Move back one step  Move forward one step  Display the start-up screen
Larger font Smaller font

Middle English

The dialects of Middle English
Middle English phonology
Changes in grammar
Sound changes
Introductions to this period

  For literature on Middle English please consult the relevant section of the Reference Guide

After the invasion of England by the Normans in 1066, the West Saxon ‘standard’, which was waning anyway due to natural language change, was dealt a death blow. Norman French became the language of the English court and clergy.

Section of the Bayeux tapistry (called after a town in Normandy where it is kept). The tapistry depicts scene from the Norman conquest of Britain including the death of King Harold.

English continued after the Norman invasion as a patois (an unwritten dialect). With the loss of England for the French in 1204 English gradually emerged as a literary language again. For the development of the later standard it is important to note (1) that it was London which was now the centre of the country and (2) that printing was introduced into England in the late 15th century. William Caxton (c. 1442-1491) was the first to introduce printing to England in 1476. He also wrote introductions to editions of works he printed.


This latter fact contributed more than any single factor to the standardisation of English. It is obvious that for the production of printing fonts a standard form of the language must be agreed upon. This applied above all to spelling, an area of English which was quite chaotic in the pre-printing days of the Middle English period.

The dialects of Middle English

The dialectal position of Middle English is basically a continuation of that of Old English. The most important extralinguistic fact for the development of the Middle English dialects is that the capital of the country was moved from Winchester (in the Old English period) to London by William the Conqueror in his attempt to diminish the political influence of the native English.

NORTHERN This dialect is the continuation of the Northumbrian variant of Old English. Note that by Middle English times English had spread to (Lowland) Scotland and indeed led to a certain literary tradition developing there at the end of the Middle English period which has been continued up to the present time (with certain breaks, admittedly).

Characteristics. Velar stops are retained (i.e. not palatalised) as can be seen in word pairs like rigg/ridge; kirk/church.

KENTISH This is the most direct continuation of an Old English dialect and has more or less the same geographical distribution.

Characteristics. The two most notable features of Kentish are (1) the existence of /e:/ for Middle English /i:/ and (2) so-called "initial softening" which caused fricatives in word-initial position to be pronounced voiced as in vat, vane and vixen (female fox).

SOUTHERN West Saxon is the forerunner of this dialect of Middle English. Note that the area covered in the Middle English period is greater than in the Old English period as inroads were made into Celtic-speaking Cornwall. This area becomes linguistically uninteresting in the Middle English period. It shares some features of both Kentish and West Midland dialects.

WEST MIDLAND This is the most conservative of the dialect areas in the Middle English period and is fairly well-documented in literary works. It is the western half of the Old English dialect area Mercia.

Characteristics. The retention of the Old English rounded vowels /y:/ and /ø:/ which in the East had been unrounded to /i:/ and /e:/ respectively.

EAST MIDLAND This is the dialect out of which the later standard developed. To be precise the standard arose out of the London dialect of the late Middle English period. Note that the London dialect naturally developed into what is called Cockney today while the standard became less and less characteristic of a certain area and finally (after the 19th century) became the sociolect which is termed Received Pronunciation.

Characteristics. In general those of the late embryonic Middle English standard.

Middle English phonology

The following sections should be seen in the context of the above one Writing and Sounds of Old English as it offers a discussion of the main changes between Old and Middle English and elaborates on some of the features of Middle English which are relevant to developments today.

LENGTHENING IN OPEN SYLLABLES This is a phonological process which started in the north of England in the 13th century and affected the high vowels /i/ and /u/ in the following century. It is one of the major sound changes of early Middle English and involves lengthening and lowering as seen in the following examples.

RETENTION OF MORA THROUGH COMPENSATORY LENGTHENING To understand Open Syllable Lengthening properly one must start with the notion of mora. A mora corresponds metrically to the quality of a short vowel; all long vowels and diphthongs are bimoric in English. The constituents of a syllable correspond to morae in metrics. One can see that in the history of English various cases of compensatory lengthening recognize that with the loss of a consonant its mora is transferred to a preceding vowel, for instance light /lɪxt/ [lɪçt] (-VCC) > /li:t/ (-V:C).

