The Anglish Alphabet

For a long time the influence which the Norman Invasion and its aftermath had on English spelling has been ignored in the Anglish project. This proposal is meant to address that. While we will never truly know how English spelling would have turned out had the Norman Invasion never happened, we can see what English spelling was like before French influence and revert to those conventions, and make educated guesses.

This project is the work of multiple people from the Anglish Discord. Main contributors are: Hurlebatte; Yose; Andwlite; Eadwine. Beware: the contents of this article based on new evidence.

REVERSIONS
Some reversions are more random, and do not fit neatly in the chart above. These include: ache to ake; scythe to sithe; island to iland; accursed to acursed; allay to alay; afford to aford; affright to afright; anneal to aneal; tongue to tung; Rhine to Rine; rhyme to rime; ghost to goast; sailor to sailer; neighbour to neighbor; harbour to harbor; guest to gest; guess to gess.


 * "The form anneal, however, derives from OE anælan; spellings of this word with NN are first attested in the 17th century, by analogy with Latinate forms such as annex (compare similar doubling of C, F, L in accursed, afford, allay)." - The History of English Spelling, Upward & Davidson

COMMENTARY
REVERSIONS

⟨c⟩ being soft (standing for /s/) is a French thing. We have found no trace of it before the Norman Invasion.

⟨ch⟩ for /tʃ/ came into English with the Norman Invasion. It appears early in English in the Ormulum, a manuscript from the 1100s. Before ⟨ch⟩ was borrowed, English developed two ways to cut down on ambiguity. First, some scribes began turning to ⟨k⟩, a letter which Anglo-Saxon scribes had previously left mostly unused. Second, some scribes would insert a silent ⟨e⟩ or a silent ⟨i⟩ after ⟨c⟩ to "trigger" its /tʃ/ value, similar to how the silent ⟨u⟩ in modern guest "triggers" the /g/ value of ⟨g⟩. Since many words which are now ambiguous if we remove ⟨ch⟩ once began with ⟨ce⟩, we believe English would have retained ⟨ce⟩ for the sake of clarifying the value of ⟨c⟩. In other words, we believe choke would be spelled ceoke today had ⟨ch⟩ never been borrowed. This is similar to how English keeps ⟨-e⟩ and ⟨gh⟩ to indicate vowel length despite those particular letters no longer standing for spoken phonemes.


 * "Anglo-Saxon scribes sometimes clarified the pronunciation of c using two devices. . . one device was to show the /tʃ/ value where it might otherwise not be apparent by inserting an E or I after the C; thus þencan could also be written þencean. The other device was to replace C by K to show its /k/ value before a front vowel: cyn 'kin', cyning 'king', cycen 'kitchen' were sometimes written kin (so contrasting with palatalized c in cin 'cin'), kyning, kicen. Likewise the genitive case of folk 'people' could be either folces or folkes 'of the people'." - The History of English Spelling, Upward & Davidson

⟨ie⟩ standing for /i/ seems to have rubbed off onto English from loanwords like piece.


 * "After the Conquest, French scribes introduced some new spellings. . . IE was used to represent /e:/. . ." - The History of English Spelling, Upward & Davidson

⟨le⟩ at the end of words seems to have rubbed off onto English from French loanwords like people.


 * "The development of people offers a good paradigm for many words ending in -LE: Lat populum > OFr poeple > ME peple > people. The final syllable of people and similar words was commonly spelt in ME with a wide variety of vowel letters, as -EL, -IL, -UL, -YL, etc. In EModE, printers showed a growing tendency to prefer the Fr -LE spelling in many words of both Franco-Lat and OE descent." - The History of English Spelling, Upward & Davidson

⟨o⟩ taking the place of ⟨u⟩ in many instances, such as sum to some, seems to be post-invasion influence.


 * "The convention of using o for earlier u begins in late Latin and is extended first to French and then to English." - Venezky, Richard L. Visible Language; Detroit, Michigan etc. Volume 10, Issue 4, pages 351-365


 * "In the handwriting of the ME period much more than in that of the OE, the letters i (and j), u (and v), n, m, and w tended to be made simply by one, two, or three short upright strokes (technically called minims) without horizontal connecting strokes at the top of bottom between minims forming parts of the same letter, and sometimes without a dot over the single minim standing for i (or j). The result was that any word containing two or more of these letters in sequence became difficult to read, a succession of, say, four minims being interpretable as nu, un, mi, wi ,im ,iw, ini, iui (ivi), nii, uii (vii), iin, or iiu (iiv). In some contemporary French dialects, o had come, in certain phonetic situations, to indicate the same sound as u; French scribes were not slow to substitute o very generally for u whenever u was etymologically called for in the neighborhood of other letters made up of minims. This practice came to be widely imitated in writing English, and hence ME sone, which was easier to read than sune. . ." - Early English: An Introduction to Old and Middle English, Clark, page 122

⟨ou⟩ was borrowed from French. Beforehand, English represented /uː/ with ⟨u⟩, and often it was marked as long with a "magic-E". Due to the Great Vowel Shift, many instances of /uː/ became /aʊ/, but in some cases the old value was preserved. These conservative instances ended up either with the English ⟨oo⟩ spelling (room), or took on the French ⟨ou⟩ spelling (bouk). For Anglish spelling, these instances of ⟨ou⟩ standing for /uː/ should only be spelled as ⟨oo⟩. In some cases ⟨ough⟩ has a French based ⟨ou⟩, but in some cases the ⟨ou⟩ stands for a naturally occurring English diphthong which arose through vowel breaking. As a general rule, words which rhyme with dough have the "English ⟨ou⟩", while words rhyming with plough and tough have the "French ⟨ou⟩".

⟨q⟩ was borrowed into regular usage under influence from French. It seems to become common in the 1200s.

⟨sh⟩ seems to be based on ⟨ch⟩ one way or another. It may be an alternative to ⟨sc⟩ which was modelled on ⟨ch⟩, or it may be a shortening of ⟨sch⟩ which was apparently modelled on ⟨ch⟩.

⟨u⟩, ⟨ue⟩, and ⟨u_e⟩ standing for /ju/ developed from French influence. English borrowed French loanwords which contained /y/, and writers chose to represent this sound with ⟨u⟩, ⟨ue⟩, and ⟨u_e⟩. Previously these spellings stood for English's native /uː/, but that phoneme began to be spelled in the French manner with ⟨ou⟩, leaving ⟨u⟩, ⟨ue⟩, and ⟨u_e⟩ open to being repurposed. Making matters more complicated, this French /y/ merged with English's native /iu/, resulting in words with /iu/ occasionally taking on spellings with ⟨u⟩, ⟨ue⟩, and ⟨u_e⟩; we write hue not hew because of this. Note that Yule should revert to yool, not become yewl.

⟨u⟩ taking the place of what was once Old English ⟨y⟩ seems to be linked to French. Although /y/ had merged with /e/ or /i/ for many English speakers by 1066, speakers in the Southwest and the West Midlands apparently retained /y/. This sound was then rendered as ⟨u⟩ in the French manner, leading to spellings like burden existing alongside spellings like birden and berden in Middle English.


 * "Bereits in ae. Zeit (um 900) war in Kent und Surrey, Essex und Suffolk y, ý (Lautwert: y, y:) zu e, e: geworden. Es blieb y, y: im Südwesten und im westlichen Mittelland, wo es nach frz. Gewohnheit u geschrieben wurde. Im östlichen Mittelland und im ganzen Norden wurde es hingegen zu i, i: entrundet. Aus der Lage Londons erklärte es sich, daß sich bei Chaucer alle drei Entwicklungen belegen lassen." - Bähr, Dieter (1997). Einführung ins Mittelenglische. UTB, Stuttgart

⟨z⟩ was rarely used in Old English, and virtually never in native words. The letter began to appear in native words after English had taken in many French loanwords which contained it.

