Final Puff

About

Remorseless Havoc consists of an ocean of algorithmically generated language. The stanzas you can see zooming in on the map are generated by an algorithm from Sea and Spar Between, a work of digital literature developed by Nick Montfort and Stephanie Strickland in 2010. They devised a grid of algorithmically generated stanzas assembled from language fragments by Emily Dickinson and Herman Melville. If you open Sea and Spar Between in a browser, about 225 trillion entries extend almost endlessly in every direction, triggering the vertigo of combinatorics in an infinity with no practically reachable boundary, created from finite source material. Montfort and Strickland say their work contains as many stanzas as there are “fish in the sea.”

Remorseless Havoc takes the metaphor literally and transfers the stanzas into a grid spanning the entire globe. Each cell of this grid represents an area of about one square kilometre (1.1 × 1.1 km at the equatorial meridian). This results in 460 million cells and just as many stanzas, distributed across the world’s oceans. Gradually, the stanzas are then deleted by the program based on the movement data of industrial fishing vessels.

All technical details can also be found on this project’s GitHub page, but the movement data comes from Global Fishing Watch. There, the transponder signals from vessels above a certain tonnage are analysed using a machine-learning algorithm that recognises characteristic turning manoeuvres, changes in speed, and dwell-time patterns. In this way, GFW identifies fishing activity and calculates it in hours per vessel in conjunction with the GPS position. The Remorseless Havoc algorithm then erases verses from the ocean for each hour in near real time, as GFW requires 72 hours to analyse the satellite data. One day, all verses will be removed from the sea — the language used in this work is finite, like the ocean’s resources.

A note on the map. Its grid of one-square-kilometre stanza cells holds approximately 460 million entries. For rendering efficiency, the map displays a coarser tile grid of roughly 10 × 10 km (0.1°); each visible tile aggregates 100 stanza cells. The three shades on the ocean encode how much of each tile has been erased: intact (all 100 stanzas still alive), partially depleted (at least one erased), fully depleted (all 100 gone). When you zoom in, each tile reveals one representative stanza drawn from its centre cell; the full stanza cell grid is only ever consulted by the backend when a fishing event actually deletes stanzas.

You can watch this happen on the map, browse all poems generated in the Fleet-View or download the day’s poetry volume under Catch of the Day — an archive of the previous day’s erasures. Each vessel assembles the stanzas it takes from the ocean on a given day into a poem, ordered from greater lexical diversity toward monotony and simplicity.

The work runs until the poetic ocean is empty. Based on current data, roughly until the year 2030. Then one last cell will remain — the “final puff” of the Leviathan, speculated upon in Chapter CV of Herman Melville’s Moby-Dick, “Does the Whale’s Magnitude Diminish?—Will He Perish?”, from which the project’s title is also drawn as a direct quotation. Of the original 225 trillion possible stanzas, a single last poem will survive. When and where it emerges, and not least what its text will be, is determined by the contingencies of the exploitation of the seas.