SYLLABIC RESTRUCTURING Recall here that a metrical foot (F = foot) refers to those syllables which stand between two stressed (S = strong, i.e. stressed, indicated by a superscript stroke: ˡ) syllables including the first stressed syllable, irrespective of the number of weak, i.e. unstressed (W = weak) syllables after it: ˡHe’s preˡdicting a ˡlandslide ˡvictory.

The labels S, ‘strong’, and W, ‘weak’, refer to the relative accentuation of the syllable. With the designations L, ‘light’, and H, ‘heavy’, the reference is to the quantity of the syllable. The correlation between strong and heavy on the one hand and weak and light on the other is in Middle English such that when a syllable is the only one in a foot then it must also be ‘heavy’, hence the lengthening of short stressed vowels after the loss of final /ə/. The entire metrical quantity of the words was retained by Open Syllable Lengthening.

VOWELS BEFORE /X/ In West Saxon there were only two recognisable variants for /x/, [h] in initial position, [x] in all other positions, irrespective of the preceding vowel. It was only towards the end of the Old English period that the variant [ç] appears as an allophone after front vowels. In Middle English this led to a diphthong with the mid front vowels /e/ and /e:/.

There is an equivalent to the diphthongisation of [e(:)ç] to [eiç] with back vowels. With the latter vowels the allophone was [x] up to early Middle English. During this period a velar glide appears before this sound, [u], the back equivalent to /i/ with [ç]. This merged with the preceding vowel and resulted in a diphthong.

THE SHIFT OF /X/ TO /F/Already by the 14th century a shift is to be found in English which is common in many other languages as well. It is the shift among fricatives from velar to labial place of articulation. In English the shift was unidirectional and represents one of the many reflexes of /-x/ in Modern English (the remaining reflexes are vocalic). Note that there are two main outputs from this shift, one with an original high vowel /ʊ/ (later lowered to /ʌ/) and one with a mid back vowel /ɔ/; occasionally the shift occurred with a mid front vowel, cf. the form dwarf which itself shows later lowering of /e/ to /a/ before /r/.

SHORT VOWEL DEVELOPMENTS The development of late Old English /y/ differed in the various dialects. In Kentish the vowel had already unrounded to /ɛ/ in the late 9th century; in the west midlands and in the south-west it was retained in the Middle English period. In the east midlands it was probably unrounded early on (after the 11th century).

late OE ME
y i    North and East Midlands
  y    West Midlands and South
  ɛ    Kent

There are many examples for the unrounding in the east midlands.

The development in the east applies to those cases where there was no phonetic conditioning. If, however, /y/ came after a labial or in the environment of /ʃ/ or before /dʒ, tʃ, ʃ/ it was retracted to /u/. Western dialects show the retraction already in the 12th century and this is responsible for many of the spellings with u to the present day.

In the west midlands and in the south a front vowel was retained longest. The spelling u there stands for /y/ and is not restricted to the environment before /dʒ, tʃ, ʃ/, cf. gult /gylt/ ‘guilt’, kun /kyn/ ‘kin’.

For Kentish /ɛ/ is attested. The vowel is the short equivalent zu /e:/ which was already to be found in Kentish instead of West Saxon /y:/. Cf. gelt /gɛlt/ ‘guilt’, ken /kɛn/ ‘kin’.

The modern standard shows forms which can be traced to the various dialects. The phonetic possibilities are /ɪ/, /ɛ/ or /ʌ/ (from earlier /ʊ/) and the spelling can be i,e or u. There are many instances of a mixture of spelling from one dialect and pronunciation from another.


busy West Midlands East Midlands
bury West Midlands Kent
merry Kent Kent
shut West Midlands West Midlands
pit East Midlands East Midlands

Apart from the above developments the short vowels of English have remained remarkably stable throughout the history of the language, for instance Old English cwic, god show the same vowels in Modern English. The two main changes which occur later are (1) /ʊ/ > /ʌ/ after the mid-17th century and (2) an earlier raising of /ɛ/ to /ɪ/ before nasals as in think [θɪk] and English [ɪŋglɪʃ].

