English questions

[Travis B.]

In the dialect here, forget and forgive often have the for- prefixes change to /frə/ [fʁˤəː], presumably to avoid [ɔːʁˤɡ] or [ʁ̩ˤːɡ]. I have not heard of this being noted anywhere as happening in any English varieties; this is just from my own personal observation. Is anyone else aware of this happening in any other English varieties? I am interested in finding out how widespread this change is.

I don't recall hearing this. However, my father (Baltimore, MD) did have sporadic /pɚ/ > /prɹ/ in words like perception.

Pronouncing perception with /prə/ [pʰʁ̥ˤə] does not sound wrong to me.

[vlad]

I have /prə/ rather than /pə/ in perform, performance, etc.

But it's an isolated example. Similar words like perceive, percussion, permission, perplex, persist, persuade are unaffected.

[Travis B.]

I have /prə/ rather than /pə/ in perform, performance, etc.

As do I, actually.

But it's an isolated example. Similar words like perceive, percussion, permission, perplex, persist, persuade are unaffected.

All these have /pər/ for me.

[Raphael]

Inspired by some current posts in the Conlangery Forum, I wonder: Does the distinction between nouns, verbs, and adjectives in English even make sense? It seems that words of each of those types can be, and often are, easily turned into words of each of the other two types.

[bradrn]

Inspired by some current posts in the Conlangery Forum, I wonder: Does the distinction between nouns, verbs, and adjectives in English even make sense? It seems that words of each of those types can be, and often are, easily turned into words of each of the other two types.

Yes, English has a strong word class distinction. The complicating factor is extensive zero-derivation: but we do know that it's a derivational process, because the semantic correspondences are so arbitrary.

(I can write more when I have more time, sorry.)

[Glass Half Baked]

I have a question about yod-coalescence. In dialects with yod-coalescence, "sure" and "sugar" are pronounced with an SH sound. But "super" is not.
I have run through every reason for this I can think of, and have come up with nothing.
1. Could it be because super is a more recent loan? No. It is attested in Middle English, before yod-coalescence happened.
2. Maybe it just never had a palatal glide? No. Sugar and sure didn't have palatal glides originally either, but they acquired them through phonological processes that should have also happened to super.
3. Perhaps it lost its glide? No. Some dialects still pronounce a glide after the s. At least those dialects should have had yod-coalescence. This yod is unlikely to be a recent innovation, because that would raise the question of where a rogue u came from that was unaffected by the great vowel shift.

Can anyone shed a little light on this mystery for me?

[sasasha]

I have a question about yod-coalescence. In dialects with yod-coalescence, "sure" and "sugar" are pronounced with an SH sound. But "super" is not.
I have run through every reason for this I can think of, and have come up with nothing.
1. Could it be because super is a more recent loan? No. It is attested in Middle English, before yod-coalescence happened.
2. Maybe it just never had a palatal glide? No. Sugar and sure didn't have palatal glides originally either, but they acquired them through phonological processes that should have also happened to super.
3. Perhaps it lost its glide? No. Some dialects still pronounce a glide after the s. At least those dialects should have had yod-coalescence. This yod is unlikely to be a recent innovation, because that would raise the question of where a rogue u came from that was unaffected by the great vowel shift.

Can anyone shed a little light on this mystery for me?

I can’t fuly illuminate you... But I imagine it’s relevant that, for me, super shares /uː/ with soon, suitor etc., none of which have a palatal glide ‒ whereas sugar has /ʊ/ and sure /ɔː/ (RP /ʊ͜ə/).

In (very much) older/upper class British speech you might indeed hear glides in super and suitor, but never /ʃ/, and never (I think) in soon. Meanwhile I am quite certain I have never heard sugar with /sj/ rather than /ʃ/, even in old recordings, whilst for sure I have heard /sj/.

I don’t know off the top of my head precisely which conditions brought about the palatal glide on /uː/ or yod-coalescence, but I can tell you that sound change is lexically diffuse. Sound changes move through the lexicon in each idiolect spreading from word to word over time: it is not the case that a speaker suddenly produces glides in every /uː/ one day, or /ʃ/ in every word where they used to have /sj/.

The sociolinguistic factors that feed sound change sometimes shut down before the change completes diffusing through a lexicon. Sometimes they complete for some speakers / speech communities and not for others. Sometimes they get interrupted or reversed by other sound changes, again, for some, but not necessarily all, speech communities.

To my understanding, we often get left with these sorts of conundrums as a result.

[anteallach]

I think it's really sure and sugar which are the exceptions here; the normal development of initial /sj/ is yod-dropping, or in some conservative dialects retention, not coalescence. My suspicion would be historical dialect mixing or, as sasasha says, a sound change which got started but never diffused beyond a handful of words.

