Talk:Voiceless velar fricative

"Bach" edit
I do not understand the most recent edit: 'the German name "Bach" when it is pronounced wrong (like usually in English).' It seems to me that the voiceless velar fricative is the correct sound for the "ch" in "Bach", but that many English-speakers substitute [k] for [x]. I hope my edit is clear. -- Lesgles 05:06, June 27, 2004 (UTC)

After seeing I wasn't quite correct about some of my English language examples for aspirated vs. unaspirated voiceless consonants, I'm not so sure about my German example either. If someone believes the aspirated/unaspirated distinction I have made here to be incorrect, by all means change it, and if someone with a better knowledge of German phonology can verify that it is correct, please leave a message here to that point. Thanks. CyborgTosser (Only half the battle) 00:53, 31 Jan 2005 (UTC)

It appears that a good amount of material on German phonology on the English Wikipedia is incorrect with respect to the Ach-Laut sound. According to this, it is more often a voiceless uvular fricative, so I will move the example there, and note that this is a dialect variation. CyborgTosser (Only half the battle) 20:46, 19 Mar 2005 (UTC)


 * Most material on German phonology anywhere fails to take the regional variation even within Standard German into account. Pluricentric language and all. David Marjanović 22:35, 24 June 2007 (UTC)

It appears an anonymous user has changed the German example without commenting on it here (as well as changing the example in voiceless uvular fricative). This has been very frustrating as I changed the example to reflect more accurately the reference above and what I was told by a German Wikipedia user. See Talk:Voiceless uvular fricative for discussion. I want to get this right. CyborgTosser (Only half the battle) 00:27, 5 May 2005 (UTC)

I'm a native speaker of both German and Dutch, and as far as I know, the ach g in both languages is the uvular vl fricative. The palatal fricative, x, is what we in the Netherlands refer to as the "zachte g", means "soft g", used in southern Dutch dialects such as Brabants. [oliver lenz]


 * A few versions of German (both dialects and regional versions of the standard) have behind /a/ and . The Alemannic dialects have  as the only realization of /x/. The rest of German lacks  entirely. I'm a native speaker of two such varieties. David Marjanović 22:35, 24 June 2007 (UTC)

Hanukkah example
I'm pretty sure that the 'ch' in Chanukkah is an uvular fricative, not a velar fricative - at least in modern Israeli hebrew... maybe that example should be taken out. (See Voiceless uvular fricative) --Mo-Al 03:43, 26 April 2006 (UTC)

This problem still exists in the "In English" section.--Mo-Al 02:54, 10 June 2006 (UTC)

Then the sound is derived from Hebrew through Yiddish instead of directly from Hebrew. A more appropriate example as a word is chutzpah, a word more distinctly connected to Yiddish culture. This sound is associated with most languages of central and eastern Europe. It is necessary for proper pronunciation of some Dutch, German, Polish, Ukrainian, and Russian place names. Pbrower2a (talk) 05:03, 18 May 2017 (UTC)

Alternate substitution in English
As an Australian, I've noted that in some exceptional cases, /x/ at the end of a word is turned into a instead of a [k]. The most obvious example I can think of is in the name of Australian celebrity David Koch, where his last name is instead pronounced, probably to prevent an unfortunate phonetic clash with "cock". I considered putting this in the article, but I'm not sure whether this is at all worthwhile for the article, as it does seem to be a very rare occurence, and indeed, I'm struggling to come up with any other examples of such a substitution. -- Kirby1024 01:09, 26 April 2006 (UTC)
 * Or perhaps he thought it was French? David Marjanović 22:35, 24 June 2007 (UTC)

I'm glad I've found the name for this. I've always just called it "that consonant I can't pronounce." (So far, it's the only consonant in any language I am convinced I am unable to produce). moink 20:50, 4 June 2006 (UTC)

It is definately the most diabolic consonant that exists. Ragzouken 16:33, 19 August 2006 (UTC)


 * I bet there are lots of consonants out there that you simply don't know… there are really difficult ones out there! :^)


 * it is ten times easier than English dental fricatives. (Maybe this will help you: It is like k, but don't let your tongue touch the soft palate, just get close enough to it to make noise.)88.101.76.122 13:41, 10 August 2007 (UTC)

