I'm certainly not such a person, but here's what Tiit-Rein Viitso in the
Routledge book on Uralic has to say on the matter:

"In all Fennic dialects [...] most words have primary stress on the
first syllable. Quadrisyllabic and longer words have secondary stress on
odd non-final syllables, i.e. there is a tendency towards trochaic
stress patterning [OK, we all knew that]. [This] is counteracted by
certain derivational suffixes that attract secondary stress , e.g.
Finnish 'pakenem`ninen, escaping. In addition, in Finnish [...],
quinquesyllabic and longer words with a short third and a long fourth
syllable have their secondary stress on the fourth syllable, e.g.
'todel'lisenl, real (sing Genitive), todelli'sessa, real (sing
Inessive).
In some Finnish dialects [there are further complications]."

So, consider yourself confirmed.

Pakeneminen, actually. But yes, that's how it works. I can think of
only two words which have normally non.initial stress, and one of them
is a recent loanword from English: "jumalauta" and "okei". Both can
have initial stress too.

I
think it's really unfortunate that an accent is used to distinguish "pêra"
("pear") from the obsolete preposition "pera" (which I know only from
dictionaries), while "fora" ("had gone/been") and "fora" ("outside",
"except") are indistinguishable in writing even though they do occur in
similar contexts and are not homophones.

- Some frequent words which wouldn't normally need a stress mark bear
one in order to be distinguished from other frequent words: pa´ra
'stops' / para 'to', da´ 'gives' / da 'of the', e´ 'is' / e 'and', and
countless others, especially monosyllables. NB a distinction is only
made if the vowels are really different.

1) When a 'stronger' stress mark ' or ^ overrides it. Example: sótão,
bênção.

There is no concept of relative strength of stress marks in portuguese.

The available diacritics either mark stress or they don't. This is a
fundamental principle of the orthography.

Breaking it is simply not an
option. For instance, they tried to break it by having adverbs formed
from -e^s words keep the ^ (*corte^smente). It didn't work.

2) For morphological reasons, when suffixes are added. Examples:
mãozinho, manhãzinha.

Just what it is that happens for morphological reasons? (I know what it
is; I'm pointing out that this item is not homomorph with the preceding
one.)

OF COURSE one can build all kinds of rules to end up with the same
result. The problem here is that a good orthographic system follows
structuring principles, which in turn ideally take advantage of the
language's phonology, phonotaxis and yes, morphology - and, at least for
portuguese, in that order. It's not just a random house of cards to
which you add at will as long as the output is the same.

To gleefully violate two important principles [...]

as these ones* perhaps suggests that your understanding of
the orthography is not the same one found in the normal
portuguese speaker.

And the reason I dwell so much on this is that I wish you
understood it so that you wouldn't mislead other people.

and in -ão one could also count the ~ as a stress mark, as
is the case in manhã etc.

NO, NO, NO, NO, NO, NO, NO, NO, NO, NO, NO, NO, NO, NO, NO, NO, NO, NO,

On May 1, 12:30 am, "Ekkehard Dengler" <ED...@t-online.de> wrote:
I
think it's really unfortunate that an accent is used to distinguish
"pêra" ("pear") from the obsolete preposition "pera" (which I know
only from dictionaries), while "fora" ("had gone/been") and "fora"
("outside", "except") are indistinguishable in writing even though
they do occur in similar contexts and are not homophones.

Oh, I wholeheartedly agree. Even if I don't think the two 'foras' are
much prone to confusion (mostly since the pluperfect is not that
common, I suppose), I suppose the matter with this issue is that the
marked form would probably be 'fôra', and for consistency all persons
would have the ^, even the P3 which isn't homograph with anything.

On Apr 19, 7:40 am, p.nummi...@suomi24.fi wrote:

How should one find pronunciation dictionaries (printed books that use
IPA)?

Is there a pronunciation dictionary for Spanish? I haven't been able
to find any. And no, I don't want a web site with audio files.

Are there words in Spanish whose pronunciation is not predictable from
their orthography? How many?

IPA isn't particularly useful in dictionaries of pronunciation,
because its symbols aren't language-specific.

You'd need to read pages
and pages of introduction to know what they specifially indicate for
each language and dialectal variation.

benli...@ihug.co.nz> wrote in message

Surely the point of IPA is that you immediately get a tolerable
aproximation.  Writing 'y' for 'ɯ' and calling the resulting scheme IPA
strikes me as dishonest.  It's also badly misleading.  (It's a fairly common
abuse for Thai.)