REJECTED IDEAS

⟨ƿ⟩ was not revived because we have not linked its disappearance to French influence. English has been using ⟨uu⟩ and ⟨w⟩ since Old English times, and ⟨ƿ⟩ persisted fairly long after the Norman Invasion. It is possible that ⟨ƿ⟩ simply looked too much like ⟨p⟩ (and blackletter ⟨þ⟩ as well) to forever withstand being replaced by clearer alternatives.

⟨ð⟩ was not revived because it seems to have died a natural death, losing out to ⟨þ⟩.

⟨þ⟩ was not revived because ⟨th⟩, a spelling convention used in Old English since the earliest manuscripts, was rapidly rising in popularity in the decades before the introduction of printing presses to England. The switch to ⟨th⟩ may have been influenced by the widespread use of French-linked ⟨ch⟩ and ⟨sh⟩, but we cannot demonstrate that.

⟨æ⟩ was not revived because it seems to have died a natural death. The short value represented by ⟨æ⟩ merged with the short vowel represented by ⟨a⟩, and took on ⟨a⟩ as its spelling. The long value represented by ⟨æ⟩ merged with ⟨ea⟩ and took on ⟨ea⟩ as its spelling. It is unlikely French had anything to do with this. As an aside, for a time some scribes chose ⟨æ⟩ rather than ⟨ea⟩ for this merged phoneme.

⟨ȝ⟩ being a new letter named Yough and lacking the ability to stand for /g/ is an outcome of the Norman Invasion. That, however, does not mean you cannot go back to drawing ⟨g⟩ in an insular manner, nor does it mean you cannot select insular style fonts while on the computer.

⟨cg⟩ is not a part of this system because nobody has demonstrated that the shift from ⟨cg⟩ to ⟨gg⟩ to ⟨dg⟩ had anything to do with French.

⟨hw⟩ is not a part of this system because it cannot be definitely linked to the Norman Invasion. Some claim that ⟨hw⟩ changed to ⟨wh⟩ to match ⟨ch⟩, ⟨sch/sh⟩, and ⟨th⟩, but it is possible that the loss of ⟨hr⟩ and ⟨hl⟩ put pressure on ⟨hw⟩ to conform instead to ⟨wr⟩ and ⟨wl⟩, two spellings that were around when the switch happened. Note that ⟨wh⟩ appears in the Early Middle English manuscript Laud Misc. 636 which lacks ⟨ch⟩ and ⟨sch/sh⟩, casting some doubt on the claim that ⟨wh⟩ is merely the result of conformity to French digraphs.


 * "The wh- spelling already appears, though infrequently, in LOE as a variant of OE hw-. In ME it is used sporadically during the 12th cent., e.g., in Peterb.Chron., a gloss, and a few names; by the end of the century it is the regular spelling in Orm., from the NEM. The spelling becomes more frequent in the 13th cent., and widespread in the 14th cent." - Middle English Compendium

⟨g⟩ does not make /j/ in this system because ⟨y⟩ seems to have taken on that job naturally. What seems to have happened is /y/ merged with /i/ in Old English, leading to ⟨y⟩ and ⟨i⟩ being interchangeable in Middle English. From here it was a small step to have ⟨y⟩ handle /j/. Some have proposed English picked up the use of ⟨y⟩ for /j/ from French, but a general lack of French loanwords containing ⟨y⟩ for /j/ makes this appear doubtful.

⟨v⟩ is not removed from in this system because ⟨u⟩ standing for /v/ was used in English before the Norman Invasion. If you choose to replace ⟨v⟩ with ⟨f⟩, we recommend dropping pointless instances of ⟨e⟩. For example, the verb live could become lif because the ⟨e⟩ does not mark the ⟨i⟩ as long, nor does it mark the ⟨f⟩ as voiced. Additionally, some words like love and dove should probably be converted to luf and duf. This is because the switch from ⟨u⟩ to ⟨o⟩ in spelling was likely driven by a desire to visually distinguish ⟨u⟩ from ⟨v⟩.

