Phonological history of English consonants

Encyclopedia : P : PH : PHO : Phonological history of English consonants

---

Consonant cluster reductions

H-cluster reductions

• The yew-hew merger is a process that causes the cluster [hj] to be reduced to [j].
• The hl-cluster, hr-cluster and hn-cluster reductions are three reductions that occurred in Middle English that caused the loss of [h] from the initial consonant clusters [hl], [hr] and [hn].

Y-cluster reductions

• Yod-dropping is the elision of the sound [j].
• Yod-coalescence changes the clusters [dj], [tj], [sj] and [zj] into [dʒ], [tʃ], [ʃ] and [ʒ] respectively.

Other initial cluster reductions

• The rap-wrap merger is a reduction that causes the historical initial cluster /wr/ to be reduced to /r/.
• The not-knot merger is a reduction that causes the historical initial cluster /kn/ to be reduced to /n/.
• The nome-gnome merger is a reduction that causes the historical initial cluster /gn/ to be reduced to /n/.

Final cluster reductions

• Final consonant cluster reduction is the nonstandard reduction of final consonant clusters in English occurring in African American Vernacular English and Caribbean English.
• The Plum-plumb merger is the reduction of the final cluster /mb/ to /m/.

Phonological history of ng

Ng-coalescence

Ng-coalescence (or the singer-finger split) is the name given to a sound change in the history of English by which word-final [g] was deleted after [ŋ] in words like sing; this sound change happened around the end of the 16th century.

As a result of Ng-coalescence, Middle English [sɪŋg] sing came to be pronounced [sɪŋ]. As well as in word-final position, Ng-coalescence was applied also in cases where a verb ending in -ng was followed by a vowel-initial suffix, so singing and singer also underwent the change. Otherwise, word-internal -ng- did not undergo coalescence and the pronunciation [ŋg] was retained, as in finger and angle. Additionally, in adjectives ending in -ng the [ŋg] is retained when the comparative and superlative suffixes are added, so younger, strongest, etc., do not show coalescence.

As a result of the differing effects of this sound change word-internally, the words finger and singer do not rhyme in most varieties of English, although they did in Middle English.

Some accents, however, do not show the full effects of Ng-coalescence as described above, and in these accents sing may be found with [ŋg], the suffix -ing may be pronounced [ɪŋg], and pairs like singer and finger may rhyme. This is particularly associated with English English accents in an area of northern England and the Midlands, including the cities of Birmingham (see Brummie), Manchester, Liverpool (see Scouse), Sheffield and Stoke on Trent. It is also associated with some American English accents in the New York area. Some of the accents of these areas may be considered to lack the phoneme /ŋ/, as the sound [ŋ] can be thought of as an allophone of /n/ before /g/ or /k/. (Wells 1982)

In some accents of the west of Scotland and Ulster, Ng-coalescence is extended to word-internal position, so that finger is pronounced /fɪŋər/.

G-dropping

G-dropping is a popular name for the substitution of /ɪn/ or /ən/ (spelt -in’, -en) for /ɪŋ/ (spelt -ing) in the English present participle and gerund due to the orthographical changes. Except in dialects which do not show NG coalescence, no sound is actually dropped; a different one is simply used (the alveolar nasal instead of the velar nasal).

The substitution is an old one in English, and derives from the generalisation of one pronunciation to what were two different morphemes in Old English: the present participle -ende and the gerund -inge. The orthography of the merged form, -ing, reflects a derivation from the Old English gerund, but the /ɪn/ pronunciation is also an old one.

It is currently a feature of colloquial and non-standard speech of all regions, and stereotypically of Cockney, Southern American English and African American Vernacular English. Historically, it has also been used by members of the educated upper-class, as reflected by the phrase huntin’, fishin’ and shootin’. That this pronunciation was once regarded as standard can also be seen from old rhymes, as for example, in this couplet from John Gay's 1732 pastoral, Acis and Galatea, set to music by Handel:

Shepherd, what art thou pursuing,
Heedless running to thy ruin?

Which was presumably pronounced "shepherd, what art thou pursuin', heedless runnin' to thy ruin" although this would sound very odd in an opera today. Such a rhyme would today be appropriate only in a comic context.

In the poetry of Jonathan Swift (1667-1745), participles consistently rhyme with words in [ɪn]:

But Weston has a new-cast gown
On Sundays to be fine in,
And, if she can but win a crown,
'Twill just new dye the lining.

The pronunciation with [ɪŋ] only became standard in the nineteenth century.

