SignSynth Facial and Body Parameter Transcription Guide

A common criticism of signed-language transcription systems is that most do not have a way to accurately record facial expressions and body movements, which are essential to the grammar of most signed languages. This system aims to satisfy that need. It is not a detailed system like Paul Ekman's notation, but just attempts to reflect the major facial expressions used by signed languages.

I have taken some principles from Don Newkirk's Literal Orthography proposal for American Sign Language. The system uses all lower-case Roman letters, and the frequency and distribution of letters approximates that of English words. However, it is distinctive enough from all other transcription systems to be clearly separated, and can thus be used with other systems to add the ability to represent facial expressions.

There is the potential for ambiguity between "bie" (eyebrows up, eyes up-right), "bie" (eyebrows up, torso forward-right) and "bie" (head tilted right, torso forward-right) if non-essential elements are omitted. This ambiguity can be clarified if we specify the mouth configuration. Thus, with a neutral mouth, they become "bien," "bnie" and "nbie." Otherwise, the system will parse "bie" the same as "bien."

Facial Expressions

Body Movements


SignSynth comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to redistribute it under certain conditions; see the GNU Public License for details.

SignSynth is created by Angus B. Grieve-Smith. Questions or comments? Email me: grvsmth at