Talk:Voiceless palatal fricative

Untitled

I added a similar note on aspirated/unaspirated in German to what I put on voiceless velar fricative. See my comments on Talk:Voiceless velar fricative for concerns about accuracy. CyborgTosser (Only half the battle) 01:54, 31 Jan 2005 (UTC)

---

How is someone suppose to play the .ogg files? -- Bob 12:21, 24 May 2005 (UTC)

See Wikipedia:Media help. — Trilobite (Talk) 19:04, 6 Jun 2005 (UTC)

Are youreally sure it is correct?

I think it should sound like ch in german "ich".88.101.76.122 17:18, 10 August 2007 (UTC)

That's identical to "dicht".--77.1.145.193 (talk) 18:54, 25 May 2012 (UTC)

That's strange «Ich» is written /iç/ but it sounds very different from here, I agree. — Preceding unsigned comment added by 2A01:E35:8B4A:86E0:10A2:2686:3A0F:9B18 (talk) 07:24, 15 February 2018 (UTC)

Standard German "ich"-Laut is exactly the sound. The audio simply isn't perfect because the speaker cannot perfectly produce the sound, which surely doesn't exist in his native language. He's quite close, but the back of the tongue seems to be a bit too low. The result appears to be somewhere between [ɕ] and [ç]. Some Germans actually say it that way, but it's not the prototypical form. 2.201.0.110 (talk) 06:24, 1 October 2024 (UTC)

Turkic vs Azeri

çörək on the k, as voiceless palatal fricative is apparent in some dialects in Eastern Turkey, like several words also in the Kurdish vocabullary (influential maybe, but Kurdish seem to have more). I've heard it also among Tatars in Russia. It's not unique to Azeri Turkic, but also present in other Turkic languages. Hence the change. VartanM (talk) 05:23, 16 February 2009 (UTC)

Are you saying that some dialects of Azeri can have this sound? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 06:29, 16 February 2009 (UTC)

I'm saying that some dialects of Turkic have this sound including Azeri. VartanM (talk) 09:52, 16 February 2009 (UTC)

If you agree that this is a sound that exists in Azeri, then don't remove it. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:11, 16 February 2009 (UTC)

Actually, I don't, I am only talking about things I've heard in the Turkish spoken in Turkey, and Kurdish speaking people. For Azeri, I don't know. What I know, is that in the Baku dialect, which is the official Azeri dialect, the K of bread does not sound voiceless palatal fricative. And it's just silly to add some dialects without defining which ones. Given Parishan's long history in adding "Azeri" everywhere, I have all the reasons to be skeptical. By adding either Turkish or Turkic, we are in the safer mode. VartanM (talk) 08:11, 17 February 2009 (UTC)

The literary dialect of Azeri is not based on the Baku dialect. If "For Azeri, you don't know", then you should not be editing this part of the article in the first place. Parishan (talk) 21:31, 17 February 2009 (UTC)

VartanM, you should try better than this. Parishan (talk) 08:12, 16 February 2009 (UTC)

I'd like too some proof of your claim. Otherwise be a good boy and revert yourself. VartanM (talk) 09:52, 16 February 2009 (UTC)

Not a problem at all: A. Damirchizadeh, Modern Azerbaijani Language: Phonetics, Orthoepy and Orthography. Maarif Publ., Baku: 1972; p. 96.

Now let us hear what you have got. Parishan (talk) 20:57, 16 February 2009 (UTC)

I'd like to see a quote please. Even if the author is not credible, as it considers several Turkic people, like those living in Iraq as Azerbaijani's. He is one of the nationalists who have worked on the fringe theories about Azeri language dating back to a thousand year ago. And since you have a long history of misusing sources, and you can't be trusted anymore You should therefor quote the section, and given the unavailability of the work, you should provide a screenshot. --VartanM (talk) 08:11, 17 February 2009 (UTC)

I do not have a long history of anything, unlike you with your long history of removing references to Azeri and making ridiculous OR claims about languages you do not know the first thing about. I will not go into the trouble of scanning the page until you present me with a source that supports your claim about Turkish, Tatar, Kurdish, Talysh and all other languages. You are contesting the existing opinion, so the burden of proof is on you. If you care about academic accuracy so much, you should know that "I've heard it", "in some cases looks like" and "as far as I am aware" do not constitute an argument. Parishan (talk) 21:11, 17 February 2009 (UTC)

Long history or no, it shouldn't be too difficult to provide a quote. Almost always, the burden is on editors wishing to include information. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:36, 17 February 2009 (UTC)

Thanks for your reply, but with all due respect, we can not use an article on Wikipedia to support a claim made in Wikipedia. Check the Azerbaijani entry for Azerbaijani on Britannica. As you can see while it uses Azerbaijani for the people, it uses Azeri for the language. The problem with the use of the term Azerbaijani here is that the term Azerbaijani has been used as a category for several different dialects, something which is controversial. One example is the language Karapapak, one of many Turkic dialect in Turkey, Azerbaijan, Georgia and as far as Kazakhstan. You can also search the term Azeri on google books and several other databases, and you will see that the term Azeri is very often clarified by adding Turkic or Turkish. It's a clarification, and there is no credible linguist who will deny that.

Parishan in fact, has been busy replacing Turkish and Turkic culture in Turkey and elsewhere to Azeri, my skepticism on everything he adds is comprehensible. When he's criticized, he resorts to accusing other members of being ignorants, here for instance the victim of his slurs was someone who actually specialized on the history of that region (as you can see, just like his accusations toward me, he answers to criticism by accusations of ignorance). Another example is the article Sari Galin, Parishan added then reverted the Azeri term by making the song an 'Azerbaijani', while as testified from Turkish sources, the song comes from Erzerum, and since Azeri is a Turkic dialect, the Turkish term will obviously be translated in Azeri too.

One other example, is the article Lingua franca, claiming Azeri having been a Lingua franca in Turkish Armenia, Kurdistan..., check the history and the number of users he has reverted. Parishan was unable to provide an adequate source to reverse the majority position, which is that it was Turkish, or at most Turkic rather than Azeri. The POV pushing isn't limited to en.wiki, see fr.wiki for example. He replaced Turkish for Kars (city is in Turkey), with Azeri.

