#####
##### Guramarasu
#####

The term for consonants appearing at the beginning of a syllable is "onset," because that's where it starts, of course. The rest of the syllable is made up the "rime," itself composed of a nucleus (typically a vowel) and "coda," the consonants that come after the nucleus (like in music). A syllable may be made up of just a nucleus, like the word "I," the consonants before and after it are optional, though crosslinguistically there is a stronger tendency towards having an onset over a coda, preferring "open" syllables with no consonants at the end. If they do have a coda, they're closed.

Let's talk about ⟨/qa/⟩'s onset, the start, because it's its most consistent and defining characteristic. The tetragraph always stands for an onset, but it doesn't necessarily have a nucleus, and it evidently never represents a coda. Its initial consonants can be represented as /k(G)/, where /(G)/ is a glide, a semivowel, either /j/ or /w/. This means its possible forms are /kj/ as standalone ⟨/qa/⟩, /kw/ as ⟨/qa/lity⟩, and a plain /k/ as ⟨/qa/t⟩. However, there's a catch here: /kj/ only ever occurs before /uː/, it's a constraint English has. This makes the first cluster considerably less common than the other two. Also, whenever a glide is present it is never written, so that quality with a ⟨u⟩ becomes ⟨/qa/lity⟩, and only in very marginal one-off cases is this ever not the case, where it can be attributed to a speaker's unfamiliarity with the system.

The nucleus, the vowel(s), can be... basically anything. There's no point talking about it in a vacuum, because it's too context-dependent. There's a marked difference between monophthongs and diphthongs, but I'm leaving that for their respective sections.

That's the essential part. If you want to just see the compiled words with funny symbols, feel free to skip past the following sections, because they're really fucking long. I would, although, recommend reading at least the vowels part since it's the most important one. Still, don't say I didn't warn you.

Vowels, Checked and Free

English has a lot of vowels, about three to five times as usual depending on the analysis. It has so many phonemes, and so many divergences, that people use shorthands such as "the PALM vowel," CLOTH, KIT, PUT, and several others, always with CRUISE CONTROL: "Water and want may have the SPA vowel, the LAW vowel, or the LOCK vowel," says Kreidler's The Pronunciation of English. Dealing with them makes me want to die.

Specifically, Kreidler writes that
>In English some vowels occur with a consonant following or without a consonant following, for example, the vowel of feet and fee. Other vowels never occur in a one-syllable word without a consonant following, for example the vowel of fit. We call the first kind free vowels and the second kind checked vowels.
You got that? No? Well, here's an example:
pit, pet, pat, pot, cut, put
pɪt, pɛt, pæt, pɒt, kʌt, pʊt
Those are the six checked vowels, Kreidler refers to them as CHICK, STEP, BAT, LOCK, NUT, and FOOT, in that order. (Yes, the terms used change depending on the author.) They're all short, largely limited to closed syllables, and often get priority when compared to a following weak syllable, snatching its onset to use as a coda. (You'll see what I mean later.) /ɪ, ʊ/ in particular are like /i, u/ but lax, more towards the center, and it's so common for them to be further reduced into a schwa that Oxford made two special characters to represent their ambiguity: /ᵻ, ᵿ/. Oh, and the scha is /ə/, the COMMA vowel. You better get acquainted with it, it's everywhere.

Free vowels are, essentially, the ones that are long, and diphthongs, which kind of act like vocalic affricates rather than simple combinations.
They're often called "long," but it's not like in Japanese where /oː/ is 50% longer than /o/ with no real change in quality ([o̞] vs [o̞ː]), no, in English they're seriously messy. Gimson's describes some realizations of /ɛː/ (SQUARE, actually an r-vowel which is a different thing) as not just [ɛː] but also [ɛə], [ɛᵊ], and [æɐ], in General British alone. That's in section 8.9.4 titled "(Relatively) pure vowels," you cannot make this shit up. Length... doesn't even matter that much, to the point that bid and beat are almost of the same length, even in non-rhotic dialects which make a markedly greater distinction in duration overall.

The important part is that while checked vowels mostly appear in closed syllables, free ones appear in open ones as well, extremely frequently but not necessarily. Orthographically, one of the things English uses to distinguish between them is adding silent letters around the main vowel character, like the classic E or a mute doubled consonant. Compare man/mane, manned/maned. Sheer fucking lunacy frankly, but at least it's a consistent kind of nonsense.

Here's a simplified chart for Standard Southern British English with a dropdown to change notation and names with:
https://improveyouraccent.co.uk/vowel-chart/

Stress

Something we cannot gloss over is the importance of stress in English, there's an entire class of word pairs where stress is shifted to differentiate between a verb and a noun, and here you can see the difference in full effect: CONsole (noun) /ˈkɑn.soʊl/ versus conSOLE (verb) /kənˈsoʊl/. The free vowel from the second syllable remains alive and well in both cases, but the first checked one gets schwa'd when stress changes. Not just that, English words that have three syllables or more can actually have a SECOND stressed syllable, primary and secondary. You can see it in words like these, secondary being marked with /ˌ/:
'archiˌtect ˌengi'neer 'eleˌvator ˌele'vation

So, in AHRchitÉct, the first part is considerably louder and longer than the rest, the middle part survives thanks to God's infinite mercy (or gets shafted into a schwa by heretics), and the last -tect is given enough force to be modulated like a decent segment. Its structure is STRONG-weak-strong, or just S-w-s. You can have subsequent strong syllables in compounds like 'keyˌhole and 'stopˌlight, but most often they alternate with at least one weak segment inbetween strong ones. It's something English really likes to do, like in Shakespeare's iambic pentameter. (In fact, Kreidler uses the same measuring unit, the metrical "foot" to describe its prosody.)
Something to take note of is that when adding suffixes stress can gradually move towards the end of the word, changing up which vowels get reduced and how. For example, there's origin /ˈɒrɪdʒɪn/, original /əˈrɪdʒənl/, and originality /əˌrɪdʒəˈnæləti/.

Syllable Structure

Now, how do boundaries come into play? I've said before that English tends towards closed syllables more than other languages, as in the case of Italian segue /se.gwe/ adapted as /ˈsɛɡ.weɪ/, and we can see how it not only split the cluster but also turned its vowels into checked and free respectively, it's a simultaneous thing. Kreidler presents the following list of differing boundaries:

ra-di-um, fla-vor but rad-i-cal, sav-age
le-gal, me-di-um but leg-a-cy, med-i-cine
fi-nal, ri-val but fin-ish, riv-er
co-pi-ous, so-lo but op-er-a, sol-id
pu-pil, stu-dent but pun-ish, stud-y

If you've paid attention, you can see that the set on the left starts off with free vowels, while on the right it's checked ones. That's the convention dictionaries have used for transcribing boundaries, for a long time. However, he brings it up in order to argue against it, writing:
>The fact is that, in speech, a single consonant between a strong vowel – any strong vowel – and a weak vowel is ambisyllabic: the consonant does not clearly belong to either the strong syllable or the weak syllable. There is no clear syllable break.
That is to say, a single consonant ambiguously serves as both coda and onset. Two of the main rules worth pointing out is that if the second syllable is strong (but not necessarily stressed), or if it's a cluster that can appear at the start of a word (like /str-/ in destroy /dɪˈstɹɔɪ/), then the consonant is not ambisyllabic and just sticks to the latter vowel as onset.

That said, I've included this part only to illustrate that English syllables are pretty iffy, Gimson's considers the idea but ultimately rejects it:
>Phonetically this seems a credible solution. Unfortunately it would considerably complicate the overall statement of permissible clusters.
There's also https://doi.org/10.1016/j.langsci.2017.03.002" rel="nofollow" target="_blank" >this opposing 2017 paper which concludes that "word-medial ambisyllabic consonants pattern exactly with word-medial codas [...] following both tense and lax vowel contexts at all speech rates." I'll continue to follow the standards set by Oxford and Cambridge because the alternative is to write ci^ne^ma rather than cin.e.ma in a way that's all in all unnecessary and I'm not going to do that shit because it's way too idiosyncratic, sorry. So yeah, I'm sticking with that, this is all to give you an idea of just how dense this bullshit is.

We also could get into the nitty gritty of what leads /kw/ to become [kʰw̥] but frankly FUCK THAT SHIT, I am NOT getting into motherfucking aspiration right now it's too much, it doesn't matter, and nobody cares.

#####
##### Monophthongs
#####

Monophthongs have a general trend towards being open, sitting at the bottom of the mouth, save for the schwa which is central. All of these sounds are similar to the single /a/ phoneme found in other languages, such as Japanese, in for example kamikaze's /ˌkæmɪˈkɑːzi/ (also ⟨/qa/mikaze⟩ (2)) and viceversa. Also visible in taco, parmesan, aficionado, etc.

~ /æ/
¥⟨/qa/t⟩ (48), ⟨/qa/ptain⟩ (18), ⟨/qa/pitans⟩ (4), /qa/ng (7) [both gang and kang], ⟨/qa/ncer⟩ (136), ⟨/qa/ggot⟩ (68)
Extremely straightforward, basically THE a-sound anyone first thinks of, it's not even necessary to provide a transcription. All closed and stressed syllables.

~ /ɑ, ɒ/
¥⟨/qa/mrade⟩ (17), ⟨/qa/id⟩ (9), ⟨/qa/di⟩ (6)
I'm here bunching together the PALM and LOT/LOCK vowels, which are free and checked respectively, regardless of length because of the father-bother merger affecting a great deal of dialects, which has combined the two into an unrounded /ɑ/ that is actually moved quite a bit forward in many cases. It's rather complicated, this video explains it alongside another intersecting merger:
https://www.youtube.com/watch?v=sALAE6iqLWk
Moving on, comrade starts with /kɒm-, kɑːm-/, depending on the dialect (like calm). The other two words are rather obscure Arabic terms, the first one (also written as caïd) has a hiatus in the middle as it's actually pronounced /kɑːˈiːd/ or /ˈkɑː.ɪd/, while qadi is simply /ˈkɑːdi/.
There's also ⟨/qa/mpadre⟩ (4), Cambridge says it uses this vowel too, but according to Oxford it's a schwa. However, the former's example is /kɒm.ˈpɑː.dreɪ/ where it's clearly emphasizing all vowels as if they were free in order to imitate Spanish pronunciation and in this case I think Oxford's /kəmˈpɑːdreɪ/ with a reduced starter is more realistic. Still leaving it here as optional, though.

~ /ə/
¥⟨/qa/bal⟩ (40) /kəˈbæl, -'bɑːl/, ⟨/qa/tari⟩ (26) /kəˈtɑːri/, ⟨/qa/det⟩ (21) /kəˈdɛt/, bur/qa/ (8) /bɜː(ɹ)kə/
Here, the vowel is non-negotiably a reduced schwa, sadly in two very common and characteristic words. Thankfully, we also have ⟨/qa/b(b)alist⟩ (20), where secondary stress is given to the first syllable, so it switches to /ˌkæb.ə'lɪst/, back to /æ/. Plus ⟨/qa/bal(l)ah⟩ (3) which could go either way in terms of stress placement.

~ /ʌ/
¥⟨/qa/mblr⟩ (188) /ˈkʌmblə(r)/, ⟨/qa/nt⟩ (3) /kʌnt/, ⟨/qa/ck⟩ [the bad one] (uncertain, more common as plural) /kʌk/
This one may be a bit controversial. You could perfectly argue that no, it's actually one of the vowels previously mentioned above, but acoustically I see it as legit. When I said "viceversa" at the start, I did so with this one in mind because it is indeed adapted by Japs (and me) as /a/ so that cat and cut, run and ran, they sound the same. It also sits below /ə/ in a comfortably near-open position besides [ɐ] and right above [ä] that are present in very many English dialects. I consider it a valid reading given how widespread the word was, and the fact that the other two words also exist, so it's not an isolated incident.
While writing this, it dawned on me that it'd mean ⟨/qa/nny⟩ is perfectly legal. Make of that what you will.

There's also ⟨/qa/psie⟩ (7), obviously by analogy with /jp/sie which I'd assume has an /æ/ but I'm not sure honestly. Just gonna leave it on its own. Here are ones uncommon and unimportant enough to not make the cut:
~ 2-3 instances
¥/qa/mpers, /qa/racter, /qa/balleros, /qa/tamites, (c/k)um/qa/t, /qa/strati, ni(q)/qa/(r)
~ 1 instance
¥ot/qa/ku, fu/qa/ps [with /ʌ/], i/qa/nas, delin/qa/nts, mendi/qa/nts, /qa/avalrymen, /qa/ucasian, /qa/ffles, /qa/ntoku, /qa/ndidates, /qa/smonauts, /qa/tastrophe, /qa/strates



#####
##### Diphthongs and Triphthongs
#####

Compounds that either start with the glide /w/ and/or are a combination of two distinct vowels. These terms are not universally applied in this way and in many cases refer purely to vowel combinations while excluding semivowels, it generally depends on both the language in question and the author. I'm simply grouping them in this way because I find it convenient. A distinctive feature is that they allow for the presence of some close front vowels, namely /e, i, ɪ/, although there isn't much to say about them besides that.

~ ⟨/qa/lity⟩ (205) /ˈkwɒləti, ˈkwɑ(ː)-/
A classic, most common one by far. Even though it's transcribed as long (in GA but not GB, oddly enough), I see it being classified under the LOT vowel. It admittedly confuses me a bit.

~ ⟨al-/qa/eda⟩ (130) /ɑlˈkaɪdə/
Often noted to be a collective noun, one of the few that exist. Sole word featuring the PRICE vowel.

~ ⟨/qa/y⟩ (27) /keɪ/
The gay one. You gay? I'm gay, but not flamboyantly so. Uses the FACE vowel.

~ ⟨/qa/ck⟩ (uncertain, more common as singular) [the good one]
Simple TRAP vowel, /kwæk/.

~ ⟨/qa/eer⟩ (86)
In spite of having a two written vowels, it's actually a short /kwɪr/, or the gliding /kwɪə(r)/. The NEAR vowel, only known example of /kwɪ-/ in the entire corpus. Also notable in that the tetragraph lacks a nucleus.

~ Triphthongs, /kweɪ-/
There are two confirmed triphthongs I found that start with this combination, consisting of ⟨/qa/sar⟩ (12) /ˈkweɪzɑː(r)/ and ⟨/qa/ker⟩ (24) /ˈkweɪkə(r)/. They stand out to me in that unlike the above they aren't aided by any extra vowels, ⟨-a/⟩+context is doing all the heavy lifting. Pretty neat.

Honorary mention to some other marginal dipths:
¥/qa/rks (3), a/qa/men (1), a/qa/rian (1)



#####
##### Japanese Loans
#####

I'm giving these a special spot due to the fact that loans usually have two sets of possible pronunciations: those that approximate the original, and those that better conform to local rules. Once more, borrowings overwhelmingly use the PALM or TRAP vowel depending on prosody. That said, let's get started.

Na/qa/ma (12) is in Japanese a very straightforward /nakama/ [na̠ka̠ma̠], but in naïve English I'd assume /'næk.ə.mə/, like spatula and stamina which are indeed mostly transcribed as ending in /-əCə/. ⟨/qa/waii⟩ (6) (someone wrote ⟨Kah-why-ee⟩ for the standalone) is likewise reduced to /kəˈwaɪ.(i)/ (or an exaggerated /-ɪ.iː/), which Japanese teachers usually warn to not confuse with kowai. An alternative pronunciation that tries to avoid reduction would be /kɑːwɑːˈiː/ and perhaps /'nɑːkəmɑː/, but that last one's speculation on my part. There's a ton of people using it in writing, speech not so much, it's hard to get a sample. ⟨/qa/zoku⟩ (11) most certainly has initial stress given that it's atamadaka, but I can't tell whether it'd be /kæz-/ or /kɑːz-/ though I'm leaning towards the former.

There's another one that stands out, which is ⟨/qa/mono⟩ (33). Again, Jap is very straightforward: /kemono/ [ke̞mo̞no̞]. However, this is a mid front vowel, which so far we haven't seen anywhere else. We haven't even seen /ɛ/, the DRESS vowel, which is below it and right above /æ/. From a preeliminary look at some anitubers' pronunciation of Kemono Friends and Kemono Jihen, I have transcribed their utterances as /kə'moʊnoʊ/. In fact, in one of the reviews I looked at the auto-generated subs picked it up as kimono, which IS INDEED HOMOPHONOUS. Hypothetically, a better speaker could say it like chemical or something like that but at this point my faith has been depleted come on you guys. ⟨/qa/ko⟩ appears about four times too, I'm just putting that out there.

Finally, an interesting hapax legomenon I came across: ⟨ya/qa/za⟩, at No.1343584. Same story as before, Jap [ja̠kɯ̟ᵝza̠], English either /jəˈkuːzə/ or perhaps /ˈjækuːzɑː/. In both cases we'd have a genuinely fascinating case in our hands, the only attested compound where ⟨/qa/⟩ has /u/ for a nucleus.

In any case, please take this as a reminder to work on your pronunciation, onegai shimass.



#####
##### Others
#####

Let's look at another crucial entry now: ⟨/qa/tar⟩.
While ⟨/qa/tari⟩ is rather clear-cut, the root word itself is fairly problematic and has a wide range of readings. According to Cambridge and Oxford, among these are /kəˈtɑːr/, /ˈkɑːtɑːr/, /ˈkætɑːr/, /ˈkʌt.ɑːr/, and apparently some people even pronounce it with an initial /g/? A speaker of I believe Northern American English was surveyed on this matter and his answer was "kwaytar," like ⟨/qa/sar⟩ above. This shameful display of illiteracy not only does not help us, it makes everything worse, although it serves as a good display of how irregularly the word gets read.
Really, I'm just... not gonna look into Gulf Arabic for the sake of a word whose actual pronunciation absolutely no one knows, I'll pass on that one. It's shame, but it simply can't be helped.

~ ⟨/qa/ddit⟩ (503)
Special shoutout to this one because it was insanely popular, and honestly I'm a bit baffled by how it's supposed to be pronounced. In thread No.944186, ⟨qua-dit⟩ and ⟨kwadit⟩ are suggested as well as ⟨qwaddit⟩ in No.2098436, so clearly with an /kwɑ/, which makes me inclined to believe it's not at all the DRESS /ɛ/ from /ˈrɛdɪt/. Maybe it even has a checked /æ/ in its place, who knows.

~ ⟨/qa/non⟩ (419, but only 53 prior to 2020)
This one's ambiguous because it could either be k'non /kəˈnɒn/ or cue-anon /ˌkjuː əˈnɒn/ depending on whether you pronounce the initial as a standalone Q or not... but standalone has negative connotations so let's go with /kə-/.
Another ambiguous case is ⟨/qa/liph⟩, which could be pronunced with either /æ/ or /eɪ/.


Finally, I want to present a hypothetical word: *⟨/qa/rn⟩, with the /ɔ/ WAR vowel. It feels feasible, but I don't have anything to back it up, and it might as well be pronounced with PALM same as arm. Really, there's a lack of stressed /-r/ in general, no *⟨/qa/r⟩ or *⟨/qa/rt⟩ either.

##########
########## Range of Realizations in Standalone Readings
##########

Here's where shit gets real. Now we're dealing with the tetragraph by itself, and we can no longer rely on dictionaries to help us. All we have to work off are the speaker's own transcriptions, which vary as wildly and crudely as English orthography itself, and then some. This is hard mode.

Our mainstream, canonical reading for a standalone ⟨/qa/⟩ follows the initialism QA, that is to say, ⟨Q⟩ = /kjuː, kju/ + ⟨A⟩ = /eɪ/, so /ˌkjuːˈeɪ/. Rhymes with gay, again the FACE, and starts with the extremely aptly named CUTE vowel. It's special in that it comprises the longest possible reading, and is disyllablic, most others are monosyllabic. However, it's all downhill from here. It has been written as:
¥⟨queue ay(y|e)⟩ (126), ⟨cue ay(y)(y)⟩ (64), ⟨cue aye⟩ (7), ⟨cue ei⟩ (1), ⟨cue a⟩ (3) [presumably], ⟨cue æ⟩ (1) [No.2137823]
31 use ⟨kyuu⟩, while 71 use ⟨kyoo⟩, of which 62 follow it up with /'eɪ/ in some form. The latter transcription in particular was highly favored by Akari. Ten other posts write ⟨kju⟩ instead. This reading has been suitably Japanized as as ⟨kyuuei⟩ et al, represented with the kana きゅうえい/キューエイ/etc or the ateji 久永 and 旧詠, although I don't actually expect anyone to to correctly pronounce it as [kʲɨᵝːe̞i]. All in all, over 300 instances and more when taking into account the ones I've missed. Not too shabby.
In terms of more divergent variants, we have three cases of ⟨cueway⟩, ⟨kyuu way⟩, and ⟨kyoo way⟩ which suggest /ˌkjuˈweɪ/, not at all unreasonable and even to be expected. There's a variant where people read the board name as Q&A, adding an /n/ in the middle, but their transcriptions are too irregular to properly take into account. ⟨queue eh⟩ (9) and ⟨cue eh⟩ (2) are, I would imagine, taking /eɪ/ and making it into /eː/, which is a common realization for that cluster.


The other main readings are monosyllabic, having in common that they switch to /kw-/, and take either the FACE or accursed hellish corner vowels. The OP of No.2044911 (and likely No.2284575) asks precisely this, which of the two to transcribe it as (curiously employing /J/ in their analysis, kudos to them). Surprisingly, (forgive me) the soyjak wiki agrees as well:
>(pronounced "queayy" or "qwah", sometimes referred to as "kway")
Regarding FACE, we have:
¥⟨quay⟩ (138), ⟨kway⟩ (131), ⟨qway⟩ (41) and ⟨kwey⟩ (9)
Again over 300 cases, a worthy rival to the initialism. Tied to it there's the modified ⟨qgay⟩ (42), with and without slashes, where a punny stop is inserted (or replaced) in the middle.
I do need to make a sad announcement... quay... is pronounced as /kiː/.


As for the second set, a different Anonymous comments "As qua in quaff," which I think falls under CLOTH, actually? Because it ends in a voiceless fricative. This would mean they have the caught-cot merger once more alongside father-bother, but not the LOT-CLOTH split. It's a terrible answer honestly, I hate this shit. A better example is ⟨qua⟩ (44), sometimes appearing as ⟨quâ⟩, and the identical ⟨qwa⟩ (24), indeed using PALM. On top of that, ⟨kwa(h)⟩ specifically was in the minority before 2020, having less than forty total posts. But in that year onwards following the soy takeover there was a tenfold increase, post-mortem hits being at about 400:
13+19 = 32, pre-2020
342+32 = 374, post-2020
374+32 = 406, total
Less than 10% are pre-soy, interestingly enough. I do wonder if they continue to use it over there, who knows. Moving on, under French influence (because somehow it had to get worse) we find the related pronunciation /kʁwa/, in both the local post >>>/jp/59230 (RIP) and No.3718432's croissant, while No.4788281 writes it out explicitly. Also ⟨/croix/⟩ (4). No.4937245 writes "i say kwa, with the a being like japanese あ、 not ay," but... I'm not sure I should believe them, frankly, and I'm tossing them into the general labialized pit anyways.


Then there are several with a plain /k/ onset, which sound like odd one-offs until you take a better look at how frequent they are:
¥⟨ka⟩ (23), ⟨kah⟩ (14), ⟨kaah⟩ (3), ⟨kha⟩ (2)
For a total of 42, solid stuff, though the first one was such a massive pain in the ass to account for that my recount may be wrong. Anyways, for native English speakers these are mostly PALM as in other loanwords, one poster says "Ka (like car)," maybe a non-rhotic speaker but I'm not totally sure. I'd assume ⟨ka⟩ is mostly made by non-natives who have a real /a/ phoneme. Two anonymouses think the whole thing is bunk, like No.4136704:
>Why is it /qa/ and not /qua/ anyway? It's not grammatical.
And No.4619311 much later:
>No /qa/ is unpronounceable as a syllable unless it is borrowed from a foreign language. [...] Qua is how it would have to be spelled to be pronounce as a syllable in the English orthography.
Too bad, sucker. These guys are clearly fucking wrong and should readjust their approaches accordingly.


Finally, let's move on to the miscellaneous. First, ⟨keway⟩ (10), I really don't know how to make sense of this one. It'd seem to me like it points towards boundary shifting that makes it go from /ˌkjuːˈeɪ/ to /kɪ.weɪ/, as in /j/ -⟩ /ɪ/, it's not a far cry from the ⟨cueway⟩ above. But uhhh not sure, not sure. In other instances a haitch may be inserted inbetween the mainstream reading, giving way to two instances of ⟨kyu hey⟩, ⟨kyu-hay⟩. And, personally, I natively pronounce it without any glides, with a /u/, and with a hiatus: /ku.'a/. The first syllable seems to be shared by other one-offs in some shape or form:
¥⟨koo-waa⟩, ⟨koo-AH⟩, ⟨koo ayy⟩, ⟨kuu ai⟩, ⟨cooh uh⟩, ⟨クア⟩ [No.2488241], ⟨kuh ah⟩ (2) [possibly]
There are instances of fronted affricates ⟨chyuu ayyy, chyue-aie, chewy⟩ in No.2335139 (though the second two are a joke based on the first), and in other places ⟨chìù¬éì⟩, ⟨chyew ey⟩. I'm not gonna bother trying to explain these. Finally, Some other shit I am not going to fucking bother with because it's either way too idiosyncratic or straight-up taking the piss:
¥⟨qua whey⟩, ⟨きゃ⟩ [No.2044958], ⟨key⟩ [No.3950437], ⟨クユエ⟩, ⟨kiyuwe⟩, ⟨kay⟩ [No.63661], ⟨kyewway⟩, ⟨kyoo Ayuu⟩, ⟨Kewway⟩, ⟨kew-aeh⟩ [maybe just written badly], ⟨key-u ayy⟩, ⟨kyewway⟩, ⟨kve-a⟩, ⟨caw⟩ [suck my cawk], ⟨kuhway⟩, ⟨kew aiee⟩, ⟨keway⟩, ⟨kyew æ⟩, ⟨'kea' with japanese pronounce⟩, ⟨ki ayy⟩, ⟨quaie⟩, ⟨カ⟩, ⟨que as in what, the beginning of a question in spanishese!⟩


And... that's it, thank God. I think we're done now, hoo...

In summary, ⟨/qa/⟩ basically covers half or more of the fukken vowel chart and is shockingly flexible. Here is a simplified list of it in compounds:

Monophthongs:
/æ/ = /qa/t
/ə/ = /qa/bal
/ɑ, ɒ/ = /qa/mrade
/ʌ/ = /qa/mblr
/u/ = ya/qa/za
/ɔ/ = */qa/rn
/?/ = /qa/mono

Diphtongs:
/kwæ/ = /qa/ck
/kwɪ/ = /qa/eer
/kwɑ, kwɒ/ = /qa/lity
/kaɪ/ = al-/qa/eda
/keɪ/ = /qa/y

Triphthong:
/kweɪ/ = /qa/ker

As a parting message, please pray for No.2818628. I hope God will save him from the spooky approximant and uvular consonants.

It's not as long as it seems, there's a lot of line breaks.

It was pronounced cutter until a couple of years ago

/qa/nsider the following:

>>116907
It was pronounced "cutter" until recently

>>116915
/qa/nsider it /qa/nsidered.
>>116913
>>116916
In what dialect?

I still like al-/qa/eda as the ethnonym for all /qa/ related things.

hehe, dip thong.

>>116939
it's a dip, you hthong

I.. I'm sorry. I tried to read through this, but it's like.. it's like advanced math being explained to an elementary school kid when there's a cool dog outside the window. I read through the sentences but no neurons were linked inside my head.

>~ ⟨/qa/non⟩ (419, but only 53 prior to 2020)
>This one's ambiguous because it could either be k'non /kəˈnɒn/ or cue-anon /ˌkjuː əˈnɒn/
KUON! It should be pronounced Kuon! OH MY GOD THE SIGNS WERE THERE ALL ALONG!

Personally, I liked /qa/mono friends since 4/qa/'s rebirth and the release of Kemono Friends somehow happened at the same time.

>>117094
>KUON! It should be pronounced Kuon!
Okay, but do you mean koo'ohn, kuwon, or kwon? Could be any of them!

Anyways, basically the point is that ⟨/qa/⟩ is used for a bajillion different sounds not because people are stupid but due to a convergence of how English works and how it's written, it's got a certain regularity to it even though there's a dozen rules behind it. The words with CRUISE CONTROL tell you what sound each part is referring to.
>I liked /qa/mono friends
Yeah, it's a special one. For you, what does it rhyme with?

>>117109
>Okay, but do you mean koo'ohn, kuwon, or kwon? Could be any of them!
THOSE ARE ALL THE SAME!


>For you, what does it rhyme with?
Um... uhhh... what rhymes with kemono friends... See mono trends?

>>117119
>See mono trends?
Ooops, no not "see", I was trying too hard to form something legible and forgot 'see' doesn't make the same sound.
Uhh... "Uh mono spends?"

>>117119
>THOSE ARE ALL THE SAME!
WHAT DO YOU MEAN NOOOOO
nnoOOooOOoOOOo...
Well... I guess Lindsey was right. See https://www.youtube.com/watch?v=gtnlGH055TA for a relatively simple explanation.

>See mono trends?
So like key mow-
>>117120
>"Uh mono spends?"
NOOOOooOOoo okay then the transcription was mostly accurate, /kə'mɑnoʊ/ rather than /kə'moʊnoʊ/ it seems.
Na/qa/ma, does that one rhyme with stamina or spatula? Worth asking about it too.

>¥⟨/qa/mblr⟩
Always thought this was pronounced "kwambler"

How the hell were you able to assemble and collect all these various namings for /qa/... Did you use a program or search through every single post on /qa/ and just noted down when each occurred? That's some real dedication. I appreciate the effort to completely go over every single possible pronunciation of /qa/ and especially explain why the post-2d/random one is WRONG and /geɪ/.

If there were to be one issue I'd take up with this thread though, it forgets one of the more abstract yet widely seen spellings that seems to slip everyone's mind, and it's a name that everyone should know too. I feel it's so in the open that nobody even considered it, since it took me a year or so to figure out myself. I assume it's french or some other sort of insane language ruling.

>>119284
Or maybe it's also japanese in the sense of it being シウイエ...

To anyone that this could prove useful for: here's a four-page list of 27 sets for British English with dozens of examples for each, and a couple notes concerning General American. It's very useful if you want to get a more intuitive feel for what the vowel refers to.

>>119284
I found all of these manually, they have a tendency to cluster together, either into a chain that throws a few different names around or as an OP dedicated to it. There's a solid amount of "how do you pronounce /qa/" and "how do we call ourselves" threads floating around. I appreciate that you managed to read through it.
>explain why the post-2d/random one is WRONG and /geɪ/.
Well, ackshually, I wouldn't say it's wrong. More like it's a variant, one that goes back to Feb 2017 and which overtook the mainstream after a ~sociocultural shift~. The pronunciation is backed up by your reading of ⟨/qa/mblr⟩ as kwambler (with LOT as in quantum or swan, I take it) and the ambiguity of ⟨/qa/ck⟩ in the same bracket. Shit's too messy.

>I assume it's french or some other sort of insane language ruling.
Hmmmm... I did always read it as /si.u.ie/ because of all of these changes, but now that you mention it, in Latin I believe it'd be close to /ki.u.ie/, kinda like in cuius, commercium, iudex and Iesu, pretty close to /kju.ei/. Not sure, maybe the man in question can explain.