Middle English phonology

Middle English phonology is necessarily somewhat speculative, since it is preserved only as a written language. Nevertheless, there is a very large text corpus of Middle English. The dialects of Middle English vary greatly over both time and place, and in contrast with Old English and Modern English, spelling was usually phonetic rather than conventional. Words were generally spelled according to how they sounded to the person writing a text, rather than according to a formalised system that might not accurately represent the way the writer's dialect was pronounced, as Modern English is today.

The Middle English speech of the city of London in the late 14th century (essentially, the speech of Geoffrey Chaucer) is used as the standard Middle English dialect in teaching and when specifying "the" grammar or phonology of Middle English. It is this form that is described below, unless otherwise indicated.

In the rest of the article, abbreviations are used as follows:

Sound inventory
The surface sounds of Chaucerian Middle English (whether allophones or phonemes) are shown in the tables below. Phonemes in bold were added across Middle English, phoneme in italics were removed during the same.

Consonants
1. The exact nature of Middle English r is unknown. This article uses $⟨r⟩$ indiscriminately.

Consonant allophones
The sounds marked in parentheses in the table above are allophones:
 * is an allophone of occurring before  and
 * For example, ring ('ring') is ; did not occur alone in Middle English, unlike in Modern English.
 * are allophones of in syllable-final position after front and back vowels, respectively.
 * Based on evidence from Old English and Modern English, and  apparently had velarised counterparts or allophones  and . These occurred after back vowels or the consonant.

Voiced fricatives
In Old English,, , were allophones of , , , respectively, occurring between vowels or voiced consonants. That led to many alternations: hūs ('house') vs. hūses ('of a house') ; wīf ('woman')  vs. wīfes ('of a woman'). In Middle English, voiced allophones become phonemes, and they are solidly established in Modern English as separate phonemes by several sources: The status of the sources in Chaucer's Middle English is as follows: The strongest distinction was between and  because of the large number of borrowings from Old French. It is also the only distinction that is consistently indicated in spelling, as $⟨f⟩$ and $⟨v⟩$ respectively. sometimes appears as $⟨z⟩$, especially in borrowings from Greek and sometimes as $⟨s⟩$. Both and  are spelled $⟨th⟩$.
 * 1) Borrowings from foreign languages, especially Latin, Ancient Greek, and Old French, which introduced sounds where they had not occurred: modern fine vs. vine (both borrowings from French); ether (from Greek) vs. either (native).
 * 2) Dialect mixture between Old English dialects (like Kentish) that voiced initial fricatives and the more standard dialects that did not. Compare fat vs. vat (both with f- in standard Old English) and fox vs. vixen (Old English fox vs. fyxen, from Proto-Germanic *fuhsa- vs. *fuhsin-).
 * 3) Analogical changes that levelled former alternations: grass, grasses, grassy and glass, glasses, glassy with replacing the original  between vowels (but to graze and to glaze, still with, originally derived from grass and glass, respectively).  Contrast wife vs. wives; greasy, still with a  in some dialects (such as that of Boston) and staff, with two plurals, analogical staffs and inherited staves.
 * 4) Loss of final, resulting in voiced fricatives at the end of a word where only voiceless fricatives had occurred. That is the source of the modern distinctions house vs. to house, teeth vs. to teethe, half vs. to halve.
 * 5) Reduction of double consonants to single consonants. That explains the contrast between kiss, to kiss (Old English coss, cyssan, with a double s) vs. house, to house with in the verb (Old English hūs, hūsian, with a single s).
 * 6) A Sandhi effect that introduced the voiced fricative /ð/ instead of original /θ/ at the beginning of unstressed function words. Contrast this with initial  vs. thistle with initial.
 * 7) A sound change that caused fricatives to be voiced when preceded by a fully unstressed syllable. This change is reflected in the modern pronunciation of the endings spelt -s (the noun plural ending, the 'Saxon genitive' ending and the ending for 3rd person present indicative), which now have phonemic shape -, having developed in Middle English from - to - and then, after the deletion of the unstressed vowel, to - (e.g. halls, tells with from earlier halles, telles). The sound change also affects function words ending in original - that are normally unstressed. Contrast this with  vs. is with ; off with  vs. of with, originally the same word; with with  in many varieties of English vs. pith with.
 * The first three sources (borrowing, dialect mixture, analogy) were already established.
 * As indicated by versification, the loss of final was normal in Chaucer's time before a vowel-initial word and optional elsewhere; it is assumed that it is a poetic relic and that the loss of final  was already complete in spoken English (a similar situation to Modern French; see e muet).
 * The reduction of double consonants was apparently about to occur.
 * The sandhi effects on unstressed function words occurred somewhat later, in the transition to Modern English.