Lastly, I won't even bother to reply to this, I'll leave Moreschi or those, who have been involved with content disruptions to handle this. I will not continue here, as any comment on Parishan edits result with this sort of intimidations and accusation of bad faith. And the discussion has become everything but constructive. VartanM (talk) 07:36, 18 February 2009 (UTC)

This is not the place to be having this discussion. As far as the initial issue of whether this sound occurs in the language in question, it's over: we've got a source and your skepticism has been addressed. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 14:33, 18 February 2009 (UTC)

I agree, and I think an academic reference was quite sufficient in light of and compared to VartanM's astoundingly deep divings into linguistics supported by nothing but personal assessments and weasel words. In any event, here is the quote:

"In the modern Azerbaijani alphabet, there is not distinct symbol to mark the voiceless equivalent of the consonant /j/. The sound is represented through the grapheme <k>. For instance, in the words çiçək, çörək, the <k> at the end is not /k/ but in fact the voiceless equivalent of the consonant /j/. For convenience purposes, we will mark this sound as /x'/. <...> However the sound recognised as the voiceless equivalent of the consonant /j/ does not appear in the syllable-initial or mid-syllable position; it can only appear in the syllable-final position. Therefore even if <k> appears at the end of a syllable or a word, it is in fact pronounced as /x'/. In general, it can be concluded that the sound /x'/ is not to be regarded as a fully-developed phoneme." Parishan (talk) 00:46, 18 February 2009 (UTC)

it is funny to see what are the main issues of disagreement here. But to the issue the voiceless palatal fricative k is clearly present in the Azeri dialects of Iran. However they are not present or less developed so in other Turkic dialects of Iran (those in Hamadan, Fars, Kerman, Esfahan). I have heard people claiming that they have heard it among the Shiite people in Turkey bordering Iran and Armenia. These people are in fact related to the Azeris and Iranians in general rather than to Istanbul or Anatolian Turks. I have not heard it in Kurdish, but I would not wonder if it exists. It is basically the same mechanism as new Persian has dropped the archaic Pahlevi final K. I mean the voiceless palatal fricative k in Azeri dialects of Iran is an intermediary stage between k and nothing (like it is the case in Persian). I do not know about Tatar but I doubt it, because Tatar has undergone no West Iranic influence in its sound system--Babakexorramdin (talk) 01:46, 17 February 2009 (UTC)

I mostly agree with you, except about Tatars. The Tatars who moved from Dagestan and settled elsewhere were influenced by non Turkic elements. The intonation in some cases looks like foreign non Turkic. The voiceless palatal fricative k is also present among the Turkic population of Iraq. My revert had more to do with Parishan's attempt to dump every dialect as Azerbaijani (note his inaccurate use of Azerbaijani rather than Azeri). It's not surprising that to support his position he uses an author who claims Iraqi Turkic and several other dialects as 'Azerbaijani'. The author in fact pushes that history as far as 9th to 10th century, when the Azari of that time contained no Turkic elements in the language. As a matter of fact, Iranian Azarbaijan dialects are different than Azeri of the republic. The dialect of Baku as far as I am aware does not have that property. While ironically this foreign property is preserved in Kars and surrounding areas and is not specific to only Shiites. Also note that once more Parishan has used the term Azerbaijani rather than the more linguistically correct term Azeri.

It is for those reasons misleading to claim 'Azerbaijani' 'some dialects,' because this fricative nature of the k, is a property not of Azerbaijani (using that word in this context is wrong in the first place), but characteristic of the people under Iranian and non-Turkic influence. Linguistically speaking the term 'Azerbaijani' is simply false, it is 'Azeri Turkish' or 'Azeric Turkic'... even then, 'some' dialects of a said language do not write the norm. From the same token, Kurdish, Talysh and probably several other languages can be added, because some dialects from those languages have the fricative.--VartanM (talk) 08:11, 17 February 2009 (UTC)

The problem with what you're saying is that "Azerbaijani" is apparantly correct enough that our article on the language is titled Azerbaijani language and at least one linguistic source uses that term as well (cited in the article now). While we can certainly be more specific on which dialects exhibit this feature (though if it's too cumbersome to be specific in the table, we needn't do so here), it is not "silly" nor is it uncommon. What is silly is claiming that we should instead call these dialects "Azeric Turkic" a term that sounds akin to "French Romance" or "Russian Slavic". Also silly is arguing that we should remove Azeri because other Turkic languages/dialects also contain it. If they do, then we can add them as well. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 09:11, 17 February 2009 (UTC)

Aeusoes1, let him be. The man does not know what he is talking about. His primary concern on Wikipedia is to strip it off any mentionings of Azeri. He lacks expertise in linguistics as seen from the fact he has not provided any academic support for any claims he has been bringing forward on this talkpage. Even Ethnologue classifies spoken dialects of the Iraqi Turkmen as dialects of Azeri. If he denies something that can be proven so easily without any reason to do so, do you really expect him to be able to contribute to this article in a valuable and constructive manner? Parishan (talk) 21:11, 17 February 2009 (UTC)

I'm not really interested in doing enough research to really corroborate your claims. He does seem to be POV-pushing but until he tries to troll me personally, I'm fine with dealing with him on a good-faith basis. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:36, 17 February 2009 (UTC)

The original language of Iranian Azerbaijan (and some areas in the South caucasus) has been a Northwestern Iranian language, similar to Talysh. This was called Azeri. Azeri Turkic is then the language which is now spoken in Iranian Azerbaijan. It is named so in order to distinguish between this and the older Azeri language. Azerbaijani Turkic is also used in Iran. This is done in order to distinguish it from other Turkic dialects of Iran. As about the tatars. Tatars speak another branch of Turkic. They are most probably Turkicized east Iranians and were living where they are now. Not in Daghestan. In Kars area the original inhabitants are Shiite Moslems, and there were also Armenians there. This area from Kars to Erzurum was lost by Iran to the Ottoman empire and they did all they could in order to replace the population by Sunni Turks and Kurds. In many cases they succeeded, in the case of Kars they could not. The fact that the editors from the republic of Azerbaijan are using the term "azerbaijani language" the whole time is in order to create the ideal image of one contry, one language, one nation etc...--Babakexorramdin (talk) 12:40, 17 February 2009 (UTC)

"Shoe" or "Hue"

The sound linked sounds like "sh" in "shoe", not like the "h" in "hue", what gives? --TiagoTiago (talk) 03:51, 26 June 2010 (UTC)

It does sound pretty sibilant. This may be because of German. I know some dialects of German have more of an alveolopalatal fricative in ich (with sibilance) than a palatal one and phonetic notation on German (particularly dictionaries) that uses ç may confuse the matter.

