Elamur Research⌕
☰
home / papers / 2026 / A Computational Forensic Linguistic Analysis of Narrative and Question-Answer Structures in Italian Police Interrogation Transcripts

A Computational Forensic Linguistic Analysis of Narrative and Question-Answer Structures in Italian Police Interrogation Transcripts

Paper · 2026 · Romane Werner, Thomas François and Sonja Bitzer

Summary

Italian police interrogation records are treated in court as faithful renderings of what suspects said. In practice, loosely regulated protocols let officers rewrite statements, and the linguistic effect of the chosen format has never been measured computationally. This student-workshop paper profiles the two formats used in Italy: narrative monologues, where questions are dropped and the account is written in the first person, and question-answer (Q-A) transcripts.

The authors took 50 publicly available interrogations of Cosa Nostra justice collaborators from Antimafia Commission archives. They annotated them automatically and extracted 147 lexical, morphosyntactic, syntactic, readability and legal features. These were compared with a five-genre Italian reference corpus using feature selection and statistical tests, with classifiers serving only as interpretable diagnostics.

Monologues show deeper embedding, more past tense and more first-person singular verbs, which suits a chronological story. Q-A texts show longer subordinate chains, more clausal complements, more pronouns and more object preposing, reflecting turn-taking. Both resemble oral and narrative discourse more than legal prose.

The work argues that no transcript is neutral and that legal NLP should model transcription genre explicitly.

Key points

  • Corpus: 25 monologic and 25 Q-A transcripts dated 1984–2018, totalling 259,067 tokens.
    • Monologic transcripts: 46,810 tokens.
    • Q-A transcripts: 212,257 tokens.
  • Reference genres:
    • Legal-lay texts from the CorIELLS Italian subcorpus.
    • Newspaper articles from Il Fatto Quotidiano.
    • Semi-structured interviews from HABLA.
    • Eight contemporary novels.
    • Parliamentary debates from ParlaMint.
  • Processing pipeline: FDO-POS tagger and the DeSR dependency parser, both adapted to legal Italian (TEMIS), producing CoNLL-style output. Manual checks found POS precision of 0.97 on narrative transcripts and 0.98 on Q-A transcripts, with most errors caused by dialectal words.
  • Feature set: 147 automatically extracted features. The approach is explanatory and favours interpretability over prediction accuracy.
  • Stated limitations: a small corpus, no multimodal cues such as prosody or pauses, and coverage of only Italian and two formats.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksPsychology · 1,649 worksSocial Psychology · 330 works

🧭 Topics 1Research subjects and languages the work is about.Legal AI · 150 works

Cite

Cite as Romane Werner, Thomas François and Sonja Bitzer. 2026. A Computational Forensic Linguistic Analysis of Narrative and Question-Answer Structures in Italian Police Interrogation Transcripts. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 4: Student Research Workshop).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.eacl-srw.45

@inproceedings{werner2026computational,
  title = {{A Computational Forensic Linguistic Analysis of Narrative and Question-Answer Structures in Italian Police Interrogation Transcripts}},
  author = {Romane Werner, Thomas François and Sonja Bitzer},
  booktitle = {Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 4: Student Research Workshop)},
  year = {2026},
  doi = {10.18653/v1/2026.eacl-srw.45},
  url = {https://doi.org/10.18653/v1/2026.eacl-srw.45}
}

