Grammar with a "G"

[Jess Tauber]

Not everything linguistic is within Grammar- ideophones, at least as
classically formulated, do not lie within clausal structure unless imported
through auxiliation or other derivational processes. This is interesting since
this type of ideophone is phonosemantically transparent, each shift of
phonological feature shifting meaning in a predictable and geometrically
symmetical way. In a way, one might even say there was only one ideophone,
with many variants.
That the most phonosemantically transparent items of vocabulary (I'll let you
decide on your own whether ideophones are "in" the Lexicon) are the least
integrated syntactically is probably no accident, as is the fact that the most
worn down, historically changed material is the most syntactically integrated.
Given that most vocabulary items likely have a phonosemantic origin
ultimately, can we posit some sort of mechanism or mechanisms whereby we can
get from the first state to the last? Is the ideophone like some sort of
neutral molecule, with all the "valence" requirements satisfied by it's own
internal parts? Once integrated into normal vocabulary, are some of these
"valences" now directed outward, allowing syntactic combination? In doing so,
is the stage set for historical change, in that there is decay in the
structural coherence and integrity of the original ideophonic molecule?
After all, what is the ultimate motivation for historical shifts of
articulatory position and manner? Why do these changes seem to resemble matrix
operations? (Note that real atomic orbitals used in  molecular binding are
analyzed in this fashion).
Ideophones also are subject to typological phenomena- dependent-marking verb-
final languages almost always have a separate word class containing these,
verb-medial languages have some ambiguity between these and some other word
class(es), and head-marking verb-initial languages mostly have "ideophones"
(I'm not sure the name is applicable here) fully integrated into one word
class (and verb-initial languages often have "precategorical" roots- word
class is assigned with morphology, not before. Interestingly, all the usual
ideophone-associated processes- robust reduplication, rhetorical lengthening,
infixation, etc., are often present in this type of rootstock as derivational
possibilities.)  It is as if there is some sort of inversion of structure. It
makes one wonder if such things might happen in the quantum world (for
instance, could one have hydrophilic micelles floating in an oily
environment?)
It would be interesting if the various levels of reality (subatomic particles,
atoms, DNA, language) which display tabularizable primitives and rules of
combination were conceptually congruent. Is this a result of inherent
structure, or are we simply viewing one possible interpretation of raw data
through the constraining filters of our language-adapted minds? The properties
of primitives and their rules of combination are not divorced from those of
the environments they are part of- at any one level, each domain "licenses"
the others, and all are born together. Thus there is the possibility of an
equation describing all the available interactions, whether quantized or
continuous. If language is like this, then only a holistic or integrative
investigative/theoretic strategy which takes into account not only evolving
formalistic structures but also the contexts of usage- all of them- will have
any hope of generating enough of the points in the painting to give the big
picture. Confirmational
support will come from neurolinguistic work, but the kind of resolution one
would need will be a long time developing.

Jess Tauber
zylogy at aol.com