Lowering of /e/ to /a/

This is a development which began in the north and spread to the south after about 1400. It is difficult to date this exactly as there is no orthographical indication of the shift. The lowering of /e/ to /a/ explains the present-day pronunciations of many proper names in England such as Derby, Hertfordshire, Berkeley (the name of the philosopher, not that of the Californian city). This shift was very common and in many cases the orthography has been adapted to the pronunciation so that these words cannot be recognised as having originally involved the shift, e.g. dark (from derk), barn (from bern), heart (from herte). The shift affected words irrespective of origin, hence some French loans also have the shift. Note that many instances did not become established and the /er/ (later /ɜ:/) pronunciation was retained.

Early Modern English
serve /sarv/  >  /sɜ:v/
certain /sartɪn/  >  /sɜ:tən/
fervent /farvɪnt/  >  /fɜ:vɪnt/

In one case this development led to a semantic distinction between two words one with the lowered vowel and one without. The word parson is a form of person with this lowering and came to mean not just any person but an ecclesiastical person and so the two forms continued with separate meanings in the standard.

THE LOSS OF FINAL /-ə/ The loss of phonetic substance in words is one of the most remarking developments in the history of English. It is already attested in Old English in the simplification of consonants. Later vowels in unstressed syllables lost their distinctiveness, then a final inflectional nasal was dropped and finally — probably by the 14th century — the remaining shwa, [ə], disappeared as can be seen in the following sequence.

drīfan /dri:van/ > /dri:vən/ > /dri:və/ > /dri:v/ (> /draiv/)

This phonetic loss always involves unstressed syllables and usually resulted in apocope (loss of endings). There are, however, instances of syncope (medial loss) and procope (initial loss). The latter can be seen quite clearly with the past participles of verbs which originally had a prefix ge- (cf. the similar prefix in German) but which was weakened progressively until it finally disappeared.

ge- /jə-/ > /i:/ > /ɪ/ > ø
OE gelufod ME yloved NE loved

This phonetic reduction had far-reaching consequences for the typology of English which gradually drifted from a synthetic type (Old English, much like German) to a more analytic type in modern times. The language developed means for compensating for the loss in manifestation of grammatical categories chiefly by a more rigid word order and by the increasing functionalisation of prepositions.

It is difficult to reconstruct the demise of final /ə/. The reason is quite simply that final -e continued to be written. The only sound proof is offered by a series of spellings in Middle English where the words have a final -e which is not etymologically justified.

cōle (OE col) ‘coal’
shīre (OE scīr) ‘county’

The loss of phonetic /-ə/ led to a refunctionalisation of the final written -e. It was henceforth used to indicate that the vowel of the preceding (stressed) syllable was long. This function has survived to the present day, cf. pane, with /ei/ from ME /a:/, and pan, with /æ/ from ME /a/.

THE LOSS OF FINAL /l/ In cases where final /-l/ is still spoken one must differentiate between those which represent a retention of an inherited /-l/ and those where the /-l/ is pronounced because it was reintroduced into the writing, e.g. ModE fault (< ME faute from French). The other cases show the loss of /l/ due to its being vocalised and absorbed into the preceding vowel, e.g. walk /walk/ > /wɑ:k/ > /wɔ:k/. The /l/ was probably realised as a velarised [ɫ] before its loss. The long vowel which resulted was later retracted and raised as indicated in the transcriptions just given, cf. other words as well, e.g. talk, stalk, baulk (the latter with an orthographic indication of the /ɔ:/ vowel). The situation is slightly different where the present-day English word shows a long low vowel. Here the /-l/ can have disappeared without necessarily having caused a diphthong, e.g.calf /ka:f/, half /ha:f/.

