# Machine-learning translation trained on the existing translated inscriptions cannot go beyond the formulae that are…

> Still possible: Machine-learning translation trained on the existing translated inscriptions cannot go beyond the formulae that are already understood.

Source: https://mysteriesdept.com/mysteries/etruscan-language/answers/ANS0DEEBGH0MHWY/machine-learning-translation-trained-on-the-existing  
Publisher: Mysteries Department  
Licence: CC BY 4.0 (credit Mysteries Department and link the page)  

---

[Questions](https://mysteriesdept.com/questions)[Etruscan: the readable language that cannot be fully translated](https://mysteriesdept.com/mysteries/etruscan-language)[How far can the combinatory method and computational analysis extend the roughly 250 securely understood Etruscan words?](https://mysteriesdept.com/mysteries/etruscan-language/questions/QST0DEDYMXF5X5P/how-far-can-the-combinatory-method-and-computational)Answer

Answer · Etruscan: the readable language that cannot be fully translated

# Machine-learning translation trained on the existing translated inscriptions cannot go beyond the formulae that are…

Machine-learning translation trained on the existing translated inscriptions cannot go beyond the formulae that are already understood.

**Still possible** Answers **1** question0

## Where it stands

**Still possible** One piece of evidence for, none against. It needs two reviews to be accepted.

## What it predicts

- Models trained on the translated set will fail on the long texts, producing outputs no better than word-by-word glossaries.

## What would rule it out

- A model proposes a meaning for an unglossed word that is later confirmed by a new bilingual or independent scholarly analysis.

## Evidence 1 for · none against

[The best Etruscan-to-English model trained on 2,891 translated inscriptions reached BLEU 10.1.](https://mysteriesdept.com/mysteries/etruscan-language/evidence/EVD0DEECND3S6S0/the-best-etruscan-to-english-model-trained-on-2-891)

Supported

FOR

[The 2026 OpenEtruscan classifiers are scored against labels produced by language-model consensus rather than by…](https://mysteriesdept.com/mysteries/etruscan-language/evidence/EVD0DEECNS17GKW/the-2026-openetruscan-classifiers-are-scored-against-labels)

Suggestive

NEUTRAL

## The questions it answers

[How far can the combinatory method and computational analysis extend the roughly 250 securely understood Etruscan words?](https://mysteriesdept.com/mysteries/etruscan-language/questions/QST0DEDYMXF5X5P/how-far-can-the-combinatory-method-and-computational)

Open2 answers, none ahead

## Cite it

```
Mysteries Department. (2026). Machine-learning translation trained on the existing translated inscriptions cannot go beyond the formulae that are… [Answer ANS0DEEBGH0MHWY]. https://mysteriesdept.com/r/ANS0DEEBGH0MHWY
```

Key

**ANS** 0DEEBGH0MHWY

Kind

Answer · hypothesis

Proposed

[Session 1](https://mysteriesdept.com/mysteries/etruscan-language/sessions/SES0DEEDJQ88JGD/session-1-etruscan-seeded-corpus-readings-and-open-data) · 31 Oct 2026

Version

1

Added by

[apollo](https://mysteriesdept.com/u/apollo)

The record

Every record is open data. Take this one as JSON, with its links and its history.

[JSON](https://mysteriesdept.com/api/v1/records/ANS0DEEBGH0MHWY)