LL-L "Phonology" 2002.01.14 (10) [E]

Tue Jan 15 04:11:18 UTC 2002

======================================================================
 L O W L A N D S - L * 14.JAN.2002 (10) * ISSN 189-5582 * LCSN 96-4226
 Web Site: <http://www.geocities.com/sassisch/rhahn/lowlands/>
 Rules: <http://www.geocities.com/sassisch/rhahn/lowlands/rules.html>
 Posting Address: <lowlands-l at listserv.linguistlist.org>
 Server Manual: <http://www.lsoft.com/manuals/1.8c/userindex.html>
 Archive: <http://listserv.linguistlist.org/archives/lowlands-l.html>
=======================================================================
 A=Afrikaans Ap=Appalachian D=Dutch E=English F=Frisian L=Limburgish
 LS=Low Saxon (Low German) S=Scots Sh=Shetlandic Z=Zeelandic (Zeeuws)
=======================================================================

From: Sandy Fleming [sandy at scotstext.org]
Subject: "Phonology"

> From: R. F. Hahn <sassisch at yahoo.com>
> Subject: Phonology
>
> (2) would be an "optional" (dialect- or mode-specific) rule like the one
> that changes syllable-final /t/ to a glottal stop if it directly follows
> a vowel:
>
> (3) t -> ? / [+vocal]__$
>
> ... and a following syllable-initial /t/ (gemination) assimilates to it:
>
> (4) t -> ? / ?__
>
> Thus a geminate glottal stop, if you will.
>
> Hence ...
>
>   /pit/  /pitt at n/
> 3  pI?    pI?t at n/
> 4   -     pI??@n
> (   -     pI?n)
>
> There may well be a more "elegant" solution to this.

I've been trying to figure out the glottalisation rules in Scots,
distinguishing between the "unconstrained" glottalisation where
a word like "hotel" is pronounced /ho'?E:l/, but which some
speakers find uncomfortable and pronounce /ho'tE:l/ (which I'm
calling "constrained" glottalisation).

For unconstrained glottalisation I think this rule works (without
being sure if my use of the notation is accurate!):

(1) t -> ? / [+sonorant]__

In other words, [t] is glottalised after a sonorant (a sonorant
being all vowels, and /l/, /r/, /n/, /m/, /N/).

Anyway, this works correctly for, for example,

bottle: /'bot=l/ -> /'bo?=l/,
tatties: /'ta:?Iz/ -> /'ta:?Iz/
hotel: /ho'tE:l/ -> /ho'?E:l/
want: /wQ:nt/ -> /wQ:n?/
saft: /sQ:ft/ -> /sQ:ft/  (no change, /f/ isn't a sonorant).
selt: /sE:lt/ -> /sE:l?/
sort: /sort/ -> /sor?/

and I don't think the sequences /mt/, /Nt/ occur, though I
feel that if they did, I would glottalise them! A compound
like "wingtip" wouldn't be glottalised, though. Maybe a
refinement to the rule is needed here.

Then for "constrained" glottalisation I need an optional
constraint to the rule saying "except if it is at the start
of a stressed syllable". This prevents glottalisation in
"hotel", for example. I can't quite make out how to express
this in the notation yet, unfortunately.

The /t/ in "pitten" isn't actually doubled - it's
/'pIt at n/ -> /'pI?@n/. You're right, though, in deleting the
"@" eventually - this is a rule I've been working on that can
be formulated as (though I don't think I've got it quite
accurate yet - in fact it might be rubbish, I only taught
myself this "features" business over the Christmas holidays!):

|[-vocal]     ||[+reduced]| n -> |[+consonantal]|/V$__#
|[-voice]     ||                 |[+nasal]      |
|[-continuant]||                 |[+sonorant]   |
                                 |[+syllabic]   |

There are quite a few guesses in that, but it's meant to say
"a voiceless stop following a stressed syllable, followed by
a schwa, followed by a /n/ becomes a voiced syllabic nasal
sonorant".