Changes in grammar

The older noun system

The noun system of Old English was quite complex with 3 genders (masculine, feminine and neuter) and 5 cases (nominative, accusative, genitive, dative, instrumental). In the history of English this was simplified considerably. The cases were reduced to nominative and genitive and the phenomenon of grammatical gender was lost.

In addition to gender and case Old English had a variety of plural types just like Modern German. The number of these has steadily declined throughout the centuries. This demise can be documented quite clearly and the reduction in diversity shows a definite sequence which can be summarised as follows:

umlaut plurals man ~ men
/r/-plurals child ~ childer
/n/-plurals ox ~ oxen
/s/-plurals stone ~ stones

It is clear from Old English that the umlaut plurals declined quite rapidly. We have words like cu with an original plural cy (compare German Kuh ~ Kühe) which later becomes cyne (cf. the Early Modern English form kine). /r/ plurals are replaced by nasal plurals in the early Middle English period as the present-day form children shows which has a nasal ending added to an original childer (/r/-plural). The nasal plurals themselves pass into decline by the late Middle English period (Chaucer still has eyen for modern eyes).

Now note that the umlaut plurals are an inherited type from pre-Old English forms of Germanic which English shares with other forms of North and West Germanic like German and Swedish. From the point of view of iconicity (a form indicates a grammatical category) the umlaut plurals fulfill their function well. The singular and the plural forms of words are clearly distinguished.

However, from the point of view of the language’s grammar the umlaut type contravenes a principle of morphology which requires that there be an isomorphic relationship (one to one in form) between the lexical root or stem of a word and any inflection added to it. Ideally, this type of situation applies to agglutinative languages. Looking at the remaining three formational types one sees that they form a scale of decreasing sonority (open, vowel-like quality): /r/ -> /n/ -> /s/. A general principle of morphology is that inflectional endings are favoured which show a high degree of phonetic salience. It is this factor which increases with progression in types just indicated.

The point being made here is that any explanation which tries to avail of either phonological or morphological arguments exclusively is doomed to failure as there was a morphological reason for loss in the first case and a phonological reason in each of the remaining three.

The effects of the above changes on the morphology of Middle English were very considerable. They led to a loss of distinctiveness among grammatical endings so that the various declensional classes of Old English collapsed, with the dative plural remaining for a while the only case — with a final nasal /-n/ — which was distinctive, but even that was reduced in the course of the Middle English period. A direct consequence of this was that the more common declensions were generalised and used productively. The two main ones are the s-type and the nasal type as seen in the Old English words stān ‘stone’ : stānas ‘stones’ and ēage ‘eye’ : ēagan ‘eyes’ respectively. For a while the nasal declension was productive as is seen in its addition to the old r-plural child : childer > child(e)ren to give the doubly marked plural which has survived to the present day. The north of the country was as always innovative and by about 1200 nouns are commonly found with a plural and a genitive singular in /-s/, this then spreading to the south somewhat later and with time it replaced virtually all nasal plurals. There are a few remainders into the time of Shakespeare — e.g. eyen, shoon, housen — but these have been brought into line with the universal s-plural so that nowadays there are only three nasal plurals remaining: oxen, children and brethren (a double plural with an umlaut of brother and a nasal ending).

There are a few other plural types of which reflexes still exist in English. Most noticeable are umlaut plurals which are the forerunners to the modern word pairs foot : feet, goose : geese, man : men, mouse : mice. These nouns are part of the core of English vocabulary and are nearly all terms for humans, parts of the body or familiar animals. Still less significant are the few examples of zero plurals, all terms for animals as in sheep : sheep, deer : deer, fish : fish (an analogical plural form fishes also exists).

Be careful to distinguish these instances of inherited plural types from cases which are derived from direct imports from Latin or Greek. Hence in Modern English one has pairs like formula : formulae, criterion : criteria which are direct loans from Latin and Greek respectively and show the plural endings typical of these languages.

