Mary-marry-merry merger

"the contrast between long and short high and mid vowels is still
maintained in initial position, so that the vowel of Mary is associated with the /ey/ of
mate, the vowel of merry with the /e/ of met and the vowel of marry with /æ/ of
mat.
Maps 49 to 51 of PEAS show that in the mid-twentieth century, a very large
part of the Eastern United States maintained a difference between Mary, cherry,
and marry. The vowel of marry appears as /æ/ in most of the eastern United States,
except for western New England (and a few points in southern Maine and New
Hampshire, western New York, and West Virginia). The vowel of Mary maintains
its upper mid quality in most of the North and the South, with the Midland and
eastern New York State showing a short vowel. In the intervening period, the two
mergers have progressed almost as far as the three preceding cases.
Map 8.4 presents the data differently from the previous three maps, since two
oppositions are involved, looking at production rather than perception.
6
The great
majority of symbols are blue circles, indicating the merger of both pairs. The red
circles represent the speakers who have both distinctions firmly in place. They are
concentrated in two areas: southeastern New England and a Mid-Atlantic region
including New York and Philadelphia, but not the rest of the Mid-Atlantic area to
the south. A third major type is shown by the green symbols: Mary merged with
merry, but marry is distinct. In the South, one can observe a wide distribution of
these green symbols, they also characterize Montreal in Canada. The rest of the
continent is dominated by a uniform distribution of blue symbols, representing
speakers with both mergers.
The belt of red symbols in the Philadelphia area is the result of a slightly
more complex phenomenon than we find in the rest of North America. While
/ey/, /e/, and /æ/ are distinct before intervocalic /r/, /e/ is not independent of /√/
in that position. Philadelphia shows a centralization of /e/ before intervocalic /r/
in words such as very, terrible, Merion as well as merry and ferry. The vowels of
ferry and furry, merry and Murray are distinct for about one third of Philadelphia
speakers, totally merged for another third, and in a state of near-merger for the
remainder. In the near-merger condition, speakers produce a consistent, statisti-
cally significant difference between the two classes, sometimes with no overlap,
but they label the two as ʻthe sameʼ in minimal pair tests and fail to discriminate
them in commutation tests (Labov, Karan, and Miller 1991).
The merger of the low and mid-front vowels before intervocalic /r/ is not
an isolated phenomenon. The tense and lax high front vowels are also widely
merged among younger speakers in pairs like nearer and mirror, spear it and
spirit. It seems clear that the erosion of contrast before intervocalic /r/ continues
the trend towards r-constriction. In r-less dialects, inter-
vocalic /r/ normally forms the onset of the second syllable, while in r-ful dialects
it is ambisyllabic and exerts a strong influence on the preceding vowel. Map 8.4
shows speakers with some degree of r-lessness as stars. It is evident that there are
very few star symbols that are blue; that is, vocalizaton of coda /r/ is associated
with retention of the contrast before intervocalic /r/"

<the vowel of Mary is associated with the /ey/ of mate>

Is it??
Isn't it rather associated with the vowel of "share" or of the verb "tear" ???

The matter here is that in unmerged dialects, the vowel of "Mary" is [e], whereas in many/most English dialects, the vowel of "mate" is [eI] or something similar such as [{I] in Australian English; only some English dialect groups, such as Scottish English, Irish English, and parts of North American English (particularly Upper Midwestern and Californian dialects) have [e(:)] as the vowel in "mate".

i am a native english speaker from the south. for me mary and merry are the same but marry is completely different.

marry has the a sound of cat but mary and merry have the e sound of bed.

I have a complete merger of "mary", "merry", and "marry", but I can't identify the vowel that I use. It doesn't sound quite the same as any vowels I have that aren't placed before an R. It sounds closest to [e], though.

I'm Mary-merry-marry unmerged. The vowels that I use in "merry" ["mEr\i] and "marry" ["m{r\i] are identical, respectively, to the vowels that I use in "met" and "mat". The vowel that I use in "Mary" doesn't sound like either [E] or [eI] for me - I transcribe this as [E@`].

I have a complete Marry-merry-marry merger, and my merged vowel is between [e] and [E] but it is much closer to my [e] than to my [E]. I would narrowly transcribe it as [e_o], particularly because my [E] is rather low and centralized while my [e], like my merged vowel, is completely uncentralized. The traditional transcription of the merged vowel as [E] or even [E_r] would be very misleading due to my [E]'s lowness and centralization.

Furthermore, such a transcription would be problematic because my merged vowel and [E] are actually contrastive in my dialect at times. For instance, I distinctly contrast "airy" as ["e_o:Ri:] (using [e_o] for the merged vowel) and "every" as ["E:Ri:] (alongside the more careful ["E:vRi:] or ["E:vr\i:]). Using [E] for the transcription of the merged vowel wold imply that they are homophones, and even using [E_r] would imply that they are rather similar in pronunciation, while in reality they are very different from each other in pronunciation.

>>I was just thinking about that yesterday. I have the same type of elision in 'every' in casual speech, and a contrast between 'airy' ["E_r@`i:] and 'every' [E:r\i:]<<

Now that I think about it, this example shows that I really do have the same rhotic postvocalically (except after /aI/, /aU/, /OI/, and sometimes /i/) and in other positions, as I do not have the variation between [@`] and [r\] that you have here.

(I used to treat /ir/ as /Ir/, but it seems that my dialect has an allophonic variation between [iR=] and [I_rR], and while I in everyday speech tend to use more [I_rR] more, primarily using [iR=] in more careful or emphatic speech, many here have the former far more frequently than I have it. Consequently, I now think that /ir/ is probably a better interpretation of this than /Ir/ in my dialect.)

(Well, that's more free variation than allophony, as saying "allophony" implies that such is determined by environment.)

I have Mary as [me4i], marry as [ma4I] and merry as [mE4I].

<<Are you Scottish?>>

Yes.

wat is a homophones for even???

It's to keep your ears warm in the winter, I have a pair. Do you?

>>I have a complete Marry-merry-marry merger, and my merged vowel is between [e] and [E] but it is much closer to my [e] than to my [E]. I would narrowly transcribe it as [e_o], particularly because my [E] is rather low and centralized while my [e], like my merged vowel, is completely uncentralized. The traditional transcription of the merged vowel as [E] or even [E_r] would be very misleading due to my [E]'s lowness and centralization.

Furthermore, such a transcription would be problematic because my merged vowel and [E] are actually contrastive in my dialect at times. For instance, I distinctly contrast "airy" as ["e_o:Ri:] (using [e_o] for the merged vowel) and "every" as ["E:Ri:] (alongside the more careful ["E:vRi:] or ["E:vr\i:]). Using [E] for the transcription of the merged vowel wold imply that they are homophones, and even using [E_r] would imply that they are rather similar in pronunciation, while in reality they are very different from each other in pronunciation.<<

I have to revise my transcription above. The transcription that I am using above is a rather conventional one, but it really is not necessarily accurate. To more accurately map such, such should be:

Conventional > Accurate

e > e_o
e_o > E_r
E > 3_+
{_r > E or E_o

Can we say many front vowels in not very much space? And of all of them, the only two can never be contrasted are [e_o] and [E_r]...

That should be "the only two that can never be contrasted" above.