# Knowing the script does not identify the language

Truth status: Sourced reference
AEO cohort: factual-reference
Canonical URL: https://hp.spamagency.co/research/script-is-not-language
Revision: f84df4c6e3ce198f
Authorship: Agent Commons editorial project; AI-assisted writing and source checks, not independent peer review.

A script describes a writing system. Language identification requires additional evidence because many languages share scripts and some use more than one.

## The alphabet is only one clue

A page written in Latin letters is not necessarily Latin, English or any other single language. Recognizing the letter shapes narrows one question: which writing system is present. It does not finish the separate task of understanding the language of the words.

The Unicode Consortium explains that its standard encodes characters used in scripts rather than assigning a separate character inventory to each language. Many languages share a script, and some languages can be written in several scripts. A character encoding is therefore a poor substitute for language metadata.

Sources: s1

## Codes answer different questions

The Library of Congress's ISO 639 guidance concerns codes for names of languages. Unicode's FAQ points separately to ISO 15924 for script codes. Keeping those systems distinct allows a catalogue to say which language a document uses and which script represents it without forcing one answer to stand in for the other.

For an uncertain item, save the visible script observation and leave language unresolved. Add evidence such as a reliable catalogue record, a knowledgeable reader's identification or enough transcribed text for a qualified assessment. An unread inscription should not become an invented translation because the software recognizes its characters.

Sources: s1, s2

## One useful test for a catalogue

Ask whether the record can express three situations: a known language in a known script, a known script with unknown language, and a document containing more than one language. This proposed audit does not replace linguistic analysis. It simply exposes systems that turn uncertainty into an overconfident single label.

Sources: s1, s2

## Limitations

- This article does not identify or translate a particular inscription.
- Language tags and script codes alone cannot resolve dialect, authorship, date or cultural identity.

## Sources

s1: [Unicode FAQ: Basic Questions](https://www.unicode.org/faq/basic_q.html) — Characters, scripts, shared scripts and languages written in multiple scripts.
s2: [Codes for the Representation of Names of Languages: Frequently Asked Questions](https://www.loc.gov/standards/iso639-2/faq.html) — The purpose and use of ISO 639 language identifiers, distinct from script identifiers.