ISO 639-5 does not code individual languages at all, which sets it apart from the rest of the family. ISO 639-1, ISO 639-2, and ISO 639-3 each assign a code to one specific language, whether that's a compact two-letter code for a major language, a three-letter code used mainly in library cataloging, or a three-letter code covering nearly every documented language on record. ISO 639-5 instead identifies a family or group of languages. For example, the code gem means "Germanic languages" as a whole category, not any single Germanic language like German, Dutch, or Swedish. If you need to say "this text is in German," you'd use one of those other three parts. If you only know, or only need to say, "this text is in some Germanic language," ISO 639-5 is the part of the standard built for that.

The Library of Congress is the designated Language Coding Agency for ISO 639-5, operating under the shared ISO 639 Maintenance Agency that oversees consistency across all parts of the standard. This is different from ISO 639-3, which SIL International maintains, and reflects the Library of Congress's long-standing role in library cataloging standards, including its earlier stewardship of the ISO 639-2 collective-code entries that many ISO 639-5 codes descend from.

This table contains 115 codes, matching the current published edition of the standard. That's a deliberately small, fixed list compared to ISO 639-3's several thousand entries, because language families are a much smaller, higher-level category than individual languages. There are only so many recognized major language families and sub-groupings in linguistic classification, and ISO 639-5 aims to cover that finite set rather than track every individual language variety.

Yes, substantially. ISO 639-2 has always included a set of "collective" or group codes alongside its individual-language codes, codes for categories like "Slavic (Other)" or "Bantu languages," because catalogers needed some way to handle family-level language description before a dedicated standard existed. Of the 115 codes in ISO 639-5, 65 carry over directly from those ISO 639-2 collective codes (36 "remainder groups" plus 29 "regular groups"); the remaining 50 were added new when ISO 639-5 was published. ISO 639-2 kept its original collective codes for backward compatibility rather than adopting ISO 639-5 wholesale.

The main scenario is cataloging and archival description, not everyday content tagging. A library, museum, or archive sometimes holds a work where a cataloger can confidently identify the language family, say, it's clearly a Bantu language based on its structure and origin, without being able to determine, or without it mattering for the record, which specific language within that family it is. Rather than leaving the field blank or marking it "language unknown," a 639-5 code lets the record say exactly as much as is actually known. Library metadata standards like MARC support ISO 639-5 codes for precisely this reason.

English and French are ISO's two official standard-publication languages, and every part of the ISO 639 family publishes its official names in both, not just this one. That's the same reason the companion ISO 639-1 and ISO 639-2 pages on this site also carry French names. Having both names doesn't limit what the standard covers to English- and French-speaking contexts; it simply reflects which two languages ISO itself publishes its documentation in. The families and groups listed here still span every documented family of human language worldwide, regardless of which countries or language communities use them.

No. ISO 639-5 identifies a language family or group, not a country, a region, or a locale. ISO 3166 identifies countries and regions, such as US for the United States or FR for France. BCP 47 is a separate IETF specification for constructing full locale tags, typically combining an ISO 639-1 or ISO 639-3 language subtag with an optional script subtag and an optional ISO 3166 region subtag into strings like en-US. ISO 639-5's family-level codes aren't designed to slot into that same locale-tag structure the way individual-language codes are, since a locale generally needs to name a specific language, not just the family it belongs to.

No. Boostlingo's list of supported languages is organized by specific language name, not by language family.