Notation systems for transcription: an empirical investigation

Area of Science:

• Linguistics
• Phonetics
• Speech Science

Background:

• Accurate transcription of spoken language is crucial for linguistic research.
• Understanding the impact of transcription systems on speech imitation is an ongoing area of study.

Purpose of the Study:

• To investigate how different transcription systems, including those with prosodic notation, affect the accuracy of speech imitation.
• To compare the effectiveness of audio recordings, typed transcripts, and combined formats for replicating speech patterns.

Main Methods:

• 90 participants listened to or read a 21-syllable utterance by Bernard Shaw.
• Participants imitated the utterance under nine conditions, varying audio, ordinary transcript, and prosodically notated transcript presentations.
• Performance was measured by approximations to Shaw's original temporal measures.

Main Results:

• Contrary to the hypothesis, prosodically notated transcripts significantly improved imitation accuracy when paired with audio.
• In the absence of audio, prosodically notated transcripts led to even greater improvements in temporal accuracy.
• Ordinary transcripts alone resulted in the poorest imitation accuracy.

Conclusions:

• Detailed prosodic notation in transcripts enhances the accuracy of speech imitation, particularly when audio is unavailable.
• The findings highlight the value of advanced transcription systems for research on spoken discourse.
• Further discussion on the utility, precision, and readability of these notation systems is warranted.