Vowels
1 The Old English sequences, produced late Middle English , apparently after passing through early Middle English : OE grōwan ('grow') → LME. However, early Middle English produced by Middle English breaking became late Middle English : OE tōh (tough') → EME  → LME. Apparently, early became  before the occurrence of Middle English breaking, which generated new occurrences of, which later became.

Monophthongs

Middle English had a distinction between close-mid and open-mid long vowels but no corresponding distinction in short vowels. Although the behavior of open syllable lengthening seems to indicate that the short vowels were open-mid in quality, according to Lass, they were close-mid. (There is some direct documentary evidence: in early texts, open-mid was spelled $⟨ea⟩$, but both  and  were spelled $⟨eo⟩$.) Later, the short vowels were in fact lowered to become open-mid vowels, as is shown by their values in Modern English.

The front rounded vowels existed in the southwest dialects of Middle English, which developed from the standard Late West Saxon dialect of Old English, but not in the standard Middle English dialect of London. The close vowels and  are direct descendants of the corresponding Old English vowels and were indicated as $⟨u⟩$. (In the standard dialect of Middle English, the sounds became and ; in Kentish, they became  and .)  may have existed in learned speech in loanwords from Old French, also spelled $⟨u⟩$, but, as it merged with, becoming  in Modern English, rather than , it can be assumed that  was the vernacular pronunciation that was used in French-derived words.

The mid-front rounded vowels likewise had existed in the southwest dialects but not in the standard Middle English dialect of London. They were indicated as $⟨o⟩$. Sometime in the 13th century, they became unrounded and merged with the normal front mid vowels. They derived from the Old English diphthongs and. There is no direct evidence that there was ever a distinction between open-mid and close-mid, but it can be assumed because of the corresponding distinction in the unrounded mid front vowels. would have derived directly from Old English, and derived from the open syllable lengthening of short , from the Old English short diphthong.

The quality of the short open vowel is unclear. Early in Middle English, it presumably was central since it represented the coalescence of the Old English vowels  and, and at the time of Middle English breaking, it could not have been a front vowel since  rather than  was introduced after it. During the Early Modern English period, it was fronted, in most environments, to in southern England, and it and even closer values are found in the contemporary speech of southern England, North America and the southern hemisphere: it remains  in much of Northern England, Scotland and the Caribbean. Meanwhile, the long open vowel, which developed later because of open syllable lengthening, was. It was gradually fronted, to successively, and , in the 16th and the 17th centuries.

Diphthongs
All of the above diphthongs came about within the Middle English era. Old English had a number of diphthongs, but all of them had been reduced to monophthongs in the transition to Middle English. Middle English diphthongs came about by various processes and at various time periods. Diphthongs tended to change their quality over time. The changes above occurred mostly between early and late Middle English. Early Middle English had a distinction between open-mid and close-mid diphthongs, and all of the close-mid diphthongs had been eliminated by late Middle English.