The use of a colloquial pronunciation which derives from a different word from the standard is not uncommon; ’em or em, a colloquial form of them, derives from Old English him of the same meaning, whereas them was a borrowing from Old Norse þeim.

H-dropping and h-adding

H-dropping

H-dropping is a colloquial term used to describe the omission of initial /h/ in words like house, heat, and hangover in many dialects of English, such as Cockney and Estuary English. The practice is generally stigmatised. The same phenomenon occurs in many other languages, such as Serbian, and Late Latin, the ancestor of the modern Romance languages. Interestingly, both French and Spanish acquired new initial [h] in mediæval times, but these were later lost in both languages in a "second round" of h-dropping. Many dialects of Dutch also feature h-dropping, particularly the south western variants.

H-dropping in English is found in all dialects in the weak forms of function words like he, him, her, his, had, and have; and, in most dialects, in all forms of the pronoun it — the older form hit survives as the strong form in a few dialects such as Southern American English and Scots. Because the /h/ of unstressed have is usually dropped, the word is usually pronounced /əv/ in phrases like should have, would have, and could have. As a result, the contractions should've, would've, and could've are often misspelt as should of, could of, and would of. This, in turn, has led to many people in the South of England erroneously pronouncing these phrases unambiguously as "should of", etc.

H-adding

The opposite of h-dropping, so-called h-adding, is a hypercorrection found in typically h-dropping accents of English. A classist stereotype, commonly found in literature from late Victorian times to the early 20^th century, holds that some lower-class people consistently drop h in words that should have it, while adding h to words that should not have it. An example from the musical My Fair Lady is, "In 'Ertford, 'Ereford, and 'Ampshire, 'urricanes 'ardly hever 'appen". In practice, however, it would appear that h-adding is more of a stylistic prosodic effect, being found on some words receiving particular emphasis, regardless of whether those words are h-initial or vowel-initial in the standard language.

Words borrowed from French frequently begin with the letter h but not with the sound /h/. Examples include hour, heir, hors d'oeuvre and honest. In some cases, spelling pronunciation has introduced the sound /h/ into such words, as in humble, hotel and (for most speakers nowadays) historic. Spelling pronunciation has also added /h/ to the Commonwealth pronunciation of herb, /hɜːb/, while American English retains the older pronunciation /ɝb/.

Elimination of velar fricatives in English

Taut-taught merger

The taut-taught merger is a process that occurs in modern English that causes /x/ to be dropped in words like thought, night, daughter etc. http://web.archive.org/web/20040703025114/http://www.uni-mainz.de/FB/Philologie-II/fb1413/roesel/seminar0203/regional_varieties/Scotland.htm*http://216.239.51.104/search?q=cache:9zbJpPgRmfMJ:www.phon.ucl.ac.uk/home/wells/English%2520accents_6.ppthttp://www1.uni-hamburg.de/peter.siemund/Articles/English%2520(Variationstypologie).pdf

The phoneme /x/ was previously distinguished as [ç] after front vowels, [x] after back vowels. [ç] and sometimes [x] was lost in most dialects with compensatory lengthening of the previous vowels. /nɪxt/ [nɪçt] > /niːt/, later > /naɪt/ "night" by the Great Vowel Shift.

/x/ sometimes became /f/, with shortening of previous vowel.

Inconsistent development of [x] combined with ambiguity of ou (either /ou/ or /uː/ in Early Middle English) produced a multi-way result for words spelled ough. Compare Modern English through /θruː/, though /ðoʊ/, bough /baʊ/, cough /kɒf/ or /kɑf/, rough /rʌf/.

The spelling gh that occurs in current orthography usually indicates that in Middle English the word in question had a velar fricative /x/ in the pronunciation. During the 16th century, this fricative was still pronounced, in some circles at least; by the seventeenth century it had disappeared in the precursor of RP, except that in some cases /x/ had been switch instead to a labiodental, /f/. Hence the spelling gh is now either silent, as in right, high, eight, bough, dough, daughter, through, or else corresponds to /f/, as in cough, rough, laugh, tough etc. The resulting multiplicity of pronunciations corresponding to orthographical ough is notorious.

In traditional-dialect of the north of England and Scotland, /x/ may remain in many of the words in which it was found in Middle English. Where it does not remain as such, it often happens that it yielded /f/ in items where standard English has zero consonant, or vice versa. Hence on the one hand we have the stereotype Scots phrase bricht moonlicht nicht (for bright moonlight night; compare standard /braIt/ etc.), and on the other hand occasional relic forms such as [trQx] for trough (RP /trQf/) and [lICt] for light'', recorded by SED in southern West Yorkshire (at localities y30 ans y21 respectively). In both cases the local accent (as opposed to traditional-dialect) lacks /x/.

