Talk:Schwa

French schwa is spoken with rounded lips (close to ø/œ), German (and afaik Dutch, English) schwa is not. Do they both count as the same schwa described in the article? 213.6.2.110 15:05, 10 Dec 2004 (UTC)

I would say no. Schwa is universally described as "neutral", which means that it is the sound that comes out when all of the speech apparati except the vocal cords are relaxed. There are a few languages that also have unvoiced schwa, such as some dialects of Inuit, but these are generally represented as the unvoiced /a/ by which they are represented in other dialects. The ø as a vowel does not indicate a specific phoneme, since in some languages it is the same as German ö fortis, and in other languages, as French œ. On the other hand, œ is itself represents (as far as I'm aware) only a schwa pronounced with tightly rounded lips. The difference is between French "deux" and "de". (I think...I'm not actually a genie of French phonetics...) TShilo12 06:50, 14 Feb 2005 (UTC)

---

I thought the schwa was in all the Germanic languages as a sound made by unstressed E. I do know that it is a sound in Catalan, which is a Romance language. (I don't know of any other Romance languages with this sound, except for possibly French.)

It might well be in many languages, but only in English, that I know of, is it called "Schwa". I already pointed out on the page that the same sound is represented in German and French in various ways, (as well as many other languages) but in those languages it is explicitly specified, has a name and a graphic representation of its own. In English, most native speakers use it by imitation, but wouldn't even believe it was a legit English sound if you asked them. In my view, that's what makes it worth an entry in the first place. After all, there's no interesting encyclopedia entry for "E". --User:Steverapaport

---

"Adept" is pronounced with the accent on the second syllable when used as a predicate adjective: "He is quite adept." But it is pronounced with the accent on the first syllable when the word is used as a noun: "He is a manual adept." In the latter case, the "a" is definitely not a schwa! I replaced this example with "about". -- Mike Hardy

My version of the English language never stresses the first syllable in adept, and the first vowel is always a schwa. -- Zoe

My pronunciation is the same as Zoe's, but I see Mike's version in the dictionaries I've checked (ex [1]). --Brion 00:41 Nov 20, 2002 (UTC)

---

As for the article, in Hebrew, the proper pronunciation is still sh'tayim (with a shwa). People may skip over it, but in proper spoken Hebrew it is still there. Danny

Can you provide an example where it is silent, then? -phma

Hebrew has two shwas, both written with the same sign. The schwa you gave is called a sheva na' and is always sounded. The silent is a sheva nah. It depends on the placement in the word. In the first syllable (a shwa is not considered a syllable grammatically) it is always sounded. after a short vowel it is silent, but the tendency in Hebrew today is not to include them so as to avoid confusion. Generally, though, the vowels are not written at all except in prayer books, children's books, and poetry. I hate reading Hebrew with vowels personally. Danny

---

I think in Catalan it's not possible to pronounce the schwa sound in an stressed v owel, but I'm neither a native speaker of Catalan, nor an expert in the language. In the text it says that it is a characteristic form Bulgarian, may be also from Catalan. If any one could check it...

grh

It's true that in Catalan & Valencian spoken on the mainland do not have stressed schwas. However, the Catalan of Mallorca does. It's found in words such as estreta (stress on penult). --Chris 20:00, 24 Jul 2004 (UTC)

this article lacks a sound sample

Schwa cleanup

(Below moved from User talk:Steverapaport)

That was a very ..er... enthusiastic cleanup, Nohat. Not a bad idea but I think you may have tossed a few of the important bits too.

I moved a fair bit of content to unstressed vowel, because a lot of the discussion on schwa wasn't about the schwa sound at all, but other reduced vowel sounds in English.

Here's what I think you missed:

1- Schwa is, in English (and Hebrew) considered a "null" sound, the equivalent of "no vowel at all". English and Hebrew speakers asked to pronounce, say "Mxpltzk", are pretty much guaranteed to use schwa at least 3 or 4 times. The sound appears in French or German too but has its own distinct identity. This is probably the reason the English sound is named after the Hebrew one, rather than the more obvious examples of the same sound in more-related French or German.

This is a fine point; I don't think it was clear in the article before I cleaned it up. The way you have explained it here would be fine on the page as far as I'm concerned.

Ok I did that, and pointed out as a bonus the reason for the lack of conflict in Hebrew between the two schwas.

2- Schwa is a sound that's never taught to English-speakers (it doesn't even have a fixed symbol) and so you need at least a few examples for English-speakers to recognize it by. You only left one, and a pedagogically-poor one at that. Anyone who doesn't know what schwa sounds like will probably go away thinking it sounds like /V/ or /A/. The US "uh" sound can also sound like these. Instead, use a drawn-out example that remains true to sound even when spoken slowly, like "third". Then mention that it shows up in many unstressed syllables like the "about" or "cycle".

