The average educated Englishman has no accurate conception of what a dialect really is, beyond a vague notion that the term covers a mass of barbarisms, corruptions, and mispronunciations of the King’s English, devoid of any order or system, and used by the illiterate rustic in a haphazard fashion with no regard to consistency. But as we have already seen in Chapter VII, in very many cases it is the standard language which contains the anomalies and the corruptions, whilst the correct forms have been handed down in the dialects where systematic sound-laws and exact grammatical rules have been regularly developed and carried out unhampered by the arbitrary rules of fashion, or the regulations of a stereotyped spelling 400 years behind the pronunciation. As Max MÜller puts it: ‘the real and natural life of language is in its dialects,’ The Science of Language, vol. i, p. 55.

A dialect may be defined as one of the subordinate forms or varieties of a language arising from local peculiarities of pronunciation, vocabulary, and idiom, or as that form or idiom of a language peculiar to a limited region or people, as distinguished from the literary language of the whole people. Hitherto we have been concerned chiefly with the second of these three characteristics of a dialect, namely vocabulary, but we will now consider in some detail the first on the list, namely pronunciation, and here we cannot fail to be struck by the wonderful uniformity and regularity of the sound-system of modern dialects.

To classify the modern dialects of a country is a difficult and unsatisfactory task. If we possessed about three hundred detailed grammars of the principal English dialects spoken in the United Kingdom, and could find hundreds of competent people willing to answer queries about difficult or doubtful points, it might be possible to furnish a classification which would be tolerably accurate. But this is a state of things never likely to be realized. Though a great deal has been done in collecting material, it is as yet insufficient to enable any one to give the exact geographical area over which many of the grammatical phenomena extend, hence the boundaries given in the classification of our dialects are more or less roughly drawn. For all practical purposes we may divide the English dialects into the following seven groups:

(1) Scottish, including n.Nhb. and n.Cum. Here literary English a has a tendency to become À before a single nasal in such words as can, man. The sound is generally represented in books by o, as con, mon. O.E. Æ (a) in originally open syllables and O.E. a have fallen together, as name, hame (O.E. nama, ham), lit. Eng. name, home. O.E. o in originally open syllables and O.E. a are still kept apart, as kol, hame (O.E. colu, ham), lit. Eng. coal, home. O.E. i and u have not been diphthongized before a following nd as in lit. Eng. O.E. u has become ? [the sound in sun] as in lit. Eng. O.E. u has generally remained, but in s.Sc. it has become ?u [the sound in cow] when final. In Sc. medial d has disappeared after n in such words as cinder, wonder. Final l has generally disappeared after a guttural vowel, as a, fu, lit. Eng. all, full. r is strongly trilled in all positions.

(2) North-country, meaning Nhb. Dur. Cum. Wm. Yks. (except sw. and s.Yks.), and the northern portion of Lancashire. O.E. i has remained before nd, e.g. a word like blind rhymes with lit. Eng. wind sb. O.E. u has generally remained, and also when followed by nd. In words like cup, summer, pound (O.E. pund), the u has the sound of the u in lit. Eng. pull. O.E. u has generally remained as in hus, ut, lit. Eng. house, out. r is uvular in Nhb. and parts of Dur. This is called ‘the Northumberland burr’.

(3) North Midland, meaning sw. and s.Yks., the southern portion of Lan. I.Ma. Chs. n.Wal. Stf. Der. Not. Lin. Rut. Lei. Shr. O.E. a has become e before g in parts of Yks. and Lan. as dreg, reg, lit. Eng. drag, rag. O.E. a(o) has in several of these dialects become u or ? before ng in such words as long, wrong. This pronunciation has been taken over into the standard language in among, -monger, mongrel. O.E. e in originally open syllables, Germanic ? and O.E. ? (= i-umlaut of a) are still kept apart in several dialects, whereas in lit. Eng. they have fallen together, e.g. steal, sleep, heal (O.E. stelan, sl?pan, h?lan beside hal).

