Pronunciations you had to unlearn

[anteallach]

[citation desperately needed]

All the sources I have to hand (OED, AHD, Wiktionary, etc.) give /ˈeːt/ as the chief pronunciation for all varieties of English and list /ˈɛt/ as--at best--an alternative in some. Maybe it's "standard" only for UK nonagenarians?

The Old English preterite is <æt>, which can yield /et/, but not /eɪt/. Onions' Oxford Dictionary of English Etymology says, "The sp. ate of the pt. depends on early ME at, which repr. a short var. of OE ǣt; the pronunc. et is usu. assoc with the sp. ate, but is perh. a shortening of pt. ēt." The pronunciation of the preterite is given as "et, eɪt".

The fact of the matter though is despite what one can dig up about etymology, /eɪt/ (whether you pronounce it with an actual diphthong or with a mid or close-mid monophthong, and regardless of whether you actually pronounce it long or not) is the standard pronunciation of ate in all English varieties today.

I'd say that both are standard in current BrE. I have /ɛt/.

[WarpedWartWars]

Croatia as [kɹo.ti.ʌ]. Apparently it's actually [kɹo.e.ʃʌ].

[Linguoboy]

Croatia as [kɹo.ti.ʌ]. Apparently it's actually [kɹo.e.ʃʌ].

Why is stress omitted from your transcription?

[Travis B.]

Croatia as [kɹo.ti.ʌ]. Apparently it's actually [kɹo.e.ʃʌ].

Why is stress omitted from your transcription?

I also wonder whether WarpedWartWars really has true [ʌ] for /ə/ or rather that they have fallen victim to the misconception that [ə] is "[ʌ]". (As a note for WarpedWartWars, unless you have the NCVS you probably don't have [ʌ] as a phone, and then only in stressed syllables.)

[Travis B.]

Croatia as [kɹo.ti.ʌ]. Apparently it's actually [kɹo.e.ʃʌ].

Also note that you should properly mark aspiration when transcribing English - e.g. I will hear initial [k] or [k] at the start of a stressed syllable not preceded by a vowel as /g/ rather than /k/ because the contrast between /g/ and /k/ in those positions is an aspiration contrast* for me and many other English-speakers rather than a voicing contrast.

* There is a reason Pinyin uses <p t k> to mark aspirated consonants and <b d g> to mark unaspirated consonants.

[Linguoboy]

Croatia as [kɹo.ti.ʌ]. Apparently it's actually [kɹo.e.ʃʌ].

Why is stress omitted from your transcription?

I also wonder whether WarpedWartWars really has true [ʌ] for /ə/ or rather that they have fallen victim to the misconception that [ə] is "[ʌ]". (As a note for WarpedWartWars, unless you have the NCVS you probably don't have [ʌ] as a phone, and then only in stressed syllables.)

I think it's pretty clear that, despite the brackets, this is not actually a narrow phonetic transcription but a broad phonemic one. That's why I was focussing on a phonemic feature (stress) rather than getting bogged down in nitpicking each and every one of the phonetic symbols (since it was the lack of stress indication which caused me to misread the transcription).

For reference, WWW, the Wiktionary transcription is /ˌkɹoʊˈeɪ.ʃə/, which the forward slashes marks as phonemic but which I would actually consider broadly phonetic (since it indicates such subphonemic details as diphthongisation and the nature of the /r/).

[Travis B.]

I think it's pretty clear that, despite the brackets, this is not actually a narrow phonetic transcription but a broad phonemic one.

On second thought, I agree.

For reference, WWW, the Wiktionary transcription is /ˌkɹoʊˈeɪ.ʃə/, which the forward slashes marks as phonemic but which I would actually consider broadly phonetic (since it indicates such subphonemic details as diphthongisation and the nature of the /r/).

I also agree that the "/ɹ/" used is phonetic detail, because the conventional phonemic representation is /r/ (particularly as there are English-speakers such as myself which don't have [ɹ] as a phone, so "/ɹ/" would be misleading).

About "/oʊ/" and "/eɪ/", to me those are just conventional transcriptions, particularly motivated by the convention by many British writers to use "/e/" to refer to /ɛ/, and hence the need to contrast such with "/eɪ/" (and if one is going to mark that as a diphthong, one might as well also mark "/oʊ/" as one too to match it).

Personally, I would phonemically mark Croatia as I speak it as /ˌkroʊˈeɪʃə/, as /oʊ/ and /eɪ/ are conventional, even though the phonetic reality of it is [ˌkʰʁˤo̞ːˈwe̞ʃə(ː)] for me.

[Linguoboy]

Personally, I would phonemically mark Croatia as I speak it as /ˌkroʊˈeɪʃə/, as /oʊ/ are /eɪ/ conventional, even though the phonetic reality of it is [ˌkʰʁˤo̞ːˈwe̞ʃə(ː)] for me.

