Our Poetry Analysis & Translation App

It performs complex analysis of poems or songs and drafts a specialized feature-aware translation aiming for a balance of formal and semantic fidelities. Should work with most literary languages; albeit, only in one direction – non-English to Worldish. You can try it right here. No need to download or install anything. Simply paste a verse/stanza into the “Query / Source Text” form, press the “RUN LYRICAL SymbCoT” button, and wait from roughly 20/30-seconds (input of 4/5-lines) to a few minutes (for 5-6 verses/stanzas). Outputs will appear in underlying boxes: Structural (+ Rhythmic, Phonaesthetic, & other…) Analyses, Purview of Linguistic Arsenals of Necessity (aka “PLAN”), Procedural Draft (sometimes), and Final Draft. Lastly, for multiple stanzas/verses, one should format inputs like so:

Translate:(#Verse 1:)A noir, E blanc, I rouge, U vert, O bleu: voyelles,
Je dirai quelque jour vos naissances latentes:
A, noir corset velu des mouches éclatantes
Qui bombinent autour des puanteurs cruelles,//(#Verse 2:)Golfes d'ombre… etc…

Now, and this may be a moot caveat, but as with most neureal artifice, it’s doubtful this tensor-brained drafter would approach your elated standards, even when working from so much formalized insight into the source, and saddled with all the veiled guidance we’ve crafted. But perhaps the key generative space is precisely here, in what is no longer a “fissure”, but a kind of suspension (like a gravity well?): no longer between languages, but between a certain suggestion invoked here – that of a poem abstracted unto “a world” – and this concrete text, this chimeric “output”, this poem partly-abstracted, like some Frankensteinian salvage of marooned universals; a poem already departed or severed from full identification with any particular language, but yet to arrive all the way to… itself. For A.I. only carries it partway: ever more than halfway, seldomly near (unless by luck). And if i.t. can’t altogether help the poem get there, then who can?! Well… all sorts of persons, actually! (Had they but time, plus some spoons, and a porous musical head &, you know, a room, plus some UBI, and OCD really helps too… But seeing as you’re already here…) Maybe you? Now that you can run your favorite poems under this handy tap of formal analysis. And, as dubious as this may seem, the analysis is largely apt and accurate (in my experience). After all, many of the analytical operations dealing with the source text are not herein entrusted to generalist LLMs, but handed off to specialized NLP (Natural Language Processing) libraries (spacy, nltk, stanza). An LLM, for its part, puts this info together, collating its findings in piles, even for a thin little quatrain! So, prepare to copy/paste.