LINGUIST List 7.1176

Tue Aug 20 1996

Sum: Initial vowels

Editor for this issue: Ljuba Veselinova <lveselinemunix.emich.edu>

Directory

1. John Atkinson, Sum: Initial vowels

On the 8th of August, as a followup to Carsten Peust's query regarding 
consonant-final languages, I posted the following query on Linguist:

>A fair number of languages in the world (including many Australian
>languages) admit only consonants (including w and y) at the
>beginning of words. Does any one of you know of languages in which
>all words and/or syllables must begin with a vowel?

Unlike Carsten, I received only two replies -- from Philip Hamilton 
<phamiltochass.utoronto.ca> and Ernest McCarus <enmumich.edu>.
The reason appears to be two-fold:

(1) Languages closely approximating the initial-vowel-only state are 
extremely rare; in fact, apparently non-existent outside Australia; so 
no one had any to contribute! As Philip says:

>It turns out that Oykangand and Olgol/Olkola and the nearby but 
>not-that-closely-related Ogh-Unydjan [all in southern Cape York] are 
>the only true V-initial languages, although some of the Arandic 
>languages [in Central Australia] come pretty close. Nothing comes 
>very close outside Australia.

(2) As Philip pointed out, a similar question ("Is Basque unusual in 
having at least 50% vowel-initial native vowels and adjectives?") was 
asked this time last year, so the topic has already received a fair 
amount of discussion on the List, before I joined it (although with no 
mention of Australian languages). To quote from the summary, by Larry 
Trask <larrytcogs.susx.ac.uk>:

> (1) Languages in which 40% or more of nouns are V-initial are not
> exceedingly rare, but they do not appear to be at all common. The
> majority of the examples come from Niger-Congo.
> 
> (2) In Niger-Congo, the initial vowels derive chiefly from
> morphological sources, probably entirely so apart from the reduction
> of CV- prefixes to V- in some languages . . .
> 
> (3) Outside of Niger-Congo, . . . loss of initial consonants is more 
> usually invoked. 

In the Paman and Arandic examples mentioned by Philip, fairly 
well-understood processes of initial-dropping are invariably 
responsible (these languages have no prefixes).

Ernest McCarus's interesting reply dealt with consonant-initial rather 
than vowel-initial languages.

Thanks to both of you.

John