The invisible code that tells the internet how to speak to you

Cultures · Published 1 October 2026 Swipe more like this →

In short: Every time you load a webpage in your native tongue, a precise, standardized string of characters is working behind the scenes. From HTML to PNG files, the IETF language tag ensures that digital content recognizes your specific language, script, and region with mathematical precision.

The IETF BCP 47 language tag is a standardized system used to identify human languages across the internet. It functions by combining various 'subtags'—small, hyphen-separated identifiers—to create a highly specific linguistic profile. For instance, a tag like 'sr-Cyrl' tells a computer that the content is in Serbian, specifically using the Cyrillic script. This system isn't just about language; it can even specify regional variants, such as 'es-419' for Latin American Spanish, or even include mathematical notation and emojis via specialized scripts like 'Zmth' or 'Zsye'.

The architecture of these tags is a complex mosaic of international standards. It pulls two-letter language codes from ISO 6/39-1, three-letter codes from ISO 639-2 and 639-3, and four-letter script codes from ISO 15924. It also incorporates geographical data from UN M.49 and country codes from ISO 3166-1. This interoperability allows a single tag to communicate a massive amount of data, such as 'nan-Hant-TW', which identifies the Minnan language using traditional Han characters as spoken in Taiwan.

Maintaining stability is a core priority for the Internet Engineering Task Force. To prevent the internet from breaking when international standards change, the specification includes 'grandfathered' tags. This means that even if an ISO code is withdrawn or its meaning is updated, the existing language tag retains its original meaning to ensure backward compatibility. While the system is highly structured, it is designed to be efficient: if a script or region is already the default for a language—such as the Latin script for English ('en')—the tag is simplified to avoid unnecessary data.

The system is constantly evolving. Originally defined in March 1995 by Harald Tveit Alvestrand in RFC 1766, the standard has undergone several major revisions. Significant updates in 2001 (RFC 3066) and 2006 (RFC 4646) introduced more complex structures, including the ability to use digits and more diverse script codes. The current iteration, RFC 5646, published in September 2009, focuses on maximizing interoperability between different ISO language registries.

General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.

Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?

Related

• Every known language gets a three-letter code, and animal calls get their own
• The 26 letters that standardized the digital world
• Unicode is the invisible language that allows your computer to speak every human tongue.
• Why are the world's two-letter internet country codes assigned the way they are?
• Why a burning ship might yell 'Yuliet alfa vore' to be saved
• How Did English Become India's De Facto National Language?
• The ancient myth of the silent, universal language
• The Indian indenture system: A post-abolition labor migration that reshaped the British Empire
• Why modern Icelanders can still read medieval sagas written eight centuries ago
• The staggering linguistic diversity of the Americas is far greater than Europe's
• The green screen terminal that defined how we talk to massive mainframe computers
• The prehistoric mother tongue that connects billions of people today

More in Languages · More in Cultures · Latest · A–Z · All topics