Multimodal transcripts allow the visualization of participants’ lower-level actions, so that the analyst has visual records of the ongoing interaction, providing them with documents for analysis.  Subsequently, multimodal transcripts allow the analyst to present their findings to others (Norris, 2004: 60).

When transcribing video data, we deal with many different modes that cannot be transcribed simultaneously. 

An initial linguistic transcript would look something like this:

(1)   Luke:  out my face

Many linguistic transcription systems then illustrate intonational changes by using keyboard symbols.  However, with today's audio software, even spoken language can be transcribed adding a visual dimension and making it more exact.

Of course, transcribing language is just one small aspect when transcribing video data.  For the clip above, we would also want to transcribe the boy's facial expression, his gaze, and the head and shoulder movements in order to construct a true multimodal transcript.