About Open Editions

The Open Editions project publishes richly annotated open source literary texts. We aim to aggregate the literary knowledge that surrounds a text, codify that knowledge, and integrate it into the text.

Read the draft article for James Joyce Quarterly where we discuss the project and its preliminary goals.

Guiding Principles

1. Openness.

We believe in openness. That means open-source textual editing, open-access publishing, open project management, and all the accompanying principles of the Free and Open Source Software movement (FOSS). These are not just technical details, but a way of life.

Anyone can become a contributor. Read the Contributing page, and get in touch with us in our chatroom.

2. Meaningful markup.

Traditional textual encodings, like HTML, describe the way a text looks, rather than what it is. Consider italicized words, for example. In early HTML, they would be written <i>introibo</i> where <i> stands for “italics.” Later HTML made that more semantic, using <emph> meaning “emphasized.” But a truly semantic markup tells us why a word is italicized. So we use <foreign xml:lang="la">introibo</foreign>, which explains that the word is italicized because it is a foreign expression, and also that this is a Latin word (la is the two-letter language code for Latin). This kind of semantic markup allows us to be explicit about what we’re reading, how we’re understanding it, and how computers should “understand” that meaning. When computers understand a text in this way, it allows humans to manipulate that data, and be able to learn more about the text programmatically. This is not meant as a replacement for traditional literary criticism, but an advancement of it.

3. Depth before breadth.

Other digital publishing projects are interested in gathering as many texts as possible. We are more interested in going as deep as possible into each text, as a way of exploring the potential for this semantic approach.

Thus, we are less interested in discussing ways to import thousands of new documents into our framework. Rather, we’re interested in the kinds of things scholars discuss in miscellanies. This is why so many of our texts are dense, like James Joyce’s Ulysses.

4. Standards and correctness.

It is more important to create a system that is robust and interoperable than to make something pretty. We would rather push clean, reusable data to GitHub than to publish a hastily-assembled website, just to have something to show off. Thus, our web views and interfaces lag behind by design.



We use the P5 textual markup specification from the Text Encoding Initiative, which has been thinking about ways to encode text since the 80s. There is a vast toolchain that surrounds TEI markup, and there are many standards.

While it is true that there are some limitations to XML, we try to mitigate those where possible. For example, a common critique of TEI is that there are too many ways to encode the same textual features. A number of projects have tried to correct this. TEI Lite and TEI Simple are subsets of TEI intended to disambiguate markup choices. We don’t use either of these, but prefer to maintain a list of conventions, documenting how we use certain types of markup.

Git and GitHub

Distributed version control is crucial to our work, since it allows us to collaborate widely, using well-tested frameworks. Collaboration is public, thereby making our intellectual labor fair and transparent.

GitHub provides a place for us to store our Git repositories, and to communicate with each other via issues. We plan and discuss our editions there.