alt-English scratchpad 2

[Nortaneous]

There's been a lot of talk about alt-English conlangs. Most of the examples IMO are too close to modern standard English - closer even than attested English dialects! - so I have decided to try to make my own. (This is also useful for Allosphere-related endeavors, since at some point I'll probably lift most of the verbal system into Rau, which is intended to be the Mid-Atlantic English of the trans-Zhjumna coast anyway.)

There are a number of developments, or at least vague tendencies, which are attested in English dialects, or at least Frisian ones. Some of them will be preserved here; others won't.
- The development of a length contrast orthogonal to the tense/lax contrast. This is attested in a handful of Frisian (and I think also German) dialects and I'm not really sure how it happened. I will probably not do this.
- The development of reduced forms of the nominative personal pronouns into consonantal prefixes on generally vowel-initial auxiliaries. It's only by historical accident that English is not French, and it seems reasonable that removing French from English (presumably by ceding the Midlands to the Norman language) should turn English into French.
- Total absence of the GVS, as in the Forth and Bargy dialect and old Irish English in general
- Partial absence of the GVS, as in Scots (where ē ī > i ai, but ō ū > ø u)
- Fortition of /θ ð/ into /t̪ d̪/ contrasting with /t d/ and reinforced by conditional shifts of /t d/ > /t̪ d̪/.
- More consonant changes than are preserved in any existing English standard:
-- The merger of /v w/ (at which point it may as well be Dutch)
-- /kl gl/ > /tl dl/ (which many Puritans had and preserved for centuries)
-- Various treatments of the labiovelars, including /kw/ > /hw/ (Shetland; cf. /hv/ > /kv/ in Icelandic) and /hw/ > /f/ (various dialects; I've been told that the old prescriptive Southern pronunciation of /hw/ was [ɸʷ], so this isn't surprising)

(Are there any dialects that didn't merge /eu iu/?)

There are also some cross-variety differences in common West Germanic processes.
- In Middle Dutch, the vowels resulting from open syllable lengthening didn't merge with the pre-existing long vowels: high vowels lowered when lengthened (similarly to English), and inherited ē ō had diphthongized to something like /eə oə/ or /eɛ oɔ/.
- Absence of high vowel breaking is attested from some German dialects in addition to outlying English dialects

Some outlying English dialects are of utility in demonstrating the range of possibilities. For example, Forth and Bargy:
- Retention of root-final affricates: ich "I", bydge "buy", lidge "lie"
- Retention of /x/: ligt "light", nickht "night"
- Absence of the GVS: dhreeve "drive", heeve "hive"; eal "eel", lear "lere (empty)" (ea /e:/ is known from Irish loanwords); laady "lady", kaake "cake", faace "face"; agyne "again", gryne "grain"; pyle "pail"
- A GVS-like shift of ū > ɛu: greound "ground", keow "cow"
- Lowering of ɛ̄: haade "head", laafe "leaf", maate "meat" (but deed "dead", leed "lead")
- Raising of ā before /m/ and /nd/: gaume "game", llawm "lamb", lhaung "long"; hone "hand", brone "brand", lone "land", sthrone "strand"
- Absence of general ā-backing: bane "bone", dhraat "throat" (this is presumably of use in dating the split)
- Probable merger of /ɛu iu/ to /iu/: vleeu "flew" (but yowe "ewe")
- ɛ > ā before r: zaareth "served", aar "their"
- Occasional lengthening: waant "want", vaat "fat", weend "wind"
- Difficult outcomes for back vowels: zoween "sowing", knouth "knows", chourch "church", gooun "gun", jooudge "judge"; zoon "soon", woode "would", zoot "soot", woork "work"; hint "hunt", piff "puff", pit "put", zin "sun"; rub "rib", vurst "first"

However I have many garden-related obligations rn; more later

[Otto Kretschmer]

When did this alt English start diverging from our one?

Earliest possible point of divergence (around 650-700) would be most interesting. I imagine alternate English that has undergone different sound and grammatical changes than our English and is even more distinct from the rest of West Germanic languages

[Man in Space]

I am curious to see how this turns out.