The Merriam-Webster definition of schwa says "... the usual sound of the first and last vowels of the English word America", an equivalent example to about. In fact, schwa does sound like /ʌ/, as it's the closest stressed vowel to schwa—in fact Merriam-Webster uses the same symbol to represent both schwa and /ʌ/. They are admittedly different—schwa is somewhat closer than /ʌ/, but the vowel of third is NOT schwa, it's open-mid central unrounded vowel, or a rhotic version thereof, depending on dialect. For the 70% of native English speakers that are American English speakers (not to meantion that vast majority of foreign learners of English, who learn American English), the vowel sound of third is nothing like schwa. Also, there is no schwa in cycle. The 'le' represent a syllabic 'l'. The details of all the non-schwa unstressed vowels are on unstressed vowel.

Now I'm learning stuff. I'll be very interested to see your changes. I've listened to the sound sample for the open-mid central unrounded vowel and it's definitely not how I'd pronounce "third". (I'm from Toronto, I speak like newscaster Peter Jennings.) Having lived the last 5 years in Europe (Italy & Sweden) I'm starting to internationalize my accent a bit and dropping what remains of my Canadian bias, but I'm sure I can be confused, too. But I thought that the schwa was closer to a mid than a back vowel, and Webster notwithstanding, I can't believe /ʌ/' at all. I'd be more likely to agree with one of the other mid-central vowels such as /3/ as an approximation.

Well, it's complicated. The sound /ʌ/ as pronounced in English is rather more central than the IPA vowel chart would have you believe. But it's still traditional to use that symbol for the vowel in cut and run, even though that the particular vowel sound is really only a touch more open than schwa. The reason M-W uses the same symbol is because they are in complementary distribution—/ʌ/ and schwa never occur in the same environment because schwa only occurs in syllables without stress and /ʌ/ only occurs in syllables with stress. I don't actually agree with M-W—I think schwa and /ʌ/ are not merely allophones that occur in stressed and stressless syllables but separate phonemes. But there is some validity to the argument because the sounds are very close, at least in American English.

On the other hand, I don't think /3/ makes for a good approximation because only about 25% of English speakers have that sound. It's likely to be confusing to the other 75% of speakers who pronounce third with a rhoticized /3/ which sounds nothing like schwa. Nohat 20:23, 12 Jan 2005 (UTC)

actually more like 70/30, but let's let that go. I've listened to the American samples at [2] and you're right that the schwa there does sound like a schwa. Could we agree for the sake of the 70 that the 'e' in "cinema" is a good example and use that? The last 'a' in 'cinema' and first 'a' in about are not such good examples because people are likely to pronounce words slowly when trying to hear for themselves how they sound, and in that case will likely pronounce both those 'a's more like /V/ and less like schwa. I'd also be happier if we used a transatlantic solution with an RP and an American example.

No, it really is 75%/25% because Canadian English (4.9%) is also rhotic. I think that your preference for 'e' in "cinema" over the 'a' in "about" for the example might be misleading. For those dialects that distinguish schwa and barred i, the 'e' in cinema would be barred i, not schwa. Maybe that's why you think they would be pronounced differently when pronounced slowly? For those dialects that don't make the distinction, they would both be pronounced the same—as schwa. Nohat 21:53, 12 Jan 2005 (UTC)

See below, this is getting too nested --steve

1. The o in harmony is a schwa.
2. The u in medium is a schwa.
3. The y in syringe is a schwa.

Steverapaport 02:48, 13 Jan 2005 (UTC)

Yes, yes, and no. I usually pronounce the first two vowels in "syringe" with the same timbre. I think the problem is that because schwa is mid, central, essentially neutral, often epenthetic, and easily affected by rhoticism and consonantal environment, it's the most unstable vowel in English. I don't think it's even possible to give one word that manifests the same schwa in all dialects of English. If you hear a West Indies English speaker with a really thick accent, sometimes you begin to doubt whether schwa is present in that dialect at all. thefamouseccles 02:02, 26 Jul 2005 (UTC)

Pronunciation of 'synonym'

Schwa is the most common vowel sound in English, the unstressed vowel in many unstressed syllables, like the 'a' in about or the 'o' in synonym. In my area - northeastern New Jersey - people rarely, but sometimes, pronounce the 'o' in synonym as in ode, although it's usually a schwa... Has anyone heard the pronunciation sin-oh-nim, and should synonym be changed to a different word to avoid any ambiguity? --Lady~Macbeth 01:06, 1 Jun 2005 (UTC)

It's possible to give any schwa a "full" pronunciation by looking at the spelling; syn-O-nym, synth-E-sis, A-bout, etc., but no dictionary that I'm aware of sanctions these pronunciations. They should be considered like any other spelling pronunciation: hypercorrections without any orthoepic authority. Nohat 04:51, 26 July 2005 (UTC)

I can pronounce synonym with a full /o/, but then it grabs the stress of the word and I end up saying sin-OH-nim, and not SIN-uh-nim. --Suicidium 20:53, 9 March 2006 (UTC)

Am I pronouncing this sound right?