Generally, the difference between a voiceless palatal fricative and the h in hue is that the latter is a voiceless approximant but those are so similar acoustically that no language contrasts the two and even trained phoneticians have difficulty with the distinction without use of recording machinery.

So we should probably get another sound, preferably from a speaker of a language other than German. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 07:59, 26 June 2010 (UTC)

I am a native Assamese language speaker. Assamese সীমা which is used as an example here is pronounced not using the sound used in the sound clip provided! Also, in the German pronunciation of Ludwig Van Beethoven ([ˈluːtvɪç fan ˈbeːtˌhoːfn̩]), I (I do not claim it, I only say I heard it like that) the sound used is not like the sound in the sound clip provided (which I hear like sh). The ç used in সীমা and the ç used in Ludwig Van Beethoven sound same to me, but the ç of the audio clip provided do not sound same to me................... What is correct and what is not?? Can anyone help?? IKHazarika (talk) 05:12, 26 April 2016 (UTC)

IKHazarika there is one interesting thing in this article. This one sound completely not like . But if we speak about Kölsch, than it is true. But it is not general German.--Divega (talk) 22:26, 21 January 2017 (UTC)

After some investigation - it is the same sound, just not in typical position (after "a"). So, everything is correct--Divega (talk) 13:45, 22 January 2017 (UTC)

hue

In English, due to yod-coalescence, [sj] and [zj] are replaced with [ʃ] and [ʒ]. Since [ʃ] and [ʒ] are from /sj/ and /zj/, [j] cannot come after [ʃ] and [ʒ]. Also, [ç] is from /hj/ due to yod-coalescence; thus, like [ʃ] and [ʒ], you cannot have [j] after [ç]. This is why hue should be [çuː] not [çjuː]. If hue is [çjuː], then Asia should be [ˈeɪʒjə] instead of [ˈeɪʒə]. --Unnecessary stuff (talk) 21:14, 13 October 2010 (UTC)

Prove it. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 03:43, 14 October 2010 (UTC)

I think my explanation is enough to show why [çj] is not valid, but if you insist so, here. See A huge query from this link. [ç] indeed is from /hj/, and it doesn't have it as [çj]. Then you prove why [çj] is valid. Why do you keep [j] only after [ç] which is also resulted from yod-coalescence, while you don't keep [j] after any other consonants resulted from yod-coalescence (such as [ʃ] and [ʒ])? --Unnecessary stuff (talk) 07:56, 14 October 2010 (UTC)

I understand your claim, but I don't buy it. I'll break it down for you. You say that [ʃ] and [ʒ] from yod-coalescence cannot have a following palatal offglide. There is variation in this, even amongst an individual speaker. What doesn't occur in most dialects of English is a contrast between e.g. [ˈeɪʒə] and [ˈeɪʒjə] where the two mean different things. However, there may be dialects that can contrast [tʃ] and [tʃj], particularly in the United Kingdom.

On top of this, the appearance of a voiceless palatal fricative in the word hue is not one of yod-coalescence (notice that our article on the phenomenon only talks about alveolar > postalveolar?). It is a matter of the /h/, which in English is simply a devoiced version of the preceding sonorant or vowel, being a devoiced [j]. The acoustic contrast between a voiceless palatal approximant and a voiceless palatal fricative is so minimal that there is speaker variation as to which one it is (just as your source says). Since you say [çj] cannot occur in English, the burden is on you to find a source that claims this. The source you've provided contradicts this claim. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:06, 14 October 2010 (UTC)

As a speaker of American English, I can assure you that I most definitely pronounce the word 'hue' as [çj̥uː]. Pronouncing it without the [j] makes it sound British, which I certainly am not. In AE there is absolutely a glide which is neither dropped nor coalesced. Sources to follow. Bearnfæder (talk) 05:34, 15 November 2012 (UTC)

The source Roach (2009:43) does not support the claims made about the English use of [ç]. There is no mention of [j] being dropped. As for the reference above to yod-coalescence involving /s/ and /z/, it should be sufficient to point out that [ç] is not either /s/ or /z/ and that it is also non-phonemic in English; there is no reason to expect [ç] to behave like these other two sounds.Bearnfæder (talk) 23:04, 17 November 2012 (UTC)

I'm writing this while I'm still trying to get my head round the many arguments that I've been reading here. I'd like to clarify what I actually say in my book (referred to above): the phonetic reality, in a word like 'hue', is that most speakers produce a voiceless palatal fricative [ç] followed by the [uː] vowel, so the pronunciation in this case is made up of two segments, [ç] and [uː]. It may sound as if there is a [j] in between these two segments, but I believe this palatal glide is simply the predictable result of the transition from consonant to vowel. However, no phonologist wants to add a /ç/ consonant to the list of English phonemes, so they use the two phoneme symbols /h/ and /j/ instead; this is partly justified on historical grounds. Thus the usual phonemic representation is /hjuː/, with three phonemic symbols representing two phonetic segments. I think this particular phonemic convention misleads a lot of people. I hope to put forward a more detailed account of this with some acoustic evidence soon. RoachPeter (talk) 17:41, 19 November 2012 (UTC)

Well, if it 'sounds' as if there is a [j] in between the two segments, and said [j] is, as you admit, a 'predictable result of the transition from consonant to vowel', then I think that settles the issue: there is definitely a [j] between the [ç] and [uː], not only does it sound as though there is but the [j] is predicted to exist. Therefore, a close transcription should include it, along with mentions of any characteristics it might have (such as being partially voiceless, etc.).Bearnfæder (talk) 13:39, 20 November 2012 (UTC)

Peter is saying that the non-glided version is more common and that our perception of a [j] is due to the underlying representation and to the movement of the articulators. One way to think of this is that there are two things going on, a switch from voiceless to voiced and a shift from palatal to velar (or front to back); the two do not necessarily occur simultaneously, so that the onset of voicing before the shift would create the perception of a stronger onglide and the onset of voicing after the shift would not. The difference is probably something that varies even with an individual speaker in the same sentence. But is something lost if we are just a little imprecise or oversimplistic on this matter? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 14:08, 20 November 2012 (UTC)

The tongue starts out in a position to create a palatal fricative; it must then move down to [uː]; during this transition, it 'passes through' a zone producing something between a palatal fricative and [uː] that can best be described as a palatal approximate—with variations on the voicing quality. There is no way to avoid the [j] except to completely pause the airstream during the transition from the [ç] to the [uː], which is something no native English speaker does, since it sounds ridiculous to do so.

