How your idiolect differs from the standard language

The best topics from Languages & Linguistics, kept on a permanent basis.
caedes
Sanci
Sanci
Posts: 24
Joined: Sun Mar 01, 2009 3:35 am
Location: Poland

Post by caedes »

A funny construction in my German dialect I discovered after having met some older family members who use the dialect much more often:

Where it's Ich kam, um sie zu sehen in Standard German for engl. "I came to see them", I heard constructions like [i bɪ̃n ͡tsʊ̃mn di ˈseː.ɜ ˈkʰɔ̜̃.mɜ], lit. Ich bin zum die sehen gekommen or I am to-the they see come, so with a gerund and a preceding personal pronoun instead of an adverbial phrase, which probably is due to analogy to verbal nouns with an incorporated noun like schiffeversenken etc.

And i finally took a look again at the allophonic realisations of /x/:
It's [ç~x] after close front vowels, [χʀ̥]/[χ̆ʀ̥] after other non-pharyngealized vowels and [ʀ̥] after pharyngealized vowels. The interesting thing now is that there's a phoneme /ʀ/, but that is usually voiced [ʀ̬], which leads to a contrastive pair [ʀ̥]/[ʀ̬] in terms of some word pairs. One would be Lercher (a family name) vs. Lehrer "teacher", namely [ˈlɛɑ̯ˤ.ʀ̥ɑˤ] vs. [ˈlɛɑ̯ˤ.ʀ̥ɑˤ].

Even more strange since my dialect normally lacks a distinction [+-voiced] at all. ~
Cuix ticpiä cuitlatl itic motzontecomauh ?

User avatar
Alces
Lebom
Lebom
Posts: 87
Joined: Sun Jan 03, 2010 6:09 pm
Location: Merseyside, England, UK
Contact:

Post by Alces »

I live near Liverpool, not in the city though. My dialect is fairly normal urban British English, with the distinctive northern English features, as well as some Scouse features.

This is a fairly complete list of how my dialect differs from RP phonologically. In terms of grammar and vocabulary, it's basically standard, with only a few minor differences.
- No trap-bath split, so 'bath' is [baf], 'laugh' is [laf], 'dance' is [dans], etc. In recently added words and loanwords though, I may have [ɑː] where Americans wouldn't have it, for example 'can't' [kɑ̃ːʔ].
- No foot-strut split. /ʌ/ never split off from /ʊ/ and is completely absent from the language for me. So 'up' [ʊpʰ].
- Some words, like 'book' or 'cook', may have /u/ instead of /ʊ/, although it varies somewhat with me (because my mum would say /buk/ but my dad would say /bʊk/).
- [ŋ] is not a phoneme; it is an allophone of /n/ before /k g/. So 'sing' is [sɪŋg].
- The verbal ending -ing has dropped the /g/, so that it is now -[ɪn].
- In words like 'nothing', 'something', 'everything', I vary between dropping the [g] and fortifying the [g] to [k]. So 'nothing' might be [nɒfɪn] or [nɒfɪnk].
- /r/ is realised as a labialised postalveolar approximant, [ɹ̱ʷ]. The labialisation is very strong, so that when I was first learning about phonetics I thought /r/ was a labial consonant. The roundedness is of a different sort than that in /w/, but I'm not sure how to describe it. Since [ɹ̱ʷ] never appears postvocalically you can analyse [ɑː ɔː ɜː] as underlying /ar ɒr ɛr/ respectively, and I think a lot of people perceive them that way. With this you get a nice tame eight-vowel system of /i ɪ ɛ a ə u ʊ ɒ/, which you can even reduce to seven if you analyse [ə] as unstressed /ɛ/, /a/ or /ɒ/.
- I'm not too sure about how to interpret my stops. /p t k/ are generally [pʰ tʰ kʰ], only becoming unaspirated after obstruents; they are still aspirated at the end of a word as far as I can tell. /b d g/ might be voiceless in some environments, but I'm not sure.
- RP /əu/ is reflected as /ɵu/.
- /l/ is only velarised before consonants. The consonant can be the first one of the next word though. So 'all of them' [ɔːləvm̩], 'all things' [ɔːɫ̩fɪŋgz].
- /t/ is glottalised, except initially, in the onset of a stressed syllable, and in clusters. So 'bottle' [bɒʔl̩], 'heart' [hɑːʔ], but 'attack' [ə'tak] and enter ['ɛntə]; 'cat' is [kaʔ] but 'cats' is [kats].
- It can be glottalised after /l/ though. For instance 'alter' [ɒɫ̩ʔə]. Oddly, /nt/ is not glottalised usually, but it is at the end of words: 'went' [wɛ̃nʔ]. The [n] in [nʔ] also tends to disappear, but I'm inconsistent about this.
- More traditional in my area is to weaken /t/ to a tap or even fully-fledged [ɹ̱ʷ] in the same environments; I don't usually do this though.
- /d/ weakens as well as /t/, but not as strongly. It either weakens towards a tap [ɾ] or an affricate [dz], but never fully. In fact I think the most common realisation would be an alveolar non-sibilant fricative, [ð̱]. The conditions are generally the same as for /t/, but I'm not very consistent about it.
- /θ ð/ have basically disappeared, although I do pronounce them properly sometimes. /θ/ has generally become /f/. /ð/ is more complicated; it becomes /v/ everywhere but the start of a word, where it survives as a sort of lax phoneme that might be realised as zero, [d], [ɾ], alveolar [ð̱], or even [l]. Alveolar [ð̱] seems to be the most common realisation so I will use that symbol.
- There are no dipthongs ending in [ə]; /ɛə/ has become /ɜː/, /ʊə/ has become /ɔː/, /iə/ is two syllables.
- /i ai ei oi/ break to /iə ajə ejə ojə/ before /l/.
- In general, word boundaries are not good barriers to sound change. Nasals often assimilate to a following consonant, and similar consonants often turn to geminates. This never applies in careful speech though.
- Certain words can be unstressed (I've not read anything about this, so my terminology might not be right), including pronouns, copulas, prepositions, and determiners. For example 'you' is [ju] stressed and [jə] unstressed; 'I'll' is [ajəl] stressed and [al] unstressed. These words are unstressed most of the time, only being stressed when they recieve special emphasis. E.g. 'what are you doing' would normally be [wɒʔəjəduɪ̃n], but if I'd just asked someone else the same question I'd say [wɒʔəjuduɪ̃n], and if they were doing something quite strange I'd say [wɒʔɑːjəduɪ̃n].
- /j/ has dropped after /l/: 'lure' /lɔː/. After /t d s z n h/ it coalesces with it producing [tʃ dʒ ʃ ʒ ɲ ç]: 'tune' [tʃuːn], 'dew' [dʒu], 'huge' [çuːdʒ].
- The pronoun 'youse' [juz], unstressed [jəz], exists as a plural 2nd person pronoun. It is not really a true plural though, since it's only used when disambiguation is necessary.
- Certain vowels, [ɪ ɛ a ʊ ɒ], are not permitted at the end of a word. So historical [ɪ] is always word finally, for example 'happy' [hapi].
- Only [ɪ ə] may appear in unstressed syllables, barring compounds and words with scientific prefixes or suffixes ('archaeology' [ˌɑːki'ɒlədʒi]. [ɪ] tends to centralise to [ɨ] when unstressed, but it's not consistent. I don't have [ɪ] in some cases where standard English apparently has it (e.g. 'roses' [ɹ̱ʷɵʊ̯zəz]); this might be a spelling pronunciation.
- Stressed /i u/ are lengthened before voiced consonants. 'seem' [sĩːm], 'bead' [biːd].
- /ai au/ are lengthened before voiced consonants and at the end of a word: 'bough' [baːʊ]. They also weaken their final element to [ɪ ʊ] when lengthened. In all other environments they are generally [ai au]; the other dipthongs always have a fully close final element.
- If a word ends in [ə ɜː ɔː ɑː] and a word beginning in a vowel follows, /r/ is inserted in between as sandhi. This also applied in the middle of words: so 'drawing' [dɹ̱ʷɔːɹ̱ʷɪn], 'draw a cat' [dɹ̱ʷɔːɹ̱ʷəkaʔ], but 'draw that' [dɹ̱ʷɔːð̱aʔ].
- The letter H is [heɪtʃ]. This is actually normal in my area, not an error.
- My /f v/ tend to be bilabial, but this is probably an idiolectal thing.
- /h/ may be dropped initially, but it varies a lot. It's most common with function words like 'he' or 'him'.