Can anyone tell me if my pronunciation of this sound, timestamped 08:20, 26 July 2005, is accurate?  Denelson83  23:51, 6 August 2005 (UTC)

Sounds exactly like the Chinese word for "hungry", which AFAIK is a perfect [ə] in IPA. --Menchi 01:27, 7 August 2005 (UTC)

I'm fairly convinced that è (饿) is a close-mid back unrounded vowel. At the very least it would seem odd if it didn't properly rhyme with words containing an initial and/or a final like hē (喝) or gen (跟). There's a sound sample in the vowel article if you're interested.

Peter ^Isotalo 08:34, 20 August 2005 (UTC)

Merger proposal

I am proposing that the section on the Mid-central unrounded vowel, in the article on the Close-mid central unrounded vowel, be moved here. From its own description, that sound is a mid vowel, not a close-mid vowel. FilipeS 14:16, 21 April 2006 (UTC)

Actually, we should just delete the mention of a mid-central vowel in the close-mid article. I'm not a big fan of mentioning mid-central vowels in the articles about close-mid vowels anyway, and there's already an article on schwa. In addition, the claims that the section in the close-mid article makes need verification. AEuSoes1 21:04, 21 April 2006 (UTC)

Schwa isn't the same thing as a mid-central vowel. kwami 06:08, 25 April 2006 (UTC)

...and you say this because...? AEuSoes1 06:05, 26 April 2006 (UTC)

Schwa is a reduced vowel. It is not necessarily mid, not necessarily central, and not necessarily unrounded, though it tends to be all of those. French schwa, for example, is a mid-front rounded vowel. On the other hand, a mid-central unrounded vowel is not necessarily reduced. (Sometimes such vowels are called 'full schwas', but 'mid-central unrounded vowel' may be considered less ethnocentric.) kwami 06:27, 26 April 2006 (UTC)

I'm sorry but, looking at the IPA vowel chart, I see a specific height and backness of [ə]. Even if the word "schwa" is ethnocentric, it is still silly to me to refuse to use the character for an actual mid-central unrounded vowel. The variation in its use is probably related to the same phenomenon that allows <ʌ> to be used instead of [ɐ], <ɹ> instead of [ɻ], etc. I really doubt that the French schwa is front. Do you have any sources to back up your claims? AEuSoes1 03:28, 27 April 2006 (UTC)

Sorry, I got that partially wrong. (It was rather late!) I once had quite a nice French phonetics book that described /ə/ as [œ], but perhaps that was intended as a nearest equivalent. The IPA Handbook has it as (rounded) [ə]. I suspect the Handbook might be the more reliable of the two.

[ə] is a mid-central vowel, undefined for roundness, and the symbol <ə> is called 'schwa'. However, the symbol <ə> does not necessarily represent the phonetic value [ə], and the name 'schwa' is used for all manner of reduced vowels other than [ə]. That is, while [ə] is undoubtedly mid-central, schwa isn't. Does that make any sense?

However, I guess this isn't too different from using /r/ for English [ɹ] ([r] is a trill, but English /r/ isn't). Maybe we just need to be clear whether we mean 'schwa' phonologically, as a reduced vowel, or phonetically, as a mid-central vowel.

Anyway, phonetic schwa/mid-central vowel isn't the same as a mid-central unrounded vowel, as its roundness is ambiguous. kwami 05:00, 27 April 2006 (UTC)

It really is quite dreadfully confusing. The world of phonetics could really use a terminological clean sweep. There should be separate terminology and symbology when talking about phonetics and when talking about phonology. Someone like a Linnaeus needs to come along and just say this is this and that's that, and create a framework for resolving terminological imprecision and ambiguity, like biology has for naming organisms. I recognize that there is value in there being vagueness in certain terms, but we should separate the vague terms and the precise terms. *sigh* Nohat 06:53, 27 April 2006 (UTC)

The current introduction to the article on the Schwa has the following:

"In linguistics and phonology, schwa can mean:

• An unstressed and toneless neutral vowel sound in any language, often but not necessarily a mid-central vowel (rounded or unrounded). Such vowels are often transcribed with the symbol ə, regardless of their actual phonetic value.
• The mid-central unrounded vowel sound in the middle of the vowel chart, stressed or unstressed. In IPA phonetic transcription, it is written as ə. In this case the term mid-central vowel may be used instead of schwa to avoid ambiguity.
• The symbol ə is itself called schwa."

Either this definition is incorrect, or a Mid-central unrounded vowel is a schwa, since it satisfies the second definition. FilipeS 13:42, 27 April 2006 (UTC)

The definition was incorrect. kwami 05:42, 28 April 2006 (UTC)

All of this makes sense except this claim:

• If precision is desired, the symbol for the close-mid central unrounded vowel may be used with a lowering diacritic, [ɘ̞].