X
In Russian, the X letter is pronounced as and not x.
 * Wrong. It can be or  depending on the surroundings, and the latter sound is similar to, but it's not the same. David Marjanović 22:35, 24 June 2007 (UTC)

Inaccure .ogg file?
I am Jewish-American, Hebrew being my second, yet semi native toungue, given that i learned it a young age. The speaker in the .ogg file is not pronouncing the hebrew version of the "ch" sound correctly. It is much rougher and sounds like throat clearing, almost.
 * Yep, the Hebrew one is uvular. See Hebrew phonology. David Marjanović 22:35, 24 June 2007 (UTC)

German example
I'm a native speaker myself, and I had never heard of the word "zach" used as an example here. I looked it up in a dictionary and it was noted as a regional, probably dialectal word. Now, I don't want to just edit it to another example, seeing how there seems to be a slight difference betweeen the voiceless velar fricative and the voiceless uvular fricative that in all honesty, I'm not sure I understood correctly. So, I propose using an example that is actually known to German speaker, unless the former only occurs in some german dialects and the typical "throat" sound is the uvular one, in which case I propose to mark is as dialectal with annotations of what dialects and that it is only used there instead of the uvular version. --91.0.68.156 19:31, 4 January 2007 (UTC)


 * It was in my oxford dictionary but as far as I know, any German word that rhymes with "zach" or "pacht" will have the voiceless velar fricative so don't be afraid to change the example word. Æµ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 03:29, 5 January 2007 (UTC)

I'm a German speaker, too, and at least I never say, but always. Maybe it's allophonic in words like "Kachel" for some speakers? Is there any reference that backs this up? &mdash; Sebastian 01:24, 16 March 2007 (UTC)


 * It's regional, say above. I always say.
 * BTW, there is no Standard German word zach, except for a rare surname. There is a word in the Austro-Bavarian dialects which could be spelled this way,, but in Standard German this appears as zäh ( in the south, in the north, roughly). David Marjanović 22:35, 24 June 2007 (UTC)
 * Shouldn't be ? — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 08:23, 29 November 2014 (UTC)

Spanish example
Are you sure the word "ojo" is pronounced with the voiceless velar fricative? I was taught the letter jota makes an /h/ sound when used as a consonant, although I suppose if you speak quickly it might sound like /oxo/. -- Mwalcoff 04:54, 27 February 2007 (UTC)


 * /h/ and /x/ are similiar sounds, especially to speakers with only one or the other in their phonetic inventory. Æµ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 07:11, 27 February 2007 (UTC)


 * It all depends on which kind of Spanish you speak. In Spain itself ojo is . David Marjanović 22:35, 24 June 2007 (UTC)


 * Yes it depends on wich dialect is spoken. /x/ appears in Mexican, Peruvian, Argentinean and north Spain dialects. In other cases as Colombian or Cuban Spanish, J sounds as /h/, but in Madrid Spain it sounds /χ/ —Preceding unsigned comment added by 189.137.241.251 (talk) 00:33, 14 June 2010 (UTC)

Incorrect pronunciation hint
"prepare to say yes. Now, without moving your mouth or tongue, instead of making the 'y' sound, expel a light burst of air-this is the voiceless velar fricative"

Would'nt this be the voiceless palatal fricative? iv changed the entry slightly so that it is more accurate, but if im wrong please correct me.

Non-sibilant
Is there a non-sibilant velar fricative? Is that what [] is? ionas68224|talk|contribs|email 09:34, 16 July 2007 (UTC)

[] has been described as a co-articulation of [] and [], although that has been disputed. 24.235.155.156 01:43, 18 August 2007 (UTC)


 * Dorsal sibilants (as radical and glottal ones) are impossible, unless you count - but they are both dorsal and coronal. That is why velar, uvular, pharyngeal, epiglottal and glottal fricatives are by definition non-sibilant. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 08:22, 29 November 2014 (UTC)