User avatar
Soap
Smeric
Smeric
Posts: 1228
Joined: Sun Feb 16, 2003 2:57 pm
Location: Scattered disc
Contact:

Post by Soap »

I sometimes say "close the lights" when I mean to say turn them off. Given my personality one might think I'm being weird on purpose, and maybe originally I picked this up just to be weird, but now it's just a matter of me speaking quicker than I think. The original expression seems to be Portuguese/Spanish.
Sunàqʷa the Sea Lamprey says:
Image

User avatar
Torco
Smeric
Smeric
Posts: 2372
Joined: Thu Aug 30, 2007 10:45 pm
Location: Santiago de Chile

Post by Torco »

Portuguese, maybe... spanish doesn't have that.

User avatar
Hakaku
Lebom
Lebom
Posts: 132
Joined: Sat Feb 03, 2007 12:55 pm
Location: 常世

Post by Hakaku »

Soap wrote:I sometimes say "close the lights" when I mean to say turn them off. Given my personality one might think I'm being weird on purpose, and maybe originally I picked this up just to be weird, but now it's just a matter of me speaking quicker than I think. The original expression seems to be Portuguese/Spanish.
More like French. "Close the lights" and "Open the lights" are expressions that are really common where I live, but that's a result of a strong francophone and bilingual influence in the region. Compare "Fermer les lumières" and "Ouvrir les lumières".
Chances are it's Ryukyuan (Resources).

TaylorS
Avisaru
Avisaru
Posts: 557
Joined: Sat Jul 05, 2008 1:44 pm
Location: Moorhead, MN, USA

Post by TaylorS »

Alces wrote:I live near Liverpool, not in the city though. My dialect is fairly normal urban British English, with the distinctive northern English features, as well as some Scouse features.