Yeah, this is why I prefer monophthongal transcriptions, i.e. /oː/ and /eː/. (I know there are dialects which are not pain-pane or toe-tow merged, but they're very rare and marginal; if I ever need to transcribe an elderly speaker from the Rhondda Valley or something, I'll introduce the distinction.)

[Travis B.]

Personally, I would phonemically mark Croatia as I speak it as /ˌkroʊˈeɪʃə/, as /oʊ/ are /eɪ/ conventional, even though the phonetic reality of it is [ˌkʰʁˤo̞ːˈwe̞ʃə(ː)] for me.

Yeah, this is why I prefer monophthongal transcriptions, i.e. /oː/ and /eː/. (I know there are dialects which are not pain-pane or toe-tow merged, but they're very rare and marginal; if I ever need to transcribe an elderly speaker from the Rhondda Valley or something, I'll introduce the distinction.)

To me much of phonemic transcription is convention. E.g. to me there is no real difference between /oː eː/ and /oʊ eɪ/. If I were to use a phonemic transcription that had NAE varieties alone in mind, I'd go further and mark those as /o e/, as historical phonemic vowel length has been largely lost in NAE. However, as there do exist good portions of English for which vowel length is not predictable from vowel quality and there are some varieties which have vowels that phonemically contrast on vowel length alone, such as AusE, it is best to still indicate either length or diphthongization in transcriptions when describing words in English in general.

[Linguoboy]

To me much of phonemic transcription is convention. E.g. to me there is no real difference between /oː eː/ and /oʊ eɪ/. If I were to use a phonemic transcription that had NAE varieties alone in mind, I'd go further and mark those as /o e/, as historical phonemic vowel length has been largely lost in NAE.

I used to be pretty militant about using Trager-Smith notation for my NAE variety but eventually I got tired of blockheads who couldn't wrap their heads around the possibility that phonemic transcriptions which used characters for something besides their God-given IPA values were equally valid and useful and gave up.

On topic, I think I used to rhyme Croat with groat before I knew better.

[Travis B.]

To me much of phonemic transcription is convention. E.g. to me there is no real difference between /oː eː/ and /oʊ eɪ/. If I were to use a phonemic transcription that had NAE varieties alone in mind, I'd go further and mark those as /o e/, as historical phonemic vowel length has been largely lost in NAE.

I used to be pretty militant about using Trager-Smith notation for my NAE variety but eventually I got tired of blockheads who couldn't wrap their heads around the possibility that phonemic transcriptions which used characters for something besides their God-given IPA values were equally valid and useful and gave up.

I for a while a long time ago insisted on using my own idiosyncratic phonemic transcription for my NAE variety out of excessive snowflakiness, until I realized that I just confused everyone and no one could then compare my pronunciations with those in other English varieties, where then I switched quite consciously to a purely conventional IPA transcription that pretty much fit the standard sort used by most North Americans. (However, I insist on not marking vowels as rhotic but rather use a separate following /r/ because to me that is excessively narrow and because an analysis with a separate /r/ works better in certain corner cases in my variety.)

On topic, I think I used to rhyme Croat with groat before I knew better.

For quite some time I was unsure as to the pronunciation, since I primarily read the word.

[WarpedWartWars]

Why is stress omitted from your transcription?

Because

I think it's pretty clear that, despite the brackets, this is not actually a narrow phonetic transcription but a broad phonemic one.

and I saw no reason to indicate stress, but here it is anyway: /ˈkro.ti.ə/ [ˈkɹo.ti.ʌ] and /kroˈe.ʃə/ [kʰɹoˈ(w)e̞(j).ʃʌ].

I also wonder whether WarpedWartWars really has true [ʌ] for /ə/ or rather that they have fallen victim to the misconception that [ə] is "[ʌ]". (As a note for WarpedWartWars, unless you have the NCVS you probably don't have [ʌ] as a phone, and then only in stressed syllables.)

I have three schwas: [ɪ], [ə] or [ʊ] (I can't really tell the difference easily), and [ʌ]. All are phonemically /ə/, though.

[Moose-tache]

I also wonder whether WarpedWartWars really has true [ʌ] for /ə/ or rather that they have fallen victim to the misconception that [ə] is "[ʌ]". (As a note for WarpedWartWars, unless you have the NCVS you probably don't have [ʌ] as a phone, and then only in stressed syllables.)