[Rounin Ryuuji]

There are a number of developments, or at least vague tendencies, which are attested in English dialects, or at least Frisian ones. Some of them will be preserved here; others won't.
- The development of a length contrast orthogonal to the tense/lax contrast. This is attested in a handful of Frisian (and I think also German) dialects and I'm not really sure how it happened. I will probably not do this.
- The development of reduced forms of the nominative personal pronouns into consonantal prefixes on generally vowel-initial auxiliaries. It's only by historical accident that English is not French, and it seems reasonable that removing French from English (presumably by ceding the Midlands to the Norman language) should turn English into French.
- Total absence of the GVS, as in the Forth and Bargy dialect and old Irish English in general
- Partial absence of the GVS, as in Scots (where ē ī > i ai, but ō ū > ø u)
- Fortition of /θ ð/ into /t̪ d̪/ contrasting with /t d/ and reinforced by conditional shifts of /t d/ > /t̪ d̪/.
- More consonant changes than are preserved in any existing English standard:
-- The merger of /v w/ (at which point it may as well be Dutch)
-- /kl gl/ > /tl dl/ (which many Puritans had and preserved for centuries)
-- Various treatments of the labiovelars, including /kw/ > /hw/ (Shetland; cf. /hv/ > /kv/ in Icelandic) and /hw/ > /f/ (various dialects; I've been told that the old prescriptive Southern pronunciation of /hw/ was [ɸʷ], so this isn't surprising)

Interesting so far; I'm not personally fond of the clusters /tl dl/; if I were to offer a unique change to substitute, /kl gl/ > /kw gw/ or /kl gl/ > /kj gj/, especially the latter (Italian-style) would be my pick.

(Are there any dialects that didn't merge /eu iu/?)

Apparently some North American and Welsh dialects preserved /iu/, but in a vacuum, I think /iu y(ː)/ > /yː y(ː)/ would be most likely, with /eu/ being left to form a new phoneme /øː/ or diphthong /øy/ > /ey/, or something like that, such that /fiu/ (few) > /fyː/ fue, but /deu/ (dew) > /døy/ > /dey/ dewe.

There are also some cross-variety differences in common West Germanic processes.
- In Middle Dutch, the vowels resulting from open syllable lengthening didn't merge with the pre-existing long vowels: high vowels lowered when lengthened (similarly to English), and inherited ē ō had diphthongized to something like /eə oə/ or /eɛ oɔ/.

I like this.

Some outlying English dialects are of utility in demonstrating the range of possibilities. For example, Forth and Bargy:
- Retention of root-final affricates: ich "I", bydge "buy", lidge "lie"
- Retention of /x/: ligt "light", nickht "night"

And this.

- Absence of the GVS: dhreeve "drive", heeve "hive"; eal "eel", lear "lere (empty)" (ea /e:/ is known from Irish loanwords); laady "lady", kaake "cake", faace "face"; agyne "again", gryne "grain"; pyle "pail"

I do question the orthography — what conditions "ee" to be pronounced [iː]?

- A GVS-like shift of ū > ɛu: greound "ground", keow "cow"
- Lowering of ɛ̄: haade "head", laafe "leaf", maate "meat" (but deed "dead", leed "lead")
- Raising of ā before /m/ and /nd/: gaume "game", llawm "lamb", lhaung "long"; hone "hand", brone "brand", lone "land", sthrone "strand"
- Absence of general ā-backing: bane "bone", dhraat "throat" (this is presumably of use in dating the split)
- Probable merger of /ɛu iu/ to /iu/: vleeu "flew" (but yowe "ewe")
- ɛ > ā before r: zaareth "served", aar "their"
- Occasional lengthening: waant "want", vaat "fat", weend "wind"
- Difficult outcomes for back vowels: zoween "sowing", knouth "knows", chourch "church", gooun "gun", jooudge "judge"; zoon "soon", woode "would", zoot "soot", woork "work"; hint "hunt", piff "puff", pit "put", zin "sun"; rub "rib", vurst "first"

However I have many garden-related obligations rn; more later

I like all these, too.

[Nortaneous]

When did this alt English start diverging from our one?

When the Norman Conquest didn't happen, since that's what everyone else does, and specifically my main motivator is that I think Justin B. Rye's what-if-the-Norman-Conquest-didn't-happen alt-English is pretty bad.