Finnish example
The Finnish example is quite simply incorrect; the voiceless_velar_fricative sound does not appear in the Finnish language. The correct pronunciation for "lahti" is with a voiceless glottal fricative. —Preceding unsigned comment added by 82.181.150.253 (talk) 17:37, 10 October 2007 (UTC)
 * While lahti is indeed a bad example (even if there were frication, this would come off as [laħti]!), [x] as an (optional) allophone is well-attested in more velar environments, such as o_k, u_k. Readding with cite… -- Trɔpʏliʊm • blah 22:50, 28 June 2011 (UTC)

This sound just doesn't exist in Finnish language. There shouldn't be a Finnish example at all. Current example "tuhka" is simply ridiculous. 82.181.244.133 (talk) 18:46, 24 February 2015 (UTC)

Irish example
The broad 'ch' in Irish is an uvular fricative (contrasting with the palatal fricative). Velar fricatives are liable to be conditioned very heavily by context; in Irish inflexion and some contexts result in the palatal fricative, but one does not hear the same variance seen as with the velar variety. —Preceding unsigned comment added by 159.134.221.224 (talk) 10:34, 12 November 2007 (UTC)

Azerbaijani example
The Azeri example is actually a borrowed word from Persian (with the same meaning), even though it's pronounced correctly. Another actual Azeri word should probably be substituted., —Preceding unsigned comment added by Z amirkhosravi (talk • contribs) 02:37, 26 November 2009 (UTC)
 * What's a better example? — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  02:54, 26 November 2009 (UTC)

Arabic
The KH in arab word "Khodhra" is not pronounced the same way as German "nacht". It is much harder, and the soft KH does'nt even exist in any varieties of Arabic that i know. It is always pronounced in the hard way - as in persian Khorram. —Preceding unsigned comment added by 87.51.211.84 (talk) 18:54, 2 December 2007 (UTC)
 * Actually, I think the German example might be wrong. But it is true that the Arabic sound is typically a bit more back in the throat than the German one. Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  20:47, 2 December 2007 (UTC)

This really looks VERY wrong to me!
I'm another German native speaker, and I can tell you: the russian 'x' letter is NOT the same as the "Bach" "ch"! Just hear a German speaker pronunce "Bach" and then listen to a Russian speaker pronunce "Michail" in his native language! The Russian sound sounds much weaker and softer. There IS a difference, and this should be considered. [edit] Just wanted to note too that Russian 'X' letter and Greek 'X' letter are more closely together than Russian and German: compare to Greek 'nyxta' (night). -andy 84.149.103.162 (talk) 23:46, 15 December 2007 (UTC)
 * German ch is uvular after a. This is why the example is not Bach.  Compare German Kuchen with Russian кухарка  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  00:21, 16 December 2007 (UTC)

The Russian sound is like (the 1st) H in German hoch, an [h] sound not the CH sound [x]. —Preceding unsigned comment added by 93.136.209.169 (talk) 13:52, 23 April 2011 (UTC)

In Dutch
I removed this source : apparantly claiming that this sounds appears in Belgian Dutch (or Belgian Dutch alone, not in northern Dutch). This is simply wrong (maybe a wrong quote). I provided a new, accessible source that states that this sounds is "typical" for northern Dutch, just like it is. I corrected it in the table but to that end, I had to delete Verhoeven.--Hooiwind (talk) 15:29, 4 May 2009 (UTC)
 * Hooiwind, can't you provide a different example? As far as I know, Belgians don't have a hard G as it were, like the "northern dialects" are said to have. So, please give a univocal example. It will be appreciated. 81.68.255.36 (talk) 22:59, 24 October 2010 (UTC)

Well nevermind..χ I searched for that one and it's clearer. But still, weird construction this is. I understand you want to be complete, but at least make a note or something then. It's confusing in my opinion and I'm Dutch! :O 81.68.255.36 (talk) 23:04, 24 October 2010 (UTC)

Goed
This is originally a dutch word (colonization) which is still being used frequently in the Netherlands and flemisch Belgium. Maybe this word should be mentioned twice in the table or both countries should me mentioned for this word. 193.190.253.160 (talk) 08:14, 24 March 2010 (UTC)
 * That's normally voiced, though. — Æµ§œš¹  [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 22:11, 24 March 2010 (UTC)