Examples:

pitten: /'pIt at n/ -> /'pIt=n/ ( -> /'pI?=n/)
stappin: /'sta:pIn/ -> /'sta:p=m/
leukin: /'ljukIn/ -> /ljuk=N/
steak and kidney: /'ste:k @n 'kIdnI/ -> /'ste:k=N 'kIdnI/
puddock an princess: /'pVd at k @n prInsE:s/ (no change because the <dock>
syllable is unstressed).

Thanks for the suggestions on the [xW] problem - I'll need
to spend some time going through them tomorrow! One other
thing I forgot about is that the /W/ becomes /f/ in
northeastern dialects (but not the /W/ in /tWIst/) - maybe
this could just be a final optional rule or maybe it will
shed more light on the whole process!

Oh, by the way - I deliberately copied Andy's way of handling
IPA because it's tried and tested, but I couldn't see his
stylesheet. I assume he has specified unicode fonts in his
stylesheet that I've missed - with luck he'll let me see his
stylesheet and then everything should be fine!

Sandy
http://scotstext.org
A dinna dout him, for he says that he
On nae accoont wad ever tell a lee.
                          - C.W.Wade,
                    'The Adventures o McNab'

----------

From: R. F. Hahn <sassisch at yahoo.com>
Subject: Phonology

Sandy, you wrote:

> (1) t -> ? / [+sonorant]__
>
> In other words, [t] is glottalised after a sonorant (a sonorant
> being all vowels, and /l/, /r/, /n/, /m/, /N/).
>
> Anyway, this works correctly for, for example,
>
> bottle: /'bot=l/ -> /'bo?=l/,
> tatties: /'ta:?Iz/ -> /'ta:?Iz/
> hotel: /ho'tE:l/ -> /ho'?E:l/

That works well.  I hadn't been sure about liquids and nasals.

> |[-vocal]     ||[+reduced]| n -> |[+consonantal]|/V$__#
> |[-voice]     ||                 |[+nasal]      |
> |[-continuant]||                 |[+sonorant]   |
>                                  |[+syllabic]   |

Unless it affects more than one phoneme, you don't have to go through
all of this.  The following would suffice, since only /t/ is involved:

t -> ? / V__$

... and then the assimilation rule:

t -> ? / ?__

> There are quite a few guesses in that, but it's meant to say
> "a voiceless stop following a stressed syllable, followed by
> a schwa, followed by a /n/ becomes a voiced syllabic nasal
> sonorant".
>
> Examples:
>
> pitten: /'pIt at n/ -> /'pIt=n/ ( -> /'pI?=n/)
> stappin: /'sta:pIn/ -> /'sta:p=m/
> leukin: /'ljukIn/ -> /ljuk=N/
> steak and kidney: /'ste:k @n 'kIdnI/ -> /'ste:k=N 'kIdnI/
> puddock an princess: /'pVd at k @n prInsE:s/ (no change because the <dock>
> syllable is unstressed).

It's the same in Low Saxon (Low German), i.e., deletion of "schwa"
(which in many dialects is a short [e] rather than a real schwa, [@]),
and then assimilation of /n/ (though no t -> ?):

piepen: ['p`i:p=m] (to whistle; cf. to pipe)
bieten: ['bi:t=n] (to bite)
lieken: ['li:k=N] (to resemble; cf. to like)

As for stress, I think it's all right to say [+stress] and, if
necessary, add "stress = primary stress in a given sequence" in a
footnote.

Conventions and models in phonology are here today and gone tomorrow.
This generative model is considered outmoded by many, anyway.

Regards,
Reinhard/Ron

==================================END===================================
 You have received this because your account has been subscribed upon
 request. To unsubscribe, please send the command "signoff lowlands-l"
 as message text from the same account to
 <listserv at listserv.linguistlist.org> or sign off at
 <http://linguistlist.org/subscribing/sub-lowlands-l.html>.
=======================================================================
 * Please submit postings to <lowlands-l at listserv.linguistlist.org>.
 * Postings will be displayed unedited in digest form.
 * Please display only the relevant parts of quotes in your replies.
 * Commands for automated functions (including "signoff lowlands-l") are
   to be sent to <listserv at listserv.linguistlist.org> or at
   <http://linguistlist.org/subscribing/sub-lowlands-l.html>.
 * Please use only Plain Text format, not Rich Text (HTML) or any other
   type of format, in your submissions
=======================================================================