Looking for languages with nasal vowels

1%homeless · Post by **1%homeless** » Thu Feb 19, 2004 2:35 am

I’d like to compare French nasal vowels with another (living) language that has nasal vowels, but I don’t know of any. Old French has quite a few more nasal vowels than modern French. I’d like to hear what those ancient nasal vowels sounded like. Yes, you could use the theory of just nasalizing the regular vowels to get nasal vowels but.... I’m having some difficulties with Modern French and that approach.

The IPA transcription is weird. This pronunciation tape also takes the same approach. The tape says to pronounce vais and nasalize it to get vin (wine). To me, the vowel in vin sounds like the ‘a’ in “cat”. Then there is fleur to get the nasal sound lundi... it sounds like the ‘a’ in “about” to me. I vary the level of nasality to test this; from full “ng” nasality to really light nasality and what the tape and other sources tell me doesn’t sound right. I mean the “o” and “a” nasal vowels are pretty close to non-nasal vowels, so there must be something that I’m missing with the other two vowels. To make matters more confusing, some parts of France pronounce lundi the same way as vin...

annis · Post by **annis** » Thu Feb 19, 2004 1:34 pm

1%homeless wrote:I’d like to compare French nasal vowels with another (living) language that has nasal vowels, but I don’t know of any.

Ah. Hmm. The only living one I can think of off hand is Navajo, which I doubt will be of much help to you.

Google to the rescue!

This has only a few nasal examples, but has a bunch of sound files for all sorts of curious sounds: sounds.

Ooh, wait: Hindi! I doubt it has the full range of vowel sounds that French has, though.

Episcopus · Post by **Episcopus** » Thu Feb 19, 2004 3:32 pm

Yes I've heard lundi to have a vin vowel too.

I hate the French people. You probably will not find many other languages with such nasal gymnastics. It is extremely annoying when, after learning a language oneself, one can understand but little of common speech since it is but a murmur, I swear. If they take the effort to speak clearly it's fine, French becomes a very natural lovely tongue. Yet if they do not it's just "huhhooohoongngngng". (I don't have the characters to write the phonetic nasal symbols)

Helen of Troy · Post by **Helen of Troy** » Thu Feb 19, 2004 4:18 pm

I feel like an idiot - I don't even know what the nasal vowel are!

Is there a phonetic transcription of these sounds? How do you pronounce the words you wrote (those Old-French words)?

1%homeless · Post by **1%homeless** » Thu Feb 19, 2004 7:19 pm

Thank you William! I'm sure they have Hindi tapes in my library. It seems like they have nasalized diphthongs too! If I remember correctly, I think Old French had nasalized diphthongs too. I did some more searching and it seems like quite a few more (East) Indian languages have nasal vowels. Non Indo-European Devanagari has them too.

Ooh, wait: Hindi! I doubt it has the full range of vowel sounds that French has, though.

Well, according to this place it Hindi has more nasal vowels than French does.
http://www3.aa.tufs.ac.jp/~kmach/hnd_la-e.htm

From the Chinatec examples, it sounds like a robot is speaking. Ijo has much better recordings. Nasal spreading is new to me, but it's hard to discriminate any nasality in the recordings of Sudanese though.

I feel like an idiot - I don't even know what the nasal vowel are!
Is there a phonetic transcription of these sounds? How do you pronounce the words you wrote (those Old-French words)?

Nasal vowels are vowels with air passing through the nose in some degree. They're not Old French words I wrote, they're modern French.

You can listen to some sounds here:
http://french.about.com/library/pronunc ... ion-un.htm

Here is the transcription: http://clauchau.free.fr/L/phonetics.html

Here is the IPA chart of vowels:
http://www2.arts.gla.ac.uk/IPA/fullchart.html

Basicially, most vowels can be nasalized, so instead of writing new vowel symbols they put a marker on top like this tilda: ~

N30F15H · Post by **N30F15H** » Thu Feb 19, 2004 9:14 pm

1. i was attacked by a french man with bellows in a creperie

2. im in top set french although im absolutly *beep* at it

3. i was gonna say french until i actually read the post

Lex · Post by **Lex** » Fri Feb 20, 2004 5:20 am

Helen of Troy wrote:I feel like an idiot - I don't even know what the nasal vowel are!