The evolution of sugar is weird in other ways as well.

[jcb]

The key thing is that most North Americans at least to my knowledge have no idea that anyone makes a distinction between two vowels in these words. (I didn't until I was well into adulthood.)

And then a selecti quidem of us grew up in a city with at START-NORTH merger and didn't realise their pronunciation of words like "forest" and "horror" was not GA until high school.

Apparently many people with the cot-caught merger have little to no awareness that other people have such a distinction, as strange as it may seem to cot-caught-unmerged people like myself.

Indeed, I thought that the spelling distinction between <o> and <au> was just an aesthetic spelling difference, like <ee> and <ea>, or <oCe> and <oa>, and nothing more. (Is there any dialect left in the English-speaking world that hasn't merged REED and READ, or LODE and LOAD?)

Even after nearly two decades of learning about linguistics, I still struggle to hear the difference between /A/ and /O/ (and /a/ and /Q/ that some dialects have). How non-COT-CAUGHT mergerers do it on a daily basis still sorta baffles me.

[sasasha]

Even after nearly two decades of learning about linguistics, I still struggle to hear the difference between /A/ and /O/ (and /a/ and /Q/ that other dialects have). How non-COT-CAUGHT mergerers do it on a daily basis still sorta baffles me.

It sort of comforts me that /ɑ/ and /ɔ/ still sound near identical to you, because of my different but somewhat relateable experience of learning Finnish and failing on a regular, long-term basis to hear accurately their /æ ɑ/ distinction, despite, of course, having one of my own. (The Finnish pair are generally a bit more central than mine.) I can hear it, but I often get tripped up trying to follow it (or produce it) in more rapid speech.

For me, FWIW, there’s a very significant length distinction betwen COT and CAUGHT.

[jcb]

Even after nearly two decades of learning about linguistics, I still struggle to hear the difference between /A/ and /O/ (and /a/ and /Q/ that other dialects have). How non-COT-CAUGHT mergerers do it on a daily basis still sorta baffles me.

It sort of comforts me that /ɑ/ and /ɔ/ still sound near identical to you, because of my different but somewhat relateable experience of learning Finnish and failing on a regular, long-term basis to hear accurately their /æ ɑ/ distinction, despite, of course, having one of my own. (The Finnish pair are generally a bit more central than mine.) I can hear it, but I often get tripped up trying to follow it (or produce it) in more rapid speech.

For me, FWIW, there’s a very significant length distinction betwen COT and CAUGHT.

(1) And I'm even a native English speaker! (Just one that merges COT-CAUGHT.)
(2) If you asked me to distinguish between /A/ and /O/ in an isolated word (COT or CAUGHT) without any sentence context, I very probably wouldn't be able to do it with any greater chance than just guessing. I would only be able to do it reliably by remembering the spelling.
(3) Oh dear, another length distinction in some English dialect that I'm unaware of? How does this one work? (Mine are both short vowels.) (I just learned about the Scottish length rule/distinction in my BARD-BARRED thread.)

[Travis B.]

Even after nearly two decades of learning about linguistics, I still struggle to hear the difference between /A/ and /O/ (and /a/ and /Q/ that other dialects have). How non-COT-CAUGHT mergerers do it on a daily basis still sorta baffles me.

It sort of comforts me that /ɑ/ and /ɔ/ still sound near identical to you, because of my different but somewhat relateable experience of learning Finnish and failing on a regular, long-term basis to hear accurately their /æ ɑ/ distinction, despite, of course, having one of my own. (The Finnish pair are generally a bit more central than mine.) I can hear it, but I often get tripped up trying to follow it (or produce it) in more rapid speech.

For me, FWIW, there’s a very significant length distinction betwen COT and CAUGHT.

(1) And I'm even a native English speaker! (Just one that merges COT-CAUGHT.)
(2) If you asked me to distinguish between /A/ and /O/ in an isolated word (COT or CAUGHT) without any sentence context, I very probably wouldn't be able to do it with any greater chance than just guessing. I would only be able to do it reliably by remembering the spelling.
(3) Oh dear, another length distinction in some English dialect that I'm unaware of? How does this one work? (Mine are both short vowels.) (I just learned about the Scottish length rule/distinction in my BARD-BARRED thread.)

Traditionally LOT (e.g. cot) is short and THOUGHT (e.g. caught) is long. It just happens that most NAE varieties have collapsed the traditional length distinction between vowels that is retained in the likes of EngE and AusE.

[anteallach]

Indeed, I thought that the spelling distinction between <o> and <au> was just an aesthetic spelling difference, like <ee> and <ea>, or <oCe> and <oa>, and nothing more. (Is there any dialect left in the English-speaking world that hasn't merged REED and READ, or LODE and LOAD?)

I don't think <oCe> vs. <oa> reflects a historic pronunciation difference at all. The traditional local dialect where I am (which I don't speak) has a distinction between the reflex of Old English /ɑː/ and that of Old English /ɔ/ lengthened in open syllables in Middle English. You might think that these would be represented by <oa> and <oCe> respectively but that's not how it works: for example both coal and hole come from open syllable lengthening but whole and goal come from the OE long vowel.

(The local reflex of the one from open syllable lengthening, found in words like coal and hole, is a diphthong of the [ɔɪ] type. You might be aware of the odd example of this, because it's the reason for the oy in the surnames Boothroyd and Murgatroyd; AIUI this Yorkshire royd is cognate with the -rhoden in the Swiss canton names Appenzell Innerrhoden and Appenzell Ausserrhoden.)

[jcb]

Indeed, I thought that the spelling distinction between <o> and <au> was just an aesthetic spelling difference, like <ee> and <ea>, or <oCe> and <oa>, and nothing more. (Is there any dialect left in the English-speaking world that hasn't merged REED and READ, or LODE and LOAD?)

I don't think <oCe> vs. <oa> reflects a historic pronunciation difference at all. The traditional local dialect where I am (which I don't speak) has a distinction between the reflex of Old English /ɑː/ and that of Old English /ɔ/ lengthened in open syllables in Middle English. You might think that these would be represented by <oa> and <oCe> respectively but that's not how it works: for example both coal and hole come from open syllable lengthening but whole and goal come from the OE long vowel.

(The local reflex of the one from open syllable lengthening, found in words like coal and hole, is a diphthong of the [ɔɪ] type. You might be aware of the odd example of this, because it's the reason for the oy in the surnames Boothroyd and Murgatroyd; AIUI this Yorkshire royd is cognate with the -rhoden in the Swiss canton names Appenzell Innerrhoden and Appenzell Ausserrhoden.)

So, there was a difference, but it wasn't reliably captured by the orthography. Great. Now I feel compelled to change English orthography so <oCe> and <oa> do represent the (un)merger properly.

[jcb]

The key thing is that most North Americans at least to my knowledge have no idea that anyone makes a distinction between two vowels in these words. (I didn't until I was well into adulthood.)

And then a selecti quidem of us grew up in a city with at START-NORTH merger and didn't realise their pronunciation of words like "forest" and "horror" was not GA until high school.

Apparently many people with the cot-caught merger have little to no awareness that other people have such a distinction, as strange as it may seem to cot-caught-unmerged people like myself.

What did you think of COT-CAUGHT mergers when you were growing up? Did you even notice them? Did you think that they just pronounced certain words different and not realize is was systemic?

[Travis B.]

And then a selecti quidem of us grew up in a city with at START-NORTH merger and didn't realise their pronunciation of words like "forest" and "horror" was not GA until high school.

Apparently many people with the cot-caught merger have little to no awareness that other people have such a distinction, as strange as it may seem to cot-caught-unmerged people like myself.

What did you think of COT-CAUGHT mergers when you were growing up? Did you even notice them? Did you think that they just pronounced certain words different and not realize is was systemic?

I was completely oblivious to the very existence of the cot-caught merger, partially that I had no awareness that anyone would merge them, but seemingly contradictorily because for me [ɑ(ː)] is an acceptable realization of both /ɑ/ and /ɔ/ (except /ɔ/ before /r/)*, so I didn't notice when people did merge the two.

* For me [ɑ(ː)] is the realization of /ɑ/ adjacent to /r w h kw gw/ (/ɑ/ otherwise is [a]), and while my normal realization of /ɔ/ not before /r/ is [ɒ(ː)], I will sporadically realize it as [ɑ(ː)], I am very used to [ɑ(ː)] as a valid realization of /ɔ/ not before /r/ as that is what my mother has for it (she has a similar realization of /ɑ/ as myself, where it is normally [a(ː)] but is [ɑ(ː)] adjacent to /r w h kw gw/). Adding to this, I am familiar with [ɑ(ː)] as a valid realization of /ɑ/ in all environments from its realization in GA.

[Glass Half Baked]

I don't have the cot-caught merger, but even I have to speak very carefully to notice the distinction. As a kid I would hear people speaking in Californian and just think "Oh, that's a funny way to say that." The low and back vowels are generally a good source of ridicule for non-standard dialects (up north they get cat by the caps stealing cwaufee), so I think subconsciously a lot of people just assume that part of the vowel space is a hopeless mess anyway.

[Glass Half Baked]

I have a question about yod-coalescence. In dialects with yod-coalescence, "sure" and "sugar" are pronounced with an SH sound. But "super" is not.
I have run through every reason for this I can think of, and have come up with nothing.
1. Could it be because super is a more recent loan? No. It is attested in Middle English, before yod-coalescence happened.
2. Maybe it just never had a palatal glide? No. Sugar and sure didn't have palatal glides originally either, but they acquired them through phonological processes that should have also happened to super.
3. Perhaps it lost its glide? No. Some dialects still pronounce a glide after the s. At least those dialects should have had yod-coalescence. This yod is unlikely to be a recent innovation, because that would raise the question of where a rogue u came from that was unaffected by the great vowel shift.

Can anyone shed a little light on this mystery for me?

I can’t fuly illuminate you... But I imagine it’s relevant that, for me, super shares /uː/ with soon, suitor etc., none of which have a palatal glide ‒ whereas sugar has /ʊ/ and sure /ɔː/ (RP /ʊ͜ə/).

In (very much) older/upper class British speech you might indeed hear glides in super and suitor, but never /ʃ/, and never (I think) in soon. Meanwhile I am quite certain I have never heard sugar with /sj/ rather than /ʃ/, even in old recordings, whilst for sure I have heard /sj/.

I don’t know off the top of my head precisely which conditions brought about the palatal glide on /uː/ or yod-coalescence, but I can tell you that sound change is lexically diffuse. Sound changes move through the lexicon in each idiolect spreading from word to word over time: it is not the case that a speaker suddenly produces glides in every /uː/ one day, or /ʃ/ in every word where they used to have /sj/.

The sociolinguistic factors that feed sound change sometimes shut down before the change completes diffusing through a lexicon. Sometimes they complete for some speakers / speech communities and not for others. Sometimes they get interrupted or reversed by other sound changes, again, for some, but not necessarily all, speech communities.

To my understanding, we often get left with these sorts of conundrums as a result.

Yeah, "it's just an exception; give up trying to think about it" was always a possibility; I just didn't want to accept it.

It's not easy to say whether super or sugar is the exception, though, since there are so few examples (remember, we're only talking about /su/ that predates the GVS, so "soon" is irrelevant).

One issue is that there may have been multiple waves of palatization. French /y/ borrowed into Middle English merged with /iw/ (itself from both native and borrowed words), and thence /ju/. This is how words like sugar got their glides, and why they're not pronounced /saugr/ today, as well as why words like "few" are pronounced the way they are. But is the /sju/ in "pseudonym" a conservation from Middle English (thus avoiding yod-coalescence), or is it a learned pronunciation based on the spelling (like "eureka")? If English is constantly making new /sju/ sequences, that could explain the lack of yod-coalescence. But it's hard to track, and in any case "super," like "pseudo" does go back to Middle English.

[zompist]

It's not easy to say whether super or sugar is the exception, though, since there are so few examples (remember, we're only talking about /su/ that predates the GVS, so "soon" is irrelevant).

FWIW super stands out as a Latin loan— compare suture, supine. The OED thinks the latter two have an optional yod in British English.

Is suit relevant? I'm not sure what the vowel was in Anglo-French—  there's an alternate spelling siwete which suggests a diphthong at least. (Modern French /sɥit/ presumably comes from /syitə/.)

[jcb]

Apparently many people with the cot-caught merger have little to no awareness that other people have such a distinction, as strange as it may seem to cot-caught-unmerged people like myself.

What did you think of COT-CAUGHT mergers when you were growing up? Did you even notice them? Did you think that they just pronounced certain words different and not realize is was systemic?

I was completely oblivious to the very existence of the cot-caught merger, partially that I had no awareness that anyone would merge them, but seemingly contradictorily because for me [ɑ(ː)] is an acceptable realization of both /ɑ/ and /ɔ/ (except /ɔ/ before /r/)*, so I didn't notice when people did merge the two.

* For me [ɑ(ː)] is the realization of /ɑ/ adjacent to /r w h kw gw/ (/ɑ/ otherwise is [a]), and while my normal realization of /ɔ/ not before /r/ is [ɒ(ː)], I will sporadically realize it as [ɑ(ː)], I am very used to [ɑ(ː)] as a valid realization of /ɔ/ not before /r/ as that is what my mother has for it (she has a similar realization of /ɑ/ as myself, where it is normally [a(ː)] but is [ɑ(ː)] adjacent to /r w h kw gw/). Adding to this, I am familiar with [ɑ(ː)] as a valid realization of /ɑ/ in all environments from its realization in GA.

Wait, so if [A] is an acceptable realization of /O/, then what is /A/ ? [a] ?