Quite apart from traditional dialect, a fair number of names in the Celtic countries contain /x/ in the local pronunciation. This is the case with Buchan and Auchtermuchty in Scotland, Amlwch /-lUx/ and Loughor /lUx@/ in Wales. A more English, less Celtic pronunciation commonly involves the replacement of this /x/ by /k/: English people and other outsiders call Buchan /bVk@n/. In Ireland /h/ is common corresponding to putative earlier /x/, as in Donaghee, Haughey, though some speakers do have a /x/ there. Thus there is a sense in which the consonant system even in Standard English, as pronounced in the Celtic countries, includes /x/, whether as a residualism going back to Middle English /x/ or as a loan phoneme from Scottish Gaelic, Welsh or Irish.

Even in England /x/ can be said to hold a tenuous and marginal position in the consonant system of educated speakers, though certainly no longer found in Standard English in words which contained it in Middle English. Here it is clearly a loan-phoneme.

Many people know of the word loch and its Scottish guttural sound. The composer Bach is often called [bA:x], with a vowel reflecting the German short [a] only poorly but a consonant which is more successful. Some English people, too, will attempt an [x] in surnames of Celtic origin such as MacLachlan. The hyperforeignism [mju:nIx] for Munich is very common; [xl] is also familiar as one of the would-be sophisticated attempts at Welsh [K], as in Llangollen.

A velar or uvular fricative has also been taken into South African English along with Afrikaans words containing it; as gogga /xQx@/.

Some accents in northern England show slightly different changes, for example night as /niːt/ and in the dialectal words owt and nowt (from aught and naught). Also, in Northern England, the distinction between wait and weight is often preserved, so those speakers lack the wait-weight merger.

Wait-weight merger

The wait-weight merger is the merger of the Middle English sound sequences /ɛi/ (as in wait) and /ɛix/ (as in weight) that occurs in most dialects of English. (Wells 1982: 192–94, 337, 357, 384–85, 498)

The main exceptions are in Northern England, for example in many Yorkshire accents, where these sequences are often kept distinct, so that wait /weːt/ is distinct from weight /wɛɪt/ and late /leːt/ does not rhyme with eight [ɛɪt].

The distinction between wait and weight is an old one that goes back to a diphthongization of Middle English /ɛ/ before the fricative /x/ which was represented by gh in English. So in words like weight /ɛ/ became /ɛɪ/ and subsequently /x/ was lost as in Standard English, but the diphthong remained.

Wait on the other hand is a Norman French loan word (which in turn was a Germanic loan) and had the Middle English diphthong [ai] that was also found in words like day. This diphthong merged with the reflex of Middle English /aː/ (as in late) and both ended up as /eː/ in the accents of parts of northern England, hence the distinction wait /weːt/ vs. weight /wɛɪt/.

Lock-loch merger

The lock-loch merger is a phonemic merger of /k/ and /x/ that is starting to occur in some Scottish English dialects, making lock and loch homonyms as /lɔk/. Many other varieties of English have borrowed foreign and Scottish /x/ as /k/, and so not all people who pronounce "lock" and "loch" alike exhibit the merger.http://www.arts.gla.ac.uk/SESLL/EngLang/research/accent/annex4.htmhttp://www.essex.ac.uk/linguistics/archive/viewconf2000/abstracts.html

The English spoken in Scotland has traditionally been known for having an extra consonant sound /x/, but that is starting to disappear among some younger speakers in Glasgow.

The observations of the merger happening were investigated by auditory and acoustic analysis for a sample of the children from Glasgow pronouncing words that traditionally have /x/ in Scottish English.

Elimination of dental fricatives in English dialects

Th-fronting

Th-fronting is a merger that occurs (historically independently) in Cockney, Newfoundland English, African American Vernacular English, and Liberian English (though the details differ among those accents), by which Early Modern English /θ, ð/ merge with /f, v/. (Wells 1982: 96–97, 328–30, 498, 500, 553, 557–58, 635)

Apparently, no accents with the merger completely merge the phonemes, because virtually all speakers of such accents know which words "should" have which sound; moreover, in many accents the two sounds appear in free variation. Where TH fronting is applied, pairs such as three/free, slither/sliver, and oath/oaf are homophonous. Th-fronting of the voiced consonant [ð] does not occur at the beginning of a word, but only intervocalically and at the end of a word.