Say the word "song", but don't pronounce the "s" or the "g" at the end. Try it again, as though you have nasal congestion. That will approximate a nasal vowel.

Lex · Post by **Lex** » Fri Feb 20, 2004 5:22 am

annis wrote:Ah. Hmm. The only living one I can think of off hand is Navajo, which I doubt will be of much help to you.

How about Yanomamo? Is that living? I believe every vowel in "Yanomamo" is supposed to be nasal.

Episcopus · Post by **Episcopus** » Fri Feb 20, 2004 12:44 pm

Just block your nose and say "song" it's fun.

Skylax · Post by **Skylax** » Fri Feb 20, 2004 7:18 pm

Portuguese too uses nasal vowels.

xn · Post by xn » Sat Feb 21, 2004 9:36 am

William and Skylax are right: Hindi and Portuguese are languages with nasal vowels and many speakers.

Episcopus: Hindi has ten different nasal vowels, as compared to French’s four (or three, depending on dialect).

Helen of Troy: the French nasal vowels can be represented in the phrase «un bon vin blanc»; the vowels’ respective IPA symbols are /œ̃/, /ɔ̃/, /ɛ̃/, and /ɑ̃/. (I hope that the IPA text survives posting!)

1%homeless: French and Portuguese have different sets of nasal vowels, so they’re not directly comparable. French and Hindi nasal vowels can be compared: French «bon» with the Hindi word for “eyebrow”, French «vin» with the Hindi word for “I”, or French «blanc» with the Hindi word for “breath”. (I don't know Hindi, but found the sample Hindi words in the Handbook of the IPA.)

Helen of Troy · Post by **Helen of Troy** » Sat Feb 21, 2004 1:59 pm

xn wrote:
Helen of Troy: the French nasal vowels can be represented in the phrase «un bon vin blanc»; the vowels’ respective IPA symbols are /œ̃/, /ɔ̃/, /ɛ̃/, and /ɑ̃/. (I hope that the IPA text survives posting!)

Well, it didn't. I'll try with that links 1%homeless gave.

More on the nasal vowels issue: from the examples you all gave (like "song" etc.) I'm still failing to comprehend the very meaning of termin "nasal vowel". Are they always followed by so-called "velar nasal" ( a group of cons. where the firs one is "n" and the second is one of the velars "c" or "g")? If so, I have few more examples to add - itl. angolo (meaning corner), itl. banca (meaning bank), itl. fianco (meaning side), serbian fem. names like Anka, Stanka ..., and so on. Have I got it right?

annis · Post by **annis** » Sat Feb 21, 2004 2:45 pm

Helen of Troy wrote:I'm still failing to comprehend the very meaning of termin "nasal vowel". Are they always followed by so-called "velar nasal"

Or you can call them "nasalized." They're not just a vowel followed by a consonant sound, but have a different mouth configuration and have a pretty distinct sound.

It's hard to explain in words. Say the first syllable of "angora" (I just saw the move Ed Wood, so excuse the angora fetish). It'll be like "ahng..." where the -ng- is like in "sing." So, keeping that "ahng" going, just ever so slightly separate the back of your tongue from the roof of your mouth, which joined are making the -ng, and move nothing else. Now keep making the -ah- sound with this very slight opening. This should be pretty close to a nasalized "ah".

xn · Post by xn » Sat Feb 21, 2004 9:16 pm

All: I wonder if success with displaying the IPA text is browser-dependent? It displays tolerably well for me, using Mozilla 1.6 on Windows 98 SE.

William: Thank you for suggesting “nasalized” — it is more descriptive than “nasal”.

