Geneword is an ongoing series of digital poetry experiments focused on the invention of new words, letters, and ultimately languages.

Index I, the first iteration of this project, generates “new words” by affixing Greek & Latin prefixes to nouns in the English language. This combinatory word-processing occurs through the poetic intervention of an algorithm & a bot we created (called Geneword) that generates a vast permutational series of selected prefixes and nouns, and can also place these “new words” randomly within a wide range of texts of different genres and forms. The tiered parameters of Index 1 are defined by two sets of nouns: the first is “all nouns,” which consists of 37,199 different nouns, and the second is a smaller set that we are calling “common nouns,” which consists of 1514 nouns. By using 168 prefixes, the algorithm produces over 9 million “new words,” not counting redundancies.

Index I is designed to be a playful provocation and serious starting point for imagining what new words for things can do. With a simple gesture, this algorithm aims to permanently expand the English language, and to find out what happens (expressively & poetically) when such new spaces are created. This feels particularly important in a time when languages of power are being used and abused to limit our vision of what is possible, rather than opening up new pathways of meaning-making and ways of being in the world. This is the power of language — a’bra c’dabra — i create as i speak.

All nouns and definitions are sourced from the WordNet database.


Tom Haviv is a writer, artist and educator based in New York City. His website is

Owen Roberts is an artist and educator based in Brooklyn, New York. His work is available at

Geneword License

Princeton University “About WordNet.” WordNet. Princeton University. 2010.

Bird, Steven, Edward Loper and Ewan Klein (2009). Natural Language Processing with Python. O’Reilly Media Inc.