The sound sample
I am a Russian native speaker and I heared the sound sample and can say that this has nothing to do with Russian х in "хвост". Please correct the pronounciation of the sample. I would say that even the sample in Voiceless glottal fricative is much closer to Russian [x].--MathFacts (talk) 18:25, 29 November 2010 (UTC)
 * There may be dialectal variation. All the sources I've seen have it as a velar fricative in standard Russian.  — Æµ§œš¹  [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  18:39, 29 November 2010 (UTC)
 * This is not a dialectal variation. What this dictor pronouns has nothing to do with standard Russian (I live and born in Moscow). I never heared any Russian to say [x] like this dictor. Also note that here the same dictor pronouns a sound which is supposed to be voiced as voiceless, so I think the problem is with the sound samples.--MathFacts (talk) 19:39, 29 November 2010 (UTC)
 * I'm not sure what to tell you. Your impression contradicts linguistic sources; while native speakers can provide valuable expertise in a number of areas related to linguistics, the phonetic particularities of consonants and vowels isn't one of them.  — Æµ§œš¹  [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  21:51, 29 November 2010 (UTC)
 * Is the voice in the sample also taken from a linguistic source so you could claim that any doubt in its correctness contradicts the sources?--MathFacts (talk) 22:57, 29 November 2010 (UTC)
 * I doubt it. You're not really arguing that the sound is incorrect, though.  You're arguing that it doesn't correlate with the Russian sound that you're familiar with.  — Æµ§œš¹  [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  04:52, 30 November 2010 (UTC)
 * Yes. And in this article it is said that the sound should be the same as in Russian "хвост". What we in fact hear is something between "x" and "s". This reminds me Japanese. I also looked at this sample . It is supposed to be a voiced counterpart of Russian "x". While Russian language has no such phoneme, I can easily pronounce the "x" with voice and that has nothing to do whith what is in the sample. What I expect is the sount like "g" but fricative. But I definitely hear something like "z".--MathFacts (talk) 18:07, 30 November 2010 (UTC)
 * Okay, well you're arguing that Russian has a glottal fricative represented by х. This is clearly not the case for everybody, though it would be interesting to find that speakers may debuccalize it to .  This is the point where one looks for sourcing.  If you or I can't find reliable sources that say something to this effect, we shouldn't put it in article space.  — Æµ§œš¹  [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  18:14, 30 November 2010 (UTC)
 * Okay, well you're arguing that Russian has a glottal fricative represented by х - No why did you decide it? What I argue is that what this speaker said has nothing to do with Russian "x". Possibly we need a better example.--MathFacts (talk) 18:19, 30 November 2010 (UTC)
 * You said "the sample in Voiceless glottal fricative is much closer to Russian [x]". If that's not saying the Russian sound is glottal rather than velar, then I don't know what you're arguing.  — Æµ§œš¹  [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  18:33, 30 November 2010 (UTC)
 * What am I arguing is that what this speakers says is farther from Russian 'x' than the samples in other articles.--MathFacts (talk) 18:44, 30 November 2010 (UTC)
 * What the speaker says is the same as the sound file given for the Russian example in this article. — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  22:09, 30 November 2010 (UTC)


 * The voiced velar fricative of definitely is the voiced counterpart of the "x"s in both хвост and хороший. I don't know where you hear the "z" in it, nor is there any "s" pronunciation in the two "x"s. --JorisvS (talk) 16:09, 1 December 2010 (UTC)
 * Well I definitely hear "z"-like sound following the link above. What I hear is something "zza, azza". I certify that the words "хвост" and "хороший" pronounced correctly according the Russian pronounciation.--MathFacts (talk) 22:14, 2 December 2010 (UTC)

MathFacts, no offense but have you considered you may have a hearing problem? I listened to all the above files and they certainly are exactly what they are labelled as. Alternatively, the speakers on your computer might be bad, have you tried a different machine? Akerbeltz (talk) 22:22, 2 December 2010 (UTC)
 * So do you think this sound is voiced ?--MathFacts (talk) 23:14, 4 December 2010 (UTC)
 * Yes, and if you're still in doubt, punch it through formant software. Akerbeltz (talk) 11:31, 5 December 2010 (UTC)
 * Sorry, I gave an incorrect link. I meant this link: http://en.wikipedia.org/wiki/File:Voiced_glottal_fricative.ogg It is supposed to be voiced but it is obviously not.--MathFacts (talk) 18:15, 5 December 2010 (UTC)

