Phonological history of the English language

The phonological history of the English language describes changing phonology of English over time, starting from its roots in proto-Germanic to diverse changes in different dialects of modern English.

Within each section, changes are in approximate chronological order.

NOTE: In the following description, abbreviations are used as follows:The time periods for many of the following stages are extremely short due to the extensive population movements occurring during the early AD period, which resulted in rapid dialect fragmentation:
*The migration of the Goths from southeast Sweden to the Baltic Sea area around AD 1, followed by the migration to southeast Romania around AD 200. (Later migrations carried the Ostrogoths eastward to the Crimea area in modern Ukraine, and carried the Visigoths westward to Spain.)
*The migration of the High German ancestors southward, starting around AD 260, and renewed in the 5th century AD.
*The migration of the Anglo-Saxons westward into Britain, starting around AD 450.

Late Proto-Germanic period (c. AD 0–200)

This includes changes in late Proto-Germanic, up to the appearance of Proto-West-Germanic c. AD 200:

*Early i-mutation: IPA|/e/ is raised to IPA|/i/ when an IPA|/i/ or IPA|/j/ follows in the next syllable.
**This occurs before deletion of any unstressed vowels; hence PIE IPA|/bereti/ > PG IPA|/bereθi/ > IPA|/beriθi/ > Goth IPA|"baíriθ" IPA|/beriθ/ "(he) carries".
** The IPA|/i/ produced by this change can itself trigger later i-mutation. Hence WG IPA|/beriθ/ > IPA|/biriθ/ > OE IPA|/birθ/ "(he) bears".
*a-mutation: IPA|/u/ is lowered to IPA|/o/ when a non-high vowel follows in the next syllable.
**This is blocked when followed by a nasal followed by a consonant, or by a cluster with IPA|/j/ in it. Hence PG IPA|/ɡulda/ > OE/NE "gold", but PG IPA|/ɡuldjanan/ > OE "gyldan" > NE "gild".
**This produces a new phoneme IPA|/o/, due to inconsistent application and later loss of unstressed IPA|/a/ and IPA|/e/.
*Loss of IPA|/n/ before IPA|/x/, with nasalization and compensatory lengthening of the preceding vowel.
**The nasalization was eventually lost, but remained through the Ingvaeonic period.
**Hence PrePG IPA|/tonɡjonom/ > PG IPA|/θankjanan/ > OE "þencan" > NE "think", but PrePG IPA|/tonktoːm/ > PG IPA|/θanxtoːn/ > IPA|/θãːxtoːn/ > OE IPA|"þóhte" > NE "thought".
*Loss of final IPA|/m/, with nasalization (eventually lost) of the preceding vowel. Hence PrePG IPA|/dʱoɡʱom/ > PG IPA|/daɡam/ > PN IPA|/daɡa/ > WG IPA|/daɡ/ "day (acc. sg.)".
*Pre-nasal raising: IPA|/e/ > IPA|/i/ before nasal + consonant. PrePG IPA|/bʱendʱonom/ > PG IPA|/bendanan/ > IPA|/bindanan/ > OE "bindan" > NE "bind" (Latin IPA|"of-fendō").
**This post-dated lost of IPA|/n/ before IPA|/x/.
**This was later extended in PreOE times to vowels before all nasals; hence OE "niman" "take" but OHG "neman".
*IPA|/ei/ > IPA|/iː/ (c. AD 100). The Elder Futhark of the Proto-Norse language still contain different symbols for the two sounds.
*Vowels in unstressed syllables were reduced or eliminated. The specifics are quite complex and occurred as a result of many successive changes, with successive stages often happening hundreds of years after the previous stage. Some specifics of the initial stage:
**Final-syllable short vowels inherited from Proto-Germanic were generally deleted. Hence Goth IPA|"baíriθ" IPA|/beriθ/ "(he) carries" < PG IPA|/bereθi/ (see above).
***This operated universally only in words of three syllables or more. In words of two syllables, final-syllable IPA|/a/ and IPA|/e/ were deleted, but IPA|/i/ and IPA|/u/ were unaffected following a short syllable (i.e. one with a short vowel followed by a single consonant.) Hence PG IPA|/daɡaz/ > Goth "dags" "day (nom. sing.)" (OE "dæg"), PIE IPA|/woida/ > PG IPA|/waita/ > Goth "wáit" "(I) know" (OE IPA|"wát"), PIE IPA|/woide/ > PG IPA|/waite/ > Goth "wáit" "(he) knows" (OE IPA|"wát"); but PIE IPA|/sunus/ > PG IPA|/sunuz/ > Goth "sunus" "son (nom. sing.)" (OE "sunu"), PIE IPA|/peku/ > PG IPA|/fehu/ > Goth "faíhu" IPA|/fehu/ "cattle (nom. sing.)" (OE "feohu"), PIE IPA|/wenis/ > PG IPA|/weniz/ > IPA|/winiz/ > OHG "wini" "friend (nom. sing.)" (OE "wine"), PIE IPA|/poːdi/ > PG IPA|/foːti/ > PreOE IPA|/føːti/ > OE IPA|"fét" "foot (dat. sing.)".
***Final-syllable IPA|/a/ and IPA|/e/ were protected in words of two syllables by following IPA|/r/ and IPA|/ns/. Hence PG IPA|/fader/ > NE "father"; PG IPA|/stainans/ > Goth "stáinans" "stone (acc. pl.)".
***Final-syllable IPA|/a/ and IPA|/e/ in two-syllable words were still present in Proto-Norse. PN IPA|/daɡaz/, Goth "dags" "day (nom. sg.)". PN IPA|/daɡa/, Goth "dag" "day (acc. sg.)".
**Final-syllable long vowels were shortened.
***But final-syllable IPA|/oː/ becomes IPA|/u/ in NWG, IPA|/a/ in Gothic. Hence PG IPA|/beroː/ > early OE "beru" "(I) carry", but Goth "baíra"; PG IPA|/ɡeboː/ > OE "giefu" "gift (nom. sg.)", but Goth "giba".
**Middle-syllable vowels of all types were unchanged; likewise in monosyllables, since they were stressed.
**"Extra-long"' vowels were shorted to long vowels. There is a great deal of argument about what is exactly going on here.
***The traditional view is that a circumflex accent arose (as in Ancient Greek) when two adjacent vowels were contracted into a single long vowel in a final syllable. This circumflexed vowel then remained long when other long vowels shortened.
***A newer view holds that "overlong" (tri-moraic) vowels arose from the contraction of two vowels, one of which was long. Furthermore, final-syllable long vowels remained long before certain final consonants (IPA|/z/ and IPA|/d/).
***The reason why such theories are necessary is that some final-syllable long vowels are shortened, while others remain. Nominative singular IPA|/-oːn/ shortens, for example; likewise first singular IPA|/-oːn/ < IPA|/-oːm/; while genitive plural IPA|/-oːn/ < IPA|/-oːm/ remains long. Both of the above theories postulate an overlong or circumflex ending IPA|/-ôːn/ in the genitive plural arising in the vocalic (PIE IPA|/o/ and IPA|/aː/, PG IPA|/a/ and IPA|/oː/) declensions, arising from contraction of the vocalic stem ending with the genitive plural ending.
***Other examples of vowels that remain long are "a"-stem and "ó"-stem nominative plural IPA|/-ôz/ < early PIE IPA|/-o-es/ and IPA|/-aː-es/; PrePG ablative singular IPA|/-ôd/, IPA|/-êd/ (Gothic IPA|"ƕadrē" "whither", "undarō" "under"); IPA|/ō/-stem dative singular PG IPA|/ɡibâi/ > Goth "gibái" "gift" (but IPA|/a/-stem dative singular PG IPA|/stainai/ > Goth "staina" "stone").

