The concept-space of poetic forms
Jorn Barger
for clarifying previous threads on 'The concept-space of rhetoric'
(actually rhetorical forms) and 'The concept-space of Aristotle's
28 topoi'.
Imagine you're writing a poem in your word-processor, and when
the first draft is done you click a 'check scansion' button
and it highlights the places where your accents don't fall right.
It could theoretically do this by using a dictionary that
identifies accented syllables for every word/entry. And you
could theoretically then load in the complete works of (eg)
William Blake and ask it to search for (eg) dactyls.
I've chosen poetic forms as a startingpoint because the 'concept
space' of poetic forms is comparatively simple:
- how many lines
- how many syllables per line
- what meter
- what rhyme-scheme
In theory, the word-processor's search-function could do
searches on any of these conceptual dimensions, or check that
a given poem matches a given form, or suggest words that fit
the rhyme-scheme and meter, or randomly generate new 'forms'
and challenge you to write a poem that fits.
Classical rhetorical forms take these dimensions and add
dozens of others. Where 'rhyme' is about two lines ending
with the same sound, alliteration is about two words starting
with the same sound, and assonance is about two words with
the same sound in the middle. Acrostics are about multiple
lines that start with particular letters (or less often
multiple _words_ instead of lines).
Palindromes reverse a whole series of letters, while
allowing spacing and punctuation to be varied-- this
rhetorical word-processor might have a special function for
exploring palindromes, and by analogy 'word-palindromes'
that repeat a series of words (syllables, phrases, concepts)
in reverse order.
One of the reasons I distrust XML is that these rhetoric-
games are infinitely fluid, where XML wants to nail
everything into boxes-within-boxes, with no overlaps.
This works okay for the simple hierarchy letter-syllable-
word-phrase-sentence-paragraph-chapter-book, but rhetoric
exploits a much richer and promiscuously-overlapping set
of concepts including rhythmic-foot and poetic line, etc.
So if you're tagging _lines_ in a poem, you can't also tag
the sentences because they don't always end at linebreaks.
If you're tagging feet, you can't also tag words. Rhymes
may only work if you break one word at the end of a line:
...orange/ ...Jorn, J/orn...
The classical rhetorical forms wander off in dozens of
other directions, mostly a lot more difficult to imagine
automating:
http://humanities.byu.edu/rhetoric/Figures/Figures-Groupings.htm
This is what I mean by (the dimensions of) the
'concept-space' of rhetoric, and it should be clear that
many classical terms are poorly defined, and many useful
rhetorical forms within the same concept-space are as yet
unnamed.
Joyce: http://www.robotwisdom.com/jaj/ulysses/eolustropes.html
XML: http://www.robotwisdom.com/web/structure.html
AI: http://www.robotwisdom.com/ai/