Japanese katakana phonetic patterns?
Sonja Elen Kisa
Latin, German, etc.) to Japanese pronunciation?
For example:
Spider-Man > supaidāman
Mozilla Firefox > mojira faiāfokkusu
centaur > kentaurosu (from Ancient Greek here)
etc.
I'd love to have a program that converts to these words, or a
dictionary of such words...
Sonja
Christian Weisgerber
> Are there any rules or patterns to convert loanwards from English (or
> Latin, German, etc.) to Japanese pronunciation?
I am fairly certain that sometime in the last 12 months I have read
a description of the rules for mapping English to Japanese phonology--
somewhere on the web, where else?--but even after thinking about
it for a day I can't remember the exact place.
--
Christian "naddy" Weisgerber na...@mips.inka.de
Joachim Pense
> Sonja Elen Kisa <son...@gmail.com> wrote:
>
>> Are there any rules or patterns to convert loanwards from English (or
>> Latin, German, etc.) to Japanese pronunciation?
>
> I am fairly certain that sometime in the last 12 months I have read
> a description of the rules for mapping English to Japanese phonology--
> somewhere on the web, where else?--but even after thinking about
> it for a day I can't remember the exact place.
>
The transformation into Japanese loandwords is based on the sound, not on
the spelling in the source language. So, if you start with a written
English text, your conversion engine must contain a complete English
dictionary, because the pronounciation cannot be reliably derived from the
written text in English. In many other source languages, you might get away
with a corpus of rules and a list of exceptions.
Joachim
LEE Sau Dan
Sonja> Are there any rules or patterns to convert loanwards from
Sonja> English (or Latin, German, etc.) to Japanese pronunciation?
Yes, I think so, because I can easily guess the English words from the
Katakanas, and I can easily guess how certain English words are
Japanized as Katakanas. The rules are not easy to describe, though. It
involves a basic knowledge of Japanese phonology.
Sonja> For example: Spider-Man > supaidāman
Sonja> Mozilla Firefox > mojira faiāfokkusu
Sonja> centaur > kentaurosu (from Ancient Greek here) etc.
Sonja> I'd love to have a program that converts to these words, or a
Sonja> dictionary of such words...
You should first realize from your examples that the conversion is not
based on the English spellings, but the pronunciation. So, the "i" in
"spider" is rendered as "ai" in Japanese, not "i". Another example:
"ll" in "mozilla" becomes a single "r". Also note that in most cases,
the conversions are based on non-rhotic English pronunciations. So,
"er" in "spider" becomes "ā" instead of "aru", and "fire" becomes "faiā"
instead of "fairu".
Therefore, to have a program to do that conversion, the program would
need a phonetic dictionary so that it can look up non-rhotic English
pronunciations for every word it is asked to convert.
Does it now sound more difficult than you have thought at first?
--
Lee Sau Dan 李守敦 ~{@nJX6X~}
E-mail: dan...@informatik.uni-freiburg.de
Home page: http://www.informatik.uni-freiburg.de/~danlee
Ruud Harmsen
<dan...@informatik.uni-freiburg.de>: in sci.lang:
>Therefore, to have a program to do that conversion, the program would
>need a phonetic dictionary so that it can look up non-rhotic English
>pronunciations for every word it is asked to convert.
Or rhotic pronunciations, because non-rhotics can be reliably derived
from these (although not always in the other direction).
--
Ruud Harmsen, http://rudhar.com
Joachim Pense
>>>>>> "Sonja" == Sonja Elen Kisa <son...@gmail.com> writes:
>
> Sonja> Are there any rules or patterns to convert loanwards from
> Sonja> English (or Latin, German, etc.) to Japanese pronunciation?
>
> Yes, I think so, because I can easily guess the English words from the
> Katakanas, and I can easily guess how certain English words are
> Japanized as Katakanas. The rules are not easy to describe, though. It
> involves a basic knowledge of Japanese phonology.
>
But that doesn't seem to be that difficult to me. It's the _English_
phonology that causes trouble.
>
> Therefore, to have a program to do that conversion, the program would
> need a phonetic dictionary so that it can look up non-rhotic English
> pronunciations for every word it is asked to convert.
>
Here, you are saying it yourself.
Joachim
Joachim Pense
Fine, but you'll need the non-rhotic ones at the end, because the Japanese
loans are based on them.
Joachim
Bart Mathias
> Christian Weisgerber (in sci.lang):
>
>> Sonja Elen Kisa <son...@gmail.com> wrote:
>>
>>> Are there any rules or patterns to convert loanwards from English (or
>>> Latin, German, etc.) to Japanese pronunciation?
>> I am fairly certain that sometime in the last 12 months I have read
>> a description of the rules for mapping English to Japanese phonology--
>> somewhere on the web, where else?--but even after thinking about
>> it for a day I can't remember the exact place.
>>
>
Well, not necessarily. A couple minor counter-examples that come readily
to (even my fading) mind:
"strike" --> 'sutoraiki' (something a labor union might call) vs.
'sutoraiku' (something a baseball umpire might call).
"Andrew" --> 'andoryuu' (so far as I have been able to determine, that
name isn't pronounced with a [...rju(w)] in any surviving dialect).
ranjit_...@yahoo.com
> Are there any rules or patterns to convert loanwards from English (or
> Latin, German, etc.) to Japanese pronunciation?
>
> For example:
> Spider-Man > supaidāman
> Mozilla Firefox > mojira faiāfokkusu
It's curious that /z/ is realized as an affricate before [i] when /s/
is realized as a fricative before [i].
> centaur > kentaurosu (from Ancient Greek here)
> etc.
>
> I'd love to have a program that converts to these words, or a
> dictionary of such words...
Has no lexicographer attempted to compile all the katakana words that
have been used in Japanese?
Joachim Pense
> Joachim Pense wrote:
>> Christian Weisgerber (in sci.lang):
>>
>>> Sonja Elen Kisa <son...@gmail.com> wrote:
>>>
>>>> Are there any rules or patterns to convert loanwards from English (or
>>>> Latin, German, etc.) to Japanese pronunciation?
>>> I am fairly certain that sometime in the last 12 months I have read
>>> a description of the rules for mapping English to Japanese phonology--
>>> somewhere on the web, where else?--but even after thinking about
>>> it for a day I can't remember the exact place.
>>>
>>
>> The transformation into Japanese loandwords is based on the sound, [...]
>
> Well, not necessarily. A couple minor counter-examples that come readily
> to (even my fading) mind:
>
> "strike" --> 'sutoraiki' (something a labor union might call) vs.
> 'sutoraiku' (something a baseball umpire might call).
>
So the mapping is not unique. A null vowel is mapped to either i or u
(because these vovels are often mute in Japanese), and o after t and d
(because the consonant would be affected otherwise). So "strike" can be
either "sutoraiki" or "sutoraiku". Both versions exist with different
particular meanings of "strike" (a common phenomenon with loanwords). But
not a counterexample against "based on the sound".
> "Andrew" --> 'andoryuu' (so far as I have been able to determine, that
> name isn't pronounced with a [...rju(w)] in any surviving dialect).
Was it palatalized in an earlier form of English?
Joachim
anal...@hotmail.com
What inquiring minds want to know is
is "Godzilla" an English word invented by the Japanese, which was then
subjected to loan-word phonology to derive "Gojira".
ranjit_...@yahoo.com
> Bart Mathias (in sci.lang):
> > Joachim Pense wrote:
> >> Christian Weisgerber (in sci.lang):
>
>
> >>>> Are there any rules or patterns to convert loanwards from English (or
> >>>> Latin, German, etc.) to Japanese pronunciation?
> >>> I am fairly certain that sometime in the last 12 months I have read
> >>> a description of the rules for mapping English to Japanese phonology--
> >>> somewhere on the web, where else?--but even after thinking about
> >>> it for a day I can't remember the exact place.
>
> >> The transformation into Japanese loandwords is based on the sound, [...]
>
> > Well, not necessarily. A couple minor counter-examples that come readily
> > to (even my fading) mind:
>
> > "strike" --> 'sutoraiki' (something a labor union might call) vs.
> > 'sutoraiku' (something a baseball umpire might call).
>
> So the mapping is not unique. A null vowel is mapped to either i or u
> (because these vovels are often mute in Japanese),
In which contexts are they mute?
Joachim Pense
> On Jul 4, 4:50 pm, Joachim Pense <s...@pense-mainz.eu> wrote:
>> Bart Mathias (in sci.lang):
>> > Joachim Pense wrote:
>> >> Christian Weisgerber (in sci.lang):
>>
>> >>> Sonja Elen Kisa <sonj...@gmail.com> wrote:
>>
>> >>>> Are there any rules or patterns to convert loanwards from English
>> >>>> (or Latin, German, etc.) to Japanese pronunciation?