West Germanic period (c. AD 200–400)

This includes changes up through the split of Ingvaeonic and High German (c. AD 400):

*Unstressed diphthongs were monophthongized. IPA|/ai/ > IPA|/æː/, IPA|/au/ > IPA|/oː/.
**Results were different in Gothic. Diphthongs remained except for absolutely final diphthongs stemming from PIE short diphthongs, which became short IPA|/a/.
**Hence PIE IPA|/sunous/ > PG IPA|/sunauz/ > Goth "sunáus", but > PWG IPA|/sunoː/ > OE "suna" "son (gen. sing.)"; PIE IPA|/nemoit/ > PG IPA|/nemait/ > IPA|/nimait/ > Goth "nimái", but > PWG IPA|/nimæː/ > OE "nime" "(he) takes (subj.)"; PIE (loc.?) IPA|/stoinoi/ > PG IPA|/stainai/ > Goth "staina", but > PWG IPA|/stainæː/ > OE IPA|"stáne" "stone (dat. sing.)"; PIE (loc.?) IPA|/ɡʱebʱaːi/ > PG IPA|/ɡebâi/ > Goth "gibái", but > PWG IPA|/ɡebæː/ > OE "giefe" "gift" (dat. sing.).
*IPA|/æː/ becomes IPA|/aː/ IPA| [ɑː] .

*Elimination of word-final IPA|/z/.
**Note that this change must have occurred before rhoticization, as original word-final IPA|/z/ did not become IPA|/r/.
**But it must have occurred after the North-West-Germanic split , since word-final IPA|/z/ was not eliminated in Old Norse, instead merging with IPA|/r/.
*Rhoticization: IPA|/z/ > IPA|/r/.
**This change also affected Proto-Norse; but in Proto-Norse, the date and nature are contested. IPA|/z/ and IPA|/r/ were still distinct in the Danish and Swedish dialect of Old Norse, as is testified by distinct runes. (IPA|/z/ is normally assumed to be a rhotic fricative in this language, but there is no actual evidence of this.)
*West Germanic Gemination of consonants except IPA|/r/, when preceded by a short vowel and followed by IPA|/j/.
*OE nominative plural IPA|/as/ (ME IPA|/s/), OS nominative plural IPA|/oːs/ may be from original accusative plural IPA|/ans/ (rather than original nominative plural IPA|/oːz/; cf. ON nominative plural IPA|/ar/), following Ingvaeonic nasalization/loss of nasals before fricatives.

=Ingvaeonic and Proto-Anglo-Frisian period (c. AD 400–475)=

This includes changes from c. AD 400 up through the split of the Anglo-Frisian languages from Ingvaeonic, followed by the split of pre-Old English from pre-Old Frisian (c. AD 475). The time periods for these stages are extremely short due to the migration of the Anglo-Saxons westward through Frisian territory and then across the English Channel into Britain, around AD 450.

*Ingvaeonic nasal spirant law: Loss of nasals before fricatives, with compensatory lengthening. Hence PG IPA|/munθaz/ > NHG "Mund" but OE IPA|"múþ", NE "mouth".
**An intermediate stage was a long nasal vowel, where nasal IPA|/ãː/ > IPA|/õː/. PIE IPA|/dontos/ > PG IPA|/tanθaz/ > OE IPA|"tóθ" "tooth". (NHG "Zahn" < OHG "zant".)
*Development of new IPA|/ɑ/-IPA|/æ/ distinction through Anglo-Frisian brightening and other changes:
**Fronting of IPA|/ɑː/ to IPA|/æː/ (generally, unless IPA|/w/ followed).
**Fronting of IPA|/ɑ/ to IPA|/æ/ (unless followed by a geminate, by a back vowel in the next syllable, or in certain other cases). Hence OE "dæg" IPA|/dæj/ "day", plural "dagas" IPA|/dɑɣɑs/ "days" (dialectal NE "dawes"; compare NE "dawn" < OE "dagung" IPA|/dɑɣunɡ/). Gothic "dags", plural IPA|"daɡós".
**Change of IPA|/ai/ to IPA|/ɑː/. PG IPA|/stainaz/ > OE IPA|"stán" > NE "stone".

