---
title: "Tracing English interference in AI-generated German (2026)"
description: "This study tests that claim for German syntax in a single, well-defined genre: news reporting. 50 references."
authors: ["Paolo Valentinelli"]
year: 2026
venue: "AI-Linguistica 5"
source: "https://ai-ling.publia.org/ai_ling/article/view/36"
doi: "10.62408/ai-ling.v5i1.36"
licence: "CC BY-NC-SA (journal licence per DOAJ; AI-Linguistica)"
licence_url: "https://creativecommons.org/licenses/by-nc-sa/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/valentinelli-2026-tracing-english-interference-ai-generated/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/valentinelli-2026-tracing-english-interference-ai-generated/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/18/) / Tracing English interference in AI-generated German

# Tracing English interference in AI-generated German

**Authors:** Paolo Valentinelli · **Year:** 2026 · **Venue:** AI-Linguistica 5 · **Licence:** [CC BY-NC-SA (journal licence per DOAJ; AI-Linguistica)](https://creativecommons.org/licenses/by-nc-sa/4.0/)

**Links:** [Publisher's page](https://doi.org/10.62408/ai-ling.v5i1.36) · [PDF](https://ai-ling.publia.org/ai_ling/article/download/36/32) · [DOI 10.62408/ai-ling.v5i1.36](https://doi.org/10.62408/ai-ling.v5i1.36) · [OpenAlex](https://openalex.org/W7124581957)

**Citations:** 1 citations (OpenAlex, 2026-10-14)

## Summary
LLM training data is dominated by English, so text generated in other languages may carry English structural habits. This study tests that claim for German syntax in a single, well-defined genre: news reporting. The author built a reference corpus of real political and general news articles from Süddeutsche Zeitung (labelled SZ) and BILD (labelled B). He then compared it with articles in which four different LLMs imitated German newspaper prose. The comparison covers two features: overall word order, and what fills the pre-field, which is the single slot before the finite verb in a German main clause. Both quantitative counts and close qualitative reading show the same pattern. The models favour subject-first SVO order and vary their sentence openings less than human journalists do. They also produce marked constructions that follow English linear sequencing instead of the German verbal bracket. Some models come close to human figures on some variables, but only partly and depending on context. The author treats this as evidence of a new kind of language contact that reduces the syntactic richness of non-English output.

## Key points
- Background figures from the paper show the imbalance in GPT-3 training data: about 93% English and only about 1.5% German. Newer models no longer disclose their data composition.
- The human baseline comes from Süddeutsche Zeitung and BILD articles, numbered SZ1–SZ25 and B1–B25. It is compared against output from four distinct LLMs.
- The reference grammar cited (Wöllstein 2022) puts the subject in the pre-field in roughly 50–55% of German main clauses. The remaining share goes to adverbials, objects, predicates and other constituents.
- The paper reports three findings:
  - a strong bias toward SVO;
  - lower syntactic variability than in human texts;
  - stylistically marked, English-like linearisation.
- "Interference" is defined statistically rather than cognitively. It means negative cross-lingual transfer, where target-language output maps one-to-one onto English syntactic templates.

## Interesting ideas
- The study compares the full distribution of a syntactic slot's fillers against a human corpus, not just grammaticality, which catches "correct but unnatural" output that error checks miss.
- The author deliberately picks a conservative, norm-bound register as the gold standard, so that any deviation cannot be excused as casual style.
- Pre-field occupation is a clean, countable proxy for word-order freedom: you classify the one constituent before the finite verb by its type.
- Interference is framed as output mirroring another language's linear templates, which makes it testable by checking for isomorphic structure.
- The paper measures reduced variability as a failure in itself, even when every sentence is well-formed.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (learn): Silver data produced by LLM teacher tiers could be checked for a skewed distribution of sentence-initial constituent types against a human treebank, as a negative control against teacher bias.
- **I2** (learn): Evaluation sets for generation or parsing could favour edited registers such as news or textbooks, so that divergences can be counted as system errors rather than noise.
- **I3** (tree): From a UD tree, the deprel of the dependent preceding the root or finite verb (nsubj, obl, obj, advmod…) gives a cheap per-sentence "fronting profile" usable for corpus diagnostics.
- **I4** (back): When translating through the meta-world, the generated sentence's constituent order could be compared with the source's, flagging outputs that simply copy the source order instead of using the target's natural options.
- **I5** (back): The generator and the text → graph → text compressor could track the diversity of sentence openings and clause orders against human text, so fluent output does not collapse into one monotonous template.

