I saw this post from @piesaac about how to use word rhythms and it reminded me that I wanted to continue my previous post around writing web tools for categorization and usage of words and their pronunciations.
Last I left off, I had the ability to look up words forwards and backwards by pronunciations for a portmantrois tool. Among other rhyming-dictionary projects that I had gotten into along the way, I realized that the phonemic pronunciations of words could be classified into vowel sounds and non-vowel sounds very easily, and that vowels were the basis by which CMU had labeled stressed and unstressed syllables of words.
I began looking up the names of poetic meter word rhythms to help classify words based on syllable number and stressed/unstressed orderings.
I had been familiar with
iambs (stressed unstressed)
com-PARE
and trochees (unstressed stressed),
AC-tor
but I learned of many more like amphibrachs (unstressed stressed unstressed),
a-COUS-tic
anapests (unstressed, unstressed, stressed),
clar-i-NET
primus paeons (stressed, unstressed, unstressed, unstressed), and lots of others.
DE-fin-ite-ly
By the end of it, I had categorized all of the words up to six syllables, opting for a shorthand of s and u letters for ones like "coeducational" that didn't fit any of the named meters on Wikipedia. This collection can be accessed below:
A large driving factor in the feedback loop of me choosing to work on a rhyming dictionary or a word rhythm dictionary was the idea of using these categories to help create random name generators with support for complex grammars.
With the ability to tap into words or names with certain rhythms to them, I was able to make tools to:
- Help me form names based off a specific requested rhythm:
- Help me determine words that had the same mouthfeel as Turpentine:
- Help me towards an ultimate goal of making a tool that generates original and absurdist limericks (or any specific template, like the verse of Eleanor Rigby) and have them follow the rhyme and rhythm scheme while also being close to grammatical.