Old English phonology

This article includes a list of general references, but it lacks sufficient corresponding inline citations. Please help to improve this article by introducing more precise citations. (July 2010) (Learn how and when to remove this message)

Part of a series on
Old English
Dialects Kentish Mercian Northumbrian West Saxon
Use Orthography (Runic alphabet, Latin alphabet) Grammar Phonology Phonological history
Literature Beowulf Anglo-Saxon Chronicle Cædmon's Hymn
History Development of Old English Influences Proto-Germanic Latin Norse Brittonic
Legacy Middle English Early Modern English Modern English Scots
v t e

Old English phonology is the pronunciation system of Old English, the Germanic language spoken on Great Britain from around 450 to 1150 and attested in a body of written texts from the 7th–12th centuries. Although its reconstruction is necessarily somewhat speculative, features of Old English pronunciation have been inferred partly from the sounds used in modern varieties of English (including dialects), partly from the spellings used in Old English literature, partly from analysis of Old English poetry, and partly from comparison with other Germanic languages.

Old English included multiple dialects: the dialect called West Saxon is the best documented in surviving texts, and so is commonly treated as a default reference in descriptions of Old English, even though it is not a direct ancestor of the modern English language (which is more closely related to the Mercian dialect).^[1]

Old English had a distinction between short and long (doubled) consonants, at least between vowels (as seen in sunne "sun" and sunu "son", stellan "to put" and stelan "to steal"), and a distinction between short vowels and long vowels in stressed syllables. It had a larger number of vowel qualities in stressed syllables (/i y u e o æ ɑ/ and in some dialects /ø/) than in unstressed ones (/ɑ e u/). It had diphthongs that no longer exist in Modern English (such as /eo̯ æɑ̯/), with both short and long versions.

The inventory of consonant surface sounds (whether allophones or phonemes) of Old English is shown below. Allophones are enclosed in parentheses.

The fricative phonemes /f θ s/ had voiced and voiceless allophones. The voiced allophones [v ð z] were used when the fricative was surrounded on both sides by voiced sounds (between vowels, between a vowel and a voiced consonant, or between voiced consonants) and immediately preceded by a syllable with some degree of stress^[2][3] (e.g. eorðe ('earth') /ˈeo̯rθe/ [ˈeo̯rðe], fæþm ('fathom') /ˈfæθm/ [ˈfæðm]). The voiceless allophones [f θ s] were used next to voiceless consonants, at the beginning and end of words, after unstressed syllables, and at the start of the second elements of compound words.^[3]

The allophones [f θ s] and [v ð z] alternated in contexts such as the following:

• ċealf ('calf, young cow') /tʃæɑ̯lf/ [tʃæɑ̯ɫf] vs. ċealfru ('calves') /ˈtʃæɑ̯lfru/ [ˈtʃæɑ̯ɫvru]
• smiþ ('blacksmith') /smiθ/ [smiθ] vs. smiþas ('blacksmiths') /ˈsmiθɑs/ [ˈsmiðɑs]
• hūs ('house' noun) /ˈxuːs/ [ˈhuːs] vs. hūsian ('to house') /ˈxuːsiɑn/ [ˈhuːziɑn]

As a rule, the distinction between the voiceless and voiced versions of these sounds was unmarked in Old English spelling: /f/, whether voiceless [f] or voiced [v], was written as ⟨f⟩; /θ/, whether voiceless [θ] or unvoiced [ð], was written as ⟨ð⟩ or ⟨þ⟩, and /s/, whether voiceless [s] or voiced [z], was written as ⟨s⟩. Certain alternative spellings existed for some sounds (e.g. [v] was sometimes written ⟨u⟩, as in Latin^[4]) but for the most part, the presence and distribution of the allophones described above is inferred from the sounds' origins and later development. However, there are some uncertainties and disputes about how to interpret the evidence. There may have been some exceptions to the distribution of voiceless and voiced fricative sounds according to these rules: some scholars have argued that the contrast had already become phonemic (if marginally so) in Old English^[5] whereas Minkova 2011, citing the absence of minimal pairs, argues that they were not lexically contrastive segments and so should be analyzed as allophones during Old English, even if their distribution was not determined solely by phonology.

Potential exceptions include words where the fricative originally stood after an unstressed vowel, but the vowel was deleted by syncope, such as strengþu^[6] or hālsian 'to take an oath': the devoicing of the preceding plosive consonant in the verb blētsian (derived from the root of blōd 'blood' with the same suffix) suggests that fricatives were at least initially voiceless in this context, although the outcomes of some words such as anfilt(e), anfealt > modern English anvil and sīþe > modern English scythe suggest that voicing took place at least sometimes after syncope.^[7]

This voicing pattern did not apply to the fricatives /x/ (spelled ⟨h⟩) or /ʃ/ (spelled ⟨sc⟩, often written ⟨sċ⟩ in modern editions).

• In contexts where other fricatives became voiced, Proto-Germanic *x came to be lost entirely in Old English, though before it was lost it caused certain sound changes such as breaking of preceding vowels.^[8] Old English did possess a voiced velar fricative sound [ɣ], which developed from Proto-Germanic *ɡ, but this is usually analyzed as a separate phoneme from /x/: the sounds were normally distinguished in spelling, with [ɣ] written as ⟨g⟩ and /x/ as ⟨h⟩, although some unetymological interchange of these spellings occurs, especially in word-final position (where the sounds seem to have merged into one phoneme by late Old English). The fricative [ɣ] seems to have instead been phonemically identified in Old English with the plosive [ɡ], also written ⟨g⟩.
• The fricative /ʃ/ developed later than other fricative sounds,^[9] as it evolved from the West Germanic cluster /sk/.^[10] It is likely that sċ was pronounced as geminate [ʃː] between vowels, and possibly also at the end of a word after a short vowel.^[2] In Old English poetry, ⟨sc⟩ between vowels seems to have been treated metrically like a cluster rather than like a single consonant.^[10]

As described above, word-initial fricatives are standardly reconstructed as voiceless, as they are for the vast majority of modern English speakers in inherited Germanic vocabulary. However, some dialects of Middle English used voiced fricatives in word-initial position, and this is also attested in some continental Germanic languages, such as Dutch and High German, although not in Frisian. Lass 1991–1993 argues that even if unambiguous written evidence of word-initial [v ð z] is only found in Middle English, their development was likely related in some manner to the voicing of word-initial fricatives in other Germanic languages, since voicing in this position seems to be a cross-linguistically 'unnatural' sound change, and therefore unlikely to develop multiple times independently. As as result, Lass considers it probable that initial ⟨f þ~ð s⟩ might have been pronounced as [v ð z] in at least some accents of Old English (suggesting this would occur in West Saxon, Kentish, and West Mercian, but not in East Mercian or Northumbrian).^[11]

In the case of the phoneme /f/, the alternation between voiceless [f] and voiced [v] originated from more complex sound changes than just voicing of original [f] between voiced sounds. In medial or final position, Old English /f/ was derived not only from Proto-Germanic *f, but also from Proto-Germanic *[β], a fricative allophone of *b. Proto-Germanic *b developed to Old English /b/ only at the start of a word, after a nasal, or when geminated. In other contexts, it became Old English /f/, pronounced either as [v] or [f] based on its position (the originally voiced fricative was devoiced before voiceless sounds or in final position):^[12]

• PG *stabaz [ˈstɑβɑz] > OE stæf /ˈstæf/
• PG *habdē > OE hæfde [ˈhævde] 'had', but PG *habjaną > OE habban [ˈhɑbbɑn] 'to have'

In contrast, the Old English phonemes /θ/ and /s/ generally descend only from Proto-Germanic voiceless *θ and *s. Proto-Germanic *[ð] (a fricative allophone of *d, sometimes derived by voicing of *θ in the context of Verner's Law) regularly developed in all positions into the Old English stop /d/, as in fæder^[13] /ˈfæder/ from Proto-Germanic *fadēr [ˈɸɑðɛːr]. Proto-Germanic *z (which existed only as the Verner's Law counterpart of *s) regularly developed to Old English /r/ (a sound change called rhotacism). As a result, some Old English verbs show alternations between /θ/ [θ~ð] and /d/ or between /s/ [s~z] and /r/, although in others this alternation was leveled, resulting in /θ/ [θ~ð] or /s/ [s~z] throughout.^[14]

Examples of Old English verbs that retained inherited /θ/–/d/ or /s/–/r/ alternations:

• snīþan, snāþ, snidon, sniden^[13][14] /ˈsniːθɑn, ˈsnɑːθ, ˈsnidon, ˈsniden/ [ˈsniːðɑn, ˈsnɑːθ, ˈsnidon, ˈsniden] from Proto-Germanic: *snīθaną, *snaiθ, *snidun, *snidanaz 'cut'
• frēosan, frēas, fruron, froren^[14] /ˈfreːo̯sɑn, ˈfræːɑ̯s, ˈfruron, ˈfroren/ [ˈfreːo̯zɑn, ˈfræːɑ̯s, ˈfruron, ˈfroren] from Proto-Germanic: *freusaną, *fraus, *fruzun, *fruzanaz 'freeze'

Examples of Old English verbs that leveled the consonant to only /θ/ or only /s/:

• wrīþan, wrāþ, wriþon, wriþen^[14] /ˈwriːθɑn, ˈwrɑːθ, ˈwriθon, ˈwriθen/ [ˈwriːðɑn, ˈwrɑːθ, ˈwriðon, ˈwriðen], versus Proto-Germanic: *wrīθaną, *wraiθ, *wridun, *wridanaz 'wrap, twist'
• lesan, læs, lǣson, lesen^[15] /ˈlesɑn, ˈlæs, ˈlæːson, ˈlesen/ [ˈlezɑn, ˈlæs, ˈlæːzon, ˈlezen], versus Proto-Germanic: *lesaną, *las, *lēzun, *lezanaz 'gather'

Old English possessed palatal consonants [tʃ, dʒ, j, ʃ], but they were represented in Old English spelling with the same letters as velar consonants or clusters [k, ɡ, ɣ, sk]:

• ⟨c⟩ represented both palatal [tʃ] and velar [k]
• ⟨g⟩ represented both palatal [j] and velar [ɣ], as well as palatal [dʒ] and velar [ɡ]
• ⟨sc⟩ represented both palatal [ʃ] and velar [sk]

Modern editors may mark the palatalized consonants with a dot above the letter: ⟨ċ⟩, ⟨ġ⟩, ⟨sċ⟩.^[2] Historically, [tʃ, ʃ, dʒ] developed from [k, sk, ɡ] by palatalization.^[16] Some cases of [j] developed from palatalization of [ɣ], while others developed from Proto-Germanic *j. Those historical sound changes resulted in certain common patterns to where the palatal sounds occurred: for example, ⟨c⟩ and ⟨g⟩ often represent palatalized sounds when followed by the front vowels ⟨i⟩, ⟨e⟩, ⟨æ⟩. (Note that Old English had palatalized ⟨g⟩ in certain words that have hard G in Modern English because of Old Norse influence such as ġiefan "give" and ġeat "gate".)