The use of the labiodental fricatives [f] and [v] for the dental fricatives [θ] and [ð] is a well known feature of the proverbial Cockney. It has recently been noted as spreading through non-standard accents in England (cf. Trudgill 1988, 43). TH fronting is one of the boundaries between Estuary English and Cockney.

Although TH fronting 'pops up' occasionally in the middle and upper (middle) class English accents as well, there is still a marked social difference between working and middle class speakers. TH fronting can therefore serve as a 'boundary marker' between Cockney and Estuary English.

TH fronting is still a feature of Cockney which is extremely rare in the other social English accents. TH fronting can (still) serve as a 'boundary marker' between Estuary English and Cockney. Furthermore, TH fronting is currently making its way into the middle class English accent and thus into Estuary English.

Here's a sample of a speaker of the Cockney accent that has th-fronting:

• http://www.gazzaro.it/accents/sound/Cockney.mp3

My dad came from Wapping and me mum came from Poplar. Me dad was one of eleven kids… and Wapping in them days really was one of the poorest parts of London. I mean they really didn't have shoes on their feet. I'm talking about seventy years ago now. Erm… and Poplar was… sli… just slightly a cut above Wapping; erm… you was either East End respectable or you was sort of East End villain, you know, and my family was respectable on both sides. But me father had a very tough time because his father died when he was nineteen, leaving him the only one working to bring up eleven brothers… ten brothers and sisters and on a Thursday night he'd sometimes go home and the youngest two would be crying in the corner and he'd say “What's the matter with them, ma?” “Oh, well, Harry, you know it's Thursday night, and you don't get paid till tomorrow.” and they literally didn't have any food in the house.

In that recording father and brother are pronounced [ˈfɑːvə] and [ˈbrʌvə]; Thursday is pronounced [ˈfɜːzdi].

Th-stopping

Th-stopping is the realization of the dental fricatives [θ, ð] as stops, which occurs in several dialects of English. In some accents, such as Hiberno-English, some varieties of Newfoundland English, some varieties of New York-New Jersey English, and Indian English, they are realized as the dental stops [t̪, d̪] and as such do not merge with the alveolar stops /t, d/. Thus pairs like tin/thin and den/then are not homophonous (Wells 1982: 428–29, 498, 500, 515). In other accents, such as Caribbean English and Liberian English, such pairs are merged (Wells 1982: 565–66, 635).

For some New Yorkers, the fricatives /θ/] and are pronounced as affricatives or stops, rather than as fricatives. Usually they remain dental, so that the oppositions /t-θ/ and [d-ð] are not lost. Thus thanks may be pronounced [θæŋks], [tθæŋks] or [t̪æŋks] in decreasing order of statusfulness, all are distinct from tanks. The [t̪] variant has a weakish articulation. the /t-θ/ opposition may be lost, exceptionally in the environment of a following /ɹ/ (making three homophonous with tree), and in the case of the word with, (so that with a may rhyme with the nonrhotic pronunciation of "bitter-bidder"; with you may be [wɪʧu], following the same yod-coalescence rule as hit you. These pronunciation are all stigmatized.

The [d-ð] opposition seems to be lost more readily, though not as readily as the Brooklynese stereotype might lead one to believe. As in many other places, initial [ð] is subject to assimilation or deletion in a range of environments in relatively informal and/or popular speech, e.g. who's there [huz (z)ɛə]; as in many other places, it is also subject to stopping there /dɛə/. This option extends to one or two words in which the /ð/ is not initial, e.g. other, which can thus become a homonym of utter-udder. But it would not be usual for southern to be pronounced identically with sudden or breathe with breed.

In African American Vernacular English, in the words with and nothing, [t] may occur corresponding to standard [ð], thus [wɪt] for with and [ˈnʌtɪn] for nothing. (Wolfram 1969:83) Th-stopping is also reported for some other non-initial [θ]s, apparantly particularly when preceded by a nasal and followed by a plosive, as keep your mout closed (Wolfram 1969:90). In initial position, [θ] occurs in AAVE just as in standard accents: thin is [θɪn], without the stopping of West Indian accents. (Wolfram 1969:130, does however mention the use of 'a lenis [t]' as a rare variant.) Stopping of initial [ð], however, is frequent making then homophonous with den.

Th-weakening

Th-weakening is a process that occurs in dialects with th-fronting where initial [ð] lenites to zero, [n], [l], [d] or the like. As a result, "this" is pronounced [ɪs], [nɪs], [lɪs], [dɪs] or the like. http://www.hi.is/~peturk/KENNSLA/87/VARS/THFronting.html

Wine-whine merger

The wine-whine merger is a merger by which the sound " />