LINGUIST List 11.1675

Wed Aug 2 2000

Disc: Writing and Speech

Editor for this issue: Scott Fults <scottlinguistlist.org>

Directory

1. Nitti45, Re: 11.1671, Disc: Writing and Speech
2. JohnPhillips, Re: 11.1671, Disc: Writing and Speech
3. Ivan A Derzhanski, Re: 11.1671, Disc: Writing and Speech

Message 1: Re: 11.1671, Disc: Writing and Speech

Dear Linguist:
 In my most recent posting to the list, I mentioned some things I had 
surmised from my limited knowledge of Celtic linguistics. These things had 
to do with the respective orthographies of various of the Celtic tongues. 
Stating that my limited acquaintance with this branch of Indo-European 
precluded me from being able to speak with authority on these matters, I 
asked whether there were anyone among the readership at large who could 
indeed speak with authority on these matters and, if so, would this person or 
these persons be so good as to confirm or contradict the validity of my 
surmises.
 It is my pleasure to inform the *Linguist* that, within a mere twenty 
minutes of your having published this posting of mine, I did in fact receive 
an authoritative answer to the questions I posed. A Richard Sproat, 
Human/Computer Interface Researcher at AT&T Laboratories, not only confirmed 
my surmises, but also had some relevant information to add as well. He has 
graciously extended to me permission to post his response to the *Linguist,* 
which response I do hope you will be so good as to publish.
 To refresh the memories of the readers I shall first reproduce, in part, 
the text of my surmises. This I shall follow with the text of the response 
to my questions at large.
 
In a message dated 8/1/00 10:42:06 AM, Richard S. Kaminski writes:

>From my...limited acquaintance with
>Welsh, I would tend to concur that its orthography seems quite regular but...
>I cannot speak with authority on this. Would anyone who can indeed do
>so, please make a posting?
> ...[I]t occurred to me after my first posting to this discussion that
>I may have been slightly off base in suggesting that Irish was the only one
>of the Gaelics that had irregular orthography. As mentioned in my more
>recent posting, I received confirmation of the irregularity of Irish Gaelic
>spelling [from another source]. I really don't know, however, if this is 
true of Scots >Gaelic or of Manx. Again, can anyone out there enlighten us?

 Now for Richard Sproat's answer:

Regarding spelling in Celtic languages:

Welsh orthography is indeed very regular.

For the Goidelic languages, it's hard to say which is the most
irregular, and things are in general more complex. Irish and Scots
Gaelic orthography derives from the original romanized Gaelic
orthography, and (like English) tends to be conservative. Both
languages have probably undergone the same amount of sound changes
(indeed, political issues aside, Scots Gaelic and Irish Gaelic are
really part of a dialect continuum, with lots of differences but
plenty of shared phonological history), so it is probably fair to say
they are on a par with respect to distance between orthography and
phonology. But if I had to make a bet, I'd actually say Scots Gaelic
may be more irregular simply because Irish underwent a mild spelling
reform at around the time the Irish Republic was founded.

Manx is a different story again, since the Manx speakers became (or
always were) completely illiterate. It wasn't until the 18th century
that missionaries developed an orthography based largely on English
spelling (there were actually two such orthographies). There is some
reason to believe that the orthography was irregular even when it was
invented, but in any event, Late Manx sound changes made it much more
so. But again it's hard to compare and say that it is more irregular
than, say, Irish.

- 
Richard Sproat Human/Computer Interface Research
rwsresearch.att.com AT&T Labs -- Research, Shannon Laboratory
Tel: +1-973-360-8490 180 Park Avenue, Room E153, P.O.Box 971
Fax: +1-973-360-8809 Florham Park, NJ 07932-0000
- --------------http://www.research.att.com/~rws/-----------------------


 Thanks are due to this gentleman for this latest entry in what has turned 
out to be (in my estimation, at any rate) a most fascinating discussion.
 Cordially yours,


 Richard S. 
Kaminski
 <Nitti45aol.com>

Message 2: Re: 11.1671, Disc: Writing and Speech

The question of the spelling systems of Welsh and Gaelic keeps cropping
up, so here is a brief description.