This is a fairly complete list of how my dialect differs from RP phonologically. In terms of grammar and vocabulary, it's basically standard, with only a few minor differences.
- No trap-bath split, so 'bath' is [baf], 'laugh' is [laf], 'dance' is [dans], etc. In recently added words and loanwords though, I may have [ɑː] where Americans wouldn't have it, for example 'can't' [kɑ̃ːʔ].
- No foot-strut split. /ʌ/ never split off from /ʊ/ and is completely absent from the language for me. So 'up' [ʊpʰ].
- Some words, like 'book' or 'cook', may have /u/ instead of /ʊ/, although it varies somewhat with me (because my mum would say /buk/ but my dad would say /bʊk/).
- [ŋ] is not a phoneme; it is an allophone of /n/ before /k g/. So 'sing' is [sɪŋg].
- The verbal ending -ing has dropped the /g/, so that it is now -[ɪn].
- In words like 'nothing', 'something', 'everything', I vary between dropping the [g] and fortifying the [g] to [k]. So 'nothing' might be [nɒfɪn] or [nɒfɪnk].
- /r/ is realised as a labialised postalveolar approximant, [ɹ̱ʷ]. The labialisation is very strong, so that when I was first learning about phonetics I thought /r/ was a labial consonant. The roundedness is of a different sort than that in /w/, but I'm not sure how to describe it. Since [ɹ̱ʷ] never appears postvocalically you can analyse [ɑː ɔː ɜː] as underlying /ar ɒr ɛr/ respectively, and I think a lot of people perceive them that way. With this you get a nice tame eight-vowel system of /i ɪ ɛ a ə u ʊ ɒ/, which you can even reduce to seven if you analyse [ə] as unstressed /ɛ/, /a/ or /ɒ/.
- I'm not too sure about how to interpret my stops. /p t k/ are generally [pʰ tʰ kʰ], only becoming unaspirated after obstruents; they are still aspirated at the end of a word as far as I can tell. /b d g/ might be voiceless in some environments, but I'm not sure.
- RP /əu/ is reflected as /ɵu/.
- /l/ is only velarised before consonants. The consonant can be the first one of the next word though. So 'all of them' [ɔːləvm̩], 'all things' [ɔːɫ̩fɪŋgz].
- /t/ is glottalised, except initially, in the onset of a stressed syllable, and in clusters. So 'bottle' [bɒʔl̩], 'heart' [hɑːʔ], but 'attack' [ə'tak] and enter ['ɛntə]; 'cat' is [kaʔ] but 'cats' is [kats].
- It can be glottalised after /l/ though. For instance 'alter' [ɒɫ̩ʔə]. Oddly, /nt/ is not glottalised usually, but it is at the end of words: 'went' [wɛ̃nʔ]. The [n] in [nʔ] also tends to disappear, but I'm inconsistent about this.
- More traditional in my area is to weaken /t/ to a tap or even fully-fledged [ɹ̱ʷ] in the same environments; I don't usually do this though.
- /d/ weakens as well as /t/, but not as strongly. It either weakens towards a tap [ɾ] or an affricate [dz], but never fully. In fact I think the most common realisation would be an alveolar non-sibilant fricative, [ð̱]. The conditions are generally the same as for /t/, but I'm not very consistent about it.
- /θ ð/ have basically disappeared, although I do pronounce them properly sometimes. /θ/ has generally become /f/. /ð/ is more complicated; it becomes /v/ everywhere but the start of a word, where it survives as a sort of lax phoneme that might be realised as zero, [d], [ɾ], alveolar [ð̱], or even [l]. Alveolar [ð̱] seems to be the most common realisation so I will use that symbol.
- There are no dipthongs ending in [ə]; /ɛə/ has become /ɜː/, /ʊə/ has become /ɔː/, /iə/ is two syllables.
- /i ai ei oi/ break to /iə ajə ejə ojə/ before /l/.
- In general, word boundaries are not good barriers to sound change. Nasals often assimilate to a following consonant, and similar consonants often turn to geminates. This never applies in careful speech though.
- Certain words can be unstressed (I've not read anything about this, so my terminology might not be right), including pronouns, copulas, prepositions, and determiners. For example 'you' is [ju] stressed and [jə] unstressed; 'I'll' is [ajəl] stressed and [al] unstressed. These words are unstressed most of the time, only being stressed when they recieve special emphasis. E.g. 'what are you doing' would normally be [wɒʔəjəduɪ̃n], but if I'd just asked someone else the same question I'd say [wɒʔəjuduɪ̃n], and if they were doing something quite strange I'd say [wɒʔɑːjəduɪ̃n].
- /j/ has dropped after /l/: 'lure' /lɔː/. After /t d s z n h/ it coalesces with it producing [tʃ dʒ ʃ ʒ ɲ ç]: 'tune' [tʃuːn], 'dew' [dʒu], 'huge' [çuːdʒ].
- The pronoun 'youse' [juz], unstressed [jəz], exists as a plural 2nd person pronoun. It is not really a true plural though, since it's only used when disambiguation is necessary.
- Certain vowels, [ɪ ɛ a ʊ ɒ], are not permitted at the end of a word. So historical [ɪ] is always word finally, for example 'happy' [hapi].
- Only [ɪ ə] may appear in unstressed syllables, barring compounds and words with scientific prefixes or suffixes ('archaeology' [ˌɑːki'ɒlədʒi]. [ɪ] tends to centralise to [ɨ] when unstressed, but it's not consistent. I don't have [ɪ] in some cases where standard English apparently has it (e.g. 'roses' [ɹ̱ʷɵʊ̯zəz]); this might be a spelling pronunciation.
- Stressed /i u/ are lengthened before voiced consonants. 'seem' [sĩːm], 'bead' [biːd].
- /ai au/ are lengthened before voiced consonants and at the end of a word: 'bough' [baːʊ]. They also weaken their final element to [ɪ ʊ] when lengthened. In all other environments they are generally [ai au]; the other dipthongs always have a fully close final element.
- If a word ends in [ə ɜː ɔː ɑː] and a word beginning in a vowel follows, /r/ is inserted in between as sandhi. This also applied in the middle of words: so 'drawing' [dɹ̱ʷɔːɹ̱ʷɪn], 'draw a cat' [dɹ̱ʷɔːɹ̱ʷəkaʔ], but 'draw that' [dɹ̱ʷɔːð̱aʔ].
- The letter H is [heɪtʃ]. This is actually normal in my area, not an error.
- My /f v/ tend to be bilabial, but this is probably an idiolectal thing.
- /h/ may be dropped initially, but it varies a lot. It's most common with function words like 'he' or 'him'.
Very interesting! I love Northern England accents! :mrgreen:

TaylorS
Avisaru
Avisaru
Posts: 557
Joined: Sat Jul 05, 2008 1:44 pm
Location: Moorhead, MN, USA

Post by TaylorS »

My analysis of my dialect:

Code: Select all

p   t   tS   k
p_h t_h tS_h k_h
f T s   S
v D z   Z
m   n        N
w   r\` j    5

i:        u:
 I       U
  e:    o:
   E   V
     a

aI aU OI E@ 8` @` 
/p k/ are intervocalically.