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

## Institutions

- [University of Trento](https://md.research.elamur.ai/free/en/universities/university-of-trento/) — University, Trento, Italy · 85 works · 15,524 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works

🧭 Topics (1): [English](https://md.research.elamur.ai/free/en/topics/english/) · 2,263 works

## Cite

Cite as: Paolo Valentinelli. 2026. Tracing English interference in AI-generated German. AI-Linguistica 5.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{valentinelli2026tracing,
  title = {{Tracing English interference in AI-generated German}},
  author = {Paolo Valentinelli},
  booktitle = {AI-Linguistica 5},
  year = {2026},
  doi = {10.62408/ai-ling.v5i1.36},
  url = {https://doi.org/10.62408/ai-ling.v5i1.36}
}
```

- [← previous TRACE: Evidence Grounding-Guided Multi-Video Event Understanding and Claim Generation](https://md.research.elamur.ai/free/en/papers/yan-2026-trace-evidence-grounding-guided-multi/)
- [next → Tracing Multilingual Knowledge Acquisition Dynamics in Domain Adaptation: A Case Study of Biomedical Adaptation](https://md.research.elamur.ai/free/en/papers/zhao-2026-tracing-multilingual-knowledge-acquisition-dynamics/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/18/)

## References

50 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Backus, Ad & Cohen, Michael & Cohn, Neil & Faber, Myrthe & Krahmer, Emiel & Laparle, Schuyler & Maier, Emar et al. 2023. Minds: Big questions for linguistics in the age of AI. Linguistics in the Netherlands 40(1). 301–308. https:/doi.org/10.1075/avt.00094.bac [↗](https://doi.org/10.1075/avt.00094.bac)
2. Bech, Gunnar. 1955. Studien über das deutsche Verbum Infinitum. København: I kommission hos E. Munksgaard. [find](https://scholar.google.com/scholar?q=Bech%2C+Gunnar%2E+1955%2E+Studien+%C3%BCber+das+deutsche+Verbum+Infinitum%2E+K%C3%B8benhavn%3A+I+kommission+hos+E%2E+Munksgaard%2E)
3. Brown, Tom B. & Mann, Benjamin & Ryder, Nick & Subbiah, Melanie & Kaplan, Jared & Dhariwal, Prafulla & Neelakantan, Arvind et al. 2020. [Language models are few-shot learners](https://md.research.elamur.ai/free/en/papers/brown-2020-language-models-few-shot-learners/). arXiv. https:/doi.org/10.48550/arXiv.2005.14165 [↗](https://doi.org/10.48550/arxiv.2005.14165)
4. Cicero, Francesco. 2023. L’italiano delle intelligenze artificiali generative. Italiano LinguaDue 15(2). 733–761. https:/doi.org/10.54103/2037-3597/21990 [↗](https://doi.org/10.54103/2037-3597/21990)
5. Cicero, Francesco. 2025. Esercizi di stile. La narrativa per l’infanzia delle intelligenze artificiali. AI-Linguistica. Linguistic Studies on AI-Generated Texts and Discourses 2(2). https:/doi.org/10.62408/ai-ling.v2i2.19 [↗](https://doi.org/10.62408/ai-ling.v2i2.19)
