Circumtext began in 2011 as a reading group and series hosted at FICTILIS’s Seattle studio, where participants gathered to read aloud texts that they did not write. At Circumtext events, “did not write” often meant texts were found, sought, or collected, transcribed, translated, and transformed, or generated and assembled, often by some process of search, aggregation, scraping, filtering, or constraint. This was before chatbots were household names and generative AI made text generation a common practice. Generative methods were for us part of longer traditions of experimental writing, creative uses of technology, and emancipatory approaches to language.

The name “circumtext” comes from the Latin prefix ‘circum’, meaning “around” or “about”, and the root ‘text’. A similar construction is used in the English words context, subtext, paratext, hypertext, et cetera. (The name was also a nod to an earlier Seattle experimental writing group, the Subtext Collective.) We wanted to wrap all of the other -text prefixes into one, and to suggest that texts worthy of consideration arise from many different sources—high and low and everything in between, and that these texts can be approached from many different angles—above, below, and all around. Hence circum-text.

As Circumtext evolves into a publishing practice, and its scope expands to include texts that may be ‘authored’ in a more traditional way, its spirit remains true to our beginnings as an experimental reading group. The name still evokes for us a way of thinking about our own writing practices and reading preferences, which might be called “circumtextual”. And there is always a sort of non-writing built into the activity of writing itself—using language inherited from previous generations, trying to shape something new from pieces that are old.