Incorrect pronunciations you have (or have had) to unlearn

[Zaarin]

That's one of the few words I encountered in speech before I encountered it in print. :p An embarrassing one of my own: despite being perfectly familiar with /ˈkʰɛɪ̯.ɒs/, I was a teenager before I associated it with chaos, which I (fortunately mentally) pronounced [ˈʧɑ.ɔʊ̯z].

[Soap]

One more, I guess ...

haphazard
(re)upholster

Both with /f/. In both cases I was probably <10 years old. For the first one, I didnt know that "hap" was a word, so i parsed it as "half-hazard" and didnt really think twice about the spelling. For the second word, the whole word was unknown to me, so I simply parsed it as a single unit saying something like /ju'fol.stər/.

Actually, I apparently have *still* been getting it wrong all this time. "upholster" is not a compound of "up" + "holster", and it apparently has a short o (/ɔ~ɒ~ɑ/ depending on dialect). That means either Ive never heard the word spoken out loud in my life, or there are other people getting it wrong, because I'm pretty sure I'd have noticed and remembered if I had heard it with the dictionary pronunciation.

Actually, it seems that Merriam-Webster and Wiktionary disagree.

https://www.merriam-webster.com/dictionary/upholster

https://en.wiktionary.org/wiki/upholster

[Zaarin]

I'm with Merriam-Webster on this one: upholster has a long /o/ and the other way sounds barbaric...

[Sumelic]

Actually, I apparently have *still* been getting it wrong all this time. "upholster" is not a compound of "up" + "holster", and it apparently has a short o (/ɔ~ɒ~ɑ/ depending on dialect). That means either Ive never heard the word spoken out loud in my life, or there are other people getting it wrong, because I'm pretty sure I'd have noticed and remembered if I had heard it with the dictionary pronunciation.

I actually haven't found a dictionary that gives /ɔ~ɒ~ɑ/. In any case, it seems pretty clear that the pronunciation with /ɔ~ɒ/ is due to pre-l merger of DOLL and DOLE. Original GOAT is also what would be expected from the etymology from "hold."

[linguoboy]

Although, this makes me wonder ... is resistance [ˈfjuɾɪl] or [ˈfju.tʰaɪ̯l]?

The first is the typical American pronunciation; the second is the typical English pronunciation.

I've been known to employ the British pronunciation at times in order to avoid homophony with feudal.

[Salmoneus]

Actually, I apparently have *still* been getting it wrong all this time. "upholster" is not a compound of "up" + "holster", and it apparently has a short o (/ɔ~ɒ~ɑ/ depending on dialect). That means either Ive never heard the word spoken out loud in my life, or there are other people getting it wrong, because I'm pretty sure I'd have noticed and remembered if I had heard it with the dictionary pronunciation.

I actually haven't found a dictionary that gives /ɔ~ɒ~ɑ/.

Dude, he actually already linked to wiktionary, which sure enough has the expected normal short-o pronunciation. [Long o in 'holster'? Bloody hell, that's almost physically impossible! A tongue twister, certainly...].

In any case, it seems pretty clear that the pronunciation with /ɔ~ɒ/ is due to pre-l merger of DOLL and DOLE. Original GOAT is also what would be expected from the etymology from "hold."

No, it's not the dole-doll merger, which I don't have, and which is quite marked I think. Interestingly, that link suggests long o in 'bolt' and 'polka', but that's wrong, as the wiktionary entries for those words shows (although it does give both options for 'bolt'). I might sometimes have a long O in 'bolt', but I think that's just non-SSBE parental influence showing through. Likewise, I might sometimes say 'colt' with a long O, but I think that's an Americanism (I'd probably only say it for the gun manufacturer, and not always then, and probably never for the horse).
So I guess the basic SSBE rule is just that O is always short before /l/ followed by a voiceless consonant or /v/.

[Travis B.]

Although, this makes me wonder ... is resistance [ˈfjuɾɪl] or [ˈfju.tʰaɪ̯l]?

The first is the typical American pronunciation; the second is the typical English pronunciation.

I've been known to employ the British pronunciation at times in order to avoid homophony with feudal.

This is a pair of words distinguished only by vowel length for me, if I use the American pronunciation of futile, i.e. futile [ˈfjʉɾɯ(ː)] versus feudal [ˈfjʉːɾɯ(ː)]. That said, I often use the British pronunciation of futile, i.e. [fjʉˈtʰaːejɯ(ː)].

[Sumelic]

Actually, I apparently have *still* been getting it wrong all this time. "upholster" is not a compound of "up" + "holster", and it apparently has a short o (/ɔ~ɒ~ɑ/ depending on dialect). That means either Ive never heard the word spoken out loud in my life, or there are other people getting it wrong, because I'm pretty sure I'd have noticed and remembered if I had heard it with the dictionary pronunciation.

I actually haven't found a dictionary that gives /ɔ~ɒ~ɑ/.

Dude, he actually already linked to wiktionary, which sure enough has the expected normal short-o pronunciation.

I missed the Wiktionary link (I misunderstood "it seems that Merriam-Webster and Wiktionary disagree" as "it seems they both give the same pronunciation, disagreeing with the one I found in the first dictionary I checked"). So that makes one dictionary that gives /ɔ~ɒ~ɑ/, against MW, AHD, Collins, Macmillan, Cambridge, and Random House/Dictionary.com giving /oʊ~əʊ/.

I'm sure the LOT pronunciation exists; all I was saying is that I don't see any reason to think it is original or to see it as "expected" unless you speak an accent that has some kind of change of historical pre-L GOAT to LOT. It was a mistake for me to equate all of these changes as the "dole-doll merger," since there are many varieties of pre-L vowel mergers with different sets of environments, just like there are many varieties of LOT > CLOTH. I don't know a good established umbrella term, however. Of course, some of these changes are part of usual present-day British English accents, just like some of the CLOTH-broadening changes part of usual present-day American English accents, but they nevertheless originated as innovations/sound changes. And are you certain the distinction between GOAT and LOT has been completely eliminated before /l/ + voiceless consonant for all SSBE speakers? This might be an area where variation still exists. See for example the following post from 2007: scolding slurry, by a speaker who groups "bolt" with "goal" and "scold" as words that may have a backed dipthong kept distinct from LOT.

"Bolt" is likewise originally GOAT, which is the only pronunciation listed in the OED (which evidently is not comprehensively up-to-date on contemporary British English); "polka" on the other hand I don't know about. It may have shown some variation between LOT and GOAT since it was originally borrowed since it is a relatively recent borrowing, and neither /oʊlk/ nor /ɒlk/ exists in native English words as far as I know.

[Boşkoventi]

Actually, I apparently have *still* been getting it wrong all this time. "upholster" is not a compound of "up" + "holster", and it apparently has a short o (/ɔ~ɒ~ɑ/ depending on dialect). That means either Ive never heard the word spoken out loud in my life, or there are other people getting it wrong, because I'm pretty sure I'd have noticed and remembered if I had heard it with the dictionary pronunciation.

I actually haven't found a dictionary that gives /ɔ~ɒ~ɑ/.

Dude, he actually already linked to wiktionary, which sure enough has the expected normal short-o pronunciation. [Long o in 'holster'? Bloody hell, that's almost physically impossible! A tongue twister, certainly...].

In any case, it seems pretty clear that the pronunciation with /ɔ~ɒ/ is due to pre-l merger of DOLL and DOLE. Original GOAT is also what would be expected from the etymology from "hold."

Begging your pardon, Sir, but I feel I must direct you to the right venerable Oxford English Diction'ry [sic], Second Edition, wherein the pronunciation of "upholsterer" is given as [ʌpˈhəʊlstərə(r)]--that is to say, with a so-called "long O". No other related terms are given pronunciations, suggesting that this one is definitive. This is not, of course, intended to imply that you are "wrong" but only that your notion of, uh, well, "expected normal" is, well, wrong.

It is, however, worth noting that the now obsolete term "upholster" (sb., meaning "upholder", "upholsterer") had a 16th century northern variant uphaldster, which hints at your ... quaint, possibly Irish-influenced (?), pronunciation.

[Sumelic]

It is, however, worth noting that the now obsolete term "upholster" (sb., meaning "upholder", "upholsterer") had a 16th century northern variant uphaldster, which hints at your ... quaint, possibly Irish-influenced (?), pronunciation.

This idea that this is "Irish-influenced" seems like a pretty wild guess. I also don't know what basis you have for calling /ɒlt/ here "quaint", a description that seems kind of hostile to me. Are you another southern British English speaker? Do you also feel like there is no possibility of variation in this area in what is considered to be a "standard" southern English accent?

The old northern spelling "uphaldster" looks to me like it represents a different development from the OE vowel in this word, which was short "a/ea". The "o" in standard English results from lengthening of "a" before the homorganic cluster /ld/, and then rounding of the resulting ā to ǭ (that is, [ɔː]), which developed to Early Modern English [oː] (which then had further possible developments like diphthongizing to [oʊ], fronting the first part to get [əʊ], etc.). This rounding of earlier ā did not apply in Scots, hence "stone" = Scots "stane", "old" = Scots "auld", and "hold" = Scots "haul(d)". I gather that Sal doesn't pronounce "old" as /ɔːld/ or /ɒld/, so his accent presumably does descend from the southern accents that had ā > ǭ. He indicated in his post that pronouncing olC as /ɒlC/ is conditioned by the voicelessness of the following consonant, which strongly suggests to me that it is a (relatively) recent change, like the shortening of "salt" to /sɒlt/, and part of the general tendency in southern British English towards merging vowels (especially back vowels) in some contexts before non-intervocalic /l/. You can see a description of how some of these mergers have developed phonetically from the historical "RP" vowel system on Geoff Lindsey's blog, http://englishspeechservices.com/blog/goose-backing/

[Salmoneus]

Begging your pardon, Sir, but I feel I must direct you to the right venerable Oxford English Diction'ry [sic], Second Edition, wherein the pronunciation of "upholsterer" is given as [ʌpˈhəʊlstərə(r)]--that is to say, with a so-called "long O". No other related terms are given pronunciations, suggesting that this one is definitive. This is not, of course, intended to imply that you are "wrong" but only that your notion of, uh, well, "expected normal" is, well, wrong.

I hate to break it to you, but many people in England don't actually consider the 2nd Edition OED's phonemic descriptions 'definitive'. I certainly don't claim that nobody has ever said it that way. Some people did use to say /b@Ult/, and /b@Ulst@/ and /h@Ult/ and whatnot. But people also used to say /klOT/! And sure, I might now and then encounter someone - an art critic transplanted from the 1940s, perhaps, or a minor ceremonial official employed personally by the Queen, or a Tory candidate whose first name is "Annunziata" or "Somerset" - who would naturally say that they /@ b@Ult Qv mO@nIN klO:T/. [because yes, the 2nd edition OED also predated the hoarse-horse merger in RP]. But those people would not be speaking what one would call modern SSBE. If you overheard them in the street, you'd raise your eyebrows and try not to stare.

Which may be why wiktionary gives the short-O interpretations, because it's just made by ordinary people, most of whom probably are unaware that you could once say, say, /h@Ult/.

To be fair, these pronunciations are less alien to modern SSBE than hoarse/horse or cloth/lot distinctions. But they're not the normal pronunciations in the modern language.

It is, however, worth noting that the now obsolete term "upholster" (sb., meaning "upholder", "upholsterer") had a 16th century northern variant uphaldster, which hints at your ... quaint, possibly Irish-influenced (?), pronunciation.

[/quote]
FWIW, actually I think the irish pronunciation is with long O.

[gmalivuk]

One more, I guess ...

haphazard
(re)upholster

Both with /f/. In both cases I was probably <10 years old. For the first one, I didnt know that "hap" was a word, so i parsed it as "half-hazard" and didnt really think twice about the spelling. For the second word, the whole word was unknown to me, so I simply parsed it as a single unit saying something like /ju'fol.stər/.

Actually, I apparently have *still* been getting it wrong all this time. "upholster" is not a compound of "up" + "holster", and it apparently has a short o (/ɔ~ɒ~ɑ/ depending on dialect). That means either Ive never heard the word spoken out loud in my life, or there are other people getting it wrong, because I'm pretty sure I'd have noticed and remembered if I had heard it with the dictionary pronunciation.

Actually, it seems that Merriam-Webster and Wiktionary disagree.

https://www.merriam-webster.com/dictionary/upholster

https://en.wiktionary.org/wiki/upholster

I don't know if it was a recent edit, but Wiktionary lists that as the UK pronunciation, whereas M-W is obviously American, hence the disagreement.

[StrangerCoug]

Interestingly, that link suggests long o in 'bolt' and 'polka', but that's wrong, as the wiktionary entries for those words shows (although it does give both options for 'bolt'). I might sometimes have a long O in 'bolt', but I think that's just non-SSBE parental influence showing through.

I have it in all three.

[Sumelic]

Interestingly, that link suggests long o in 'bolt' and 'polka', but that's wrong, as the wiktionary entries for those words shows (although it does give both options for 'bolt'). I might sometimes have a long O in 'bolt', but I think that's just non-SSBE parental influence showing through.

I have it in all three.

@StrangerCoug: I think you're a North American English speaker, though, right? That's what would be expected for you. Sal's use of terms like "expected" and "wrong" are all relative to his accent/what he perceives as SSBE.

As far as I know, no one argues that it's standard for North American English speakers to have /ɑl/ or /ɔl/, rather than /oʊl/, in these words. (That's why I was confused by Soap's comment about getting the pronunciation of "upholster" "wrong", since my impression is that Soap has a North American accent.)

To the extent that there has been any substantive disagreement in this thread, it seems to be about whether standard Southern British English accents now unanimously have the shortened pronunciation /ɒl/ before voiceless consonants (which is what Sal is saying) or whether it is still possible to hear /oʊl/ in this context (from an ordinary, non-pretentious speaker; obviously it's not relevant whether this pronunciation is used by people like Sal's dismissive caricature who uses "RP" pronunciations that are well-known to be obsolete in SSBE like "orphan" for "often" and "hoarse" split from "horse"). So far, Sal seems to be the only contributor who actually speaks British English, so there's been no evidence in this thread that /oʊl/ + voiceless consonant exists in standard contemporary British English (as I said earlier, it clearly exists and is standard in North American English).

The only thing that makes me a little skeptical about the idea that /oʊlt/ etc. are now nonexistent in SSBE is that it takes time for sound changes to spread and for more conservative speakers to die out--my impression is that the shortening of THOUGHT in this context (in words like "salt") started and became standard earlier (which seems to be supported by the general pattern of dictionary transcriptions; most dictionaries of British English show short pronunciations for "salt" etc. but not for "bolt" etc.) and according to this 2010 post from John Wells the shortening of the vowel in "salt" was not entirely complete, even among younger speakers, in a group he polled some time "recent" relative to the post. I think nowadays short "salt" is pretty clearly SSBE, but my point is that it took some time for this to run to "completion" (by which I mean, not literally 100%, but the point where it's clear that one pronunciation is "standard" and the other is not), and I find it easy to imagine the more recent shortening of "bolt" etc. is currently still in a stage where there is some variation among speakers of varieties that could be called "SSBE". It's not always possible to pick out a single pronunciation that everyone will agree is "standard".

[Vijay]

For the longest time, I thought the first long vowel in both Jawaharlal and Varanasi was on the third syllable, not the second. It still feels very odd to try to pronounce Varanasi the way it's supposed to be pronounced in Hindi. Benares is so much easier.

EDIT: Never mind. The first syllable of Varanasi has a long vowel, too. Well, that's easier!

[Salmoneus]

Interestingly, that link suggests long o in 'bolt' and 'polka', but that's wrong, as the wiktionary entries for those words shows (although it does give both options for 'bolt'). I might sometimes have a long O in 'bolt', but I think that's just non-SSBE parental influence showing through.

I have it in all three.

@StrangerCoug: I think you're a North American English speaker, though, right? That's what would be expected for you. Sal's use of terms like "expected" and "wrong" are all relative to his accent/what he perceives as SSBE.

Because the original link I was responding to was talking about "RP" in the present tense. RP in the present tense, outside of technical contexts, generally refers to SSBE.

To the extent that there has been any substantive disagreement in this thread, it seems to be about whether standard Southern British English accents now unanimously have the shortened pronunciation /ɒl/ before voiceless consonants (which is what Sal is saying) or whether it is still possible to hear /oʊl/ in this context (from an ordinary, non-pretentious speaker; obviously it's not relevant whether this pronunciation is used by people like Sal's dismissive caricature who uses "RP" pronunciations that are well-known to be obsolete in SSBE like "orphan" for "often" and "hoarse" split from "horse"). So far, Sal seems to be the only contributor who actually speaks British English, so there's been no evidence in this thread that /oʊl/ + voiceless consonant exists in standard contemporary British English (as I said earlier, it clearly exists and is standard in North American English).

The only thing that makes me a little skeptical about the idea that /oʊlt/ etc. are now nonexistent in SSBE is that it takes time for sound changes to spread and for more conservative speakers to die out--my impression is that the shortening of THOUGHT in this context (in words like "salt") started and became standard earlier (which seems to be supported by the general pattern of dictionary transcriptions; most dictionaries of British English show short pronunciations for "salt" etc. but not for "bolt" etc.) and according to this 2010 post from John Wells the shortening of the vowel in "salt" was not entirely complete, even among younger speakers, in a group he polled some time "recent" relative to the post. I think nowadays short "salt" is pretty clearly SSBE, but my point is that it took some time for this to run to "completion" (by which I mean, not literally 100%, but the point where it's clear that one pronunciation is "standard" and the other is not), and I find it easy to imagine the more recent shortening of "bolt" etc. is currently still in a stage where there is some variation among speakers of varieties that could be called "SSBE". It's not always possible to pick out a single pronunciation that everyone will agree is "standard".

I wouldn't say that the long-o pronunciations are nonexistent; I'd say they're more common than 'orfen' and the like. But I'd say they're now considerably non-standard in SSBE, and I'd say that many people my age and younger would be unfamiliar with them, or at least would need their attention drawn to them before realising they existed.

I would say, I'm not sure who Wells was surveying for the LPD exactly; it's not necessarily the same group I think of as 'SSBE speakers'. He may be including speakers of other dialects - the post you link to talks about "varieties of English English", not specifically about SSBE. Even if his survey is of modern "RP" speakers, it may not be SSBE strictly speaking, since there are speakers (both native and second-dialect) of derivatives of RP outside of core SSBE territory (and I'm not sure how that's defined exactly anyway). Wells himself, for instance, is from Lancashire, and I wouldn't call the higher register or higher class "RP" of the north a form of modern "SSBE".

So, for example, Wells finds 1/3 of 'young' people with long 'salt' and 2/3rds of older people with it. Whereas my experience would be approximately 0% of young people with it, and even among my parent's generation I don't think it's anywhere like normal. I'm thinking of dinner parties my parents have ocasionally - I think there's one guy in his eighties who might ask for 'sorlt', who has an extremely RP-ish accent, but I think everyone else even of that generation I know would have a short vowel there.

Also interestingly: one of the words originally in dispute was 'polka'... and Wells in that post actually lists 'polka' as having a short vowel. And he's an 80-year-old conservative-RP guy who still says "sorlt"! [Of the other words he mentions: I still have the long vowel in 'extol' and 'scold', but not in 'bolster'. Well, actually I may have long O in 'bolster' when it's used as a noun, because that's a word I probably haven't heard outside of old cassettes of people reading books in my childhood...] His comments below note that 'the short vowel in 'polka' was noted alongside the older long vowel as early as 1917.

Intriguingly, "scald" is a problematic one for me - I can have it as either short or long. [Again, I normally only have that shortening before voiceless consonants, off the top of my head, I think - "bald" is still long. To me, oddly, both uniform long ("sorlt") and uniform short ("bolld") sound 'old-fashioned'...]

[finlay]

do you... pronounce the L in polka? is it not /po:kə/ dots?

also i definitely have long O in bolt, troll, etc, but it's easier to tell in a scottish accent. if you're writing long O as /əʊ/ i have real trouble pronouncing [əʊɫ], like is that how you'd pronounce it?

[Salmoneus]

do you... pronounce the L in polka? is it not /po:kə/ dots?

Yes to the first, and no to the second. In that Wells post, he raises the possibility that some Americans might not pronounce the L, but seems to dismiss it as implausible. However, there's an American and a Canadian in the comments section who report not having the L, and indeed wiktionary brackets the L for its 'US' pronunciation. Maybe (as often seems the case) you just pronounce things the American way?
Regarding 'poker dots' specifically, though, this may just be a folk etymology? Do you pronounce the L when discussion polkas outside of the context of dots?

also i definitely have long O in bolt, troll, etc, but it's easier to tell in a scottish accent.

Ahh, now, "troll" is a different one. If that's been 'shortened', it's not part of the same shift, which doesn't apply before final L outside of the dole/doll merger - that is, although I have a short 'o' in 'polka', I have a long 'o' in 'dole'. I also seem to recall reading that originally the noun and verb had different pronunciations - "troll" meaning "walk" or "fish" (and now "be a dickhead") was long to rhyme with "stroll", being a borrowing from French, whereas the monster, borrowed from north germanic, was short. So it's possible that the variation is due to different analogies? That is, some may lengthen the vowel in the monster to match the verb, and some may shorten it in the verb contrariwise.

if you're writing long O as /əʊ/ i have real trouble pronouncing [əʊɫ], like is that how you'd pronounce it?

I'd say long O varies between /@U/ and /oU/ in 'SSBE', with the former associated either with older RPyer or younger Estuaryer accents. Mine is, I think, closer to /oU/, although not as close as some people's, although of course it's hard to tell exactly. Either way, it's definitely normally a diphthong. However, weird things happen before coda /l/ ('goat' and 'goal' do NOT have the same vowel for me at all, though they're the same phoneme. And I'm honestly not sure 'goal' and 'dole' have the same vowel either, come to think of it, but maybe I've just overthinking it). And I'm not honestly sure how much of what's going on is quality of the preceding vowel and how much is 'darkness' of the lateral.

[Travis B.]

I pronounce polka as [ˈpʰokə(ː)], i.e. with no /l/. Were I to pronounce it /l/ it would be *[ˈpʰoʊ̯kə(ː)].

[Zaarin]

do you... pronounce the L in polka? is it not /po:kə/ dots?

Yes to the first, and no to the second. In that Wells post, he raises the possibility that some Americans might not pronounce the L, but seems to dismiss it as implausible. However, there's an American and a Canadian in the comments section who report not having the L, and indeed wiktionary brackets the L for its 'US' pronunciation. Maybe (as often seems the case) you just pronounce things the American way?
Regarding 'poker dots' specifically, though, this may just be a folk etymology? Do you pronounce the L when discussion polkas outside of the context of dots?

As an American, I'd say having an /l/ in polka sounds extremely alien to me, and I don't have L-vocalization.

[Sumelic]

My natural pronunciation of "polka" as a NA speaker is/po:kə/, and I think I'd always use that in "polka dot". The dance might have /l/ in careful speech, but not necessarily.

As I said earlier, I would guess that for at least some speakers, such as Wells, /ɒlk/ in "polka" is not the result of a shortening sound-change, but simply a fairly straightforward spelling-pronunciation of the sequence "olk~olc" in loanwords. This sequence always represents /oʊk/ with vocalized /l/ in native English words, but this kind of l-vocalization is not productive for new borrowings. There are no native English words with /oʊlk/ (as there are with /oʊlt/ and /oʊld/) to provide analogies for lengthening the vowel and retaining the l. So it might be expected that some people would default to using the pronunciation /ɒlk/, just as /ælk/ is more common than /ɔːlk/ for "alk~alc" in recently borrowed or restored words (like alcohol, alkane, or one of the restored-l pronunciations of falcon).

I think Wells might be incorrect about "extol" traditionally having long o. Walker (1791) gives it with the short vowel of "not", which seems consistent with its etymology from Latin extollere. That said, most contemporary dictionaries of American English show only the long-vowel pronunciation; the only one I'm currently aware of that shows /ɛkˈstɑl/ as a possibility is Collins, and I'm not sure they're that reliable a source for this since they're primarily a British English dictionary (also, if /ɛkˈstɑl/ is at all common, I would expect the variaont /ɛkˈstɔl/ to also exist, because it seems for a number of American English speakers historical "short o" before tautosyllabic /l/ goes into the CLOTH rather than LOT lexical set).

[alynnidalar]

Whaaaaa... there's people who pronounce the "l" in "polka"? I definitely don't, and I genuinely don't remember ever hearing someone pronounce the "l" who wasn't either being intentionally silly, a non-native English speaker/otherwise unfamiliar with the word, or made fun of for it immediately afterward.

For context, I grew up next to a town in which polka music was common (they're very German there), so it legitimately was a word I heard a bit growing up.

EDIT: I don't make any distinction between "polka" (music/dance) and "polka dots".

[Sumelic]

Whaaaaa... there's people who pronounce the "l" in "polka"? I definitely don't, and I genuinely don't remember ever hearing someone pronounce the "l" who wasn't either being intentionally silly, a non-native English speaker/otherwise unfamiliar with the word, or made fun of for it immediately afterward.

Yes. There are even people who have /l/ in "folk".

[Zaarin]

Whaaaaa... there's people who pronounce the "l" in "polka"? I definitely don't, and I genuinely don't remember ever hearing someone pronounce the "l" who wasn't either being intentionally silly, a non-native English speaker/otherwise unfamiliar with the word, or made fun of for it immediately afterward.

Yes. There are even people who have /l/ in "folk".

I actually have a split here where the genre of music has an /l/ and all other uses of the term do not.

[spanick]

I probably include the /l/ in Polka (music) as often as I omit it, although I always omit it from polka dot (honestly, I didn't even realize that how it was spelled until now). Same goes for folk.

If I had to guess, I'd say it's hypercorrection on my part.