On the same page of the same book, Roach mentions /hw/ creating [ʍ]; when 'which' is produced as [ʍɪʧ] it has every bit the same glide and approximate characteristics as when produced as [wɪʧ]—that is, when the opening of /w/ assumes the voiceless fricative features of the /h/, it does not lose its own features of being an approximate and glide. The same is true of the /j/ in 'hue'; the /j/ isn't lost just because part of it takes on voiceless fricative characteristics of the /h/—it may be slightly altered, but to to say it's lost entirely is ridiculous.Bearnfæder (talk) 14:55, 20 November 2012 (UTC)

With that logic, [çu] would be phonetically impossible. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 15:53, 20 November 2012 (UTC)

Not impossible, just non-English.Bearnfæder (talk) 18:54, 20 November 2012 (UTC)

I'm sorry, but there's a problem in your argument. What you've described, a complete pause in the airstream, would be [çʔu]. I ask you again, what is lost if we don't transcribe the natural transition from consonant to vowel? I say nothing and a phonetics expert seems to agree. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 19:14, 20 November 2012 (UTC)

Vowels don't have to begin with glottal stops. A pause in the airstream does not necessitate a hard start for the vowel. What is lost is that [çjuː] and [çuː] are two different productions; one of them is English, the latter one is not.Bearnfæder (talk) 23:18, 20 November 2012 (UTC)

And we have two trained, reputable phoneticians who say you're wrong about what is and is not English. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 00:36, 21 November 2012 (UTC)

I'd like to hear one of the recordings Mr. Roach used in his analysis that led him to conclude 'hue' is pronounced [çuː]. How about the word 'queue'? Is it pronounced [cçuː]? Show us the source material; back the claim. Let's wait for Roach to get back to us with his acoustic evidence.Bearnfæder (talk) 01:47, 21 November 2012 (UTC)

I’ll try to clear this question up. 1. In all cases where a consonant precedes a vowel, apart possibly from a glottal stop, there will inevitably be an acoustically detectable transition from the consonant to the vowel. In fact, if you edit out the transition and play the resulting sound file to listeners, they will often hear a click where the transition should be, and report the item as sounding artificial. 2. In the case of ‘hue’ that we are discussing, the consonant is [ç] and the vowel is [u]. You could if you wanted label the transition between the two as a palatal approximant (note the spelling – it’s not “approximate”) [j] so that the transcription of the word would be [hju], but the [j] would be vacuous because the transition is entirely predictable from the palatal place of articulation of the fricative. 3. It would be possible to do finer and finer phonetic transcription until you reached the absurd level of writing a discrete label for each cycle of the sound-wave. Everyone would agree that you eventually reach a point of diminishing returns and do not go on slicing the salami thinner and thinner. For me, the suggested [j] element in the transcription of ‘hue’ is not wrong, but simply redundant. 4. I am asked to produce acoustic evidence. I have a load of spectrograms here, but I can’t very well paste them with their audio files into a WP talk page. The evidence is all around you, and what I am saying is just ordinary phonetic theory. Let’s take some data that’s available to anyone – go to www.dictionary.com and listen to ‘hue’, ‘human’, ‘huge’ . You will hear a burst of noise for the [ç] fricative followed by an [u] vowel. You may also hear, or think you hear, a [j]-like transition between the two sounds. You will certainly not hear any [h] sound at the beginning. If you want to check what there is in these recordings in physically real terms, make spectrograms and look at the spectral evidence (there are plenty of programs for this available for free download) for yourself. But the presence of detectable transitions in the acoustic record does not settle the question of whether that transition should be recorded in a transcription – that is a matter of judgement for the phonetician doing the analysis. 5. The word ‘queue’ is a red herring, and not relevant to the topic under discussion. RoachPeter (talk) 09:56, 21 November 2012 (UTC)

1. The transition from [ç] to [u] in no way predictably involves a [j]; in the same way that the transition from [s] to [u] predictably involves a nothing in sew. It is only predictable in the sense that English speakers know the word is /hju/ and that they need to produce a [j] in their transition from their realization of /h/ to their realization of /u/. It is not at all predictable on phonetic grounds, as both [çu:] and [çju:] are realistically possible sound combinations. You give a false impression to people not intimately familiar with the English language when you transcribe the realization of /hju/ as [çu:].

If, as you claim, the [j] is present, then a transcription should contain it, as its presence is not at all predictable—leaving it out is sloppy transcription.

2. When /hw/ becomes [ʍ] its gliding characteristics are not lost; the same is true of when /hj/ becomes [ç]; we are only unfortunately lacking an IPA symbol that can concisely represent the sound realized from /hj/. And as I said already, the transition is not at all predictable; that you claim it is shows only your inability to imagine all the other possible ways for a speaker transition from [ç] to [u].

Your lack of imagination is not convincing evidence of anything other than your lack of imagination.

3. Please look up Wikipedia's page on Reductio ad absurdum. We are talking about one extra symbol that adds immensely to the clarity of the issue being discussed.

4. You don't need to produce acoustic evidence anymore. You've already admitted that a [j] is present; the next step is for you to admit that transcribing it is actually meaningful and necessary.

5. The word queue is entirely relevant as it involves the same process of /h/→[ç] acting on the aspiration. Your refusal to discuss the example only shows how aware you are becoming of your error on this matter.Bearnfæder (talk) 15:41, 2 December 2012 (UTC)

That's enough. Two sources have been provided, both of which say that [çu] is possible for English speakers. You've provided none. All you're doing now is being rude. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 16:01, 2 December 2012 (UTC)

If I want to be rude, I can point out that you clearly haven't read the discussion, as Roach has already admitted to the presence of [j] and so its inclusion on the main page would be only appropriate as it would most satisfactorily settle the dispute. Currently, Roach has only been explaining why he left out [j] in his transcriptions; he has not argued that [j] should be left out—i.e., [çju:] is not incorrect and would even represent a more accurate description of what is going on since, as Roach admits, there is indeed a [j] present.

To further my rudeness, I can ask you how your degree in Creative Writing makes you a more qualified arbitrator on this matter than either my or Roach's degrees in linguistics and backgrounds in phonetics and phonology.