6. De Cesare, Anna-Maria. 2007. Le funzioni del passivo agentivo. Tra sintassi, semantica e testualità. Vox Romanica 66. 32–59. [find](https://scholar.google.com/scholar?q=De+Cesare%2C+Anna%2DMaria%2E+2007%2E+Le+funzioni+del+passivo+agentivo%2E+Tra+sintassi%2C+semantica+e+testualit%C3%A0%2E+Vox+Romanica+66%2E+32%E2%80%9359%2E)
7. De Cesare, Anna-Maria. 2023a. Giorgia Meloni, Meloni o la Meloni? La codifica degli antroponimi femminili in biografie generate da ChatGPT e pubblicate su Wikipedia. Lingue e culture dei media 7(1–2). 1–20. https:/doi.org/10.54103/2532-1803/22388 [↗](https://doi.org/10.54103/2532-1803/22388)
8. De Cesare, Anna-Maria. 2023b. Assessing the quality of ChatGPT’s generated output in light of human-written texts: A corpus study based on textual parameters. CHIMERA: Revista de Corpus de Lenguas Romances y Estudios Lingüísticos 10. 179–210. [https://revistas.uam.es/chimera/article/view/17979](https://revistas.uam.es/chimera/article/view/17979) (last accessed on 09/01/2026). [↗](https://revistas.uam.es/chimera/article/view/17979)
9. De Cesare, Anna-Maria. 2024. Nuove dinamiche di contatto linguistico. Le “impronte digitali” dell’inglese nell’italiano generato da LLM. Lid’O. Lingua Italiana d’Oggi XXI: 67–91. [find](https://scholar.google.com/scholar?q=De+Cesare%2C+Anna%2DMaria%2E+2024%2E+Nuove+dinamiche+di+contatto+linguistico%2E+Le+%E2%80%9Cimpronte+digitali%E2%80%9D+dell%E2%80%99inglese+nell%E2%80%99italiano+generato+da+LLM%2E+Lid%E2%80%99O%2E+Lingua+Italiana+d%E2%80%99Oggi+XXI%3A+67%E2%80%9391%2E)
10. De Cesare, Anna-Maria. 2025. LLM referential chain generation: A qualitative case study based on Italian biographies produced by GPT-4. Linguistik Online 136(4). 25–52. https:/doi.org/10.13092/8tppdv47 [↗](https://doi.org/10.13092/8tppdv47)
11. De Cesare, Anna-Maria & Garassino, Davide & Agar Marco, Rocío & Albom, Ana & Cimmino, Doriana. 2016. Sintassi marcata dell’italiano dell’uso medio in prospettiva contrastiva con il francese, lo spagnolo, il tedesco e l’inglese. Berlin: Peter Lang. [↗](https://doi.org/10.3726/978-3-653-05938-0)
12. Deutscher Presserat. 2025. Publizistische Grundsätze (Pressekodex). Berlin: Deutscher Presserat. [find](https://scholar.google.com/scholar?q=Deutscher+Presserat%2E+2025%2E+Publizistische+Grunds%C3%A4tze+%28Pressekodex%29%2E+Berlin%3A+Deutscher+Presserat%2E)
13. Eberhard, David M. & Simons, Gary F. & Fennig, Charles D. 2025. Ethnologue: Languages of the World. 28th ed. Dallas: SIL International. [https://www.ethnologue.com](https://www.ethnologue.com) (last accessed on 21/12/2025). [↗](https://www.ethnologue.com)
14. Grattafiori, Aaron & Dubey, Abhimanyu & Jauhri, Abhinav & Pandey, Abhinav & Kadian, Abhishek & Al-Dahle, Ahmad & Letman, Aiesha et al. 2024. [The Llama 3 herd of models](https://md.research.elamur.ai/free/en/papers/grattafiori-2024-llama-3-herd-models/). arXiv. https:/doi.org/10.48550/arXiv.2407.21783 [↗](https://doi.org/10.48550/arxiv.2407.21783)