Helen of Troy: Yes, “nasal vowel” is not a well-defined term. We’re trying to refer to adding a resonance from one’s nasal cavities to a particular sound; that sound could be either a consonant or a vowel. Your list of nasal consonants is valid — one could add the first consonant of the English word “emphatic” to the list — but I’m not familiar enough with Italian or Serbian (or their dialects) to know if they also use nasalized vowels such as French and Portuguese do. In the latter two languages, nasalized vowels don’t require nasalized consonant sounds to follow, e.g. French «saint», Portuguese «são«, although they often have silent orthographic consonants that follow, e.g. French «son», Portuguese «som«.

xn

N30F15H · Post by **N30F15H** » Sat Feb 21, 2004 9:39 pm

whenever my mum does any french she says Haw he haw (all nasalised) which is very annoying

NOTE: this is my chance to evil: american came from english (the rest doesnt matter for this) and nasalised is spelt with an s not a z

xn · Post by xn » Sat Feb 21, 2004 9:52 pm

N30FI5H: Good point on your mum’s French: someone doing an impersonation of Maurice Chevalier will often insert the seemingly obligatory «hon hon hon» interjection, which is replete with nasalised vowels. And you are correct: “nasalised” is spelt with an “s”. Similarly, “nasalized” is spelled with a “z”.

1%homeless · Post by **1%homeless** » Sat Feb 21, 2004 9:56 pm

(I just saw the move Ed Wood, so excuse the angora fetish).

There is no need to excuse angora. I haven't encountered angora, but it must have a lovely feel to it.

That was an awesome movie.

Well, I made a little progress with French vowels and phonetics in general. Basically the main difference between fleur and vais is lip rounding. Can I assume that is why lundi is pronounced the same way as vin? The lip rounding was lost in lundi? Wow, I didn't think Portuguese had nasal vowels, let alone nasal diphthongs. I thought I would not find a European language with nasal vowels, let alone a Romance language!

benissimus · Post by **benissimus** » Sun Feb 22, 2004 2:16 am

N30F15H wrote:NOTE: this is my chance to evil: american came from english (the rest doesnt matter for this) and nasalised is spelt with an s not a z

Many of these distinctly American spellings are actually more etymologically accurate

Lex · Post by **Lex** » Sun Feb 22, 2004 5:33 pm

xn wrote:All: I wonder if success with displaying the IPA text is browser-dependent? It displays tolerably well for me, using Mozilla 1.6 on Windows 98 SE.

I am using IE5, on W2K, and most of the characters you posted show as boxes instead of the IPA characters. Is there a particular font that you have to use to display it?

Perhaps we should decide on the official TextKit way to represent IPA? An official IPA font? But as with Greek fonts, can we rely on them working on everyone's computers? I've found a web page here that has several different ASCII-based IPA encoding schemes, and this page has links to free IPA fonts (and a sound reference page). Any thoughts?

xn · Post by xn » Sun Feb 22, 2004 10:52 pm

Lex: I’m using the Default/Normal/Arial combination of Font Settings in Post a reply. Does Internet Explorer 5 support displaying HTML numeric entities above 255? (It looks like the bulletin board software stored the IPA characters that I’d entered as the numeric entity corresponding to the appropriate Unicode character, which is a reasonable approach.)

All: If deciding the method of IPA representation is subject to a vote, then I’d have to vote for Unicode (i.e. HTML entities that correspond to the appropriate Unicode character). With Unicode, one could have archaic, polytonic, and monotonic Greek; Latin with macrons and breves; IPA; and almost any other character in the same message on any platform, as long as one typeface supports the characters in their Unicode positions.

A good freely available Unicode typeface is Gentium from sil.org, which has Linux, Macintosh, and Windows versions available. It doesn’t contain archaic Greek characters such as digamma and koppa, but it does have polytonic Greek, monotonic Greek, Latin, IPA, and many other characters. (The current version has an italic, but doesn’t have a boldface. The next version, with boldface, is supposed to be released later this year.)

xn

chad · Post by **chad** » Sun Feb 22, 2004 11:53 pm

as a side issue (about a dead language), after working through all the books on how latin sounded in the golden age, i discovered that it's as nasal as french, i was quite surprised after hearing for years that church-style latin pronunciation in movies, which is all un-nasal and somber. e.g. the closest english equivalent to the start of livy book 1 is (with accented syllables in capitals):

iam primum omnium...

