The Swedish phoneme notorious for having its own dedicated IPA symbol, /ɧ/

The Swedish phoneme that is notorious for having its own dedicated IPA symbol, /ɧ/, that is — the so-called sje -sound. What is up with that, anyway?

Historically, this phoneme arises as the palatalized equivalent of regular old /s/, as is probably suggested by the orthography ( sj uk /ɧʉk/ ‘sick’, skj uta /ɧʉta/ ‘to shoot’, stj ärn /ɧæɳ/ ‘star’) — and in standard Finland Swedish it is indeed still found as a humble [ɕ]. But elsewhere across the Swedish dialects, once enroached on by a new /ɕ/ making its entrance (this is the so-called tje -sound, from the palatalization of plain /t/ and /k/; still [tɕ] in Finland Swedish), it proceeded to drift the fuck way back where…

This does not mean that we’re dealing with some unique phonetic outcome. Regular IPA could still deal with whatever you throw at that, provided you listen carefully enough. Official IPA charts do attempt to define [ɧ] as a specific sound, the coarticulated fricative [ʃ͡x]. And yes, that is one possible realization of /ɧ/. Which would be wild enough on its own already, but nope: that is just the beginning. The range of allowed realizations is far wide open, and almost any voiceless fricative that does not get confused with the four stable fricatives of Swedish (/f/, /s/, /ɕ/, /h/) is fair game.

The variation is not really free though, but mostly dialectal and idiolectal. E.g. standard Swedish prefers retroflex [ʂ], perhaps with optional rounding or velarization as an embellishment. This normally also merges with [ʂ] from earlier /ɾs/. Other speakers, though, can e.g. back the sound further into something more guttural. For example, take this lovely rendering of “chance” with a labialized uvular affricate , courtesy of what’s probably the best-known Swedish-language song out there:

[li̹sːna‿ɔ‿læːɾ | misːa‿intɛ‿ q͡χʷ ænsɛn]

This kind of a pronunciation is generally typical of more southern Sweden, often also going together with the existence of a “guttural R” phoneme /ʁ/ (in place of earlier /ɾ/, much as has happened also in the southern neighbors of the language, Danish and German).

Other speakers yet may take other routes of embellishment, such as instead exaggerrating the rounding, and ending up with what’s basically an emphatic labial fricative. My personal favorite sje -sound overall is one of these: [f͇ʷ], the labialized (rounded) labioalveolar fricative, found e.g. in southwestern insular Finland Swedish.

( Labioalveolars are not reported to be phonemic per se anywhere, but they’re reasonably straightforward consonant sounds: fricatives articulated analogously to the more common labiodentals [f], [v], by pulling one’s lower lip inward to the mouth. This time simply further in than usual.)

But even more interestingly, you can get yourself understood also with more unpopular values of /ɧ/ yet. The lateral fricative [ɬ] could work; the pharyngeal fricative [ħ], yep, probably; whistled sibilant [ȿ], definitely.

All this poses certain problems for phonological theory. Within an invidual speaker’s grammar, a value such as like [χ] would probably have to be considered indeed something like [+back]. But the derivation that allows this to be identified with some completely different sound used by some other speaker cannot really fall back on some stable and reasonable phonological underlying value: we would likely have to use some ad hoc feature like [+miscellaneous] (or, if we wanted to follow good naming practices, maybe [+is_sje]) that has no explicit phonetic definition.

It therefore appears necessary to instead think in terms of speaker knowledge of sound correspondences : a Swede using [χ] and a Swede using [ʂ] can very well tell the two sounds apart (making them phonologically distinct), but will be acutely aware of their etymological identity, allowing comprehension. (You say [χʉ:k] where I say [ʂʉːk], etc.) We call this situation a diaphoneme : in essence, a grups of two or more distinct phonemes in distinct varieties, which are regardless historically (”diachronically”) unambiguously connected.

And, admittedly now speculating, this may very well have also driven the fact that /ɧ/ almost never merges with the other fricatives of Swedish. Not even the most outlandish values; if you dropped a consonant like [f͇ʷ] into an unassuming language just hanging out there, you would likely end up with that simplifying to plain [f] pretty soon. But within Swedish though, there is perhaps enough pressure to keep your sje -sound distinct and legible.

So there is definitely grounds for setting up a unique symbol /ɧ/ — even if it only ends up being useful when considering different varieties of Swedish all at once (and it probably does not belong on the International Phonetic Alphabet).

A better question, however, might be why aren’t other similar cases around the world treated in the same fashion…? They definitely exist: For example,Pashto has both a very similar diaphoneme [ʃ] ~ [ʂ] ~ [x] ~ [χ], and also its voiced counterpart [ʒ] ~  [ʐ] ~ [ɣ] ~ [g]. Or, closer by, Finnish has a series of opening diphthongs ie yö uo , which however vary between close-to-mid [ie yø uo] (standard and Eastern Finnish), close-to-open-mid [iɛ yœ uɔ] (many Western dialects), and close-to-open [iæ~iɑ yæ uɑ] (Southwestern dialects).

And while answering that question would take us too far off-topic here, I would think that, to a first approximation, the answer needs to be sought more from an analysis of different linguistic research traditions, than from any real-world substantial differences…