Hello there! This article is mainly just for fun as well as to test how well my perception of phones is in narrow transcriptions. I'm basically going to be attempting how to transcribe the realizations of the standard English phonemes in my idiolect by being as narrow as possible with my transcriptions. This'll mean I'll be going all out in terms of my liberal usage of the extended IPA as well. Anyhow, I hope this is an enjoyable read albeit a little out there in terms of topic.
How I realize the standard English consonants in my idiolect. My definition of a consonant in this article, broadly speaking is literally anything that can form the onset and coda of a syllable.
This section will mainly go into any notable differences between my realizations of phonemes in terms of PoA (place of articulation), and the broad transcriptions given in the Wikipedia page for English phonemes.
To start with something a little basic, the ways in which I release my stops are purely alveolar. This also extends to my pronunciation of the /l/ sound, so the super narrow transcriptions are as follows: [t͇, n͇, l͇]. This doesn't really deviate from the standard, though. Just wanted to use the extended IPA for fun. You might have already noticed, however, that the phone [s͇] is starkly absent from the list above. This is because of something interesting I've noticed in my normal speech after becoming a little more aware of the PoA in phonetics and about the more orthodox places of articulation in standard English. Namely, that rather than pronouncing my /s/ with my tongue up, I actually pronounce them with my tongue down! This actually extends to all the sibilants. Now, this explanation is a little too general for the goal of this article, so let's take a dive into what the narrow transcription would be.
As stated earlier, I realize my /s/'s with my tongue down rather than with my tongue up. More precisely, the top of my tongue comes in contact with the back of my bottom teeth! So in reality, the way I'm realizing my /s/'s is a little more like a voiceless laminal alveolar fricative, as in [s̻]. This also applies to /z/, which is just a voiced version of that, so I'd transcribe it narrowly as [z̻]. This isn't all, however, as the tongue-down articulation extends to /ʃ/ and /ʒ/ as well! And my realizations of these I would say are particularly peculiar, to say the least.
This is because rather than simply applying the tongue down rule and pronouncing them as a [ɕ] and [ʑ] respectively, the PoA is actually lower than the bottom teeth. That is to say that the tip of my tongue actually comes in contact with the back of my gums, or in fancy medical terms, my mandibular lingual gingiva! What actually constricts the air and produces that iconic hissing sound in these post-alveolar sibilants is actually the blade of my tongue rather than the front of it. And due to the hissing sound that a simple [ɕ] and [ʑ] produces being a little too high in pitch, the lowering of my tongue tip compensates by lowering the pitch of the phones to something which sounds more similar to the standard articulation of them. This means that I actually realize them phonetically as [ɕ̞] and [ʑ̞]! The affricated forms also have the PoA of the initial plosive element lowered to the gums as well, resulting in something phonetically more like [t̻̞͡ɕ̞] and [d̻̞͡ʑ̞] as opposed to [d͡ʒ] and [t͡ʃ].
Moving onto something a little more dialectal than ideolectal, the way I produce the voiced alveolar approximant, aka the weird English rhotic /ɹ/, is more bunched, which is fairly common in North America in general. Though there isn't really a standardized symbol for the bunched r, it's typically transcribed as [ɹ̈], denoting the tongue's more centralized position.
As is standard with nearly all English varieties, I also affricate /t/ and /d/ before my /ɹ/ as well, resulting in /tɹ/ looking a little more like [t̻̞͡ɕ̞ɹ̈]. That's all in terms of my unorthodox realizations of the major phonemes in English I'm afraid, however, we still have a few more marginal consonants left!
The glottal stop, /ʔ/, is basically only present in the word uh-oh though it is essentially present before initial vowels and between word boundaries. I don't pronounce any French loans with the marginal nasal vowels, however, interestingly enough, I find myself pronouncing [x] marginally solely in the word Amazigh for no apparent reason. The only theory I have for why it is the way it is likely due to that being the first pronunciation of it I was exposed to from linguistics videos and whatnot.
Something interesting to note which is characteristic of most North American varieties would be the lack of a true light l, [l]. I've noticed that my realizations of /l/ allophonically fluctuate between [lˤ ~ lˠ], with the pharyngealized version being the norm before all vowels except for the high vowels, where the darkness is a little weaker, being more akin to weak velarization.