"yahng PREE MOHNG nee oohng"... (note -um is elided)

catullus 1.1 goes:

cui dono lepidum nouum libellum

"kwee DOH noh LEH pee doong NOH woohng lee BEH loohng"

bizarre hey? cheers, chad.

xn · Post by xn » Mon Feb 23, 2004 4:44 am

chad: Thanks for noting that — I wasn’t previously aware of the “reconstructed classical” pronunciation. (Good ol’ Google…) It does make the evolution of e.g. Latin ·pāne· to French «pain» or Portuguese «pão« seem less…evolved. I was also surprised to discover that Latin ·magnus· would rhyme with English “hang noose” (i.e. no /g/ sound)!

xn

1%homeless · Post by **1%homeless** » Mon Feb 23, 2004 6:49 pm

after working through all the books on how latin sounded in the golden age

I'd like to know which books these are. I've never encountered the suggestion that there were nasalized vowels in the Golden Age of Latin. I'm wondering if these books were written by French writers...

Velar nasals like xn mentioned I've seen. In the combinations velar+n or n+velar, the velars "should" change to velar nasals. The change might also apply with m combinations too... but the book Vox Latina says about the velar nasal+m pronunciation: "though it cannot be safely recommended."

chad · Post by **chad** » Tue Feb 24, 2004 11:52 pm

hi, i can refer you to this site:

http://www.skidmore.edu/academics/class ... mple2.html

it refers to Prof Sonkowsky's reconstructed latin. He's American, not French by the way

I'd go out to the bookstore today and get the info straight from Sonkowsky's book, but sydney is flooding with rain... the harbour looks like 1800's london and i'm warm in my office hehe

cheers, chad.

Lex · Post by **Lex** » Wed Feb 25, 2004 2:41 am

xn wrote:Lex: I’m using the Default/Normal/Arial combination of Font Settings in Post a reply. Does Internet Explorer 5 support displaying HTML numeric entities above 255?

It displays the quote mark code you use. But I have fiddled with the various font and encoding settings that IE5 has to offer, and I can't make it show those IPA characters to save my life.

xn wrote:All: If deciding the method of IPA representation is subject to a vote, then I’d have to vote for Unicode (i.e. HTML entities that correspond to the appropriate Unicode character).

We went through this before when deciding how to represent Greek. The main factor was compatibility with as many systems as possible. What it came down to is:

1) Enough platforms don't handle <fill in name of favorite Greek font here> that most fonts weren't acceptable.

2) The world doesn't seem to be quite ready for nothing but Unicode.

3) An ASCII encoding scheme that anybody can enter and read with relative ease on any OS/browser combination (once you learn the scheme, of course) seemed the way to go.

4) BetaCode is a quasi-standard and is known already by those Greek geeks who frequent mailing lists, so it seemed the code to use.

5) The fact that the SPIonic font uses a variation of BetaCode as its encoding scheme clinched it. Those whose systems can handle SPIonic can read actual Greek letters, while those whose systems can't can still read the BetaCode.

6) Once you learn BetaCode, which isn't really that hard, it's a heck of a lot easier to type in Greek than it would be using HTML entity codes, or so I would think.

These reasons are why I thought that a similar ASCII encoding scheme/font combination would be best for IPA, if such a combination exists.

xn wrote:A good freely available Unicode typeface is Gentium from sil.org

Thanks. Free Greek fonts are always a good thing.

xn · Post by xn » Wed Feb 25, 2004 4:44 am

Lex: Try visiting http://www.sover.net/~xn/IPA.xhtml , and follow the link on this page to download Gentium. (This page also has a link to an ASCII encoding for IPA.) Install Gentium, and reload the page. If IE5 supports XHTML, then the page should also be a conclusive test on its ability to display IPA. As an aside, do you prefer IE5 to IE6?

The world will _never_ be ready for nothing but Unicode; after all, Morse code is still being used and updated! (A character code for “@” was recently added to its list, specifically for e-mail address support.) ASCII encoding schemes are frequently easy to enter, but reading the encoded text often isn’t. Me, I usually (but not always

) cut and paste Unicode characters rather than memorize and type their HTML entity equivalents.