Canepari says Russian /x/, is in fact pronounced as [h]. http://venus.unive.it/canipa/pdf/HPr_08_Russian.pdf —Preceding unsigned comment added by 93.136.209.169 (talk) 13:49, 23 April 2011 (UTC)
 * I can't find that claim. What page is the statement on? — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  15:05, 23 April 2011 (UTC)
 * On page 294. That would certainly explain things. It doesn't seem implausible to me, either.
 * I also agree that the sample for a voiced glottal fricative isn't great, especially the second repetition. It is at best very partially voiced. To be fair, I find it really difficult to reproduce that sound myself, despite phonetic training.
 * However, the samples for velar fricatives are clearly not alveolar but indeed velar. As I mentioned above, velar fricatives have a tendency to be post-velar, though, and the samples seem to be truly velar, which may explain why it's possible to mishear them as alveolar, especially when you're used to a "velar" fricative that is really glottal. So if the sample that is supposed to be a voiced glottal fricative sounds like the Russian /x/ to MathFacts, I wouldn't be surprised. --Florian Blaschke (talk) 21:10, 17 June 2011 (UTC)
 * Whoops, I just notice that I was misled by the phonetical symbols used. What the text is really claiming is that the /x/ is an approximant (instead of a fricative), not glottal. It uses non-standard symbols ([h] with diacritics) for voiceless velar and palatal approximants, which it claims are the actual realisations of Russian /x/. However, this doesn't change the situation too much; it still makes Russian /x/ (especially the velar variant) similar to /h/ and explains the confusion. --Florian Blaschke (talk) 21:16, 17 June 2011 (UTC)

doesn't sound like a fricative
in the sound file it sounds like a devoiced trill. I know some russians and they pronounce it as a devoiced trill too. is it possible that this sound has been misidentified? if it was truly a fricative, then it would sound more like an h or an s, not a kh.n same goes for the uvular fricative, i think it's a trill too. —Preceding unsigned comment added by 217.158.134.45 (talk) 12:20, 17 March 2011 (UTC)
 * It's not very likely that this has been misidentified. I believe what you're hearing is air turbulence.  A trill occurs when articulators vibrate, which is not what's going on here. — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  12:48, 17 March 2011 (UTC)
 * It's also possible that the IP is familiar with [x] as an allophone of [ʀ]. I can discern no friction in the sample. It is true, however, that uvular fricatives like [ʁ] can turn out (slightly) trilled, as it is difficult to produce a fricative with the uvula without causing it to vibrate a little. However, for velar fricatives this is much less common. --Florian Blaschke (talk) 20:15, 17 June 2011 (UTC)
 * Listening to the sample again, I agree that there is what sounds like a vibration at the start of the sound, but it is only very slight. If it were stronger, it would almost make the sound resemble an affricate. Keep in mind that velar fricatives are typically articulated somewhat farther back than a typical [k] (between [a] vowels), but not quite as far back as the uvula (velar and uvular fricatives are still clearly distinct phones), so they could be described as "post-velar". In that position, a slight vibration is possible, but still less likely or strong than at the uvular position. --Florian Blaschke (talk) 20:24, 17 June 2011 (UTC)

Old English

 * Copied from here

Per this edit summary: the subpage in my userspace is something I developed as a central location to clearly indicate what consensus is on the format of the tables at pages like voiced velar fricative. If you take a close look at the introductory paragraph of this subpage, you'll find that there is a link to the archived original discussion about the formatting; there are also links at the bottom that link to discussions that amend the prescribed format. Since you've asked where the consensus is, I'm assuming that you missed this.

