# How Surprising Is Historical Italian to Language Models? Tokenization Tax, Comprehension Tax, and a Simple Mitigation

arXiv: https://arxiv.org/abs/2606.27275 · HTML: https://mlsanity.com/paper/2606.27275/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 16 references that are in the map.
- First posted on arXiv 25 Jun 2026.

## Builds on (most cited first, 9 of 9)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Neural Machine Translation of Rare Words with Subword Units](https://mlsanity.com/paper/1508.07909/) (2015)
- [Extracting Training Data from Large Language Models](https://mlsanity.com/paper/2012.07805/) (2020)
- [SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing](https://mlsanity.com/paper/1808.06226/) (2018)
- [Diachronic Word Embeddings Reveal Statistical Laws of Semantic Change](https://mlsanity.com/paper/1605.09096/) (2016)
- [Diachronic word embeddings and semantic shifts: a survey](https://mlsanity.com/paper/1806.03537/) (2018)
- [Set the Clock: Temporal Alignment of Pretrained Language Models](https://mlsanity.com/paper/2402.16797/) (2024)
- [Tokenization is Sensitive to Language Variation](https://mlsanity.com/paper/2502.15343/) (2025)
- [Mapping Overlaps in Benchmarks through Perplexity in the Wild](https://mlsanity.com/paper/2509.23488/) (2025)

## Abstract (arXiv)

Large language models (LLMs) are increasingly critical to digital library workflows, yet their ability to process historical language remains poorly understood. Historical difficulty is typically treated as a monolithic barrier, conflating orthographic variation, linguistic distance, and pretraining exposure. In this paper, we propose a diagnostic framework that decomposes this difficulty into four distinct dimensions: tokenization cost, predictive uncertainty (surprisal), semantic robustness, and context sensitivity. We evaluate this framework on three datasets spanning three centuries: (1) a newly curated corpus of 17th-century Italian texts (1610-1689) digitized from original page images; (2) canonical 19th-century Italian "I Promessi Sposi" serving as a high-exposure control; and (3) 18th-century Russian civil print books as a contrastive orthographic stress test. Our results reveal a distinct dissociation between encoding cost and comprehension. While Russian and early modern Italian incur comparable tokenization penalties (25-30% inflation), their predictive difficulty diverges sharply. 17th-century Italian is on average 2.4 times more surprising than its modern equivalent - with academic prose reaching 3.2 times - whereas Russian shows only a modest increase. But predictive uncertainty does not imply representational degradation: embedding similarity remains robust (> 0.85) across all datasets, confirming that models can represent historical meaning even when generation is unstable. Finally, we demonstrate that a minimal temporal context prompt reduces historical surprisal by approximately 60%, offering a simple, model-agnostic mitigation. These findings suggest that while historical text imposes a consistent encoding tax, digital libraries can safely deploy LLMs for semantic retrieval tasks, provided that generative applications are carefully adapted.