Given the preference for BetaCode, it sounds like the best of both worlds would be achieved by creating an applet that would accept BetaCode (or SPIonic BetaCode) as input and produce (HTML entity) Unicode as output; this would allow both ease of entry and non-typeface-dependent storage and display…of nasalized vowels

.

xn

Lex · Post by **Lex** » Fri Feb 27, 2004 11:19 pm

xn wrote:Lex: Try visiting http://www.sover.net/~xn/IPA.xhtml , and follow the link on this page to download Gentium. (This page also has a link to an ASCII encoding for IPA.) Install Gentium, and reload the page. If IE5 supports XHTML, then the page should also be a conclusive test on its ability to display IPA.

I installed Gentium. The IPA symbols on the page you linked to show up OK, but not the ones here on textkit. I even tried to wrap them in a Gentium font tag, and see if they showed up properly in a post preview window. No luck. I looked at the source for the page, and I suspect that the board would have to hacked a bit to accomodate Gentium the way it now does SPIonic.

xn wrote:As an aside, do you prefer IE5 to IE6?

Not especially, but it's what was installed with the OS. I could download IE6 or Mozilla, but I have dialup, and so would rather not. Which is beside the point, since the goal is to be friendly to many platforms, not make all platforms have to conform to textkit.

xn wrote:ASCII encoding schemes are frequently easy to enter, but reading the encoded text often isn't.

True. The IPA encodings look like they would be hard to read. But Betacode isn't all that bad.

xn wrote: Given the preference for BetaCode, it sounds like the best of both worlds would be achieved by creating an applet that would accept BetaCode (or SPIonic BetaCode) as input and produce (HTML entity) Unicode as output; this would allow both ease of entry and non-typeface-dependent storage and display…of nasalized vowels .

Well... the Betacode/SPIonic method already works passably well for Greek, and I'm not sure if Jeff wants to go to all the trouble of writing an applet, just to display nasalized vowels.

He might be willing to hack the board to accomodate Gentium, though, if we ask nicely. Even then, though, one might still have to post both HTML entity Unicode and ASCII encoded IPA, if he wants those who can't use the font to be able to read his text.

Cleo · Post by **Cleo** » Sat Feb 28, 2004 3:16 am

I'm French speaking (not French nationality) and I have no idea how anyone can hear the same sound in "lundi" and "vin"!
I asked DH and he too was totally surprised. To our French-trained ears, they're totally different.

With my meager attempt at learning Latin using an English book, I am discovering that ears can be trained to discriminate sounds, based on your native language. I am not a linguist by far, so this took me by surprise!
All the explanations I kept reading about long vowels and short vowels are meaningless to me. I just don't *hear* those. But I sure hear the difference between "lundi" and "vin"!

Oh, the fun and pleasure of learning languages...

xn · Post by xn » Sat Feb 28, 2004 3:29 am

Lex: I took a look at the underlying HTML here, and came up with a reason why I can see IPA HTML entities but you can’t: it’s due to the <style> element specifying

font,th,td,p { font-family: Verdana, Arial, Helvetica, sans-serif }

In my case, since it didn’t find IPA glyphs in either Verdana, Arial, or Helvetica, it displayed them in Lucida Sans Unicode, a sans-serif typeface with IPA glyphs which is installed on my computer. My guess is that you currently don’t have any sans-serif typefaces with IPA glyphs installed on your computer. Gentium, being a serif typeface, wouldn’t be used without a change in the above <style> element. Lucida Sans Unicode can be downloaded from http://www.phon.ucl.ac.uk/home/wells/lsansuni.ttf — it’s about 300 kiB.

