| (T.B.A.) |
(chio bei nonth)
|Spoken in:||(somewhere in North America) ((samva en Nof Maweka))|
|Conworld:||A possible future timeline (arguments to the contrary are welcome)|
|Total speakers:||none yet|
|Basic word order:||SVO|
|Tropylium||one very late evening in spring 2008|
A tentative name for a future Anglic language. May be switched to a proper geographic-based one eventually.
Owes much to Futurese.
- 1 Grand Master Plan
- 1.1 Initial dialect features
- 1.2 Initial syntactic changes
- 1.3 Pre-Angrex
- 1.4 Proto-Angrex
- 1.5 Old Angrex to Standard Angrex
- 2 Phonology
- 3 Orthography
- 4 Vocabulary
Grand Master Plan
Initial dialect features
Note that these features' current geographical distribution does not necessarily limit Ängrex's, since a few of them are still spreding.
- Yod-dropping: Early Modern English /iʊ/ becomes /uː/ (GOOSE), preceded by /j/ if there's no initial consonant. Likewise /iʊɹ/ → /ʊɹ/. Typical exceptions, such as sugar, sure, apply.
- The cot-caught, father-bother and pin-pen mergers apply, i.e. /ɒ ɔː/ both merge into /ɑː/, and /ɪ/ before a nasal into /ɛ/.
- The following rhotic vowels are distinguished: /ɪɹ ɛɹ əɹ ɑɹ ɔɹ/ (NEAR SQUARE NURSE START NORTH). /oɹ/ (FORCE) merges into /ɔɹ/; /ʊɹ/ (CURE) merges varyingly† with /ɔɹ/ or the disyllable /uəɹ/, or plain /əɹ/ in certain lexemes.† Mergers apply also before medial /ɹ/: e.g. /e.ɹ æ.ɹ/ → /ɛɹ/, /ʌ.ɹ/ → /əɹ/, /i.ɹ/ → /ɪɹ/.
- Collapse to a three-vowel system in final unstress'd position: /iː/ (HAPPY) becomes /eɪ/ (FACE), while /ə(ɹ)/ (COMMA, LETTER) becomes /ɑː/. /oʊ/ remains. (Phonetically, these are basic [e a o].)
- Unstress'd /ə/ and /ɪ/ remain mostly separate when pretonic, sometimes† also when posttonic.
- Shwa + sonorant combinations are realized as syllabic consonants [m̩ n̩ ɫ̩], especially† word-finally.
The following changes are best considered phonetic detail, since they do not disturb the phonological system.
- /ɑ æ iː uː ʊ/ → [a eə ɪj ʏw ʏ]
- We'll nevertheless write the quality of the GOOSE and FOOT vowels as /ʉ ʏ/ from here on.
- [ʊ] remains before coda /l/
- need to elaborate on the fate of /æ/ per environment, difthong decay, and on other pre-sonorant mergers (aU aI > æə aə / _S
- Length is by now secondary to quality (and won't be marked, either).
- Whine-wine merger: /ʍ/ merged into /w/.
- /tɹ tj dɹ dj/ are affricated to /ʧɹ ʧ ʤɹ ʤ/. (Note that palatalization of /sɹ sj zj/ is older and applies to all English varieties, AFATAK.)
- Medial flapping of /t d/ to [ɾ] when posttonic. No glottalization develops (except, as widespred, to zero between a fricativ and a syllabic consonant — soften [sɑfn̩] → <sáfang>, rustle [ɹʌsɫ̩] → <wúso>)
- Interdental loss: /θ ð/ become /t d/ in onset position, /f v/ in coda.
- eater [ˈɪjɾa], ether [ˈɪjta]
- Coda /d/ is lost after /n/.
- The cluster /ns/ (and possibly /nz/ - unclear due to later changes & the only common examples being bronze and frenzy) inserts epenthetic /t/ (or /d/) to become /nts/ (or /ndz/). Plurals and possessivs in //-S// unaffected, as are, by this stage, any prefixes in /(C)Vn-/ (in-, con-, etc.)
Initial syntactic changes
- Gonna and wanna
- got becomes a compulsory auxiliary in possessiv constructions, while the have element is increasingly cliticized
- Cliticization of copulas's also common, with (t)here's taking place of (t)here're
- Imperfect forms of strong verbs are eliminated and replaced by the perfect. (Some cases proceed in the opposite direction, mostly when the perfect includes -en in addition to any vowel changes, but still to the effect of only allowing one past form.)