(4) South Midland, meaning Nhp. War. Wor. Hrf. Mon. s.Wal. Glo. Oxf. Bck. Bdf. Hrt. Mid. Hnt. O.E. a(o) has become u or ? before ? in long, wrong, &c. O.E. a has become a before sp, ss, st, as in lit. Eng. gasp, grass, fast. Initial shr has become sr, as in srimp, srivel, lit. Eng. shrimp, shrivel.

(5) East-country, meaning Cmb. Nrf. Suf. Ess. O.E. a has become a before sp, ss, st. O.E. y has become e, as pet (O.E. pytt), lit. Eng. pit, but this e is rapidly disappearing through the influence of the standard language. It has been adopted into lit. Eng. in evil, fledge, merry (O.E. yfel, -flycge, myrige). O.E. ? has become i, as mis (O.E. m?s), lit. Eng. mice.

(6) South-country, Ken. Sur. Sus. Brks. O.E. a has become a before sp, ss, st. O.E. Æ(a) in originally closed syllables has become e in parts of Kent, as bek, thet (O.E. bÆc, ÞÆt), lit. Eng. back, that. Initial Þr has become dr, as dri, lit. Eng. three. Initial and medial v has become w in Ken. and e.Sus.

(7) South-west-country, meaning I.W. Hmp. Wil. Dor. Som. Dev. Cor. O.E. Æ(a) has become ? before sp, ss, st. O.E. or in the combination or + consonant has become a in such words as corn, storm. This also occurs in Group 6 above. O.E. i has generally become e before ng or nk, especially in Wil. and Dev. as theng, drenk, lit. Eng. thing, drink. A d has been developed between l—r, r—l, n—r, as palder, madl, tailder, kander, lit. Eng. parlour, marl, tailor, corner. Initial f and s have become v and z in native words in Wil. Dor. Som. Dev. Initial Þ [the sound in thin] has become Ð [the sound in then] in sm. Hmp. I.W. Wil. Dor. Som. Dev. e.Cor. Initial Þr has become dr.

The above are the main distinguishing features of the phonology of the dialects as taken in groups, but no such list can adequately represent the range of pronunciation in the dialects taken individually. The extent of this range can be shown by taking a list of common standard English words, where the number of different ways in which they are known to be pronounced in the various dialects has been carefully counted and registered, e.g. all (20), both (27), chamber (23), close (33), clothes (29), coat (20), cold (31), cow (20), cucumber (35), daughter (36), do (17), done (24), earth (44), father (35), gate (30), good (21), have (24), hold (37), home (44), house (29), night (22), oats (30), old (42), one (21), potato (46), so (24), through (29), whole (33), wrong (22).

The evidence of the pronunciation of words in the different English dialects is of great importance to the student of English philology, as he is thereby often enabled to explain anomalies in the standard language. To take only one instance: philologists have been at a loss to explain why the word oven in lit. Eng. does not rhyme with cloven. The O.E. recorded form is ofen parallel to the past participle clofen, yet while the latter word has followed the normal development, the former has the development not of an original O.E. o, but of u. Now the collected evidence of the dialects goes to show that there must have been beside the recorded O.E. ofen an unrecorded form *ufen from which lit. Eng. oven is quite regularly developed, for the o representing an older u is no more than the old French spelling with which we are familiar in such words as love, come, son, &c.

We can best compare the phonology of the dialects with that of the standard language by examining the vowels and consonants categorically, and noting some of the differences in development. The following is merely a rough outline of the subject, and some of the phonological points noticed in the classification of the dialects will not here be repeated.

Vowels.—(1) a. The sound Æ which is regular in lit. Eng. in close syllables such as back, thatch, is rare in the dialects, occurring chiefly in e. and s.Cy. The majority of the dialects have a in this position. The a in open syllables which has become ei in lit. Eng. as in name, shake, has become e in Sc. n.Cy. and Midl. In s.Sc. and nearly all the other dialects it has become diphthongized to e? or i?, but ai [the sound in time] in Hrt. Lon. Ess. and se.Kent.