As I said in a previous edit summary, anyone is welcome to go to Wikipedia talk:WikiProject Phonetics and discuss the issue again, bringing up the merits of a change in this policy/MOS. Since I'm not interested in engaging in an edit war, I'd like to leave it to either you or User:Wareh to remove Old English yourselves. You should at least understand that the "modern spoken languages" stipulation is agreed upon by the community and going against that is going against what others have agreed to. — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 22:41, 20 April 2011 (UTC)


 * What I saw in the linked discussion was you proposing your template (a tiny part of which was the modern spoken languages only rule) and a couple other people not objecting. If there were a MOS page or policy page saying no dead languages, I would certainly bring it up there, but there is not.  This conversation should really be happening on the talk page of the article in question so I hope you don't mind me moving it there. -- Selket Talk 23:42, 20 April 2011 (UTC)
 * Actually, a better place is Wikipedia talk:WikiProject Phonetics, since this affects multiple articles. — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  23:49, 20 April 2011 (UTC)

Excluding information about vowels and consonants in dead languages, which are equally a matter of WP:RS discussion as the same vowels and consonants in living languages, violates WP:NPOV. I can't see how that's not binding on the content question. On the formatting question, well, format it differently (i.e. make one chart for living languages and one chart for dead languages) if you like, but there also doesn't even seem to have been significant discussion about the presentation of dead languages. Even if you show us the consensus to exclude dead languages (I couldn't find it), I repeat that that is an argument for how the occurrence of phones in dead languages is presented and cannot justify keeping the information out of the encyclopedia. Wareh (talk) 23:58, 20 April 2011 (UTC)
 * I don't really want to dwell on this issue of prior consensus too hard, since you two clearly would like to discuss the matter in a way that could form a new consensus (or at least forge a stronger articulation of a justification for the status quo), but a proposal by one editor and approval by others is exactly the sort of thing that overtly establishes consensus. In addition to the discussion I had with User:CyborgTosser over the course of several weeks, minutiae-oriented comments by others like User:J. 'mach' wust ("Wouldn't it be more common to put the meaning between single quotation marks?") and User:Nohat ("please use class="wikitable" in the table headers") clearly show that the details of the proposal were not glossed over.  Moreover, there are numerous editors active within the WikiProject's talk page who, by remaining silent, gave a form of implicit consent to the proposal.  If someone had an issue with "modern standard languages only" they would have brought it up.  While others have, in the past, added dead or artificial languages to these pages, this is the first time anyone has expressed disagreement upon being informed of the style guideline.  Again, there is established consensus, but this doesn't mean we can't form a new one.
 * Now, onto the merits of this particular feature. For me, there are three important issues at play here.


 * Not all dead languages have clarity in their phonetics. Even a language as well-documented as Ancient Greek (by Ancient Greek phoneticians, no less) has a lot of ambiguity about the nature of certain sounds.  It gets even worse when the language itself isn't even attested (e.g. Proto-Indo-European) or for which our linguistic knowledge is guided by linguistic reconstruction (Latin is a good example of this).  The key here, I think, is that the language needs to have been studied and analyzed by the kinds of methods that we currently deem of appropriate quality; namely, that by phoneticians or other kinds of linguists, especially those equipped with sound recording technology. In fact, although Ubykh has been dead for several decades now, there had been enough detailed structural linguistic research on it that it basically counts as a "modern" language.
 * The wording also carries a prohibition against artificial languages. This provides a fairly strong barrier against the inclusion of non-notable conlangs.  In addition to being without actual empirical study (instead, information on such languages is usually abstract and semi-prescriptivist guides), these artificial languages are nearly universally without native speech communities so that any proficient speaker's pronunciation will be guided by their native phonology.  Even Lojban, which has a stated policy of having "phonetic spelling" allows variation in pronunciation for speakers of languages with different phonologies.  Barring conlangs also keeps us from having to parse the distinction between an actual, fully constructed language, and a fictional language that is really just a unique alphabet with a smattering of vocabulary used in a fantasy novel or movie. Understanding this criteria for excluding conlangs should, I think, help guide us in finding exceptions to this (Esperanto being one of them).
 * Finally, as Wareh points out (here and in my talk page), we shouldn't exclude the kind of information under discussion. Having only modern spoken natural languages is really just a prohibition for the tables.  Editors are, of course, free to add prose discussing anything relevant, including historical distribution.  There was a similar issue a few years back about languages like Abkhaz and Ubykh that have nearly a dozen uvular fricatives, which is difficult to represent in the table with the one-example-per-language prohibition. It is, perhaps, hard to see how one might expand on the prose of these articles since most of them are so bare-bones, but take a look at voiceless glottal fricative, Sje-sound, and labiodental nasal, which add some prose about phonetic distribution.  Prose can even go in the occurrence section of articles, such as at voiceless dental plosive.  While it's tempting to do so, pigeonholing information into table form is, in my opinion, not the way to form quality article content.  Take a look at Spanish phonology before and after I converted the information from a table into prose.  Doesn't it look much better?  (It also, I think, made it easier to add additional information to beef up the article).  — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  03:17, 21 April 2011 (UTC)