- one becomes a compulsory pronoun in constructions involving adjectivs as referents
- gimme reanalyzed as a monotransitiv "to part with"
By now the American English speech community is increasingly fracturing. Register-switching between vernacular and classical English starts to resemble diglossia similar to the current (2009 CE) situation in the Arabophone world, and with the standard language taking the role of the lingua franca, the dialects are free to diverge.
Further spred of certain changes
- Flapping of /d/ is extended to word-final posttonic position. /t/ persists as an unreleased [t̚].
- /nʃ/(and /nʒ/?), prefixal /n/+/s/ and tautosyllabic /mf/ also undergo epenthesis (therefore eg. <thancang> from tension)
- The pattern of singular /-əs/ : plural /-aɪ/ is generalized to all applicable words, and then some.
- "Some" includes especially polysyllables ending in /-əz/, and monosyllables ending in /-s/, which will be now pluralized as /-saɪ/ (but not those with irregular plurals, e.g. mice)
More vowel changes
(Needs splitting between this and the previous stage)
- /eə/ further raised to /ɪə/
- Reduction of unstressed vowels continues. In closed syllables, sufficiently unstress'd† /i e/ typically reduce to /ɪ/, /ʉ/ to /ʏ/, /ɛ æ ɑ/ to /ə/. Pretonic initial /ə/ (but not /ɪ/) tends to be lost entirely.
- Before a vowel, unstressed /i ʉ/ reduce to /j w/. The sequences /i(j)ə ʉ(w)ə/ (e.g. idea, doer) coalesce to /ɪə ʊə/.
- Unreduced /æʊ ɔɪ/ become biphonemic /ɪə.ɔ ʊə.ɛ/ (word-finally /ɪə.o ʊə.e/). Unreduced /aɪ/ persists.
- /ʌ/ becomes essentially /ə/ even when stress'd.*
- /ɑ/ → /ʌ/ before a nasal, paralleling the situation with front vowels (does FOOT + nasal exist??)
- /ɪɹ/ (NEAR, MIRROR) becomes /ɪəɹ/, still monosyllabic but now equal in quality to former /æ/ follo'd by /ɹ/ (bear in mind that former /æɹ/, as in carry, has turned into /ɛɹ/ tho).
- /ɔɹ/ raises to /ʊɹ/, then similarly breiks to become /ʊəɹ/; /əɹ/ backs to take its place as new /ɔɹ/.
- Syllabic /n/ → /ŋ/ in word-final position
One of Angrex's most caracteristic features.
- The main chain consists of /ɹ/ → /w/, /w/ → /ʋ/, /v/ → /b/ (a push-pull chain with the 2nd change coming into place the last). Only the onset position is affected.
- /w/ remains after a velar, a testament to its imbudgency.* (So is the shift from /ɹ/ directly to /ʋ/ in the British descendants of English, but let's not go there.)
- Coda /v/ → /ʋ/, further becoming syllabic if another consonant precedes
- Former coda /ɹ/ or semivowels don't count as consonants anymore
- Coda /l/ does, by which the changes affecting it can be dated as being later
- Syllabic /ʋ̩/ is also produced from /əv/ (principally: of, 've)
- /v/ before a syllabic consonant is treated as a coda.
- Coda /l/ separates as syllabic after a (phonetic) semivowel, ie. after the difthongal vowels /i ʉ ɪə ʊə aə aɪ/.
- …after which it vocalizes: → /w/, the syllabic form → /ʊ/.
- New cases of stress'd /ʊ/ develop from /ow/ before an additional coda consonant.
- Summarizing, sail will help pull all old dull coal → <xiu veu haup phou o or du khu>; boil cool eel tile pal → <bwáyo khyowo eyo thayo pháo?>; fillen valve → <fewang bou> (For fillen, see the new syntax section)
*Yes, I made that up. Er, I mean, it demonstrates the increasingly productiv usage of certain suffixes in Angrex. ;)
An era of estabilishment of isoglosses and fledgling linguistic boundaries. This particular family
Liquid shifts, pt. 2
Also very contributing to the general flavor. Dateable as somewhat later than the labial shift tho.
- Onset /l/ → /ɾ/ after another consonant (even when separated by a syllable breik)
- except before an existing /ɾ/ later in the word
- Regressiv bilabial/labiodental harmony between /w ʋ/, including also their syllabic counterparts /ʊ ʋ̩/.†
- Due to the later change from coda /ʋ/ to /w/, this doesn't come up much in Angrex itself, but it does explain how wary → <vávi> or rewind → <wegán>, as well as the origin of some morphological elements.†
This series of sound changes forms major isoglosses among the Anglic languages.
