# The Cross-Lingual Cost: Retrieval Biases in RAG over Arabic-English Corpora

arXiv: https://arxiv.org/abs/2507.07543 · HTML: https://mlsanity.com/paper/2507.07543/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Generation and context integration, one of 15 directions in Retrieval-augmented generation.
- Builds on 11 of its 29 references that are in the map.
- First posted on arXiv 10 Jul 2025.

## Placed in

- Retrieval-augmented generation / [Generation and context integration](https://mlsanity.com/rag/generation-and-context-integration/)

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

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation](https://mlsanity.com/paper/2402.03216/) (2024)
- [MLQA: Evaluating Cross-lingual Extractive Question Answering](https://mlsanity.com/paper/1910.07475/) (2019)
- [MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answering](https://mlsanity.com/paper/2007.15207/) (2020)
- [XOR QA: Cross-lingual Open-Retrieval Question Answering](https://mlsanity.com/paper/2010.11856/) (2020)
- [Llama2Vec: Unsupervised Adaptation of Large Language Models for Dense Retrieval](https://mlsanity.com/paper/2312.15503/) (2023)
- [One Question Answering Model for Many Languages with Cross-lingual Dense Passage Retrieval](https://mlsanity.com/paper/2107.11976/) (2021)
- [Understanding Cross-Lingual Alignment -- A Survey](https://mlsanity.com/paper/2404.06228/) (2024)
- [Don't Use LLMs to Make Relevance Judgments](https://mlsanity.com/paper/2409.15133/) (2024)
- [Investigating Language Preference of Multilingual RAG Systems](https://mlsanity.com/paper/2502.11175/) (2025)
- [Multilingual Retrieval-Augmented Generation for Knowledge-Intensive Task](https://mlsanity.com/paper/2504.03616/) (2025)

## Built on by (most cited first, 5 of 5)

- [Linguistic Nepotism: Trading-off Quality for Language Preference in Multilingual RAG](https://mlsanity.com/paper/2509.13930/) (2025)
- [All Languages Matter: Understanding and Mitigating Language Bias in Multilingual RAG](https://mlsanity.com/paper/2604.20199/) (2026)
- [ChemCLIR-Bench: Benchmarking Cross-Lingual Information Retrieval in Multilingual Chemical Patents](https://mlsanity.com/paper/2609.23231/) (2026)
- [Where's Waldo? Query-language Preference under Cross-lingual Knowledge Disparities](https://mlsanity.com/paper/2610.00606/) (2026)
- [Retrieval Disagreement as a Signal: Diagnosing and Steering Multilingual Dense Embedders](https://mlsanity.com/paper/2610.18526/) (2026)

## Abstract (arXiv)

Cross-lingual retrieval-augmented generation (RAG) is a critical capability for retrieving and generating answers across languages. Prior work in this context has mostly focused on generation and relied on benchmarks derived from open-domain sources, most notably Wikipedia. In such settings, retrieval challenges often remain hidden due to language imbalances, overlap with pretraining data, and memorized content. To address this gap, we study Arabic-English RAG in a domain-specific setting using benchmarks derived from real-world corporate datasets. Our benchmarks include all combinations of languages for the user query and the supporting document, drawn independently and uniformly at random. This enables a systematic study of multilingual retrieval behavior. Our findings reveal that retrieval is a critical bottleneck in cross-lingual domain-specific scenarios, with substantial performance drops occurring when the user query and supporting document languages differ. A key insight is that these failures stem primarily from the retriever's difficulty in ranking documents across languages. Finally, we propose two simple retrieval strategies that address this source of failure by enforcing equal retrieval from both languages or by translating the query, resulting in substantial improvements in cross-lingual and overall performance. These results highlight meaningful opportunities for improving multilingual retrieval, particularly in practical, real-world RAG applications.