Old English period (c. AD 475–900)

This includes changes from the split between Old English and Old Frisian (c. AD 475) up through historic early West Saxon of AD 900:

*Breaking of front vowels
**Most generally, before IPA|/x/, IPA|/w/, IPA|/r/ + consonant, IPA|/l/ + consonant (assumed to be velar IPA| [ɹ] , IPA| [ɫ] in these circumstances), but exact conditioning factors vary from vowel to vowel
**Initial result was a falling diphthong ending in IPA|/u/, but this was followed by diphthong height harmonization, producing short IPA|/æ̆ɑ̆/, IPA|/ɛ̆ɔ̆/, IPA|/ɪ̆ʊ̆/ from short IPA|/æ/, IPA|/ɛ/, IPA|/ɪ/, long IPA|/æɑ/, IPA|/eo/, IPA|/iu/ from long IPA|/æː/, IPA|/eː/, IPA|/iː/. (Written "ea", "eo", "io", where length is not distinguished graphically.)
**Result in some dialects, for example Anglian, was back vowels rather than diphthongs. West Saxon "ceald"; but Anglian "cald" > NE "cold".
*IPA|/ɪ̆ʊ̆/ and IPA|/iu/ were lowered to IPA|/ɛ̆ɔ̆/ and IPA|/eo/ between 800 and 900 AD.
*By the above changes, IPA|/au/ was fronted to IPA|/æu/ and then modified to IPA|/æa/ by diphthong height harmonization.
**PG IPA|/draumaz/ > OE IPA|"dréam" "joy" (cf. NE "dream", NHG "Traum"). PG IPA|/dauθuz/ > OE IPA|"déaþ" > NE "death" (Goth IPA|"dáuθus", NHG "Tod"). PG IPA|/auɡoː/ > OE IPA|"éage" > NE "eye" (Goth IPA|"áugō", NHG "Auge").
*IPA|/sk/ was palatalized to IPA|/ʃ/ in almost all circumstances. PG IPA|/skipaz/ > NE "ship" (cf "skipper" < Dutch "schipper", where no such change happened). PG IPA|/skurtjaz/ > OE "scyrte" > NE "shirt", but > ON "skyrt" > NE "skirt".
*IPA|/k/, IPA|/ɣ/, IPA|/ɡ/ were palatalized to IPA|/ʧ/, IPA|/j/, IPA|/ʤ/ in certain complex circumstances, described in detail on the Old English page.
**This change, or something similar, also occurred in Old Frisian.
*Back vowels were fronted when followed in the next syllable by IPA|/i/ or IPA|/j/, by i-mutation (c. 500 AD).
**i-mutation affected all the Germanic languages except for Gothic, although with a great deal of variation. It appears to have occurred earliest, and to be most pronounced, in the Schleswig-Holstein area (the home of the Anglo-Saxons), and from there to have spread north and south.
**This produced new front rounded vowels IPA|/œ/, IPA|/øː/, IPA|/ʏ/, IPA|/yː/. IPA|/œ/ and IPA|/øː/ were soon unrounded to IPA|/ɛ/ and IPA|/eː/, respectively.
**All short diphthongs were mutated to IPA|/ɪ̆ʏ̆/, all long diphthongs to IPA|/iy/. (This interpretation is controversial. These diphthongs are written "ie", which is traditionally interpreted as short IPA|/ɪ̆ɛ̆/, long IPA|/ie/.)
**Late in Old English (c. AD 900), these new diphthongs were simplified to IPA|/ʏ/ and IPA|/yː/, respectively.
**The conditioning factors were soon obscured (loss of IPA|/j/ whenever it had produced gemination, lowering of unstressed IPA|/i/), phonemicizing the new sounds.
*More reductions in unstressed syllables:
**IPA|/oː/ became IPA|/ɑ/.
**Germanic high vowel deletion eliminated IPA|/ɪ/ and IPA|/ʊ/ when following a heavy syllable.
*Palatal diphthongization: Initial palatal IPA|/j/, IPA|/ʧ/, IPA|/ʃ/ trigger spelling changes of "a" > "ea", "e" > "ie". It is disputed whether this represents an actual sound change or merely a spelling convention indicating the palatal nature of the preceding consonant (written "g", "c", "sc" were ambiguous in OE as to palatal IPA|/j/, IPA|/ʧ/, IPA|/ʃ/ and velar IPA|/ɡ/ or IPA|/ɣ/, IPA|/k/, IPA|/sk/, respectively).
**Similar changes of "o" > "eo", "u" > "eo" are generally recognized to be merely a spelling convention. Hence WG IPA|/junɡ/ > OE "geong" IPA|/junɡ/ > NE "young"; if "geong" literally indicated an IPA|/ɛ̆ɔ̆/ diphthong, the modern result would be *"yeng".
**It is disputed whether there is Middle English evidence of the reality of this change in Old English.
*Initial IPA|/ɣ/ became IPA|/ɡ/ in late Old English.