That's really the main thing I have a problem with. I'm also not so sure about the ambiguous rounding of schwa. There are three other IPA vowels that don't have a specific counterpart in roundness on the chart-[ʊ] [æ] [ɐ]-but their rounding is not ambiguous (although the first is used in English for an unrounded counterpart). AEuSoes1 06:12, 28 April 2006 (UTC)

[ʊ] and [æ] are defined as rounded and unrounded in the IPA Handbook ("near-close near-back rounded vowel" and "near-open front unrounded vowel"), but [ɐ] and [ə] are not ("near-open central vowel" and "mid central vowel"). Open vowels seldom have rounding contrasts, so that isn't much of an issue for [ɐ], but [ɐ] could be rounded too.

With [ɘ̞] you're clearly saying that the vowel is unrounded. With [ə] you're not (it could be [ɘ̞] or [ɵ̞]), and moreover only [ə] has the connotation of being a reduced vowel. For example, the Handbook says of French [ə] that it has "some rounding". kwami 08:39, 28 April 2006 (UTC)

So, if I'm understanding this correctly, the definition of schwa does not state its roundedness. It's an ambiguous notion (underspecified, I think is the term). Kwami's reading seems to be that this implies that a schwa cannot be unrounded. My interpretation is that it may be rounded or unrounded. If the vowel of French you mentioned can be denoted [ə] even though it's rounded, then I don't see why the other vowel we're discussing can't be denoted [ə] even though it's unrounded. (Is there no rounding/unrounding diacritic that can be added to the [ə], for a more specific description?) FilipeS 11:54, 28 April 2006 (UTC)

Sorry, that's not what I meant to say. [ə] could be either [ɘ̞] or [ɵ̞]. That is, either rounded or unrounded, without specification. That means both [ɘ̞] and [ɵ̞] are more specific than [ə], and are therefore not equivalent to [ə]. Yes, you could use the 'more rounded' and 'less rounded' diacritics with <ə>, though I imagine a pedant might object "'more rounded' than what?", since it's not relative to a fixed reference point. It would be obvious what you mean, though, so I'd say go for it. kwami 17:19, 28 April 2006 (UTC)

I've moved it to it's own article, Mid central unrounded vowel. Voortle 15:48, 22 June 2006 (UTC)

"A" in about

Is it common for the "a" in about to be a schwa? I doubt this, it's more like a mid-central ɐ than ə. Can someone verify this, please? — ዮም (Yom) | contribs • Talk 00:49, 3 June 2006 (UTC)

It is schwa in the sense of mid-central unstressed vowel of indeterminate specific value. Nohat 01:21, 3 June 2006 (UTC)

Check any dictionary that provides a phonetic enough environment to distinguish between the two. AEuSoes1 02:04, 3 June 2006 (UTC)

On the definition of schwa again

• "The mid-central vowel sound in the middle of the vowel chart, stressed or unstressed. In IPA phonetic transcription, it is written as ə. In this case the term mid-central vowel may be used instead of schwa to avoid ambiguity."

Considering what was said when I proposed a merger, above, it seems that the definition would be clearer if it said that a schwa may be "stressed or unstressed, and rounded or unrounded". FilipeS 19:53, 26 June 2006 (UTC)

Article split

I think this article should be split into two - schwa (letter) and mid central vowel. Schwa is a letter of Azerbaijani and Chechen alphabet and do not represent mid central vowel (actually it represents near-open front unrounded vowel). --Hello World! 11:28, 7 August 2006 (UTC)

You've seen Schwa (Cyrillic), right? AEuSoes1 11:40, 7 August 2006 (UTC)

Azerbaijani uses Latin alphabet. Chechen uses both sets.--Hello World! 11:58, 7 August 2006 (UTC)

• Support. Schwa really seems to be the name of a letter, a grapheme. It’s not a vowel per se, but the sign which stands for one in a number of languages and in IPA. — Jéioosh 20:42, 7 August 2006 (UTC)
• Support. In addition to above, other vowels have alternate names (wedge, epsilon, open o, barred i, etc) but these aren't the article names. If it's good enough for Cyrillic, it's good enough for Latin. AEuSoes1 23:49, 7 August 2006 (UTC)
• Oppose. The name "schwa" is from Hebrew, and can mean both the letter and the vowel. I think people who look for the schwa are likely to search for both. --Kjoonlee 00:49, 8 August 2006 (UTC)
• Oppose. I do support the creation of an article Schwa (letter) (or possibly Schwa (grapheme)). The way to do it (IMHO) is to rename (move) Schwa (Cyrillic) to Schwa (letter) and edit the article to make it address both Cyrillic and Latin use. The disambiguation paragraph in the present article could put the meaning as a letter earlier, as in "schwa can mean: • A [[Schwa (letter)|letter]] occurring in the alphabets of various Turkic and other languages. • ...". Alternatively, the article could start off with: "This article is about the schwa as a phoneme. For the letter used in various alphabets, see Schwa (letter)." Further, if there is to be a new "split", the phonemic article should be called something like Schwa (phoneme) and not Mid central vowel. But both are problematic: as an unstressed vowel the schwa is often not mid-central, and also not necessarily a phoneme. Also this requires setting up a separate disambiguation page.  --Lambiam^Talk 01:08, 8 August 2006 (UTC)

