# Speaking Multiple Languages Affects the Moral Bias of Language Models

arXiv: https://arxiv.org/abs/2211.07733 · HTML: https://mlsanity.com/paper/2211.07733/

## Key facts

- Cited by 34 papers in the map (data through 14 Oct 2026).
- Builds on 36 of its 64 references that are in the map.
- First posted on arXiv 14 Nov 2022.

## Builds on (most cited first, 30 of 36)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference](https://mlsanity.com/paper/1704.05426/) (2017)
- [Language Models as Knowledge Bases?](https://mlsanity.com/paper/1909.01066/) (2019)
- [COMET: A Neural Framework for MT Evaluation](https://mlsanity.com/paper/2009.09025/) (2020)
- [SemEval-2017 Task 1: Semantic Textual Similarity - Multilingual and Cross-lingual Focused Evaluation](https://mlsanity.com/paper/1708.00055/) (2017)
- [XNLI: Evaluating Cross-lingual Sentence Representations](https://mlsanity.com/paper/1809.05053/) (2018)
- [Semantics derived automatically from language corpora contain human-like biases](https://mlsanity.com/paper/1608.07187/) (2016)
- [Aligning AI With Shared Human Values](https://mlsanity.com/paper/2008.02275/) (2020)
- [Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillation](https://mlsanity.com/paper/2004.09813/) (2020)
- [CCNet: Extracting High Quality Monolingual Datasets from Web Crawl Data](https://mlsanity.com/paper/1911.00359/) (2019)
- [AraBERT: Transformer-based Model for Arabic Language Understanding](https://mlsanity.com/paper/2003.00104/) (2020)
- [Social Chemistry 101: Learning to Reason about Social and Moral Norms](https://mlsanity.com/paper/2011.00620/) (2020)
- [Negated and Misprimed Probes for Pretrained Language Models: Birds Can Talk, But Cannot Fly](https://mlsanity.com/paper/1911.03343/) (2019)
- [Efficient Natural Language Response Suggestion for Smart Reply](https://mlsanity.com/paper/1705.00652/) (2017)
- [Challenges and Strategies in Cross-Cultural NLP](https://mlsanity.com/paper/2203.10020/) (2022)
- [Commonsense Knowledge Mining from Pretrained Models](https://mlsanity.com/paper/1909.00505/) (2019)
- [Process for Adapting Language Models to Society (PALMS) with Values-Targeted Datasets](https://mlsanity.com/paper/2106.10328/) (2021)
- [Large Pre-trained Language Models Contain Human-like Biases of What is Right and Wrong to Do](https://mlsanity.com/paper/2103.11790/) (2021)
- [Probing Pre-Trained Language Models for Cross-Cultural Differences in Values](https://mlsanity.com/paper/2203.13722/) (2022)
- [German's Next Language Model](https://mlsanity.com/paper/2010.10906/) (2020)
- [Moral Stories: Situated Reasoning about Norms, Intents, Actions, and their Consequences](https://mlsanity.com/paper/2012.15738/) (2020)
- [Multilingual Alignment of Contextual Word Representations](https://mlsanity.com/paper/2002.03518/) (2020)
- [On the Language Neutrality of Pre-trained Multilingual Representations](https://mlsanity.com/paper/2004.05160/) (2020)
- [Inducing Language-Agnostic Multilingual Representations](https://mlsanity.com/paper/2008.09112/) (2020)
- [Improving Pretrained Cross-Lingual Language Models via Self-Labeled Word Alignment](https://mlsanity.com/paper/2106.06381/) (2021)
- [Dual Conditional Cross-Entropy Filtering of Noisy Parallel Corpora](https://mlsanity.com/paper/1809.00197/) (2018)
- [RobeCzech: Czech RoBERTa, a monolingual contextualized language representation model](https://mlsanity.com/paper/2105.11314/) (2021)

