EXMARaLDA 101: How to Transcribe Audio and Video

Written by

in

EXMARaLDA (an acronym for Extensible Markup Language for Discourse Annotation) is an open-source, Java-based software system used by linguists and researchers worldwide to create, manage, and analyze spoken language corpora (collections of audio and video recordings of natural speech and their corresponding transcripts).

Originally developed at the University of Hamburg, the platform is platform-independent (working across Windows, Mac, and Linux) and relies on XML to ensure long-term data sustainability and interoperability with other transcription tools. The EXMARaLDA system consists of three main core tools: 1. The Partitur-Editor (Score Editor)

What it does: This is the primary transcription and annotation tool.

How it works: It uses a “partitur” (musical score) visualization, which represents utterances, non-verbal behaviors, and gestures of multiple speakers in separate horizontal tiers.

Key feature: This score notation makes it easy to visually track and analyze overlaps, interruptions, and simultaneous events (e.g., when two people are speaking or laughing at the same time). It also links text intervals to precise timestamps in external audio or video files. 2. The Corpus Manager (Coma)

What it does: This acts as the database and organizational hub.

How it works: It allows researchers to bundle and organize primary media files, secondary transcripts, and complex speaker data into a unified, searchable corpus.

Key feature: It securely organizes metadata (such as speaker age, gender, mother tongue, and the setting of the communication) separately from the transcripts, meaning you can easily link multiple speech events to the same speaker without duplicating records. 3. The Query Tool (EXAKT) What it does: This is the analysis and search tool.

How it works: EXAKT is a KWIC (Keyword in Context) concordancer. It allows researchers to search for specific words, phrases, or linguistic patterns (using Regular Expressions) across an entire corpus of spoken language.

Key feature: When you find a match in the search results, you can click on it to see the exact interactional context and instantly play the synchronized audio or video recording of that specific moment. Why Researchers Use It New official version – EXMARaLDA

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *