Talk:Swedish phonology/Archive 3

Latest comment: 5 years ago by Jonteemil in topic Förd(ö/u)mmande
Archive 1 Archive 2 Archive 3 Archive 4

The short e

Shouldn't there be an "e" listed among the short vowels as well, although it's comparatively rare? I'd guess most speakers pronounce "ett" (the cardinal) and "ätt" differently. Or is that non-standard?Kallerdis 13:48, 15 May 2007 (UTC)

Different pronunciations for ett and ätt seems to be very rare in modern Swedish. The vowel might be pronounced differently from one dialect to another, but they're still homonyms, not a minimal pair. I'm sure there are dialects that differentiate, but if that's the case, I would like to know which those dialects are.
Peter Isotalo 15:10, 15 May 2007 (UTC)
At least I would distinguish them -- maybe it comes from my native dialect (Western "östgötska"). Wouldn't you distinguish the cardinals ett or en from the determiners (which rhyme with ätt and än)?Kallerdis 14:36, 6 June 2007 (UTC)
I certainly wouldn't. According to Claes Garlén (Svenskans fonologi, Studentliteratur 1988) the merger of short e and short ä is characteristic of the dialects of Finland and eastern Sweden, so presumably the best place to look for dialects maintaining the distinction would be Västsverige. Orcoteuthis 19:13, 15 July 2007 (UTC)
Kallerdis, if you're sure about this, you should record a sample for us as it would make a very nice illustration for the article. Preferably in the context of a sentence; a suggestion is to use a sentence like jag sa XXX igen, a common phrase use by many phoneticians when recording samples for surveys. I've always been curious about which dialects that actually still have both short and long /e/.
Peter Isotalo 20:26, 15 July 2007 (UTC)
I don't make a huge difference, but there certainly is a difference between how I say "ätt" and "ett". (SE Småland here, btw). Heh, I remember from school when dealing with spelling, how there were a section in our textbook for different sounds which were at risk of being misspelled - skj/stj/sk/sch/..., k/kj/tj, o/å and so on. There was also a section about when to use "ä" and when to use "e" which we skipped due to "only people in Stockholm confuses those sounds anyway" :D Of course that's exaggerated, but e vs. ä never seemed to really confuse us, so I think there got to be something which told us when to use which, and I have difficulty seeing what else than pronunciation would accomplish that. \Mike(z) 21:19, 9 September 2007 (UTC)
Well, ok, there are exceptions. verk and värk, I pronounce the same. men and män: it feels like I pronounce them virtually the same, but not exactly. Lacking recording equipment, it's hard for me to tell how much one would hear of that difference, though. However, ätt and ett are definitely distinct. \M
The issue is also whether other speakers can hear the difference or not.
As an immigrant to Sweden, where I've lived for well over 40 years by now, one of which I studied linguistics, I can say that I rarely, or maybe very rarely, hear the difference, but more often one can see the difference in the speaker's face. The lips produce a more rounded figure for short /ä/ and a more smiling position for short /e/.
This does, however, not apply to vowels followed by /r/.
Without any doubt the article ought to be corrected on this point.
My personal experience is from Uppsala, Lund, Ljungby, Norrköping, Jönköping and Motala.
/Henrik Schmidt —Preceding unsigned comment added by 80.251.192.4 (talk) 04:02, 4 December 2007 (UTC)

Plosives

Aren't the plosives unaspirated in Finland Swedish? JdeJ 20:10, 23 September 2007 (UTC)

I don't think so, but many people with Finnish as mother tongue probably have this artefact when speaking Swedish.
Jens Persson (217.211.10.14 (talk) 13:17, 29 March 2008 (UTC))

References and citations

I've marked this page with {{refimprove}} though perhaps {{morefootnotes}} could be better assuming that the sources at the bottom are exhaustive to the sources used (though I'm certain the source for the vowel chart isn't listed). I don't doubt the information, though there aren't many explicit citations backing up the information on the page. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 18:12, 16 June 2008 (UTC)

This is a good opportunity to look over the article. For example, when looking a bit closer, I noticed that a lot of material in the article was the result of the insistence of certain users back in 2005. This includes just most information concerning standard pronunciation. I've edited this section down quite a bit now.
As for the request, I find it difficult to interepret it. What vowel chart is being refered to? More specific examples need to be provided, or it'll just be yet another vague discussion about general citation policies.
Peter Isotalo 17:09, 19 June 2008 (UTC)
The example I gave, the vowel chart image, is from the Handbook of the IPA but HIPA isn't listed at the bottom. Like I said, it could simply be that we just need inline citations if the references at the bottom can be used adequately to source any of the statements in the article. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 19:47, 19 June 2008 (UTC)
HIPA is listed under "References".
Peter Isotalo 06:05, 20 June 2008 (UTC)
Well, if you want to look there. I'll change it to the other template. I guess it's more an issue of inline citations. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 07:47, 20 June 2008 (UTC)
Is Garlén (2004) supposed to be Garlén (1988), Engstrand (2004) or is the reference missing at the end? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 19:29, 20 June 2008 (UTC)
It was the name that was wrong, so it was Engstrand.
You've inserted Harvardref-templates throughout the article, but I'm rather skeptical to the use of these templates myself. They're only really useful when the number of sources starts becoming extremely high. They also inject code into the article that makes it more difficult for non-regulars to understand and edit, and requires more maintenence. Also, the current citations are formatted with a blue link that ends inside of a parenthasis. That's very poor layout. And why do the templates specify that the book was "written in" a certain city? The geographical location is the seat of the publisher, not the location where the book was written.
Peter Isotalo 21:42, 20 June 2008 (UTC)
I don't know the specifics of why the Harvard reference template is the way it is. Harvard referencing is the way I've been doing it at other articles. There are variations on the harvard inline citation, including:
If we are anticipating the article's growth, Harvard referencing would be a good thing, though I guess if you really don't like the Harvard referencing system, we can undo it (though I don't see high maintenence as that compelling a reason to not have HR). — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 21:55, 20 June 2008 (UTC)
My beef is with templates, not any specific referencing standard. The latter can be done just as easily without the former. You're right that maintenence is not a particularly strong argument, but gratuitous use of code is. It makes the editing process much less transparent and adds very little to the article. I don't anticipate that the number of references will grow dramatically in the forseeable future either. And no matter who is at fault for poorly designed templates, they're still poorly designed and should not be used until they have been fixed. I'd appreciate it if you reverted back to the format without the templates.
Peter Isotalo 23:06, 20 June 2008 (UTC)

I've added some citations now. Should I remove the citation template or are there more statements that would require citing?

Peter Isotalo 12:32, 11 July 2008 (UTC)

Technically, yes. Most claims are still uncited. I'll give some examples.
  • "Contrary to the situation with Danish or Finnish, there can't be said to exist any completely uniform nation-wide spoken Standard Swedish. Instead there are several regional standard varieties (acrolects or prestige dialects), i.e. the most intelligible or prestigious forms of spoken Swedish, each within their area."
  • "Sweden actors, singers and TV-personalties are often advised to "neutralize" their dialects by assimilating Central Standard Swedish pronunciation due to the high concentration of political and media power within the Stockholm region."
  • "In Sweden, the high-prestige Central Swedish varieties often go under the name of rikssvenska ("National Swedish"), a term that in Finland indicates all varieties of Standard Swedish spoken in Sweden as opposed to Finland."
  • "As a rule, although not without exceptions, the place of articulation varies between long and short vowels."
  • "Unstressed vowels are always short"
  • "Unstressed /ɛ/ is rendered as [ə] (schwa) in most dialects, and a lowering of vowels is very common before /r/ and the various retroflex assimilations resulting from it."
  • "Various patterns of diphthongs occur in different dialect groups."
It goes on. I don't know enough about Swedish to question any of this (except for the phrase "place of articulation" when it comes to vowels), so I'm not calling for any removals. Just citations. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 19:26, 11 July 2008 (UTC)
I agree with the statements about Standard Swedish, and I'll try to dig up a general source on it. I don't agree with statements about the phonetic details, though. The latter are cited, just not on a sentence-to-sentence (or sub-clause-to-sub-clause) basis. And complaigning about statements like those about the variation in place of articulation among vowels is slightly illogical, since this is really just a comment of the rather obvious facts presented in the vowel chart. I would greatly appreciate if you made an effort to present a more cohesive request to motivate the presence of the tag. I get the feeling that this is turning into just another issue of "cite everything I ask for or the article stays tagged". That usually only leads to more footnotes for people who aren't even interested in verifying them, not better articles.
Peter Isotalo 16:07, 12 July 2008 (UTC)
The phrase "place of articulation" is one used for consonants, not vowels. Although proponents of the one mouth theory would wish to use all the same phrasing for both vowels and consonants, it's disputed whether, for instance, [i] is palatal or alveolar. I suppose this is more an issue of wording than sourcing, since saying something like "short vowels tend to be more open" or something to that effect is indeed obvious.
You say that the others are cited, just not on those particular sentences. Where are they cited? As a reader of the article, I don't know where these statements come from. Are they all from the same source? Is it Engstrand (2004)? Why don't we source each statement?
Your last point is quite perplexing. It sounds like you're saying that, because this is just an encyclopedia and nobody cares, we shouldn't hold up to higher standards of verifiability. I guess I'll simply say that I disagree. I happen to think citations are pretty important and when I expanded upon pages on other phonology articles, I would've greatly appreciated if earlier editors had cited their sources. I've got a section on my user page about the importance of citations. Although it's not specifically addressing this issue, it might help you see my perspective on the matter. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 21:08, 12 July 2008 (UTC)
This is exactly why I've complain about lack of cohesion. For example, the issue of "place of articulation" for the vowels is apparantly an issue of wording, not sourcing, and simply adding a citation won't solve anything. Try motivating yourself when picking random examples instead of making me have to guess what you want.
As for referencing standards, we're only in disagreement about how detailed citations need to be, not that we need them. Please don't imply that I'm against referencing.
Peter Isotalo 08:10, 13 July 2008 (UTC)
There's a sentence attempting to make an obvious statement but so poorly worded that it made me ask "who says that?"; only upon your response did I realize that I should have been asking "how can we reword this?" This is not an uncohesive approach; given the information I had at the time, it seemed at first that it was an uncited claim.
I agree on what we're in disagreement on and I certainly didn't say or imply that you are against referencing at all. I prefer higher standards and you prefer lower standards because you don't think it's worth it to do the extra work.
Let me present a situation to you. Let's say an editor comes along to the article as it is and adds a bit of info right before the sentence "Various patterns of diphthongs occur in different dialect groups." Let's say this sentence is not backed up by Engstrand (2004). Even if they cite it, there will be a great deal of confusion as to what the sources are saying such that one would have to look at the sources themselves to find out where the information; inline citations are designed to reduce that need. Citing each claim will reduce the chances of this happening. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 00:14, 14 July 2008 (UTC)
I don't think it's quite worth the effort to cite each and every statement just so we could hypothetically stop someone from inserting bogus info. It could just as well backfire and look like existing citations support the newly inserted info. Or people could simply use bogus citations, which is especially difficult to check when they're in a relatively obscure language like Swedish. In my view, careful monitoring of articles is the only way to avoid that, rather than growing a forest of footnotes.
Now, could we get back to improving the article? How do you propose that that sentence be reworded?
Peter Isotalo 09:16, 14 July 2008 (UTC)
Actually, the situation I put forth was about a good faith inclusion.
What about "Most short vowels are more open than their corresponding long vowel"? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 17:59, 14 July 2008 (UTC)

endolabial vs exolabial

Hey Peter,

Roundedness states that back rounded vowels tend to be exolabial/pursed, whereas front rounded vowels tend to be endolabial/compressed. That seems right to me, but I'm not too familiar with front rounded vowels. If we take it at face value, that would mean that Swedish u is a typical compressed front rounded vowel, as in French u and German ü, and that it is Swedish y that is unusual in being pursed. If this is true, then we need to revamp the front rounded vowel articles. Also, how does this effect the Swedish and Norwegian central rounded vowels? Are they still atypical in being compressed, or would pursed central vowels be atypical? kwami (talk) 18:51, 31 October 2008 (UTC)

Dental or alveolar?

Are the Swedish /t/ and /d/ truly dental, as the article says, or are they alveolar? —Angr 15:33, 9 January 2009 (UTC)

jord, /juːrd/, "soil, earth"

Why do you insist on having the non-assimilated pronunciation /rd/ in the examples table even though he clearly pronounces it as /ɖ/? Mindcooler (talk) 12:14, 23 July 2009 (UTC)

Because it's a phonemic transcription, not a phonetic one. +Angr 12:35, 23 July 2009 (UTC)
Also, "he" who? Most sound examples I've seen don't reeally match the transcription to begin with, and that's in features more obvious (to me) than /rd/ vs /ɖ/... --LjL (talk) 13:04, 23 July 2009 (UTC)
"He" is me, the one who recorded the samples for this article.
/ɖ/ does not have phonemic status in Swedish. This is explained in the second paragraph below "/r/-realizations and the retroflexes". Since the article treats Swedish phonology in general, the transcriptions have to follow phonological rules, and no analysis of Swedish that I've seen presents any retroflexes as phonemes. A transcription of the actual realization of my particular dialect can be added, but then it'd have to specified as being phonetic and written in brackets. In this case it would probably be something like [ʝuːɖ].
If there are any transcription errors currently specific examples would be very useful.
Peter Isotalo 16:39, 23 July 2009 (UTC)
I didn't claim there were. While I'd find a phonetic transcription of your speech interesting to have in principle, I also fear it's fall into WP:Original research, since I've often found just which sounds one is uttering is as widely subjective a thing as anything can be... --LjL (talk) 16:41, 23 July 2009 (UTC)
This doesn't need to be a guessing game if we just stick to what has actually been described concerning Central Swedish. For example, in the IPA handbook, Englund bases his analysis on his own dialect which is basically the same as my own.
Peter Isotalo 04:55, 25 July 2009 (UTC)

Hey Peter, can you tell if the photo at roundedness has it right? Can you identify which vowel (i, y, u) is which? Thanks, kwami (talk) 11:20, 25 July 2009 (UTC)

Indeed, the description of Standard Swedish phonology is clearly biased in this article. Contrary what Peter Isotalo claims, the retroflex consonants are phonemic just like their respective non-retroflex counterparts. Thus, "/ɖ/ does not have phonemic status in Swedish" is a flase statement. The words herde 'cowherd, shepherd' and Hede - only differentiated by the retroflexness of the second consonant - constitute a minimal pair claiming this undisputed fact. Of course, in Southern Sweden and in Finland-Swedish this is not the case, but those are dialects, not the Standard Swedish spoken by 85% of the Swedish speaking population. // JiPe (81.226.216.67 (talk) 19:55, 19 September 2009 (UTC))

Why don't you find us a source that actually considers the /r/-assimilations to be phonemes in their own right instead of flinging around annoying accusations?
Peter Isotalo 06:59, 20 September 2009 (UTC)

b, d, g as approximants?

In the first table under "Consonants", /b d g/ are shown as having approximant realizations. That's news to at least this native speaker, and if true really ought be expanded upon in the sections about allophony. Orcoteuthis 19:32, 15 July 2007 (UTC)

Being a native speaker rarely means that one is full aware of the exact phonetic gestures of one's own language, especially for adults who are heavily influenced by orthography. It's actually almost easier to notice these things as a non-native. In general, though, approximation of plosives is a very common feature of many languages and in Swedish occurs colloquial speech. In this case it's taken straight out of Engstrand's Fonetikens grunder.
Peter Isotalo 20:22, 15 July 2007 (UTC)
It should still be expanded on in the article's prose. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 21:04, 7 April 2008 (UTC)
Quote please!--91.148.159.4 (talk) 20:37, 1 February 2010 (UTC)

Audio samples

The man doing the pronunciation samples has an almost extreme Stockholm dialect. One should know that unlike many other languages, the Swedish that is spoken in the capital is distinct from what is considered Standard or High Swedish. As such, I think that this should either be stressed or there be made available other samples. 213.112.137.175 17:01, 14 August 2007 (UTC)

All pronunciations below are made by a Swedish male, age 25, in a variety of Central Standard Swedish spoken in the greater Stockholm region. I think that's clear enough. And if you think that's an extreme Stockholm dialect, you can't possibly know much about how people in Stockholm speak. I should also point out that what most Swedes refer to as rikssvenska (one term for Standard Swedish) is actually just the more neutralized Stockholm dialects, rather than some bogus "neutral" Swedish. Linguists tend to prefer the term standardsvenska and include all major regional dialects in the concept of Standard Swedish.
Peter Isotalo 17:52, 14 August 2007 (UTC)
You live in Stockholm yourself, no wonder that you don't notice. As for what Standard Swedish is, try a pre-war Swedish film, for example Hets. Since the teacher speaks Standard Swedish and some of the students, on the other hand, speak with a Stockholm accent, there is ample contrast to be heard between them. That Standard Swedish is merely a neutral form of today's Stockholm dialect is nonsense.
And no, the "warning" is not clear enough, as, firstly, it contains the phrase "Central Standard Swedish" and presents the samples as done in a mere variety of it, while we are in fact speaking of a dialect exclusive to Stockholm, and secondly because foreign readers very easily will assume that the dialect spoken in the capital is the neutral or prestige dialect. Therefore I am reverting your edit until additional samples have been submitted to which the current ones can be compared. 213.112.137.175 15:20, 15 August 2007 (UTC)
There is no dialect neutral Swedish. In fact, there are no neutral prestige dialects whatsoever; they are all influenced by one form of geographical vicinity or another. What most people, especially people without any knowledge of linguistics, perceive as rikssvenska is actually just a rather standard Central Swedish regional dialect. This is rather easily proved by simply trying to find dialects outside of this area that actually sound the same. However, it is extremely common for people who live Stockholm to believe that their own speech is the epithome of Standard Swedish as long as they don't sound like the stereotypical 45-year-old cab driver from Södermalm.
Please read the article Standard Swedish to get your bearings on the topic. If you don't trust that, check out the article rikssvenska in Nationalencyklopedin. If you want to insist on these type of disclaimers, you better back up your claims with sources, since this article is rather well-referenced.
Peter Isotalo 16:28, 15 August 2007 (UTC)


This article in its current stand has several good points, but its main failure is that it seems to be written by people who overestimate the importance and/or dominance of the Stockholm-Uppsala varieties, but at the same time some paragraphs correctly point out that this language, despite its small size, has several localized high status varieties. Particularly compared to nearby languages such as French, British English and European Castillian, this is a peculiar feature, although not totally different from the situation for German and Dutch?

Much of this article's content, and particularly the audio samples, whould fit better into an article on "Central Standard Swedish phonology".

/Henrik Schmidt —Preceding unsigned comment added by 80.251.192.4 (talk) 04:19, 4 December 2007 (UTC)

I recommend that this issue is addressed by expanding the information and recording more samples.
Peter Isotalo 13:33, 11 December 2007 (UTC)
I completely agree with Henrik Schmidt here. The phonology describes Central Standard Swedish (the dialect spoken by "educated" people in the Mälardalen region), and the audio files represent some middle-class variety of Standard Stockholmish. Living in Jämtland, I can accept the Central Standard Swedish phonology description, but the awkward audio samples seem out-of-place in a serious article on Swedish phonology. Maybe they belong to an article on middle-class white Standard Stockhomish?
Jens Persson (217.211.10.14 (talk) 13:15, 29 March 2008 (UTC))
Standard Stockholmish is one form of standard Swedish and it is certainly the predominant form of Central Standard Swedish. I totally don't see why it would be less appropriate for an illustration of Standard Swedish in a "serious article" than any other regional standard. "Standard Teacher's Swedish" as found in pre-war films such as Hets is not a better candidate, it is terribly old-fashioned nowadays and only remains of it linger on (least of all in Central Sweden). Also, like it or not, Stockholm Swedish is in fact the single most influential variety, even though "refined" regional standards do hold their stand.--91.148.159.4 (talk) 20:36, 1 February 2010 (UTC)

Just a small note here on an audio sample that I believe need re-recording. The "västkustskt" recording is cut off at the beginning, and sounds more like "ästkustskt". 83.255.186.163 (talk) 11:38, 27 December 2007 (UTC)

Ugh... I can't believe they still haven't fixed this problem. The audio file is fine, but your player for some reason chooses to truncate the beginning of it. Try using Quicktime or Winamp instead.
Peter Isotalo 11:49, 27 December 2007 (UTC)


  • Ahem, shouldn't pronunciation and transcription match? That's certainly not the case with hel, /heːl/, "whole" —Preceding unsigned comment added by Fnugh (talkcontribs) 12:57, 27 April 2009 (UTC)
Say what? 惑乱 Wakuran (talk) 16:21, 6 February 2010 (UTC)

Unique consonant sound?

It the voiceless palatal-velar fricative really that unique? I think it exists in at least some Norwegian dialects. I know a Norwegian who has no problem pronouncing it. She claims that the voiceless palatal-velar fricative also exists in Norwegian. She also denies that Swedish has any language sounds missing in Norwegian. The voiceless palatal-velar fricative probably exists in the Serbo-Croatian languages (Serbian, Croatian and Bosnian) judging from Swedish journalist Christian Palme's description of their version of the Latin alphabet. This sound may also exist in Italian through my dad's pronunciation of “gelateria” (“ice cream parlor”) may not be entirely accurate.

2009-03-20 Lena Synnerholm, Märsta, Sweden. —Preceding unsigned comment added by 212.247.167.70 (talk) 20:05, 20 March 2009 (UTC)

You're not confusing this sound with the voiceless alveolopalatal fricative are you? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 22:07, 20 March 2009 (UTC)

No, I am very aware of the difference since the two are separate phonemes in Swedish.

2009-03-29 Lena Synnerholm, Märsta, Sweden. —Preceding unsigned comment added by 213.114.151.101 (talk) 15:59, 29 March 2009 (UTC)

Wouldn't surprise me if it's found in Norwegian, though if it's in one of the standard languages we should mention that. I'm skeptical of Serbian, though. That would have been noticed. kwami (talk) 06:36, 30 March 2009 (UTC)

Alright, I remembered my dad’s pronunciation of “gelateria” incorrectly. (“Gelateria” means “ice cream parlour” in Italian.) He pronounces it ɕɛlateriːa, not ɧɛlateriːa as I thought. So Italian should be removed from the list. If the voiceless palatal-velar fricative does not exist in Serbian how do you pronounce the Š in Milošević?

2009-05-17 Lena Synnerholm, Märsta, Sweden. —Preceding unsigned comment added by 213.114.144.67 (talk) 15:25, 17 May 2009 (UTC)

Š in Milošević: Postalveolar, though whether truly ʃ or ʂ as in Russian I don't know. kwami (talk) 15:47, 17 May 2009 (UTC)
It's sounds very similar to the Russian sounds as I recall.
Peter Isotalo 05:44, 18 May 2009 (UTC)
We cheat and call him "Milåsevittj" [miˈlo:sevitʃ]. Although I've heard [miloˈse:vitʃ] and [miloˈʃe:vitʃ], prob. by someone who know some kind of proper pronuciation. Rursus dixit. (mbork3!) 16:32, 7 February 2010 (UTC)

The Swedish journalist Christian Palme has written a book named “Om ondskan i vår tid” (“On the evilness in our time” in Swedish). His introduction ends with a list on how to pronounce the letters Č, Ć, Š, Ž and C in Croatian and Bosnian. Š is described as the initial sound in “skymning” (“dusk”) which is the voiceless palatal-velar fricative. Anyone who has any objection?

2009-08-28 Lena Synnerholm, Märsta, Sweden.

Palme almost certainly meant the other widespread Swedish pronunciation of the consonant in "skymning", the one used in Northern Swedish, sometimes in Finland Swedish and also by some speakers in Central Swedish, also heard on older recordings. I.e. something like [ʂ], the same consonant as in "kurs". Certainly not /ɧ/, I have to concur with the others in the following section that Serbian Š sounds nothing like /ɧ/, while it is pretty similar to Russian and Swedish [ʂ]. --91.148.159.4 (talk) 20:08, 1 February 2010 (UTC)

Thank you for the explanation. Anyone who has any objection against the voiceless palatal-velar fricative existing in Bosnian or Croatian?

2010-03-08 Lena Synnerholm, Märsta, Sweden. —Preceding unsigned comment added by 212.247.167.71 (talk) 18:58, 8 March 2010 (UTC)

I do. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 21:46, 8 March 2010 (UTC)
So do I, as should have been clear from what I had written.--91.148.159.4 (talk) 22:11, 31 March 2010 (UTC)

ʂ

How do you pronounce “ʂ”?

2009-05-23 Lena Synnerholm, Märsta, Sweden. —Preceding unsigned comment added by 212.247.167.70 (talk) 20:21, 23 May 2009 (UTC)

It's a voiceless retroflex fricative, very similar to sh in English shot, though the shape that the tongue makes is different. Take a look at retroflex consonant. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:33, 23 May 2009 (UTC)
I'm skeptical about the supposed pronunciation of the /rs/-assimilations, though I have seen no academic discussion about it. When /r/ merges with /t, d, l, n/ the result is without doubt retroflex. It's clearly audible and if you have experience with phonetics and have a knack for pronunication, you can clearly feel the retroflex action of the tongue. /rs/, however, doesn't seem to be much different from your average [ʃ], and I'd say it's almost identical to the Standard German pronunciation of the fricative in Schwalbe or schon. I've studied a bit of Mandarin, so I'm familiar with the quite distinct retroflex fricative there, and it really doesn't sound all that similar to the Swedish /rs/-sound.
Peter Isotalo 12:55, 25 May 2009 (UTC)
Hmm. When I watched My Life as a Dog when I heard the boy say the name "Tarzan", it sounded to me very much like [taʐan], with a quite Indic-like retroflex sound very different from the usual English [ʒ]. +Angr 13:17, 25 May 2009 (UTC)
There are several types of retroflex consonants, so it could very well be that Peter is cued to tell the difference between the Hindi-like retroflex of Mandarin and the Tamil-like retroflex of... Tamil, I guess. Swedish could very well have the latter, though I can barely tell the difference between retroflex and a postalveolar. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 15:11, 25 May 2009 (UTC)

If “ʂ” is pronunced more like “sh” how do you pronunce the Ž in Karadžić?

2009-08-16 Lena Synnerholm, Märsta, Sweden. —Preceding unsigned comment added by 212.247.167.70 (talk) 18:01, 16 August 2009 (UTC)

[kâradʒitɕ]. Similar, except that it's voiced.
Peter Isotalo 18:31, 16 August 2009 (UTC)

Yes, it was him I thought about. Apparently, there are several similar sounds which I have hard to tell apart if I only have written descriptions. But if I heard them I would have no problem tell the differences between them.

2009-08-23 Lena Synnerholm, Märsta, Sweden.

Standard?

Section Standard pronunciation carefully explains:

Contrary to the situation with Danish or Finnish, there is not a uniform nation-wide spoken Standard Swedish. Instead there are several regional standard varieties (acrolects or prestige dialects), i.e. the most intelligible or prestigious forms of spoken Swedish, each within their area. Within Sweden, actors, singers and TV-personalties are often advised to "neutralize" their dialects by assimilating Central Standard Swedish pronunciation.

So the section "Standard pronunciation" says: "there is no Standard pronunciation", but X and Y "neutralize by attaining Central Standard pronunciation". Nice, except according to my world view construction rule codex, contradictions are ERROR! Whassup really? Rursus dixit. (mbork3!) 16:15, 7 February 2010 (UTC)

What is "Southern Standard Swedish"? Never ever heard about it. Is it "rikssvenska med sydlig accent"? Then it should be "Standard Swedish with Southern Accent". Rursus dixit. (mbork3!) 16:37, 7 February 2010 (UTC)
Sorry, I mean "Standard Swedish with southern accent" (note case!) Rursus dixit. (mbork3!) 16:39, 7 February 2010 (UTC)
There's zero sourcing in that section, so I don't know how accurate it is. It seems from the context that the section is mixing up "dialect" and "accent". How do you suggest we change it? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 04:20, 8 February 2010 (UTC)
I think we use "prestige dialects" for non-standard Standards. Rursus dixit. (mbork3!) 19:56, 13 May 2010 (UTC)
I fixed the section. It still needs citations, but to me it doesn't feel far from truth, except I think really finding sources will be a tough one. Rursus dixit. (mbork3!) 20:04, 13 May 2010 (UTC)

ʏ̹β not ʉβ

Pronouncing u: as [ʉβ], seem very weird to me. Instead I believe u: is really pronounced [ʏ̹β], where ʏ̹ is an overrounded ʏ. Rursus dixit. (mbork3!) 20:10, 13 May 2010 (UTC)

The primary distinction between /ʉ/ and [ʏ] is the type of lip-rounding, not necessarily the vowel height or backness. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 21:40, 13 May 2010 (UTC)
I mean that [ʏ̹β] or maybe [y̹β] for u: have the compressed type of roundedness, while the [yɥ] for y: has the protuded kind of roundedness.
I think I've seen the habit for using [ʉ:] for long [u:] before, then maybe as a front compressedly lip-rounded [ʏ], but short u is a central vowel protudedly lip-rounded vowel. Why then using the same phonetical symbol for two very different sounds? Rursus dixit. (mbork3!) 22:06, 13 May 2010 (UTC)
Yes, "ʉ" is an unfortunate convention. I like Ladefoged's use of the small cursive w sub-diacritic for w-like protruded labialization (without velarization); if [ʏ] is generally compressed in other languages, then [ʏβ̞] could be used for ut and [ʏ̫] for ylle.
It's odd that, given the huge influence that Swedish has had on the design of the IPA, that there is no fixed diacritic for this distinction! — kwami (talk) 22:08, 13 May 2010 (UTC)
But the phonetical symbol originally used for u: is not yet implemented in IPA, I can paint it here:
***       ***
 **        **
 **        **
 **   **   **
  ** **** **** 
   ***  ***  ***
I don't quite understand why it is not yet implemented. Rursus dixit. (mbork3!) 22:18, 13 May 2010 (UTC)
Is this a recent proposal? Part of the problem might be that if German ue and French u are also compressed, then this letter would need to be used for them as well, conflicting with <y>, which is quite entrenched. — kwami (talk) 22:31, 13 May 2010 (UTC)

I suspect part of the reason is that even though /ʉ:/ can be phonetically front, it "acts" phonologically as a non-front vowel like its short counterpart, which is central both phonetically and phonologically (it doesn't cause palatalization and fricativization of /g/ and /k/). Also, I'm not sure it's really front ([ʏ]) in everybody's pronunciation.--91.148.159.4 (talk) 17:08, 2 July 2010 (UTC)

Too few vowels

Standard Swedish actually have at least 19 vowels: 16 pure vowels, two half-vowels and one diphthong. I have made a list of them in alphabetical order.

1. A (long): pronounced as “a” in “father”.

2. A (short): pronounced as “u” in “under”.

3. au: pronounced as “ow” in “owl”.

4. E (long): pronounced as the first “e” in “here”.

5. E (short): pronounced as “e” in “red”.

6. I (long): pronounced as “ea” in “read”.

7. I (short): pronounced as “i” in “wit”.

8. O (long): pronounced as “o” in “open”.

9. O (short): pronounced as “u” in “put”.

10. U (long): the close central rounded vowel pronounced long.

11. U (short): the close central rounded vowel pronounced short.

12. Y (long): pronounced as “ue” in French “rue” (“street”).

13. Y (short): pronounced as “u” in French “Gustave”.

14. Å (long): pronounced as “a” in “call”.

15. Å (short): pronounced as “o” in “corn”.

16. Ä (long): pronounced as “a” in “dad”.

17. Ä (short): pronounced as “a” in “hat”.

18. Ö (long): pronounced as “i” in “girl”.

19. Ö (short): pronounced as “a” in “about”.

Please tell me if any of my descriptions of the sounds are wrong!

2007-02-17 Lena Synnerholm, Märsta, Sweden.

Standard Swedish does not have phonemic diphtongs. In the southern dialects long vowels are very marked diphtongs, but they are limited primarily to Skåne and the surrounding areas. The dialects around Stockholm also have diphthongized long vowels, but like with the souther diphtongs, they are not used in the rest of the country, and they are not as noticeable as the southern ones. But these two are separate dialect areas and don't really coincide. The instance of the use of an /au/-diphthong is limited to a few fairly recent loanwords, like aula, paus and foul ("auditorium", "pause" and "foul" respectively), so saying that Standard Swedish has diphthongs isn't really accurate.
So that brings the total down to 18, and short "e" and "ä" coincide in pretty much all dialects (even if they can be pronounced differently from one region to another). Very few people in Swedish actually distinguish män "men" men "but" or värk "pain" and verk "work, act".
As for the number of phonemes in Standard Swedish, it's 17 for pretty much all of Sweden. To my knowledge, there's not much dispute about this in the linguistic community as far as I know, even if it's common to claim that there's still a 9-vowel system with long and short pairs being used. This description is historically accurate and it's clearly visible in the spelling of many words, but it no longer accurately describes how modern Swedish is actually spoken. And the existence of two semi-vowels in Swedish is definitely new to me. It's not in any of the major phonologies.
If you're still unsure about this, I recommend checking the references provided at the bottom of the article.
Peter Isotalo 18:32, 17 February 2007 (UTC)

":So that brings the total down to 18, and short "e" and "ä" coincide in pretty much all dialects "

Depends what you mean with "pretty much all dialects". They are separate in most of Småland (I myself pronounce "rett" and "rätt" differently) and from what I can tell much of Norrland as well. In Svenska Språknämndens "Svenska skrivregler" from 2000 they transcribe short e as "e" and simply notes that the vowel has merged with ɛ in most of eastern Sweden. /Viktor —Preceding unsigned comment added by 83.254.249.46 (talk) 23:19, 28 April 2008 (UTC)

Yeah, the analysis seems based on the ortography rather than the spoken language. The distinction between "pure vowel" and "half-wovel" in particular, looks strange. The article Labial-palatal approximant mentions "Swedish: yla [yɥla] "howl" although that might be secondary. 惑乱 分からん 02:54, 20 February 2007 (UTC)
That [yɥ] is a not uncommon realization of /y:/. There are analogous realizations for all the long high vowels.
Peter Isotalo wrote:
"Very few people in Swedish actually distinguish män "men" men "but" or värk "pain" and verk "work, act"."
Modern Swedish orthography isn't really 100% relevant on this. The short e vs short ä in western dialects is based on what the vowel quality and consonant quantity was in Old Swedish. The words verk and värk have the same vowel from an etymological point of view (OSwe verk [vɛrk] for both). In any case, the r will complicate things in your example. The same goes for män and men, they had the same vowel in Old Swedish (menn [mɛnː] vs men [mɛn]), so one wouldn't expect any difference here either. A more relevant minimal pair is e.g. ett vs ätt, where the former comes from Old Swedish étt [eːtː] and the latter from Old Swedish ǽtt [æːtː].
So, the e vs ä only exists where the spelling respects the etymology - if Old Swedish had no difefrence in spelling, then modern Swedish dialects can't be expected to either.
Jens Persson (217.211.10.14 (talk) 13:42, 29 March 2008 (UTC))

I accept your dialect argument about the diphthongs. You are probably also right about the number of phonemes. (I don't make any difference between “verk” and “värk” ether...) About the “half-vowels”, I actually did not know the English term. It was not in my dictionary ether so I made a literal translation of the Swedish term (“halvvokal”). It stands for a group of language sounds that can be used as vowels yet have some consonant properties. Except for the close central rounded vowel I only know about one such sound: the one written with the first letter in the Hebrew alphabet. I hope I have explained it clear enough for you to understand!

2007-02-22 Lena Synnerholm, Märsta, Sweden.

A close central rounded vowel (which is considerably more front in the Swedish spoken around Stockolm) is not a semi-vowel per se. In fact, no vowels are. What you're referring isn't really a separate sound, but a phonological process which occurs in many languages. For example Mandarin and in the diphthongs of German. However, this occurs only in a few loan words in Swedish. I can't speak for minor dialects, but I'm pretty sure it's fairly marginal.
Peter Isotalo 15:43, 22 February 2007 (UTC)

Either I have misunderstood the concept or it is not called “half-vowel” (or “semi-vowel”) in English.

2007-02-24 Lena Synnerholm, Märsta, Sweden.

I think you have misunderstood(?) The Swedish Wikipedia article on halvvokal seems to be about the same concept as semi-vowel. I think the scope of the article is about "standard, native" phonology, not about exceptions in loanwords and dialects. (Dialects could be mentioned, though, but they would require their own sections.) 惑乱 分からん 22:45, 25 February 2007 (UTC)
I think the comment about aleph in the Hebrew alphabet refers to the glottal stop. It isn't mentioned because it's not considered a separate phoneme in Swedish, i.e. there are no words with change in meaning depending on whether they're pronounced with or without a glottal stop. 惑乱 分からん 22:52, 25 February 2007 (UTC)

Alright, I misunderstood the concept of semi-vowels from start.

2009-08-27 Lena Synnerholm, Märsta, Sweden.

Even though the discussion seems to be obsolete in time, I just have to react against the wrong result of the discussion: Swedish realizes almost all "long vowels" with phonemic diphtongs: [ɒː] by [ɒ͡a], [oː] by [ʊ͡a, [u:] by [ʊw], ["ʉ:"] by [ʏɥ], [ɛ:] by [ɛ͡a], etc. etc... That the IPA convention instead describe them as long vowels is just a phoneticists convention. Rursus dixit. (mbork3!) 17:04, 17 October 2010 (UTC)

Long consonants

Correct me if I am wrong, but doesn't Swedish (except Scanian)have long consonants? In the IPA transcriptions I see no explicit use of length marks on consonants which in Swedish for sure are long. Oh, wanna know my source for my claim concerning the existence of long consonants? What about 7,000,000 native non-Scanian swedes!? // Jens Persson (90.227.153.229 (talk) 20:49, 5 October 2008 (UTC))

The phonologies don't seem to be very clear on exactly how long consonants are realized in Swedish. They are not considered to be phonologically interesting, since they co-vary with the vowel length: within syllables long vowels are followed by short consonants and vice versa. There seem to be some exceptions, but I don't know exactly how to describe this. Neither do I know how the various dialects differ from one another. Do you have any sources on the issue?
Peter Isotalo 10:39, 9 October 2008 (UTC)
From a theoretical point of view, since long and short vosels are differentiated in quality as well ([i(ː)] vs [ɪ], [e(ː)] vs [ɛ] etc), one could remove the length mark for vowels as well. Thus, dag and dagg would be represented phonetically by [dɑg] and [dag], respectively. Don't you agree? I am just trying to be consistent here.
Jens Persson (90.230.148.95 (talk) 20:42, 29 November 2008 (UTC))
That's how it's done for American English. But we should be hesitent to do this unless linguistic literature on Swedish does this. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 00:40, 30 November 2008 (UTC)
I guess foreign literature on Swedish phonology doesn't emphasize consonant length since Swedish (apart from South Swedish = historical East Danish) is one of very few languages with long consonants, which makes such a concept difficult to explain and use.
Jens Persson (81.235.129.174 (talk) 22:55, 30 November 2008 (UTC))
I really believe the existence of long consonants should at least be mentioned. It doesn't have to be phonemic to deserve mentioning! One can definitely hear that "kappa" has a very long [p] in the speech sample on the article. As a native speaker of Italian, which doesn't have vowel length but does have geminate consonants, I cannot readily hear the vowel contrasts, but the consonantal ones are exceedingly clear (at least in the dialect that speaker uses, and many others I've heard). Also, shouldn't long consonant be marked in at least one of the two phone(m|t)ic transcriptions given (by the way, is "broad" and "narrow" a way of saying "phonemic" vs "phonetic"... or what?) LjL (talk) 21:22, 16 June 2009 (UTC)
I suppose it wouldn't hurt to mention the long consonants as a phonetic trait. However, Swedish isn't comparable to Italian where the difference between wors like fato ("fate") and fatto ("fact") depends just about entirely on consonant length. Vowel length and quality as well as tonal aspects are more important in Swedish. As far as I know the convention among Swedish phonologists is to indicate length for vowels but not for consonants.
Peter Isotalo 12:26, 2 July 2009 (UTC)
I understand that, but if we have two separate transcriptions - one phonemic, one phonetic -, then the phonetic one should show prominent phonetic traits.
And I assure you that if one is trained to hear consonant length because it's distinctive for them, well, it's quite prominent in Swedish.
If you look at the Italian phonology article, you'll see that long vowels are indicate for words like "fato" (actually I think that specific word is used as an example) when transcribing phonetically.
After all, the phonemic side of things is partly a matter of convention... one could easily say that vowel length is distinctive in Italian, and consonant length is not, since they're found in complementary distribution after all. Sure, Swedish adds the fact that vowel quality is mostly as important as length, while in Italian, it's really just length.
Second issue: if we indicate consonant length, then how do we indicate it - using ":" or by doubling the consonant? The latter is done in the Italian phonology and other Italian-related articles, mostly, I believe, to make it clear that "part" of the consonant is meant to belong to the preceeding syllable. Folkets lexikon uses ":", though, and maybe it makes more sense for Swedish.
LjL (talk) 12:54, 2 July 2009 (UTC)
In Italian, it's definitely the consonant length, not the vowel length, that's distinctive, because (I'm pretty sure) only stressed vowels are lengthened before singleton consonants in Italian. A word like amare is [aˈmaːre] with a short [a] in the first syllable, isn't it? +Angr 13:40, 2 July 2009 (UTC)
No - Italian is commonly analyzed as having more than two (namely, three or sometimes four) vowel lengths. One common analysis is "long", "half-long" and "short". Stressed+SingleConsonant=Long, Stressed+DoubleConsonant=HalfLong, Unstressed+SingleConsonant=HalfLong, Unstressed+DoubleConsonant=Short. LjL (talk) 13:48, 2 July 2009 (UTC)
I think you mean "yes" rather than "no", since you're agreeing with him. kwami (talk) 06:26, 3 July 2009 (UTC)
I don't think I am. Keep in mind that Italian stress is usually analyzed not only as length (also as loudness, namely), so you know whether or not a syllable is stressed regardless of the above. LjL (talk) 13:34, 3 July 2009 (UTC)
Using the length marker for marking long consonants in phonetic transcription is probably the least complicated (and controversial). The issue about long consonants being split between syllables is trickier. I don't recall reading anything specific about it concerning Swedish, but I guess we could always mail a question to the linguists at Stockholm or Gothenburg University and see if they know anything.
Peter Isotalo 07:34, 3 July 2009 (UTC)

As a native speaker of Swedish I can tell that consonant length is unimportant for the meaning of Swedish words.

2009-08-27 Lena Synnerholm, Märsta, Sweden.

That doesn't help much. First, being a native speaker doesn't necessarily give you an advantage in interpreting the phonology of the language (actually, I found that it's often the other way around: being a native speaker "traps" people into using spelling for phonetics and other things); second, could you still say "it's unimportant" if we removed vowel length? The question of phonemicity is whether the vowels are phonemic and consonants are not, or vice versa. To my Italian ear (trained to long consonants, but untrained in vowel length), two Swedish words like "pika" and "picka" are more easily distinguished by the consonant length than the vowel length or quality (both sound like /i/ to me, in the accents I've heard). --LjL (talk) 19:28, 27 August 2009 (UTC)
Also that consonant length doesn't affect the meaning of words doesn't mean that it isn't worth covering. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:43, 27 August 2009 (UTC)
There's also no point in arguing that consonant length is phonemic in Swedish, because it's simply not analyzed as such. The differences in vowel quality only confirms that. And while I agree that being a native speaker doesn't really add authority in analyzing one's native language, Lena does has a point here: to native speakers of Swedish, vowel length (in combination with the change in quality) is the noticeable difference, not consontant length. That you, LjL, as a native Italian speaker interpret consonant length as being more relevant while not hearing the subtle difference in vowel length and quality only confirms what we've already been saying about the differences between the two languages.
Peter Isotalo 22:11, 27 August 2009 (UTC)
The doubts seem to really revolve about vowel quality: as far as quantity goes, even Italian has a short/long vowels distinction (complicated by stress) in complementary distribution with consonant length; just as Swedish does, I'd say. So when I said that, as an Italian speaker, I "hear consonant length", I could probably just as well have said that I hear vowel length; I'm just used to consider the former relevant. Why? Perhaps one reason is that long consonants, and not long vowels, are clearly marked in spelling; but then again, that's the same that happens in Swedish. So the only difference I can see is there's a change in vowel quality in Swedish, while there seems to be none in Italian. In any case, I simply cannot concur with the original claim that consonant length is "unimportant" to meaning: consciously or not, I'm pretty sure it's at least as important as vowel length and quality: they work together, and that's hardly something unusual among the world's languages. On the other hand, I'll again endorse Ƶ§œš¹'s point that non-phonemicity doesn't make a phenomenon irrelevant. --LjL (talk) 23:14, 27 August 2009 (UTC)

My pont is that Swedish does not make any differences between long and sort consonants. I know this becuse Swedes tend to pronounce Finnish long and short consonants wrongly. When a consonant follows a long vowel it is typically prounced short regardless of whether it is short or long in Finnish. Similary, consonants following a short vowel is typically prounced long. If we should describe the use of long and short consonants in Swedish I think a description of the complementary distribution shuld be enough.

