Dick:
> And says:
>
> >I have a very clear idea of the difference in principle between
> derivation
> >and inflection. Derivation is where one stem contains another. Inflection
> >is where the phonological shape ('whole') of a word is
> determined not only
> >by its lexical identity but also by other, syntactic, properties.
> >
> >But I really struggle in applying this in practise.
>
> ## I think this is another view that I sometimes hold, or at least similar
> to one I sometimes hold, and maybe this is the one I hold at the moment.
I think we basically agree. Word-and-Paradigm inflectional morphology is
one respect in which I remain true to my WG roots!
> A derived form is just like an ordinary underived form in its grammar,
> whereas an inflected one is grammatically special. E.g.:
>
> derivation: "un-tidy" is just an adjective, grammatically
> indistinguishable
> from "tidy" or "big"
Yes.
> inflection: "tidi-er" is a comparative adjective, grammatically different
> from "tidy" and "big" because it allows "than" as complement.
Two problems here. First the particular example: I (and possibly you) have
held that "tidier" is a pronunciation of "more tidy". So that wouldn't
be inflection.
Even then, the "grammatically different" criterion is weak. For example,
one could say that that perfects are merely ordinary verbs that take
a special shape when complement of nonpossessive auxiliary _have_. The
other properties of the verb remain unchanged. 3sg marking on present
tense verbs is similar: it is applied iff the subject has certain
properties.
Note that on these analyses, there is no Perfect or 3sg "inflection" in
the Chet-Dick sense of "morphosyntactic category". There is morphological
inflection trigged by properties "is complement of aux HAVE" and
"has subject that is singular (etc etc)".
> As for Passive: this must be an inflection because it's grammatically
> unique - e.g. "taken" is grammatically different from "takes" and from
> "taking".
>
> This is what I used to tell my first-year students, and I think it was
> right. Are there any problems or uncertainties in this definition?
I don't see why derived forms shouldn't have grammatically unique
properties *in principle*. However, if it is correct (as I have claimed
hitherto) that in English all productively derived forms are ordinary
adjectives, nouns or verbs; so because of this it does happen that derived
forms can't be grammatically unique.
As for passives, this still leaves the problem of adjectives being derived
from them. That shouldn't be possible, unless passives are derived.
--And.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|