Up through Chaucer's English (c. AD 900–1400)

*Vowels were lengthened before IPA|/ld/, IPA|/mb/, IPA|/nd/, IPA|/rd/, probably also IPA|/nɡ/, IPA|/rl/, IPA|/rn/, when not followed by a third consonant.
**This probably occurred around AD 1000.
**Later on, many of these vowels were shortened again; but evidence from the Ormulum shows that this lengthening was once quite general.
**Remnants persist in the Modern English pronunciations of words such as "child" (but not "children", since a third consonant follows), "field" (plus "yield", "wield", "shield"), "climb", "find" (plus "mind", "kind", "bind", etc.), "fiend", "found" (plus "hound", "bound", etc.).
*Vowels were shortened when followed by two or more consonants, except when lengthened as above.
**This occurred in two stages, the first stage affecting only vowels followed by three or more consonants.
*Inherited height-harmonic diphthongs were monophthongized by the loss of the second component, with the length remaining the same.
*IPA|/æː/ and IPA|/ɑː/ became IPA|/ɛː/ and IPA|/ɔː/.
*IPA|/æ/ and IPA|/ɑ/ merged into IPA|/a/.
*IPA|/ʏ/ and IPA|/yː/ were unrounded to IPA|/ɪ/ and IPA|/iː/.
*IPA|/ɣ/ became IPA|/w/ or IPA|/j/, depending on surrounding vowels.
*New diphthongs formed from vowels followed by IPA|/w/ or IPA|/j/ (including from former IPA|/ɣ/).
**Length distinctions were eliminated in these diphthongs.
**Diphthongs also formed by the insertion of a glide IPA|/w/ or IPA|/j/ (after back and front vowels, respectively) preceding IPA|/x/.
**Many diphthong combinations soon merged.
*Trisyllabic laxing: Shortening of stressed vowels when two syllables followed.
**This results in pronunciation variants in Modern English such as "divine" vs "divinity" and "south" vs. "southern" (OE "súðerne").
*Middle English open syllable lengthening: Vowels were usually lengthened in open syllables (13th century), except when trisyllabic laxing would apply.
*Remaining unstressed vowels merged into IPA|/ə/.
*Initial clusters IPA|/hɾ/, IPA|/hl/, IPA|/hn/ were reduced by loss of IPA|/h/.
*Voiced fricatives became independent phonemes through borrowing and other sound changes.
*IPA|/sw/ before back vowel becomes IPA|/s/; IPA|/mb/ becomes IPA|/m/.
**Modern English "sword", "answer", "lamb".
**IPA|/w/ in "swore" is due to analogy with "swear".

Up to Shakespeare's English (c. AD 1400–1600)

