when approaching the human sciences
is to discover combinatorial primitives
that generate (any) rich semantics
before they explode
and the tragicomedy
of pomo frenchy 'theory'
is that combinatorial explosions
are embraced as a virtue
so semantic primitives are
actively disparaged
so as we explore
discourse as lists
we need to tiptoe over
eggshell minefields
trusting primitives as steppingstones
and one such
primitive discourse-list
is the counterintuitively heterogeneous
A, B, +, A+B
because most story sequences
involve changing relationships
among compositions of simpler elements
and recounting relationships
stereotypically requires accounting first
for those simpler elements (A, B)
and then too
perhaps
for the the relationship '+' itself
(more often omitted as obvious)
and only then finally
the combination A+B
less stereotypically
these list-steps
can be freely permuted
(cf transformational grammar
and mathematical commutation)
as when Gravity's Rainbow
opens with Pirate (B)
followed by Pirate spying on Slothrop (B+A)
followed by Slothrop (B)
and perhaps spying-in-general (+)
but if our goal
semantic-web-like
is to tease out each text's
semantic understructure
we have to acknowledge every text's
infinite annotatability
A has an infinitely detailed
backstory timeline
debated finitely
within a subtimeline
of the Great Conversation
along with every topic
that might shed light on A
with our current author X
(with her own infinitely detailed
timeline backstory)
writing in style S
(with its own timeline backstory)
using words W (with their own etymological histories)
X possessing a finite knowledge
of the Great Conversation about A
identifying gaps in that knowledge
being a primary priority
of Conversational followups
and (A B + A+B)
is promiscuously associative as well
(A B + A+B C * C*A+B)
so imposing strict hierarchies
of stereotyped primitive lists
is probably beside the point
.