Ah, but there are separate articles for A and А, E and Е, as well as I and І despite homography and similar use. I forgot to ask, though. Is the letter called "schwa" in Azeri or is it a completely different name? Omniglot[3] would have me believe that it is called "schwa." That might be important in the discussion. AEuSoes1 02:49, 8 August 2006 (UTC)

Then we should likewise have a (not redirected) article ə for the letter in Azeri and an article ə for the letter in Kazakh and other languages. An article called Schwa (letter) that does cover the use of the grapheme ə in the Latin alphabet but not in Cyrillic is not a good idea. --Lambiam^Talk 17:53, 17 August 2006 (UTC)

Some comments. Near-open front unrounded vowel and Æ are two seperate articles because the former is a sound/phoneme and the latter is a letter. There should be another article called Æ (Cyrillic) (ru:Ӕ (кириллица)) too (It's an Ossetic letter).--Hello World! 15:01, 18 August 2006 (UTC)

If you do support the creation / split of articles, please vote as support. The naming of the articles can be negotiated then. --Hello World! 10:49, 8 August 2006 (UTC)

• Support. I think it's quite clear that if a word can mean more than one thing, the meanings should be reached through a disambiguation page. From the three meanings in the article I suggest these titles: Schwa (letter), Schwa (phoneme) and Mid central vowel. Even though "schwa" can mean "mid central vowel", it's more analogous with the rest of the vowel sounds to use the latter title. --Starylon 10:12, 8 August 2006 (UTC)
• Support. The grapheme and the phonological use are separate, and could usefully be separated. Schwa (Cyrillic) could be merged into the article on the grapheme (or used as its base), though AEuSoes1's comment does indicate that there is a precedent for having a separate article for the Cyrillic form. — Haeleth Talk 18:32, 8 August 2006 (UTC)
• Support splitting this article into one discussing the unstressed vowel and one discussing the letter. Letters and sounds are different things and should not generally be discussed on the same page. —Felix the Cassowary 07:25, 3 September 2006 (UTC)

How should we do next? --Hello World! 09:32, 13 August 2006 (UTC)

Well, I guess someone should create the article Schwa (letter) for starters. I think we'll need an admin to adjust the redirect. AEuSoes1 18:29, 13 August 2006 (UTC)

• Before the split/cleanup/everythingelse to this article, I would want to know this first - does “schwa” equal to “mid-central vowel”? If yes, that's fine; but if no, there should be a page (in a narrower sense) taking about “mid-central vowel”, and another page for the history of schwa (may be a 3rd page for using the word “schwa” to describe another sound), and finally schwa as a separate letter in Azərbaijani alphabet) --Hello World! 16:52, 16 August 2006 (UTC)

• As far as “mid-central vowel” is concerned, please be advised that that vowel is as common in Bulgarian as it is in English, occurs both in stressed and unstressed position, and is represented by the special Cyrillic letter 'ъ' (that letter is traditionally called 'Er Golyam', i.e. 'Big Er'). Apcbg 10:36, 17 August 2006 (UTC)

As far as I know, the Estonian, French and Japanese Wikipedia have already have an article about the letter. And finally I have written down the Chinese and the English version today. --Hello World! 14:34, 25 September 2006 (UTC)

Awesome. Thanks for doing the work. AEuSoes1 19:50, 25 September 2006 (UTC)

Hebrew?

I think that שְוָא is actually an Aramaic form, and that the Hebrew form is שָוְא (as found in the Hebrew Bible at Exodus 23:1, etc. The forms are related, but distinct... AnonMoos 20:54, 2 November 2006 (UTC)

"The Schwa sound" -- misleading title

If the discussions in this talk page have shown anything, it's that there is no unique schwa sound. There are several schwa-like sounds, depending on the language, and in some cases within a single language. "Schwa" is a function, not a concrete sound. FilipeS 21:01, 2 November 2006 (UTC)

I see what you mean. What would a better title be? "Schwa as a reduced vowel"? No... there are languages with stressed schwa. Perhaps it's okay if the title is misleading as long as the section details the ways in which the title is misleading. This way the reader gets an in depth coverage of the issues with schwa. Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 01:01, 3 November 2006 (UTC)

Not function, but set. -Pgan002 07:53, 22 January 2007 (UTC)

How about just "sound"?... FilipeS 18:09, 3 November 2006 (UTC)

Ooh, Schwa as a phone might be good. That way we cover both phoneme and allophone. Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:35, 3 November 2006 (UTC)

What do you think of "schwa as a neutral vowel"? A little circular, but the term "neutral vowel" is often used for "schwa". FilipeS 15:00, 5 November 2006 (UTC)

Yeah, that is a bit circular, but it works. Go for it! Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 19:48, 5 November 2006 (UTC)

