Pronunciations you had to unlearn

[Zaarin]

I have /ɪ/ in pidgin, /n̩/ in pigeon--but in rapid speech I probably have /n̩/ in both.

[Vijay]

Pretty sure they're homophones for me, probably with [n̩]

[Xwtek]

I had to unlearn the pronunciation for either pidgin or pigeon. They're both either [pɪdʒɪn] (British) or [pɪdʒən] (US). I used to pronounce the former as the [pɪdʒɪn] and the latter as [pɪdʒən]. Since I align more to US English (despite Indonesian usually align more to British), I'll use the latter.

I'm a native speaker of BrE and have schwa (or possibly syllabic [n]) in the second syllable of pigeon. The first dictionary I checked (Chambers) shows both pronunciations with the one with schwa first. So no correction necessary!

Then, there is something wrong with this dictionary. On the other hand, now I'm confused whether it's OK not to treat them as homophone.

[Vijay]

Sure, it's okay. Anyone can edit Wiktionary as far as I'm aware, so of course it's prone to errors.

[Ryusenshi]

The difference between /ɪ/ and /ə/ in unstressed syllables is such a mess... Geoff Lindsay, in English After RP, notes that RP used to have more unstressed syllables with /ɪ/, where other BrE varieties have /ə/ instead. For instance, invisibility would be /ɪnˌvɪzɪˈbɪlɪtɪ/ in RP, but /ɪnˌvɪzəˈbɪlətiː/ in SSBE.

I tend to follow the spelling (so /ə/ for pigeon, /ɪ/ for pidgin) and not bother too much about it. Who could blame a non-native speaker for being inconsistent, where native speakers themselves don't agree with each other?

[alynnidalar]

TIL some people pronounce the two differently at all...

[Pabappa]

"Rosa's roses" is an example of a minimal pair for the two, but it's complicated by the fact that some people also distinguish between /ə/ and /ʌ/ in unstressed syllables, and have /ʌ/ for the final vowel in Rosa.

[Travis B.]

"Rosa's roses" is an example of a minimal pair for the two, but it's complicated by the fact that some people also distinguish between /ə/ and /ʌ/ in unstressed syllables, and have /ʌ/ for the final vowel in Rosa.

Mind you that this involves morpheme boundaries, and for me at least, the realization of the weak vowel is sensitive to them.

[anteallach]

"Rosa's roses" is an example of a minimal pair for the two, but it's complicated by the fact that some people also distinguish between /ə/ and /ʌ/ in unstressed syllables, and have /ʌ/ for the final vowel in Rosa.

Mind you that this involves morpheme boundaries, and for me at least, the realization of the weak vowel is sensitive to them.

Yes, and also the distribution of the two vowels varies among speakers with the distinction (as already seen with pigeon and pidgin, which form a perfectly good minimal pair for me with no morpheme boundaries). My accent has /ə/ in roses (and other -es and -ed inflectional endings where some other accents have /ɪ/) so any distinction between it and Rosa's is down to the morpheme boundary.

[Richard W]

Then, there is something wrong with this dictionary. On the other hand, now I'm confused whether it's OK not to treat them as homophone.

Then put it right! There's a lot wrong with it, but it's mostly correctable.

[Salmoneus]

Then, there is something wrong with this dictionary. On the other hand, now I'm confused whether it's OK not to treat them as homophone.

Then put it right! There's a lot wrong with it, but it's mostly correctable.

Well, 'right' is subjective, innit? It can't very well list every possible pronunciation of every word!

In this case, there are five problems:

a) the weak vowel merger results in unstressed /I/ and /@/ merging in any case, so the two words are homophones. Dialects with the WVM include most English speakers in the southern hemisphere, and most Americans, as well as some English people.

b) among people without the WVM there can still be variability between dialects. In particular, the replacement of many /I/s with /@/ is one of the vowel shifts that distinguishes SSBE from RP (along with things like happy-tensing and the debroadening of a bunch of broad As). This isn't a clearcut shift, however but one still in progress - in general, I think more /@/ is associated with lower-class, younger and more southerly speakers. So, in traditional RP these words were homophones. For many, probably most modern SSBE speakers, they're not. But for some, they still are.