Although palatalization was originally a regular sound change, later vowel changes and borrowings meant that the occurrence of the palatal forms was no longer predictable. Thus, the palatals and the velars ultimately became separate phonemes. But there is some debate about when the contrast became phonemic, as well as about when the palatal counterparts of [k ɡ] evolved to affricates [tʃ dʒ] as opposed to palatal plosives [c ɟ].^[17] The unetymological use of the letter ⟨c⟩ in the spellings ⟨orcgeard⟩ and ⟨feccan⟩, found around 900 AD for older ortġeard and fetian, is commonly interpreted as evidence that palatal ċ had become [tʃ], and so could be used to represent an affricate that is presumed to have developed in these words via a change of [tj] to [tʃ].^[18] The velar and palatal sounds came to be in surface contrast before the vowel ⟨æ⟩ after the sound change of i-umlaut.^[19]

However, because word-initial palatal ċ and velar c alliterate in English poetry up through at least the late tenth century, Minkova 2014 assumes that they were still allophones of a single phoneme before 1000.^[10] Likewise, word-initial palatal ġ and velar g alliterate with each other in early Old English verse (before the latter changed to [ɡ], circa 950 AD^[20]), which Minkova 2014 interprets as evidence that [j] and [ɣ] constituted allophones at this point in time,^[21] despite the existence of /j/ from Proto-Germanic.^[20] Lass 1994 assumes that [j], [ɣ] and [ɡ] were all allophones of a phoneme /ɡ/ at one point during the history of Old English.^[9]

Palatalized ⟨sċ⟩, according to Minkova 2014, may have still been pronounced as a cluster [sc] rather than as a unitary consonant [ʃ] in some dialects at the end of Old English.^[10]

The affricate [dʒ] was found only in restricted contexts: it did not occur at the start of a word, and occurred medially or finally only after a nasal or in contexts where it was (at least originally) geminated. It was therefore in nearly complementary distribution with [j].^{[citation needed]} However, phonetic [nj] occurred as the result of syncope in some words such as menġan.^[2] The transcription in this article ignores such exceptional cases and treats [dʒ] as an allophone of /j/.

• senġan ('to sing') /ˈsenjan/, pronounced [ˈsendʒɑn] (from *sangijan)
• bryċġ ('bridge') /bryjj/, pronounced [bryddʒ] (from *bruggjō < *bruɣjō)

In circumstances where the palatal affricates [tʃ] and [dʒ] came to be followed by another consonant due to syncope of an intervening vowel, they were eventually replaced with the corresponding velar plosives, [k] and [ɡ] respectively. (Ringe & Taylor 2014 assume this replacement occurred before the palatalized variants had developed into affricates.^[22] Campbell 1959 assumes that such consonants were never affricated, but transcribes them as palatal in Old English.^[23]) The affricates do seem to have been used before other consonants in compound words, e.g. in bryċġ-bōt 'bridge-repairing' and seċġ-lēac 'sedge-leek, rush-garlic'.^[24]

There may also have been a palatal fricative [ç], but only as a contextual allophone of /x/ (see below).

The voiceless velar plosive [k] was typically spelled ⟨c⟩. It alternated in some circumstances with the voiceless palatal affricate [tʃ], also spelled ⟨c⟩, as described above.

The voiced velar fricative [ɣ] was in mostly complementary distribution with the voiced velar plosive [ɡ]: in early Old English, the latter did not occur at the start of a word, and occurred medially or finally only after a nasal or as a geminate (the latter was rare, since West Germanic gemination caused palatalization).

However, phonetic [nɣ] occurred as the result of syncope in some words such as syngian.^[2] The transcription in this article ignores such exceptional cases and treats [ɡ] as an allophone of /ɣ/. In late Old English, [ɡ] came to be used in initial position as well,^[25] and for this time period the underlying phoneme can be analyzed as /ɡ/, with [ɣ] an allophone used after a vowel.^{[citation needed]}

In late Old English, [ɣ] was devoiced to /x/ at the ends of words. That and the palatalization referred to above made the phonemes /ɣ/, /j/, and /x/ alternate in the inflectional paradigms of some words.

• dæġ ('day') /ˈdæj/, dæġes (GEN.SG) /ˈdæjes/ vs. dagas (NOM.PL) /ˈdɑɣɑs/ (also dagung ('dawn') [ˈdɑɣuŋɡ])
• burg, burh ('castle') /burɣ/ > /burx/, vs. burgum (DAT.PL) /ˈburɣum/, vs. byrġ (NOM.PL) /byrj/

The letter ⟨h⟩ represented a single consonant phoneme that is generally assumed to have had the following allophones:^{[26][27][28][2]}

• a voiceless glottal fricative [h], used at the start of a word^[26][2] (or at the start of a morpheme in compound or prefixed words).
• a voiceless velar fricative [x], used in most other positions.^[2]
• a voiceless palatal fricative [ç], used after front vowels^[28] (or possibly only after stressed front vowels^[2]). The use of the value [ç] in this position is supported by developments in English pronunciation seen from the thirteenth century onward: original /x/ sometimes became /f/ after a back vowel (e.g. rough, tough, trough), but this change is never seen after a front vowel. That is explained if it is assumed that the allophone [x] sometimes became [f], but the allophone [ç] never did so.^[29] Lass 1994 considers it uncertain that [ç] was used already in Old English, whereas Hogg 1992 considers it certain that ⟨h⟩ had developed a palatal version, like other velar consonants.^[26]

The underlying phoneme can be analyzed as /x/,^[26] at least in early Old English:^[28]

• hund ('dog') /xund/, phonetically [hund]
• cniht ('boy') /knixt/, phonetically [kniçt]

[ŋ] is an allophone of /n/ occurring before [k] and [ɡ].^[30] Words that have final /ŋ/ in standard Modern English have the cluster [ŋɡ] in Old English.

• sincan ('sink') /ˈsinkɑn/ > [ˈsiŋkɑn]
• hring ('ring') /xrinɡ/ > [r̥iŋɡ]

The exact nature of Old English /r/ is not known.^[31] It may have been an alveolar approximant [ɹ], as in most Modern English accents; an alveolar flap [ɾ]; or an alveolar trill [r].