>> >>> I am fairly certain that sometime in the last 12 months I have read
>> >>> a description of the rules for mapping English to Japanese
>> >>> phonology-- somewhere on the web, where else?--but even after
>> >>> thinking about it for a day I can't remember the exact place.
>>
>> >> The transformation into Japanese loandwords is based on the sound,
>> >> [...]
>>
>> > Well, not necessarily. A couple minor counter-examples that come
>> > readily to (even my fading) mind:
>>
>> > "strike" --> 'sutoraiki' (something a labor union might call) vs.
>> > 'sutoraiku' (something a baseball umpire might call).
>>
>> So the mapping is not unique. A null vowel is mapped to either i or u
>> (because these vovels are often mute in Japanese),
>
> In which contexts are they mute?
>
Typically between voiceless plosives. They are often pronounced
as "voiceless vowels" in such contexts. This properly of becoming mute is
generalized to other situations (like voiced plosives) when importing
words.
Joachim
Joachim Pense
AFAIK the original film was "Gojira", and it was transformed into "Godzilla"
by Americans.
Joachim
Christian Weisgerber
> > "Andrew" --> 'andoryuu' (so far as I have been able to determine, that
> > name isn't pronounced with a [...rju(w)] in any surviving dialect).
>
> Was it palatalized in an earlier form of English?
If you mean, was there a /j/, then yes, I'm under the impression
there was.
Wikipedia suggests that Welsh English mostly does not have yod-dropping
even after /r/, but I don't know if it extends that far.
Sonja Elen Kisa
wrote:
Well, I could input the English word in X-SAMPA or similar.
benl...@ihug.co.nz
In the "pocket" size Kenkyusha which I normally use, the "katakana
words" are written in katakana, while the other words are in romaji.
All the o.p. would need to do is learn katakana (not hard), and there
would be thousands of examples, from which the rules could be worked
out "as an excercise for the student". The rules are basically quite
simple, though not without exceptions. And you have to remember that
English is not necessarily the only source language involved. E.g. /
biiru/ 'beer' does not fit with the generally non-rhotic English
source, so was more likely borrowed from German.
Ross Clark
LEE Sau Dan
>> I'd love to have a program that converts to these words, or a
>> dictionary of such words...
ranjit> Has no lexicographer attempted to compile all the katakana
ranjit> words that have been used in Japanese?
You'll need a Japanse->English dictionary. That's a large database,
even if you only extract the Katakana -> English subset.
What the OP wants, maybe, a small set of rules -- realized in the form
of a program -- that can accurately predict the mapping.
Joachim Pense
>>>>>> "ranjit" == ranjit mathews@yahoo com <ranjit_...@yahoo.com>
>>>>>> writes:
>
> >> I'd love to have a program that converts to these words, or a
> >> dictionary of such words...
>
> ranjit> Has no lexicographer attempted to compile all the katakana
> ranjit> words that have been used in Japanese?
>
> You'll need a Japanse->English dictionary. That's a large database,
> even if you only extract the Katakana -> English subset.
>
> What the OP wants, maybe, a small set of rules -- realized in the form
> of a program -- that can accurately predict the mapping.
>
>
I suppose the rules from English (or any other) pronounciation to Katakana
spelling are not difficult to formulate. But I expect writing the English
in SAMPA is more work than doing the Kanification manually.
Joachim
Bart Mathias
> ranjit_...@yahoo.com (in sci.lang):
>
>> On Jul 4, 4:50 pm, Joachim Pense <s...@pense-mainz.eu> wrote:
>>> So the mapping is not unique. A null vowel is mapped to either i or u
>>> (because these vovels are often mute in Japanese),
>> In which contexts are they mute?
>>
>
> Typically between voiceless plosives. They are often pronounced
> as "voiceless vowels" in such contexts. This properly of becoming mute is
> generalized to other situations (like voiced plosives) when importing
> words.
And finally, after a voiceless plosive, in a word that is accented but
not on the final mora.
After a voiceless fricative before a voiceless anything, or end of
accented-but-not-finally word, they tend to vanish entirely (with
lengthening of the fricative if one is speaking slowly enough).
But I'm not familiar with the concept of "null vowels."
Bart Mathias
Bart Mathias
> On Jul 2, 6:42 pm, Sonja Elen Kisa <sonj...@gmail.com> wrote:
>> Are there any rules or patterns to convert loanwards from English (or
>> Latin, German, etc.) to Japanese pronunciation?
>>
>> For example:
>> Spider-Man > supaidāman
>> Mozilla Firefox > mojira faiāfokkusu
>
> It's curious that /z/ is realized as an affricate before [i] when /s/
> is realized as a fricative before [i].
/d/ and /z/ merged before the high vowels long ago, so /zi/ = /di/ =
[(d)Zi]. (The [d] part can be very weak intervocalically.)
The distinction is largely reborn in modern borrowings. Many people can
call DDT [di:di:ti:] instead of [de:de:te:] (erstwhile /ti/ had of
course gone to [tSi]). There is some sort of food thing called "Cook-Do"
(?) [kuk:udu] (those [u]s are a lazy approximation of the real unrounded
backish voiced vowel).
Bart Mathias
Bart Mathias
Sort of depends on what you mean by "Gojira." It was ゴジラ for sure,
but I find no evidence that it was romanized that way.
I had always assumed that the "Godzilla" spelling was one of those
clever Japanese inventions (cf. the car names "Mazda," "Miata," etc.)
but I find no evidence for that belief either.
Bart Mathias
Oliver Cromm
>> Has no lexicographer attempted to compile all the katakana words that
>> have been used in Japanese?
All? Hardly possible. This is a very active area of the language, and
there are yearbooks of "new expressions" that are mostly katakana-words.
> And you have to remember that
> English is not necessarily the only source language involved. E.g. /
> biiru/ 'beer' does not fit with the generally non-rhotic English
> source, so was more likely borrowed from German.
Some compounds of "beer" have "biya": biyah�ru, biya-daru. This is
unpredictable from either English or German.
While it is mostly predictable how a word would be transcribed if
created now, some older loanwords don't fit older rules. Some
transcriptions are based on British, some on American, some on erroneous
pronunciations.
--
There are two ways of constructing a software design. One way is to make
it so simple that there are obviously no deficiencies. And the other way
is to make it so complicated that there are no obvious deficiencies.
The first method is far more difficult. -- C.�A.�R. Hoare
benl...@ihug.co.nz
> * "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> >> Has no lexicographer attempted to compile all the katakana words that
> >> have been used in Japanese?
>
> All? Hardly possible. This is a very active area of the language, and
> there are yearbooks of "new expressions" that are mostly katakana-words.
>
> > And you have to remember that
> > English is not necessarily the only source language involved. E.g. /
> > biiru/ 'beer' does not fit with the generally non-rhotic English
> > source, so was more likely borrowed from German.
> Some compounds of "beer" have "biya": biyahôru, biya-daru. This is
> unpredictable from either English or German.
Why unpredictable? /bia/ or /pia/ is exactly the form you find in most
Pacific island languages, where it comes from English. English
loanwords in Japanese are generally arhotic, e.g. baa, kaado, so it's
the appearance of /r/ in /biiru/ that's anomalous. /biya/ seems to
follow the normal pattern.
Ross Clark
Joachim Pense
>
> In the "pocket" size Kenkyusha which I normally use, the "katakana
> words" are written in katakana, while the other words are in romaji.
> All the o.p. would need to do is learn katakana (not hard), and there
> would be thousands of examples, from which the rules could be worked
> out "as an excercise for the student". The rules are basically quite
> simple, though not without exceptions. And you have to remember that
> English is not necessarily the only source language involved. E.g. /
> biiru/ 'beer' does not fit with the generally non-rhotic English
> source, so was more likely borrowed from German.
>
But German is non-rhotic, too. if biiru was token from German, it would be
biaa, the same you would expect from non-rhotic English.
Joachim
Ruud Harmsen
in sci.lang:
>But German is non-rhotic, too.
Many variants are. Not all of them.
benl...@ihug.co.nz
> benli...@ihug.co.nz (in sci.lang):
The only other German loanwords I can think of offhand are doitsu
(German) and arubaito (part-time job). The latter tends to confirm
that Japanese were hearing German rhotically in the period when the
borrowing was taking place.
Ross Clark
Joachim Pense
> Sat, 11 Jul 2009 11:41:50 +0200: Joachim Pense <sn...@pense-mainz.eu>:
> in sci.lang:
>
>>But German is non-rhotic, too.
>
> Many variants are. Not all of them.
>
Standard German is. AFAIK Swiss German dialects pronounce the r in Bier, but
I don't know of any more that do.
Joachim
Joachim Pense
In most German dialects, and also in standard pronounciation, the r in
Arbeit is not pronounced, or only weakly pronounced. The expected Japanese
version should be aabaito, in my opinon. I don't know why the r becomes ru
here. I don't know how these German loanwords came into the Japanese
language historically; maybe some hypercorrect spelling pronounciation was
involved?
