LINGUIST List 2.836

Wed 04 Dec 1991

Disc: Inflection vs derivation

Editor for this issue: <>

Directory

1. Greg Stump, Inflection vs derivation

In recent work, I have argued that there are at least two distinct
subclasses of cases in which--contrary to Greenberg's 28th
universal--inflection may appear inside of derivation; the two
subclasses correlate with a distinction between category-changing
derivation and category-preserving derivation. In those cases in which
category-changing derivation applies to an inflected form, the
inflection does not contribute to the morphosyntactic feature content of
the derived form (and is therefore irrelevant to its syntax); e.g.
Breton _ler-ou_ (n.) `sock-s', _dilerou_ (adj.) `without socks on' (lit.
`socksless'). For additional examples of this and other such processes
in Breton, see my article `Breton Inflection and the Split Morphology
Hypothesis',
_Syntax_and_Semantics_23:_The_Syntax_of_the_Modern_Celtic_Languages_,
ed. by R. Hendrick, Academic Press, 1990.
In those cases in which category-preserving derivation applies to an
inflected form, the inflection does contribute to the morphosyntactic
feature content of the derived form; that is, inflections that are
morphologically realized `inside of' an outermost layer of
category-preserving derivation (or compounding) function logically as if
they were `outside of' this layer. Instances of this phenomenon are
legion, and include the Georgian case cited by Harry Bochner (1984); the
Oneida case mentioned by Clifford Abbott (1981); the first of the two
Slave cases discussed by Keren Rice (1985) [the second of which seems to
belong with _dilerou_, etc.]; Sanskrit cases like _pari-n.i:-_ `marry'
(lit. `lead around'), imperfect _pary-a-n.ayat_ `he married'; German
cases like _Kindchen_, plural _Kinderchen_; English [ un- [ happi -er
]]; and, I believe, the case of Russian verbs in _s'/s'a_ raised by John
Nerbonne. A detailed formal account of this phenomenon is developed in
my article `A Paradigm-based Theory of Morphosemantic Mismatches', to
appear in the December issue of _Language_.
All of this presumes a well-defined distinction between
category-changing and category-preserving derivation; in the _Language_
article, I present specific proposals about the nature of this
distinction (which ends up not being the simplistic distinction which
the terms `category-changing' and `category-preserving' could be taken
to imply). I'd like to know more about the Munda evidence cited earlier
by David Stampe.