I don't know about circular, but it's not clear. I have to learn what a neutral vowel is to even understand the title. That's not very helpful. If possible, use a simpler title. I think "Sounds" is a much clearer title. The section talks about the set of sounds represented by the concept. Or, use "Vowels". Both of these titles indicate that there is more than one sound. -Pgan002 07:53, 22 January 2007 (UTC)

examples

There are problems with every example presently given in the article.  :-( Sofa doesn't end with a /ə/, it ends with a /ʌ/, and sounds really silly if pronounced with a /ə/ (it actually sounds like how a German might pronounce it)... The "e" in "taken", the "i" in "pencil", the "o" in convince and the y in sibyl are actually all placeholder vowels for /ṇ/ and /ḷ/. The "e" in "the" and the "o" in "synonym" are /ɨ/ not /ə/. I think in an overeager effort to demonstrate that schwa isn't represented by any vowel, a systematic effort was made to demonstrate that it is represented variously by all vowels, and scholarship got lost somewhere along the way... I obviously regard such an effort as frivolous, but if it's felt to be necessary, perhaps the examples should be stacatto, antecedent, gallinule, eloquent, and incredulous. Including an example with a "y" is a bit tougher to come up with off the top of my head, but then again, the "need" to have it is mere pedantry regardless: "y" has never had a sound of its own as long as English has been a written language--it's nothing more than a hangover resulting from the etymologically-friendly spellings preferred by English. (In the case of sibyl, from Greek...) Tomer^talk 04:20, 13 November 2006 (UTC)

I'm not sure why you say that Sofa doesn't end with ə. English, at least in the dialects I've been exposed to, puts [ə] and [ʌ] in complementary distribution. The former in unstressed syllables and the latter in stressed syllables. This is one of the major inconsistencies in the phonemic transcription of English that the resident phonologist at my university chooses to fix by transcribing both as /ə/. I agree in large part to your criticism of the example words. Nasals and liquids muddy up the process of indicating to the reader the ways in which schwa is represented in English. I like your new example word suggestions but check a dictionary (preferably one that uses IPA) to, you know, satisfy the whole NOR thing. Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 07:03, 13 November 2006 (UTC)

I'll see what I can find...upon further reflection, regarding your use of the term "complementary", I think the "complementary" pronunciations may have more to do with open as opposed to closed syllables...which is probably why I find the example of "sofeh" to sound so alien... More later perhaps?  :-p Tomer^talk 07:40, 13 November 2006 (UTC)

(Upon further reflection, a refinement would indicate "open final" as opposed to simply "open"...and v.v.) Tomer^talk 07:42, 13 November 2006 (UTC)

Dictionary.com has an IPA pronunciation guide. I'd check it out myself but I'm a bit busy with schoolwork. Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 09:16, 13 November 2006 (UTC)

Okay I looked up the words myself in Oxford. Staccato works, but there's nothing wrong with about or sofa. Antecedent has an unstressed /ɪ/ which is [ɨ] in my dialect, probably [ɪ] in RP and so not a good example for what we're looking for (this may be the case for many or all e's and i's outside of foreign borrowings). Eloquent works but incredulous has [ʊ], not a schwa. Gallinule has the same problems as the current example words since the i precedes a sonorant. Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 00:48, 14 November 2006 (UTC)

There are a few issues here. The original commentator appears to be confusing phonetics with phonemics. [ɨ] is used as a particular realisation of /ə/ in some American dialects. But in some American dialects (possibly the same ones), /ə/ and /ʌ/ are actually the same vowel phonemically and they’re written differently because of tradition/other dialects. In very few English dialects is /ʌ/ actually pronounced as [ʌ]: It’s usually substantially more central. This is almost certainly the case for American speakers, so a /ʌ/ would be a good example of [ə] for them.

Futhermore, in (it seems to me) many dialects, word final /ə/ may be pronounced somewhat low: [ɐ] or even [ä] (i.e. a near-low or fully low central vowel). This certainly happens in my dialect where I hear no substantial difference between my utterence-final /ə/ and the utterence-final /a/ of languages like Italian or Spanish. The pronounciation may well generally correspond to the pronunciation of /ʌ/ in that particular dialect.

Also, in English /ə/ is essentially an unstressed vowel usually with no contrast for height, rounding or backness. Hence, the only thing your mouth/tongue actually needs to do to signal the vowel /ə/ between consonants is create sufficient space for a vowel. In practice, then, [ɨ] will actually be a very common pronunciation in many contexts. But this is in no way perculiar to English and so doesn’t hinder the utility of an example like ‘synonym’ /'sɪnənɪm/ or /sɪnənəm/. Even languages with only two vowels transcribed /ə/ and /a/ will realise /ə/ like [ɨ] (and very likely [i, e, ə, u, o] in different contexts). The fact that it’s essentially unspecified in English is also what allows the vowel to be completely elided and a word like ‘taken’ pronounced as [teɪkṇ].

This is obviously a bit different for languages/dialects which use /ə/ as a stressed vowel and contrast it with vowels like /i, u, e, o, a/. New Zealand English and Bulgarian are examples of this I think.