Joachim
Peter T. Daniels
Can you be certain that the words came from German and not Dutch? The
Dutch had the monopoly on Western contact with Japan for a couple of
centuries.
Ruud Harmsen
in sci.lang:
>Ruud Harmsen (in sci.lang):
>
>> Sat, 11 Jul 2009 11:41:50 +0200: Joachim Pense <sn...@pense-mainz.eu>:
>> in sci.lang:
>>
>>>But German is non-rhotic, too.
>>
>> Many variants are. Not all of them.
>>
>
>Standard German is. AFAIK Swiss German dialects pronounce the r in Bier, but
>I don't know of any more that do.
Austrians? Ex-Prussians? Niedersachser?
BTW, in Standard German it is also pronounced, but as an open shwa.
Bier doesn't sound as Bie.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>> > The only other German loanwords I can think of offhand are doitsu
>> > (German) and arubaito (part-time job). The latter tends to confirm
>> > that Japanese were hearing German rhotically in the period when the
>> > borrowing was taking place.
>>
>> In most German dialects, and also in standard pronounciation, the r in
>> Arbeit is not pronounced, or only weakly pronounced. The expected Japanese
>> version should be aabaito, in my opinon. I don't know why the r becomes ru
>> here. I don't know how these German loanwords came into the Japanese
>> language historically; maybe some hypercorrect spelling pronounciation was
>> involved?
>
>Can you be certain that the words came from German and not Dutch? The
>Dutch had the monopoly on Western contact with Japan for a couple of
>centuries.
True. Arbeid is the Dutch word and only the last diphthong is slightly
different than in German (although details vary in both). Rhoticity
varies in both languages and also historically.
ranjit_...@yahoo.com
wrote:
> >>>>> "ranjit" ==ranjitmathews@yahoo com <ranjit_math...@yahoo.com> writes:
>
> >> I'd love to have a program that converts to these words, or a
> >> dictionary of such words...
>
> ranjit> Has no lexicographer attempted to compile all the katakana
> ranjit> words that have been used in Japanese?
>
> You'll need a Japanse->English dictionary. That's a large database,
> even if you only extract the Katakana -> English subset.
Do you know whether there is a dictionary I can download to extract
the Katakana->English subset?
> What the OP wants, maybe, a small set of rules -- realized in the form
> of a program -- that can accurately predict the mapping.
If the Japanese prefer non-rhotic pronunciation, the mapping might be
from OED pronunciation to Katakana spelling.
Joachim Pense
>
> BTW, in Standard German it is also pronounced, but as an open shwa.
> Bier doesn't sound as Bie.
I think Bia or Biä comes closest.
Joachim
LEE Sau Dan
Joachim> But German is non-rhotic, too. if biiru was token from
Joachim> German, it would be biaa, the same you would expect from
Joachim> non-rhotic English.
"biaa"? Why would you suggest a long "a"?
LEE Sau Dan
benlizro@ihug> The only other German loanwords I can think of
benlizro@ihug> offhand are doitsu (German) and arubaito (part-time
benlizro@ihug> job). The latter tends to confirm that Japanese were
benlizro@ihug> hearing German rhotically in the period when the
benlizro@ihug> borrowing was taking place.
What's the pitch-stress pattern for <arubaito>? Does it mimic the
stress pattern of "Arbeit"?
Ruud Harmsen
<dan...@informatik.uni-freiburg.de>: in sci.lang:
>>>>>> "benlizro@ihug" == benlizro@ihug co nz <benl...@ihug.co.nz> writes:
>
> benlizro@ihug> The only other German loanwords I can think of
> benlizro@ihug> offhand are doitsu (German) and arubaito (part-time
> benlizro@ihug> job). The latter tends to confirm that Japanese were
> benlizro@ihug> hearing German rhotically in the period when the
> benlizro@ihug> borrowing was taking place.
>
>What's the pitch-stress pattern for <arubaito>? Does it mimic the
>stress pattern of "Arbeit"?
--
Ruud Harmsen, http://rudhar.com
Joachim Pense
>>>>>> "Joachim" == Joachim Pense <sn...@pense-mainz.eu> writes:
>
> Joachim> But German is non-rhotic, too. if biiru was token from
> Joachim> German, it would be biaa, the same you would expect from
> Joachim> non-rhotic English.
>
> "biaa"? Why would you suggest a long "a"?
>
Because I recall that the English -er ending is often transformed to -aa in
Japanese.
Joachim
ranjit_...@yahoo.com
> * "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> >> Has no lexicographer attempted to compile all the katakana words that
> >> have been used in Japanese?
>
> All? Hardly possible. This is a very active area of the language, and
> there are yearbooks of "new expressions" that are mostly katakana-words.
Are these new expressions coined on the fly without the involvement of
a language institute?
> > And you have to remember that
> > English is not necessarily the only source language involved. E.g. /
> > biiru/ 'beer' does not fit with the generally non-rhotic English
> > source, so was more likely borrowed from German.
>
> Some compounds of "beer" have "biya": biyahôru, biya-daru. This is
Oliver Cromm
> On Jul 11, 10:32�am, Oliver Cromm <lispamat...@yahoo.de> wrote:
>> * "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
>>>> Has no lexicographer attempted to compile all the katakana words that
>>>> have been used in Japanese?
>>
>> All? Hardly possible. This is a very active area of the language, and
>> there are yearbooks of "new expressions" that are mostly katakana-words.
>>
>>> And you have to remember that
>>> English is not necessarily the only source language involved. E.g. /
>>> biiru/ 'beer' does not fit with the generally non-rhotic English
>>> source, so was more likely borrowed from German.
>> /
>> Some compounds of "beer" have "biya": biyah�ru, biya-daru. This is
>> unpredictable from either English or German.
>
> Why unpredictable? /bia/ or /pia/ is exactly the form you find in most
> Pacific island languages, where it comes from English. English
> loanwords in Japanese are generally arhotic, e.g. baa, kaado, so it's
> the appearance of /r/ in /biiru/ that's anomalous. /biya/ seems to
> follow the normal pattern.
I didn't mean unpredictable in theory, I meant that it is not following
the conventional patterns of English-to-Katakana transcription, where I
would expect "ia" instead of "iya", e.g. "kyaria" (carrier, career).
However, both variants seem to coexist, e.g. "ria" (rear of a car), but
also "riya ca-" (bicycle trailer, a Japanese-English coinage), or both
"iyaringu" and, occasionally, "iaringu" for earring.
So in yet another slightly different sense, it's unpredictable.
--
er (Tristan) hiez im ein t�renkleit
an der stete machen:
von wunderl�chen sachen
einen roc selts�n getan
und eine gugelen daran
H. V. FREIBERG Tristan
Oliver Cromm
> On Jul 10, 6:32�pm, Oliver Cromm <lispamat...@yahoo.de> wrote:
>> * "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
>>>> Has no lexicographer attempted to compile all the katakana words that
>>>> have been used in Japanese?
>>
>> All? Hardly possible. This is a very active area of the language, and
>> there are yearbooks of "new expressions" that are mostly katakana-words.
>
> Are these new expressions coined on the fly without the involvement of
> a language institute?
AFAIK there is no language institute of significant influence in Japan.
A word can have "quasi-official" status, just like in English, by
appearing in reputable dictionaries, of which there are several
competing ones from private publishers.
Thousands of Katakana expressions are coined each year. Most of them
never gain popularity. but if they do, they will make it into
dictionaries of neologisms, and if they stay in use long enough, into
mainstream dictionaries.
--
gugelgesang, m., m�nchsgesang, pejorativ
GRIMM, Deutsches W�rterbuch
LEE Sau Dan
Oliver> I didn't mean unpredictable in theory, I meant that it is
Oliver> not following the conventional patterns of
Oliver> English-to-Katakana transcription, where I would expect "ia"
Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
Why "ky-"?
Oliver Cromm
>>>>>> "Oliver" == Oliver Cromm <lispa...@yahoo.de> writes:
>
> Oliver> I didn't mean unpredictable in theory, I meant that it is
> Oliver> not following the conventional patterns of
> Oliver> English-to-Katakana transcription, where I would expect "ia"
> Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>
> Why "ky-"?
With the following a, the usual representation of the vowel /�/ (SAMPA
/{/ Kirshenbaum /&/).
--
die pfaffen mit ihren platten,
die m�nch mit ihren gugeln schon
bergreihen 26 ndr.
benl...@ihug.co.nz
> * "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> > On Jul 11, 10:32 am, Oliver Cromm <lispamat...@yahoo.de> wrote:
> >> * "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> >>>> Has no lexicographer attempted to compile all the katakana words that
> >>>> have been used in Japanese?
>
> >> All? Hardly possible. This is a very active area of the language, and
> >> there are yearbooks of "new expressions" that are mostly katakana-words.