But I don't want to be rude and I am trying not to be; as far as I am concerned Roach and I are two professionals having a professional dispute; we both have reasons for the positions we hold and we are both trying to support our positions with evidence—which at present is just ourselves. The only one I can see who is being rude here is you, as you periodically interject to tell us to stop arguing and just leave things the way they are; but your contributions in this way are completely unproductive and don't actually serve to settle the disputed content.Bearnfæder (talk) 16:42, 2 December 2012 (UTC)

For what it is worth I have most certainly been investigating this matter by examining my own spectrograms, and it is painfully obvious that they involve [j]. My recording of hue shows the fricative followed by a voiced sound with formants at about 300,2350 (corresponding to my production of /j/ or /i/); the second formant slowly transitions to a value of about 1150 by word's end, corresponding to my production of /u/.

And this is why I originally wanted to see the spectrogram evidence as I thought it might shed light on why we were coming to different conclusions about the presence of [j]; but since Roach already admits to the presence of [j], then I suppose there really is no reason for him to bother finding a way to upload anything unless he'd like to compare spectrograms, which might be fun.Bearnfæder (talk) 16:42, 2 December 2012 (UTC)

This conversation began when an editor said that hue could not be [çjuː] but failed to provide sourcing. You make the opposite case, that hue can't be [çuː] and similarly lack sourcing. I believe both are incorrect, as Roach and Wells say that the pronunciation varies.

That you yourself pronounce it [çjuː] doesn't mean that it can't be [çuː] in English. More importantly, even if you and I take a look at the spectrograms that Roach transcribes as [çuː] and even if we both see a [j] element to it, that would still constitute original research. That is verifiability over truth. The sourcing we have says that it's both and, given that fact, there's no need to get bent out of shape over whether the [j] is there or not.

And since you're fairly new here, I'd like to point out that saying what you would say if you wanted to be rude, is being rude. Comment on content, not character. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 18:17, 2 December 2012 (UTC)

I'd like to finish off by taking issue with Bearnfæder's claim to be a fellow professional in this field. To cite just two of his pearls of wisdom from above: "in the same way that the transition from [s] to [u] predictably involves a nothing in sew", and "The word queue is entirely relevant as it involves the same process of /h/→[ç] acting on the aspiration". No first year student of Linguistics would make howlers like this. It would be a waste of my time to take any further part in this "discussion". RoachPeter (talk) 18:31, 2 December 2012 (UTC)

You can fling dishonest personal attacks against me all you want, folks. People who actually do the research will be able to decide whose conclusions stand on the facts and whose are wishful thinking; the readers can find their way here, read the evidence presented, and come to their own conclusions. 'Sew' was obviously a typo, and no one assuming good faith would take it as anything else. Just remember, folks, the world is watching you.

The evidence is out there for anyone who wants to look. Enjoy your 'facts'.Bearnfæder (talk) 19:14, 2 December 2012 (UTC)

hee hee

Whenever people on the net or wikipedia talk about this sound appearing in English, they just talk about h being followed by a y-semivowel. Why not the actual ee-vowel. H before Y sounds just about the same as before EE, since EE and Y are basically the same sound. Some people make this sound when laughing (or to be exact, giggling). And its onomatopoeia is hee-hee. Just like "s" "hiss", "z" "buzz" ,"v" "voom" .

Basically, [hi]/[hj] and [çi]/[çj] mean the same thing and transcribe the same set of sounds. [h] is just, as someone's already said above, "a devoiced version of the preceding sonorant or vowel, being a devoiced [j] The acoustic contrast between a voiceless palatal approximant and a voiceless palatal fricative is so minimal that there is speaker variation as to which one it is". Friction varies freely in the h-sound before [j] or [i], and hardly anyone distinguishes voiceless fricatives and approximants. It's all up to whether you want to be general [h] or specific [ç].— Preceding unsigned comment added by Farazcole (talk • contribs)

Yes, the /h/ in a word like heed is probably similarly [ç]-like as that in hue. I think the difference is that a lot of speakers may also drop the yod so that hue is pronounced [çu] and thus forming a minimal pair with you. It may be the case that it's more constricted in words like hue than in heed — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:10, 18 November 2010 (UTC)

Well, it doesn't really matter at all if you write [hi/hj] in dictionaries/wiktionaries or [çi/çj] because they mean the same thing (respectively). It could be written using [hi/hj] for normal people, but [çi/çj] for geeks.

hiacynt in polish

hiacynt in polish - h is x! —Preceding unsigned comment added by 79.191.66.5 (talk) 19:34, 19 November 2010 (UTC)

It's not [x] (velar), but [xʲ] (post-palatal). Alternative transcriptions include [x̟] and [ç̠]. Peter238 (talk) 20:51, 23 November 2015 (UTC)

In the words "mojito" and "fajita" (both from Spanish), the 'j' is meant to be pronounced [x], but probably it's pronounced [ç]. To an English speaker, they sound like "mohito" and "fahita" respectively, since [h] also sounds like [ç] before "ee" in English. The page itself says that [ç] exists as an allophone of [x] or [h] near front vowels, and that includes English.

voiceless palatal fricative

@Aeusoes1

Is it really necessary to put an English example "hue", because [h] is the voiceless version of what comes after it. You can still transcribe 'hue' as [hju:], even if the actual sound is palatal. And if you still think the 'h' in 'hue' can't be transcribed as [h], then what about 'heat'? —Preceding unsigned comment added by 90.217.215.78 (talk) 20:26, 30 December 2010 (UTC)

It wouldn't be very fair to omit English just because the palatal fricative might be a voiceless approximant for a number of speakers. That's true of most of the languages listed here. The sources I've seen don't say anything about words like heat. There is apparently more constriction for hue than heat. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:44, 30 December 2010 (UTC)

Oh yes. Is this because the [i:] in 'heat' is actually a diphthong: [hɪit]? —Preceding unsigned comment added by 90.217.215.78 (talk) 21:16, 30 December 2010 (UTC)

It's probably because [j] is more constricted than [i]. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 21:21, 30 December 2010 (UTC)

But can you still transcribe "hue" as [hju:], even if you're transcribing someone who pronounces it with a voiceless palatal fricative? —Preceding unsigned comment added by 90.214.72.249 (talk) 17:06, 31 December 2010 (UTC)

It depends on the speaker and how narrow you want to transcribe it. Phonemically, it's most often considered a cluster /hj/. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:11, 31 December 2010 (UTC)

But isn't [h] the voiceless version of the following sound. In [hju:], [h] is the voiceless version of [j], which should be the voiceless palatal approximant. But does any English speaker pronounce it with a fricative? —Preceding unsigned comment added by 90.214.72.249 (talk) 17:19, 31 December 2010 (UTC)

Sorry, I kept deleting the English example because I wasn't sure whether [ç ~ ç˕] should be considered a true allophone of [h]. [h] can sound like [x˕] before [u], [ʊ] and [o] and like [ħ˕] before [ɑ] or [ɒ]. [h] does have a true allophone, [ɦ], because [ɦ] isn't voiceless, while [h] must be voiceless. [ç ~ ç˕], [x˕] and [ħ˕] are all voiceless and can be considered as the sounds which can be transcribed with [h].