In some dialects (typically southeastern English English e.g. RP, Cockney), I think), /ə/ and /ɪ/ still contrast to some extent in unstressed syllables, and this is what Aeusoes observes when he notes the Oxford described ‘antecedent’ as having ‘an unstressed /ɪ/’. As I said, this is not generally the case in former colonial dialects like Australian or American, and any variation between [ɨ] and [ə] that occurs in those dialects is almost certainly free or allophonic variation. This contrast, I’m informed, is neutralised before /l/, so a word like ‘happily’ still manages to have i=ə /'hæpəli/.

So, long-and-short, none of the objections are objections. The schwa in English is a neutral vowel, so it’s absolutely not a surprise that it won’t always be realised as the canonical mid central vowel. Some of the words are poor examples because in some dialects there’s two unstressed vowels, which is an objection to some of the examples, but these can almost certainly be worked around.

—Felix the Cassowary 02:49, 14 November 2006 (UTC)

To begin with, I should clarify, I did not begin this discussion with the interest of forcing change nearly as much as forcing discussion, regarding the given examples. My primary protest had to do with the apparent thrust of the examples in demonstrating the diversity of spelling for the shwa sound in English, rather than in giving representative examples thereof. I highly doubt that someone pronouncing sofa with a schwa is going to cause confusion even where I live, I was simply stating that such a pronunciation sounds "odd". You could pronounce it "sofer" "sofie" or even "sofayei", and you'd no doubt be understood--the chief question would be why you're not calling it a "couch". Hell, you could even call it a "davenport", and probably still be understood; until you started calling it a "chesterfield", and then you'd be sent back to Canadia [sic].

Second, I'm not confusing phonemes with phonetics. From the above, that should be clear. I have striven long and hard to determine why "the neutral vowel" is pronounced two different ways in certain circumstances, and while I don't specifically address it in my manifesto on the subject, have wondered whether or not it's worth the time and effort to bother to differentiate between the two, specifically because the differences are not phonemic. What I have determined is that you can't sound "normal" by interchanging the two willy-nilly, and that they seem to represent two classes of reduced vowels, which, as it happens, are relatively easy to classify, especially when liquids and nasals aren't introduced to the question.

Moving on, my use of "gallinule" as an example does have an admitted weakness, which is dialect-specific in nature. In the kind of English I speak, the /n/ in "gallinule" is not palatalized, so the syllables are quite distinctly ga·lli·nule, without the possibility of ga·lin·[y]ule, and the incumbent possibility of syllabification of the /n/.

All of that aside, there is a tremendous difficulty with this whole topic inasmuch as the indeterminate/neutral vowel[s] in English depend not only upon idiolect and circumstance, but also the fact that the more you try to nail down how things are "really" pronounced, the more you affect the subject you're trying to study, in something of a Heisenbergian sense.

For example, if I say "say the word 'syllable'", everyone will presumably pronounce the "y" as [ɪ]. If we then go to "say the word 'syllabification'", most people are likely to pronounce the "y" as [ɨ], but if you ask them to say the sentence "The Ls in syllabification are syllabificationimized [sic]", they're more likely to pronounce the "y" in "syllabification" as a -/ḷ/ and the "y" in the made-up word "syllabificationimized" as [ɨ] if not [ɪ]. I.e., you destroy your research by asking the simplest question meant to garner information. Human speech is, as all of us here should be well aware, an incredibly fluid [and for those of us who love order, therefore, incredibly conflusterpating] phenomenon.

Another incredibly frustrating thing in trying to nail this all down is that we are all literate, as are the vast majority of any test subjects we or other researchers on this topic might interview, which also screws things up. For us, and for our testees, [no pun intended], there is a natural inclination when being "listened to", to "standardize" or pronunciation as much as possible, which means, in my dialect, pronouncing even a real schwa as one of the "standard" 11 vowels for my dialect [ʌ a ɛ e æ ɪ i o ɔ ʊ u], and squishing [ə] and [ɨ] into one of those vowels. [This doesn't bother to take into account that no speaker of "my kind of English" even realizes that [ɚ] is [ṛ] or that [ɝ] is [ṛ:], etc.]

All of that pontification aside, Felissowary's example of /'hæpəli/ is inaccurate...since it's actually /'hæpɨli/ where I live.  :-p

I'd like to shuck WP:NOR for this, but right now I think it's probably the foremost thing preventing us from getting into an edit war over the topic... :-p Cheers [and good night!], Tomer^talk 09:26, 14 November 2006 (UTC)

I'd like to contrast the idea that [ɨ] is a realization of [ə] in some American dialects. I myself can distinguish between roses and rosa's. It's possible that [ɨ] is the epinthetic vowel of choice, but I've seen languages with [ʏ] as the epinthetic vowel of choice. Also, the actual height and backness is not quite as important in the section because it's the vague "neutral vowel" idea. In French, the schwa is slightly rounded. Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 21:39, 14 November 2006 (UTC)

