Re: THEORY: Verbs go irregular before our very eyes!

John Cowan wrote:


>Apparently Maori has quite recently undergone a transition that has made
>almost all its verbs technically irregular.  Consider these data sets:
>
>Set I:
>
>Active verb     Passive verb    Gloss
>awhi            awhitia         'embrace'
>hopu            hopukia         'catch'
>aru             arumia          'follow'
>tohu            tohungia        'point out'
>mau             mauria          'carry'
>wero            werohia         'stab'
>
>There is a old (pre-Polynesian) and strong constraint that all Maori
>syllables are (C)V, which disallows final consonants.  ("wh" and "ng" are
>digraphs, not clusters.)  The neatest explanation of Set I, though, is that
>the underlying forms are AWHIT, HOPUK, ARUM, TOHUNG, MAUR, WEROH, and then
>the rules applied are:
>
>1)      To make the passive voice, apply the suffix "-ia".
>2)      Remove any final consonant.

That's correct from the historical point of view (whether native speakers
would agree is another matter-- "you just have to LEARN which verb takes
which suffix" just as we learn sing-sang-sung, bring-brought etc.).  Proto
Oceanic, the grand-daddy of all Polynesian and Melanesian languages, is
reconstructed with final consonants, though some are in doubt, due to a lot
of analogical re-shuffling in individual langs.-- e.g. the putative cognate
of "awhit" in Tongan, or Fijian, will not necessarily form its passive
with -t-.  But by and large, these consonants in Oceanic languages
correspond with each other, if not always to the Proto Austronesian
original.

I think the _ia_ part is a Polynesian innovation; it doesn't correspond to
anything much in Fijian or elsewhere (IIRC).
(snips)
>But Set III is the kicker:
>
>"Rootless" verbs (those newly invented, or created from nouns, or
>borrowed from other languages) are made passive with the suffix
>"-tia".
>
>Now these verbs cannot all have underlying forms in -T.  Certainly
>not the borrowings, where there was no /t/ in the source language!
>But if the underlying form ends in a vowel, why is not the suffix
>"-a" applied?
>
>The final explanation, I think, is that the rule-based mechanism
>no longer works in present-day Maori.  Instead, there are now separate
passive
>suffixes "-tia", "-kia", "-ngia", "-hia", etc., of which "-tia" is the
regular
>ending, and the others are lexically specific, just like English strong
verbs.
>Of course, only a few verbs, the new ones, have become regularized as
>yet.  We can probably expect Maori, if it survives, to have some of the
>rarer verbs losing their irregular inflections in favor of -tia.
Yes, quite likely.  Similar things are happening in those Indonesian
languages that have lost or severely neutralized final C contrasts.
Buginese /-?/ > -k-, -s-, or -r- before certain verbal suffixes. (It's
immediate parent had at least final */t k s r l/).  Matthes' 1850s
Dictionary gives e.g. /nipi?/ 'thin', /nipisi/ 'make thin' and most of my
40year+ informants agreed. (It's etymologically correct, Malay tipis/nipis.)
One young guy in his 20s insisted it was /nipiri/. (And some verbs even have
a -r- and/or a -k- or -s- form, with slightly different meanings.  Analogy
at work.)