15. Guo, Yanzhu & Conia, Simone & Zhou, Zelin & Li, Min & Potdar, Saloni & Xiao, Henry. 2025. [Do Large Language Models have an English accent? Evaluating and improving the naturalness of multilingual LLMs](https://md.research.elamur.ai/free/en/papers/guo-2025-do-large-language-models-have/). In Che, Wanxiang & Nabende, Joyce & Shutova, Ekaterina & Pilehvar, Mohammad Taher (eds.), Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 3823–3838. Vienna: Association for Computational Linguistics. https:/doi.org/10.18653/v1/2025.acl-long.193 [↗](https://doi.org/10.18653/v1/2025.acl-long.193)
16. Hofstätter, Alexandra. 2020. Entwicklungen in der deutschen Nachrichtensprache. PhD. Dissertation. München: Ludwig-Maximilians-Universität München. [↗](https://doi.org/10.5282/edoc.26652)
17. Holmberg, Anders. 2015. Verb Second. In Kiss, Tibor & Alexiadou, Artemis (eds), Syntax – Theory and Analysis. An International Handbook. Volume 1, 342– 383. Berlin: De Gruyter Mouton. https:/doi.org/ 10.1515/9783110377408.342 [↗](https://doi.org/10.1515/9783110377408.342)
18. Informationsgemeinschaft zur Feststellung der Verbreitung von Werbeträgern e. V. (IVW). 2024. IVW - Quartalsausweisung. [https://quartalsausweisung.ivw.de/(last](https://quartalsausweisung.ivw.de/(last) accessed on 18/08/2025). [↗](https://quartalsausweisung.ivw.de/(last)
19. Johnson, Rebecca L. & Pistilli, Giada & Menédez-González, Natalia & Duran, Leslye Denisse Dias & Panai, Enrico & Kalpokiene, Julija & Bertulfo, Donald Jay. 2022. The ghost in the machine has an American accent: value conflict in GPT-3. arXiv. https:/doi.org/10.48550/arXiv.2203.07785 [↗](https://doi.org/10.48550/arxiv.2203.07785)
20. Koch, Peter & Oesterreicher, Wulf. 1985. Sprache der Nähe – Sprache der Distanz. Mündlichkeit und Schriftlichkeit im Spannungsfeld von Sprachtheorie und Sprachgeschichte. Romanistisches Jahrbuch 36(1). 15–43. https:/doi.org/10.1515/9783110244922.15 [↗](https://doi.org/10.1515/9783110244922.15)
21. Kurz, Josef & Müller, Daniel & Pötschke, Joachim & Pöttker, Horst & Gehr, Martin. 2010. Stilistik für Journalisten. Wiesbaden: VS Verlag für Sozialwissenschaften. https:/doi.org/10.1007/978-3-531-92492-2 [↗](https://doi.org/10.1007/978-3-531-92492-2)
22. Linden, Peter. 2000. Wie Texte wirken: Anleitung zur Analyse journalistischer Sprache. 2., aktualisierte und erw. Aufl. Berlin: ZV. [find](https://scholar.google.com/scholar?q=Linden%2C+Peter%2E+2000%2E+Wie+Texte+wirken%3A+Anleitung+zur+Analyse+journalistischer+Sprache%2E+2%2E%2C+aktualisierte+und+erw%2E+Aufl%2E+Berlin%3A+ZV%2E)
23. Linden, Peter. 2023. Handbuch Stilsicher schreiben: Wie Sie wirkungsvolle Texte formulieren. Berlin: Duden. [find](https://scholar.google.com/scholar?q=Linden%2C+Peter%2E+2023%2E+Handbuch+Stilsicher+schreiben%3A+Wie+Sie+wirkungsvolle+Texte+formulieren%2E+Berlin%3A+Duden%2E)
24. LiveBench. [https://livebench.ai/#/](https://livebench.ai/#/) (last accessed on 18/08/2025). [↗](https://livebench.ai/#/)