Intervocalic /t t_h/ are [4] when followed by an unstressed vowel,

Intervocalic /nt nt_h/ are [4~] when followed by an unstressed vowel.

Coda /nd/ is [n:].

Coda /np_h/ is [?k] and nasalization of the previous vowel.

Coda /nt_h/ is [?] and nasalization of the previous vowel.

Coda /ntS_h/ is [?tS] and nasalization of the previous vowel.

Coda /nk_h/ is [?k] and nasalization of the previous vowel.

coda /p_h t_h tS_h k_h/ are [?p ? ?tS ?k].

Initial /k_h/ is [kx] in stressed syllables.

/tj t_hj/ become [tS tS_h] (gotcha, dontcha, etc.).

Initial /T D/ harden to [t_d d_d].

/v D z Z/ devoice when at the end of a word, but the glottis is held lower then in /f T s S/.

Initial and intervocalic /st zt/ becomes /s_j: z_j:/

/n/ assimilates to the following consonant.

/n/ elides before fricatives, nasalizing the previous vowel.

unstressed /N/ becomes [n]

/r\`/ is always pharyngealized. Is labialized at the beginning of a word. is [R_/?] after back vowels and /a/.

Alveolar plosives become post-alveolar apical affricates when followed by [r\`]

Alveolar fricatives, /n/, and /5/ become post-alveolar apical fricatives, [n`], and [l`] when followed by [r\`]

/5/ is [M\_l] when following back vowels and /a/.

/h/ is [C] and [x] when followed by /i:/ and /u:/

unstressed /a E V/ are [@].

Unstressed /I U/ are [1]

/u: U/ front to the middle when following coronal consonants.

/o:/ is [O:] when followed by /r\`/.

/a/ is [Q] when followed by /5/.

Historical /{r Er eIr/ are [Er\`] (Mary-Marry-Merry Merger).

In unstressed syllables /aI aU E@/ are [A@ {@ E]
Last edited by TaylorS on Sun Jun 06, 2010 7:21 pm, edited 1 time in total.

User avatar
äreo
Avisaru
Avisaru
Posts: 326
Joined: Sun Jul 01, 2007 10:40 pm
Location: Texas

Post by äreo »

Idiolectal analysis (of myself and others) is something I find really interesting, so I'm working (again) on a really, fully elaborate description of my 'lect that I'll post when I finish.

Another question, though, is my point: does anyone else notice a significant presence of high-rising terminals in their speech?

User avatar
finlay
Sumerul
Sumerul
Posts: 3600
Joined: Mon Dec 22, 2003 12:35 pm
Location: Tokyo

Post by finlay »

Alces wrote: - The verbal ending -ing has dropped the /g/, so that it is now -[ɪn].
Careful now; this is an extant feature in all varieties of English and has been for the past 1000 years, although it's strongly stratified socially and how this works varies wildly from area to area. I'd be surprised if you used this variant completely exclusively.

But anyway, you've got it back to front: the /g/ was 'added' in the early modern period, essentially, even though yours is one of the few accents that actually has it. (Usually to be more accurate we'd just say that the place of articulation has changed, as the 'g' is a spelling artifact.) There was also a distinction between the -ing ending and the -in ending inherited from Old English, as well as a distinction between the verbal endings and other -ings, which you appear to maintain. I can't tell if you do the first one, and I doubt it. But apparently there are places in the world (rednecks was the example) where they do maintain it. (How do you pronounce "wedding"? "Building"?)

As for the rest of your post, a couple of points/questions:
- I've never understood why people spell the plural pronoun 'youse' that way. I mean, it's clearly a plural -s morpheme. 'yous' would be the way I'd do it.
- [ i ] in happy, I think, is the historical one, and it's RP that messes it up (again)

- Most of the phenomena you've described have conventional names:
-- /r/ as "sandhi" in non-rhotic accents: Linking-R or Intrusive-R (you've got the second, which means you'll insert an R where there is none in the spelling; conservative RP has Linking-R only)
-- dropping /h/ is, unsurprisingly, "H-Dropping" (again, there's not an accent in the world that doesn't drop H's in unstressed function words).
-- /θ ð/ becoming /f v/, /t d/, /f d/, or any number of other combinations: "TH-fronting"
-- glottalising /t/ is known as T-glottaling, again unsurprisingly.

- When you analyse the vowels as a "nice 8-vowel system" of whatever, bear in mind that I've never found dropping diphthongs out of these that acceptable, personally. You've got at least 13 vowels with them and at least 16 with the three that you dropped out for dubious reasons... spelling should not be conflated with punctuation, and no, I don't think the /r/ really is there in your accent. Since it only shows up under a regular phonological rule (intrusive r) at the end of a morpheme, I wouldn't then conclude that it's there underlyingly. You've also presumably got /ɔ:/ in THOUGHT, for example (oh yeah, look up the Standard Lexical Sets by Wells, they're a fun way to classify your vowels).

- When you phonemicise, try and make it as simple as possible. Frankly I think that [əu] vs [ɵu] is rather a minute difference and not worth marking in the phonemic transcription. In fact, probably not worth marking in a phonetic transcription unless you're trying to be as narrow as possible. (To be fair, this is further conflated by the fact that I'm used to just writing /o/, which is also acceptable for accents like yours or RP, incidentally. I occasionally get confused with all the /əu/s that get thrown around...)