*Loss of most remaining diphthongs.
**IPA|/ai/ (and former IPA|/ɛi/, merged into IPA|/ai/ in Early Middle English) became IPA|/ɑː/ before the Great Vowel Shift.
**IPA|/ou/ (and former IPA|/ɔu/, merged into IPA|/ou/ in Early Middle English) became IPA|/oː/ and IPA|/ei/ became IPA|/eː/ after the shift causing the long mid mergers.
**IPA|/au/ became IPA|/ɔː/ after the shift.
**The dew-new merger: IPA|/u/ and /iu/ merger, and they then become IPA|/juː/ after the shift.
**The joy-point merger: IPA|/i/ and /oi/ merge, so that "point" and "joy" now have the same vowel.
**The rein-rain merger: /ai/ and /ei/ merge, so that "rain" and "rein" are now homonyms.
**The dew-duke merger: /y/ and /iu/ merge, so that "dew" and "duke" now have the same vowel.
**IPA|/oi/ remained.
**A few regional accents, including some in Northern England, East Anglia, South Wales, and even Newfoundland, monophthongization has not been complete, so that pairs like "pane" /"pain" and "toe"/"tow" are distinct. (Wells 1982, pp. 192–94, 337, 357, 384–85, 498)
*IPA|/x/ (written "gh") lost in most dialects causing the taut-taught merger.
*Great Vowel Shift; all long vowels raised or diphthongized.
**IPA|/aː/, IPA|/ɛː/, IPA|/eː/ become IPA|/ɛː/, IPA|/eː/, IPA|/iː/, respectively.
**IPA|/ɔː/, IPA|/oː/ become IPA|/oː/, IPA|/uː/, respectively.
**IPA|/iː/, IPA|/uː/ become IPA|/əi/ and IPA|/əu/, later IPA|/ai/ and IPA|/au/.
**New IPA|/ɔː/ developed from old IPA|/au/ (see above).
**Note that IPA|/ɔː/, IPA|/oː/, IPA|/uː/, IPA|/au/ effectively rotated in-place.
**IPA|/ɛː/, IPA|/eː/ are shifted again to IPA|/eː/, IPA|/iː/ in Early Modern English, causing merger of former IPA|/eː/ with IPA|/iː/; but the two are still distinguished in spelling as "ea", "ee".
*Loss of IPA|/ə/ in final syllables.
*Initial cluster IPA|/ɡn/ loses first element; but still reflected in spelling.
*/kn/ reduces to /n/ in most dialects, causing the not-knot merger.
*/wr/ reduces to /r/ in most dialects, causing the rap-wrap merger.
*Doubled consonants reduced to single consonants.

Up to the American–British split (c. AD 1600–1725)

*At some preceding time after Old English, all IPA|/r/ become IPA|/ɹ/.
**Evidence from Old English shows that, at that point, the pronunciation IPA|/ɹ/ occurred only before a consonant.
**Scottish English has IPA|/r/ consistently.
* The foot-strut split: Except in northern England, IPA|/ʊ/ splits into IPA|/ʊ/ (inconsistently after labials), as in "put", IPA|/ʌ/ (otherwise), as in "cut".
*Ng coalescence: Reduction of IPA|/nɡ/ in most areas produces new phoneme IPA|/ŋ/.
*Palatalization of IPA|/tj/, IPA|/sj/, IPA|/dj/, IPA|/zj/ produces IPA|/ʧ/, IPA|/ʃ/, IPA|/ʤ/, and new phoneme IPA|/ʒ/ (for example "measure", "vision"). Received Pronunciation resisted against this kind of coalescence until the 20th century.
**These combinations mostly occurred in borrowings from French and Latin.
**Pronunciation of "-tion" was IPA|/sjən/ from Old French IPA|/sjon/, thus becoming IPA|/ʃən/.
*Long vowels inconsistently shortened in closed syllables. (Modern English "head", "breath", "bread", "blood", etc.)
* The meet-meat merger: "Meet" and "meat" become homonyms in most accents.
* Changes affect short vowels in many varieties before an IPA|/r/ at the end of a word or before a consonant
** IPA|/a/ as in "start" and IPA|/ɔ/ as in "north" are lengthened.
** IPA|/ɛ/, IPA|/ɪ/ and IPA|/ʌ/ merge, hence most varieties of Modern English have the same vowel in each of "fern", "fir" and "fur".
** Also affects vowels in derived forms, so that "starry" no longer rhymes with "marry".
** Scottish English unaffected.
*IPA|/a/, as in "cat" and "trap", fronted to IPA| [æ] in many areas.
**But backed, rounded, and lengthened to IPA|/ɔː/ before syllable-final (that is, velarized) IPA|/l/ (IPA| [ɫ] ). Modern English "tall", "talk", "bald", "salt", etc. But IPA|/ɑ/ in "-alm", IPA|/æ/ in "-alf".
**New phoneme IPA|/ɑ/ develops from IPA|/alm/ ("calm" IPA|/kɑm/) and in certain other words, for example "father" IPA|/fɑðə(ɹ)/.
**Most varieties of northern English English, Welsh English and Scottish English retain IPA| [a] in "cat", "trap" etc.
*Loss of IPA|/l/ in IPA|/lk/, IPA|/lm/, IPA|/lf/ (see above).
*The pane-pain merger: The words "pane" and "pain" become homophones in most accents.
*The toe-tow merger: The words "toe" and "tow" become homophones in most accents.