Is /ɧ/ a sibilant? It sounds like a sibilant to me but I'm not sure. I didn't find any evidence so I really have nothing else to include in my question.

The sj-sound (Swedish: sj-ljudet [ˈɧêːˌjʉːdɛt]) is a voiceless fricative phoneme found in most dialects of the sound system of Swedish. It has a variety of realisations, whose precise phonetic characterisation is a matter of debate, but which usually feature distinct labialization.


Since the question is apparently about the letter ɧ and not a sound of a specific dialect of a language, we have to answer the question first according to some system of standards. It is part of the IPA: to get an authoritative view of what the thing sounds like, go to the International Phonetic Association (webpage) and see what the experts say. Look for ɧ here and play the 4 recordings. What the experts produce seems to satisfactorily fall into the category of "sibilant". You can also put your faith in Wiki and listen to a performance of ɧ, which is not a sibilant. I would be inclined to dismiss Wiki contributions, but the speaker's audio chart is linked in the IPA web page, which may be accidental, or it may be an official sanction. By a majority vote, I'd say that the IPA use is "a sibilant".

However, I have heard a bunch of Swedish people speaking their language and have notes them using a funny quasi-rounded fronted [x] in words like skit, sju, skyr, sjö. You can hear samples on Forvo, which sound very much like the minority pronunciation on the Wiki page. I believe that the producer of the minority pronunciation is a speaker of Swedish, therefore his pronunciation sounds like it does in Swedish. This is similar to the problem that expert renditions of [ħ,ʕ] do not sound like those sounds in Arabic. It's not that the experts are wrong, it's that IPA is a standard against which specific language sounds are measured, so Arabic "ʕ" is most like IPA [ʕ] – IPA [ʕ] is not defined in terms of Arabic, or Swedish, or whatever language may be generally thought to have a certain sound.

The much deeper question is, what is the history of ɧ in the IPA. The letter seems to have been added to the 1947 chart, judging from the IPA's historical chart page. Associated with any symbol should be some phonetic description, in 1947 described simply as combination of ʃ and x. Auditory standards were established in (a few) classes, since the internet was not functioning at the time. One could read Maître Phonétique in the surrounding years to see what discussion there is that leads to the current system: it is doubtful that there are any living participants from the era that could testify on the matter.

  • 1
    As far as I know, the IPA symbol was specifically added to represent the Swedish phoneme, and the description as a “combination of x and ʃ” is highly unfortunate, since such a sound, a postalveolar sibilant with simultaneous velar frication, is not a possible realisation of the phoneme in any variant of Swedish (nor indeed, it seems, attested in any known language at all). In practice, it is something so paradoxical as an IPA symbol with no inherent realisation, which makes its inclusion practical, but questionable. – Janus Bahs Jacquet Feb 3 at 0:07
  • IMO the main problem is the fluidity of the auditory standard in general. It's like having 6 different meter bars that converge at "about a meter". But I doubt that is something that can be changed. – user6726 Feb 3 at 15:30
  • That is obviously a problem with trying to assign any sort of base value to anything auditory – but it seems to me that in general, the IPA does try. Each symbol is supposed to represent a base sound that can be described as accurately as can reasonably be expected by delineating place of articulation and such things, and all uses of that symbol depart from that basis. In the case of /ɧ/, it seems there quite simply is no basis. You can’t describe /ɧ/ in auditory terms at all, because it represents half a dozen very different sounds, from velar to labial, rounded and unrounded, etc. – Janus Bahs Jacquet Feb 3 at 16:00
  • Likewise with ʕ. The lacuna that needs filling is a collection of sound files with examples of sj-ludet in Swedish dialects. If you know of one... – user6726 Feb 3 at 16:56
  • I’m afraid I don’t. I can (roughly) place the different variants on a map, but that’s just from personal experience talking to Swedes over many years… – Janus Bahs Jacquet Feb 3 at 17:05

Given that this is the Linguistics Stack Exchange I am going to assume you are using IPA and IPA notation. The diagonals "/ /" indicate a phonological sound. It doesn't make much sense to ask about the sound of a phoneme, since sound is the territory of phonetics, which is indicated with square brackets "[ ]". I am with whoever made the "3 to 2" comment. It's not a sibilant.

Sibilants, also known as grooved fricatives, have the defining characteristic of being a "fricative articulated with the central line of the tongue slightly grooved." (Cambridge Dictionary of Linguistics, 2013). The defining characteristics of [ɧ] are that it's a voiced glottal fricative, produced deep in the throat. In the case of the Wikipedia quote discussing the grapheme <sj> (phonetically rendered in the word <sj-ljudet> as [ˈɧêːˌjʉːdɛt]) it does state that labialization often occurs. However, this is a secondary-articulation and isn't related to the tongue or its shape.

If you find this to be helpful, I would be most grateful if you marked this as the answer and gave it an upvote. Thanks so much.

  • 2
    Actually if you look at the IPA chart, it describes ɧ as being simultaneous ʃ and x. The voiced glottal fricative is [ɦ]. The use of // vs. [] in the field is highly variable: some people use // for everything, some for underlying form, some for "phonemic" form. Over 50% of phonological theories do maintain that phonological representation have phonetic properties. It is btw not spelled uniformly, cf. skit bra /ɧit brɑ/. – user6726 Sep 5 '20 at 23:41
  • 1
    This is completely incorrect. The sj-sound is not a voiced glottal fricative. It is neither voiced, nor glottal in any of the dialectal variants in Sweden. It is unvoiced everywhere; in the south and centre of Sweden, it has a velar-ish place of articulation combined with a (post)alveolar one, which is usually non-sibilant; in the north, it is mostly a postalveolar or slightly retroflex sibilant with little to no velar coarticulation. – Janus Bahs Jacquet Sep 6 '20 at 6:31
  • My apologies to all that are participating. I missed the bottom hook and mistook it for a different sound, rendering my conclusions irrelevant. If the OP is indeed asking about the varied phonetic renderings of the Sweedish grapheme <sj>, the question should be modified to that effect. If so, I have little to contribute, not knowing much about Sweedish. Replying to user6726, I do of course recognize the strong link between phonetics and phonology. I have not encountered the variation in annotation in my area of expertise. Thanks, from a Stack Exchange newbie who's trying to do it right. – larkale07 Sep 7 '20 at 6:37
  • 3
    If you discover you’ve misread the question and your answer is rendered misleading, you can always choose to delete your answer – that way you don’t risk misleading future visitors, and you also get any reputation points back that you may have lost from downvotes (though you also lose points you’ve gained from upvotes). – Janus Bahs Jacquet Sep 8 '20 at 20:58
  • @larkale07 "Swedish" has only 1 "e". FTFY. – matt Oct 9 '20 at 14:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.