Also, IPA dates as pronunciations change.
Not unless there are phonemic changes, in which case  both systems
will date.
Isn't /�/ for Maori <wh> unnatural when the typical pronunciation is now
more like [f]?
Who would use inverted-w in a phonemic transcription?

If a Maori dictionary had been written using the IPA early enough, it would
use /�/, and by your arɡument that IPA does not date without phoneme
chanɡes, the transcription would stick.

On Apr 19, 4:43 am, "Peter T. Daniels" <gramma...@verizon.net> wrote:

On Apr 19, 7:40 am, p.nummi...@suomi24.fi wrote:

How should one find pronunciation dictionaries (printed books that use
IPA)?

Is there a pronunciation dictionary for Spanish? I haven't been able
to find any. And no, I don't want a web site with audio files.

Are there words in Spanish whose pronunciation is not predictable from
their orthography? How many?

Izquierda. Is <z> [T] or [s]? Is <i> [j] or [i]?

IPA isn't particularly useful in dictionaries of pronunciation,
because its symbols aren't language-specific.

Is it useful for dictionaries of (standard) pronunciation of a single
language? How would IPA be used to describe to an  Anglophone that
"would've" is different from "wood of"?

You'd need to read pages
and pages of introduction to know what they specifially indicate for
each language and dialectal variation.

I've used devnAgri to describe pronunciations of proper names (eg.,
Schwartzaneger, Natalya. Dallas) to Indians who didn't have Hindi for
a native language. IPA is finer; so it would seem that it ought to be
better for describing pronuciations. It is not clear that it is,
though, especially for describing pronunciations to people who have
learnt IPA from respellings. For example, one who assumes a certain
phonetic value of [e] by seeing IPA respellings of English words in
the latest OED (would) wrongly pronounce a German's IPA respelling of
Beethoven as (non-rhotic) Beartoven.

benli...@ihug.co.nz> wrote in message

news:414006ff-0bde-402b-99b0-6c04fcdf3f20@m1g2000pre.googlegroups.com...



On Apr 29, 11:19 am, "Richard Wordingham" <jrw0...@yahoo.co.uk> wrote:
benli...@ihug.co.nz> wrote in message

news:bf7ef403-3e71-4ad5-9b75-e09c6967872b@z24g2000prf.googlegroups.com....

On Apr 28, 12:50 pm, "Richard Wordingham" <jrw0...@yahoo.co.uk> wrote:
'bomb' ?? Are you talking about the 'bum' pronunciation (given as an
alternative in OED Online, but not mentioned in any other source I've
looked at, and previously unknown to me)?
Or the American variation
between -a- and -o- type vowels in this and many other words. I don't
see that either system handles either problem neatly.
I was thinking of the latter. Are you saying that there is too much
selectional variation in these words?
I think there's a lot, from what I read. (I tend not to notice it, and
I find it mysterious because it doesn't correspond to any contrast in
my dialect.) But my point was that modified-orthography doesn't deal
with this any better than IPA-phonemic.

Modified orthography can deal with the correspondences, so long as it
doesn't get overwhelmed with vast numbers of different correspondence sets..
IPA with diacritics for different correspondence sets is not IPA.

'cloth' : vs. 'clawth'? My COD (7th ed) lists them as alternatives. If
you tried to subsume them under a modified spelling, it would work for
'broth', but not for 'moth' or 'Goth'.
Enlighten me. What's the difficulty?
It was you that suggested there was a difficulty. Remember you
suggested these words as problems for IPA-based phonemic
representation. I'm pointing out that the same problems exist for
modified orthography systems. In this case you end up having to list
the variants in both systems, or, if you try to represent them by a
special notation or rule, it applies to maybe two words.

A possible example of that

selection rule as 'cloth'. The solution, of course, is a judicious mix of
the two systems.

The Oxford Dictionary of English
Etymology captures the variation by writing /klɔ̀þ/ but /mɔþ/, which is
bastardised IPA.
You seem to feel that the "funny letters" of IPA like open-o and
inverted-v must be preserved from "bastardization" and used in
precisely the phonetic senses defined by D.Jones, even in a phonemic
orthography. But IPA also includes all 26 letters of the normal
alphabet. Do we have to observe the same taboo with those?