 * Always eager for compromise (which of course I can't guarantee would be acceptable to anyone else), I'd like to focus on the apparent common ground when you agree, "we shouldn't exclude the kind of information under discussion." You will never find me wanting to sweep the uncertainties of reconstructions, etc., under the rug: my only interest (I hope) is in a complete & detailed encyclopedia reflecting what the reliable sources say.
 * The obvious good-faith follow-up to this is to ask you to agree always to retain the claims about dead languages, with whatever qualification you deem appropriate. If disagreeing editors had moved, instead of removing, content, I doubt there would be such concern.  Let me put this as a direct question: Will you agree to take the trouble to figure out how to retain any dead-language phonological information put in tables somewhere in the article, rather than simply removing it?  This would mostly satisfy me.
 * But frankly, once we're retaining the information, isn't the table the most natural place to put it? In other words, if we bring up the point for centralized discussion & consensus -- "Given that everyone agrees phonological reconstructions of dead languages should be treated in the encyclopedia, should we exclude them from our existing tables, even with appropriate qualifications or disclaimers ("reconstructed," "possibly," "probably" etc.), and list them in some separate-but-equal way?" -- I become skeptical that any normal Wikipedian audience, whatever their disciplinary POV, would end up with exactly what you propose.  Look, on the consensus issue, the bottom line (IMO) is that two editors specifically disagreeing with this aspect of your scheme outweighs a discussion in which the issue was never specifically considered or debated on both sides.
 * In sum, (1) great that we agree the content should be somewhere, because otherwise huge WP:NPOV problem, (2) really accepting #1 means not lazily removing but helping establish the information in the articles, when added by other editors, in a way that meets your approval, (3) given #1 and #2 is there really any better stylistic approach than putting the dead languages in the tables? Wareh (talk) 15:19, 21 April 2011 (UTC)
 * P.S. If you don't agree about the table, I think it's probably a good idea to try to gather editors interested in both living and dead languages to a centralized discussion specifically on this point of how dead languages' phonology is presented. If this apparently conflicts with my "would mostly satisfy me" above, the point is that I need to feel there are sound reasons for the presentation style adopted.  I'm persuadable but not persuaded. You have convinced me that there are some considerations weighing on the side of restricting to living languages in the charts, but the messy issue of creating a separate-but-equal other place for the dead-lang. info is what makes that look unsound in practice. Wareh (talk) 15:35, 21 April 2011 (UTC)
 * I'm inclined to agree with Wareh. I also do like the idea that we not include constructed languages unless there is a good reason to do so.  The thing that makes dead languages important for a page like this is the way they can interact with modern languages.  This sound was in OE but lost except in a few rather peculiar Scottish words.  I was thinking of adding Biblical Hebrew to this page for much the same reason.  -- Selket Talk 19:41, 21 April 2011 (UTC)
 * For anyone else interested, I have started a discussion on the matter of including dead languages in these tables at WikiProject Linguistics. — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 20:04, 10 January 2012 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Voiceless velar fricative. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20131211020607/http://www.kent.ac.uk/secl/ell/staff/amalia-arvaniti/docs/Greek%20Phonetics%20-%20The%20State%20of%20the%20Art.pdf to http://www.kent.ac.uk/secl/ell/staff/amalia-arvaniti/docs/Greek%20Phonetics%20-%20The%20State%20of%20the%20Art.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 01:59, 12 December 2017 (UTC)

Once again, a very obscure word is the Romanian example
The Romanian example is hram - a word that I have definitely never heard, and which is defined as "patronal feast of a church". I have overheard a lot of Romanian in my lifetime, this is a very common phoneme. I am not a native speaker so I won't change the example, but please: Use a word that is not super rare. I'm almost 100% sure a native Romanian speaker could find an example that uses this phoneme that is one or two syllables and is one of the 500 most common words. Fluoborate (talk) 11:35, 22 June 2018 (UTC)