- Do you fricativise stops? Pronounce 'back' as [bax], for instance? This is pretty common in Liverpool.
- Do you front the GOOSE vowel (/u/)? I don't know if this happens in Liverpool for sure but I suspect it might.

And finally, "- In general, word boundaries are not good barriers to sound change. Nasals often assimilate to a following consonant, and similar consonants often turn to geminates. This never applies in careful speech though. " is only even applicable in this thread because the Standard Variety is a carefully constructed lie that nobody speaks unless they're speaking "carefully". Lol.

TaylorS
Avisaru
Avisaru
Posts: 557
Joined: Sat Jul 05, 2008 1:44 pm
Location: Moorhead, MN, USA

Post by TaylorS »

IIRC final /I/ is the archaic pronunciation, "happy-tensing" of final /I/ to /i:/ is a phenomenon of the last 250 years in various dialects.

User avatar
Nortaneous
Sumerul
Sumerul
Posts: 4544
Joined: Mon Apr 13, 2009 1:52 am
Location: the Imperial Corridor

Post by Nortaneous »

turns out schwa dropping/devoicing ("tomorrow" [t_hmAr@U]) is pretty common around here; it's not just me, as I thought earlier. also, modal "can" [gN=]. I don't have this ([k_hEn]~[k_hn=]) except before velars, I think, but a lot of people have it everywhere in unstressed position. nasal assimilation isn't all that common for me anyway (I have [n] in "input", for example), so I guess that makes sense.
TaylorS wrote:IIRC final /I/ is the archaic pronunciation, "happy-tensing" of final /I/ to /i:/ is a phenomenon of the last 250 years in various dialects.
this is what I thought. IMD it's anywhere from [1] (maybe even [M]?) to to , but it's never long (never gets any stress).
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

User avatar
finlay
Sumerul
Sumerul
Posts: 3600
Joined: Mon Dec 22, 2003 12:35 pm
Location: Tokyo

Post by finlay »

TaylorS wrote:IIRC final /I/ is the archaic pronunciation, "happy-tensing" of final /I/ to /i:/ is a phenomenon of the last 250 years in various dialects.
Meh, couldn't be bothered to look it up. Also it's weird when you write it like that, with a long marker. It's, like, unstressed and can't be long... :?

User avatar
Soap
Smeric
Smeric
Posts: 1228
Joined: Sun Feb 16, 2003 2:57 pm
Location: Scattered disc
Contact:

Post by Soap »

finlay wrote:
TaylorS wrote:IIRC final /I/ is the archaic pronunciation, "happy-tensing" of final /I/ to /i:/ is a phenomenon of the last 250 years in various dialects.
Meh, couldn't be bothered to look it up. Also it's weird when you write it like that, with a long marker. It's, like, unstressed and can't be long... :?
I think that;s a matter of definition. Wikipedia says it "may" have been a boomerang change, /ī/ to /ĭ/ to /ī/, which went all the way in the USA but not in some parts of Britain.
Sunàqʷa the Sea Lamprey says:
Image

User avatar
AnTeallach
Lebom
Lebom
Posts: 125
Joined: Tue Jan 17, 2006 12:51 pm
Location: Yorkshire

Post by AnTeallach »

Alces wrote: - /r/ is realised as a labialised postalveolar approximant, [ɹ̱ʷ]. The labialisation is very strong, so that when I was first learning about phonetics I thought /r/ was a labial consonant. The roundedness is of a different sort than that in /w/, but I'm not sure how to describe it. Since [ɹ̱ʷ] never appears postvocalically you can analyse [ɑː ɔː ɜː] as underlying /ar ɒr ɛr/ respectively, and I think a lot of people perceive them that way. With this you get a nice tame eight-vowel system of /i ɪ ɛ a ə u ʊ ɒ/, which you can even reduce to seven if you analyse [ə] as unstressed /ɛ/, /a/ or /ɒ/.
How would you treat words like Sahara or aural in this analysis? [a] and [ɒ] can presumably appear in that environment, so there is a potential contrast (not that I can think of a minimal pair right now).

I do agree that a lot of people think of [ɑː] as somehow "having an r in it", e.g. they'll say that southerners pronounce "bath" as "barth", but I tend to think this is more to do with how the sound is usually spelt in non-rhotic Northern English than the actual phonology...
- /t/ is glottalised, except initially, in the onset of a stressed syllable, and in clusters. So 'bottle' [bɒʔl̩], 'heart' [hɑːʔ], but 'attack' [ə'tak] and enter ['ɛntə]; 'cat' is [kaʔ] but 'cats' is [kats].
- It can be glottalised after /l/ though. For instance 'alter' [ɒɫ̩ʔə]. Oddly, /nt/ is not glottalised usually, but it is at the end of words: 'went' [wɛ̃nʔ]. The [n] in [nʔ] also tends to disappear, but I'm inconsistent about this.
- More traditional in my area is to weaken /t/ to a tap or even fully-fledged [ɹ̱ʷ] in the same environments; I don't usually do this though.
- /d/ weakens as well as /t/, but not as strongly. It either weakens towards a tap [ɾ] or an affricate [dz], but never fully. In fact I think the most common realisation would be an alveolar non-sibilant fricative, [ð̱]. The conditions are generally the same as for /t/, but I'm not very consistent about it.
Interesting that you have the fricative realisation of /d/ but not /t/.
The letter H is [heɪtʃ]. This is actually normal in my area, not an error.
I think this is becoming quite common generally. It's presumably partly a hypercorrection of h-dropping, but there's supposedly an Irish Catholic connection (to the extent that it's a known shibboleth in Northern Ireland).

