O n f e r e n c e

Multimodal transcriptions 
 
In recent years some interest has developed around the multifaceted aspect of media. 
Some authors (O'Halloran,2004; Baldry, 2004; Baldry & Thibault, 2001; Jewitt, 2006) 
have suggested several multimodal transcription methods. These authors suggest 
different transcriptions for different multimodal texts. Baldry (2004), for instance, 
describes the multimodality that occurs in car advertisements; Williamson (2007) studies 
multimodality in the press and 
O’Halloran (2004) has carried
out a multimodal analysis in 
films. 
 
In the field of classroom research, Pujolà (2000) devises a transcribing method to 
describe the on- and off-screen interaction of language learners while engaging with 
computers. However, there is a lack or research when it comes to transcribing what 
occurs in a 3D virtual world a class where the student and the teacher take the form of 
an 'avatar' to represent themselves in the MUVEs. 
 
The 3M Transcription 
The aim of this paper is to propose a method for transcribing the interaction that occurs 
in a MUVE language classroom setting. The method was developed after recording, 
observing and analyzing foreign language classes carried out in a 3D MUVE, Second Life. 
The 3M transcription method (Pujolà & Palomeque, 2010) works at two levels of analysis: 
a macro level in which the sequences of the classroom interaction are observed and a 
micro level in which specific instances are described in detail. This transcription method is 
based on a coding system which facilitates the storage and retrieval of information. 
 
Macro 
At the macro level of transcription, as in any other transcription method, contextual 
information is gathered such as the nature of the course, its location 
–
a difference 
between face-to-face classes and MUVE classes is that the class can be held in any 
setting within the virtual world
–
and the participants of the course. However, the 
information about the participants is much more detailed than in other transcriptions as 
the researcher gathers information such as the clothes the participants are wearing, their 
skin and hair color, etc. Also, it should be bared in mind that a participant can choose 
what kind of avatar they want ranging from a female or male avatar to a child or animal-
shaped avatar. There is a wide range of options open to the imagination and expertise of 
each participant. This information may be irrelevant in a face-to-face classroom, but it 
gains relevancy in a MUVE context where these details can give the researcher clues to 
the participants’' level or proficiency in the MUVE as well as the reasons behind the 
choice of a certain shape, costume or item of clothing.
Figure 1 illustrates a highly abridged sample of a macro level of the transcription. It 
represents a whole class and it is organized so that the table fits the page and, thus, the 
reader can grasp the essence of the whole of the class easily. It is read like a musical 
score, it is organized in rows that are read from left to right.
The class is divided into scenes. A scene changes when there is a change in class stage 
and/or class activity or when there is a significant location or camera angle change. The 
scenes are numbered and on top of the scene the reader can see the duration of the 
scene. Apart from scenes, there are transitions. Transitions mark a change in scenario or 
continuous movement of the camera for a certain number seconds. Transitions usually 
occur when walking from one activity to the next or when teleporting to a different 
location. 

---

-287- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be 
To make this transcription easier to read and understand, the scenes have then been 
grouped into classroom activities (A1, A2…) and these activities have been named. Thus, 
it is easier to locate a specific scene within the context of an activity. 
The macro transcription has two aims, the first one, as mentioned before, is for the 
reader to get the ‘big picture’ of the sequence of the class. The second aim is to focus on 
those sequences where the researcher is going to develop their micro transcription 
analysis and locate it within the context of the whole class. To mark the selected scenes 
for the micro analysis, a snapshot of the picture has been added to the scene and the 
cells have been shaded. 

---

-288- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be