What about 'heat'? Is the "h" sound a fricative, approximant, or neither of them? —Preceding unsigned comment added by 90.214.72.249 (talk) 17:37, 31 December 2010 (UTC)

Yes, speakers do pronounce it as an actual fricative, though there is some variation in this. This is what the sources say.

I haven't found anything about other allophones of English /h/ though the ones you posit are certainly likely. I don't know what you mean by "true allophone." An allophone is an allophone. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:52, 31 December 2010 (UTC)

A true allophone is when someone says another sound instead of the original sound. For example, [ŋ] is an allophone of [n] before [g] and [k] in hundreds of languages. However, [n] only represents an alveolar nasal, not a velar nasal. We can't transcribe the 'n' as in 'bank' with [n].

Back to [h]. The [h] sound in "heat" isn't the same as the [h] sound as in "high". The h in heat is actually [ç˕], but it CAN be transcribed as [h], because [h] is the voiceless version of the following sound. I'm not sure whether the [ç] in hue should be considered an allophone of [h] because of the definition of [h] (h isn't a definite sound). Is the "h" sound as in 'heat' a fricative, approximant, or neither of them? —Preceding unsigned comment added by 90.214.72.249 (talk) 23:32, 31 December 2010 (UTC)

It's not a fricative. Whether it's an approximant kind of depends on your definition of approximant. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:47, 31 December 2010 (UTC)

Does anyone pronounce [h] as a fricative before 'ee'? —Preceding unsigned comment added by 90.214.72.249 (talk) 23:48, 31 December 2010 (UTC)

Anyone amongst millions of speakers? Probably, but I haven't seen a source that talks about it. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 14:11, 1 January 2011 (UTC)

Can you transcribe the 'h' in 'heat' as [ç] even when the actual sound is [ç˕]?? I said the h as in hue is [ç˕], but you said that there's so little difference between [ç] and [ç˕], that both sounds can be transcribed as just [ç]. 90.219.106.12 (talk) 20:51, 1 January 2011 (UTC)

I don't think [ç˕] in heat is even accurate for most speakers. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 21:24, 1 January 2011 (UTC)

So is the [h] in heat even weaker than an approximant? 90.219.106.23 (talk) 19:18, 2 January 2011 (UTC)

Probably. It seems like you really want to transcribe the [h] in heat as [ç] but I've never seen any source do that. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:20, 2 January 2011 (UTC)

Why is it [çu:] in the example rather than [çju:]? 90.206.29.100 (talk) 19:00, 3 January 2011 (UTC)

It's apparently the more common pronunciation. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 19:40, 3 January 2011 (UTC)

So, you don't put offglides in phonetic transcriptions? —Preceding unsigned comment added by 90.219.106.4 (talk) 19:05, 4 January 2011 (UTC)

You do, but in this case there are people who pronounce it with an offglide and people who don't. The latter is, according to the sources, more common. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 19:08, 4 January 2011 (UTC)

So how should the h-sound in "heat" be transcribed? —Preceding unsigned comment added by 90.214.72.200 (talk) 17:18, 29 January 2011 (UTC)

For English, I'd put it as [h]. I haven't seen anything else. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 18:03, 29 January 2011 (UTC)

But what about the 'h' as in 'hue'? —Preceding unsigned comment added by 90.206.29.114 (talk) 19:15, 30 January 2011 (UTC)

Hue has the glide because it's followed by the back vowel, no? hue, cue, gubernatorial, pew all hinge on the [u] that follows. Hence no glide in heat because it's followed by a front vowel? Or am I thinking to Scots Gaelic? Akerbeltz (talk) 19:42, 30 January 2011 (UTC)

I'm talking about the 'h' sound as in 'hue'. Should it be transcribed as [h] or [ç]? —Preceding unsigned comment added by 90.206.29.114 (talk) 20:03, 30 January 2011 (UTC)

We've already been over this. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:28, 30 January 2011 (UTC)

I mean, is it OK to transcribe the h-sound as in 'hue' as [h] in dictionaries etc.? — Preceding unsigned comment added by Farazcole (talk • contribs) 13:04, 31 January 2011 (UTC)

Depends entirely on where the transcription's going to be. Even without going into the "etc.", dictionaries have different goals depending on what the lexical scpe and target audience are. Both options have their places. --Trɔpʏliʊm • blah 19:48, 31 January 2011 (UTC)

@aeusoes1: What do you mean by [j] being more constricted than [i]? — Preceding unsigned comment added by Farazcole (talk • contribs) 13:46, 1 February 2011 (UTC)

Have you checked out our articles on semivowels or approximants? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 14:22, 1 February 2011 (UTC)

But why is the pronounciation on the page without the [j]? Without the [j], hue should just be pronounced like 'who'. — Preceding unsigned comment added by Farazcole (talk • contribs) 11:23, 4 February 2011 (UTC)

We've been over this long enough. If you don't get it by now, nothing I say will help you. This is not a general discussion forum and what we're discussing has no impact on this or any other article. Regards. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 12:59, 4 February 2011 (UTC)

Sorry to dig up the corpse of this discussion after such a long time, but looking at the table of occurrences of the voiceless palatal fricative WRT English, it does say that [ç] is an allophone of /h/, and this is a notion that I have never seen elsewhere. It implies that there is no phonological contrast involved in the difference between 'hue' and 'who'. The usual analysis is that [ç] is a segment that serves as the phonetic realization of the phoneme sequence /hj/. Should I change the table? RoachPeter (talk) 15:42, 15 June 2014 (UTC)

incorrect sound

The sound sample is incorrect. It's ɕ, not ç.--Ancient Anomaly (talk) 04:04, 11 January 2011 (UTC)

