[Lexicog] semantic domains

[Wayne Leman]

--- In lexicographylist at yahoogroups.com, "yahganlang"
<phonosemantics at e...> wrote:
Basque historically has been losing its small handful of light verbs,
which are the only verbs in the language which may inflect.

Many Australian languages exhibit similar behavior, though the number
of such inflectable light verbs may be much larger and more
semantically specific. Fine semantic specification is achieved by
using a large number of uninflected "particles", which form the bulk
of the lexicon in these languages.

In this kind of language one often sees an excess of phonosemantic
transparency, at least in expressives, but also often within the
other parts of speech as well (such as nouns, adjectives (when
distinct from nouns) and adverbs, which can be compounded with the
light verbs for semantic specification.

So how many true verbs are there in such languages? They tend to be
heavy on the "nouny" end of the continuum, often heavily case marked,
often SOV in constituent order.

Interestingly there seems to be a cycle here typologically- as the
total number of true inflectable verbs decreases, one sees the
concomitant rise of the "particle" class, with corresponding
expansion of expressives as a class, giving semantic specification.
The death of the light verb system, which corresponds to the rise of
a new and much larger manner specific inflectable verb system after
fusion and reanalysis, renders expressives for the most part
expendable. I would be curious to know whether the other parts of
speech used as "particles" also can undergo similar reduction.
Perhaps noun classificational systems parallel in some way the
dynamics of light verb systems, but might be out of phase with regard
to the interactions between expressives and grammaticalized items.

I have elsewhere hypothesized that expressives in fact are part of a
pragmatic system opposed to the grammaticalized one function for
function, an "antigrammar" if you will.

Jess Tauber
--- End forwarded message ---