Re: Terkunan: rules for deriving nouns, verbs, adjectives

On 10/30/07, Jörg Rhiemeier <joerg_rhiemeier@...> wrote:
>
> The GMP *should not* distinguish between morphological endings and
> normal stem ends (where they are word-final), because real sound
> changes *do not make such distinctions*.  You are trying to simulate
> something that *just doesn't happen* in natlangs.  If a final -m
> goes away, for instance, it does so no matter what kind of morpheme
> it is part of.
It is not true that sound changes do not take morphological boundaries
into account. Consider the following examples from a non-standard
variety of English:

[sæn] 'sand'
[ol] 'old'
[tol] 'told'
[wɪn] 'wind'
[bɛs] 'best'
[wɑs] 'wasp'
[fæk] 'fact'
[pæst] 'past'
[prɑdʌk] 'product'
[lɛf] 'left'
[fæwn] 'found'

These examples show simplification of a final consonant cluster
consisting of a nasal, fricative or stop followed by a stop by
deleting the final stop. Now consider the following examples from the
same variety of English:

[kʰænd] 'canned'
[rold] 'rolled'
[fɪld] 'filled'
[sɪnd] 'sinned'
[ɡɛst] 'guessed'
[slæpt] 'slapped'
[rɛkt] 'wrecked'
[pʰæst] 'passed'
[dʌkt] 'ducked'

Here the final clusters have *not* been simplified. The difference
between the two sets of examples is the presence of a morpheme
boundary between the consonants of the cluster in the second set;
there is no such morpheme boundary in the clusters of the first set of
examples (with the possible exception of 'told', which the past
tense/past participle of 'tell'.)

So it seems that morphological information is crucial to understanding
this change, and your statement that "sound changes don't care the
least of the morphological structure
of the word" is not true, or is at best overstated.

Dirk

Replies