c) as a result, for many people these sounds may actually be in free variation, or may shift to accomodate to the person one is speaking to. Personally, I absolutely and immediately think of 'pigeon' as having /@n/... but if I really think about saying it with /In/, particularly if I were talking to someone like my father who does say it that way... it doesn't seem totally alien to me.

d) since these are unstressed vowels, which struggle to even be heard, let alone distinctly analysed, for many tokens of these words it would probably be hard to pinpoint exactly what sound is found. Actually, in this specific pair it's easier than it often is - because /@n/ reduces to syllabic /n/, and /In/ doesn't - although that in turn means that even phonetic [@n] would probably be heard as phonemic /In/.

e) on the other hand, that makes it hard to distinguish between "this person has /I/ instead of /@/" and "this person doesn't reduce their /@n/ sequences to syllabic nasals". Since the process of reduction is itself not complete or universal, the former can easily be hidden as the latter.

e) self-analysis is notoriously unreliable, but in this case there's an unusually big issue with spelling. A lot of the shift /I/ > /@/ seems to be based on spelling pronunciations, to the extent that I think most young SSBE speakers, like me, have probably internalised this as a 'rule', to the extent that it can be genuinely hard, given a word, to distinguish what we really say from what we just think we're meant to say.




Anyway: short version, yes it's OK to have them as homophones. Most Americans, and older British standards, agree on that. However, a lot of younger English people would be surprised, as I briefly was, to learn than anyone had them as homophones, because this isn't a salient sociolinguistic shibboleth. But also because even if you said them as homophones, then given how this is complicated by the nasal syllabification issue, and how little attention is paid to unstressed vowels anyway, most of us wouldn't even notice...

[Richard W]

Well, 'right' is subjective, innit? It can't very well list every possible pronunciation of every word!

As the declared intention is for the English Wiktionary to list every frequent spelling of every well-established word in every language, I don't see why it shouldn't attempt to list every frequent pronunciation, at least at the phonemic level. For example, it includes 'seperate' as well as 'innit'.

In practical terms, it's already creaking under the strain of recording all the words with the same spelling on one page, and so far it only has 111 languages for 'a'.

I'm noi disputing your problem list.

[Linguoboy]

Wait, so the conventional pronunciation of the ancient city of Tanis is with /eː/, not with /aː/ or /æ/? And Abydos is /əˈbaidɔs/, not */ˈæbɪdɔs/?

[anteallach]

Wait, so the conventional pronunciation of the ancient city of Tanis is with /eː/, not with /aː/ or /æ/? And Abydos is /əˈbaidɔs/, not */ˈæbɪdɔs/?

Both follow the usual rules for pronouncing classical names in English, as far as I can see. Now, I haven't internalised these rules very well, and the spelling Abydos is ambiguous anyway, and would probably instinctively pronounce them the same way that you did.

[KathTheDragon]

I say /ˈtænɪs/ and /ˈæbiˌdəʊs/, and /ˈteɪnɪs/ and /əˈbɐɪdɒs/ are utterly incomprehensible to me.

[Raphael]

Wait, so the conventional pronunciation of the ancient city of Tanis is with /eː/, not with /aː/ or /æ/? And Abydos is /əˈbaidɔs/, not */ˈæbɪdɔs/?

Both follow the usual rules for pronouncing classical names in English,

TIL that there are usual rules for pronouncing classical names in English.

[Vlürch]

Wait, so the conventional pronunciation of the ancient city of Tanis is with /eː/, not with /aː/ or /æ/? And Abydos is /əˈbaidɔs/, not */ˈæbɪdɔs/?

Both follow the usual rules for pronouncing classical names in English,

TIL that there are usual rules for pronouncing classical names in English.

Yeah, at least for Latin there are, but I'm not sure if there are for other languages. I think more or less the same rules are mostly applied to all Greek words and names, too, not just ones filtered through Latin? No idea if there are any set "rules" for Sumerian, Akkadian, Sanskrit, etc. words and names, but probably not.