So let's compare

• ç Voiceless palatal fricative
  ç
• ɕ Voiceless alveolo-palatal fricative
  ɕ

Also: same source. -DePiep (talk) 22:18, 1 February 2011 (UTC)

That is: different files, uploaded by the same editor on the same day. -DePiep (talk) 13:05, 4 February 2011 (UTC)

Yeah, that definitely sounds wrong to my ears too. Too much sibilance. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 13:19, 4 February 2011 (UTC)

I suggest anyone but me can upload a better version. ;-) -DePiep (talk) 14:24, 4 February 2011 (UTC)

Still wrong, a year later... --213.47.74.51 (talk) 20:56, 12 February 2012 (UTC)

I tried to remove it, but it's used in some templates and I can't figure out which ones they are, so I've just submitted a deletion request at the Commons. - M0rphzone (talk) 11:42, 27 November 2012 (UTC)

I recorded the sound. I agree that there's a lot of sibilance, but I have no trouble spotting the difference. Completely removing the sample solves absolutely nothing, though. Unless I beat you to it, just make a better recording.

Peter ^Isotalo 00:45, 29 November 2012 (UTC)

"Unless I beat you to it, just make a better recording." It'd be a good idea to let people overwrite the file then. --Lfdder (talk) 02:11, 12 February 2013 (UTC)

Were you implying that I've actually managed to stop ppl from doing so?

Peter ^Isotalo 23:01, 22 December 2013 (UTC)

No, I didn't know about permission levels at the time. — Lfdder (talk) 12:44, 30 December 2013 (UTC)

Added a link to a recording, to the introduction. I hope my pronunciation is OK, as this is not a sound found in my language, and I just tried to pronounce it right based on the knowledge I have of the articulation. I have no idea how to change the template, though :/. Edralis 22 December 2013 (UTC)

Good work, Edralis. Considering we actually have a female recorder for once, I think it'd be appropriate to keep both files. I'm surprised no one has thought of this before.

Btw, I'm still very unconvinced about claims about "too much" sibilance, especially since there's an obviously difference compare to my recording of the alveolo-palatal sibilant.

Peter ^Isotalo 22:59, 22 December 2013 (UTC)

None of those are unfortunately a palatal fricative. The one by Edralis but a palatalized laminal voiceless alveolar fricative. Try the sound file at wikt:hue for an actual voiceless palatal fricative. --JorisvS (talk) 12:00, 30 December 2013 (UTC)

So what makes you convinced that this particular recording is the genuine article? It's a honest question, btw. All we've done so far is to tell each other that we know best because... we feel we know best.

Peter ^Isotalo 12:29, 30 December 2013 (UTC)

From exposure to languages that have been described to have these sounds, I suppose. I agree with JorisvS's assessment. — Lfdder (talk) 12:32, 30 December 2013 (UTC)

Here's my attempt at it: https://www.dropbox.com/s/k314u6q7x9jltnu/New%20Recording%202.m4a. If people think it's ok, I'll put it on Commons. — Lfdder (talk) 12:38, 30 December 2013 (UTC)

Well, I wouldn't have a hard time claiming familiarity myself (Norwegian, Japanese, German, English), and I'd guess that Edralis would claim something similar. But how do we solve this without simply resorting to claims that our personal judgement is the most correct or *shudder* voting?

Peter ^Isotalo 12:48, 30 December 2013 (UTC)

Maybe by saying that we must include [ç] of a particular language as produced by a native speaker. — Lfdder (talk) 14:03, 30 December 2013 (UTC)

The second one in the file is good, the first one isn't. A palatal is a palatal, but there are a lot of sounds that are in fact not really palatal. --JorisvS (talk) 14:04, 30 December 2013 (UTC)

The first one is louder, but the formants are nearly identical. — Lfdder (talk) 14:21, 30 December 2013 (UTC)

Can anyone tell me what the bandwidth is on these recordings? I can't read .ogg files into my spectrum analyzer, so I can't check, but I suspect (from listening to them) that the upper frequency limit may be around 8kHz, which is too low for accurate assessment of fricative sounds. If so, there's not much point in arguing about whether there is too much "sibilance" or talking about "formants" in the fricative spectrum. As for what should be the method for judging whether a recording contains the "correct sound", we are surely here talking about sounds of the IPA chart, and so looking for a speaker of a particular language that is claimed to use the sound is not as relevant as finding a recording made by someone who would be recognized by the IPA as competent to make the sound, i.e. someone who has undergone a rigorous course of training in practical phonetics. I believe there are quite a few "talking IPA charts" with high-quality recordings made by professional phoneticians such as Peter Ladefoged (former President of the IPA), which could be mined for examples. Having said that, the current recording for the voiceless palatal fricative sounds OK to me. RoachPeter (talk) 17:44, 30 December 2013 (UTC)

The file I linked to is 44k. You can convert it to some other format in Audacity to load it up in Praat. Is the licensing of those recordings compatible with Wikipedia's? — Lfdder (talk) 18:04, 30 December 2013 (UTC)

What RoachPeter said, pretty much. The point is that we shouldn't be trying to analyze sounds ourselves. At the very least we should compare what we have with recordings made by trained phoneticians.

Peter ^Isotalo 02:11, 31 December 2013 (UTC)

how are we gonna compare if not by some kind of analysis? — Lfdder (talk) 03:36, 31 December 2013 (UTC)

"Hue", really?