User avatar
finlay
Sumerul
Sumerul
Posts: 3600
Joined: Mon Dec 22, 2003 12:35 pm
Location: Tokyo

Post by finlay »

Soap wrote:
finlay wrote:
TaylorS wrote:IIRC final /I/ is the archaic pronunciation, "happy-tensing" of final /I/ to /i:/ is a phenomenon of the last 250 years in various dialects.
Meh, couldn't be bothered to look it up. Also it's weird when you write it like that, with a long marker. It's, like, unstressed and can't be long... :?
I think that;s a matter of definition. Wikipedia says it "may" have been a boomerang change, /ī/ to /ĭ/ to /ī/, which went all the way in the USA but not in some parts of Britain.
But it's , not or [i:]......

Travis B.
Sumerul
Sumerul
Posts: 3570
Joined: Mon Jun 20, 2005 12:47 pm
Location: Milwaukee, US

Post by Travis B. »

finlay wrote:
Soap wrote:
finlay wrote:
TaylorS wrote:IIRC final /I/ is the archaic pronunciation, "happy-tensing" of final /I/ to /i:/ is a phenomenon of the last 250 years in various dialects.
Meh, couldn't be bothered to look it up. Also it's weird when you write it like that, with a long marker. It's, like, unstressed and can't be long... :?
I think that;s a matter of definition. Wikipedia says it "may" have been a boomerang change, /ī/ to /ĭ/ to /ī/, which went all the way in the USA but not in some parts of Britain.
But it's , not or [i:]......

You are confusing realized sounds with historical phonemes, and mind you that there are dialects such as my own which do have vowel length distinctions in unstressed syllables; of course, mine are not related to historical vowel length, but incidentally enough, all final syllables in my dialect that did not contain fortis obstruents in their codas historically and which do not precede words that had fortis obstruents or /h/ historically in the initial syllables are long (even though they are still generally shorter than long vowels in stressed syllables, mind you).

User avatar
finlay
Sumerul
Sumerul
Posts: 3600
Joined: Mon Dec 22, 2003 12:35 pm
Location: Tokyo

Post by finlay »

Travis B. wrote:
finlay wrote:
Soap wrote:
finlay wrote:
TaylorS wrote:IIRC final /I/ is the archaic pronunciation, "happy-tensing" of final /I/ to /i:/ is a phenomenon of the last 250 years in various dialects.
Meh, couldn't be bothered to look it up. Also it's weird when you write it like that, with a long marker. It's, like, unstressed and can't be long... :?
I think that;s a matter of definition. Wikipedia says it "may" have been a boomerang change, /ī/ to /ĭ/ to /ī/, which went all the way in the USA but not in some parts of Britain.
But it's , not or [i:]......

You are confusing realized sounds with historical phonemes, and mind you that there are dialects such as my own which do have vowel length distinctions in unstressed syllables; of course, mine are not related to historical vowel length, but incidentally enough, all final syllables in my dialect that did not contain fortis obstruents in their codas historically and which do not precede words that had fortis obstruents or /h/ historically in the initial syllables are long (even though they are still generally shorter than long vowels in stressed syllables, mind you).

Huh? Erm, historical phonemes? How are they relevant? I'd argue that this is an area where phonemes don't do a terribly adequate job of describing the phenomenon fully.

Travis B.
Sumerul
Sumerul
Posts: 3570
Joined: Mon Jun 20, 2005 12:47 pm
Location: Milwaukee, US

Post by Travis B. »

finlay wrote:
Travis B. wrote:
finlay wrote:
Soap wrote:
finlay wrote: Meh, couldn't be bothered to look it up. Also it's weird when you write it like that, with a long marker. It's, like, unstressed and can't be long... :?
I think that;s a matter of definition. Wikipedia says it "may" have been a boomerang change, /ī/ to /ĭ/ to /ī/, which went all the way in the USA but not in some parts of Britain.
But it's , not or [i:]......

You are confusing realized sounds with historical phonemes, and mind you that there are dialects such as my own which do have vowel length distinctions in unstressed syllables; of course, mine are not related to historical vowel length, but incidentally enough, all final syllables in my dialect that did not contain fortis obstruents in their codas historically and which do not precede words that had fortis obstruents or /h/ historically in the initial syllables are long (even though they are still generally shorter than long vowels in stressed syllables, mind you).

Huh? Erm, historical phonemes? How are they relevant? I'd argue that this is an area where phonemes don't do a terribly adequate job of describing the phenomenon fully.

I normally refer to historical phonemes as in forms that present realizations are descended from diachronically; I do not really believe in phonemes at all as a synchronic phenomenon, as it poorly models actual sound change and phonological phenomena. In particular, I tend to refer only to historical phonemes and present-day phones when referring to my own dialect, as phonemic-type analyses of my dialect tend to be rather scary*.

* They tend to involve either phonemic vowel length decoupled from vowel quality along with phonemic vowel nasalization and phonemic pitch accent, which happen to be largely not consciously accessible to speakers or complex layers of (often quite contrived) phonological processes constructed to allow "explaining away" the former type of analysis, but which run into their own problems because many of the above sorts of processes tend to be at least somewhat lexicalized in practice, implying that, were one to follow a phonemic-type model, the former type of analysis is right after all**. In the end, there are simply too many inconsistencies and complications from trying to force my dialect's phonology into a phonemic-type model, to the point that one gets far more consistent results if one just gives up on trying to analyze it all in terms of phonemes in the first place. (Oh, and just to make things worse, even with the latter type of phonemic-type model of my dialect, you cannot explain away pitch accentuation for historically contracted forms, no matter how hard you try.)

** The diachronics also just work out much nicer if you assume that vowel length and nasalization were frozen at a relatively early date****, likely several generations ago, than if you try to keep them as synchronic phenomena. (Vowel length alternations in final syllables of morphemes are best treated as morphophonological in nature, mind you; in practice there are a lot of inconsistencies introduced by assuming that such is actually a synchronic phonetic-level phenomenon***.)

