LINGUIST List 13.565

Fri Mar 1 2002

Sum: Voice Onset Time and Stress

Editor for this issue: Marie Klopfenstein <marielinguistlist.org>


Directory

  1. Joaquim Brand�ode Carvalho, VOT and stress

Message 1: VOT and stress

Date: Mon, 25 Feb 2002 20:04:33 +0100
From: Joaquim Brand�ode Carvalho <jbrandaoext.jussieu.fr>
Subject: VOT and stress

Last week I posted the following question to the List:

- ---------

Dear colleagues,

I am working on the possible relationships between VOT and stress. I 
am concerned in particular with alternations as the one shown by 
proto-Germanic (Verner's law) or by chinook (E. Sapir, Sound patterns 
in language, 1925) :

WApul "night" / waBULmax "nights"
(capitals = stressed syllable)

My questions are :

(1) Do you know of other examples of such alternations?

(2) Do you know of cases in which VOT either behaves differently :
i.e. *WAbul / *waPULmax,
or involves another feature :
e.g. *WAphul / *waPULmax ?

(3) In any case, could you please let me know the basic rule 
governing stress directionality in the language(s) at issue 
(left>right or right>left)?

Thank you very much for your help. I'll post a summary of the responses.

- ---------

I wish to thank those who responded to my query :

Michael Johnstone
Robert Port
Daniel Loehr

As can be seen, the number of responses is rather small, and only one 
deals with VOT proper. Are the facts involved by Verner's law and 
Chinook alternations so rare in the world languages? Furthermore, two 
of the responses mention related facts in English phonology that, in 
some way, seem contradictory with respect to the WApul / waBULmax 
pattern. I was aware of such phenomena, and this was one of the 
reasons of my query. Here are the three responses received. I shall 
add a brief comment at the end of each one.

- -------------

[Robert Port]
The question that occurs to me is that you don't really mention anything
about VOT. Have you looked at it? That is, is the P in wapul aspirated or
unaspirated?
On the assumption that you are a Portuguese speaker, I am assuming that the
P is unaspirated and the B is fully voiced. This alternation reminds me of
Korean where they have an unaspirated stop in initial position that 
becomes fully
voiced intervocalically. But the alternation you report here is definitely
unusual.
[End of quote]

I suppose, according to Sapir's quotation, that Chinook [p] is not 
aspirated. In this sense, this alternation resembles that of Korean, 
where only the 'weakest' member of the well-known triplet /ph, p*, p/ 
has voiced allophones. However, as far as I remember, voiced 
alternants occur intervolically therein, regardless of any prosodic 
factor such as stress, but I may be wrong.

- --------------

[Daniel Loehr]
This may not be quite what you're looking for, but there is the familiar
American English alternation:

 atom ['aeD m] ('ae' = low front vowel, 'D' = flap, '' = schwa)
 (stress on the first syllable, and the 't' is a voiced flap)

 atomic [ 'tam ik]
 (stress on the second syllable, and the 't' is voiceless
aspirated)

This is an alternation in the other direction from your example - i.e. the
change is from voiced to voiceless when you add the suffix. The stress
directionality would be from left to right in this case, since both words
are following an English rule of stress (for these words anyway) of stress
on the penultimate syllable. Adding a suffix moves the stress to the right.

If I remember correctly (I don't have my notes in front of me), the usual
explanation for this is that flapping in American English is sensitive to
foot boundaries. In "atom" the 't' is inside a foot and is flapped. In
"atomic" the 't' is at a foot boundary and is not flapped.

Again, I'm not sure if this is relevant to your inquiry, but thought you
might be interested.
[End of quote]

Indeed, this is interesting, since, as Daniel Loehr notes, the weak 
term of the alternation ([D]) occurs in the context where Chinook has 
the strong (voiceless) obstruent. Hence my question (which is a mere 
conjecture) about the directionality of stress in Chinook-type 
languages, if any.

- --------------

[Michael Johnstone]
Pre-glottalisation of stops in certain dialects of English e.g. in London
is sensitive to stress. This pre-glottalisation (a creaky end to a
preceding period of voicing, transcribed here as [?]) only occurs with
[p,t,k] when followed by a consonant or pause. (This covers also the
affricate [tS].)

However, pre-glottalisation cannot occur when stress follows, IF the stop
+ following consonant could form a valid syllable onset.

Hence 'kitchen' [?tS], 'diploid' [?pl], 'beetroot' [?tr] with preceding
stress, vs. 'achieve' [tS], 'deploy' [pl], 'betray' [tr] with following
stress. This produces alternations in related words such as 'apply/
application'.

But 'atlas' [?tl] with preceding stress and 'atlantic' [?tl] with
following stress. (Note that [tl] is not a valid onset.)

This is similar to those accents where complete replacement of [t] with
[?] occurs - 'better' [?] and preceding stress, but 'attack' [t] and
following stress; and accents where pre-glottalisation occurs before
vowels - 'paper' with [?p] but 'appear' [p].

In these examples of pre-glottalisation and glottal replacement, both
phenomena occur word-finally but not word-initially, and word-medially
before unstressed vowels but not stressed ones.

The Verner's / Chinook pattern is similar to this in as much as the
voiceless consonant occurs word-medially before unstressed vowels, and
voicelessness is elsewhere associated with word-final consonants (German,
Russian etc.)

The link, I suppose, is increasing articulatory effort when a stressed
vowel follows, vs. decreasing effort when an unstressed vowel or pause
follows.
[End of quote]

Here, the relationship between the English facts and the Chinook 
alternation may not be opposite, as in the previous case, but, 
assuming that pre-glottalisation or glottal replacement are weakening 
phenomena, one can ask whether final devoicing is to be viewed as a 
lenition process. The same question arises from [p]-preservation in 
WApul, while voiced consonants are usually said to involve less 
effort than their voiceless counterparts.

- ---------------------

In sum, too many things remain problematic about these 
stress-sensitive alternations, which is, of course, exciting for the 
phonologist.
Once again, many thanks to those who responded.
- 

Joaquim Brandao de Carvalho
320, rue des Pyr�n�es
75020 Paris France
Tel./fax : 01 43 66 95 24
(If calling from outside France, please replace
the prefix '0' with the country number '33'.)
jbrandaoext.jussieu.fr

Departement de linguistique
Faculte des Sciences Humaines et Sociales - Sorbonne
Universite Rene Descartes - Paris V
CNRS : UMR 7018, GDR 1954

** If you send an attachment,
please use the RTF format **
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue