English questions

[Travis B.]

As this thread has already talked about before, in American English, some dialects raise /{/ before /n/.

My question: Has any dialect raised this /{/ so far that it's merged with /ej/? (If so, I guess we could call it the "man-mane(-main) merger".)

It’s before the velar nasal instead, but otherwise I do this.

I too raise /{/ before /N/ (and /g/), but it's to [{j], except in "hang" which indeed has /ej/. But, unlike many people, I don't raise /{/ before /n/.

I raise TRAP before most cases of /ŋ/ to [e̞(ː)] so as to merge with FACE (but not in pancake or Bernanke) and sometimes I allophonically raise TRAP before /g/ to [ĕ̞ə̯̆]~[e̞ə̯].

[jcb]

It’s before the velar nasal instead, but otherwise I do this.

I too raise /{/ before /N/ (and /g/), but it's to [{j], except in "hang" which indeed has /ej/. But, unlike many people, I don't raise /{/ before /n/.

I raise TRAP before most cases of /ŋ/ to [e̞(ː)] so as to merge with FACE (but not in pancake or Bernanke) and sometimes I allophonically raise TRAP before /g/ to [ĕ̞ə̯̆]~[e̞ə̯].

You have /N/ in "pancake"?

[Darren]

You have /N/ in "pancake"?

I think all L1 English varieties do

[Nortaneous]

You have /N/ in "pancake"?

I think all L1 English varieties do

Not with slashes unless GenAm [eə] is also phonemic - otherwise how to explain p[eə]ncake but b[æj]nker? And it's a transparent compound.

/ŋ/ can almost certainly be eliminated with syllabification tricks, or contrastive secondary stress tricks (/ij ʌw ɝ/ vs. /ɨj əw ər/?) if you don't like the possibility of ambiguous or contrastive syllabification. But I don't think there's any reason to postulate phonemic /eə/ for GenAm. (It's obviously phonemic in some Northern Mid-Atlantic dialects, though.)

[Darren]

You have /N/ in "pancake"?

I think all L1 English varieties do

Not with slashes unless GenAm [eə] is also phonemic - otherwise how to explain p[eə]ncake but b[æj]nker? And it's a transparent compound.

/ŋ/ can almost certainly be eliminated with syllabification tricks, or contrastive secondary stress tricks (/ij ʌw ɝ/ vs. /ɨj əw ər/?) if you don't like the possibility of ambiguous or contrastive syllabification. But I don't think there's any reason to postulate phonemic /eə/ for GenAm. (It's obviously phonemic in some Northern Mid-Atlantic dialects, though.)

ah yes i should of clarified i meant [ŋ]

[jcb]

I think all L1 English varieties do

Not with slashes unless GenAm [eə] is also phonemic - otherwise how to explain p[eə]ncake but b[æj]nker? And it's a transparent compound.

/ŋ/ can almost certainly be eliminated with syllabification tricks, or contrastive secondary stress tricks (/ij ʌw ɝ/ vs. /ɨj əw ər/?) if you don't like the possibility of ambiguous or contrastive syllabification. But I don't think there's any reason to postulate phonemic /eə/ for GenAm. (It's obviously phonemic in some Northern Mid-Atlantic dialects, though.)

ah yes i should of clarified i meant [ŋ]

My point is that I have neither /N/ or [N], but /n/ and [n]. The word's status as a transparent compound prevents assimilation.
"pancake" = /p{n.kejk/ = ["pʰ{n.cejc]

[Travis B.]

In the English here /n/ regularly assimilates across morpheme and word boundaries to following plosives with regard to POA; e.g. I have [ˈpʰɛ̆ə̯̆ŋˌke̞ʔk]. This is such that as a kid I thought the Hoan Bridge in Milwaukee was the "Home Bridge", because the /n/ in Hoan assimilates as [m] very consistently (and Mayor Hoan was before my time).

[Richard W]

My point is that I have neither /N/ or [N], but /n/ and [n]. The word's status as a transparent compound prevents assimilation.
"pancake" = /p{n.kejk/ = ["pʰ{n.cejc]

I would describe it as semantically opaque, and I seem to have either in this word. For me, the rejection of final /ŋɡ/ in a word is a fairly weak rule (but stronger than the rejection of final /mb/). I would say that English 'dentals' assimilate their point of articulation fairly freely.

[Travis B.]

In very careful speech I may have [n] in pancake, but I wonder how much of that is the force of orthography at work.

[jcb]

My point is that I have neither /N/ or [N], but /n/ and [n]. The word's status as a transparent compound prevents assimilation.
"pancake" = /p{n.kejk/ = ["pʰ{n.cejc]

I would describe it as semantically opaque, and I seem to have either in this word. For me, the rejection of final /ŋɡ/ in a word is a fairly weak rule (but stronger than the rejection of final /mb/). I would say that English 'dentals' assimilate their point of articulation fairly freely.

I suppose that it's not as transparent as a word like "blackbird", but it's still made up of two common english words, and is more transparent than a word like "cranberry".

[Travis B.]

My point is that I have neither /N/ or [N], but /n/ and [n]. The word's status as a transparent compound prevents assimilation.
"pancake" = /p{n.kejk/ = ["pʰ{n.cejc]

I would describe it as semantically opaque, and I seem to have either in this word. For me, the rejection of final /ŋɡ/ in a word is a fairly weak rule (but stronger than the rejection of final /mb/). I would say that English 'dentals' assimilate their point of articulation fairly freely.

I suppose that it's not as transparent as a word like "blackbird", but it's still made up of two common english words, and is more transparent than a word like "cranberry".

To me at least, assimilation of /n/ is not blocked by transparent compounds, after all it productively operates across word boundaries quite consistently for me.

[Richard W]

I suppose that it's not as transparent as a word like "blackbird", but it's still made up of two common english words, and is more transparent than a word like "cranberry".

But I don't consider pancakes to be cakes.

[Nortaneous]

I suppose that it's not as transparent as a word like "blackbird", but it's still made up of two common english words, and is more transparent than a word like "cranberry".

But I don't consider pancakes to be cakes.

A witch doctor is neither a doctor nor a witch, and a fuckwind is neither a wind nor a fuck.

[jcb]

I suppose that it's not as transparent as a word like "blackbird", but it's still made up of two common english words, and is more transparent than a word like "cranberry".

But I don't consider pancakes to be cakes.

And people don't generally live in lighthouses.

[zompist]

I suppose that it's not as transparent as a word like "blackbird", but it's still made up of two common english words, and is more transparent than a word like "cranberry".

But I don't consider pancakes to be cakes.

A witch doctor is neither a doctor nor a witch,

And people don't generally live in lighthouses.

All these examples depend on thinking in prototypes. Sure, a pancake doesn't look much like a chocolate layer cake. It is "a breadlike food made from a dough or batter that is usually fried or baked in small flat shapes and is often unleavened", as MW has it.

People don't live in doghouses, publishing houses, or the Houses of Parliament either, but these are easily understood extensions of the word.

A witch doctor, in the sense of shaman, isn't a M.D., but is a person you go to for healing or counseling. The combo suggests "supernatural physician" which is pretty much what is meant.

[jcb]

But I don't consider pancakes to be cakes.

A witch doctor is neither a doctor nor a witch,

And people don't generally live in lighthouses.

All these examples depend on thinking in prototypes. Sure, a pancake doesn't look much like a chocolate layer cake. It is "a breadlike food made from a dough or batter that is usually fried or baked in small flat shapes and is often unleavened", as MW has it.

People don't live in doghouses, publishing houses, or the Houses of Parliament either, but these are easily understood extensions of the word.

A witch doctor, in the sense of shaman, isn't a M.D., but is a person you go to for healing or counseling. The combo suggests "supernatural physician" which is pretty much what is meant.

Yes, this is what I meant. "blackbird" doesn't have this problem, so it's fully transparent. Next are compounds like "pancake" and "lighthouse", which I guess we can call semi-transparent. Then there are compounds like "cranberry" and "acorn", which have otherwise dead morphemes in them, and are liable to be reanalyzed with living morphemes ("eggcorn"). Then there are words that were originally compounds, then stopped being seen/pronounced as compounds, then became compounds again by force of spelling and/or reanalysis, like "waistcoat/weskit" and (sort of) "posthumous".

[Starbeam]

In the case of acorn both components are living morphemes, they've just been obscured by sound changes.

[Travis B.]

In the case of acorn both components are living morphemes, they've just been obscured by sound changes.

Actually, acorn apparently isn't a compound at all -- according to Etymonline, it goes back to OE æcern and is cognate with ON akarn, Dutch aker, LG Ecker, StG Ecker (loan from Low German?), and Gothic akran, and the idea that it comes from OE ac (ModE oak) and corn is a 15th/16th century folk etymology.

[Starbeam]

Huh, thank you! The more you know

[Travis B.]

If it were such a compound you'd expect, if it were cognate, StG Eichkorn (which is a surname and does not refer to anything like acorns).