In addition, something which is fairly prevalent in North American dialects and my own speech is the loss of the yod, essentially just the /j/ glid found before the tense vowel /u/ (Which can be observed by the name of the letter as well, /ju/). Interestingly enough however, this loss of yod is impartial as we can see from the name of the letter, as at least in most North American varieties, it was only lost before alveolar consonants /t/, /d/, /n/, /s/, /z/, /l/ as well as /θ/. It is fairly alive and well after the rest of the consonants (which historically had it) and when split off from the consonant before it, being the sole coda of the preceding syllable, even causing /hj/ to be realized more as a [ç] which is unsurprisingly the allophone of /h/ before /i/ as well. This means that I pronounce "new" /njuː/ as [nʊ̜̈ü̯], yet pronounce continue /kənˈtɪnjuː/ as [kʰn͇̩ˈt͇ʰɪn.jʊ̜̈ü̯]
Continuing to stick to features which are tell-tail features of North American varieties, the allophonic change of both my /d/'s and /t/'s into the voiced alveolar flap [ɾ] intervocalically is absolute, meaning that outside of careful speech, "ladder" and "latter" are complete homophones. This same [ɾ] can also be found variably in my realizations of the coda /d/ while my /t/ is simply not released, being realized as a [t̚] instead. This allophonic non-releasing of my voiceless coda stops isn't complete however, as it's a little more variable for the coda /p/, and is very rarely done for the coda /k/. In addition to this, my voiced codas, including /d/occasionally, are somewhat devoiced, however not completely, giving rise to something more similar to [d̥] [b̥] [g̥] word finally.
As a little conclusion to this section on codas, due to the fact I'm geographically located more northwest of the continent, my variety of English is rhotic, meaning that the coda /ɹ/ is preserved.
Going into features which are a little more prevalent across all English varieties, my dialect has the wine-whine merger. This merger, as the name entails, completely merges [hʷ] and [w] into [w], with the distinction only being retained in more conservative varieties of English generally.
As with nearly all English varieties, I also have allophonic aspiration present in my voiceless plosives word initially and in stressed syllables, which is allophonically absent when the plosive is the second consonant of a consonant cluster and generally weaker or nonexistent in unstressed syllables.
Something fairly common cross-linguistically as well, is the merging of the PoA of my nasals before labio-dental consonants, turning both /m/ and /n/ into [ɱ] variably.
Though not as numerous as the features described above, I thought it would be interesting to discuss the addition of and deletion of certain phonemes in specific environments as well.
For starters, there is a helping plosive sound which is inserted between nasal consonants and voiceless sibilants which merges its PoA with that of the nasal before it. This is present even when the nasal and sibilant are separated by syllable boundaries, hence why /hɛns/ is more like [hɛnts] and /səm.θɪŋ/ sounds more like [səm.pθɪŋ].
That being said, this isn't the full picture when trying to transcribe my speech as narrowly as possible, as there is another example of epenthesis within the word [səm.pθɪŋ], which is the diphthongization of [ɪ] to [ɪi̯] before [ŋ].
On to the deletion of sounds for a change of pace, I seem to have medial cluster reductions most commonly in word final positions, with /kts/ turning it into more of a [ks] in rapid speech.
The realizations of vowels in my speech. Going off of the loose definition of consonants, I'm considering anything that can form the nucleus of a syllable to be a vowel.
A lot of my vowels are actually a lot more centralized and lowered, with /eɪ/ like an [ëi̯], /oʊ/ like an [ɵ̜ü̯] and /ʊ/ as more of an [ɯ̞]. This also extends to my realizations of some of the non-diphthongized tense vowels, with /æ/ [æ̈ː]. /i/, however, remains as [iː] as does /ɑː/ as [ɑː], only being long which is characteristic of the tense vowels. Something worth noting however, is that it seems that the traditional long /u/ vowel is potentially going through a process of breaking, becoming more diphthong-esque, as I phonetically realize it as [ʊ̜̈ü̯], with a clear movement from unrounded to rounded lips, similar to [ɵ̜ü̯].