The following processes produced the above diphthongs:
 * Reinterpretation of Old English sequences of a vowel followed by Old English (which became  after back vowels, and  after front vowels), or with pre-existing, :
 * OE weġ ('way') → EME
 * OE dæġ ('day') → ME  → LME
 * Middle English breaking before ( after back vowels,  after front vowels)
 * Borrowing, especially from Old French

Phonological processes
The following sections describe the major phonological processes occurring between written Late West Saxon (the standard written form of Old English) and the end of Middle English, conventionally dated to around 1500 AD.

Homorganic lengthening
Late in Old English, vowels were lengthened before certain clusters:, , , ,. Later on, the vowels in many of these words were shortened again, giving the appearance that no lengthening happened; but evidence from the Ormulum indicates otherwise. For details see Phonological history of Old English: Vowel lengthening.

Stressed vowel changes
Late West Saxon (the standard written form of Old English) included matched pairs of short and long vowels, including seven pairs of pure vowels (monophthongs),, and two pairs of height-harmonic diphthongs, and. Two additional pairs of diphthongs, and, existed in earlier Old English but had been reduced to  and , respectively, by late Old English times.

In the transition to Middle English, this system underwent major changes, eliminating the diphthongs and leaving only one pair of low vowels, but with a vowel distinction appearing in the long mid vowels:
 * The diphthongs simplified to  and, respectively.  Subsequently, the low vowels were modified as follows:
 * and merged to a single central vowel.
 * and rose to  and, respectively.
 * The diphthongs and  (as in OE frēond 'friend') respectively simplified to new front-round vowels  and  (yielding /frøːnd/ 'friend').  Everywhere except in the southwest,  and   were soon respectively unrounded to  and  (yielding Middle English freend /freːnd/ 'friend'); in the southwest, it took 200 or 300 years for this process to take place, and in the meantime the sounds were spelled $⟨o⟩$ in texts from the southwest.
 * The front-rounded vowels and  unrounded to  and, respectively, everywhere but in the southwest (former West Saxon area) and southeast (former Kentish area).
 * In the southwest, the front-rounded vowels and  remained, and were spelled $⟨u⟩$.
 * In the southeast, the vowels had already been unrounded to and, respectively, in Old English times, and remained as such in Middle English.

This left an asymmetric system consisting of five short vowels and six long vowels , with additional front-rounded vowels  in the southwest area. Some symmetry was restored by open syllable lengthening, which restored a long low vowel.

Reduction and loss of unstressed vowels
Unstressed vowels were gradually confused in late Old English, although the spelling lagged behind, due to the existence of a standardized spelling system. By early Middle English, all unstressed vowels were spelt $⟨e⟩$, probably representing. Also in late Old English, final unstressed became ; during the Middle English period, this final  was dropped when it was part of an inflectional syllable (but remained when part of the root, e.g. seven, or in derivational endings, e.g. written). Around Chaucer's time, final was dropped; judging from inflectional evidence, this occurred first when the following word began with a vowel. A century or so later, unstressed also dropped in the plural and genitive ending -es (spelled -s in Modern English) and the past ending -ed.

These changes steadily effaced most inflectional endings, e.g.: In the last two examples, the stressed vowel was affected by open-syllable lengthening.
 * OE mētan → ME meete(n) → LME → NE meet
 * OE wicu → ME weeke → LME → NE week
 * OE nama → ME nāme → LME → NE name

Vocalization of and development of new diphthongs
The sound, which had been a post-vocalic allophone of , became vocalized to. This occurred around the year 1200.

A new set of diphthongs developed from combinations of vowel+ (either from or from pre-existing ) or vowel+ (from pre-existing ), and also due to borrowing from French – see Diphthongs above.

Breaking
During the 12th or 13th century, a vowel was inserted between a front vowel and a following  (pronounced  in this context), and a vowel  was inserted between a back vowel and a following  (pronounced  in this context). Short was treated as a back vowel in this process (the long equivalent did not occur in the relevant context). See H-loss, below.

