There are many stories to tell about our motivations for creating Eli Review. One goes back to my dissertation research at Purdue, where I studied the way experienced teachers adapted to new technologies in the classroom. The headline from that study: technologies that support writing have pedagogies built into them. Often, those pedagogies clash with those the teacher has. The takeaway from this is straighforward, but hardly simple. Rhetorics can live not only in books and articles, but in machines and software. When they do, they can transform practice in powerful ways.
What would it mean for us to build rhetorics and pedagogies into our own writing environments? That has been one of the driving questions of my career. And so it is one of the stories behind Eli. Eli is not the first system I had a hand in creating to embody a theory of rhetoric. But it is, by just about any standard, the most successful attempt. Eli is used every day by hundreds or even thousands of teachers and students. More than all the people who've ever read an article or book chapter I wrote, to be sure. Likely more than all the people who've ever heard me speak about research, theory, or pedagogy.
Eli embodies many theories about the way writing works. Far more than I can usefully wedge into an article and still have the article remain coherent. In Eli's deep structure - the data formats that represent people, texts, and actions in the system - lie counterintuitive ideas about textuality that mash the ideas of Derrida, Barthes, and Iser together. Eli doesn't see texts that are part of reviews as coherent wholes. Rather, a text is a surface upon which the social actions of a review unfold. Texts provide an ability to index actions, relate them to one another, and provide a view of what would ordinarily be invisible social interactions back to writers, reviewers, and review coordinators. And while texts in Eli are palimpsests, they are also social actors of a sort. That is, we understand that over the course of a review, texts assume the agency of various human users for whom they stand in. And when I say we "understand" this, I mean we've built representations of review activity that explicitly support what, in Actor-Network Theory terms, are moments of inscription, enrollment, and interresement in the course of a review. We even have an algorithm for evaluating review success based on this progression, a novel measure of review *activity* rather than review comments or texts.
Eli also carries our ideas about how people learn to write. One radical one is that people can learn as much or more from reviewing work by others as they can by writing. Another, less radical thought for most writing teachers that is still difficult to act upon in a traditional classroom is that where writers *do* learn from writing, they do so most effectively from revision. Eli provides an environment where review can play a more prominent role in learning to write simply by making the coordination of review activity far easier and faster to do. Eli also provides an environment where writers can get more high-quality feedback more often, allowing more of their writing time to be spent on revising.
If Eli were a book, I could never get an editor to allow me to put all the ideas that make it work into one volume. It would appear to stray, to lose focus, or to vacillate wildly from the minutae of poststructuralist theories of language to the currently-in-fashion ontological perspectives of ANT to the pragmatic ideas of writing researchers like Britton and Hillocks. In Eli, they live as comfortably as an engine, drive train, and electronic ignition controls in a contemporary car, systems that stem from ideas of different eras, with different theories, all working seamlessly together.