>
> >>> And you have to remember that
> >>> English is not necessarily the only source language involved. E.g. /
> >>> biiru/ 'beer' does not fit with the generally non-rhotic English
> >>> source, so was more likely borrowed from German.
> >> /
> >> unpredictable from either English or German.
>
> > Why unpredictable? /bia/ or /pia/ is exactly the form you find in most
> > Pacific island languages, where it comes from English. English
> > loanwords in Japanese are generally arhotic, e.g. baa, kaado, so it's
> > the appearance of /r/ in /biiru/ that's anomalous. /biya/ seems to
> > follow the normal pattern.
>
> I didn't mean unpredictable in theory, I meant that it is not following
> the conventional patterns of English-to-Katakana transcription, where I
> would expect "ia" instead of "iya", e.g. "kyaria" (carrier, career).
> However, both variants seem to coexist, e.g. "ria" (rear of a car), but
> also "riya ca-" (bicycle trailer, a Japanese-English coinage), or both
> "iyaringu" and, occasionally, "iaringu" for earring.
>
> So in yet another slightly different sense, it's unpredictable.
Do you know if /-ia/ vs /-iya/ is a stable contrast in spoken
Japanese? Or is it just an available orthographic difference which is
more or less haphazardly chosen?
Ross Clark
ranjit_...@yahoo.com
> * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>
>
> > Oliver> I didn't mean unpredictable in theory, I meant that it is
> > Oliver> not following the conventional patterns of
> > Oliver> English-to-Katakana transcription, where I would expect "ia"
> > Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>
> > Why "ky-"?
>
> /{/ Kirshenbaum /&/).
Some Americans have a [j] on-glide before [&].
Oliver Cromm
RP "beer" doesn't sound like "bee", either. "Non-rhotic" doesn't mean Rs
are simply dropped.
--
Give a man a fish and you feed him for a day; teach him to use the 'Net
and he won't bother you for weeks.
Oliver Cromm
>>>>>> "benlizro@ihug" == benlizro@ihug co nz <benl...@ihug.co.nz> writes:
>
> benlizro@ihug> The only other German loanwords I can think of
> benlizro@ihug> offhand are doitsu (German) and arubaito (part-time
> benlizro@ihug> job). The latter tends to confirm that Japanese were
> benlizro@ihug> hearing German rhotically in the period when the
> benlizro@ihug> borrowing was taking place.
>
> What's the pitch-stress pattern for <arubaito>? Does it mimic the
> stress pattern of "Arbeit"?
No, there is an accent (voice-up) on the "ba" (IIRC). However, I don't
find that Japanese accent in loanwords generally tries to mimic the
accent of the source language.
--
Q: What do computer engineers use for birth control?
A: Their personalities.
Oliver Cromm
Used to, but that is currently changing. The young generation writes
"purinta" and not "purinta-" any more. I prefer "oriba" to the
traditional "oriba-", but the writing of this name is quite ingrained.
--
WinErr 008: Erroneous error. Nothing is wrong.
Bart Mathias
> [...]
> Do you know if /-ia/ vs /-iya/ is a stable contrast in spoken
> Japanese? Or is it just an available orthographic difference which is
> more or less haphazardly chosen?
I can't think of any words that are differentiated that way (in native
Japanese, "...ia" would be impossible, outside of compounds), even in
writing. It's a distinction that I would certainly hate to have to learn.
I have a vague memory of the Monbusho stepping in and decreeing that the
word borrowed from "piano" (whatever language they got that from) would
henceforth be "piano" and not "piyano." My dictionary has only "piano,"
but Google suggests "piyano" is still used.
I was about to suggest that the ("accepted") "...iya" vs. "...ia"
distinction largely comes down to: if it came from English [IR] it's
"...iya," ("biya," "iyaringu," "hiyaringu," etc.) and [i(j)R] results in
"...ia," but after all, "kyariya" (when it's "career") and "suki-ya-"
("-" for lengthening) = "one who plies the snowy slopes" mess up that
hypothesis. In my Shinmeikai, "gia" is defined as "giya."
Bart Mathias
Bart Mathias
> [...]
>
> The only other German loanwords I can think of offhand are doitsu
> (German) and arubaito (part-time job). The latter tends to confirm
> that Japanese were hearing German rhotically in the period when the
> borrowing was taking place.
Those words are from German, according to my Shinmeikai (to respond to a
doubt further down the thread).
There is a lot of medical vocabulary from German. "arerugi-" (allergy),
"rentogen" (X-ray), etc.
But the Japanese did first learn Western medicine from the Dutch. "mesu"
(scalpel) is attributed to Dutch.
Bart Mathias
Bart Mathias
> [...]
>
> But German is non-rhotic, too. if biiru was token from German, it would be
> biaa, the same you would expect from non-rhotic English.
Terrifying news! I was taught to pronounce the "r" (in the back of my
mouth) in college (both of my teachers were natives), and my _Brockhaus
Illustrated_ G-E/E-G dictionary gives "bi:r."
Bart Mathias
Bart Mathias
> * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>
>>>>>>> "benlizro@ihug" == benlizro@ihug co nz <benl...@ihug.co.nz> writes:
>> benlizro@ihug> The only other German loanwords I can think of
>> benlizro@ihug> offhand are doitsu (German) and arubaito (part-time
>> benlizro@ihug> job). The latter tends to confirm that Japanese were
>> benlizro@ihug> hearing German rhotically in the period when the
>> benlizro@ihug> borrowing was taking place.
>>
>> What's the pitch-stress pattern for <arubaito>? Does it mimic the
>> stress pattern of "Arbeit"?
>
> No, there is an accent (voice-up) on the "ba" (IIRC). However, I don't
> find that Japanese accent in loanwords generally tries to mimic the
> accent of the source language.
Actually a rise in tone on a third mora wouldn't be possible (there's an
up or a down after the first, and only one down per word). There is a
drop after the (high) "ruba," however.
Foreign words are (most?) often given a sort of default accent, with a
drop in pitch after the third mora from the end. The kind of thing that
has made words like "choKOREeto" and "aRUMINIumu" hard for me to learn.
Bart Mathias
Oliver Cromm
> Oliver Cromm wrote:
>> * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>>
>>>>>>>> "benlizro@ihug" == benlizro@ihug co nz <benl...@ihug.co.nz> writes:
>>> benlizro@ihug> The only other German loanwords I can think of
>>> benlizro@ihug> offhand are doitsu (German) and arubaito (part-time
>>> benlizro@ihug> job). The latter tends to confirm that Japanese were
>>> benlizro@ihug> hearing German rhotically in the period when the
>>> benlizro@ihug> borrowing was taking place.
>>>
>>> What's the pitch-stress pattern for <arubaito>? Does it mimic the
>>> stress pattern of "Arbeit"?
>>
>> No, there is an accent (voice-up) on the "ba" (IIRC). However, I don't
>> find that Japanese accent in loanwords generally tries to mimic the
>> accent of the source language.
>
> Actually a rise in tone on a third mora wouldn't be possible (there's an
> up or a down after the first, and only one down per word). There is a
> drop after the (high) "ruba," however.
You're right, of course. The third syllable is the most prominent (to
me), with the pitch falling after it. This does not correspond to the
stress in German.
--
Software gets slower, faster than hardware gets faster.
--Wirth's law
benl...@ihug.co.nz
Generally, or just after velars? The treatment of /&/ as /ya/
specifically after /k/ and /g/ is also found in some Atlantic creoles,
e.g. gal > gyal, can't > kyan.
Ross Clark
benl...@ihug.co.nz
Does Shinmeikai (or any other source) have dates or citations for
these early European loans?
Ross Clark
LEE Sau Dan
>> BTW, in Standard German it is also pronounced, but as an open
>> shwa. Bier doesn't sound as Bie.
Oliver> RP "beer" doesn't sound like "bee", either. "Non-rhotic"
Oliver> doesn't mean Rs are simply dropped.
The "r" here is not dropped, but vocalized as schwa [@].
LEE Sau Dan
>> No, there is an accent (voice-up) on the "ba" (IIRC). However, I
>> don't find that Japanese accent in loanwords generally tries to
>> mimic the accent of the source language.
Bart> Actually a rise in tone on a third mora wouldn't be possible
Right. In Tokyo Japanese, the first and second mora must be different
in pitch. Only 2 different pitches are identified. The other rule is
that in the same word, there is at most 1 drop in pitch. So, a rise in
pitch on the 3rd mora is impossible.
Bart> (there's an up or a down after the first, and only one down
Bart> per word). There is a drop after the (high) "ruba," however.
There is a catch: The drop could occur AFTER the LAST mora. So, it
won't be audible, unless you stick a suffixing particle to it. The
particle will then be pronounced with a low pitch, showing that the drop
occurs after the last mora of the preceding stem. (Sounds like
"liaison"?)
Joachim Pense
>>>>>> "Oliver" == Oliver Cromm <lispa...@yahoo.de> writes:
>
> >> BTW, in Standard German it is also pronounced, but as an open
> >> shwa. Bier doesn't sound as Bie.
