List of phonemes in Proto-Celtic, Celtic, Proto-Irish? (fwd)

[Andrew Carnie]

---------- Forwarded message ----------
Date: Mon, 04 Aug 1997 10:21:26 -0400
From: Antony Dubach Green <green at fas.ag-berlin.mpg.de>
To: CELTLING at mitvma.mit.edu
Subject: Re: List of phonemes in Proto-Celtic, Celtic, Proto-Irish?


>Can anyone suggest a phonetic list of the sounds of Proto-Celtic, Celtic,
>Gaulish, Q-Celtic, P-Celtic, Proto-Irish, and/or Old-Irish?
>
>In particular, is there any evidence of the 'ng' phone?  Were 'd' and 'g'
>once allomorphs of the same phoneme?  Did the stridents 'z' 'sh' or 'ch'
>exist?  And how about those interdental-fricative, 'th' sounds -- both
>voiced and unvoiced?  Lastly, did a 'p' sound exist early on?  I express a
>bit of confusion on the last point, being that I have read in _How the
>Irish Saved Civilization_ that the sound 'P' was non-existent in Early
>Irish.

I don't think anyone has ever published a list of the reconstructed phonemes
of Proto-Celtic.  For Old Irish, the best source is Thurneysen's Old Irish
Grammar, and of course Pedersen's Vergleichende Grammatik der keltischen
Sprachen and its abbreviated English translation by Lewis, A Concise
Comparative Celtic Grammar, will tell you how to get from Proto-IE to the
attested old Celtic languages.

But I can tell you what the probable consonant phonemes of Proto-Insular
Celtic were:

Voiceless stops:         t   k kw
Voiced stops:          b d   g gw
Voiceless fricative:     s
Nasals:                m n
Liquids:                 l r
Glides:                  y     w

As you see, there was no 'p'.  IE 'p' was lost in Celtic, but 'kw' became
'p' in some languages, including Brittonic and Gaulish.  (If you believe in
P-Celtic, this happened once, namely in Proto-P-Celtic; but if, like me, you
don't believe in P-Celtic, it happened independently in the two branches.)

'z' existed as an allophone of 's' before voiced stops.  There was no 'sh'
or 'ch' like sounds (alveopalatal fricatives or affricates).

'N' (velar nasal) was an allophone of 'n' before the velar and labiovelar
stops.  It didn't become a phoneme until the sequence 'Ng' was simplified to
'N' sometime during the Old Irish period.

The interdental fricatives arose first as allophones of t and d in leniting
environments (basically, after vowels), and later became phonemes once
lenition was no longer phonologically predictable.

'd' and 'g' have always been separate phonemes, and they still are.  In the
Middle Irish period, the lenited correspondent to 'd' shifted from the
voiced dental fricative 'dh' to the voiced velar fricative 'gh', and thus
merged with the existing 'gh' that was the lenited correspondent to 'g'.

>Also, where might one find good data and research on the earliest ogham
>markings, both on the islands, and on the continent?  I understand that
>chalk stone was unearthed near Denmark(?) in the late 19th century which
>had evidence of 'umistakable ogham'.  Were there any other continental
>discoveries of ogham?

I can't help you with this, sorry.  Maybe someone else on the list can, or
try CELTIC-L.

>Curiously, might there be a convention for describing the IPA sounds in
>ascii text?  I find rather risky the use of english letter combinations to
>represent IPA sounds  -- as in the first paragraph.

>>From time to time people propose conventions, but they never catch on with
other people.  Your best bet is to use English orthographic practices, in
combination with articulatory descriptions of the sounds in question.
(I.e., tell people, "I'm using 'N' to represent the velar nasal" or "I'm
using 'ch' to represent the voiceless alveopalatal affricate".)

adh mor,
Tonio Green
----------------------------------------------------------------
Antony Dubach Green                   green at fas.ag-berlin.mpg.de
Zentrum fuer Allgemeine Sprachwissenschaft
Jaegerstrasse 10-11
10117 Berlin, Germany          GO MAIRE NA TEANGACHA CEILTEACHA!
Tel.(49 30) 20 192 574              BYWIO YR IEITHOEDD CELTAIDD!
----------------------------------------------------------------