25. LMArena. [https://lmarena.ai/](https://lmarena.ai/) (last accessed on 18/08/2025). [↗](https://lmarena.ai/)
26. Lünenborg, Margreth. 2013. Boulevardisierung. In Bentele, Günter & Brosius, Hans-Bernd & Jarren, Otfried (Hrsg.). Lexikon Kommunikations- und Medienwissenschaft. Wiesbaden: Springer. 33. [find](https://scholar.google.com/scholar?q=L%C3%BCnenborg%2C+Margreth%2E+2013%2E+Boulevardisierung%2E+In+Bentele%2C+G%C3%BCnter+%26+Brosius%2C+Hans%2DBernd+%26+Jarren%2C+Otfried+%28Hrsg%2E%29%2E+Lexikon+Kommunikations%2D+und+Medienwissenschaft%2E+Wiesbaden%3A+Springer%2E+33%2E)
27. Matthies, Jochen. 2025. Agensdemotion und wissenschaftliche Verfasserreferenz im Kontext des KI-gestützten Schreibens – Eine Pilotstudie. In Lykhnenko, Olha & Matthies, Jochen & Roelcke, Thorsten & Teufele, Lisa (eds.), Deutsch als Fremdsprache – Mehrsprachigkeit, Fachkommunikation, Digitalisierung und Kultur: Internationale Nachwuchstagung 2024 an der Technischen Universität Berlin, 91–114. Berlin, Heidelberg: Springer. https:/doi.org/10.1007/978-3-662-70351-9_5 [↗](https://doi.org/10.1007/978-3-662-70351-9_5)
28. McEnery, Tony & Hardie, Andrew. 2012. Corpus linguistics: Method, Theory and Practice. Cambridge: Cambridge University Press. [find](https://scholar.google.com/scholar?q=McEnery%2C+Tony+%26+Hardie%2C+Andrew%2E+2012%2E+Corpus+linguistics%3A+Method%2C+Theory+and+Practice%2E+Cambridge%3A+Cambridge+University+Press%2E)
29. Meyn, Hermann. 1985. Massenmedien in der Bundesrepublik Deutschland. Überarbeitete Neuauflage. Berlin: Colloquium Verlag. [find](https://scholar.google.com/scholar?q=Meyn%2C+Hermann%2E+1985%2E+Massenmedien+in+der+Bundesrepublik+Deutschland%2E+%C3%9Cberarbeitete+Neuauflage%2E+Berlin%3A+Colloquium+Verlag%2E)
30. Mittelberg, Ekkehart. 1967. Wortschatz und Syntax der Bild-Zeitung. Marburger Beiträge zur Germanistik. Marburg: N. G. Elwert Verlag. [find](https://scholar.google.com/scholar?q=Mittelberg%2C+Ekkehart%2E+1967%2E+Wortschatz+und+Syntax+der+Bild%2DZeitung%2E+Marburger+Beitr%C3%A4ge+zur+Germanistik%2E+Marburg%3A+N%2E+G%2E+Elwert+Verlag%2E)
31. Newman, Nic & Fletcher, Richard & Eddy, Kirsten & Robertson, Craig T. & Nielsen, Rasmus Kleis. 2023. [find](https://scholar.google.com/scholar?q=Newman%2C+Nic+%26+Fletcher%2C+Richard+%26+Eddy%2C+Kirsten+%26+Robertson%2C+Craig+T%2E+%26+Nielsen%2C+Rasmus+Kleis%2E+2023%2E)
32. Reuters Institute Digital News Report 2023. Oxford: Reuters Institute for the Study of Journalism. [find](https://scholar.google.com/scholar?q=Reuters+Institute+Digital+News+Report+2023%2E+Oxford%3A+Reuters+Institute+for+the+Study+of+Journalism%2E)
33. Newman, Nic & Fletcher, Richard & Robertson, Craig T. & Ross Arguedas, Amy & Nielsen, Rasmus Kleis. 2024. [find](https://scholar.google.com/scholar?q=Newman%2C+Nic+%26+Fletcher%2C+Richard+%26+Robertson%2C+Craig+T%2E+%26+Ross+Arguedas%2C+Amy+%26+Nielsen%2C+Rasmus+Kleis%2E+2024%2E)