Yet another victim of the /ə/ fallacy. If I had a nickel for every time I heard this one.
There is no /ə/. The STRUT vowel, in every significant variety of NAE, is [ʌ]. People have convinced themselves that this is what schwa sounds like, but use your brain for two seconds. Schwa is an allophonic realization of unstressed vowels in NAE. You can run a simple experiment. Pronounce words like "codable, Thibodeaux, bandages," as quickly as you would in normal speech. Notice that reduced vowel in the middle syllable? Now carefully pronounce the word "nut." Is it the same vowel? No, obviously not. Because nobody says "beuhckle euhp before you drive the treuhck." Generations have been brainwashed into believing that the STRUT vowel is schwa, but it just demonstrably isn't. Schwa isn't even a separate phoneme in most varieties of NAE.

Climate change is real. STRUT is not a schwa. Get over it.

[Moose-tache]

[ʌ]... phonemically /ə/

This is backwards. [ə] occurs as an unstressed allophone of a number of different vowels, while /ʌ/ is contrastive against all other stressed vowels. You say [bʌt] when speaking carefully, and only say [bət] when speaking quickly and the word has attached to a larger prosodic unit that drags the stress elsewhere. Also, [bʌt] has one origin and is maximally contrastive (i.e. it can't be confused with "bait," "bot," etc.), while [bət] could be the reduced version of almost anything, as in "Talbot" or "gibbet."

[Travis B.]

I also wonder whether WarpedWartWars really has true [ʌ] for /ə/ or rather that they have fallen victim to the misconception that [ə] is "[ʌ]". (As a note for WarpedWartWars, unless you have the NCVS you probably don't have [ʌ] as a phone, and then only in stressed syllables.)

Yet another victim of the /ə/ fallacy. If I had a nickel for every time I heard this one.
There is no /ə/. The STRUT vowel, in every significant variety of NAE, is [ʌ]. People have convinced themselves that this is what schwa sounds like, but use your brain for two seconds. Schwa is an allophonic realization of unstressed vowels in NAE. You can run a simple experiment. Pronounce words like "codable, Thibodeaux, bandages," as quickly as you would in normal speech. Notice that reduced vowel in the middle syllable? Now carefully pronounce the word "nut." Is it the same vowel? No, obviously not. Because nobody says "beuhckle euhp before you drive the treuhck." Generations have been brainwashed into believing that the STRUT vowel is schwa, but it just demonstrably isn't. Schwa isn't even a separate phoneme in most varieties of NAE.

Climate change is real. STRUT is not a schwa. Get over it.

Well there is a phonemic /ə/ but it only exists only in unstressed syllables (e.g. in commA) and it emphatically is not STRUT, which is not a schwa.

[Ares Land]

Oh, today I learned Epstein is pronounced /ˈɛpstiːn/ but Einstein is /ˈaɪnstaɪn/.

[Linguoboy]

Oh, today I learned Epstein is pronounced /ˈɛpstiːn/ but Einstein is /ˈaɪnstaɪn/.

These names are a mess in English. Take Fierstein, which according to German spelling conventions would be /fiːrstain/ but which Harvey Fierstein pronounces /fairstiːn/.

In general, German gentile names have /ai/ and Ashkenazic Jewish names have /iː/, but there are plenty of exceptions.

[Travis B.]

Oh, today I learned Epstein is pronounced /ˈɛpstiːn/ but Einstein is /ˈaɪnstaɪn/.

These names are a mess in English. Take Fierstein, which according to German spelling conventions would be /fiːrstain/ but which Harvey Fierstein pronounces /fairstiːn/.

In general, German gentile names have /ai/ and Ashkenazic Jewish names have /iː/, but there are plenty of exceptions.

The traditional pronunciation of <ei> in German names here is /ai/~/əi/ but as Linguoboy says there are many exceptions, especially in Ashkenazi names (note that Albert Einstein himself was an Ashkenazi Jew).

[WarpedWartWars]

This is backwards. [ə] occurs as an unstressed allophone of a number of different vowels, while /ʌ/ is contrastive against all other stressed vowels. You say [bʌt] when speaking carefully, and only say [bət] when speaking quickly and the word has attached to a larger prosodic unit that drags the stress elsewhere. Also, [bʌt] has one origin and is maximally contrastive (i.e. it can't be confused with "bait," "bot," etc.), while [bət] could be the reduced version of almost anything, as in "Talbot" or "gibbet."

The first ways I think of for pronouncing "Talbot" and "gibbet" are [ˈtʰæɫ.b(ə)ʔ] and [ˈgɪ.bɪʔ]. BTW what does "Talbot" refer to here?

[Raphael]

The traditional pronunciation of <ei> in German names here is /ai/~/əi/ but as Linguoboy says there are many exceptions, especially in Ashkenazi names (note that Albert Einstein himself was an Ashkenazi Jew).

Keep in mind that Einstein only migrated to the US in his fifties, unlike people with similar names who had been born and raised there.