Open-syllable lengthening
Around the 13th century, short vowels were lengthened in an open syllable (i.e. when followed by a single consonant that in turn is followed by another vowel). In addition, non-low vowels were lowered: →,  → ,  → ,  →. This accounts, for example, for the vowel difference between staff and the alternative plural staves (Middle English staf vs. stāves, with open-syllable lengthening in the latter word). This process was restricted in the following ways:
 * 1) It did not occur when two or more syllables followed, due to the opposing process of trisyllabic laxing.
 * 2) It only occasionally applied to the high vowels and, e.g. OE wudu → ME  → wood; OE wicu → ME  → week.  Most instances of  and  remained as such, e.g. OE hnutu → NE nut, OE riden → NE ridden.

The effects of open-syllable lengthening and trisyllabic laxing often led to differences in the stem vowel between singular and plural/genitive. Generally these differences were regularized by analogy in one direction or another, but not in a consistent way:
 * ME path, pāthes → NE path, paths, but ME whal, whāles → NE whale, whales
 * ME crādel, cradeles → NE cradle, cradles, but ME sādel, sadeles → NE saddle, saddles

Trisyllabic laxing
In late Old English, vowels were shortened before clusters of two consonants when two or more syllables followed. Later in Middle English this process was expanded, and applied to all vowels when two or more syllables followed. This led to the Modern English variations between divine vs. divinity, school vs. scholarly, grateful vs. gratitude, etc. In some cases, later changes have led to apparently anomalous results, e.g. south vs. southern with only two syllables (but at the time that trisyllabic laxing applied). This change is still fairly productive in Modern English.

Pre-cluster shortening
In late Old English, vowels were shortened before clusters of three consonants:
 * OE gāst → NE ghost ; OE gāstliċ → NE ghastly
 * OE ċild → NE child ; OE ċildru + OE -an → NE children
 * OE gōd → NE good; OE gōdspell → NE gospel

As shown by ghastly, this shortening occurred before the raising of OE to EME, which occurred in the transition to Middle English.

Later in Middle English, vowels were shortened before clusters of two consonants, except before and in some cases where homorganic lengthening applied. Examples:
 * OE cēpte → kept (cf. OE cēpan → keep)
 * OE mētte → met (cf. OE mētan → meet)

Reduction of double consonants
Double (geminated) consonants were reduced to single ones. This took place after open syllable lengthening; the syllable before a geminate was a closed syllable, hence vowels were not lengthened before (originally) doubled consonants. The loss of gemination may have been stimulated by its small functional load&mdash;by this time there were few minimal pairs of words distinguished solely by the single vs. double consonant contrast.

H-loss
The phoneme, when it occurred in the syllable coda, is believed to have had two allophones: the voiceless palatal fricative , occurring after front vowels, and the voiceless velar fricative , occurring after back vowels. The usual spelling in both cases was $⟨gh⟩$, which is retained today in words like night and taught.

These sounds were lost during the later Middle English and Early Modern English eras. The timing of this process was dependent on dialect; the fricatives were still pronounced in some educated speech in the 16th century, but they had disappeared by the late 17th. Loss of the fricatives was accompanied by some compensatory lengthening or diphthongization of preceding vowels. In some cases, the velar fricative developed into ; as such the preceding vowel was shortened, and the  of a diphthong was absorbed. Some developments are illustrated below:
 * OE niht ('night') → ME  →  → NE  (by the Great Vowel Shift)
 * OE hlæhhan ('to laugh') → ME → LLME  → ENE  → NE
 * OE tōh ('tough') → ME → LLME  → NE

This variable outcome, along with other variable changes and the ambiguity of the Middle English spelling $⟨ou⟩$ (either or  in Early Middle English) accounts for the numerous pronunciations of Modern English words in -ough- (e.g. though, through, bough, rough, trough, thought, with -ough- pronounced  respectively).

spelled -gh- is realized as even today in some traditional dialects of northern England and more famously Scots. Some accents of northern England that lack the, instead exhibiting special vowel developments in some such words; for example, night as (sounds like neat) and in the dialectal words owt and nowt (from aught and naught, pronounced like out and nout, meaning 'anything' and 'nothing').