The spellings ⟨hw⟩, ⟨hl⟩, ⟨hn⟩, ⟨hr⟩ probably represented two-phoneme clusters, /xw, xl, xn, xr/, where /x/ was pronounced [h] (its usual allophone in syllable-initial position). In this context, /w, l, n, r/ may have been pronounced as voiceless sonorants^[32] [ʍ, l̥, n̥, r̥]. The status of ⟨hw⟩, ⟨hl⟩, ⟨hn⟩, ⟨hr⟩ as clusters rather than unitary segments in Old English phonology is supported by their alliteration in poetry with each other and with prevocalic [h]^[33] /x/.

There is an alternative hypothesis that holds that (at least in later periods) ⟨h⟩ in these sequences was not pronounced as an independent consonant sound, but was only a diacritic marking the voicelessness of the following sonorant.^[34] Original /xl, xn, xr/ would merge with plain /l n r/ by early Middle English.^[35] The merger of /xr/ and /r/ seems to have been completed earliest, by the middle of the eleventh century, based on frequent interchange of the spellings ⟨hr⟩ and ⟨r⟩ in glosses from that time period.^[36] The merger of /xn/ and /n/ was probably complete by the start of the thirteenth century.^[35] The spellings ⟨hr⟩ and ⟨hn⟩ had both fallen out of use by around 1250.^[35] The merger of /xl/ and /l/ may have taken somewhat longer to complete, as the spelling ⟨hl⟩ and an alternative spelling ⟨lh⟩ are attested in some Middle English texts,^[37] one of the latest being the Ayenbite of Inwyt (written in 1340 by a Kentish English speaker who was probably born during the thirteenth century).^[38] At least some of these mergers may have begun earlier. Old English scribes occasionally omitted the letter ⟨h⟩ in words starting with these clusters.^[34] A merge of the cluster /xw/ with /w/ is also attested in some historical and many current varieties of English, but has still not been completed, as some present-day speakers distinguish the former as [ʍ]. There is evidence of alliteration between ⟨hw⟩ and ⟨w⟩ in some Old English poems.^[39]

/l r/ apparently had velarized allophones [ɫ] and [rˠ] or similar sounds when they were followed by another consonant or were geminated. That is suggested by the vowel shifts of breaking and retraction before /l r/, which could be cases of assimilation to a following velar consonant:

• *lirnian > liornian > leornian [ˈleo̯rˠniɑn] ('learn')
• *erþǣ > eorþǣ > eorþe [ˈeo̯rˠðe] ('earth')
• *fællan > feallan [ˈfæɑ̯ɫɫɑn] ('to fall')

Based on phonotactic constraints on initial clusters, Fisiak 1967 proposed interpreting ⟨wr⟩ and ⟨wl⟩ as digraphs representing the velarized sounds in prevocalic position, in which case the distinction would be phonemic, as exhibited by minimal pairs such as wrīdan [ˈrˠiːdɑn] "to grow" vs. rīdan [ˈriːdɑn] "to ride" or wlītan [ˈɫiːtɑn] "to look" vs. lītan [ˈliːtɑn] "to bend".^[40] However, this hypothesis is inconsistent with orthoepic and orthographic evidence from the Early Modern English era,^[41] as well as borrowings into and from Welsh, which has [wl] and [wr] as genuine initial clusters. Furthermore, in Old English poetry, ⟨wr⟩ and ⟨wl⟩ can alliterate with each other as well as with ⟨w⟩ followed by a vowel, as in "Wēn' ic þæt gē for wlenco, nalles for wræcsīðum" (Beowulf 338).^[42]

Old English had a moderately large vowel system. In stressed syllables both monophthongs and diphthongs had short and long versions, which were clearly distinguished in pronunciation. In unstressed syllables, vowels were reduced or elided though not as much as in Modern English.

Old English had seven or eight vowel qualities, depending on dialect, and each could appear as a long or a short monophthong. An example of a pair of words that are distinguished by vowel length is god [god] ('god') and gōd [goːd] ('good').

The front mid rounded vowel /ø(ː)/ merged into /e eː/ in the West Saxon dialect,^[43] but can be found in some other dialects such as Anglian.

The long–short vowel pair /æ æː/ developed into the Middle English vowels /a ɛː/, with two different vowel qualities distinguished by height, and so they may have had different qualities in Old English as well.^[44]

The short open back vowel /ɑ/ before nasals was probably rounded to [ɒ], as is suggested by the fact that the word for "person", for example, is spelled as mann or monn.^[44]

In unstressed syllables, only three vowels (/ɑ e u/) were distinguished.^[45] /æ, e, i/ were reduced to /e/; /ɑ, o/ were reduced to /ɑ/, and /u/ remained. Unstressed /e, u/ were sometimes pronounced or spelled as [i, o] in closed syllables, as in hāliġ and heofon.^[46]

All dialects of Old English had diphthongs. Diphthongs were written with digraphs composed of two vowel letters and pronounced by gliding from one vowel quality to another within a single syllable. The two main spellings used to represent diphthongs were ⟨ea⟩ and ⟨eo⟩.^[47][48]

• ⟨ea⟩ was pronounced like [æɑ], gliding from the sound of the vowel ⟨æ⟩ to ⟨a⟩.^[49][50] (It is possible that the spelling ⟨ea⟩ was used in place of ⟨æa⟩ because the latter would have essentially been a trigraph ⟨aea⟩.^[51][52])
• ⟨eo⟩ was pronounced like [eo], gliding from the sound of the vowel ⟨e⟩ to ⟨o⟩.^[49][53]

Diphthongs could be short or long.^[54] A short diphthong had the same length as a short single vowel, and a long diphthong had the same length as a long single vowel.^[55] As with monophthongs, length was not marked in writing in Old English manuscripts, but is inferred from other evidence, such as a word's etymological origins or the pronunciation of its descendants. Modern editions conventionally mark long diphthongs with a macron on the first letter: e.g. long ⟨ēa⟩, ⟨ēo⟩ in contrast to short ⟨ea⟩, ⟨eo⟩.^[56] In this article, short diphthongs such as ea are transcribed like [æɑ̯], and long diphthongs such as ēa are transcribed like [æːɑ̯] (with a length marker after the first letter).

