The Text Encoding Initiative (TEI) published its first detailed recommendations for the encoding and transcription of all manner of written and spoken materials, using an extensible SGML framework, in 1994. This was widely used by many academic research projects and digitization initiatives and was also very influential in many corpus creation projects (e.g. BNC, Parole, Multex, Silfide) as well as in defining EAGLES and other standards. A membership consortium was formed in 2000 to continue maintenance and development of the TEI, and the most recent recommendations (P5) are expressed using a modular XML framework, which supports most kinds of linguistic annotation.