*** You get the problem of pesky silent consonants still effecting changes in vowel length even across word boundaries...

**** The lower bound for this is the voicing assimilation/fortition of historical /v/ in have to and historical /z/ in has to, as these are preceded by short vowels rather than the long vowels that one would expect if vowel length were set before these became /f/ and /s/ respectively.

User avatar
Soap
Smeric
Smeric
Posts: 1228
Joined: Sun Feb 16, 2003 2:57 pm
Location: Scattered disc
Contact:

Post by Soap »

Sorry, I should have just used SAMPA. I dont know the British rules of vowel length, so I figured I'd just obscure them by using dictionary-style ī and ĭ instead, but that caused more trouble than it solved.
Sunàqʷa the Sea Lamprey says:
Image

Yng
Avisaru
Avisaru
Posts: 880
Joined: Fri Jul 03, 2009 3:17 pm
Location: Llundain

Post by Yng »

I've noticed that I have an awful tendency towards lenition intervocalically, particularly across word boundaries:

'no thanks' is [nohanks]

I glottal-stop all the time, but this isn't particularly rare. My r flicks between a tap and that weird approximant thing most English speakers have depending on register. I also seem to keep spitting out things that sound awfully similar to ejectives, which is something I've heard other people claim to do on more than one occasion, although they seem awfully out of place...

User avatar
Alces
Lebom
Lebom
Posts: 87
Joined: Sun Jan 03, 2010 6:09 pm
Location: Merseyside, England, UK
Contact:

Post by Alces »

AnTeallach wrote:How would you treat words like Sahara or aural in this analysis? [a] and [ɒ] can presumably appear in that environment, so there is a potential contrast (not that I can think of a minimal pair right now).

I do agree that a lot of people think of [ɑː] as somehow "having an r in it", e.g. they'll say that southerners pronounce "bath" as "barth", but I tend to think this is more to do with how the sound is usually spelt in non-rhotic Northern English than the actual phonology...
Although I guess there are people who would pronounce those words /səharə/ or /ɒrəl/, I'd always pronounce them /səhɑːrə/ and /ɔːrəl/.

It's true that the idea of [ɔː ɜː ɑː] being /ɒr ɛr ar/ is probably due to spelling, but the way I see it phonology is basically arbitrary when it comes to this sort of thing, and the analysis works in my idiolect.
AnTeallach wrote:Interesting that you have the fricative realisation of /d/ but not /t/.
Yes; it seems that Scouse has influenced me to pronounce the alveolar stops more like fricatives sometimes, but T-glottalisation has overriden it for /t/.

User avatar
finlay
Sumerul
Sumerul
Posts: 3600
Joined: Mon Dec 22, 2003 12:35 pm
Location: Tokyo

Post by finlay »

Soap wrote:Sorry, I should have just used SAMPA. I dont know the British rules of vowel length, so I figured I'd just obscure them by using dictionary-style ī and ĭ instead, but that caused more trouble than it solved.
Oh I ignored that; my point is that it can't really be /i:/ or /I/, because, well, it's neither... :?

Travis: decoupled from examples half your words don't make sense. :P

Travis B.
Sumerul
Sumerul
Posts: 3570
Joined: Mon Jun 20, 2005 12:47 pm
Location: Milwaukee, US

Post by Travis B. »

finlay wrote:Travis: decoupled from examples half your words don't make sense. :P
Well, a lot of the underlying logic behind what I stated there is rather hard to understand as a whole, because it tends to involve edge cases where the typical rules one would assume of English phonology really do not work, or matters of distributions of realizations of different words that strongly imply that such is not a matter of mere allophony despite what one would normally assume. However, I can try to explain as best as I can below:

My dialect, obviously, is descended from English dialects that had distinctive vowel length unrelated to any kind of allophonic vowel length, and which lacked phonemic vowel nasalization or pitch accentuation independent of mere distinctive stress. At some point contracted forms gained pitch accentuation independent of mere distinctive stress through some combination of pitch variation tied to distinctive stress interacting with loss of syllables and intonation, but just when and how this occurred is still very unclear to me. Also, the parent dialect to my own on one hand had the vowel mergers common to most General American-like NAE dialects spoken today but lacked any merger between unstressed intervocalic /t/ and /d/ or between unstressed intervocalic /n/, /nt/, or /nd/, unlike many NAE dialects today; these non-mergers are rather important to the diachronics of my dialect, it should be noted.

At this point, for instance, one would get something like:

Aren't they cute vowels and consonants?
ˈɑ̃̂ːr̃nt ˈðeː ˈkʰjuːt ˈvaʊ̯əɫz ˈæ̃nd ˈkʰɑ̃ːnsɨ̃nɨ̃nts

At some point historical distinctive vowel length was replaced with allophonic vowel length and many instances of /n/ in syllable codas were lost, yielding something along the lines of:

Aren't they cute vowels and consonants?
ˈɑ̃̂r̃t ˈðe ˈkʰjut ˈvaʊ̯əːɫz ˈæ̃ːnd ˈkʰɑ̃nsɨ̃ːnɨ̃ts

As I actually realize such today, that would be:

Aren't they cute vowels and consonants?
[ˈɑ̃̂ʁ̃ˤʔt ̚ ˈd̥e ˈk̟ʰjuʔ ˈvɑɔ̯wʊːz ˈɛ̞̃ːŋː ˈkʰãntsɨ̃ːnɨ̃ʔts]

Now, the typical traditional analysis is that this is mere allophony at work, aside from the pitch accent on aren't. For a lot of words, this seems to work fine. The problem, though, is that in corner cases this does not work. Consider an example from a thread on Unilang:
Travis B. wrote:
Mutusen wrote:How is “it’d” pronounced? Like “it”, or should I try to pronounce a D after a T?
I generally pronounce it'd as [ˈɪ̂d̥], contrasting with it [ˈɪʔ] and id [ˈɪːd̥]. My pronunciation of that'd is similar, being [ˈd̥ɛ̞̂d̥]~[ˈðɛ̞̂d̥]~[ˈɛ̞̂d̥] (with other variations upon the initial consonant), contrastic with that [ˈd̥ɛ̞ʔ]~[ˈðɛ̞ʔ]~[ˈɛ̞ʔ] (with other variations upon the initial consonant); for the same of comparison, also compare with dad [ˈd̥ɛ̞ːd̥].