Some dialects had additional diphthongs:

• io, īo were found in Anglian dialects of Old English, but had merged with eo, ēo in West Saxon. The quality of ⟨io⟩ is hypothesized to have been [iu].
• ie, īe were found only in Early West Saxon.^[57] The quality of ⟨ie⟩ is disputed: some scholars reconstruct it as [iy], but others argue it was [ie] or [iə]. In later West Saxon, it merged with ⟨y⟩ or ⟨i⟩.

Thus, the inventory of diphthongs in Anglian dialects was as follows:^[58]

The table above displays how Old English vowel digraphs are commonly interpreted, but there are various debates about the pronunciation and phonemic analysis of these spellings.

The phonetic realization of Old English diphthongs is controversial.^{[59][60][61][47]}

During the 20th century, various academic articles^[62] disputed the reconstruction of "short diphthongs", arguing that they were actually monophthongs (on the phonetic level, the phonemic level, or both). However, in response to these proposals, further arguments have been made in support of the proposition that short digraphs did in fact represent phonetic diphthongs.^[63][64][65] Hogg 1992 argues that a contrast between long and short diphthongs is not necessarily phonologically implausible, noting it is attested in some modern languages, such as Scots, where the short diphthong in tide /təid/ contrasts with the long diphthong in tied /taid/.^[66] In contrast, Minkova 2014 considers the evidence for the phonemic status of short diphthongs to be unconvincing and prefers to analyze short ⟨ea⟩, ⟨eo⟩ as allophones of /æ, e/, or at most, as semi-contrastive entities that never became completely distinct phonemes from the corresponding short monophthongs.^[60]

Assuming vowel digraphs were in fact pronounced as phonetic diphthongs, they may have been the 'falling' type^[67] (where the first portion of the diphthong was more prominent, and the second part was a non-syllabic offglide.^[52] Alternatively, both components may have been more or less equal in prominence.^[68]

The primary feature that distinguished ⟨ea⟩ from ⟨eo⟩ seems to have been the height of the first component of the diphthong: the start of ⟨ea⟩ sounded like ⟨æ⟩ /æ/ whereas the start of ⟨eo⟩ sounded like ⟨e⟩ /e/.^[52] The second component of any diphthong (whether original or from breaking) seems to have originally been high back rounded [u] (or [u̯]).^[69][70] Diphthongs seem to have still ended in this quality at the time when i-umlaut occurred.^[71] Fulk 2014 assumes the qualities [æu̯ eu̯ iu̯ æːu̯ eːu̯ iːu̯] continued to be used into Old English for ea eo io ēa ēo īo respectively, but acknowledges that their values may have been different in late Old English.^[61] Ringe & Taylor 2014 assume that by the 9th century, the second component of ea had become lowered and unrounded (aside from in the minority of regions where the alternative spelling ⟨eo⟩ was used for this diphthong).^[69] Both components of [æɑ̯] are low vowels and both components of [eo̯] are mid vowels. Lass & Anderson 1975 propose that Old English diphthongs were height-harmonic as a rule, that is, that both parts of any diphthong had the same vowel height (high, mid or low).^[72][73] The reconstruction of io as [iu] and early West Saxon ie as [iy] is consistent with this principle of height harmony.^[74][75] However, Ringe & Taylor 2014 do not find height harmony convincing as a general rule, arguing that the later development of ie īe points instead to the value [iə̯ iːə̯].^[76] Some other scholars have reconstructed ⟨ea⟩ and ⟨eo⟩ as ending in an unrounded schwa-like glide in Old English.^[a] However, there is evidence that Old English eo io ēo īo had rounded outcomes in some dialects of Middle English.^[61]

Old English diphthongs have several origins. Long diphthongs developed partly from the Proto-Germanic diphthongs *au, *eu, *iu and partly from Old English vowel shifts, while short diphthongs developed only from Old English vowel shifts. Here are examples of diphthongs inherited from Proto-Germanic:

• PG *dauþuz > OE dēaþ 'death'
• PG *deuzą > OE dēor 'animal' > Modern English deer
• PG *biumi > Anglian bīom, West Saxon bēo '[I] am'

Three vowel shifts produced diphthongs: breaking, back mutation, and palatal diphthongization. Breaking caused Anglo-Frisian short *æ, *e, *i to develop into the short diphthongs ea, eo, io before /x, w/ or a consonant cluster beginning with /r, l/. Anglo-Frisian long *ǣ, *ī developed into the diphthongs ēa and īo before /x/:

• PG *liznōjaną > Anglo-Frisian *lirnian > Anglian liornian, West Saxon leornian 'learn'
• PG *nāh > AF *nǣh > Old English nēah 'near'

Back mutation changed i, e and sometimes a to io, eo and ea before a back vowel in the next syllable:

• PG *sebun > AF *sefon > OE seofon 'seven'
• PG *niþanē > OE neoþan 'from beneath'

Palatal diphthongization changed æ, ǣ, e, ē to the diphthongs ea, ēa, ie, īe respectively after the palatalized consonants ġ, sċ, and ċ:

• PG *gebaną > AF *jefan > ġiefan 'give'

In addition, the back vowels a, o, u (long or short) could be spelled as ⟨ea⟩, ⟨eo⟩, ⟨eo⟩ respectively after ċ, ġ, or sċ. However, rather than indicating the development of a diphthong, these spellings might have just been a convention for marking palatal consonants before a back vowel,^[77] since the modern English descendants of such words do not display the typical evolution of the diphthong ⟨eo⟩ to a front vowel:

• PG *jungaz > OE ġeong 'young' (Modern English /jʌŋ/)
• PG *skuldē > sċeolde 'should' (Modern English /ʃʊd/)