>
> Oliver> RP "beer" doesn't sound like "bee", either. "Non-rhotic"
> Oliver> doesn't mean Rs are simply dropped.
>
> The "r" here is not dropped, but vocalized as schwa [@].
>
>
Same as in German. Depending on the region, the schwa may come out as an a
or an ä.
Joachim
John Atkinson
> On Jul 14, 2:55 am, "ranjit_math...@yahoo.com"
> <ranjit_math...@yahoo.com> wrote:
>> On Jul 13, 2:15 am, Oliver Cromm <lispamat...@yahoo.de> wrote:
>>
>>> * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>>
>>>>>>>>> "Oliver" == Oliver Cromm <lispamat...@yahoo.de> writes:
>>
>>>> Oliver> I didn't mean unpredictable in theory, I meant that it is
>>>> Oliver> not following the conventional patterns of
>>>> Oliver> English-to-Katakana transcription, where I would expect
>>>> "ia" Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>>
>>>> Why "ky-"?
>>
>>> (SAMPA /{/ Kirshenbaum /&/).
>>
>> Some Americans have a [j] on-glide before [&].
>
> Generally, or just after velars? The treatment of /& / as /ya /
True. But better thought of as /& / becoming /a / everywhere, and then /k /
and /g / becoming /ky / and /gy / before /a /. See below.
> e.g. gal > gyal, can't > kyan.
More accurately, [gya:l]. [kya:n] (long "a", I think).
I don't think these two are examples of / &/ in the lexifying language of
these creoles, which would surely have been some some variety from southern
England, which I assume would have /g3:l / and /ka:nt /.
As I understand it, Jamaican (and most other Altantic creoles) has [a] for
our [&] in _all_ circumstances, and Jamaican (but not most others, ) inserts
y after k and g whenever the vowel in standard English was either /& / or
/a(:) / -- in Jamaican, <cat> = [kyat], <gas> = [gyas], <cart> = [kya:t],
<garden> = [gya:dn].
(OTOH, y isn't inserted before [a] when it derives from standard English /O
/ or /A. / -- <corn> = [ka:n], <caught> = [ka:t], <cot> = [kat] -- which are
minimal pairs with <can't>, <cart>, and <cat> resp.)
John.
John.
Bart Mathias
>>>>>> "Bart" == Bart Mathias <mat...@hawaii.edu> writes:
>
> >> No, there is an accent (voice-up) on the "ba" (IIRC). However, I
> >> don't find that Japanese accent in loanwords generally tries to
> >> mimic the accent of the source language.
>
> Bart> Actually a rise in tone on a third mora wouldn't be possible
>
> Right. In Tokyo Japanese, the first and second mora must be different
> in pitch. Only 2 different pitches are identified. The other rule is
> that in the same word, there is at most 1 drop in pitch. So, a rise in
> pitch on the 3rd mora is impossible.
>
>
> Bart> (there's an up or a down after the first, and only one down
> Bart> per word). There is a drop after the (high) "ruba," however.
>
> There is a catch: The drop could occur AFTER the LAST mora.
A wee bit misstated, that. It is true that the "accented" mora of a word
may be the last mora, but there won't be a inaudible drop in pitch after
it, ever.
> So, it
> won't be audible, unless you stick a suffixing particle to it. The
> particle will then be pronounced with a low pitch, showing that the drop
> occurs after the last mora of the preceding stem. (Sounds like
> "liaison"?)
One of the reasons (I may have left it out of a discussion of the matter
in an earlier thread?) that I consider particles, like jodoshi, to be
suffixes.
Bart Mathias
>> There is a lot of medical vocabulary from German. "arerugi-" (allergy),
>> "rentogen" (X-ray), etc.
>>
>> But the Japanese did first learn Western medicine from the Dutch. "mesu"
>> (scalpel) is attributed to Dutch.
>>
>> Bart Mathias
>
> Does Shinmeikai (or any other source) have dates or citations for
> these early European loans?
I could squeeze Shinmeikai into my front pocket. There are some usage
examples, but I don't recall ever seeing a citation, and dates would be
restricted to words where the date is part of the definition, if it
includes any such words (I just tried "Meiji-ishin," but neither it nor
"Meiji" are listed).
I would expect to find both dates and citations in Shogak(u)kan's
(should be a macron on that "o" but I don't know how to do it)
_Kokugodaijiten_'s 24 volumes, but my access to that is 5K away, and I
can't get my bike out of the closet anymore.
Bart Mathias
benl...@ihug.co.nz
> benli...@ihug.co.nz wrote:
> > On Jul 14, 2:55 am, "ranjit_math...@yahoo.com"
> > <ranjit_math...@yahoo.com> wrote:
> >> On Jul 13, 2:15 am, Oliver Cromm <lispamat...@yahoo.de> wrote:
>
> >>> * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>
> >>>>>>>>> "Oliver" == Oliver Cromm <lispamat...@yahoo.de> writes:
>
> >>>> Oliver> I didn't mean unpredictable in theory, I meant that it is
> >>>> Oliver> not following the conventional patterns of
> >>>> Oliver> English-to-Katakana transcription, where I would expect
> >>>> "ia" Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>
> >>>> Why "ky-"?
>
> >>> (SAMPA /{/ Kirshenbaum /&/).
>
> >> Some Americans have a [j] on-glide before [&].
>
> > Generally, or just after velars? The treatment of /& / as /ya /
> > specifically after /k / and /g / is also found in some Atlantic creoles,
>
> True. But better thought of as /& / becoming /a / everywhere, and then /k /
> and /g / becoming /ky / and /gy / before /a /. See below.
>
> > e.g. gal > gyal, can't > kyan.
>
> More accurately, [gya:l]. [kya:n] (long "a", I think).
>
> I don't think these two are examples of / &/ in the lexifying language of
> these creoles, which would surely have been some some variety from southern
> England, which I assume would have /g3:l / and /ka:nt /.
>
> As I understand it, Jamaican (and most other Altantic creoles) has [a] for
> our [&] in _all_ circumstances, and Jamaican (but not most others, ) inserts
> y after k and g whenever the vowel in standard English was either /& / or
> /a(:) / -- in Jamaican, <cat> = [kyat], <gas> = [gyas], <cart> = [kya:t],
> <garden> = [gya:dn].
>
> (OTOH, y isn't inserted before [a] when it derives from standard English /O
> / or /A. / -- <corn> = [ka:n], <caught> = [ka:t], <cot> = [kat] -- which are
> minimal pairs with <can't>, <cart>, and <cat> resp.)
>
internal change, is it?
I assume gyal is from "gal" rather than "girl". And the vowel in the /
ar/ sequence is apparently treated the same way.
Why this should happen _just_ after velars remains to be explained.
I forgot to mention that there are traces of the same thing in
Melanesian Pidgin, e.g. gammon > giaman, and probably cap(tain) > Tok
Pisin kiap "district officer".
Ross Clark
John Atkinson
> On Jul 14, 11:45 pm, "John Atkinson" <johna...@bigpond.com> wrote:
>> benli...@ihug.co.nz wrote:
>>> On Jul 14, 2:55 am, "ranjit_math...@yahoo.com"
>>> <ranjit_math...@yahoo.com> wrote:
>>>> On Jul 13, 2:15 am, Oliver Cromm <lispamat...@yahoo.de> wrote:
>>
>>>>> * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>>
>>>>>>>>>>> "Oliver" == Oliver Cromm <lispamat...@yahoo.de> writes:
>>
>>>>>> Oliver> I didn't mean unpredictable in theory, I meant that it is
>>>>>> Oliver> not following the conventional patterns of
>>>>>> Oliver> English-to-Katakana transcription, where I would expect
>>>>>> "ia" Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>>
>>>>>> Why "ky-"?
>>
>>>>> (SAMPA /{/ Kirshenbaum /&/).
>>
>>>> Some Americans have a [j] on-glide before [&].
>>
>>> Generally, or just after velars? The treatment of /& / as /ya /
>>> specifically after /k / and /g / is also found in some Atlantic
>>> creoles,
>>
>> True. But better thought of as /& / becoming /a / everywhere, and
>> then /k / and /g / becoming /ky / and /gy / before /a /. See below.
>>
>>> e.g. gal > gyal, can't > kyan.
>>
>> More accurately, [gya:l]. [kya:n] (long "a", I think).
>>
>> I don't think these two are examples of / &/ in the lexifying
>> language of these creoles, which would surely have been some some
>> variety from southern England, which I assume would have /g3:l / and
>> /ka:nt /.
>>
>> As I understand it, Jamaican (and most other Altantic creoles) has
>> [a] for our [&] in _all_ circumstances, and Jamaican (but not most
>> others, ) inserts y after k and g whenever the vowel in standard
>> English was either /& / or /a(:) / -- in Jamaican, <cat> = [kyat],
>> <gas> = [gyas], <cart> = [kya:t], <garden> = [gya:dn].