Except "what if the Norman Conquest didn't happen" is boring, not to mention difficult - why wouldn't the center of gravity in England move toward London? In order to preserve West Saxon, which is easier to work with, you probably have to keep the division between Wessex and Mercia. It seems reasonable, then, for London to fall to the Normans, and for the Kingdom of Wessex to have its own capital - probably Gloucester, which I assume would be something like /dløʃtər/.

I do question the orthography — what conditions "ee" to be pronounced [iː]?

The Forth and Bargy dialect was written phonetically in accord with the conventions of the London standard. (It would be useful to know which Old English dialect it's primarily descended from, but I don't know that yet. I've read that a lot of them were from Pembrokeshire.)

Apparently some North American and Welsh dialects preserved /iu/, but in a vacuum, I think /iu y(ː)/ > /yː y(ː)/ would be most likely, with /eu/ being left to form a new phoneme /øː/ or diphthong /øy/ > /ey/, or something like that, such that /fiu/ (few) > /fyː/ fue, but /deu/ (dew) > /døy/ > /dey/ dewe.

Those dialects preserved /iu/ as the reflex of both MidE /eu iu/. My impression is that that merger was universal (and it seems to have preceded ā-backing because it appears to be present in Forth and Bargy); are there any dialects (presumably of Scots) in which "dew" and "few"/"new" have different vowels? (edit: this is the difference between Scots vowels 14a and 14b)

(Incidentally, weend for "wind" is actually regular, and the standard varieties have irregular reversal of lengthening, presumably by influence from derivatives like "windmill" and "windy".)

[Rounin Ryuuji]

I do question the orthography — what conditions "ee" to be pronounced [iː]?

The Forth and Bargy dialect was written phonetically in accord with the conventions of the London standard. (It would be useful to know which Old English dialect it's primarily descended from, but I don't know that yet. I've read that a lot of them were from Pembrokeshire.)

Oh, I misapprehended; I had thought those were hypothetical words for your alternate English.

[Nortaneous]

It seems also possible that, in Forth and Bargy, lack of ɛ̄-lowering and ā-backing instead show an Early Middle English merger of ǣ ā > ā, paralleling the short vowels.

Anyway, here's a first attempt at a vague timeline.

1000s:
1059: An outbreak of plague in Scandinavia killed the king, Harald Sigurdsson, and his son Magnus. The kingship passed to his other son, Olaf.
1066: After an unsuccessful raid of heavily fortified Pevensey, William the Mad landed near Camber and defeated Harold Godwinson in the exceptionally bloody Battle of Rye, but failed to receive the submission of English leaders and resolved to capture London; however, he was killed in the Battle of Southwark.
1067: Robert the Conqueror was crowned king of England at Westminster Abbey. Harold Godwinson retreats to Winchester.
1068-1070: Western forces repel Robert the Conqueror's forces; Sweyn II of Denmark captures York.
1078: Sweyn II of Denmark dies; in the ensuing succession conflict, Canute IV establishes himself as king of Northumbria.
1089: Robert the Conqueror dies in a hunting accident. Henry I succeeds him as king of Mercia.

The reinstated Winchester standard has:
/p b t d tʃ (dʒ) k g/
/f θ s ʃ x/
/m n l r j w/
/ɑ e ø o i ɨ y u/
/ǣ ɑ̄ ē ø̄ ō ī ɨ̄ ȳ ū/
/æi ɑi ei/
/æw ɑw ew øw ow iw yw/

The vowel ie, of unknown quality, has backed to [ɨ]. Short æ a have merged to produce /ɑ/; eo ēo have become /ø ø̄/, and ea ēa have become /ɑ ǣ/. Homorganic lengthening has applied. /ɣ/ has become /g/ word-initially and /j/ or /w/ elsewhere. Initial /f θ s/ voice.

Before loss of /ɣ/, schwa is epenthesized into coda clusters of the form liquid + non-sibilant fricative; the resulting clusters /əf əx/ ~ /əv əɣ/ become /əw əx/.

1100s-1500s:
The great destruction wrought by William the Mad and Robert the Conqueror leads to extensive Norman settlement of Mercia and Kent; as a result, the Norman language displaces the English there. Henry I dies and is succeeded peacefully by William Adelin, who tries and fails to capture Northumbria but succeeds in pushing the Welsh Marches west. Dermot MacMurrough makes a very big mistake; consequently, the Danes invade Ireland.

Various things presumably happen in later centuries.

Short ie merges into i; ea ēa merge into æ ǣ, and eo ēo become ø ø̄. Short vowels centralize. /ɑ̄/ is diphthongized to /ɑu/ before /m nd ld lk lg/.

The usual Germanic changes of open syllable lengthening and pre-cluster shortening apply. (These are admittedly absent in some German dialects, but are AFAICT an almost pan-Germanic parallel development.) /dʒ/ is phonemicized. Vj and Vw sequences coalesce into diphthongs.

Open syllable lengthening, however, has some complications: since short vowels have centralized, the new long vowels do not merge with the short vowels. However, /ei øu ou/ merge into new /ɪ̄ ʏ̄ ʊ̄/, and /æi æu eu/ become /ei eu iu/.

/p b t d tʃ dʒ k g/
/f v θ ð s z ʃ x/
/m n l r j w/
/ɑ ɛ œ ɔ ɪ ʏ ʊ/
/ǣ ɑ̄ ē ø̄ ō ī ȳ ɨ̄ ū/
/ɛ̄ œ̄ ɔ̄ ɪ̄ ʏ̄ ʊ̄/ (new vowels from open syllable lengthening and diphthong coalescence)
/ei ɑi/
/eu ɑu iu/

This is, of course, entirely too many vowels. /ǣ ɛ̄/ merge, as do /ɔ̄ ɑ̄/. /ɪ̄ ʏ̄ ʊ̄/ develop into centralizing diphthongs /iə yə uə/. /ū/ breaks into /ou/, /ɨ̄/ lowers to /ɤ̄/, /œ̄/ breaks to /ay/, and short /u/ lowers to /o/.

/ɑ ɛ œ ɔ o i y/
/ɑ̄ ɛ̄ œ̄ ē ø̄ ɤ̄ ō ī ȳ/
/ei ɑi/
/ay/
/eu ɑu ou iu/
/iə yə uə/

/s/ backs to /ʃ/ syllable-finally after a high vowel or offglide. /h/ is lost from the cluster /hn/. The short vowels /ɛ i/ round to /œ y/ adjacent to a postalveolar or following /w/; after this, /hw/ becomes /f/. /hl hr/ are preserved.

The orthography of the time has not yet standardized, but there are some general trends:

/p b t d tʃ dʒ k g/ <p b t d ch gh~cgh~cgi~cg c~k g>
/f v θ ð s z ʃ x/ <f v þ þ s~ss z~s sch gh>
/m n l r j w/ <m n l r i~j~y ƿ~v~u>
/ɑ ɛ œ ɔ o i y/ <a e~ea eo oa o~u i~y io~yo>
/ɑ̄ ɛ̄ ē ø̄ ɤ̄ ō ī ȳ/ <aa~ae ea ee eo ie oo ij~y io~yo>
/ei ɑi/ <ei~ey ai~ay>
/ay/ <eou~auy>
/eu ɑu ou iu/ <eu au ou iu~yu>
/iə yə uə/ <ie~ia ioe~yoe ue~oe~ua>

[Otto Kretschmer]

I thought of a scenareio in which the Visigoths never go to Spain but settle in England instead.

You then end up with England being populated by Goths and possibly smaller contingents of Angles and Saxons.

Another interesting scenario is an EARLIER settlement of Britain by Germanic peoples. Let's say the Roman invasion of Britain either fails or is repelled after a rebellion leaving the land weaker and depopulated. The Germanic tribes then move in and we end up with Britain that is mostly Germanic by 150-200 AD rather than 650 AD - several more centuries of linguistic change in isolation

[anteallach]

- In Middle Dutch, the vowels resulting from open syllable lengthening didn't merge with the pre-existing long vowels: high vowels lowered when lengthened (similarly to English), and inherited ē ō had diphthongized to something like /eə oə/ or /eɛ oɔ/.

In the traditional dialects of part of northern England where the FLEECE and GOAT vowels derive from open syllable lengthening of early ME /ɛ/ and /ɔ/ you often get /ɛɪ/ and /ɔɪ/, e.g. coal and hole (and also lose, where the standard development is irregular) with /ɔɪ/, eat and meat with /ɛɪ/. The originally long vowels have different developments.