The modern phoneme most commonly appears today in the typically Scottish word loch and in names such as Buchan. Here the is usual in Scotland, although the alternative  is becoming more common among some younger speakers. The same is true in Wales, in names such as Loughor. English speakers from elsewhere may replace the in such cases with, but some use  in imitation of the local pronunciations (as they may in certain foreign words such as Bach, Kharkiv, Sakhalin, chutzpah, etc.).

Great Vowel Shift
The Great Vowel Shift was a fundamental change in late Middle English (post-Chaucer) and Early Modern English that affected the pronunciation of all of the long vowels. The high vowels and  were diphthongized, ultimately producing the modern diphthongs  and, and all other vowels were raised.

Diphthong loss
Although not normally considered a part of the Great Vowel Shift, during the same time period most of the pre-existing Middle English diphthongs were monophthongized:
 * → ENE →  → NE
 * → ENE
 * → ENE → NE

The remaining diphthongs developed as follows:
 * , → ENE  → NE .   is still used in Welsh English.
 * , → NE

Vowel equivalents from Old English to Modern English
For a detailed description of the changes between Old English and Middle/Modern English, see the article on the phonological history of English. A summary of the main vowel changes is presented below. The spelling of Modern English largely reflects Middle English pronunciation.

Monophthongs
This table presents the general developments. Many exceptional outcomes occurred in particular environments: vowels were often lengthened in late Old English before, , ; vowels changed in complex ways before , throughout the history of English etc. Vowels were diphthongized in Middle English before , and new diphthongs arose in Middle English by the combination of vowels with Old English ,  → , and  ; for more information, see the section below. The only conditional development considered in detail below is Middle English open-syllable lengthening. In the column on modern spelling, CV means a sequence of a single consonant followed by a vowel.

The Modern English vowel usually spelled (Received Pronunciation:, General American: ) does not appear in the above chart. Its main source is late Middle English < early  and, which come from various sources: Old English  and  (claw < clawu, law < lagu); diphthongization before  (sought < sōhte, taught < tāhte, daughter < dohtor); borrowings from Latin and French (fawn < Old French faune, Paul < Latin Paulus). Other sources are Early Modern English lengthening of before  (salt, all); occasional shortening and later re-lengthening of Middle English  (broad <  < brād); and in American English, lengthening of short o before unvoiced fricatives and voiced velars (dog, long, off, cross, moth, all with  in American dialects that still maintain the difference between  and ).

As mentioned above, Modern English is derived from the Middle English of London, which is derived largely from Anglian Old English, with some admixture of West Saxon and Kentish. One of the most noticeable differences among the dialects is the handling of original Old English. By the time of the written Old English documents, the Old English of Kent had already unrounded to, and the late Old English of Anglia unrounded  to. In the West Saxon area, remained as such well into Middle English times and was written  in Middle English documents from the area. Some words with the sound were borrowed into London Middle English, where the unfamiliar was substituted with :
 * gild < gyldan, did < dyde, sin < synn, mind < mynd, dizzy < dysiġ, lift < lyft etc. show the normal (Anglian) development;
 * much < myċel shows the West Saxon development;
 * merry < myriġ shows the Kentish development;
 * bury < byrġan has its spelling from West Saxon but its pronunciation from Kentish;
 * busy < bysiġ, build < byldan, buy < bycġan have their spelling from West Saxon but their pronunciation from Anglian.

Some apparent instances of modern for Old English  are actually regular developments, particularly if the  is a development of earlier (West Saxon)  from i-mutation of, as the normal i-mutation of  in Anglian is ; for example, stern < styrne < *starnijaz, steel < stȳle < *stahliją (cf. Old Saxon stehli). Also, some apparent instances of modern for Old English  may actually be from the influence of a related form with unmutated : sundry < syndriġ, influenced by sundor "apart, differently" (compare to sunder and asunder).