>>
>> (OTOH, y isn't inserted before [a] when it derives from standard
>> English /O / or /A. / -- <corn> = [ka:n], <caught> = [ka:t], <cot> =
>> [kat] -- which are minimal pairs with <can't>, <cart>, and <cat>
>> resp.)
>>
> But if it's dependent on the lexifier vowel, it's not a creole-
> internal change, is it?
I don't know. Not sure what "creole-internal" means here. There's the
complication that A. and O: merge with a and a: only in meso- and basilectal
Jamaican, remaining unchanged in the acrolect. And all these registers
interact with each other, synchronically.
> I assume gyal is from "gal" rather than "girl".
Yes, that sounds right. The NURSE vowel merges with the STRUT vowel in
Jamaican, so "girl" shouldn't give "gyal" or "gya:l"
> And the vowel in the /
> ar/ sequence is apparently treated the same way.
> Why this should happen _just_ after velars remains to be explained.
Why? Palatisation of velars is common as dirt. Old French also had it
before [a] (cattu- > chat). The thing that perhaps remains to be explained
is why it occurs _just_ before these particular vowels in Jamaican, and not
before other front vowels, in words like 'kit" and "get".
FWIW, glide-insertion also occurs in Jamaican between labial stops and [ai]
(<point> = [pwaint], <boy> = [bwai]), but only if the vowel corresponds to
Standard /Oi /, not when it corresponds to Standard /ai / (<pint> =
[paint], <buy> = [bai]).
> I forgot to mention that there are traces of the same thing in
> Melanesian Pidgin, e.g. gammon > giaman, and probably cap(tain) > Tok
> Pisin kiap "district officer".
Any idea how far back this goes? Is it the case in all varieties of
Melanesian pidgin? How about other Pacific and Australian pidgins? (AFAIK,
"gammon" doesn't have "gy-" in modern Aboriginal English.)
John.
Oliver Cromm
> On Jul 14, 2:55�am, "ranjit_math...@yahoo.com"
> <ranjit_math...@yahoo.com> wrote:
>> On Jul 13, 2:15�am, Oliver Cromm <lispamat...@yahoo.de> wrote:
>>
>>> * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>>
>>> >>>>>> "Oliver" == Oliver Cromm <lispamat...@yahoo.de> writes:
>>
>>> > � � Oliver> I didn't mean unpredictable in theory, I meant that it is
>>> > � � Oliver> not following the conventional patterns of
>>> > � � Oliver> English-to-Katakana transcription, where I would expect "ia"
>>> > � � Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>>
>>> > Why "ky-"?
>>
>>> With the following a, the usual representation of the vowel /�/ (SAMPA
>>> /{/ Kirshenbaum /&/).
>>
>> Some Americans have a [j] on-glide before [&].
>
> Generally, or just after velars? The treatment of /&/ as /ya/
> specifically after /k/ and /g/ is also found in some Atlantic creoles,
> e.g. gal > gyal, can't > kyan.
Now that you mention it, it occurs in Japanese after velars only. Never
thought about it before, even though I had the intuition on when to use
it. Given that it's paralleled in another language, there should be an
underlying phonetic reason.
--
If the aeroplane industry had advanced at the same rate as the computer
industry, today's planes could circumnavigate the world in ten seconds,
be two inches long, and crash twice a day.
Peter Moylan in alt.usage.english
Oliver Cromm
A good argument. However, the drop can also occur, e.g., with following
"da". Would you consider that a suffix, too, then?
--
binsaugnesseln, hummelsk�l, m�nchspfeffer, gugelk�pf, minbr�derm�ntz
FISCHART
GRIMM, Deutsches W�rterbuch
Sonja Elen Kisa
ranjit_...@yahoo.com
> On Jul 14, 2:55 am, "ranjit_math...@yahoo.com"
> <ranjit_math...@yahoo.com> wrote:
> > On Jul 13, 2:15 am, Oliver Cromm <lispamat...@yahoo.de> wrote:
>
> > > * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>
> > > >>>>>> "Oliver" == Oliver Cromm <lispamat...@yahoo.de> writes:
>
> > > > Oliver> I didn't mean unpredictable in theory, I meant that it is
> > > > Oliver> not following the conventional patterns of
> > > > Oliver> English-to-Katakana transcription, where I would expect "ia"
> > > > Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>
> > > > Why "ky-"?
>
> > > With the following a, the usual representation of the vowel /æ/ (SAMPA
> > > /{/ Kirshenbaum /&/).
>
> > Some Americans have a [j] on-glide before [&].
>
> Generally, or just after velars?
<dad> [dj&:d].
So, not just after velars.
benl...@ihug.co.nz
<ranjit_math...@yahoo.com> wrote:
> On Jul 13, 7:36 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
>
>
>
> > On Jul 14, 2:55 am, "ranjit_math...@yahoo.com"
> > <ranjit_math...@yahoo.com> wrote:
> > > On Jul 13, 2:15 am, Oliver Cromm <lispamat...@yahoo.de> wrote:
>
> > > > * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>
> > > > >>>>>> "Oliver" == Oliver Cromm <lispamat...@yahoo.de> writes:
>
> > > > > Oliver> I didn't mean unpredictable in theory, I meant that it is
> > > > > Oliver> not following the conventional patterns of
> > > > > Oliver> English-to-Katakana transcription, where I would expect "ia"
> > > > > Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>
> > > > > Why "ky-"?
>
> > > > With the following a, the usual representation of the vowel /æ/ (SAMPA
> > > > /{/ Kirshenbaum /&/).
>
> > > Some Americans have a [j] on-glide before [&].
>
> > Generally, or just after velars?
>
> <dad> [dj&:d].
> So, not just after velars.
What type of American is this?
benl...@ihug.co.nz
> How does a word like <ancient> get Japanicized?
Applying my own mental version of "the rules" I would guess /
einshento/.
Ross Clark
Oliver Cromm
> How does a word like <ancient> get Japanicized?
I think "ancient" starts with /eI/ (in most varieties of English).
Stressed /eI/ is normally transcribed as long e, so I would suggest
"e-nshento".
Eijiro suggests "einshento, enshento, enshiento".
<http://eow.alc.co.jp/ancient/UTF-8/>
Well, with e- and ei, we have another area of pretty free variation.
Whether they are even pronounced differently depends on the speaker, I
suppose.
--
The nice thing about standards is that you have so many to choose
from; furthermore, if you do not like any of them, you can just
wait for next year's model.
Andrew Tanenbaum, _Computer Networks_ (1981), p. 168.
Oliver Cromm
> * Sonja Elen Kisa <son...@gmail.com> wrote:
>
>> How does a word like <ancient> get Japanicized?
>
> Eijiro suggests "einshento, enshento, enshiento".
"enshiento" smells of spelling pronunciation to me, by the way.
I just see that "ancient" has a /tS/ in BE - never noticed that.
--
A computer will do what you tell it to do, but that may be much
different from what you had in mind. - Joseph Weizenbaum
benl...@ihug.co.nz
> benli...@ihug.co.nz wrote:
> > On Jul 14, 11:45 pm, "John Atkinson" <johna...@bigpond.com> wrote:
> >> benli...@ihug.co.nz wrote:
> >>> On Jul 14, 2:55 am, "ranjit_math...@yahoo.com"
> >>> <ranjit_math...@yahoo.com> wrote:
> >>>> On Jul 13, 2:15 am, Oliver Cromm <lispamat...@yahoo.de> wrote:
>
> >>>>> * LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
>
> >>>>>>>>>>> "Oliver" == Oliver Cromm <lispamat...@yahoo.de> writes:
>
> >>>>>> Oliver> I didn't mean unpredictable in theory, I meant that it is
> >>>>>> Oliver> not following the conventional patterns of
> >>>>>> Oliver> English-to-Katakana transcription, where I would expect
> >>>>>> "ia" Oliver> instead of "iya", e.g. "kyaria" (carrier, career).
>
> >>>>>> Why "ky-"?
>
Yah, and I don't know what the historical records suggest. I was
assuming an early merger of all these vowels to JC /a(:)/, after which
a differential glide formation would not be possible.
>
> > I assume gyal is from "gal" rather than "girl".
>
> Yes, that sounds right. The NURSE vowel merges with the STRUT vowel in
> Jamaican, so "girl" shouldn't give "gyal" or "gya:l"
>
> > And the vowel in the /
> > ar/ sequence is apparently treated the same way.
> > Why this should happen _just_ after velars remains to be explained.
>
> Why? Palatisation of velars is common as dirt. Old French also had it
> before [a] (cattu- > chat). The thing that perhaps remains to be explained
> is why it occurs _just_ before these particular vowels in Jamaican, and not
> before other front vowels, in words like 'kit" and "get".
So again, we can account for it at the creole-formation interface.