34. Reuters Institute Digital News Report 2024. Oxford: Reuters Institute for the Study of Journalism. [find](https://scholar.google.com/scholar?q=Reuters+Institute+Digital+News+Report+2024%2E+Oxford%3A+Reuters+Institute+for+the+Study+of+Journalism%2E)
35. Newman, Nic & Ross Arguedas, Amy & Robertson, Craig T. & Nielsen, Rasmus Kleis & Fletcher, Richard. 2025. [find](https://scholar.google.com/scholar?q=Newman%2C+Nic+%26+Ross+Arguedas%2C+Amy+%26+Robertson%2C+Craig+T%2E+%26+Nielsen%2C+Rasmus+Kleis+%26+Fletcher%2C+Richard%2E+2025%2E)
36. Reuters Institute Digital News Report 2025. Oxford: Reuters Institute for the Study of Journalism. [find](https://scholar.google.com/scholar?q=Reuters+Institute+Digital+News+Report+2025%2E+Oxford%3A+Reuters+Institute+for+the+Study+of+Journalism%2E)
37. OpenAI. 2023. OpenAI Data Partnerships. [https://openai.com/index/data-partnerships/](https://openai.com/index/data-partnerships/) (last accessed on 18/08/2025). [↗](https://openai.com/index/data-partnerships/)
38. OpenAI. 2025. Best practices for prompt engineering with the OpenAI API. OpenAI Help Center. [https://help.openai.com/en/articles/6654000-best-practices-for-prompt-engineering-with-the-openai-api?](https://help.openai.com/en/articles/6654000-best-practices-for-prompt-engineering-with-the-openai-api?) (last accessed on 18/08/2025). [↗](https://help.openai.com/en/articles/6654000-best-practices-for-prompt-engineering-with-the-openai-api?)
39. Phoenix, James & Taylor, Mike. 2024. Prompt Engineering for Generative AI. O’Reilly Media, Inc. [find](https://scholar.google.com/scholar?q=Phoenix%2C+James+%26+Taylor%2C+Mike%2E+2024%2E+Prompt+Engineering+for+Generative+AI%2E+O%E2%80%99Reilly+Media%2C+Inc%2E)
40. Qin, Libo & Chen, Qiguang & Zhou, Yuhang & Chen, Zhi & Li, Yinghui & Liao, Lizi & Li, Min & Che, Wanxiang & Yu, Philip S. 2025. A survey of multilingual large language models. Patterns 6(1). 101118. https:/doi.org/10.1016/j.patter.2024.101118 [↗](https://doi.org/10.1016/j.patter.2024.101118)
41. Raabe, Johannes. 2013a. Boulevardpresse. In Bentele, Günter & Brosius, Hans-Bernd & Jarren, Otfried (Hrsg.). Lexikon Kommunikations- und Medienwissenschaft. Wiesbaden: Springer. 33–34. [find](https://scholar.google.com/scholar?q=Raabe%2C+Johannes%2E+2013a%2E+Boulevardpresse%2E+In+Bentele%2C+G%C3%BCnter+%26+Brosius%2C+Hans%2DBernd+%26+Jarren%2C+Otfried+%28Hrsg%2E%29%2E+Lexikon+Kommunikations%2D+und+Medienwissenschaft%2E+Wiesbaden%3A+Springer%2E+33%E2%80%9334%2E)
42. Raabe, Johannes. 2013b. Qualitätszeitungen. In Bentele, Günter & Brosius, Hans-Bernd & Jarren, Otfried (Hrsg.). Lexikon Kommunikations- und Medienwissenschaft. Wiesbaden: Springer. 288. [find](https://scholar.google.com/scholar?q=Raabe%2C+Johannes%2E+2013b%2E+Qualit%C3%A4tszeitungen%2E+In+Bentele%2C+G%C3%BCnter+%26+Brosius%2C+Hans%2DBernd+%26+Jarren%2C+Otfried+%28Hrsg%2E%29%2E+Lexikon+Kommunikations%2D+und+Medienwissenschaft%2E+Wiesbaden%3A+Springer%2E+288%2E)