(2) e. O.E. e of whatever origin has in close syllables generally had the same development in the dialects as in the standard language, but in many of the s.Sc. e. and sw.Cy. dialects it has become Æ [the sound in hat, man]. O.E. e of whatever origin, has in originally open syllables generally had the same development in Sc. n. and s.Cy. as in the standard language, i.e. it has become i, but in the s.Midl. e.Cy. and sw.Cy. dialects it has mostly become e, and in the other dialects it has generally been diphthongized into ei or i?, the former occurring especially in the w. and s. portions of Yks., in Lan. n.Stf. and Nhp., and the latter in the remaining portions of Yks. Lan. and in Lin. s.Oxf. and w.Wil.

(3) i. This vowel has generally had the same development in the dialects as in the standard language, but in s.Sc. n.Nhb. n.Cum. Der. and w.Som. it has become e. In most Sc. dialects except in the south, it has become a kind of mixed vowel somewhat resembling the e in German Gabe.

(4) u. This vowel has had the same development in Sc. n.Nhb. n.Cum. e. s. and sw.Cy. and in some of the s.Midl. dialects as in the standard language, but in the n.Cy. and many of the n.Midl. dialects O.E. u has generally remained unchanged. In some of the n.Midl. and many of the s.Midl. dialects it has become Ù, a sound formed with the lips more open than for u, and which acoustically resembles an o-sound. It should be noted that those dialects which have ? or Ù, generally also have it in those words where the standard language has u, as in bull, put.

(5) y. This vowel has generally had the same development as in the standard language, but in Ken. e.Sc. and e.An. it has regularly become e, which was a characteristic feature of these dialects already in the M.E. period.

(6) o. In close syllables. This vowel has generally had the same development in the dialects as in the standard language, but in the m.Sc. s.Midl. s. and sw.Cy. dialects there is a tendency to lengthen the vowel in monosyllables, and in some dialects there is also a tendency to change o to a especially before a following p and ft, as shap, tap, craft, lit. Eng. shop, top, croft.

o. In originally open syllables. In the development of this vowel the dialects differ entirely from the standard language. In the southern portions of Yks. and Lan. it has become oi (parallel with the development of e to ei, v. (2) above), but in all the other dialects it has become long close o or has become diphthongized to u? (often written o?). It should be noted that in Lan. ne.Der. and all the dialects north of the Humber the development of O.E. o in open syllables and O.E. a is still kept apart, whereas in all the other parts of England the two sounds have fallen together.

(7) a. In all the dialects north of the Humber this vowel has had the same development as O.E. a, Æ, in open syllables, i.e. it has become e, e?, or i?(ia), whereas in the dialects south of the Humber the regular development is generally the same as for O.E. o in open syllables.

(8) ? (= Germanic ?, W.S. ?, Anglian e). This vowel has generally had the same development in the dialects as in the standard language, i.e. it has become i, but in the southern half of England it has not unfrequently become e or i?, rarely ei, and these diphthongs also occur sporadically as far north as Yorkshire.

(9) ? (= i-umlaut of a). This vowel has generally had the same development as the preceding one, except that the e and i? extend over a much wider area, which shows that many dialects still keep these two sounds apart (?¹ and ?²).

(10) e. This vowel has mostly become i in the dialects just as in the standard language, but ei beside i occurs in nw.Yks. s.Chs. and Lei., and i? beside i in m.Yks. s.Midl. and sw.Cy.

(11) i. O.E. i appears as a diphthong in all the dialects except in those of e. and se. Yks. m. and s.Lan. where we have a. In Sc. and Nhb. it is mostly ei, but ai is also not uncommon, especially in Frf. Per. Lth. and Edb.; n.Cy. ai; in the Midlands, e. and s.Cy. it is generally oi or a diphthong closely resembling oi; and in sw.Cy. ?i, which is approximately the same as in the standard language.

(12) o. The normal development of this vowel is generally Ü or Ö (rarely u or oe, but i in ne.Sc.) in Sc.; u in e.Cy.; u beside oe in sw.Cy.; iu beside i? in n.Cy., but sw.Yks. ui; and u, more rarely iu, in the Midlands; u, in s.Cy.

