# When Simpler Is Better: Evaluating Translation Pipelines for Medieval Latin Manuscripts

arXiv: https://arxiv.org/abs/2607.03836 · HTML: https://mlsanity.com/paper/2607.03836/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Document understanding and multimodal RAG, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 15 of its 41 references that are in the map.
- First posted on arXiv 4 Jul 2026.

## Placed in

- Multimodal LLMs / vision-language models / [Document understanding and multimodal RAG](https://mlsanity.com/mllm/document-understanding-and-multimodal-rag/)

## Builds on (most cited first, 15 of 15)

- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://mlsanity.com/paper/2105.13626/) (2021)
- [Achieving Human Parity on Automatic Chinese to English News Translation](https://mlsanity.com/paper/1803.05567/) (2018)
- [olmOCR 2: Unit Test Rewards for Document OCR](https://mlsanity.com/paper/2510.19817/) (2025)
- [Mitigating Copy Bias in In-Context Learning through Neuron Pruning](https://mlsanity.com/paper/2410.01288/) (2024)
- [Language Drift in Multilingual Retrieval-Augmented Generation: Characterization and Decoding-Time Mitigation](https://mlsanity.com/paper/2511.09984/) (2025)
- [Measuring the Effect of Transcription Noise on Downstream Language Understanding Tasks](https://mlsanity.com/paper/2502.13645/) (2025)
- [Translating Hanja Historical Documents to Contemporary Korean and English](https://mlsanity.com/paper/2205.10019/) (2022)
- [Not All Languages are Equal: Insights into Multilingual Retrieval-Augmented Generation](https://mlsanity.com/paper/2410.21970/) (2024)
- [An Assessment of the Impact of OCR Noise on Language Models](https://mlsanity.com/paper/2202.00470/) (2022)
- [TRIDIS: A Comprehensive Medieval and Early Modern Corpus for HTR and NER](https://mlsanity.com/paper/2503.22714/) (2025)
- [LITERA: An LLM Based Approach to Latin-to-English Translation](https://mlsanity.com/paper/2504.10660/) (2025)
- [SozKZ: Training Efficient Small Language Models for Kazakh from Scratch](https://mlsanity.com/paper/2603.20854/) (2026)
- [Towards Inclusive NLP: Assessing Compressed Multilingual Transformers across Diverse Language Benchmarks](https://mlsanity.com/paper/2507.19699/) (2025)

## Abstract (arXiv)

Despite remarkable progress in machine translation, Vision Language Models (VLMs) struggle on historical manuscripts, a domain that stresses core Natural Language Processing (NLP) capabilities: low-resource transliteration, archaic vocabulary, and noisy input signals. We present a systematic framework for evaluating the full image-to-translation pipeline on medieval Latin manuscripts, a setting in which scribal shorthand, ligatures, and parchment degradation expose failure modes that are invisible in clean-text benchmarks. Benchmarking on the CATMuS Latin dataset reveals a specialization gap: domain-specific Optical Character Recognition (OCR) models reduce character error rate by up to 4.3× compared to general-purpose VLMs, despite operating at orders of magnitude fewer parameters. We introduce the Interpres-Parallel-Corpus (IPC), a novel dataset comprising 1,383 aligned manuscript image lines, transcriptions, and expert translations, the first of its kind for medieval Latin. Our experiments uncover a complexity paradox: the simplest pipeline, a specialized OCR model feeding directly into a VLM, outperforms all multi-component variants. Adding retrieval-augmented generation (RAG) or post-OCR correction introduces prompt saturation and error propagation that degrade aggregate translation quality. These findings offer both a new benchmark and practical guidance for deploying translation systems in low-resource historical settings.