2009-08-28 Lena Synnerholm, Märsta, Sweden. —Preceding unsigned comment added by 212.247.167.70 (talk) 18:53, 28 August 2009 (UTC)

I'd like to add that the Lexin Swedish-English online dictionary marks vowel length as well as consonant length, which is unusual and suggests the significance of consonant length in Swedish apart from other languages. Moreover, the words känna and tjäna are as much minimal pairs for a long vowel as for a long consonant. There is a stark gemination of the n's in känna that only occurs in English and German between words but is frequent in Russian, Korean, Italian and Japanese.99.55.156.65 (talk) 00:08, 19 September 2009 (UTC)

Quote: "There's also no point in arguing that consonant length is phonemic in Swedish, because it's simply not analyzed as such"
Not true. Some Swedish phonologists (e.g. Tomas Riad, in his Structures in Germanic Prosody) do analyse Swedish consonant length as underlying and vowel length as secondary.--Anonymous44 (talk) 21:08, 1 February 2010 (UTC)
I agree. The article now points out that there are different analyses to the phenomenon of consonant length. It is true, however, that consonant length is not independent of vowel length. I also suspect that Peter's analysis is the more common one. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 00:02, 2 February 2010 (UTC)
Lena Synnerholm's claim:
Swedish does not make any differences between long and sort consonants
Oh, yes it does. Either the vowel is long or the following consonant or consonant cluster is long. The cases like in Finnish where a short vowel is followed by a short consonant have disappeared from Swedish by lengthening the vovel, [gata] becale [gɒːta]. The case in Finnish where a long vowel is followed by a long consonant (cluster) did never occur, as well as the over-long consonants in Sami. Rursus dixit. (mbork3!) 17:26, 17 October 2010 (UTC)

If it is true that "within syllables long vowels are followed by short consonants and vice versa", then long consonants are already being marked by a short vowel and whenever you see a double consonant in spelling, you know a short vowel precedes it. This is what is meant by phonological; if the pattern holds, then it is better to memorize the pattern rather than redundantly specify both. Using a term in mathematics, any variable you specify should be a free variable.75.73.93.176 (talk) 19:35, 29 October 2010 (UTC)

Info on rounding

The article on Roundedness states:

Swedish is unusual in that dialects of it make a phonemic distinction between the two types, having unrounded, endolabial, and exolabial front close-mid vowels. Some varieties of Dutch make the same distinction. There is no dedicated IPA diacritic to represent this contrast, and without disambiguation both the word "rounded" and the symbols for the rounded vowels are understood to refer to exolabial rounding.

So this article(Swedish phonology) needs an explanation of the different sorts of lip-rounding in Swedish, especially that for / ʉ /. Anyone?

Marquetry28 (talk) 02:41, 16 June 2008 (UTC)

The funniest thing is, the sound files about long 'u' & long 'y' sound reversely to what their descriptions say: it's the long 'u' that sounds exolabial (i.e. like in German or French) and long 'y' endolabial (not very dissimilar to long 'i'), not vice versa. 89.231.116.65 (talk) 20:23, 22 February 2011 (UTC)

The ʂ sound for rs

I found the ʂ sound missing in the list of fricatives, so I included it, with a description. You see, all front-R using Swedes use ʂ for words spelled with rs, which is an important fact that was missing in the article. But Peter Isotalo just deleted my entire contribution. Why, Peter? Are you unaware of this rather obvious fact or did you misinterpret?

Also, I thought it would be interesting to add an account of the fact that there is a tendency among Swedes to pronounce in initial /ɕ/ close to [ʂ]: känna, /ʂɛn:a/, "feel".

I was bewildered you found nothing in my contributions worth saving.

—Preceding unsigned comment added by Aikclaes (talkcontribs) 08:13, 17 September 2008 (UTC)

[ʂ] as an allophone of the phoneme /ɧ/ is already mentioned in the article as well as one of the more common realizations of /rs/. If you look at the individual sounds listed in the tables, they are all phonemes, and that's the usual way to organize things in phonologies. The many varying allophones are listed under each of the phonemes. So in that case you simply repeated information that was already present under other headings.
As for the realization of initial /ɕ/ that would be "close to [ʂ]", you added this information without any references. I don't know if this might be a feature that is specific to certain regions, but I am myself highly skeptical to the claim, and that is why I removed it. If you feel this is something that has been investigated by linguists, please feel free to present a reliable source and I will gladly reinsert the fact.
Peter Isotalo 13:11, 17 September 2008 (UTC)
Wrong again! [ʂ] is in Swedish used as the sound gotten when assimilating [r] and [s] into one sound, it is never an allophone of [ɧ] (sj, stj and their kin), while the differently pronounced [ʃ] is a Svea allophone of the Göta [ɧ]. The sound of [ʂ] is clearly "darker" than the sound of [ʃ] (pronounced like sh in ship). I've never seen so much erroneous claims in any talk page. Rursus dixit. (mbork3!) 17:15, 17 October 2010 (UTC)
The [ʂ] sound is used in lots of words in all dialects, e.g. "börs", "törs", "först", "fars" (two words with the same spelling but not the same pronunciation), "mors" (again two words) and "morsk". (Stefan2 (talk) 19:43, 11 October 2011 (UTC))
Added this sound to the list of consonants. The three examples "sjok", "kjol" and "börs" clearly use different consonant sounds, although foreigners often have difficulties telling them apart. (Stefan2 (talk) 19:56, 11 October 2011 (UTC))

Palatalization of velars

In Polish (my native tongue) it's partially distinctive. In Swedish, apparently, it's not, and this is perhaps why it hasn't been covered so far, but it's very prominent when you're listening to the samples here and in the pronunciation guide linked to hereunder. It seems to be obligatory everywhere but for directly before the stressed vowel (unless it's k+a). 'många', for instance, sounds to me almost like ['mɔŋ'ɲa]; 'regn' like ['rɛŋʲn], 'språk' like ['spruɔkʲ], 'glas' like ['gʲlɒ:s]. Surely most other European language don't do such things. 89.231.116.65 (talk) 21:06, 22 February 2011 (UTC)

There seem to be also other minor issues not covered by the article, namely possible pronunciation of 'j' as [ʝ] and possible devoicing of 'v' after a voiceless consonant (as in 'tvistade' in the tale). 89.231.116.65 (talk) 21:51, 22 February 2011 (UTC)

I think it (i.e."tvistade") has to do with fortis and lenis consonants with fortis being phonotactically prohibited in certain clusters. Fortis are aspirated t, p, k, and f, the rest being lenis. V in tvistade does not denote a [v], rather a lenis [f]. I’m just analyzing the clusters as I see them with my own phonetic knowledge and some vague memory of a fortis vs. lenis discussion in a book on Swedish phonetics so I don't have any reliable source, sorry. As for the palatalization I suppose it is quite prominent in my own speech at least, but since, like you say, it's not phonemic I haven't really given it much thought. It is a most intriguing subject though.. Might pester someone at the university (of Uppsala) about it...Servus Triviae (talk) 15:16, 19 May 2011 (UTC)

As a fellow Slav, I second the observation about palatalization of velars, although I've never seen it explained officially anywhere.--91.148.159.4 (talk) 23:45, 7 January 2012 (UTC)

The short "i".