(13) u. O.E. u has generally remained in Sc. and n.Cy. (but ?u in s.Sc. when final) and n.Lin. It has become a in s. and sw.Yks. and the greater part of Der. and Not.; ? in Lan., ?u in the Midlands, especially in the northern portions, and sw.Cy.; eu in the southern portions of the Midlands, e. and s.Cy. and parts of sw.Cy.

(14) ?. This vowel has generally had the same development as O.E. i, but it has become i in the eastern counties and also in Glo. Bdf. e.Sus. Dev. and Cor.

(15) O.E. e?a. This diphthong has generally had the same development as O.E. ? (= i-umlaut of a).

(16) O.E. e?o has generally had the same development as O.E. e.

Consonants.—(1) The Semi-vowels. (a) w. Initial w has generally remained before vowels, but in parts of Sc. Midl. e.An. and sw.Cy. it has disappeared in certain words, mainly where it stands before a following u, such as woman, wonder, wood, wool, wound, &c. There are no examples in the dialects of initial w being changed to v before a following vowel. This sound-change, characteristic of the language spoken by Mr. Samuel Weller and his father—‘ven’, ‘vay’, ‘svear’, ‘anyveres’, &c.—seems to have been invented by Dickens. The converse, namely, the change of initial v to w, does occur in Bck. Nrf. Suf. Ess. Ken. e.Sus., and Dickens would have heard this pronunciation—wery, very, wenter, venture—used by the class of person typified in Sam Weller, but there is no authority for the change of w to v, and it can only be described as ‘artist’s licence’. An initial w has often arisen in the dialects through a falling diphthong having become a rising diphthong, e.g. in such words as wome, wum, woats, wold, lit. Eng. home, oats, old. This accounts for the w in the place-names Woking, Wokingham, which within living memory were pronounced Oaking, Oakingham, and for the pronunciation of lit. Eng. one, once, and the spelling whole. Initial hw has become f in ne.Sc. in such words as what, wheat, wheel, &c. Initial kw has often become tw in n.Cy. dialects, in such words as twilt, lit. Eng. quilt. A w has often been developed before a back vowel preceded by a consonant, especially a labial, more rarely when preceded by a guttural, dental, nasal, or liquid. This w is chiefly confined to the s.Midl. s. and sw. dialects when the preceding consonant is a labial, as bwone, bwoy, pwoizn, lit. Eng. bone, boy, poison. Medial w has generally disappeared in words compounded with -ward, -worth, as awkward, backward, pennyworth, &c. It has also generally disappeared in always, and in somewhat.

(b) j. This consonant is represented in modern English spelling by y. An initial j has often arisen in the dialects through a falling diphthong having become a rising diphthong, as jabl, jek, ji?r, lit. Eng. able, ache, ear. Many educated people in the south of England make no difference in the pronunciation of ear and year. A s.Midl. s.Cy. saying to express a long period of time is ‘years and years and donkey’s ears’. A medial j has often been developed after a consonant. In many cases the change has been caused by a falling diphthong having become a rising diphthong, e.g. gjardin, kjetl, lit. Eng. garden, kettle.

(2) The Liquids. (a) l. Medial l has often disappeared, especially in the combinations ld, lf, lh, lk, lp, ls, and lt, e.g. in such words as bald, bulk, pulpit, false, bolt. Final l has often disappeared after a guttural vowel, especially in the Sc. Ir. n.Cy. and n.Midl. dialects, e.g. in such words as all, fool, pull, small, wool.