Peter Schrijver has theorized that Old English breaking developed from language contact with Celtic languages. He says that two Celtic languages were spoken in Britain, Highland British Celtic, which was phonologically influenced by British Latin and developed into Welsh, Cornish and Breton, and Lowland British Celtic, which was brought to Ireland at the time of the Roman conquest of Britain and became Old Irish. Lowland British Celtic had velarization like Old and Modern Irish, which gives preceding vowels a back offglide. That feature came by language contact to Old English and resulted in backing diphthongs.^[78]

Early West Saxon ie, īe developed from i-mutation of ea, ēa or io, īo (at the time of i-mutation, the merger of the latter with eo, ēo seems to have not yet occurred).^[79] By the time of Alfred the Great, the diphthongs ie, īe were apparently monophthongized to a vowel known as "unstable i", whose pronunciation is still uncertain. It later went on to merge with /y yː/ according to spellings such as gelyfan for earlier geliefan and gelifan ('to believe').^[80] (According to another interpretation, however, the "unstable i" may simply have been /i/, and the later /y/ can be explained by the fact that Late West Saxon was not a direct descendant of Early West Saxon. See Old English dialects.) That produced additional instances of /y(ː)/ alongside those that developed from i-mutation and from sporadic rounding of /i(ː)/ in certain circumstances (e.g. myċel 'much' from earlier miċel with rounding perhaps triggered by the rounded /m/). All instances of /y(ː)/ were normally unrounded next to ⟨c⟩, ⟨g⟩ and ⟨h⟩: hence ġifan from earlier ġiefan 'to give'.

In dialects other than West Saxon, i-mutation instead turned ea, ēa into e, ē and left io, īo unchanged.^[79]

As in modern English, there was a distinction in Old English between stressed and unstressed syllables. Stress typically could be found only on the leftmost syllable of a root morpheme, which coincided in morphologically simple words with the leftmost syllable of the word: e.g. yfel 'evil',^[81] pronounced [ˈyvel]. Non-initial syllables within a morpheme were unstressed.

Inflectional suffixes are inferred to have been fully unstressed, based on the absence of alliteration involving these syllables^[82] (although in words with multiple unstressed syllables in a row, such as fremedon [ˈfremedon] 'they did', it is possible that there was some kind of alternating rhythm).^[81] Fully unstressed syllables did not contain long vowels or diphthongs.

When a simple word was extended by a derivational suffix, or when two roots capable of standing as free words were combined to form a compound, the primary stress fell on the first syllable of the leftmost root. However, there may have been secondary stress in some circumstances on the first syllable of the later element.^[81] In Old English verse, the first root of a compound participates in alliteration, whereas the second root of a compound can be involved in alliteration only as a supporting element, if it starts with the same consonant as the first root.^[83] Derivational suffixes and the second elements of compound words appear to display a wider range of vowel contrasts than inflectional suffixes: for example, a diphthong can be seen in the second syllable of the word spelled ⟨arleas⟩^[84] 'honorless' derived from the morphemes ār 'honor' and lēas 'devoid of, bereft of' (as a suffix, '-less'). Since vowel length was not written in Old English, it is less clear to what extent long vowels may have been shortened, or conversely, analogically restored, in such derivational suffixes.^[84]

In the case of prefixed words, the primary stress did not always fall on the first syllable of the word, but could instead fall on the first syllable of the root after the prefix.^[81] The prefixes ġe- and be- were always fully unstressed, and the prefix for- was nearly always unstressed.^[b] In contrast, the prefixes and- and ed- always received primary stress. Other prefixes seem to have generally received primary stress in nouns or adjectives, but not in verbs or adverbs.^[86]

Phonotactics is the study of the sequences of phonemes that occur in languages and the sound structures that they form. It usually represents consonants in general with the letter C and vowels with the letter V and so a syllable such as 'be' is described as having CV structure. The IPA symbol that shows a division between syllables is the dot [.].

Old English stressed syllables were structured as (C)³V(C)⁴; that is, one vowel as the nucleus with zero to three consonant phonemes in the onset and zero to four consonant phonemes in the coda. An example of a stressed syllable with the minimal number of phonemes would be ǣ 'law, statute', whereas an example of a stressed syllable with nearly the maximum number of phonemes^{[citation needed]} would be bringst (the syncopated second-person singular present form of the strong verb bringan 'bring'^[87]).

Onset clusters typically consist of a obstruent /p, b, t, d, k, ɣ, s, ʃ, f, θ, x/ followed by a sonorant /m, n, r, l, w/, although /s/ is allowed as a third element before voiceless stops, and /w/ is allowed before /r, l/. The consonants /j, tʃ/ occur only on their own. (If /n̥, r̥, l̥, ʍ, rˠ, ɫ/ are accepted as their own phonemes, the same can be said of these consonants and of /x/, but these are normally analyzed respectively as /xn, xr, xl, xw, wr, wl/.) Some have proposed analyzing clusters of /s/ and a voiceless stop as single segments.^[88][89] In Old English alliterative poetry, a word-initial sequence of /s/ + voiceless stop alliterates only with itself^[21] (with or without a following liquid): that is, ⟨st⟩ and ⟨str⟩ count as a match, as do ⟨sp⟩ and ⟨spl⟩, but ⟨st⟩ and /sp/ do not alliterate with each other.