Concerning the proposal to merge mid central unrounded vowel back into Schwa

Strongly oppose. "Schwa" clearly has a broader sense than "mid central unrounded vowel". Plainly put, not all so-called "schwa sounds" are the mid central unrounded vowel. The common terminology is unfortunately confusing, giving the wrong impression that people are talking about a single sound, when in fact "schwa" seems to be more like a placeholder term for any reduced vowel. See also the reasons why the two articles were split in the first place. FilipeS 16:41, 21 November 2006 (UTC)

I've never liked the idea of having separate articles for every conceivable phonetic entity. Most of these articles on sounds don't even grow beyond a short description and a few examples. Say what you want about phonetic pin-point accuracy, but I find this fragmenting of sound articles to be an extremely unpractical way of providing encyclopedic information on phonetics. I feel we should stick to about one article per IPA symbol, not because it's a reflection of reality, but because it's practical. Kinda the point of establishing the IPA to begin with. Having twice the articles is a lot more difficult to manage than simply pointing out in the articles that language sounds are not precise entities and may vary.

Like many times before, I really feel I must stress that the importance of trying to focus on general readership, not peers and experts.

Peter ^Isotalo 02:21, 17 December 2006 (UTC)

The schwa is not necessarily a mid-central vowel, even though the symbol for the mid-central vowel (which may itself be rounded or unrounded!) is often used indiscriminately for all schwas. The ambiguity of the word "schwa" is therefore a source of great confusion, and that is why I remain a strong supporter of dissecting this concept as much as possible. Whether you do that in one, two, or five articles is less important, but I think that, at this stage of Wikipedia, and given the huge confusion that surrounds this term, separate articles would contribute to avoiding further conflations of non-synonymous terms. FilipeS 13:20, 17 December 2006 (UTC)

I believe that the dichotomy between exact phonetics and inexact phonology is rather misleading, since human speech production per se is vary varied. No matter how many articles we split off, it'll still be an approximation. The problem is not that the concept of a schwa is vague, but that the IPA is a theoretical entity intended to be reasonably exact, but never 100% correct. If anything, it's supposed to be practical, and splitting off one article per minor variant is anything but practical. Except for those who are immersed in it, of course...

Peter ^Isotalo 13:36, 17 December 2006 (UTC)

Sure, every IPA symbol actually corresponds to a broad area in the vowel chart, not to an exact point (and things get even murkier when the same symbol is used to represent quite different sounds, just because they're considered dialects of the same language, or just because it's traditional to use that symbol even though it's not the most accurate one). But the so-called "schwa" is even more ambiguous than that. In this Talk Page, people have described it as:

• a mid-central unrounded vowel;
• a mid-central rounded vowel;
• a near-open central vowel.
• a close central unrounded vowel.

If the IPA has different names for all four of these, why should Wikipedia treat them as the same? FilipeS 13:54, 17 December 2006 (UTC)

The French schwa

I have studied French phonology and the language itself for several years although I am not a native speaker, but my understanding is that the [ə] in words like de and le is realized as [ø] when stressed (as in when the word is pronounced in isolation or enunciated for clarity). The vowel is indeed more neutral when spoken as part of a full sentence. Actually, I have read that many experts do not even consider the schwa a phoneme in French. This means a phonemic transcription of le is actually just /l/. Schwas are inserted according to a regular pattern basically to keep clusters of three or more consonants from occurring (one common exception is parce que), but this is phonetic rather than phonemic.--NeantHumain 20:17, 23 November 2006 (UTC)

You got a source? I've read that French schwa is rounded and that it is somewhat low, but that it's a central vowel. Not the front close-mid vowel. Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 23:07, 23 November 2006 (UTC)

Turkish?

Is there a schwa in Turkish? Is this written using an e with two dots over it? -Pgan002 08:01, 22 January 2007 (UTC)

Yes there is — written by a dotless 'i'; the 'e' with two dots over it is the Albanian schwa I believe. Apcbg 08:38, 22 January 2007 (UTC)

No, a dotless i in turkish represents /ɯ/. Not a schwa. As far as I know, Turkish doesn't have this sound or at least doesn't represent it in its writing. Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 00:23, 23 January 2007 (UTC)

List of languages and examples

Can someone who knows enough about this make a list of languages that have schwa sounds, with examples from each? -Pgan002 08:15, 22 January 2007 (UTC)

Toneless?

The article currently says "An unstressed and toneless neutral vowel sound in any language, often but not necessarily a mid-central vowel." What is meant by "toneless"? Right now the word links to Tone (linguistics), but I'm doubtful that's what was actually meant. More likely, "toneless" was being used in its old sense of "atonic," i.e. "unstressed." I don't think it's possible, in a tonal language, for any vowel to not have a tone. 24.159.255.29 22:39, 17 February 2007 (UTC)

"In General American, schwa is one of the two vowel sounds that can be rhotacized." Who says only two vowel sounds can be rhotacized? What is the other one? R-colored vowel says "The r-colored vowels of General American are written with vowel-r digraphs. Any vowel can be used: ...." In my experience, this seems to be true. 24.159.255.29 22:44, 17 February 2007 (UTC)