French jota?
"French	jota	[xɔta]	'jota'"

Jota is not a French word! You mean: Spanish. Even for loanwords, this sound is not used in French, and French people are not able to make it naturally. — Preceding unsigned comment added by 87.91.51.235 (talk) 01:55, 14 December 2019 (UTC)

X sound doesn't exist in Nepali or Hindustani (Hindi/Urdu)
This page is about 'X' sound, which is written as "ch" in German or as the Greek letter "χ". That sound clearly doesn't exist in Hindustani and Nepali. The only Indo-Aryan language to have this sound is Assamese.

Use Google translate to translate the English word 'Book' in German language & listen how its pronounced. Google translate too agrees that there's no X sound in Hindustani or Nepali. The example given for Nepali is pronounce as "Sakha" (branch in English). The example of Hindustani is pronounced as "Khushi" (happy in English). So where is the X sound here ? — Preceding unsigned comment added by Tizen03 (talk • contribs) 16:44, 8 September 2020 (UTC)
 * @Tizen03 Actually, the [x] sound is present in Urdu and I've heard Urdu speakers use that a lot (including words like Khushi) so your point that it doesn't occur there isn't correct. I can understand the points about Hindi and maybe Nepali as it isn't quite as common in them, however, the phonology article for Hindi & Urdu (Hindustani phonology) does say that many Hindi speakers do use [x] although some may replace it with [kʰ] while Nepali phonology does mention [x] can occur as an intervocalic allophone of [kʰ]. In any case, I've mentioned here that in Hindi, [x] may be sometimes replaced with [kʰ] while for Nepali, I've mentioned that [x] is an allophone of of [kʰ] so that should solve this problem. Please read those two phonology articles I've mentioned carefully and if you still believe that information is wrong, feel free to discuss the matter in the talk pages of those two phonology articles. Broman178 (talk) 19:34, 29 May 2021 (UTC)
 * The fact that you still state that its not present in Hindustani (Urdu is a Hindustani language after all) despite me giving evidence in my comment does show you either didn't read my comment carefully or just ignored parts of it because I have heard Urdu speakers use [x] in "Khushi" (I have never heard an Urdu speaker use [kʰ] for "Khushi", its usually been [x] while for Hindi I have heard [kʰ] for Khushi). I have no wish to further comment on your points on Hindi & Nepali because you might be correct for them, especially for Nepali, but your assumption that [x] is not present in Urdu is wrong, especially as it is a main phoneme of Urdu (see the Phonology section in the Urdu language article) - I just used Google Translate now between Hindi and Urdu with words like Khushi, Khan and Kharab and while the voice sample in Hindi used [kʰ] (or the kh sound as you might say if you don't want to use IPA terms), the Urdu sample clearly used [x] for all of them. Like I said in my previous comment, it would be better to discuss this issue in the talk pages for Hindustani phonology & Nepali phonology if you still believe its not a phoneme for them (even though there are sources for them in both articles). And I might also point on a side note, it would be very helpful to add your signature onto your comments here (by finding "Sign your posts on talk pages" and clicking the icon) as other editors will know you made those comments. Broman178 (talk) 12:05, 30 May 2021 (UTC)

Georgian
As a native speaker of Georgian I've never heard anyone pronounce "ხ" in "ჯოხი" as a velar fricative, most of the time it is pronounced as an uvular fricative [χ] or maybe slightly fronted [χ˖] when next to a front vowel like /i/ or /ɛ/, the same is true for "ღ" as well which is pronounced as [ʁ~ʁ̟]. 2A01:CB10:65:400:4DCF:2CA7:7201:97D2 (talk) 17:16, 22 May 2022 (UTC)

sound files
.ogg file for voiceless voiceless velar fricative is also the one posted as voiceless velar approximant. 2A00:23C5:BA8A:F401:F517:4F4:EFC2:5E8E (talk) 21:26, 20 October 2023 (UTC)
 * Yes, I'm gonna remove it. Edit: Can't remove it; it's protected. :/ Emdosis (talk) 09:28, 28 May 2024 (UTC)