What is rendering the Swedish short “i” as [ı] in IPA based on? This is the same sound as the short “i” in English "hit" or German "mit" and if you ask me these sounds are in no way similar to the Swedish sound. To me the Swedish short “i” is simply the same (in quality) as the long “i” i.e it should be [i]. This is not the first time I’ve seen this and I’m curious to know where it comes from, I feel it is an obvious error and yet it keeps on being repeated in various forums. Nothingbutmeat 16:28, 30 July 2007 (UTC)

It's slightly, but not considerably different. For one thing, it's slightly lower. If you want sources, then check Garlén or Engstrand. The latter, btw, is the professor of phonetics at Stockholm University. If you think he's mistaken on this point, you can always email him and I'm sure he'll explain himself.
Peter Isotalo 17:57, 30 July 2007 (UTC)
I can only say that the spoken example for "sill" given in the article doesn't sound like [ı] at all to me, but rather much more like [i]. LjL (talk) 21:16, 16 June 2009 (UTC)
It is pronounced exactly like “i” in English "hit". Rursus dixit. (mbork3!) 18:13, 7 February 2010 (UTC)
It's not. Check your ears :) --89.79.88.96 (talk) 13:00, 29 April 2012 (UTC)
89.79.88.96 is right, it is definitely not the same. It's considerably more front than the English /ɪ/, it's blatant on the recording. --89.79.88.109 (talk) 20:31, 20 August 2012 (UTC)

Allophones of short e and u

In Le suédois en vingt leçons, it's claimed that Swedish has two allophones of the short e (page 11), and two of the short u (page 13). The words rutin and hund are listed as having different realizations of the short u, and bett and fem for short e, for example.

Does anyone have any other sources for this? I would say that in my speech, there isn't any difference, but I'm aware of the problem with self-analysis. Mats (talk) 12:05, 21 October 2009 (UTC)

There is such a thing as a short [ʉ] in words like musik and rutin and it's clearly audible. I think it may be interchangable with [ɵ], but I'm not sure. I'm pretty sure I can find references to it in some Swedish dictionaries. The difference between fem and lett, however, I far more sceptical to. I'll see if I check it.
Peter Isotalo 14:57, 21 October 2009 (UTC)
Yet another native view... This may be a distiction not always recognized by younger people, but lett and lätt are often pronounced differently. More examples of the former are bett, sett, ledd, bedd, sedd, hem, rem and of the latter are bättre, fem, femton, sätt, tätt, häst, fäst, while fest could go with either - besides the even more open short ä found in northern dialects which is quite different. Pronouncing fem like bett may sound a bit childish, but the other way around noone would notice. In any case, the use varies with dialect and age. --81.216.218.158 (talk) 04:26, 24 November 2012 (UTC)

Carl Gustaf Rossby

Could someone add the Swedish language pronunciation for Carl-Gustaf Rossby and Loreen (singer)? — Preceding unsigned comment added by 91.238.215.134 (talk) 09:17, 9 July 2012 (UTC)

  Done Rursus dixit. (mbork3!) 12:13, 17 May 2014 (UTC)

Stress and pitch

The section's description of the tone accents and especially of the Central Swedish ones is very inaccurate and much of it seems to be based on descriptions from a century or half a century ago. First it is stated that the acute has a low tone as opposed to the high tone of the grave, then, in contradiction, it is stated that the acute has a fall as opposed to the double fall of the grave. This type of naive phonetic descriptions have been discarded for many decades now, since the effects of sentence intonation and compression have been successfully distinguished from the tone accents proper. Googling for just about any recent publication about the tone accents would show that and give a more correct picture. I'm not editing, because the mess is such that I would have to rewrite the section nearly from scratch, and I don't feel like it. --91.148.159.4 (talk) 19:17, 3 October 2010 (UTC)

Are you disputing that Central Swedish has a low acute tone and high grave tone or are you pointing out a seemingly self-contradictory depiction. The statement about Central Swedish's tones is cited by a very recent source (which can perhaps be used to fix the section) so it's probably not antiquated. Perhaps the wording isn't clear, as the sentence beginning "Generally, the grave accent is characterized by..." compliments the sentence "The actual realizations of these two tones varies from dialect to dialect." — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 19:53, 3 October 2010 (UTC)

I'm pointing out a contradiction and I'm also disputing the actual claim. The contradiction is obvious, because the text says, later, that the acute has a single fall and the grave has a double fall, which suggests a difference not of tone value but of the number of tones involved. However, it's not just that one of the claims is wrong: in fact, both are. Specifically:

1.The source saying that the acute is Low and the grave is High has either been misunderstood or is wrong - it may be recent, but it's not devoted specifically to the tone accent. To say that the acute has a Low and the grave has a High is inaccurate, because in fact the grave also contains the same Low as the grave, only timed later. The typical melody of the acute in isolation is L*HL, and the typical melody of the grave in isolation is H*LHL (* stands for the first tone of the stressed syllable). Thus the difference is an one extra High in the beginning of the grave. But even this does not always hold true, because there may be a H preceding the L* of the acute (there is some debate about this particular point, though).

2.To say that the acute is falling and the grave is doubly falling is only right if you look casually at the phonetics of words in isolated pronunciation. In fact, the only fall of the acute and the second fall of the grace are both due to a Low boundary tone, which is present only phrase-finally and hence has nothing to do with the tone accent. The last Highs of the acute and of the grave are the result of so-called focal accents, which occur mostly when focus is signalled and aren't really parts of the tone accent proper either. Thus, at most one could say that the acute is Low in the stressed syllable, potentially with a High before the stressed syllable, while the grave has a High-Low fall in the stressed syllable. Most of this has been known since Gösta Bruce's dissertation from 1977, which is seminal and has been cited in pretty much every work on the Swedish word accents since then. Again, simple googling can prove this. --91.148.159.4 (talk) 17:49, 4 October 2010 (UTC)

Okay, well you and I are both too busy to fix it ourselves but the section is tagged. Are there any other works you recommend for fixing the section? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 19:14, 4 October 2010 (UTC)
Other works besides Bruce 1977? Well, I can't think of anything online that focuses exactly on describing the accents as such, but the basic, widely accepted premises about the Central Swedish word accents just have to be mentioned in any modern work on the topic. For example, here are a couple of PhD theses that I've found available online and that all have sections on the basic form of the tone accents according to Bruce: [1], [2], [3], [4]. Here's also an article by Tomas Riad, which also deals with tone in Central Swedish, among other dialects. There are many more such sources scattered on the web which could be used for that section. --91.148.159.4 (talk) 20:51, 4 October 2010 (UTC)
I don't know about the rest, but whoever added "In Standard Central Swedish, for example, the acute accent has a low tone while the grave accent has a high one.[62]" probably just made a typo - it's the other way around. Maybe just remove that bit since the difference is explained in more detail (and correctly) below. 85.225.51.30 (talk) 20:06, 27 August 2014 (UTC)

Too few sources!

Too few sources for the article! Duden 90 is not enough. While my reading thru indicates that most of it is pretty correct, there should preferrably be more sources to support the article better. Rursus dixit. (mbork3!) 17:35, 17 October 2010 (UTC)

What do you mean by "Duden 90"? What claims that haven't already been tagged need verification? — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 17:44, 17 October 2010 (UTC)
Duden 90 is this book: Mangold, Max (1990). Das Aussprachewörterbuch (in German) (3rd ed.). Dudenverlag. ISBN 3-411-20916-X. Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 16:56, 3 November 2014 (UTC)

Complementary quantity

Excerpt from the wikipedia article on 'Swedish phonology': "Central Standard Swedish (and presumably many other dialects) features a rare "complementary quantity" feature[61] wherein a phonologically short consonant follows a long vowel and a long consonant follows a short vowel; this is true only for stressed syllables and all segments are short in unstressed syllables" (Swedish phonology)

I have no sources to back this up (but if you look I think you will find them because I'm pretty sure this is common knowledge among those in the know) but this "rare" complementary quantity is not limited to dialects, It's Swedish standard. All stressed syllables must be long, that is either contain a long vowel or a long consonant. All unstressed syllables must be short, that is, in correct standard Swedish. There might be some dialect out there, which does not share this feature, but I doubt it.

Wellingtonhoots (talk) 12:09, 11 March 2011 (UTC)

The feature is rare amongst languages, not rare amongst other dialects (hence the "many other dialects" parenthetical). How should we reword it to make it clearer? — Ƶ§œš¹ <sall>[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 16:37, 11 March 2011 (UTC)
How about removing the whole dialect-'disclaimer' and just write that it's Swedish standard? Not only is it not rare in other dialects but it's simply the national standard of the Swedish language. — Preceding unsigned comment added by Wellingtonhoots (talkcontribs) 10:07, 18 March 2011 (UTC)
What confuses me is that the section implies that all Swedish syllables are either VC: or V:C. Is the final consonant of tack really phonetically a geminate, identical to the consonant in tacka? And are there no words at all with a long vowel followed by two consonants? For example, does hel with a long vowel (V:C) alternate with helt with a short vowel (VCC) or does it remain long (V:CC)? CodeCat (talk) 19:52, 11 March 2013 (UTC)
AFAIK, tack, tacka, and helt would all be VC: (unless tack is actually V:C); there are no V:C: words except in the dialects that allow them. — Ƶ§œš¹ [ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ] 22:18, 11 March 2013 (UTC)


Curious what other languages have "complementary length". I think Middle English might've, given how vowels were lengthened before single consonants; German may have as well. — Eru·tuon 21:06, 13 April 2015 (UTC)

@Erutuon: Standard Eastern Norwegian (AFAIK doesn't apply for /r/, which is often a flap even when "long"). Peter238 (talk) 21:29, 13 April 2015 (UTC)

Lowered mid front vowels

@Peter Isotalo: On this page you wrote that lowered /øː/ (i.e. [œː]) is the usual pronunciation in the southwest. If so, what about /ɛ, ɛː, œ/? Are they lowered to [æ, æː, œ]? — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 20:54, 22 March 2015 (UTC)

Not sure. Haven't really noticed anything. But that's just my impression.
Peter Isotalo 23:13, 16 August 2015 (UTC)
Thanks! Peter238 (talk) 01:33, 17 August 2015 (UTC)

Why aren't the retroflexes included in the consonants?