The same standards, yes. I'm not asking for a narrow transcription.

A grave accent signifies a low tone! That is why I regard distinguishing
diacritics loke this as a bastardisation of IPA. Usinɡ an accent is not out
of place in a modified orthoɡraphy, but is for IPA.

This would
pretty much make any Roman-based phonemic transcription unworkable.

Surely the point of IPA is that you immediately get a tolerable
aproximation.

strikes me as dishonest. It's also badly misleading. (It's a fairly common
abuse for Thai.)

Note that the annotated spelling system will need other interpretation
rules, e.g. to explain that <ar> with ligature in 'farce' sounds the
same as <ah> in 'grass'. or that ligatured <ir>, <er>, and <ur> in
'first', 'terse' and 'curse' all sound the same, unless you're
Scottish.

That's the difference between use by a native speaker and by a non-native
speaker.
Well, even native speakers aren't necessarily going to be familiar
with all dialects.

No, but once you become acquainted with 'educated Scottish English' (I may
have the jargon term wrong - I mean Standard English with a Scottish
accent), the dictionary can advise you of cases where the spelling does not
accord with the spelling, as well as covering cases such as _word_ by giving
its pronunication as /wurd/. And it immediately gives the right information
to a speaker of educated Scottish English.

What would these native speakers be doing with this information? Would
they be straining their ears and vocal cords to capture some elusive
distinction in pronunciation between <er> and <ir>, which isn't there?
The problem with
an IPA system is that it will say that 'farce' and 'grass' rhyme if it is
based on RP or similar.
Why is that a problem? If it's based on RP, they do.

Not in educated Scottish English or General American.

Also, IPA dates as pronunciations change.
Not unless there are phonemic changes, in which case both systems
will date.
Isn't /�/ for Maori <wh> unnatural when the typical pronunciation is now
more like [f]?
Who would use inverted-w in a phonemic transcription?

If a Maori dictionary had been written using the IPA early enough, it would
use /�/, and by your arɡument that IPA does not date without phoneme
chanɡes, the transcription would stick.

You then have to anchor the phonemes by reference to particular words.
The
vowels of British 'moat' and non-rhotic British 'bare' are particular
cases - 'IPA' actually becomes just the script for another writing
system.
??

Brian also raised this point. In the Oxford Dictionary of English
Etymology, the pronunications are given as /mout/ and /bɛəɹ/. For many
Britons, the pronunications are now /məʊt/ and /bɛ�/. (/ɹ/ is being used to
code an r-sound in rhotic accents and nothing in non-rhotic accents.)

It's bad enough that the inverted
'v' is traditionally misused for a central vowel.
I think this is a particularly sad case of IPA symbols effectively coming
to
be used arbitrarily.
There may be something we can agree on. I think some IPA-based English
transcriptions suffer from spurious
phonetic precision. I would not like to have to get along without
schwa or inverted-v in writing English, but I think epsilon, upsilon,
and that horrible turned-script-a thing (Cardinal 13) are unnecessary
if you are talking about the phonemics of English-in-general.

You can certainly get away without upsilon if you accept that 'pure' vowels
can be divided into short and long vowels, but epsilon is not so easy to
dispense with. There are native British English speakers who have /e�/ for
standard /ei/ (or /eɪ/ to be narrower); I have already mentioned /ɛ�/.

Cardinal [ʌ] is unrounded [o], so usinɡ it for a central vowel is wronɡ. It
leads to such monstrosities as usinɡ [ʌ] for the typical Indian short /a/.

This is obviously a very sore point for you, and I don't want to make

On May 1, 10:25 pm, "ranjit_math...@yahoo.com"

ranjit_math...@yahoo.com> wrote:
On Apr 19, 4:43 am, "Peter T. Daniels" <gramma...@verizon.net> wrote:

On Apr 19, 7:40 am, p.nummi...@suomi24.fi wrote:

How should one find pronunciation dictionaries (printed books that use
IPA)?

Is there a pronunciation dictionary for Spanish? I haven't been able
to find any. And no, I don't want a web site with audio files.

Are there words in Spanish whose pronunciation is not predictable from
their orthography? How many?
Izquierda. Is <z> [T] or [s]? Is <i> [j] or [i]?
Doesn't that depend on the dialect?