An interesting phenomenon which came as a surprise to me was the fact that my phonetic realizations of /ə/ and /ʌ/ were actually different, despite me previously believing they were phonetically merged in my variety. That being said, they're phonologically more like allophones of one another, with /ʌ/ appearing in stressed syllables and /ə/ elsewhere. Additionally, this /ʌ/ vowel also keeps in line with the general centralization of my vowels, being realized closer to [ɐ]. The rest of the lax vowels are produced essentially the same as the phonological transcriptions. Interestingly enough, there's a general tendency for /ə/ within word boundaries to be realized as [ɪ] outside of open initial and final word positions.
As with my consonants, my variety of English also has its fair share of mergers and variation in production, albeit, to a far larger extent.
A couple of mergers which are exceedingly prominent in the west coast of North American varieties would be the cot-caught merger and the father-bother merger, in which /ɑː/, /ɒ/ and /ɔː/ all coalesce into [ɑː], causing words which traditionally had different vowels to rhyme with one another or become homophones all together.
As for the general allophonic tendencies among all my vowels, there is a marked lengthening of all vowels regardless of tenseness before voiced codas. Among other allophonic phenomena occurring before certain coda consonants, vowels before nasal codas, and oftentimes nasal consonants in general are somewhat nasalized as well.
With the phonological monophthongs out of the way, I'll be taking a look at my own diphthongs along with vowels with the coda r.
To begin, my diphthongs seem to be slightly higher and or centralized than what is represented in the IPA, with /aʊ/ being more similar to [æ̈ü̯] and /aɪ/ being closer to [äi̯].
Something interesting to note however is the fact that they have different realizations before voiceless consonants, turning into [ɐü̯] and [ɐi̯] respectively. This phenomenon is applied prior to the intervocalic flapping of /t/ and /d/, actually becoming the sole denominator in differentiating the two without context, such as in rider and writer. (Rendered into my narrow transcription style as [ɹ̈äi̯.ɾɹ̩̈] [ɹ̈ɐi̯.ɾɹ̩̈] respectively)
Notably, /ɔɪ/ is also raised to something a little closer to [öi̯] as is/ɔːr/to [öɹ̩̈], with the initial vowel of the diphthongs being far rounder and back than that of [ɵ̜ü̯].
As with many North American variaties has gone through an incomplete shift from/ɒr/ to [ör], with some words such as "tomorrow" and "sorry" retaining the /ɒr/ as an [ɑr]. That being said, for the word "sorry", there is a fair amount of fluctuation between [ɑr] and [ör], though there is definitely a leaning towards the former.
A general allophonic process observed before my retained coda r is the shortening of tense vowels before it, along with the actual cutting of the final vowel segment in diphthongic vowels, such as [ëi̯] being rendered as [ëɹ̈] before it. /o/, however, is fairly exceptional in this regard, however, this could be explained by the fact that it was originally actually an /ɔ/ rather than an /o/, with the cot-caught merger not affecting it because the vowel was before said coda r.
That being said, vowels before the coda r have their fair share of mergers within my normal speech, as /ær/, /ɛər/ and /ɛr/ have completely merged into [ëɹ̈] along with /ɪər/, /ɪr/ and /iːr/ merging into [iɹ̈], resulting in the iconic homophony of marry, merry and Mary in most west coast varieties.
Other than that, /ʊər/ is generally rendered as [üɹ̈], however interestingly enough, can variably be changed to or [ɹ̩̈] if [üɹ̈] is preceded by a yod. (e.g. "cure" /kjʊəɹ/ being pronounced variably as [kʰjüɹ̈ ~ kʰjɹ̩̈])
The reason why it has fully morphed into a syllabic [ɹ̩̈] is due to the fact that the schwa along with /ʊ/ /ɜː/ and /ʌ/ are completely lost before sonorants, turning the consonant after it into a syllabic consonant which becomes the nucleus of the syllable itself. This process can also turn the onset of the syllable into the coda of the syllable before it if the resulting consonant cluster is permitted, concequently blocking the intervocalic flapping of /t/ and /d/. An example of such is "mountain" /maʊn.tən/ (Phonetically [mæ̈ü̯n.tɪn͇]) which shifts to [mæ̈ü̯n.tn͇̩] and further to my normal pronunciation of [mæ̈̃ü̯̃nt͇̚.n͇̩] along with the allophonic nasality. An example of the the intervocalic flapping getting blocked is in the word kitten /ˈkɪtən/, which shifts to [kʰɪt͇̚.n͇̩] as opposed to the expected *[kʰɪ.ɾn͇̩].