- The trigger is the loss of /s/ before another consonant. This leads to the phonemicization of aspiration of voiceless stops in the onset — as well as of word-initial /ɾ/.
- Medial stops are also geminated when /s/ is lost.
- Coda stops are normally unreleased, with the exception of the position before another obstruent, or a non-tautosyllabic consonant such as [ɫ̩] (syllables with a consonantal nucleus may not contain an onset) or [m] (stop+/m/ not being a valid initial cluster). After its inception by the previous change, aspiration develops also in this position, in stressed syllables only.
- Any coda affricates are 'fcors released, but generally don't develop aspiration. An exception, most likely analogy-motivated, is the cluster /ʧt/ when resulting from /ʧ/ plus the past tense suffix //d//.
- Additionally, /pf ts/ → /pʰ tʰ/ in all positions.
- This system is then muddled by application of anti-Grassman's Law, i.e. in a word containing two aspirates within one syllable of one another, the latter is deaspirated.
- /h/ triggers deaspiration regularly; it's however medially lost in all cases (e.g. megahertz → <magát>)
- /h/ is also lost initially before any high vowel or glide.
- However, in syllables with an initial lo vowel or /ɾ/, any possible aspiration in the coda is transferred to become an intrusiv initial /h/ (e.g. opt → <háp>, slacks → <hrákhs>)
- All coda stops after an obstruent or a homorganic nasal are lost. The first part of this change is common enuff even in contemporary English dialects, but the consequences for aspiration sho that at least in /pt kt/ this must postdate the development of phonemic aspiration.
- To compensate for the increasingly stronger allophonic palatalization of ordinary coronals before front vowels, the postalveolar sibilants become apical in articulation, in enunciated speech even retroflex. These retain their inherent labialization, and a folloing /w/ is assimilated. An /f/ or /v/ from former /w/, however, is metathesized, giving eg. <pháfcúk> from patchwork, <defxa> from dishwasher (note the haplology) and <pháse(u)jri> from passageway.
- Mid vowels become long high vowels: /e o/ → /jiː uː/
- Near-high vowels become short mid vowels: /ɪ ʏ ʊ/ → /e ø o/
- furthermore /ø ej/ → /jo jej/
- Rhotic vowels also become long mid (or lo) vowels: /ɑɹ oɹ əɹ eɹ/ → /ɑː oː ɘː ɛː/
- Difthongs in -ə smooth into long lo vowels: /ɪə ʊə aə/ → /jaː waː aː/
- There are two theories to what's going on in here with the front vowels. One is that this is simply palatalization and the glide found in eg. cube → <khiobo> is excrescent in nature; the other is that /j/'s are actually broken off vowels, but simply lost in most situations. The behavior is at any rate identical to that of inherited instances of /Cj/ in most, perhaps all cases.
- No glide nor palatalization appears after a liquid (/l ɾ ʋ/), a postalveolar, or a bilabial.
- Non-liquid alveolars reddily palatalize: /tʰ t d s z n/ + /ʲ/ → /tɕʰ tɕ dʑ ɕ ʑ ɲ/
- This occurs also after /iː j/.
- A plain [j] appears after velars and /f/.
- /w/ palatalizes to /ɥ/ when occurring after a front vowel or palatal consonant; no effect is seen elsewhere.
- Word-initially (or after a vowel?) [j] remains when resulting from /ø ɪə/.
- In contrast to all the palatalization hijinx, there's no interplay between /w/ and consonants to note. *
Old Angrex to Standard Angrex
Yet more vowel shifts
- A sequence of /ɛ a ɑ/ plus a non-close vowel becomes a long vowel. /i u/ turn into [j w] if adjacent to a vowel.
- The sequences /əj əw ɑw/ become [i u oː], even before another vowel. This counterfeeds the previous monofthongization. (Equivalently: before a vowel /əj əw/ become [ij uw]).
- Vowels are additionally lengthen'd:
- always word-finally
- in open stressed syllables, if the next vowel is short†
- Vowels are shorten'd:
- in superhevvy syllables (with a CC coda), except those resulting from a vowel sequence
- Syllabic consonants are lost: [ʋ̩ m̩ n̩ ŋ̩] → [u am an aŋ] (always short)
- This postdates lengthening/shortening, both due to the consistency of resultant length, and since consonants between a syllable with a vocalic nucleus and one with a consonantal nucleus† are assigned to the vocalic syllable if possible, increasing its weight. Frex action → <yakhshang>, not *<yákhshang>; and kitten → <kherang> has short [e], not long [eː].