The reduction in morphological variation which is found with nouns applied to other word-classes as well. Adjectives lost their endings so that the previous distinction between a strong and weak declension — as with Modern German dichter Nebel and der dichte Nebel — was lost.

Equally one can notice a loss in grammatical gender in the transition from Old to Middle English. The older stage of the language showed three genders as in Modern German, masculine, feminine and neuter, distributed on arbitrary grounds, e.g. the word wīf was neuter (cf. German das Weib). There were three forms of the definite article þe, and sēo. By the end of the Middle English period there was only one form, the modern the (which derives from Old English þe). The consequence of the loss of grammatical gender is that it was replaced by natural gender in most instances. There are examples in Modern English in which another gender is used — for instance, a feminine reference is used for technical objects such as cars, planes or ships — but this is more an analogical extension of natural gender rather than a survival of grammatical gender.

The older verb system

The verbal system in Middle English was subjected to inflectional decay just as much as the nominal area. What is of greatest interest for the present-day language is the set of developments which led to a more complex system of tenses in English. To begin with recall that the inherited Germanic system had two tenses, the present and the perfect. As inflections were disappearing any expansion of tenses could only be achieved by the use of analytic devices, in this case in the increased deployment of auxiliaries. In Modern English there are two types, the primary auxiliaries be and have along with do which arose later. The secondary auxiliaries are modals, in particular two shall and will, which are used for the formation of the future and which developed out of full verbs with lexical meanings originally, i.e. shall‘to be obliged to’ and will ‘to want to’, something which they have not entirely lost to this day.

The primary auxiliaries were originally distinguished according to verb type. Have was the default used for the perfect and be was employed with verbs of motion and state, much as in Modern German. This lasted for some considerable time and in Shakespeare, for example, we find this distribution of auxiliaries still, e.g. Our revels now are ended (at the end of The Tempest) with a form of be rather than have as one would expect today.

In Old English there was a further verb in the auxiliary system, namely weorþan which was used to form the passive (note that this is etymologically related to German werden which has the same purpose). In the course of the Middle English period this verb disappeared and its functions were taken over by be.

The verb ‘be’ consists of two main lexical roots, bēon and wesan, both of which are inherited from Indo-European and represented in Modern English by be and was, were nowadays. The is form stems from a third source and the plural form are represents a Scandinavian borrowing which entered northern dialects and was later adopted into southern English.

The continuous tenses, so familiar from Modern English, are a Middle English development. Old English had indeed -ing forms but these were nominal in character (later called gerunds), for instance ræding meant more the act of reading as in German Das Lesen des Buches hat lange gedauert. In addition in Old English the gerundial form usually had the ending -end(e) or -and(e) and indicated more a state than a process or action, again compare German as in Er ist lebend herausgekommen. In the Middle English period the -nd forms were supplanted fully by -ing forms. In addition there developed a truly continuous tense type, most probably from a construction which has the meaning of being engaged in an action. Consider the sentence type He was sleeping. This can be traced back to a type He was asleeping, where the a- prefix is itself derived from the preposition on so that the sentence originally meant ‘He was on sleeping’, i.e. he was engaged in the act of sleeping, much as in German Er war am Schlafen. The (unstressed) prefix a- was lost in the course of time and the phrase was reanalysed by later generations as expressing a continuous tense rather than engagement in a certain action or being in a given state. The stative use is still found in adverbial phrases in English (and also in German) Shaking with fear, he entered the room. Zitternd vor Angst, betrat er das Zimmer.

Sound changes

ALLOPHONES TO PHONEMES Word-medial fricatives [v, z, ð] were allophones in Old English, e.g. līf ‘life’ ended in [-f] whereas libban — cf. the preterite form with the voiced fricative lifde — had an internal voiced fricative [-v-]. However with the loss of the inflectional endings the verb was reduced to a single syllable and the final /-v/ now contrasted with the final /-f/ of the noun, hence the change in status of voice among fricatives in Middle English which from then on distinguished phonemes.