[Pabappa]

Hebrew at least seems to have no clear rule. I at least think I remember hearing Capernaum with final /nɔm/ in church as a kid, but Wikipedia lists the two pronuncs as (/kəˈpɜːrniəm, -neɪəm/). We seem to have a difficult time with hiatus in English, especially unstressed hiatus. Using an /i/ in this word is merely following the precedent of existing loans like karaoke.

edit: oh, right, I made the same mistake with Emmaus, though in that case, I was saying it with /au/ because it's a college in the USA and I had no idea it was a Biblical place name.

For comparison, Sabaoth always keeps its /e/, according to Wiktionary. Canaan and Isaac are just pronounced with schwas.

edit: but yet, Esau has /ɔ/, ... I dont know if thats because it ends in that vowel or if it's just due to greater familiarity.
edit2: also, Saul .... but that's a monosyllable so maybe that's a factor too.

[Kuchigakatai]

I wish the traditional English pronunciation of Latin was still alive. It was hard to understand for anyone else who knew Latin but was not used to it, but the pronunciation was LOVELY.

Quō ūsque tandem abūtēre, Catilīna, patientiā nostrā?
[ˈkʰwoʊ ˈɐskwi ˈtʰændɪm ˌæbjʉˈtʰiɚi | ˌkʰæɾəˈlaɪnə | ˌpʰeɪsiˈɛnʃə ˈnɑstʃɹə]

Quam diū etiam furor iste tuus nōs ēlūdet?
[ˈkʰwɑm ˈdiju ˌɛʃəm ˈfjuɹɚ ˌɪsti ˈtʰʉwəs ˈnɑs iˈlʉɾɪt]

Quem ad fīnem sēsē effrēnāta iactābit audācia?
[ˈkʰwɛm æd ˈfaɪnɪm ˈsizi ˌɛfɹəˈneɪɾə jækˈtʰeɪbɪɾ ɑˈdeɪʃə]

[Estav]

I wish the traditional English pronunciation of Latin was still alive. It was hard to understand for anyone else who knew Latin but was not used to it, but the pronunciation was LOVELY.

Quō ūsque tandem abūtēre, Catilīna, patientiā nostrā?
[ˈkʰwoʊ ˈɐskwi ˈtʰændɪm ˌæbjʉˈtʰiɚi | ˌkʰæɾəˈlaɪnə | ˌpʰeɪsiˈɛnʃə ˈnɑstʃɹə]

Quam diū etiam furor iste tuus nōs ēlūdet?
[ˈkʰwɑm ˈdiju ˌɛʃəm ˈfjuɹɚ ˌɪsti ˈtʰʉwəs ˈnɑs iˈlʉɾɪt]

Quem ad fīnem sēsē effrēnāta iactābit audācia?
[ˈkʰwɛm æd ˈfaɪnɪm ˈsizi ˌɛfɹəˈneɪɾə jækˈtʰeɪbɪɾ ɑˈdeɪʃə]

For English words, and names used in English, I prefer pronunciations based on the traditional English pronunciation of Latin because I find it a more consistent system than the alternatives. But I haven't met many people who prefer the sound of the traditional English pronunciation for reading Latin text aloud. I'm curious about a few of the pronunciations you mentioned here—I don't want to grill you on what may have been just a quick and approximate transcription, but I'm interested in learning if some of the details represent traditional features that I am not aware of.

• [ˈtʰændɪm] : why ɪm rather than ɛm or əm? For some reason, those feel more natural to me, and those are the pronunciations that Wikipedia mentions for final unstressed -em.
• Catilīna, effrēnāta [kʰæɾəˈlaɪnə, ˌɛfɹəˈneɪɾə] : likewise, why ə instead of ı here? Wouldn't the weak vowel merger or its absence affect medial and final syllables similarly?
• kʰwɑm : why [ɑm] and not [æm] here?
• etiam [ɛʃəm] : This seems like an interesting exception to the usual lengthening of e before -tia-. Does it have [ɛ] because of influence from the pronunciation of "et"?
• iactābit [jækˈtʰeɪbɪɾ] : Did the replacement of <j~i> [dʒ] with <i> [j] occur before other reforms to the pronunciation of Latin in the English-speaking world?