LINGUIST List 3.549

Fri 03 Jul 1992

Disc: Morphology and Phonology

Editor for this issue: <>

Directory

1. , double morphology
2. John S. Coleman, 3.546 Innateness, Dissimilation Summary

My last posting on double morphology led to seven more reactions (thanks to
Anthony Aristar, E. Broselow, Greg Stump, Sally Thomason, Karen Wallace, and
two anonymous colleagues).
Aristar points out that double morphology is "fairly common in situations where
a form has become morphologically opaque". The following cases were mentioned:

(1) Afrikaans kind 'child' has the plural kind-er-s, with the old plural
 suffix -er and the new suffix -s.
(2) The well-known case of English child-re-n.
(3) In Aramaic, the mpl suffix combined with the 3msg possessive suffix
 to produce -o. -hi marked 3msg after vowel-stems. In later Aramaic,
 the mpl/3msgPos suffix is -ohi.
(4) In Amharic, words with Geez plurals can take the regular Amharic plural
 suffix -oc as well, e.g. gize-yat-oc 'times'
(5) In Bolivian Quechua, words with the Spanish plural suffix -s can
 take the regular Quechua plural suffix -kuna as well: wasi-s-niy-kuna
 'my houses'

It was also pointed out that English words like fixer-upper, putter-oner
are usually used tongue-in-cheek, so they are still felt as violations of
the norm.

While Anthony Aristar concludes that the elsewhere condition is "at best
misguided," Greg Stump mentions his recent papers (NNLT 1989, Yearbook of M.
1991, Syntax & Semantics 23, Language 1991), where he presents an account
of the Elsewhere Condition where double marking of this kind is not ruled out.

For a more critical evaluation of the Elsewhere Condition, see also
Richard Janda & Maria Sandoval. 1984. Elsewhere in morphology. IULC.

Martin Haspelmath, Free University of Berlin

Richard Goerwitz asks me:

> Where is it you
> feel natural phonologists are claiming uniquely elegant solutions that are in
> fact neither unique nor elegant? ...
> Perhaps it's just me, but this
> seems a rather presumptuous way of handling the matter.

It seems Richard missed the earlier posting to which I took objection,
which explicitly claimed that Natural Phonology was the ONLY theory
to address phonetic data. Now, whatever merits Natural Phonology's analyses
may have --- and I acknowledge it has some attractions! --- it cannot in
all seriousness claim to be the ONLY theory to
address phonetic data. It seems to me that as far as most versions
of contemporary phonology are concerned, there is a canon of phenomena,
some phonetic-ish, some more arbitrary, which all theories pick over. I
invited Natural Phonologists
to provide an example of some phonetic phenomenon which they alone have
analyzed, a simple request to back up the initial claim with a single
example. Why does Richard take such exception to me asking this? If someone
makes a controversial claim, surely the onus is on them to provide an
example or two in support of that claim. Presumptuous? Has it now become
presumptuous to invite one's distinguished colleagues to present
evidence and arguments in support of their claims?

--- John Coleman