The system status of voiced and voiceless fricatives was strengthened by the fact that with French loans instances of voiced initial fricatives now occurred in English, adding to the functional load — and hence to the systematic importance — of these segments.

vertu ‘virtue’ vileynye ‘villainy’ zēle ‘zeal’

Another development relevant to this issue is the loss of distinctive consonant length in Middle English. Recall that in Old English consonants could be long or short, e.g. /-s-/ [-z-] and /-ss-/ [-ss-] were phonemically distinct. cf. Offa (proper name), missan ‘miss’, siþþan ‘sit’ which show geminates (long consonants) for all three fricative types in Old English. In the Middle English period this distinction begins to be lost so that the instances of voiceless word-medial geminates were reduced to simple segments but remained voiceless, so that the original phonetic contrast between voiced medial non-geminate versus voiceless medial geminate was now reduced to a simple distinction between voiced and voiceless fricative in medial position, thus strengthening the phonemic significance of voice for these fricatives.

Later initial /θ-/ in grammatical formatives such as the, there, that, etc. was softened to [ð] because of the unstressed character of the words. Thus the series /v-, ð-, z-/ in initial position was completed.

THE WORD FOR ‘WOMAN’ This derives from Old English wīfmann ‘woman’ + ‘man’ (German ‘Weib’ + ‘Mann’). The plural was formed by umlaut on the second element, i.e. wīfmenn (cf. German Mann : Männer). Now in the course of time English lost the distinction between /a/ and /e/ in unstressed syllables — both collapsing to schwa /ə/ — and in addition in this form the fricative /f/ was lost before the nasal /m/ and the vowel of the first syllable was shortened, so that the form was then /wɪmən/ and homophonous in the singular and plural in many varieties of English. Now the initial /w/ had a retracting influence on the first vowel so that a form arose with a high back vowel rather than a high front vowel, i.e. /wʊmən/. This came to be used for the singular of the noun and the original pronunciation for the plural form, so that one had /wʊmən/-SG versus /wɪmən/-PL. With the lowering of /ʊ/ in the 17th century, most words adopted a value /ʌ/ for the former high vowel. However, the pronunciation of the singular of woman was retained as /wʌmən/ (probably under the influence of the preceding /w/, cf. would /wʊd/). Note that the spelling in modern English, woman : women, implies that the plural has a change in the vowel of the second syllable. This is not the case, the vowel of the first syllable changes, the spelling is just to form a parallel orthographical case to man : men (here the phonetic alternation is indeed between /æ/ and /e/ for the singular and plural respectively).

Compensation for the loss of inflections

If grammatical categories were indicated in Old English via inflections then the loss of the latter implied that something took their place. The answer to the question ‘what?’ is simple: word order and the increased functionalisation of prepositions. In Old English the order S - O - V (Subject - Object - Verb) was common but with the loss of inflections the indentification of Subject and Object was not always that simple. For this and other contributory reasons the order S - V - O (Subject - Verb - Object) became more usual in the course of the Middle English period. The order V - S - O (Verb - Subject - Object) which was also found in Old English declined in frequency, remaining most tenaciously after adverbs where it is still sometimes found today as in Hardly had he left the room when she rang.

The increased use of prepositions served the function of rendering sentences unambiguous. A simple example illustrates this. The German sentence Er schrieb ihr einen Brief has variants like Ihr schrieb er einen Brief or Einen Brief hat er ihr geschrieben all of which are possible because the inflected forms of the pronouns and the object noun are unambiguous with regard to sentence function. In English there are two equivalents to the sentence He wrote her a letter and He wrote a letter to her, the former uses word order to indentify the sentence elements functionally — indirect object precedes direct object — and the latter employs a preposition to to identify the indirect object. For topicalisation as in the German examples modern English has to resort to intonational strategies (stressing the highlighted element) or to the syntactic device of clefting which retains the prepositional object but moves it to the front by embedding it into a dummy sentence: It's to her that he wrote the letter.

Introductory books on Middle English