In Standard Swedish, there, quite obviously, are the retroflexes [ɳ], [ɖ], [ʈ], [ʂ], and [ɭ]. I don't think it's very representative of the Swedish language to exclude them because they aren't consonants on their own. They aren't mentioned a lot at all. Why? — Preceding unsigned comment added by Sönsvall (talkcontribs) 13:46, 6 September 2015 (UTC)

In English, there are spirated stops [pʰ], [tʰ] and [kʰ], but those aren't mentioned in the English table either. This is because they are not phonemes: they don't act to distinguish sounds from each other. The retroflexes in Swedish are allophones of r + consonant combinations. CodeCat (talk) 14:02, 6 September 2015 (UTC)
Sönsvall, those retroflexes are assimilations. They appear in words like bärs or flärd ("flamboyance") and also in compounds like karlslok or barskåp, but they also manifest themselves word boundaries in phrases like för sin dotters skull. It easily becomes something like [fœːʂɪndɔtːɛʂkɵl] when I say it in a running sentence, and I speak Central Standard Swedish with a tinge of stockholmska.
In Finland and southern Sweden the /r/ is not assimilated in any context. So retroflex consonants aren't even relevant to all Swedish-speakers.
Peter Isotalo 15:25, 6 September 2015 (UTC)

This table is showing what sounds are in Standard Swedish, so it doesn't matter if they aren't in southern Swedish and Finnish Swedish. They do distinguish words, such as "varda" and "vada". You could just as well skip the "ŋ" too, since it's just n and g put together, and we might as well also just skip "ɧ" and write "sj" in the IPA transcriptions. Isn't this table supposed to show what sounds are in Standard Swedish? It's is very weird to leave them out just because they are assimilations. ŋ and ɧ are also assimilations. — Preceding unsigned comment added by Sönsvall (talkcontribs) 15:44, 6 September 2015 (UTC)

Standard Swedish includes all major regional variants. You're thinking about the antiquated definition of rikssvenska that defines anything outside Stockholm and Uppsala as dialectal. The modern definition used by linguists includes anything that can be readily understood by other native Swedish speakers as Standard Swedish. But even if that wasn't the case, it's still not a phoneme since retroflexes only appear in very specific situations where /r/ is involved, including across word boundaries. /ŋ/ doesn't behave in the same way since there are no example in contemporary Swedish where /n/ and /g/ actually assimilate. You're confusing orthography with phonology in this case.
And /ɧ/ is neither a retroflex nor an assimilation. It's pretty clear from the content of this article, so read it more carefully. I also recommend checking out sources like Garlén (1988) or Engstrand (1999). Engstrand has also written an excellent basic introduction to phonetics called Fonetikens grunder. It explains many of these phonological processes.
Peter Isotalo 16:34, 6 September 2015 (UTC)


Questions about pitch accent

I would like to restructure and expand the section about pitch accent, mostly to rearrange the information in a way that makes it easier to get an overview (at least to me). But there are some things I would like to clarify first, not being a native speaker:

  • Only multisyllable words have pitch distinctions?
  • What happens with words of three or more syllables? The article describes that the grave accent has its second pitch drop spread over all unstressed syllables, does the same apply when more syllables are added? And what about the acute, is it the same (first high, remaining syllables low)?
  • What happens with words with stress on a syllable other than the first? Do these words have tonal distinctions, and if so how is this realised?
  • Likewise, what happens to compound words? Do all parts of the compound receive their own pitch pattern, so that a word could have grave accent on the first syllable but acute on the second?
  • How are loanwords handled?

I would appreciate it if a native speaker or someone else who knows this could answer these questions for me. CodeCat (talk) 16:12, 4 July 2014 (UTC)

In both Central Standard Swedish and Standard Eastern Norwegian, words with one syllable are toneless, even if they are stressed. For the word to have a pitch accent, there need to be two or more syllables. This is unlike Danish, where stød can occur in monosyllabic words. Peter238 (talk) 09:43, 15 October 2014 (UTC)
@CodeCat:
  • Tone accents in words longer than two syllables work according to similar rules, but they are complicated by secondary stress and whatnot. With more than two syllables, though, there aren't really any minimal pairs (like and/ande or tomt/tomte in definite form).
  • Stress and tone accent are different aspects of Swedish prosody. If stress is on the second syllable (or beyond) in a two-syllable word, there simply is no accent 2. In longer compounds, it's more complicated, though after a certain number of syllables, I believe words tend to adhere to fixed patterns relating to the number of syllables. I don't believe this varies much due to the pitch accent of constituent parts.
  • Loanwords fit into the overall phonological system, but it's more important how they fit into Swedish morphology. Verbs loaned from English, for example, always get a basic -a ending attached, like mejla ("to e-mail") or dejta ("to date"). They are no different from core Swedish verbs like flyga ("to fly") or vara ("to be"). Two-syllable noun loanwords are pretty much always accent 1, I think. There might be the occasional that might be accent 2, but they're rare since they don't tend to match phonologically with Swedish accent 2 nounds.
The article could certainly be elaborated regarding this. Maybe not every imaginable tone pattern, but some basics.
Peter Isotalo 17:06, 17 August 2015 (UTC)
Is it true that all Swedish compounds are accent 2? Because that's not the case in Oslo Norwegian, in which the compounds are almost randomly assigned either one of the tonemes (see Kristoffersen (2000)), which is a major annoyance to foreign learners (or at least surely looks like it). Peter238 (talk) 22:47, 17 August 2015 (UTC)
I should stop pretending to actually know the minutia of these things. I can recommend these sources, though.[5][6]
Peter Isotalo 01:46, 18 August 2015 (UTC)
I'll take a look at these, thanks. Peter238 (talk) 22:37, 19 August 2015 (UTC)

@Peter Isotalo: I couldn't access Riad's book, but take a look at what I found on Reddit: [7] (it's from Riad's book). It's exactly what I needed! Peter238 (talk) 05:26, 3 September 2015 (UTC)

Added to the article (as well as Norwegian phonology) per WP:FAIRUSE. Peter238 (talk) 23:24, 5 January 2016 (UTC)

Ulla-Britt Söderlund

So what happens with a name such as Ulla-Britt Söderlund – do the two first names, each with single (acute) tone, combine to form a single name with double (grave) tone: /ˈɵlːəˈbrɪtː/? Is she /ˈsøːdərˈlɵnd/? Or should that /ə/ be /ɛ/, even though it is realized as a schwa? And should it be /rl/ or /ɭ/?

Her surname is pronounced [²søːdɛrlɵnd], maybe with a secondary stress on the last syllable, but I'm not sure. I think southerners would pronounce it [²søːdɔʁlɵnd], with an [ɔ]-schwa (so to say) because of the following uvular [ʁ].
I don't know what to make of the Forvo pronunciation of Ulla-Britt, but it's completely possible that natives treat it as one word ([²ɵlːabrɪtː] - again, maybe the last syllable has a secondary stress), but it's just as possible that the user who uploaded that file simply mispronounced the words. Maybe @Peter Isotalo: will know. Mr KEBAB (talk) 03:44, 9 April 2017 (UTC)
Forvo pronunciation is correct. Hyphenated double first names in Swedish are considered just one name. There's a whole bunch of these that were common names given to children born in the 1940s-1960s: Britt-Inger, Anna-Maria, Maj-Britt, Anna-Lena, Kjell-Åke, Jan-Inge, Björn-Åke, etc. You can't split them up. If someone has two first names, it's not hyphenated and are considered separate.
"Söderlund" has an accent 2 pattern with the secondary stress on "lund". The "e" is pretty much a schwa since it's unstressed.
Peter Isotalo 09:52, 23 September 2017 (UTC)

Förd(ö/u)mmande

”In some pronunciations, traditionally characteristic of the varieties spoken around Gothenburg and in Östergötland, but today more common e.g. in Stockholm and especially in younger speakers, [œ] and [ɵ] merge into [œ]. Words like fördömande ('judging', pronounced [fœˈɖœmːandə] in Standard Swedish) and fördummande ('dumbing', pronounced [fœˈɖɵmːandə] in Standard Swedish) are then often pronounced similarly, if not identically.”

Can someone confirm this? I’m from Stockholm county but I use the Standard Swedish pronunciations for both words. Although I do pronounce /ur:/ and /ör:/ the same. Dörr and surr are thus rhymes according to me, both pronounced /ɵrː/. Maybe it is this the source means? How do you pronounce dörr and surr?Jonteemil (talk) 12:50, 22 June 2018 (UTC)