[Moose-tache]

I'm confused by the diphthongs of Winchester OE, the starting point of your project. What are all those Vw diphthongs, and whence do they come? What happened to the real diphthongs eo and ea? And what is "ie" supposed to represent? It can't be the OE diphthong spelled similarly, because that was absent from the Wessex dialect of OE.

[Nortaneous]

Not sure if it's lame to preserve front rounded vowels or not.

What are all those Vw diphthongs, and whence do they come?

from vowels followed by /w/ and back vowels followed by /ɣ/, same as they did in Middle English

What happened to the real diphthongs eo and ea?

got lost in editing, but they become ø æ, same as they did in Middle English

And what is "ie" supposed to represent? It can't be the OE diphthong spelled similarly, because that was absent from the Wessex dialect of OE.

no, it was absent from every dialect but West Saxon

[Nortaneous]

1500s-1800s:

More historical events happen; fill in later.

ɑ backs and lengthens to ɔ̄ between /w/ and a non-coronal and before /lC rC sC m n g/. The merger of /hw/ into /f/ actually occurs after this.

/iu/ merges into /ȳ/ after labials and /eu/ elsewhere.

/ð/ becomes /d/ before /r/.

There is some occasional interchange of intervocalic /v w/, especially before schwa. This results in the redevelopment of the diphthongs /iu øu yu/, which all merge to produce /øu/. Final schwa is mostly lost.

The introduction of the printing press to the Kingdom of Wessex, as well as an influx of Flemish and Norman printers, lead to some degree of orthographic regularization; notably, /y/ becomes written as <u>. The West Saxon letters thorn and wynn (eth having been displaced by thorn centuries prior) are generally replaced by <th> and the relatively new letter <w>; however, <þ ð> were to be restored in the later Romantic period. Vowel length comes to be written in accordance with the rules of Dutch.

The first stage of the Great Vowel Shift occurs - namely, the merger of ī ȳ into ei øu, and the raising of /iə/ to /ī/. This stage is reflected in the orthography; however, the later raising of /yə uə/ to /ȳ ū/ is not. (Possibly this is due to Dutch influence, however.)

/p b t d tʃ dʒ k g/ <p b t d ch gi~cgh c~k g>
/f v θ ð s z ʃ x/ <f v þ þ s~ss z~s sch gh>
/m n l r j w/ <m n l r y w>
/ɑ ɛ œ ɔ o i y/ <a e eo o i u>
/ɑ̄ ɛ̄ ɔ̄ ē ø̄ ɤ̄ ō/ <aa~ae ea oa ee eo ie oo> (the variant spellings /ɑ̄ ɛ̄ ɔ̄/ <aCe eCe oCe> are eliminated)
/ei ɑi/ <ei ai>
/ay/ <eu>
/eu øu ɑu ou/ <ew~eou uu aw~au~aou ow~ou>
/iə yə uə/ <ij ue oe>

/wr/ becomes /vr/; /kl gl/ become /tl dl/; /hl/ fortites to /kl/. None of these changes are reflected in the spelling. The cluster /hr/, written <rh>, is preserved, and reinforced by the occasional loan from Welsh. /tw dw/ become /p b/.

The remainder of the Great Vowel Shift applies:
- Short ɔ o > o u
- iə yə uə > ī ȳ ū
- Remaining ɑ backs unconditionally to ɒ; consequently, ɛ lowers, and œ lowers and unrounds, to æ
- ay > æy
- o unrounds to ɤ, including in diphthongs; øu fronts to ɤy
- ɛ̄ ɔ̄ shorten in tautomorphemic closed syllables except when followed by liquids

æ ɒ ɤ i y u
ɑ̄ ɛ̄ ɔ̄ ē ø̄ ɤ̄ ō ī ȳ ū
ei ɑi
æy øy
eu ɑu ɤu

[Nortaneous]

The Lord's Prayer (a reasonably conservative sample text) in various years:

1000: (I think I've corrected this to be proper West Saxon)
Fæder ūre
þū þē eart on heofonum
Sīe þīn nama ġehālgod
Tō becume þīn riċe
Ġeworþe þīn willa
On eorðan swā swā on heofonum
Urne ġedæġhwamlīċan hlāf syle ūs tō dæġ
And forgief ūs ūre gyltas
Swā swā wē forgiefað ūrum gyltendum.
And ne ġelæd þū ūs on costnunge
Ac alȳs ūs of yfele.