(b) r. In Sc. and the greater part of Irel. and the northern parts of Nhb. and Cum. r has a strong trill. In Nhb. and parts of n.Dur. it is a uvular r, not unlike the French r. It is often called ‘the Northumberland burr’. In all the s. and sw. dialects it is a reverted or retracted r, the trill being indistinct and less sharp than for the Sc. r. Similarly in these dialects the l is reverted. In the rest of England r has had practically the same development as in the standard language. When a word ends in and the next word begins with a vowel, a ‘euphonic’ r is generally inserted to avoid a hiatus, in the s.Midl., eastern, southern, and south-western dialects, as aidi?r ?v it, idea of it, Ser?r An, Sarah Ann, lor ?v I?gl?nd, law of England. And an r is sometimes inserted medially, as drorin, drawing. This insertion of ‘euphonic’ r is not confined to dialect speakers, it is quite common among educated people in the s.Midl. and s. counties, and seems to be spreading gradually further north. r has often undergone metathesis, especially in the sw. dialects in ap?n, t?ild?n, g?rn, h)und?d, p?ti, &c., lit. Eng. apron, children, grin, hundred, pretty, &c.

(3) The Nasals. (a) m. This consonant has generally remained unchanged in all positions except where after consonants it has become vocalic, as in bodm, botm, kindm, &c., lit. Eng. bottom, kingdom, &c.

(b) n. Initial n has remained in nadder (O.E. n?dre), napron (O.Fr. naperon), nauger (M.E. nauger), lit. Eng. adder, apron, auger. In the various dialects there is a large number of words which have an inorganic initial n. It has arisen partly from the n of the indefinite article an, and partly from the n of the possessive pronoun mine; the latter is especially the case in words denoting relationship, as n-oration, a great noise or clamour, n-urchin, a hedgehog, n-awl, n-aunt, n-uncle, cp. ‘Nuncle Lear, nuncle Lear, tarry and take the fool with thee,’ Lear, I. iv. 338. The n in lit. Eng. nickname (M.E. ekename), newt (O.E. efeta) is of this origin. The normal form evet is common in the dialects of southern England. In a few words n has been developed before medial d?; [the final sound in bridge], as porind??(r, a coarse pot or mug used for porridge, sosind??(r, sausage, cp. lit. Eng. messenger, passenger, for messager, passager. In the n. and n.Midl. dialects medial n has disappeared in unaccented syllables as Li?ki??(r, Lincolnshire, Robisn, Robinson, &c. In a few words, mainly in n.Cy. dialects, final n occurs contrary to the usage of the lit. language; these are: aivin (O.E. ifig, ifegn), ivy, holin (O.E. holen, holegn), holly, miln (O.E. mylen), mill, ratn (O.Fr. raton), rat, slon (O.E. slah, sla, plur. slan), sloe.

The guttural ? [the final sound in hang], written n in O.E., only occurred before the gutturals g and c. In stressed syllables medial ?g has become ? in Sc. Irel. n.Cy. n.Midl. and parts of Ken. Sus. and Som., as fi??r, finger, si?l, single, &c. ? has become n before a following dental in lenÞ, length, strenÞ, strength, in Sc. Irel. and n.Cy. The n is also very common in other parts of England, but beside it there exist the forms le?Þ, le?kÞ; stre?Þ, stre?kÞ. The forms with k are often used by educated people in the Midlands. Medial ? in unstressed syllables has generally disappeared, as Bebitn, Bebbington, Notig?m, Nottingham, &c. Final unstressed ? has generally become n in all the dialects, as in evenin(g), farthin(g), mornin(g), sendin(g), and similarly in all present participles and verbal nouns in -ing. In parts of Lan. Chs. Der. when dialect speakers try to talk ‘fine’ they generally substitute ?k for ? in all present participles and verbal nouns in -ing. The same thing can often be heard among educated speakers in those parts.

(4) The Labials. (a) p. This consonant has generally remained in all positions the same as in the standard language.

(b) b. This consonant hardly ever occurs in any of the dialects between m—l or m—r in such words as bramble, thimble, chamber, number. The word marble appears in almost all the dialects as marvl. The form pipl, pebble, occurs in some s. and sw. dialects, cp. O.E. papol- beside M.E. pibble-, pobble.

(c) f. Initial voiceless f has become the voiced spirant v in e.Hrf., parts of Glo., w.Brks. Wil. Dor. Dev. Som. The change must have taken place at a very early period because it is confined almost exclusively to native words, hence it must have taken place before the influx of French words. Three examples of this dialect peculiarity have been incorporated into lit. Eng., viz. vixen, vat, vane (O.E. fyxen, fÆt, fana).