Palatalization of velars before /i/ and /e/ is too universal to be
perceived, whereas that before /&/ is less usual, and is represented
by glide insertion.
>
> FWIW, glide-insertion also occurs in Jamaican between labial stops and [ai]
> (<point> = [pwaint], <boy> = [bwai]), but only if the vowel corresponds to
> Standard /Oi /, not when it corresponds to Standard /ai / (<pint> =
> [paint], <buy> = [bai]).
>
> > I forgot to mention that there are traces of the same thing in
> > Melanesian Pidgin, e.g. gammon > giaman, and probably cap(tain) > Tok
> > Pisin kiap "district officer".
>
> Any idea how far back this goes? Is it the case in all varieties of
> Melanesian pidgin?
"gammon" has the glide in all MP varieties; in "captain" it only
occurs in this item with shifted meaning, but note TP /kepten/
"captain". I think some /ia/ or /ya/ sequences must have either fused
to /e/ or lost the glide, historically. (Have not checked all my early
documentary sources.) There is also what appears to be a symmetrical
process when a velar follows (original) /&/, as in "flag" > /flaeg/,
"back" > /baek/.
How about other Pacific and Australian pidgins? (AFAIK,
> "gammon" doesn't have "gy-" in modern Aboriginal English.)
In a hasty trawl through my old notes I don't find it outside the MP
group (including Torres Straits). Nor, interestingly, does it seem to
turn up in any of the borrowings into the Pacific vernaculars, where /
&/ generally > /a/, occasionally /e/.
Ross Clark
Bart Mathias
> * Bart Mathias <mat...@hawaii.edu> wrote:
>
>> LEE Sau Dan wrote:
>>> Bart> (there's an up or a down after the first, and only one down
>>> Bart> per word). There is a drop after the (high) "ruba," however.
>>>
>>> There is a catch: The drop could occur AFTER the LAST mora.
>> A wee bit misstated, that. It is true that the "accented" mora of a word
>> may be the last mora, but there won't be a inaudible drop in pitch after
>> it, ever.
>>
>>> So, it
>>> won't be audible, unless you stick a suffixing particle to it. The
>>> particle will then be pronounced with a low pitch, showing that the drop
>>> occurs after the last mora of the preceding stem. (Sounds like
>>> "liaison"?)
>> One of the reasons (I may have left it out of a discussion of the matter
>> in an earlier thread?) that I consider particles, like jodoshi, to be
>> suffixes.
>
> A good argument. However, the drop can also occur, e.g., with following
> "da". Would you consider that a suffix, too, then?
Of course! I mentioned jodoshi above.
This partictular jodoshi happens to derive from a particle -de plus the
front part of a verb, a(ru), but the indepence of a verb was lost with
the juncture.
I've always thought it interesting that some of the things Japanese
grammarians consider particles (if we equate that term to "joshi") are
automatically treated as suffixes in romanization, e.g. "mi-te"
('look-ing/look and'), "ike-ba" ('if ... go[es]').
Bart Mathias
> * Sonja Elen Kisa <son...@gmail.com> wrote:
>
>> How does a word like <ancient> get Japanicized?
>
> I think "ancient" starts with /eI/ (in most varieties of English).
> Stressed /eI/ is normally transcribed as long e, so I would suggest
> "e-nshento".
>
> Eijiro suggests "einshento, enshento, enshiento".
> <http://eow.alc.co.jp/ancient/UTF-8/>
>
> Well, with e- and ei, we have another area of pretty free variation.
> Whether they are even pronounced differently depends on the speaker, I
> suppose.
Although Samuel E. Martin might agree with you if I recall our
discussion correctly, I'm persuaded that there are some cases where [e:]
instead of [ei] is fairly wrong. These include "tamei\ki" ("\" here
marks drop in pitch) = 'sigh' and "e\i" = 'ray (fish).'
Bart Mathias
> * Oliver Cromm <lispa...@yahoo.de> wrote:
>
>> * Sonja Elen Kisa <son...@gmail.com> wrote:
>>
>>> How does a word like <ancient> get Japanicized?
>> Eijiro suggests "einshento, enshento, enshiento".
>
> "enshiento" smells of spelling pronunciation to me, by the way.
>
> I just see that "ancient" has a /tS/ in BE - never noticed that.
That would also be true for all of us Americans who don't/can't
distinguish "cents" and "sense."
ranjit_...@yahoo.com
My cousins. One born and raised in greater Detroit and one near
Topeka, Kansas.
Brian M. Scott
<mat...@hawaii.edu> wrote in
<news:h3m18q$t2i$2...@news.eternal-september.org> in sci.lang:
> Oliver Cromm wrote:
[...]
>> I just see that "ancient" has a /tS/ in BE - never noticed that.
> That would also be true for all of us Americans who don't/can't
> distinguish "cents" and "sense."
Some Americans have /'eIN(k)S@nt/; I suspect that some of
them are among those who don't distinguish 'cents' and
'sense'.
Brian
LEE Sau Dan
>> One of the reasons (I may have left it out of a discussion of the
>> matter in an earlier thread?) that I consider particles, like
>> jodoshi, to be suffixes.
Oliver> A good argument. However, the drop can also occur, e.g.,
Oliver> with following "da". Would you consider that a suffix, too,
Oliver> then?
Yes, I do, because of that drop.
LEE Sau Dan
Oliver> * Sonja Elen Kisa <son...@gmail.com> wrote:
>> How does a word like <ancient> get Japanicized?
Oliver> I think "ancient" starts with /eI/ (in most varieties of
Oliver> English). Stressed /eI/ is normally transcribed as long e,
Oliver> so I would suggest "e-nshento".
So, in Katakana: エインシェント 'e i n shye n to'.
I'd agree with this Katakanaization.
BTW, is there any reason why this "a" (in English) is a 'long a'? Any
other examples where "an-" is pronounced with a 'long a'?
Sonja Elen Kisa
> a description of the rules for mapping English to Japanese phonology--
> somewhere on the web, where else?--but even after thinking about
> it for a day I can't remember the exact place.
Let me know if you find it!
Sonja Elen Kisa
> <http://eow.alc.co.jp/ancient/UTF-8/>
Cool, then that's the only tool I need to look up the answers to my
puzzle!
PaulJK
>>>>>> "Oliver" == Oliver Cromm <lispa...@yahoo.de> writes:
>
> Oliver> * Sonja Elen Kisa <son...@gmail.com> wrote:
> >> How does a word like <ancient> get Japanicized?
>
> Oliver> I think "ancient" starts with /eI/ (in most varieties of
> Oliver> English). Stressed /eI/ is normally transcribed as long e,
> Oliver> so I would suggest "e-nshento".
>
> So, in Katakana: エインシェント 'e i n shye n to'.
> I'd agree with this Katakanaization.
>
> BTW, is there any reason why this "a" (in English) is a 'long a'? Any
> other examples where "an-" is pronounced with a 'long a'?
As much as keep I trying, I can't think of any polite words,
just these two: "anal" and "anus". :-)
Are those two pronounced that way to distinguish them from
"annal" and "annus"?
pjk
LEE Sau Dan
>> BTW, is there any reason why this "a" (in English) is a 'long a'?
>> Any other examples where "an-" is pronounced with a 'long a'?
PaulJK> As much as keep I trying, I can't think of any polite words,
PaulJK> just these two: "anal" and "anus". :-)
Since these 2 words are related, they should count as one. :)
My dictionary gives 2 pronuncications for each, /en.../ and /eIn.../.
But I've heard the first more.
ranjit_...@yahoo.com
Oops; the one in Kansas was born in Grand Rapids, Michigan.
Christian Weisgerber
[ancient]
> > BTW, is there any reason why this "a" (in English) is a 'long a'? Any
> > other examples where "an-" is pronounced with a 'long a'?
>
> As much as keep I trying, I can't think of any polite words,
> just these two: "anal" and "anus". :-)
"angel"
--
Christian "naddy" Weisgerber na...@mips.inka.de
Oliver Cromm
Now I wonder whether I would perceive those t's as part of segmental
phonology, or maybe I automatically dismiss them as some kind of
production artefact, and that's why I "didn't notice". And that could
have to do with my native tongue. In my (and the standard) variety of
German, certain contexts don't allow /s/, and while I can hear and
produce it, as long as I'm alert, I noticed numerous other German
speakers producing /ts/ in the effort of avoiding falling back to /z/.
--
*Hardware* /n./ The parts of a computer that can be kicked
Bart Mathias
> PaulJK <paul....@paradise.net.nz> wrote:
>
> [ancient]
>>> BTW, is there any reason why this "a" (in English) is a 'long a'? Any
>>> other examples where "an-" is pronounced with a 'long a'?
>> As much as keep I trying, I can't think of any polite words,
>> just these two: "anal" and "anus". :-)
>
> "angel"
And that counts in a way that "anal" and "anus" (and for some, "anile")
don't: the "an" is a syllable.