Welsh spelling is almost phonemic, though the pronunciation it
represents is an extremely formal one which would only be heard in a
sermon or a very formal public speech or in reading aloud. The spelling
system has a continuous tradition probably dating back to Roman times,
and has changed gradually to reflect pronunciation and fashion, while
keeping fairly closely to the phonemic principle throughout. Loan words
throughout the ages have been written in Welsh spelling. There is
something of a controversy at the moment over what should be done with
international scientific words: some people follow the tradition of
spelling them in Welsh (e.g. cilomedr for kilometre), others keep the
English spelling.

Manx spelling is very irregular. It is not a historical system, but
dates from the 18th century when printing began in Manx. The basic idea
is to spell a Manx word as a word of roughly the same pronunciation
would be spelt in English. In earlier times Manx seems to have had a
different system, rather like the Welsh system, which reflected
pronunciation accurately. Presumably in very early times Manx used the
general Gaelic writing system described below, though I think no
examples are known to survive.

Scottish and Irish Gaelic both use a common Gaelic system, though a
recent reform in Irish has done away with some of the silent letters.
To exaggerate a bit (though not much), the spelling represents the
pronunciation of a thousand years or more ago. Hence pronunciation is
largely predictable from spelling, but spelling is not predictable from
pronunciation. Gaelic spelling gives the impression of being worse than
it is because of the profusion of h's, but these just represent what
were written as accents over the letters of the Gaelic alphabet (used
up until the 1960's). Old loan words are generally integrated into the
system; modern loan words tend to have exceptional spellings.

John Phillips
Dept. of Linguistics,
Yamaguchi University,
Japan

Message 3: Re: 11.1671, Disc: Writing and Speech

Quoth Nitti45aol.com (Richard S. Kaminski):
> From my (also) limited acquaintance with Welsh, I would tend
> to concur that its orthography seems quite regular but, again,
> I cannot speak with authority on this. Would anyone who can
> indeed do so, please make a posting?

Yes, it is as regular (and as faithful to the sound) as seems to
be possible for an orthography that strives to accommodate several
equally standard dialects at once (though not a huge variety).

> As mentioned in my more recent posting, I received confirmation
> of the irregularity of Irish Gaelic spelling. I really don't know,
> however, if this is true of Scots Gaelic or of Manx.

The orthography of Scottish Gaelic follows the same principles
as that of Irish; if anything, it is a little more baroque, not
having been subjected to the reform of 1948 that made Irish look
a trifle less frightening. The Manx orthography was developed by
an Englishman who had no idea what written Gaelic looked like, and
it is based on English, with all that this implies.

Gaelic and English spelling are problematic in very different ways,
however. Gaelic has an unusually low sound-to-letter ratio (many
letters are written that are no longer pronounced), and different
sequences of letters can have the same sound, so the transition
frm pronunciation to spelling can be difficult. Not so in the
opposite direction: the sound of a written word can be predicted
with a fair degree of certainty. Accents are very sensibly used
to mark long vowels, and the sound values of the individual letters
are on the whole closer to what they are in virtually all other
Roman-written languages (& IPA) than the ones assigned in English.

I shan't say that it is wrong to describe either IG or SG as an
orthographically irregular language, but if we do so, a different
description (a more emphatic one) is sorely needed for English.

- 
<fa-al-_haylu wa-al-laylu wa-al-baydA'u ta`rifunI
 wa-as-sayfu wa-ar-rum.hu wa-al-qir.tAsu wa-al-qalamu>
 (Abu t-Tayyib Ahmad Ibn Hussayn al-Mutanabbi)
Ivan A Derzhanski <http://www.math.bas.bg/~iad/>;
H: cplx Iztok bl 91, 1113 Sofia, Bulgaria <iadmath.bas.bg>
W: Dept for Math Lx, Inst for Maths & CompSci, Bulg Acad of Sciences