## Built on by (most cited first, 30 of 34)

- [Assessing Cross-Cultural Alignment between ChatGPT and Human Societies: An Empirical Study](https://mlsanity.com/paper/2303.17466/) (2023)
- [Having Beer after Prayer? Measuring Cultural Bias in Large Language Models](https://mlsanity.com/paper/2305.14456/) (2023)
- [NLPositionality: Characterizing Design Biases of Datasets and Models](https://mlsanity.com/paper/2306.01943/) (2023)
- [Are Multilingual LLMs Culturally-Diverse Reasoners? An Investigation into Multicultural Proverbs and Sayings](https://mlsanity.com/paper/2309.08591/) (2023)
- [Goldfish: Monolingual Language Models for 350 Languages](https://mlsanity.com/paper/2408.10441/) (2024)
- [Exploring Multilingual Concepts of Human Value in Large Language Models: Is Value Alignment Consistent, Transferable and Controllable across Languages?](https://mlsanity.com/paper/2402.18120/) (2024)
- [DOSA: A Dataset of Social Artifacts from Different Indian Geographical Subcultures](https://mlsanity.com/paper/2403.14651/) (2024)
- [The Potential and Challenges of Evaluating Attitudes, Opinions, and Values in Large Language Models](https://mlsanity.com/paper/2406.11096/) (2024)
- [Bridging Cultural Nuances in Dialogue Agents through Cultural Value Surveys](https://mlsanity.com/paper/2401.10352/) (2024)
- [Whose Morality Do They Speak? Unraveling Cultural Bias in Multilingual Language Models](https://mlsanity.com/paper/2412.18863/) (2024)
- [Values, Ethics, Morals? On the Use of Moral Concepts in NLP Research](https://mlsanity.com/paper/2310.13915/) (2023)
- [Decoding Multilingual Moral Preferences: Unveiling LLM's Biases Through the Moral Machine Experiment](https://mlsanity.com/paper/2407.15184/) (2024)
- [Decolonial AI Alignment: Openness, Vi'sedsa-Dharma, and Including Excluded Knowledges](https://mlsanity.com/paper/2309.05030/) (2023)
- [The Echoes of Multilinguality: Tracing Cultural Value Shifts during LM Fine-tuning](https://mlsanity.com/paper/2405.12744/) (2024)
- [CURE-Med: Curriculum-Informed Reinforcement Learning for Multilingual Medical Reasoning](https://mlsanity.com/paper/2601.13262/) (2026)
- [A "Perspectival" Mirror of the Elephant: Investigating Language Bias on Google, ChatGPT, YouTube, and Wikipedia](https://mlsanity.com/paper/2303.16281/) (2023)
- [Analyzing Political Bias in LLMs via Target-Oriented Sentiment Classification](https://mlsanity.com/paper/2505.19776/) (2025)
- [Unveiling and Mitigating Bias in Large Language Model Recommendations: A Path to Fairness](https://mlsanity.com/paper/2409.10825/) (2024)
- [Made-in China, Thinking in America:U.S. Values Persist in Chinese LLMs](https://mlsanity.com/paper/2512.13723/) (2025)
- [Histoires Morales: A French Dataset for Assessing Moral Alignment](https://mlsanity.com/paper/2501.17117/) (2025)
- [A Survey on Moral Foundation Theory and Pre-Trained Language Models: Current Advances and Challenges](https://mlsanity.com/paper/2409.13521/) (2024)
- [Scopes of Alignment](https://mlsanity.com/paper/2501.12405/) (2025)
- [How Gender Interacts with Political Values: A Case Study on Czech BERT Models](https://mlsanity.com/paper/2403.13514/) (2024)
- [Better Aligned with Survey Respondents or Training Data? Unveiling Political Leanings of LLMs on U.S. Supreme Court Cases](https://mlsanity.com/paper/2502.18282/) (2025)
- [Camellia: Benchmarking Cultural Biases in LLMs for Asian Languages](https://mlsanity.com/paper/2510.05291/) (2025)
- [The Moralization Corpus: Frame-Based Annotation and Analysis of Moralizing Speech Acts across Diverse Text Genres](https://mlsanity.com/paper/2512.15248/) (2025)
- [Leveraging Wikidata for Geographically Informed Sociocultural Bias Dataset Creation: Application to Latin America](https://mlsanity.com/paper/2603.10001/) (2026)
- [Beyond English: Evaluating Automated Measurement of Moral Foundations in Non-English Discourse with a Chinese Case Study](https://mlsanity.com/paper/2502.02451/) (2025)
- [Do LLMs Need to Think in One Language? Correlation between Latent Language and Task Performance](https://mlsanity.com/paper/2505.21458/) (2025)
- [Is a Prestigious Job the same as a Prestigious Country? A Case Study on Multilingual Sentence Embeddings and European Countries](https://mlsanity.com/paper/2305.14482/) (2023)

## Abstract (arXiv)

Pre-trained multilingual language models (PMLMs) are commonly used when dealing with data from multiple languages and cross-lingual transfer. However, PMLMs are trained on varying amounts of data for each language. In practice this means their performance is often much better on English than many other languages. We explore to what extent this also applies to moral norms. Do the models capture moral norms from English and impose them on other languages? Do the models exhibit random and thus potentially harmful beliefs in certain languages? Both these issues could negatively impact cross-lingual transfer and potentially lead to harmful outcomes. In this paper, we (1) apply the MoralDirection framework to multilingual models, comparing results in German, Czech, Arabic, Chinese, and English, (2) analyse model behaviour on filtered parallel subtitles corpora, and (3) apply the models to a Moral Foundations Questionnaire, comparing with human responses from different countries. Our experiments demonstrate that, indeed, PMLMs encode differing moral biases, but these do not necessarily correspond to cultural differences or commonalities in human opinions. We release our code and models.