References

49 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Anabel Borja Albi. 2013. A genre analysis approach to the study of the translation of court documents. Linguistica Antverpiensia, New Series – Themes in Translation Studies, 12. find
  2. Giuseppe Attardi. 2006. Experiments with a multilingual non-projective dependency parser. Proceedings of the Tenth Conference on Computational Natural Language Learning, pages 166–170. find
  3. Anis Bawarshi and Mary Reiff. 2010. Genre: An Introduction to History, Theory, Research and Pedagogy. Parlor Press, Anderson. find
  4. Vijay Bhatia. 1993. Analysing Genre: Language Use in Professional Settings. Longman, London. find
  5. Douglas Biber and Susan Conrad. 2009. Register, Genre and Style. Cambridge University Press, Cambridge. find
  6. Younes Bouchlaghem, Yassine Akhiat, and Souad Amjad. 2022. Feature selection: A review and comparative study. 10th International Conference on Innovation, Modern Applied Sciences and Environmental Studies, 351. find
  7. Karl Branting, Craig Pfeifer, Bradford Brown, Lisa Ferro, John Aberdeen, Brandy Weiss, Mark Pfaff, and Bill Liao. 2020. Scalable and explainable legal prediction. Artificial Intelligence and Law, 22(2):213– 238. find
  8. Dominique Brunato. 2014. Complessità necessaria o stereotipi del ’burocratese’? un’indagine sulla leggibilità del linguaggio amministrativo da una prospettiva linguistico-computazionale. XIII Congresso della SILFI. find
  9. Dominique Brunato, Andrea Cimino, Felice Dell’Orletta, Simonetta Montemagni, and Giulia Venturi. 2020. Profiling-UD: A tool for linguistic profiling of texts. Proceedings of the 12th Conference on Language Resources and Evaluation, pages 7145–7151. find
  10. Dominique Brunato and Felice Dell’Orletta. 2017. On the order of words in italian: A survey on genre vs complexity. Proceedings of the Fourt International Conference on Dependency Linguistics. find
  11. Lucia Busso. 2021. Lexicon and grammar in legal-lay language: A quantitative corpus study on italian. Studi Italiani di Linguistica Teorica e Applicata, 51:5– 32. find
  12. Anna Bussu. 2016. Gathering evidence: Problems, training requirements, and good practices in the italian judicial police force. Police Practice and Research, 17(5):394–407. find
  13. Jie Cai, Jiawei Luo, Shulin Wang, and Sheng Yang. 2018. Feature selection in machine learning: A new perspective. Neurocomputing, 300:70–79. find
  14. Xinying Chen and Miroslav Kubát. 2025. Genre variation in dependency types: A two-level genre analysis using the czech national corpus. Proceedings of the Eighth International Conference on Dependency Linguistics, pages 84–92. find
  15. Andrea Cimino, Martijn Wieling, Felice Dell’Orletta, Simonetta Montemagni, and Giulia Venturi. 2017. Identifying predictive features for textual genre classification: the key role of syntax. Proceedings of the Fourth Italian Conference on Computational Linguistics CLiC-it 2017. find
  16. Eleonora Cocciu, Dominique Brunato, Giulia Venturi, and Felice Dell’Orletta. 2018. Gender and genre linguistic profiling: A case study on female and male journalistic and diary prose. Proceedings of the Fifth Italian Conference on Computational Linguistics. find
  17. Malcolm Coulthard and Alison Johnson. 2007. An Introduction to Forensic Linguistics: Language in Evidence. Routledge, London. find
  18. Felice Dell’Orletta. 2009. Ensemble system for part-of-speech tagging. Proceedings of EVALITA, 9:1–8. find
  19. Felice Dell’Orletta, Simonetta Montemagni, and Giulia Venturi. 2013. Linguistic profiling of texts across textual gernes and readability levels: An exploratory study on italian fictional prose. Proceedings of Recent Advances in Natural Language Processing, 26(4):471–495. find
  20. Felice Dell’Orletta, Simonetta Montemagni, and Giulia Venturi. 2014. Assessing document and sentence readability in less resourced languages and across textual genres. ITL - International Journal of Applied Linguistics, 165(2):163–193. find
  21. Paul Drew and John Heritage. 1992. Analysing talk at work: An introduction. In Drew Paul and Heritage John, editors, Talk at Work: Interaction in Institutional Settings, pages 3–65. Cambridge University Press, Cambridge. find
  22. Andrea Dömötör, Tibor Kákonyi1, and Zijian Gyózó Yang. 2022. What’s your style? automatic genre identification with neural network. Computación y Sistemas, 26(3):1293–1299. find