Is [çuː] perhaps a BrEng pronunciation? I'm from Southern Ontario, and I pronounce it [hjuː]. I can't imagine [çuː] or [çjuː] or whatever in NAmEng, any more than [t͡ʃuːn]. This example can only be confusing to (at least) NAmEng speakers. Curly Turkey ⚞¡gobble!⚟ 12:00, 20 August 2014 (UTC)

It does appear to be so, which is why it's now listed as a BrEng pronunciation. (That's how I say it, incidentally.) Double sharp (talk) 12:54, 28 December 2015 (UTC)

I'd be really astonished if your pronunciation really was [hju:], a syllable containing three segments. That would mean starting with a voiceless glottal fricative [h], then making a voiced palatal approximant [j], then finally a vowel [u]. It's easy to demonstrate by means of spectrograms that English speakers don't make this complicated sequence, but produce just the two segments, a voiceless palatal fricative followed by a vowel. You can hear this pronunciation in any on-line dictionary that gives pronunciations (e.g. Forvo); if you want to see the spectrograms I can put some up for inspection. I'm not saying that there aren't other pronunciations of 'hue' - some English speakers have [ju:], and round my part of the world (Norfolk, UK) we have [hu:]. But mostly, the belief that we say [hju:] is the result of people misinterpreting the phonemic representation /hju:/ which uses /h/ and /j/ combined just for the sake of economy, avoiding the need for a voiceless palatal fricative phoneme. RoachPeter (talk) 15:30, 28 December 2015 (UTC)

(Because I realise I wasn't clear about that: I say [çuː]. Are you replying to me or Curly Turkey? I don't think he would say [hju:] either. It's probably not [çuː], but the three-segment syllable [hju:] is implausible as a phonetic representation of what he actually says.) Double sharp (talk) 15:33, 28 December 2015 (UTC)

Sorry, I thought you meant that you say [hju:]. BTW, where is [çuː] listed as a BrEng-only pronunciation? RoachPeter (talk) 15:38, 28 December 2015 (UTC)

In the table in the article under "Occurrence", where [çuː] for "hue" is listed explicitly under BrEng (and thus implying that it is not a possible AmEng pronunciation). Double sharp (talk) 08:49, 29 December 2015 (UTC)

No, it doesn't imply that. It's just that the source doesn't talk about other dialects in which it'd occur, that's all. Martin sv 85 (talk) 07:35, 11 March 2016 (UTC)

Scots

Doesn't Scots also make use of this sound? From our articles on the subject, it seems like words like licht "light, bright" are usually pronounced with [x], i. e., like in Dutch, but in Southern Scots, the pronunciation does seem to be with [ç], i. e., like in (Northern) Standard German. --Florian Blaschke (talk) 16:39, 5 July 2016 (UTC)

Finnish example

I removed the Finnish example "vihko" for the sake of being completely ridiculous. I am a native Finnish speaker, and I don't know why such an example existed in a first place - either somebody has made an error or just trolling. — Preceding unsigned comment added by 128.214.200.142 (talk) 15:22, 27 March 2017 (UTC)

Please explain. I didn't understand your explanation. For a non native: why is it wrong. Thanks.— Preceding unsigned comment added by 2A01:E35:8B4A:86E0:10A2:2686:3A0F:9B18 (talk) 07:21, 15 February 2018 (UTC)

It wasn't wrong, and it's quite funny that almost all Finns are unaware they have this sound in their language because they've been brainwashed into believing that each Finnish sound is always represented by the same letter and that each letter represents only one sound. So they all really believe they say the same sound in Helsinki, vihko, and Lahti even though the h represents three completely different sounds.

https://en.wiki.x.io/wiki/Finnish_phonology#Consonants

The phoneme /h/ has glottal and fricative allophones. In general, at the end of a syllable it is pronounced as a fricative whose place of articulation is similar to the preceding vowel: velar [x] after a back vowel (/a o u/), palatal [ç] after a high front vowel (/i y/). Between vowels a breathy or murmured /ɦ/ can occur:

• vihma [ˈʋiçmɑ], lyhty [ˈlyçty]
• mahti [ˈmɑxti], kohme [ˈkoxme], tuhka [ˈtuxka]
• maha [ˈmɑɦɑ]

--Espoo (talk) 07:44, 12 October 2019 (UTC)

@Espoo: No need to blame native speakers for being unaware of allophones in their own language. That's more usual than being aware of them, no matter the language. Kbb2 (ex. Mr KEBAB) (talk) 17:30, 12 October 2019 (UTC)

Yes, i know, which is why it's so surprising that nobody reverted the good faith deletion carried out more than two years ago on a page edited by several linguists. And the info i added here will perhaps help the native Finnish speaker who indignantly removed the info and some others understand the specific problem and also realize they shouldn't think they're experts on the analysis and description of other sounds of their own language either. --Espoo (talk) 21:15, 12 October 2019 (UTC)

References

For German

For German, you give the audio «Nicht», to illustrate /ç/, but it would be really useful to give also the audio for «Milch», «ich», as it sounds a bit different (longer). — Preceding unsigned comment added by 2A01:E35:8B4A:86E0:10A2:2686:3A0F:9B18 (talk) 07:20, 15 February 2018 (UTC)

I don't really think it's longer. However, the sequence /lç/ is a bit difficult, so different kinds of assimilations can occur. The most standard form is [mɪlʲç], which may assimilate to [mɪʎç] and be weakened further to something like [mɪi̯ç]. Maybe this is what you hear as "longer". Other speakers insert a [t], thus [mɪltʲç]. 2.201.0.110 (talk) 06:36, 1 October 2024 (UTC)

Greek missing

Greek has palatals [c, ɟ, ç, ʝ] that contrast with velars [k, ɡ, x, ɣ] before /a, o, u/, but in complementary distribution with velars before front vowels /e, i/. --Espoo (talk) 09:15, 12 October 2019 (UTC)

@Espoo: Greek [ç] is post-palatal like the whole "palatal" series. Check the second table under the main one. Kbb2 (ex. Mr KEBAB) (talk) 17:28, 12 October 2019 (UTC)

Standard Chinese example doesn't have ç

I think this is a misplaced copy-paste or similar error. --78.63.237.25 (talk) 18:30, 28 August 2021 (UTC)

Inconsistent Unicode Codepoint/Encoding across Wikipedia

This article has the Voiceless palatal fricative as the precomposed character U+00E7 while Phonetic symbols in Unicode has the codepoint as U+0063 U+0327 ('c' and the combining cedilla character). Is there a particular encoding that is "preferred" for use in the IPA, or should both be noted in both this article and Phonetic symbols in Unicode? Bhbuehler (talk) 03:15, 24 September 2021 (UTC)

They're canonically equivalent, so it shouldn't matter which you use. — kwami (talk) 11:24, 1 October 2024 (UTC)

Open-mid front rounded vowel?

Why does this page have the open-mid front rounded vowel on it? Is this a copying error? Or does the voiceless palatal fricative have something to do with the open-mid front rounded vowel? Potato2357 (talk) 05:14, 23 October 2024 (UTC)

Vandalism. — kwami (talk) 05:28, 23 October 2024 (UTC)

I see. Thank you! Potato2357 (talk) 05:36, 23 October 2024 (UTC)