EXMARaLDA

EXMARaLDA, standing for Extensible Markup Language for Discourse Analysis, is a free, open-source software suite for transcription, annotation, and analysis of spoken language data. It is designed to handle multimodal, time-aligned transcripts of conversations, interviews, and field recordings, with a focus on linguistic research and language documentation.

The data model relies on an XML-based format that stores media references (audio or video), speaker information, and multiple annotation tiers. The format supports hierarchical transcription structures and precise time alignment to media, enabling researchers to capture speaker turns, events, and discourse features in a flexible, extensible way. This facilitates reproducibility and sharing of annotated corpora across projects.

EXMARaLDA comprises several core components, including a Transcription Editor for creating and editing transcripts, and a Data Manager for organizing projects and media assets. It provides import and export filters to interoperate with other tools and formats, and supports multi-speaker dialogues, language variation, and non-speech annotations. The suite is designed to be used in a typical workflow that starts with project creation, media import, tier construction, annotation, timing alignment, quality control, and finally export for further analysis or publication.

The software is cross-platform, implemented in Java, and runs on Windows, macOS, and Linux. Projects are stored as portable bundles that link transcripts with their media files, aiding collaboration and long-term accessibility. EXMARaLDA is utilized in academic linguistics, sociolinguistics, and language documentation to build annotated spoken language corpora and support qualitative and quantitative analyses.