  23. Diana Eades. 2010. Verbatim courtroom transcripts and discourse analysis. In Hannes Kniffka, editor, Recent Developments in Forensic Linguistics, pages 241–254. Peter Lang, Frankfurt am Main. find
  24. Anita Eerland and Tessa van Charldorp. 2022. The influence of police reporting styles on the processing of crime related information. Frontiers in Communication, 7. find
  25. Tomaz Erjavec. 2024. Multilingual comparable corpora of parliamentary debates in Parlamint 4.1. find
  26. Helen Fraser. 2003. Issues in transcription: Factors affecting the reliability of transcripts as evidence in legal cases. Forensic Linguistics, 10(2):203–226. find
  27. John Gibbons. 2003. Forensic Linguistics: An Introduction to Language in the Justice System. Wiley, Hoboken. find
  28. Elena Semino and Short Mick. 2004. Corpus Stylistics: Speech, Writing and Thought Presentation in a Corpus of English Writing. Routledge, London. find
  29. Isabelle Guyon and André Elisseeff. 2003. An introduction to variable and feature selection. Journal of Machine Learning Research, 3:1157–1182. find
  30. Roger Shuy. 1998. The Language of Confession, Interrogation and Deception. Sage, New York. find
  31. Kate Haworth. 2010. Police interviews in the judicial process: Police interviews as evidence. In Coulthard Malcolm and Johnson Alison, editors, The Routledge Handbook of Linguistics, pages 241–254. Routledge, Abingdon. find
  32. Martha Komter. 2006. From talk to text: The interactional construction of a police record. Research on Language and Social Interaction, 39(3):201–228. find
  33. Martha Komter. 2019. The Suspect’s Statement: Talk and Text in the Criminal Process. Cambridge University Press, Cambridge. find
  34. Martha Komter. 2022. Institutional and academic transcripts of police interrogations. Frontiers in Communication, 7. find
  35. Pietro Lucisano and Maria Emanuela Piemontese. 1988. Gulpease: una formula per la predizione della leggibilità di testi in lingua italiana. Scuola e Città, pages 110–124. find
  36. Chiara Sinatra. 2014. Il passaggio dall’oralità alla scrittura in ambito forense e giudiziario. Quadernos AISPI, 4:197–212. find
  37. Efstathios Stamatatos, George Kokkinakis, and Nikos Fakotakis. 2000. Automatic text categorization in terms of genre and author. Computational Linguistics, 26(4):471–495.
  38. Jiliang Tang, Salem Alelyani, and Huan Liu. 2014. Feature selection for classification: A review. Data Classification: Algorithms and Applications, pages 37–67. find
  39. Nesrine Triki. 2023. Narratorial techniques in Tunisian police and court transcripts: A forensic linguistic approach. Text and Context. find
  40. Giulia Venturi. 2012. Investigating legal language peculiarities across different types of italian legal texts: An nlp-based approach. find
  41. IAFL Porto 2012 Proceedings. find
  42. Jacqueline Visctonti. 2010. Lingua e diritto: Livelli di analisi. LED Edizioni Universitarie, Milano. find
  43. Viviana Masa and Matteo Pandimiglio. 2020. Linguistica, diritto e variazione: uno sguardo al linguaggio delle sentenze in italia. Proceedings of Linguaggi settoriali e specialistici: sincronia, diacronia, traduzione, variazione. find
  44. Stefano Ondelli. 2014. Ordine delle parole nell’italiano delle sentenze: alcune misurazioni su corpora elet-tronici. Informatica e diritto, 23(1):13–39. find
  45. Jacqueline O’Toole. 2018. Institutional storytelling and personal narratives: Reflecting on the "value" of narrative inquiry. Irish Educational Studies, 37(2):175– 189. find
  46. Giovanna Pianese. 2008. Analisi linguistica comparativa di un corpus di testi del dominio giuridico: Sentenze penali italiane e francesi a confronto. Università degli Studi di Napoli Federico II, Napoli. find
  47. Emma Richardson, Kate Haworth, and Felicity Deamer. 2022. For the record: Questioning transcription processes in legal contexts. Applied Linguistics, 43(4):677–697. find
  48. Marina Santini. 2007. Automatic genre identification: Towards a flexible classification scheme. BCS IRSG Symposium: Future Directions in Information Access. find
  49. Thomas Schmidt and Kai Wörner. 2012. Multilingual Corpora and Multilingual Corpus Analysis. John Benjamins, Amsterdam. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/werner-2026-computational-forensic-linguistic-analysis-narrative/