I didn’t intend to imply that using IE6 rather than IE5 would be a prerequisite for conformance to textkit; I simply was curious why you used version 5. I only have dialup as well, but when I wish to download a large file, I’ll start it when I’m planning on doing some other activity for a period of time (usually a household chore of some sort).

xn

benissimus · Post by **benissimus** » Sat Feb 28, 2004 3:32 am

xn, your apostrophes are showing up as coding for some reason.

xn · Post by xn » Sat Feb 28, 2004 4:08 am

Cleo: Oops, I missed that portion of 1%homeless’ message. My ears were raised in the States, yet I hear the vowels in «lundi» and «vin» as being distinct. (I am also fortunate enough to live within range of Sherbrooke radio stations.) I can’t think of many French words that might be distinguished solely by vowel length — perhaps some words with o-sounds, e.g. «sot» vs. «seau»? Your ears are certainly better judges than mine would be!

benissimus: I noticed that as well. Perhaps including CSS in a reply affects entity transformation?

xn

Lex · Post by **Lex** » Mon Mar 01, 2004 10:45 pm

xn wrote:Lex: I took a look at the underlying HTML here, and came up with a reason why I can see IPA HTML entities but you can't: it's due to the <style> element specifying

font,th,td,p { font-family: Verdana, Arial, Helvetica, sans-serif }

In my case, since it didn't find IPA glyphs in either Verdana, Arial, or Helvetica, it displayed them in Lucida Sans Unicode, a sans-serif typeface with IPA glyphs which is installed on my computer. My guess is that you currently don't have any sans-serif typefaces with IPA glyphs installed on your computer. Gentium, being a serif typeface, wouldn't be used without a change in the above <style> element.

I checked, and I do have Lucida Sans Unicode available on my system. I even changed my default web page font in the IE settings to Lucida Sans Unicode, but still no luck! I think that's because I also have Arial and Verdana installed, so that it defaults to one of them (and I don't plan on uninstalling them).

When I use the SPIonic face tags, like this:

Code: Select all

χ

Here's what I see in the page's source:

Code: Select all

<span style="font-family: 'SPIonic'; line-height: normal">x</span>

That successfully shows up as a Greek chi, since I have SPIonic installed. But when I try to specify another font, such as Gentium, like this:

Code: Select all

[face=Gentium]x[/face]

I see this in the source (and on the page):

Code: Select all

[face=Gentium]x[/face]

The board must have a hack that translates SPIonic face tags to span tags (and probably also Arial and Verdana, since they are also in the font drop-down box in the post composition screen), but it doesn't do the same for Gentium. So, in order for posters to force a local standard IPA font to be used like we do with SPIonic, I still think Jeff would have to hack the board again. And even if he does, the post will still not be readable for those who cannot use Gentium (or whatever the font is) on their machines. For them to read it, ASCII encoding would still be needed as well. So, to have both a nice pretty font, and universal readability, you'd have to write the IPA twice -- once in HTML entity codes wrapped in Gentium tags, and once in ASCII encoding.

Here's my opinion, for what it's worth. It seems to me that this makes an IPA font hack not worth the trouble. As ugly as it might be, it seems a straight ASCII encoding is the way to go. I don't know that we'll have that much call for IPA, anyhow, so I doubt that an "official" encoding scheme is necessary, either. Just pick a flavor that you like, and specify what scheme you're using in your post. And to be friendly to those who aren't familiar with the scheme, you could post a link to an ASCII to IPA key, such as the one off the omniglot.com IPA page that I linked to earlier.

annis · Post by **annis** » Mon Mar 01, 2004 10:53 pm

Lex wrote:Here's my opinion, for what it's worth. It seems to me that this makes an IPA font hack not worth the trouble. As ugly as it might be, it seems a straight ASCII encoding is the way to go.

There is a standard way to map ASCII to IPA: SAMPA and X-SAMPA.

Ken_Teng · Post by **Ken_Teng** » Tue Mar 02, 2004 2:58 am

Perhaps Sanskrit and Chinese[/quote]

chad · Post by **chad** » Tue Mar 02, 2004 3:11 am

sorry, what's this long discussion actually about (for the non-programmers)?. is it, ideally, changing textkit's forum font to gentium?

please don't do that. i had spionic installed especially by IT, and i can't install any more fonts/applets/programs on my computer. we don't have access to our fonts directory or local drives, and can't download.

thanks, chad.

xn · Post by xn » Tue Mar 02, 2004 5:15 am

lex: I have Verdana and Arial installed as well. Given that you also had Lucida Sans Unicode installed, it sounds as though it’s a browser issue: Mozilla 1.6 knows not to use a given typeface in a font-family rule when it doesn’t have a glyph for a particular character, but apparently Internet Explorer 5 will use the first installed typeface, whether it has the appropriate glyph or not.

