# An Investigation of Translationese in the Generations of Multilingual Large Language Models

arXiv: https://arxiv.org/abs/2608.17399 · HTML: https://mlsanity.com/paper/2608.17399/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 31 references that are in the map.
- First posted on arXiv 18 Aug 2026.

## Builds on (most cited first, 11 of 11)

- [PaLM 2 Technical Report](https://mlsanity.com/paper/2305.10403/) (2023)
- [The State and Fate of Linguistic Diversity and Inclusion in the NLP World](https://mlsanity.com/paper/2004.09095/) (2020)
- [Do Multilingual Language Models Think Better in English?](https://mlsanity.com/paper/2308.01223/) (2023)
- [Do Multilingual LLMs Think In English?](https://mlsanity.com/paper/2502.15603/) (2025)
- [Translationese as a Language in "Multilingual" NMT](https://mlsanity.com/paper/1911.03823/) (2019)
- [Post-editese: an Exacerbated Translationese](https://mlsanity.com/paper/1907.00900/) (2019)
- [A Survey on Human Preference Learning for Large Language Models](https://mlsanity.com/paper/2406.11191/) (2024)
- [Lost in Literalism: How Supervised Training Shapes Translationese in LLMs](https://mlsanity.com/paper/2503.04369/) (2025)
- [Found in Translation: Reconstructing Phylogenetic Language Trees from Translations](https://mlsanity.com/paper/1704.07146/) (2017)
- [Measuring Spurious Correlation in Classification: 'Clever Hans' in Translationese](https://mlsanity.com/paper/2308.13170/) (2023)
- [Translating away Translationese without Parallel Data](https://mlsanity.com/paper/2310.18830/) (2023)

## Abstract (arXiv)

Text which has been translated from another language tends to carry with it evidence of translationunicodex2014hence, it is often referred to as textittranslationese. Multilingual large language models (MLLMs) generate text in a variety of languages. However, it is still unclear if MLLMs' generations resemble internal translation (from English or, potentially, other languages) and, thus, result in translationese. Here, we ask the following research questions: (1) Does text generated by MLLMs resemble translationese? (2) How does translationese produced by MLLMs differ from translationese produced through direct translation? We leverage established indicators of translated text to evaluate text generated by state-of-the-art MLLMs in five languages, comparing to both non-translated and human-written baselines in order to isolate translationese from other kinds of interference. Through the use of high-accuracy classification models, analyses of variance on individual linguistic features, and the collection of human annotations in a subset of two languages (German and Spanish), we assess the translationese content of MLLM generations and examine the key features that distinguish MLLM-generated text from typical translation-related interference.