1200: (very little in the way of grammatical change yet bc lazy)
vǣdər ūrə
ðū hwyltʃ ært ɔn hœ̄vənəu
zɨ̄ ðīn nɑ̄mə jəhɑləd
ðīn rɪ̄tʃə bəkʊ̄mə
ðīn wilə jəwɔ̄rðə
ɔn œ̄rðən swɑ̄ ɔn hœ̄vənəu
urnə jədeilītʃən hlɑ̄f jɨ̄v ūs tō dei
ɑnd vərjɨ̄v ūs ūrə gyltəs
swɑ̄ wē vərjɨ̄vəθ ūrəu gyltəndəu
ɑnd nə jəlæd θū ūs nāwixt ɔn kɔ̄stnəŋgə
ɑk əlȳs ūs ɔf yvələ

Replacement of þe with hwylc, loss of various forms
-əx incl. -ɣə etc. > -ə, but after cluster shortening
-əm > -əw, although the inflection is in a state of decay already
Trisyllabic laxing probably exists but is analogicaly leveled out
Homorganic lengthening also exists, before /st mb nd ŋg ld rd rð/, and has the same effect as open syllable lengthening

1450:
ourə vɛ̄dər hwyltʃ ɑrt ɔn hayvnə
zɤ̄ ðīn nɑ̄mə jəhɑləd
ðīn riətʃə bəkuəmə
ðīn wylə jəwɑ̄rðə
ɔn ayrðən ɑlsəw ɔn hayvnə
orən jədeilītʃə hlɑ̄f jɤ̄v ouʃ tō dei
ɑnd vərjɤ̄v ouʃ ourə gyltəs
swɑ̄ wē vərjɤ̄vəθ ourəu gyltəndə
ɑnd jəlɑd θou ouʃ nɑ̄xt ɔn kɑ̄stnəŋgə
ɑk əlȳʃ ouʃ ɔf yvlə

Oure Veder huylch art on heowne
zie þijn naame jehalled
þijn ryeche bekoeme
þijn wylle jewaarþe
on eourþen allsew on heowne
Oren jedeilijche lhaaf jieu ousch to dey
and veryieu ousch ourew guyltende
and yeladd þou ousch naaght on kaastninge
ack alyos ousch of jowle

swɑ̄ replaced in some circumstances with æl-swɑ̄, which is then reduced to æl-səw
dative lost except in pronouns
ornə > orən
hlɑ̄f is not replaced
æ > ɑ, which i forgot to mention

1650: (Early Modern)

There should be some lexical churn here but, again, lazy, I'll get to it

ɤur vɛ̄dər fyl ɔ̄rt ō hayvə
zɤ̄ ðein nɑ̄m jəhɑləd
ðein rītʃ bəkuəm
ðein wyl jəwɑ̄rd
ōn ayrd ɔ̄ls ō hayvə
jɤ̄u ɤuʃ ðə dei ɤur jədeilītʃ klɑ̄f
ɔn vərjɤ̄u ɤuʃ ɤur gyltəs
swɑ̄ wē vərjɤ̄uθ ɤurəu gyltəndəs
ɔn jəlɑd θou ouʃ nɑ̄x ō kɑ̄stnəŋgə
ɑk əløyʃ ouʃ of ɤuəl

Our Veader ful oart oo Heuve
zie theyn Naam yhalled
theyn Rijch bekoem
theyn Wul ywaard
oon Eurd oals oo Heuve
you ousch thedey our ydeylijch Lhaaf
on veryou ousch our Gultes
swa we veryouth ourow Gultendes
on ylad thou ousch naagh oo Caastninge
ack alws ousch of Ouel

final /n/ is lost with compensatory lengthening somewhere
ch is lost in grammatical words except it's the light prefix of the first person, more later
many mistakes due to circumstances

[Otto Kretschmer]

It reminds me of West Frisian
.. why?