With this panel, the Technical Council publicly announces the beginning of serious work on a P6 version of the TEI Guidelines. We want to discuss how we arrived at this decision and how we plan to proceed, and we welcome input from the community at this crucial stage.

That there have only been five major versions of the TEI Guidelines to date seems a sign of robust permanence. P1, P2, and P3 appeared within roughly a decade while TEI was under development in the 1990s and early 2000s. The change from P3 to P4 marked the move from Standard Generalized Markup Language (SGML) to eXtensible Markup Language (XML). The move from P4 to P5 introduced the system for organizing elements and content models by subscription to classes, and more fully embraced XML namespaces, non-DTD schema languages, and URI identity referencing.

Launching a new proposal indicates a fundamental change, and we do not make this decision lightly. Following the "Birnbaum principle", “such transitions should only happen in response to the emergence of new technologies (e.g., the transition from SGML to XML) or the development of a new architecture that conveys substantial benefit (e.g., the development of the new class system)”.

In 2007, P5 v. 1.0.0 included 503 elements; today P5 v. 4.11.0 has 588 elements organized in 21 distinct modules and an intricate system of model and attribute classes. Council finds that it is often very difficult to introduce new features, fix longstanding issues, and reduce redundancy in TEI P5 without having to disrupt the class structure, and scholarly developers often find it difficult to adapt the TEI's highly complicated schema to real projects. It is too easy for a project to break conformance with the TEI while its developers do not fully understand good reasons why conformance is broken. We want customization to be simpler for our scholarly community, and we are re-evaluating and giving ourselves a fresh start with P6 to reconsider how we construct our content models and our schema. If we are successful, the P6 we design should be a form of TEI that is more tractable for all of us to learn, apply, customize, and develop.

Discussing the TEI long ago, C. Michael Sperberg-McQueen is recalled to have said, "schemas exist on a continuum from enabling to enforcing." Perhaps we can understand the TEI as a balance of enabling flexibility and constraints that we support as a community. Our Guidelines have been refined over many years with many voices from around the world. While the TEI's early forms reflected a strong emphasis on European print and manuscript works, we are no longer the same TEI we once were: our community is more diverse and our projects represent a wider range of texts, including "born digital" resources, writing on objects, and non-Western scripts. Our community has strengthened the TEI by introducing a wider range of features and methods for text encoding than originally anticipated by our first editors.

Because so much has changed, the Technical Council finds that it is time to re-evaluate our priorities, to revise the TEI in such a way as no longer to be backwards-compatible with P5, yet to continue the expression of the TEI in eXtensible Markup Language (XML). Council agrees that XML is the most robust way to express our community's encoding principles, but that there is much to improve and change about our methods of modeling information.

Council will begin serious work on an approach to P6 during a three-day face-to-face meeting in Würzburg, Germany in March 2026. At the time of this writing, that meeting is two weeks in the future, but Council plans to address topics about P6 that we want to share and discuss with the TEI community at the 2026 conference. We propose a panel involving members of the TEI Technical Council to share highlights and key points of controversy that emerge from our face-to-face session.

Together with the audience, the panelists plan to address a number of crucial questions, including but not limited to the following:

• Why is it time for P6? Why should we not just continue to add feature requests to P5?

• As we develop P6, we plan to continue supporting P5 in "maintenance mode". What will that mean?

• Will there be a period of accepting feature requests to support P5 while developing P6?

• Will the elements we know from P5 persist in P6? Are we going to try to reduce or expand the number of elements?

• Will models and classes persist in P6, but not in a way that's compatible with P5? Or will we create an entirely different system of organizing TEI elements and attributes?

• Where and how is P6 being developed? How will our supporting infrastructure change?

• Will there be parallel serializations alongside XML?