Possible phonological changes (was: Rate of change)

Larry Trask larryt at cogs.susx.ac.uk
Thu Jun 14 09:15:38 UTC 2001


--On Friday, June 8, 2001 7:47 am +0300 Ante Aikio
<anaikio at mail.student.oulu.fi> wrote:

> Just out of curiosity: how strange are the changes in these Pama-Nyungan
> languages?

OK.  First, a correction.  I gave Barry Blake credit for working this stuff
out, but I was quoting from memory.  In fact, it was Ken Hale who did the
initial work, and then it was Terry Crowley who did the undergrad
dissertation which figured out the nasty cases.  Apologies all round.

The changes I had in mind were the initial-dropping languages of Australia.

In a typical Pama-Nyungan language, the phoneme inventory is small; there
are no fricatives; every word must begin with a single consonant; and
monosyllabic word-forms are prohibited.  All of these statements fail to
hold in the exceptional languages.

What happened in the exceptional languages was this.  First, all
word-initial consonants were lost.  Second, in most cases, the first mora
of the following vowel was also lost.  So, for example, before we consider
the rest, original *<guda> becomes <da>, while original *<guuda> becomes
<uda>.  But there's more.

Phonetic features of the lost segments, such as labiality and palatality,
were copied onto the remaining material in one way or another (and lost
velars often produced rhoticity in the surviving material).  In addition,
there were a number of lenitions in the surviving consonants (often
producing fricatives), but also some fortitions (especially of nasals and
liquids).  Finally, some word-final material was frequently also lost or
gained.

The changes in any given language are usually highly regular, though it
took a long time to work out the rules, and in fact it took a long time
even to identify the cognates.  It is not obvious, for example, that
Mbabaram <dOg> 'dog' (O = turned c; the word is amusingly identical to
English 'dog'), Mpalityanh <twa> 'dog', and Ntra?ngith <u?a> 'dog' are
cognate and all derived from original *<gudaga> -- which survives unchanged
in Yidiny (hiya, Steve).

In Yinwum, original *<wanta-> 'put' becomes <ntra->, and original *<yinta->
'(to) spear' becomes <nti->.  In LinngithiG (G = gamma), original
*<ngungku> 'there' becomes <ko>, and original *<ngantu> 'where' becomes
<tro>.  You can see why it took a while to find the cognates.

For an overview of this, see pp. 195-207 of this book:

R. M. W. Dixon (1980), The Languages of Australia, Cambridge UP.

For more information, see several of the papers in this book:

P. Sutton (ed.) (1976), Languages of Cape York, Canberra: Australian
Institute of Aboriginal Studies.

> Most examples of bizarre changes that I know are from various Samoyedic
> languages, e.g. *w- > q-, *j- > q-, *V- > ngV-, *mp > ngf and even *s- >
> k- before /e/ and /i/! But I'd be most interested in other examples of
> uncommon phonetic developments.

Look up the entry for 'unnatural change' in my dictionary:

R. L. Trask (2000), Dictionary of Historical and Comparative Linguistics,
Edinburgh UP.

This cites, with references, a number of odd-looking changes (N = [eng]):

  t > k
  -b- > -k-
  l > Ng
  w > ts-, -nts-
  ts > kw OR p OR f
  t > w
  d > N
  l > N

The examples come from Austronesian, Papuan and Athabaskan.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk

Tel: (01273)-678693 (from UK); +44-1273-678693 (from abroad)
Fax: (01273)-671320 (from UK); +44-1273-671320 (from abroad)



More information about the Indo-european mailing list