Of course, do not go by what I say here as a guide to English pronunciation. Really.

(One other note - the pitch accentuation is not there for it'd and that'd where the -'d is from had or did rather than would, where then such are [ˈɪd̥] and [ˈd̥ɛ̞̂d̥]~[ˈðɛ̞d̥]~[ˈɛ̞d̥] respectively. However, the vowels are still short rather than long in such.)
In this case, this type of traditional analysis really is not workable. Using historical phonemes without any vowel length associated with them, but with pitch accentuation, we would get:

it'd from it would: /ˈɪ̂td/
it'd from it had or it did: /ˈɪtd/
it: /ˈɪt/
id: /ˈɪd/
that'd from that would: /ˈðæ̂td/
that'd from that had or that did: /ˈðætd/
that: /ˈðæt/
dad: /ˈdæd/

Note that /ˈɪtəd/ for it'd and /ˈðætəd/ for that'd do not work, as such should invariably result in [ˈɪ̂ə̯d̥] and [ˈd̥ɛ̞̂ə̯d̥] respectively due to the nature of intervocalic consonant elision in my dialect, but such do not. Yet, obviously, neither /ˈɪt/ nor /ˈɪd/ work for it'd and neither /ˈðæt/ nor /ˈðæd/ work for that'd, as should be clear from the post I quoted. At the same time, though, positing coda /td/ as an underlying form turns out to be very contrived, as stop clusters of that sort almost invariably turn into geminates in my dialect; /ˈɪ̂td/ ought to actually give something like [ˈɪ̂ʔtː] or at least [ˈɪ̂ʔt] and likewise /ˈðæ̂td/ ought to give something like [ˈd̥ɛ̞̂ʔtː] or at least [ˈd̥ɛ̞̂ʔt], but those are clearly not actual realized forms for it'd and that'd in my dialect. There are no other cases of /t/ just disappearing in such a fashion in my dialect before another stop, so hence such would clearly be a rule contrived to explain the realized form in a way that fits traditional English phonology rather than a rule that actually fits how the phonology of my dialect works as a whole.

I could go on further, but this just gives a bit of insight into the phonological issues present in the analysis of my dialect. I could go on and on and on, but that would make for a very long post that I really do not have the time to write right now.

User avatar
finlay
Sumerul
Sumerul
Posts: 3600
Joined: Mon Dec 22, 2003 12:35 pm
Location: Tokyo

Post by finlay »

The morpheme boundary could be playing into it, then, or something. This is allowed when doing phonology...... :P The way I see it, you've got a bimorphemic word and it's quite likely that you're joining the two together, meaning that you're quite likely to have an underlying form of whatever "it" is + whatever "'d" is. This can then change differently from other clusters of the sort because there's a morpheme boundary. Just sayin'.

I'm curious, though, as to where the hell else you have a cluster like that. They... like... don't exist.

oh and, sound files. Please.

Travis B.
Sumerul
Sumerul
Posts: 3570
Joined: Mon Jun 20, 2005 12:47 pm
Location: Milwaukee, US

Post by Travis B. »

finlay wrote:The morpheme boundary could be playing into it, then, or something. This is allowed when doing phonology...... :P The way I see it, you've got a bimorphemic word and it's quite likely that you're joining the two together, meaning that you're quite likely to have an underlying form of whatever "it" is + whatever "'d" is. This can then change differently from other clusters of the sort because there's a morpheme boundary. Just sayin'.

I'm curious, though, as to where the hell else you have a cluster like that. They... like... don't exist.
But that is pretty much my point right there - at least in my dialect, in corner cases like this the conventional phonological rules that one normally thinks of English as having just do not, well, work without come major contriving on the part of whoever is doing the analysis thereof. The idea that my it'd and that'd actually contain /td/ is just that - a contrivance to get around the idea that an English dialect cannot have distinctive vowel length independent of distinctive vowel quality in a general fashion. I honestly do not believe that it'd or that'd in my dialect have anything like /td/ in them; I was rather just positing /td/ as a potential analysis to illustrate how to maintain something that resembles a traditional English phonology requires positing analyses that, when actually thought about, make no sense from a phonological standpoint.

It is much easier to analyze this case if one presumes that, synchronically, vowel quality is decoupled from both vowel length and the lenis/fortis value of the synchronically-next obstruent before the next vowel, with the apparent coupling between vowel length and the lenis/fortis value of synchronically-next obstruent before the next vowel in many ways being a matter of how things happened diachronically combined with very strong analogy. However, the problem with that is that one cannot coin arbitrary new words or loan arbitrary new loans which have arbitrary vowel length in them; rather, the vowel length can only be assigned through the same historical pathways that existing words had their vowel lengths assigned. And it is this, amongst other things, that makes me really doubt a phonemic-type analysis thereof, as everything works best from a phonemic-type standpoint if vowel quality, vowel length, and following consonants are all decoupled from each other, yet this is clearly not consciously accessible at all, with people coining new words only having access to two degrees of freedom, vowel quality and vowel length coupled with the fortisness/lenisness of the next obstruent before the next vowel, unlike what one would assume to be the case in a phonemic-type model.

Post Reply