Gail Jefferson 1996. A case of transcriptional stereotyping. Journal of Pragmatics 26:159-170
Seminar with Celia Roberts (King’s College, London), Johannes Wagner (SDU, Kolding), Martha Sif Karrebæk (Copenhagen), Hartmut Haberland (RUC) and Brian MacWhinney (Carnegie Mellon University, Pittsburgh)
Programme
- 9:45 Welcome
- 10:00-12:00 Morning session
Celia Roberts: Issues of representation
Panel and general discussion
- 13:00-15:00 Afternoon session
Johannes Wagner and Brian MacWhinney: Difficult issues in transcriptions and some of the ways they have been solved
Panel and general discussion
This seminar takes its point of departure in some issues that have been discussed within the CALPIU research network and research center and elsewhere. It is easy to agree that interactional data – even in telephone interviews and focus groups – are the full communicative and interactional event, not the ‘visible marks’. Bloomfield (Language, p. 21) wrote ‘Writing is not language, but merely a way of recording language by means of visible marks’. In the same way a transcription is not our data, but merely a way of recording our data by means of visible marks. But apparently the transformation of linear data (as we can record them in video or audio recordings) need to be transformed into a
two-dimensional transcript before we can work with them in a serious manner. But what happens on the way from the data to the transcription? Some reduction and reconstruction, not to talk about decontextualization and recontextualization. This implies decisions both on the side of the transcriber and on the side of the software developer.
These decisions are neither trivial nor mechanical, one can say: they are political.
Within CALPIU we have decided to stick to using the CHAT format for transcription and its associated software, CLAN. One of the reasons is that CLAN makes the relationship between data and transcription more transparent than other software packages. CLAN indeed helps the transcriber to produce a text (the transcript), but not only that. CLAN does not encourage the analysts to forget the video or audio recording of the data but reminds them constantly of the source of the transcript in the data by linking it back to the recording (by synchronizing data recordings and data transcriptions). Hence CLAN emphasizes that the transcript is not the data and what is being analyzed is not visible marks (on paper or on a screen), but the interactive event.
The panel participants present their views on issues of representation and transcription of data and invite the participants to discuss them.
Relevant literature that is recommended (but not required) as preparatory reading:
Celia Roberts n.d. Issues in transcribing spoken discourse.
http://www.kcl.ac.uk/content/1/c6/01/81/04/part1.pdf
Cécile Vigouroux 2009. The making of a scription: a case study on authority and authorship. Text and Talk 29(5):615-637
Gail Jefferson 1996. A case of transcriptional stereotyping. Journal of Pragmatics 26:159-170