I'm not sure what LEE meant by "this 'a'" in reference to "ancient," but
if it's OK to put a consonant in front, there's "(ar)range," "mange,"
"change," etc.
benl...@ihug.co.nz
Well, that makes it a little more geographically coherent.
So what's the range of environments in which this occurs? Have you
checked to see whether Labov's onto it?
Ross Clark
benl...@ihug.co.nz
Oh, and I forgot to ask: How old are these people?
Ross Clark
PaulJK
> Christian Weisgerber wrote:
>> PaulJK <paul....@paradise.net.nz> wrote:
>>
>> [ancient]
>>>> BTW, is there any reason why this "a" (in English) is a 'long a'? Any
>>>> other examples where "an-" is pronounced with a 'long a'?
>>> As much as keep I trying, I can't think of any polite words,
>>> just these two: "anal" and "anus". :-)
>>
>> "angel"
Ahhh, good one.
It has a number of related words, from angel cake and angelfish
to angel dust. However, I notice that "angelic" is pronounced with
/en.../, not /eIn.../. Now, why is that?
Is it just a matter of syllabic stress? If "an-" or "a-" forms
a stressed syllable it's pronounced /eIn.../, if it's unstressed
it's an /en.../.
Would that be the answer to LSD'd original question:
"BTW, is there any reason why this "a" (in English) is a 'long a'?"
> And that counts in a way that "anal" and "anus" (and for some, "anile")
> don't: the "an" is a syllable.
True.
:-) In defence of "anal" and "anus", I say "-al" and "-us" is just
morphology, so "an" is the root.
> I'm not sure what LEE meant by "this 'a'" in reference to "ancient," but
> if it's OK to put a consonant in front, there's "(ar)range," "mange,"
> "change," etc.
I assumed, by "an-" he meant first two letters in a word.
pjk
LEE Sau Dan
>>> "angel"
PaulJK> Ahhh, good one.
PaulJK> It has a number of related words, from angel cake and
PaulJK> angelfish to angel dust. However, I notice that "angelic" is
PaulJK> pronounced with /en.../, not /eIn.../. Now, why is that? Is
PaulJK> it just a matter of syllabic stress? If "an-" or "a-" forms
PaulJK> a stressed syllable it's pronounced /eIn.../, if it's
PaulJK> unstressed it's an /en.../.
Animal?
Any?
Ant?
>> I'm not sure what LEE meant by "this 'a'" in reference to
>> "ancient," but if it's OK to put a consonant in front, there's
>> "(ar)range," "mange," "change," etc.
PaulJK> I assumed, by "an-" he meant first two letters in a word.
Yes.
Brian M. Scott
<paul....@paradise.net.nz> wrote in
<news:h3p38i$7qv$1...@news.eternal-september.org> in sci.lang:
> Bart Mathias wrote:
>> Christian Weisgerber wrote:
[...]
>>> "angel"
> Ahhh, good one.
> It has a number of related words, from angel cake and angelfish
> to angel dust. However, I notice that "angelic" is pronounced with
> /en.../, not /eIn.../.
/�n.../, actually.
> Now, why is that?
It's a regular phenomenon: <grateful> but <gratitude>,
<sole> but <solitude>, <divine> but <divinity>. In early
Middle English stressed long vowels in antepenults were
shortened before consonants.
[...]
> Would that be the answer to LSD'd original question: "BTW,
> is there any reason why this "a" (in English) is a
> 'long a'?"
It's long because short stressed vowels followed by /ld/,
/mb/, /nd/, /ng/, /rd/, /rl/, /rn/, [rz], or [r�] were
lengthened in late Old English unless a third consonant
followed the consonant group. That's why <child> has
so-called 'long i' (i.e., /aI/), while <children> has a
short vowel in the first syllable.
[...]
Brian
ranjit_...@yahoo.com
born in the 70s.
ranjit_...@yahoo.com
I can't think of any other words. This was when they were young. Their
accents are not all that pronounced now.
> Have you
> checked to see whether Labov's onto it?
No.
> Ross Clark
Nathan Sanders
"PaulJK" <paul....@paradise.net.nz> wrote:
> Bart Mathias wrote:
> > Christian Weisgerber wrote:
> >> PaulJK <paul....@paradise.net.nz> wrote:
> >>
> >> [ancient]
> >>>> BTW, is there any reason why this "a" (in English) is a 'long a'? Any
> >>>> other examples where "an-" is pronounced with a 'long a'?
> >>> As much as keep I trying, I can't think of any polite words,
> >>> just these two: "anal" and "anus". :-)
> >>
> >> "angel"
>
> Ahhh, good one.
For me, also "angle", "ankle", and "anchor", but I realize many people
pronounce them with something closer to [&].
Nathan
--
Nathan Sanders
Linguistics Program
Williams College
http://wso.williams.edu/~nsanders/
Nathan Sanders
LEE Sau Dan <dan...@informatik.uni-freiburg.de> wrote:
> >>>>> "PaulJK" == PaulJK <paul....@paradise.net.nz> writes:
>
> >>> "angel"
>
> PaulJK> Ahhh, good one.
>
> PaulJK> It has a number of related words, from angel cake and
> PaulJK> angelfish to angel dust. However, I notice that "angelic" is
> PaulJK> pronounced with /en.../, not /eIn.../. Now, why is that? Is
> PaulJK> it just a matter of syllabic stress? If "an-" or "a-" forms
> PaulJK> a stressed syllable it's pronounced /eIn.../, if it's
> PaulJK> unstressed it's an /en.../.
>
> Animal?
> Ant?
[&].
> Any?
[E] or [I], depending on the speaker (as a Georgian, I used to have
[I], but nowadays I usually have something more like [E]). Some
people seem to have something lower than [E], around [&]. Off-hand, I
don't know if "any" and "Annie" are homophones for them, and if not,
how "Annie" is pronounced differently (it's ordinarily [&]).
ranjit_...@yahoo.com
> benli...@ihug.co.nz (in sci.lang):
> > In the "pocket" size Kenkyusha which I normally use, the "katakana
> > words" are written in katakana, while the other words are in romaji.
> > All the o.p. would need to do is learn katakana (not hard), and there
> > would be thousands of examples, from which the rules could be worked
> > out "as an excercise for the student". The rules are basically quite
> > simple, though not without exceptions. And you have to remember that
> > English is not necessarily the only source language involved. E.g. /
> > biiru/ 'beer' does not fit with the generally non-rhotic English
> > source, so was more likely borrowed from German.
>
> But German is non-rhotic, too. if biiru was token from German, it would be
> biaa, the same you would expect from non-rhotic English.
In some non-rhotic Englishes, bier/beer is [bj3:] and Bierman is
[bj3:m@n].
> Joachim
PaulJK
>>>>>> "PaulJK" == PaulJK <paul....@paradise.net.nz> writes:
>
> >>> "angel"
>
> PaulJK> Ahhh, good one.
> PaulJK> It has a number of related words, from angel cake and
> PaulJK> angelfish to angel dust. However, I notice that "angelic" is
> PaulJK> pronounced with /en.../, not /eIn.../. Now, why is that? Is
> PaulJK> it just a matter of syllabic stress? If "an-" or "a-" forms
> PaulJK> a stressed syllable it's pronounced /eIn.../, if it's
> PaulJK> unstressed it's an /en.../.
> Animal?
> Any?
> Ant?
Yeah, that kills it.
pjk
PaulJK
> On Fri, 17 Jul 2009 17:45:40 +1200, PaulJK
> <paul....@paradise.net.nz> wrote in
> <news:h3p38i$7qv$1...@news.eternal-september.org> in sci.lang:
>> Bart Mathias wrote:
>>> Christian Weisgerber wrote:
>
> [...]
>
>>>> "angel"
>
>> Ahhh, good one.
>
>> It has a number of related words, from angel cake and angelfish
>> to angel dust. However, I notice that "angelic" is pronounced with
>> /en.../, not /eIn.../.
>
> /�n.../, actually.
I was trying for /�/ on my keyboard, it came out as /e/,
but now I know I have to type Ctrl e a.
>> Now, why is that?
>
> It's a regular phenomenon: <grateful> but <gratitude>,
> <sole> but <solitude>, <divine> but <divinity>. In early
> Middle English stressed long vowels in antepenults were
> shortened before consonants.
>
> [...]
>
>> Would that be the answer to LSD'd original question: "BTW,
>> is there any reason why this "a" (in English) is a
>> 'long a'?"
>
> It's long because short stressed vowels followed by /ld/,
> /mb/, /nd/, /ng/, /rd/, /rl/, /rn/, [rz], or [r�] were
> lengthened in late Old English unless a third consonant
> followed the consonant group. That's why <child> has
> so-called 'long i' (i.e., /aI/), while <children> has a
> short vowel in the first syllable.
Thanks, Brian.
pjk