Analysing Argumentation Structures

A Minimalistic Annotation Scheme for Argumentation Structures (v1.0)

These guidelines provide clarifications, maxims and hints that will help annotators to analyse the argumentation structure in argumentative texts.

Overview

This website contains MAnnoSAS, a minimalistic annotation scheme for argumentation structures. The annotation scheme is designed to narrow the degree of variability among (independently working) annotators. In other words, it helps to increase inter-annotator reliability when identifying argumentation structures.

The annotation scheme is based on concepts and insights from argumentation theory. It is minimalistic in the sense that it is parsimonious in the number of categories introduced and the theoretical baggage. The design of this scheme is motivated by the following question: How can we employ tools from argumentation theory to minimise the degree of interpretation without requiring annotators to be experts in argumentation theory?

Important context information about aims, scope and intended use of this annotation scheme can be found under https://github.com/xylomorph/mannosas.

License

This work is licensed under a Creative Commons Attribution 4.0 International License.

CC BY 4.0

Citation

BibTeX citation:

@article{cacean_2026,
  title = {Analysing Argumentation Structures - A Minimalistic Annotation Scheme.},
  author = {Cacean, Sebastian},
  year = {2026},
  month = september,
  doi = {10.5281/zenodo.22709968},
  langid = {english},
  url = {https://sebastiancacean.de/mannosas/},
}

NoteFor attribution, please cite this works as, for instance:

Cacean, S. (2026). Analysing Argumentation Structures - A Minimalistic Annotation Scheme. https://doi.org/10.5281/zenodo.22709968

Credits

MAnnoSAS (v1.0) was developed as part of my PhD Thesis “Content Analysis of Argumentation Structures - The Role of Reliability in Argument Mapping”.

Further Development

MAnnoSAS (v1.0) will be subject to further improvements (which will be published in new versions). If you have feedback and suggestions to improve the annotation scheme, feel free to open a GitHub issue and/or create a pull request.