The Text Encoding Initiative: evolution of a standard

For the last three decades and more, the Guidelines of the Text Encoding Initiative, the TEI, have served as a key reference point  for the builders, describers, organisers, and annotators of linguistic corpora and of other resources in the Digital Humanities. The TEI claims to be a practical response to the tension between the needs of beginners, seeking simple well understood solutions to common problems, and the requirements of others, seeking to push beyond the established frontiers of research. In this talk, I will try to summarize some of its ideas as they have developed over time, focussing primarily on their adaptability and extensibility. For, despite its name, the TEI is not concerned solely with text. Even in its earliest versions, there are well-elaborated proposals for the handling of bibliographic metadata, transcriptions of speech, and abstract linguistic analyses. Two guiding principles – the application of Ockham’s razor and a modifiable modular architecture – underly its evolution as one of the great efforts towards interdisciplinarity of our time. In fact, we may think of the TEI as a constructor set, a framework which facilitates the making of a system well adapted to the needs of its user, but also comprehensible to other people or systems. I suggest that the reason for the longevity of the TEI is to be found in its capacity for mutation and evolution and explore the implications of this design goal for the notion of conformance and standardisation.