Skip to content
ISO 639-2 code · ISO 639-2 language codes

tup, the ISO 639-2 code for the Tupi languages, and how to use it

Published Formatory editorial team
http://id.loc.gov/vocabulary/iso639-2/tup

Tupi languages

tup is a collective code with no ISO 639-1 equivalent; it groups the Tupian languages and is carried into ISO 639-5, while ISO 639-3 codes the individual languages.

The Tupian family is one of the largest language families of South America, spread from the Atlantic coast of Brazil across the Amazon basin into Paraguay, Bolivia and beyond. ISO 639-2 does not code most of its members one by one. It offers a single collective code, tup, with the label Tupi languages (French tupi, langues), published by the Library of Congress at http://id.loc.gov/vocabulary/iso639-2/tup.

What a collective code is for

ISO 639-2 has around 480 entries, and roughly a seventh of them are collective codes for families or groups. They exist for a practical reason: when the standard was built from library practice, catalogers needed a valid value for material in languages too small or too little documented to get an individual code. A collective code says honestly “a language of this family” without inventing precision.

The rule follows directly. tup is a fallback, used only when no individual code applies or when the specific language cannot be determined from the item.

Individual codes come first

Among Tupian languages, ISO 639-2 has an individual entry for Guarani (grn, ISO 639-1 gn), the language shared by most of the population of Paraguay. Guarani material therefore never takes tup.

For the many other Tupian languages, from Mundurukú to Kamaiurá, ISO 639-3 provides individual codes, and the Guarani entry itself becomes a macrolanguage there, with Paraguayan Guarani and other varieties coded separately. A repository whose schema accepts ISO 639-3 or BCP 47 tags should use those codes. Where the schema is fixed to the ISO 639-2 list, as in MARC 21, tup plus a 546 note naming the language is the correct pattern.

Old Tupi and the colonial record

The best-known use of tup in library catalogs is for Old Tupi, the coastal language that Portuguese and French observers met in the sixteenth century. Jesuit missionaries wrote grammars and catechisms in it (José de Anchieta’s grammar was printed in 1595), and a later form, Língua Geral, served as a common language in parts of colonial Brazil. Nheengatu, a modern descendant, is still spoken in the Amazon region.

Those historical books are valuable and widely digitized, and their records typically combine tup with Portuguese (por) or Latin (lat), as in the example above. Digitization projects that add full text should note that the spelling of colonial Tupi follows Portuguese or Spanish conventions of its time, which varies from book to book.

Typical errors

  • Coding Paraguayan or Bolivian Guarani as tup because “Guarani is a Tupi language”. It is, and it has its own code.
  • Using tup for Tupi-language place names in Portuguese texts. Brazilian toponyms such as Itaquaquecetuba or Paraíba come from Tupi, but a Portuguese text containing them is Portuguese.
  • Assuming tup is a valid code in ISO 639-1. There is no two-letter form for any collective group.

The same logic of collective fallback applies to the code for artificial languages and to North American families such as the Caddoan languages, of which only Caddo has an individual ISO 639-2 code.

Example in RDF

@prefix dcterms: <http://purl.org/dc/terms/> .

<https://example.org/rare-books/catechism-1618>
    dcterms:title "Catechism in the language of the Brazilian coast" ;
    dcterms:language <http://id.loc.gov/vocabulary/iso639-2/tup> ,
                     <http://id.loc.gov/vocabulary/iso639-2/por> ;
    dcterms:description "Text in Old Tupi with Portuguese." .

Questions about tup

What does the language code tup mean?

tup is the ISO 639-2 collective code for "Tupi languages", the large Tupian family of South America. It identifies a group of languages, not a single one. The Library of Congress publishes it at http://id.loc.gov/vocabulary/iso639-2/tup.

Should Guarani be coded as tup?

No. Guarani has its own ISO 639-2 code, grn, with the ISO 639-1 form gn. Paraguayan Guarani material should take grn. The collective code tup is for Tupian languages that have no individual ISO 639-2 code.

Which code fits Old Tupi texts from colonial Brazil?

ISO 639-2 has no individual code for Old Tupi, so library records use tup. ISO 639-3 has an individual code for the historical Tupi language, which is the more precise value in schemas and BCP 47 tags that accept it.

Is tup the same as a Tupi-Guarani code?

Tupi-Guarani is one branch inside the wider Tupian family. ISO 639-2 has only the one collective code, tup, for the whole family, so it covers Tupi-Guarani languages and the other branches alike.

Can tup appear in a BCP 47 tag?

Yes. tup is registered in the IANA subtag registry as a collection subtag, so it is well formed. For web content in a specific Tupian language, the individual ISO 639-3 subtag is better because software can do something with it.

More from ISO 639-2 language codes

All 17 in this group