Unpalatalized /sk/ and /skr/ did not occur as a rule at the start of a word, since in inherited vocabulary, original */sk/ came to be palatalized in this position regardless of what sound followed it. However, they are hypothesized to have been retained as onsets word-medially in certain contexts: e.g. /skr/ probably occurred in malscrung (based on the forms of the related Middle English malskren and Modern English masker^[90]). Kuhn 1970 assumes that /sk/ was found at the start of the word scolere, from Latin scholārius, but Campbell 1959 transcribes it as sċolere;^[91] the form sċrift from Latin scrīpt- shows that palatalized sċ- could come to be used at the start of Old English words taken from Latin.^[10]

The onset was optional, so syllables could start with a vowel phoneme. In Old English poetry, stressed syllables starting with vowel phonemes all alliterate with each other (regardless of whether the vowels are the same or different), which suggests that phonetically, a glottal stop consonant [ʔ] could be inserted in this position.^[92]

The syllable nucleus was always a vowel in stressed syllables.^[93] Stressed monosyllabic words always ended in either a consonant or a long vowel (whether a long monophthong or long diphthong): this can be stated in terms of stressed words having at least two moras of length.^[55] In words of two or more syllables, it was possible for the stressed syllable to end in a short vowel (called a light syllable), although two-syllable words more often had a heavy first syllable (one that ended in a consonant or long vowel).^[94]

It is possible that certain sonorant consonants, such as /n/ or /l/, could serve as the nucleus of an unstressed syllable.^[95][55] However, it is difficult to determine whether or in which contexts consonants were syllabic in Old English, because the relevant forms show variable spelling (a vowel letter, presumably representing an epenthetic vowel sound, could often be inserted before the sonorant)^[55] and variable behavior in verse.^[96]

In general, Old English permitted similar kinds of clusters of coda consonants as modern English.^[97] Most coda clusters in simple words started with a sonorant or /s/.^[98]

Long (geminate) consonants seem to have become simplified to single consonants when not between vowels. However, [dʒ] (analyzed above as long /jj/) did not merge with single /j/ in this context, but remained a distinct coda.

The following tables show some examples of coda clusters that could occur in Old English, while not necessarily constituting an exhaustive list.

†It is assumed that geminate consonants such as /ll/, /nn/, /mm/ were simplified by the Old English period to single consonants when entirely in a syllable coda.

‡The final /l/ in words ending in /nl/, /ml/, /wl/ could potentially become syllabic or have an epenthetic vowel inserted before it; see below. This possibly could apply also to the final /n/ in /ln/.

Although /j/ might be categorized as a resonant, it had non-resonant allophones, and so will be listed alongside obstruent consonants in the tables below.

Some codas with an obstruent preceded by more than one resonant are attested, often as the result of syncope, e.g:

• /rnθ/, as in irnþ
• /rmθ/, /rms/, as in yrmþ, wyrms
• /rld/, as in weorld (a sycopated variant form of weorold)
• /lmd/, /lmθ/, as in cwylmd, cwylmð (syncopated variant forms of cwylmed and cwylmeð, the past participle and third person present singular of cwylman = cwielman)

The following additional two-obstruent coda clusters may rarely occur:

• /fs/ was normally replaced with /ps/, but there are some variant spellings with final ⟨fs⟩, such as wæfs, which may represent the original pronunciation.^[104]
• /xs/ was normally replaced with /ks/,^[105] but some variant spellings with final ⟨hs⟩ can be found, such as þreahs for þreax.
• /sp/, as in ġeresp, hosp, hyspte (past of hyspan). This could become /ps/ by metathesis, as in crisp~cyrps 'curly' from Latin crispus.
• /sk/, as in frosc. Uncommon compared to sċ /ʃ/, and often varies with ⟨x⟩ /ks/ (as in frox) by metathesis.
• /fd/, /jd/, as in ġenæfd, ġehyġd
• /tʃt/ might occur at the end of some Kentish verb forms spelled with ⟨ct⟩, namely ofðreċt and ġehyðlǣċt.^[106] (Campbell 1959 marks these with ⟨ċt⟩ but assumes ⟨ċ⟩ was phonetically a palatal stop, rather than affricate, when followed by another consonant.^[107] Hogg & Fulk 2011, p. 219 transcribe them with dotless ⟨ct⟩, assuming that preconsonantal ⟨ċ⟩ had already been replaced with velar ⟨c⟩ in Old English.)
• /θs/ was apparently regularly assimilated to /ss/ (which in turn was degeminated in syllable-final position to /s/), as in the nominative forms of the nouns bliss and liss, which would etymologically be blīþs and līþs.^[108] Some variant spellings can be found that do not show this simplification, but it is unclear whether the consonant was ever restored in pronunciation.^[109]

Additional possible three-obstruent clusters include:

• /ksθ/, as in wixð (third-person singular present form of wexan, smoothed variant of weaxan), if not simply a morphological spelling for /kst/.

Because of the loss of certain vowels in final syllables, Proto-West-Germanic came to have words ending in sequences of an obstruent consonant followed by a resonant consonant:^[110] for example, Proto-Germanic *xlaxtraz developed to Proto-West-Germanic *xlaxtr.^[111] In the past, it was sometimes assumed that a resonant consonant in such a position must necessarily be syllabic. This assumption is false: there are languages where a syllable can end in an obstruent followed by a resonant, as demonstrated by modern Icelandic, where vatn, býsn, segl, gísl are all monosyllables.^[112] There is evidence that this type of coda cluster eventually became disallowed in Old English, because many such words show a spelling with a vowel letter inserted before the consonant, such as hleahtor. However, some words could be spelled with or without an inserted vowel letter in Old English, raising the question of whether there was also variation between different pronunciations. Based on the treatment of such words in poetry, Fulk 1989 argues that their pronunciation changed either during or shortly before the time period when Old English literature was written: when not etymologically preceded by a vowel, resonant consonants in this position were generally nonsyllabic in early Old English verse, whereas in late Old English verse, they came to be syllabic (or preceded by an epenthetic vowel). Fulk finds that the syllabic pronunciations are generally used consistently in poetry from the ninth century or later.^[113] The development of a syllabic pronunciation seems to have been affected by the identity of the resonant, the identity of the consonant preceding the resonant, and the weight of the syllable.