Does IE5 support alternate stylesheets? If it does, I could modify my IPA page to specify different typefaces with different stylesheets, which would allow you to explore IE5’s behavior.

Regarding the chi, it shows up for SPIonic because the Latin character x is mated to a Greek chi glyph in SPIonic. For a typeface with Unicode support such as Gentium, which pairs Latin characters with Latin glyphs and Greek characters with Greek glyphs, one would have to use the Greek character chi (or corresponding HTML numeric or character entity) to display the Greek glyph chi:

Code: Select all

&#x03C7;

or

Code: Select all

&chi;

It does sound as though the BBS software would need to be modified to accomodate additional [face]-to-typeface conversions; I’m not familiar with the software, so I don’t know how difficult it would be. In the best case it would just be adding another typeface (e.g. Gentium) to the text file that currently contains Arial, SPIonic, and Verdana.

The apparent behavior of IE5 does complicate the possibility of using Greek characters (as opposed to Latin characters with Greek glyphs); the BBS software would have to avoid specifying typefaces without Unicode support in the “default” font-family rule to avoid a problem with IE5.

chad: The discussion started on why IPA characters in board messages were visible to me but not for lex. Gentium usage was suggested only as a possible method by which the difference between my viewing and lex’ viewing could be determined. There is no plan to replace SPIonic with any other typeface — not least because if there were, the SPIonic mismatch of Latin character to Greek glyph would require massive message-text conversion.

xn

Lex · Post by **Lex** » Tue Mar 02, 2004 10:35 pm

annis wrote:
Lex wrote:Here's my opinion, for what it's worth. It seems to me that this makes an IPA font hack not worth the trouble. As ugly as it might be, it seems a straight ASCII encoding is the way to go.
There is a standard way to map ASCII to IPA: SAMPA and X-SAMPA.

There's also Kirshenbaum, Coutts-Barrett, Branner, and Carrasquer, listed here. That's the great thing about standards; there are so many of them to choose from!

BTW, xn, I think that was William's subtle way of telling us that he, as moderator, would prefer that we use SAMPA/X-SAMPA as our ASCII encoding scheme.

annis · Post by **annis** » Tue Mar 02, 2004 10:58 pm

Lex wrote:There's also Kirshenbaum, Coutts-Barrett, Branner, and Carrasquer, listed here. That's the great thing about standards; there are so many of them to choose from!

Wow. I had no idea there were competing schemes. I've only ever seen SAMPA since the conlang people love it.

BTW, xn, I think that was William's subtle way of telling us that he, as moderator, would prefer that we use SAMPA/X-SAMPA as our ASCII encoding scheme.

It might be nice for us to agree on something, but I have no good reason to endorse SAMPA except my familiarity with it.

So that wasn't actually a hint.

Lex · Post by **Lex** » Tue Mar 02, 2004 11:14 pm

annis wrote:
Lex wrote:There's also Kirshenbaum, Coutts-Barrett, Branner, and Carrasquer, listed here. That's the great thing about standards; there are so many of them to choose from!
Wow. I had no idea there were competing schemes.

I had no idea about any of them, until I did a google search for this thread, so don't feel bad.

annis wrote:
BTW, xn, I think that was William's subtle way of telling us that he, as moderator, would prefer that we use SAMPA/X-SAMPA as our ASCII encoding scheme.
It might be nice for us to agree on something, but I have no good reason to endorse SAMPA except my familiarity with it. So that wasn't actually a hint.

OK. Well, since I'm not familiar with any of them (or with proper IPA with a font like Gentium, for that matter), it matters little to me. I just thought some sort of convention would be nice, when there is call for using IPA. SAMPA is as good as any.

benissimus · Post by **benissimus** » Tue Mar 02, 2004 11:15 pm

How many people here can understand IPA, ahem!

Textkit Greek and Latin Forums