How would IPA be used to describe to an Anglophone that
"would've" is different from "wood of"?
One would hope that it wouldn't, since it isn't.

On Apr 29, 11:19 am, "Richard Wordingham" <jrw0...@yahoo.co.uk> wrote:
benli...@ihug.co.nz> wrote in message

news:bf7ef403-3e71-4ad5-9b75-e09c6967872b@z24g2000prf.googlegroups.com...

On Apr 28, 12:50 pm, "Richard Wordingham" <jrw0...@yahoo.co.uk> wrote:

'bomb' ?? Are you talking about the 'bum' pronunciation (given as an
alternative in OED Online, but not mentioned in any other source I've
looked at, and previously unknown to me)?

Or the American variation
between -a- and -o- type vowels in this and many other words. I don't
see that either system handles either problem neatly.

I was thinking of the latter. Are you saying that there is too much
selectional variation in these words?

I think there's a lot, from what I read. (I tend not to notice it, and
I find it mysterious because it doesn't correspond to any contrast in
my dialect.) But my point was that modified-orthography doesn't deal
with this any better than IPA-phonemic.

'cloth' : vs. 'clawth'? My COD (7th ed) lists them as alternatives. If
you tried to subsume them under a modified spelling, it would work for
'broth', but not for 'moth' or 'Goth'.

Enlighten me. What's the difficulty?

It was you that suggested there was a difficulty. Remember you
suggested these words as problems for IPA-based phonemic
representation. I'm pointing out that the same problems exist for
modified orthography systems. In this case you end up having to list
the variants in both systems, or, if you try to represent them by a
special notation or rule, it applies to maybe two words.

The Oxford Dictionary of English
Etymology captures the variation by writing /klɔ̀þ/ but /mɔþ/, which is
bastardised IPA.

You seem to feel that the "funny letters" of IPA like open-o and
inverted-v must be preserved from "bastardization" and used in
precisely the phonetic senses defined by D.Jones, even in a phonemic
orthography. But IPA also includes all 26 letters of the normal
alphabet. Do we have to observe the same taboo with those?

This would
pretty much make any Roman-based phonemic transcription unworkable.

Note that the annotated spelling system will need other interpretation
rules, e.g. to explain that <ar> with ligature in 'farce' sounds the
same as <ah> in 'grass'. or that ligatured <ir>, <er>, and <ur> in
'first', 'terse' and 'curse' all sound the same, unless you're
Scottish.

That's the difference between use by a native speaker and by a non-native
speaker.

Well, even native speakers aren't necessarily going to be familiar
with all dialects.

What would these native speakers be doing with this information? Would
they be straining their ears and vocal cords to capture some elusive
distinction in pronunciation between <er> and <ir>, which isn't there?

The problem with
an IPA system is that it will say that 'farce' and 'grass' rhyme if it is
based on RP or similar.

Why is that a problem? If it's based on RP, they do.

Also, IPA dates as pronunciations change.
Not unless there are phonemic changes, in which case both systems
will date.

Isn't /�/ for Maori <wh> unnatural when the typical pronunciation is now
more like [f]?

Who would use inverted-w in a phonemic transcription?

You then have to anchor the phonemes by reference to particular words.
The
vowels of British 'moat' and non-rhotic British 'bare' are particular
cases - 'IPA' actually becomes just the script for another writing
system.

??

It's bad enough that the inverted
'v' is traditionally misused for a central vowel.

I think this is a particularly sad case of IPA symbols effectively coming
to
be used arbitrarily.

There may be something we can agree on. I think some IPA-based English
transcriptions suffer from spurious
phonetic precision. I would not like to have to get along without
schwa or inverted-v in writing English, but I think epsilon, upsilon,
and that horrible turned-script-a thing (Cardinal 13) are unnecessary
if you are talking about the phonemics of English-in-general.

No, but once you become acquainted with 'educated Scottish
English' (I may have the jargon term wrong - I mean
Standard English with a Scottish accent),

the dictionary can advise you of cases where the spelling
does not accord with the spelling,

as well as covering cases such as _word_ by giving its
pronunication as /wurd/.

And it immediately gives the right information to a
speaker of educated Scottish English.

The problem with an IPA system is that it will say that
'farce' and 'grass' rhyme if it is based on RP or
similar.

Why is that a problem? If it's based on RP, they do.

Not in educated Scottish English or General American.