• Word-final sequences of a consonant + ⟨r⟩ had come to be spelled practically always with an epenthetic vowel letter in Old English, e.g. æcer, fōdor, fæġer, although in early Old English poetry these epenthetic vowels do not always count as syllables.^[114]
• Word-final sequences of a consonant + ⟨l⟩ could be spelled with or without an epenthetic vowel depending on the identity of the consonant. An inserted vowel letter is never seen in ⟨rl⟩, and usually not in ⟨tl⟩, ⟨dl⟩, ⟨þl⟩, ⟨sl⟩, ⟨fl⟩, ⟨ġl⟩, but is seen often in other combinations, e.g. ⟨pl⟩, ⟨bl⟩, ⟨gl⟩, ⟨wl⟩, ⟨nl⟩.^[115] In poetry, ⟨l⟩ never scans as its own syllable after /d t f s/.^[d][116]
• Word-final sequences of a consonant + ⟨n⟩ show considerable variability. Spellings with an epenthetic vowel seem to be generally uncommon in words ending with a short vowel + single consonant + ⟨n⟩, e.g. swefn, reġn, stemn, although a few spellings with inserted ⟨e⟩ are attested early on (namely efen and ofen) and others are attested in late West Saxon.^[117] In cases where word-final ⟨n⟩ is preceded by more than one consonant, or by a syllable containing a long vowel or diphthong, the inclusion of an epenthetic vowel letter is variable, possibly affected by dialect.^[118]
• Word-final sequences of a consonant + ⟨m⟩ never scan as a separate syllable in poetry,^[116] and are never spelled with an epenthetic vowel letter in early West Saxon: e.g. ⟨māþm⟩, ⟨fæþm⟩, ⟨bōsm⟩, ⟨botm⟩, ⟨wæstm⟩, ⟨breahtm⟩. In texts from other time periods or dialects, spellings with ⟨um⟩ (e.g. late West Saxon ⟨māþum⟩, ⟨bōsum⟩) or ⟨em⟩ (e.g. Mercian ⟨westem⟩) can be found except for in words ending with ⟨rm⟩ or ⟨lm⟩.^[119]

Like Frisian, Old English underwent palatalization of the velar consonants /k ɣ/ and fronting of the open vowel /ɑ ɑː/ to /æ æː/ in certain cases. Old English also underwent vowel shifts that were not shared with Old Frisian: smoothing, diphthong height harmonization and breaking. Diphthong height harmonization and breaking resulted in the unique Old English diphthongs io, ie, eo, ea.

Palatalization yielded some Modern English word pairs in which one word has a velar and the other has a palatal or postalveolar. Some of these were inherited from Old English (drink and drench, day and dawn), and others have an unpalatalized form loaned from Old Norse (skirt and shirt).

Old English had four major dialect groups: Kentish, West Saxon, Mercian and Northumbrian. Kentish and West Saxon were the dialects spoken south of a line approximately following the course of the River Thames: Kentish in the easternmost portion of that area and West Saxon everywhere else. Mercian was spoken in the middle part of England and was separated from the southern dialects by the Thames and from Northumbrian by the River Humber. Mercian and Northumbrian are often grouped together as "Anglian". The largest differences occurred between West Saxon and the other groups and occurred mostly in the front vowels, particularly the diphthongs.

The early history of Kentish was similar to Anglian, but sometime around the 9th century, all front vowels^[ambiguous] æ, e, y (long and short) merged into e (long and short). This discussion concerns the differences between Anglian and West Saxon, with the understanding that Kentish, other than where noted, can be derived from Anglian by front vowel merger. The primary differences were the following:

• Original (after Anglo-Frisian brightening) ǣ was raised to ē in Anglian but remained in West Saxon. That occurred before other changes such as breaking and did not affect ǣ caused by i-umlaut of ā: dǣlan ('to divide') < *dailijan appears the same in both dialects, but West Saxon slǣpan ('to sleep') appears as slēpan in Anglian. (Note the corresponding vowel difference in the spelling of "deal" < dǣlan vs. "sleep" < Anglian slēpan.)
• The West Saxon vowels ie/īe were caused by i-umlaut of long and short ea, eo, io and did not appear in Anglian. Instead, i-umlaut of ea and rare eo are spelled e, and i-umlaut of io remains io.
• Breaking of short /æ/ to ea did not happen in Anglian before /l/ followed by a consonant; instead, the vowel was retracted to /ɑ/. When mutated by i-umlaut, it appears again as æ (vs. West Saxon ie): Anglian cald ('cold') vs. West Saxon ċeald.
• The merger of eo and io (long and short) occurred early in West Saxon but much later in Anglian.
• Many instances of diphthongs in Anglian, including the majority of those caused by breaking, were turned back into monophthongs again by the process of "Anglian smoothing", which occurred before c, h, g, alone or preceded by r or l. That accounts for some of the most noticeable differences between standard (West Saxon) Old English and Modern English spelling: ēage ('eye') became ēge in Anglian; nēah ('near') became Anglian nēh and was later raised to nīh in the transition to Middle English by the raising of ē before h (hence nigh in Modern English); nēahst ('nearest') become Anglian nēhst, shortened to nehst in late Old English by vowel-shortening before three consonants (hence next in Modern English).

Modern English derives mostly from the Anglian dialect, rather than the standard West Saxon dialect of Old English. However, since London sits on the Thames, near the boundary of the Anglian, West Saxon and Kentish dialects, some West Saxon and Kentish forms have entered Modern English. For example, bury has its spelling derived from West Saxon and its pronunciation^[ambiguous] from Kentish (see below).

All dialects of Old English seem to have shared palatalization as a sound change, including Northumbrian.^[120] Forms in Modern English with hard /k/ and /ɡ/ in which a palatalized sound would be expected from Old English appear to be influenced by Scandinavian.

The prologue to Beowulf:

Fæder ūre

Recording of the Lord's Prayer in reconstructed Old English

---

Problems playing this file? See media help.

The Lord's Prayer:

Wikimedia Commons has media related to Anglo-Saxon pronunciation.