⟨gh⟩ is retained, here is our reasoning. To start with, insular ⟨g⟩ often stood for [ɣ] in Old English. [x] was usually written with ⟨h⟩ and sometimes with ⟨ch⟩. In Middle English the distinction between [x] and [ɣ] seems to have fully collapsed, and insular ⟨g⟩ became the standard spelling for both, although some scribes went with ⟨h⟩ and others with ⟨ch⟩. Eventually ⟨gh⟩ overcame ⟨g⟩, and although some link this to French influence, we have not yet ruled out alternative possibilities, such as ⟨gh⟩ being modelled on ⟨ch⟩. It goes without saying that one may write ⟨gh⟩ in an insular style, giving it a non-Carolingian appearance like ⟨ȝh⟩.

SUNDRY
ATTESTATIONS

Use of ⟨u_e⟩ before ⟨ou⟩ was borrowed:
 * c1175(?OE) HRood (Bod 343)4/16 : Sonæ swa he ðes wateres swetnysse ifelde, þa wearð he swiðe bliðe on his mode & mid lude stefne to ðare ferde clypian ongan.
 * a1200 Trin.Hom.(Trin-C B.14.52)89 : Ðo þe after him comen remden lude stefne, þus queðinde.
 * c1275(?c1250) Owl & N.(Clg A.9)314 : Þat nis noȝt soþ ich singe efne Mid fulle dreme & lude stefne.
 * a1225(c1200) Vices & V.(1) (Stw 34)43/21 : Ðe gastliche hierdes..sculen..stieren ðo ðe bieð in ðare woreld, and ec..ðe bieð ute.
 * a1275(?c1150) Prov.Alf.(Trin-C B.14.39)122/487 : He þat is ute bi-loken, he is inne sone forȝeten.

Use of ⟨u⟩ for /v/ before 1066:
 * London, British Library, Cotton Charters viii. 16
 * London, British Library, Cotton Tiberius A. XIII, ff. 1-118, f. 116v (s. xi) ?
 * London, British Library, Add. Ch. 19796 ?

Use of ⟨u⟩ for /v/ in the 1100s:


 * a1131 Peterb.Chron.(LdMisc 636)an.1128 : God ælmihtig haue his milce ofer þæt wrecce stede.
 * a1131 Peterb.Chron.(LdMisc 636)an.1128 : . . . Dauid. ..

Use of "E-I insertion"?:


 * c1175(?OE) Bod.Hom.(Bod 343)88/20 : Ic underȝite þæt ic wulle underȝyten & þencean, & ic wulle þæt ic underȝite & munen.
 * Cd. Th. 26, 18; Gen. 408. : Onginnaþ ymb ða fyrde þencean

⟨ea⟩ becoming ⟨æ⟩ rather than ⟨æ⟩ becoming ⟨ea⟩?:


 * c1275(?a1200) Lay.Brut (Clg A.9)5692 : Ofte heo letten grundhat læd [Otho: leod] gliden heom an heore hæfd.

CREDITS


 * Credit goes to Frith for pointing out how magic-E on ⟨u⟩ used to make the native long-U sound.
 * Credit goes to ShrekBeeBensonDCLXVI on the Anglish Reddit for sharing the idea that ⟨tch⟩ was invented to distinguish the new French pronunciation of ⟨ch⟩ from the old one.
 * Credit goes to Henry Bane of Calques of the Anglish Discord for pointing out that the loss of ⟨hr⟩ and ⟨hl⟩ could have influenced the switch from ⟨hw⟩ to ⟨wh⟩.
 * Credit goes to Eadwine of the Old English Discord for pointing out that replacing ⟨ie⟩ with ⟨ea⟩ rather than ⟨ee⟩ is not supported with evidence as far as we can tell.
 * Credit goes to Yose of the Anglish Discord for pointing out that ⟨th⟩ was rising in popularity in the decades before the printing press, and for pointing out that Iceland was able to procure printing blocks for ⟨þ⟩ (implying that Englishmen could have as well if they wanted to).
 * Credit goes to the YouTube channel "Middle-English Manuscripts" for helping me find the Anglo-Saxon charters which contain ⟨u⟩ making /v/.
 * Credit goes to Andwlite of the Anglish Discord for finding sources on the ⟨o⟩ to ⟨u⟩ switch, and for alerting me to the ⟨y⟩ to ⟨u⟩ thing.