- Coda /ʋ/ becomes /w/, while onset /ʋ/ becomes /v/ or /f/, modulo voicing of a possible preceding consonant.
- Lo front vowels collapse: /aː/ merges into /ɑː/ except when word-final, while /ɛ a/ merge as /æ/. (This counterfeeds the monofthongization of /ɑw/.)
- Shwa is inserted after word-final voiced stops and affricates. (?)
- Also postdates prosodic shortening/lengthening.
- Unstressed shwa then proceeds to become [e] if there is an immediately preceding palatal, and [o] if there is an immediately preceding labial glide or retroflex.
- Stressed shwas become [ɨː] when long, [ɑ] when short.
- Coda /j/ is lost between front vowels and palatals, likewise coda and onset /w/ between labial vowels and labials, with compensatory lengthening if the vowel precedes.
- Voiced affricates lenite to fricativs: /dz dʐ dʑ/ → /z ʐ ʑ/. Before nasals, however, the opposite development is found.
- Of two nearby prevocalic /w/ (and possibly /j/?), the 2nd dissimilates → [ɣ] → /ɡ/.
- Certain awkward consonant clusters are resolved:
- /n/ after an alveolar stop becomes /ɾ/
- /nɾ/ becomes /nː/
- Before another obstruent, coda /k/ and /kʰ/ become [x], similarly /p(ʰ) b/ become /f v/.† Word-final /k/ tends to be lost,
- Any obstruents stranded between a nasal and another obstruent, or between two obstruents, drop out.
Distribution particulars not work'd out, other than that Stop Merger and Back Chainshift occur in the same branch.
- /n/ → /ɾ/ after other stops as well
- Coda /ɾ/ → /l/†
- Alternately, other stops may spirantize in this position as well, leading to correspondences such as [xn] ~ [kɾ] for standard [kn]
- /fw fj hɾ/ → /xw xj xɾ/, leading to phonemic /x/
- Stop merger: aspirated/geminate → voiceless; voiced/unaspirated → voiced
- Long back(ish) vowel chain shift: /ɑː/ → /oː/ → /uː/ → /ɨː/ → /au/ → …
- Alternately, difthong simplification: /iu eu/ → /yː øː/, loss of vowel length with a split of /eː e oː o/ → /e ɛ o ɔ/
Difthongs iu eu ei au ai ou oi (ui??)
Per IPA, except:
- Obviously (?), <r> is /ɾ/, <y> is /j/ and a digraph with <h> signifies aspiration.
- <c x j> are /tɕ ɕ ʑ/ before an orthographic <i> (itself silent before another vowel), /tʂ ʂ ʐ/ elsewhere.
- A silent <r> occurs between a retroflex and /i/. E.g. jail → <jriu> /ʐiw/, dale → <jiu> /ʑiw/.
- need to decide what to do with coda position
- [x] is treated as an allophone of /kʰ/ and written to the effect.
- <ñ ng> are /ɲ ŋ/.
- The lo vowels are <a á> = /æ ɑ/; unstressed, <a> varies from [a] to [ə]. <ú> is /ɨ/.
- The acutes here derive from a superscript <h>. *
- Vowel length is normally not written, but if needed, macrons can be used (acute + macron is generally substituted by a circumflex).
- vún, chiou, thwei, fwá, fayu, sekh, sabang, ic, nán, than, elabang, thfau, thoceñ, fwáceñ ... thfanni ... ánjor, chiaosang ... melang, belang, chelang ...
- fús, sakhang, thor, fwáf, fef, sekhf, sabangf etc.
- súnni, múnni, chioji, wanji, thoji, fwairi, xiarji
- jámawi, fabawi, mác, ipo, mi, joun, jolai, ágús, súftamba, úkhtuba, núbamba , dúsamba
- vac, gwi, bra; war, wágenjo, yalu, gweñ, thokwas, sayang, brou, annegu, phopo, majanna; pheng, bwang
- saram, magnúsam, jalmeñam, pharsam, khyausam, khum, ájang, khápa, zeng, seuba, then, phlárenam, gor, mokyowi, lar, yowiñam
- chayanggo, wakhcianggo, kwa, phannúgang, alef, sokho; pheumer, khyobo, khun
- phwabet, khwapo, sajang, anthang, lournang, khyapang, mija, khono, januo, jarmuo
- ija, yáfweka, acchrila, màweka, yowop, anthákhte