43. Schmitt, Uta. 2004. Diskurspragmatik und Syntax: die Funktionale Satzperspektive in der französischen und deutschen Tagespresse unter Berücksichtigung einzelsprachlicher, pressetyp- und textklassenabhängiger Spezifika. Frankfurt am Main etc.: Peter Lang. [find](https://scholar.google.com/scholar?q=Schmitt%2C+Uta%2E+2004%2E+Diskurspragmatik+und+Syntax%3A+die+Funktionale+Satzperspektive+in+der+franz%C3%B6sischen+und+deutschen+Tagespresse+unter+Ber%C3%BCcksichtigung+einzelsprachlicher%2C+pressetyp%2D+und+textklassenabh)
44. Shi, Freda & Suzgun, Mirac & Freitag, Markus & Wang, Xuezhi & Srivats, Suraj & Vosoughi, Soroush & Chung, Hyung Won et al. 2022. [Language Models are Multilingual Chain-of-thought Reasoners](https://md.research.elamur.ai/free/en/papers/shi-2022-language-models-multilingual-chain-thought/). ICLR 2023. [↗](https://doi.org/10.48550/arxiv.2210.03057)
45. SPIEGEL-Verlag (ed.). 2021. Die SPIEGEL-Standards: Haltung, Recherche, Schreibweise. Hamburg: SPIEGEL-Verlag. [find](https://scholar.google.com/scholar?q=SPIEGEL%2DVerlag+%28ed%2E%29%2E+2021%2E+Die+SPIEGEL%2DStandards%3A+Haltung%2C+Recherche%2C+Schreibweise%2E+Hamburg%3A+SPIEGEL%2DVerlag%2E)
46. Tavosanis, Mirko. 2024. Valutare la qualità dei testi generati in lingua italiana. AI-Linguistica. Linguistic Studies on AI-Generated Texts and Discourses 1(1). https:/doi.org/10.62408/ai-ling.v1i1.14 [↗](https://doi.org/10.62408/ai-ling.v1i1.14)
47. Weinrich, Harald. 2007. Textgrammatik der deutschen Sprache. 4. Aufl. Hildesheim: Olms. [find](https://scholar.google.com/scholar?q=Weinrich%2C+Harald%2E+2007%2E+Textgrammatik+der+deutschen+Sprache%2E+4%2E+Aufl%2E+Hildesheim%3A+Olms%2E)
48. White, Colin & Dooley, Samuel & Roberts, Manley & Pal, Arka & Feuer, Ben & Jain, Siddhartha & Shwartz-Ziv, Ravid et al. 2025. LiveBench: A Challenging, Contamination-Limited LLM Benchmark. arXiv. https:/doi.org/10.48550/arXiv.2406.19314 [↗](https://doi.org/10.48550/arxiv.2406.19314)
49. Wöllstein, Angelika. 2022. Duden: die Grammatik (Der Duden in zwölf Bänden, Band 4). 10., völlig neu verfasste Auflage. Berlin: Dudenverlag. [find](https://scholar.google.com/scholar?q=W%C3%B6llstein%2C+Angelika%2E+2022%2E+Duden%3A+die+Grammatik+%28Der+Duden+in+zw%C3%B6lf+B%C3%A4nden%2C+Band+4%29%2E+10%2E%2C+v%C3%B6llig+neu+verfasste+Auflage%2E+Berlin%3A+Dudenverlag%2E)
50. Zheng, Lianmin & Chiang, Wei-Lin & Sheng, Ying & Li, Tianle & Zhuang, Siyuan & Wu, Zhanghao & Zhuang, Yonghao et al. 2024. [LMSYS-Chat-1M: A large-scale real-world LLM conversation dataset](https://md.research.elamur.ai/free/en/papers/zheng-2023-lmsys-chat-1m-large-scale/). arXiv. https:/doi.org/10.48550/arXiv.2309.11998 Abbreviations B: C: L: M: O: SZ: BILD Anthropic’s Claude 3.5 Sonnet Meta’s llama-3.1-405b-instruct Mistral’s mistral-large-2407 OpenAI’s o1 Süddeutsche Zeitung [↗](https://doi.org/10.48550/arxiv.2309.11998)

*Updated 1 November 2026*