To conclude, here're a few IPA charts to visualize the contents and give examples to summarize the contents of the article a little more clearly. Also, side note though all phones will be represented as opposed to phonemes to keep in line with being as narrow as possible.
(Based solely on the PoA differences)
| bilabial | labio-dental | dental | alveolar | alveolo-palatal | palatal | velar | glottal | |
|---|---|---|---|---|---|---|---|---|
| stop | [pʰ ~ p ~ (p̚)], [b ~ b̥] | [t͇ʰ ~ t͇ ~ t͇̚], [d̳ ~ d̳̥] | [kʰ~ k ~ (k̚)], [g ~ g̥ ] | [ʔ] | ||||
| fricative | [f], [v ~ v̥] | [θ], [ð ~ ð̥] | [s̻], [z̻ ~ z̻̥] | [ɕ̞], [ʑ̞ ~ ʑ̞̥] | [ç] | (x) | [h] | |
| affricate | [t̻̞͡ɕ̞], [d̻̞͡ʑ̞ ~ d̻̞̥͡ʑ̞̥] | |||||||
| nasal | [m] | [ɱ] | [n͇] | [ŋ] | ||||
| approximant | [ɹ̈] | [j] | [ɰʷ] | |||||
| lateral approximant | [l͇ˤ ~ l͇ˠ] | |||||||
| flap | [ɾ] |
| front | central | back | |
|---|---|---|---|
| high | [iː] | ||
| near-high | [ɪ] | [ɯ̞] | |
| mid-high | |||
| mid | [ə] | ||
| mid-low | |||
| near-low | [æ̈ː] | [ɐ] | |
| low | [ɑː] |
| high | near-high | mid-high | mid | mid-low | mid-low | near-low | low | syllabic | |
|---|---|---|---|---|---|---|---|---|---|
| high | [ʊ̜̈ü̯] | [ɪi̯] | [ɵ̜ü̯] , [ëi̯] | [æ̈ü̯ ~ ɐü̯], [äi̯ ~ ɐi̯] | |||||
| sonorants | [iɹ̈], [üɹ̈] | [öɹ̈], [ëɹ̈] | [ɑɹ̈] | [ɹ̩̈], [l̩ˤ], [n͇̩], [m̩] |
Finally, just to wrap things up and for fun, here's a transcription which summarizes everything up fairly well. The text is Article 1 of the Human Declaration of Human Rights
All human beings are born free and equal in dignity and rights.
[ʔɑl͇ˤ ˈçʊ̜̈ü̯.mn͇̩ ˈbiː.ɪi̯ŋz̻ ʔɑɹ̈ böɹ̈n͇ fɹ̈iː ʔn͇̩(d̳̥) ˈʔiː.kwl̩ˤ ʔɪn͇ ˈd̳ɪg.n͇ɪ.ɾiː ʔn͇̩(d̳̥) ɹ̈ɐi̯t͇s̻ ‖]
They are endowed with reason and conscience and should act towards one another in a spirit of brotherhood.
[ðëi̯ ʔɑɹ̈ ʔɪn͇.ˈd̳æ̈ü̯ɾ wɪθ ˈɹ̈iː.zn͇̩ n͇̩(d̳̥) ˈkʰɑn͇.ɕ̞n͇̩t͇s̻ | ʔn͇̩(d̳̥) ɕ̞ɯ̞ɾ ʔæ̈kt t͇ʰə.ˈwöɹ̈d̳z̻̥ wɐn͇ ʔə.ˈn͇ɐ.ðɹ̩̈ ʔɪn͇ ʔɐ ˈspiɹ̈.ɹ̈ɪt͇̚ ʔəv̥ ˈbɹ̈ɐ.ðɹ̩̈.hɯ̞ɾ ‖]