(5) The Dentals. (a) t. The initial combinations tr and str have become tÞr, stÞr, or Þr, sÞr in Irel. Wm. e. and se. Yks. e. em. and s.Lan. I.Ma., as tÞri, Þri, tree, stÞrit, sÞrit, street. Medial t between vowels and vowel-like consonants has become d in the sw. dialects, as bodl, bottle, kedl, kettle; bodm, bottom, occurs also in Sc. and n.Cy. dialects, but this goes back to a form bodan which existed beside botm already in O.E. The t in French words which has become t? [the sound of the medial consonant in nature] in lit. Eng. through the influence of the following Ü has remained unchanged in the dialects, as pikt?(r, picture, fi?t?(r, feature. Final t has disappeared in many dialects after voiceless consonants, especially in the combination st; finally after k and p it has disappeared in all Sc. dialects, as fak(t, korek(t, temp(t. Examples of the loss of t after s occur in all parts of Sc. Irel. and Eng. especially in such words as beast, joist, last, next. In a few instances a t has been added after n, f, or s, as sam?nt, sermon, sud?nt, sudden, vamint, vermin, teligraft, telegraph, aist, ice, naist, nice, w?nst, once, tweist, twice. This excrescent t occurs in certain words in the standard language, e.g. against (M.E. ageines), amidst (M.E. amiddes), behest (O.E. h?s), betwixt (O.E. and M.E. betwix), whilst (M.E. whiles), ancient (Fr. ancien), pheasant (O.Fr. faisan).

(b) d. Intervocalic d followed by r in the next syllable became in the first instance Ð in all dialects, as blaÐ?(r, bladder, konsiÐ?(r, consider, foÐ?(r, fodder, puÐ?(r, powder, &c., in addition to the words which have Ð in the standard language, as father, gather, mother, weather, &c. (O.E. fÆder, gÆdrian, modor, weder, &c.). Examples of the Ð forms begin to appear about the year 1500, but the change has never been consistently carried out in the literary language, whilst in the dialects its operation has been regular. Where exceptions seem to occur they are due either to the influence of the standard language or to the sound-change given below. This Ð from d (O.E. fÆder, &c.) fell together with O.E. Ð in the same position (O.E. feÐer, &c.), and underwent all further changes in common with it. It has thus become (1) d beside dÐ n.Cum. Wm. and parts of Yks. and Lan., (2) d in sn.Sc. n.Cy. and se.Cy. dialects. The words burden (O.E. byrÞen) and murder (O.E. myrÞran) had a spirant already in O.E. The forms with Ð are still very common in Sc. Irel. n.Cy. Lan. Stf. Der. e.An. Medial d very seldom occurs in any of the dialects between n—l or n—r in such words as bundle, candle, gander, thunder, &c. Medial d has regularly disappeared after n in the Sc. dialects except in those of the south, as sin?r, cinder, w?n?r, wonder, &c. Final d has a tendency in all dialects except those of the e. and se. counties to become t in words of more than one syllable, especially after n and r, as bi-jont, beyond, &c. Final d has generally disappeared after n in Sc., but in the southern counties of Sc. it has only disappeared in the conjunction and, the present participles, and in the pret. and pp. of strong verbs whose present ends in -nd. This loss of final d in the pret. and pp. of verbs like bind, find, grind is quite regular in Sc. Irel. and the north and north Midl. counties.

(c) Þ. Initial Þ has generally remained voiceless except in pronouns and the adverbs derived from them, as in the lit. language. The definite article has undergone various changes. It has become (1) t in me.Nhb. Cum. Wm. n. e. nm. sw. and s.Yks. nw.Lan. n.Lin. (2) Þ in m. and se.Lan. wm.Stf. (3) t, Þ sm. and w.Yks. n. em. sw. and s.Lan. Chs. n.Stf. Der. Not. (4) d? Ken. Sus. (5) d, t w.Dur. ne.Yks. (6) d, t, Þ nw. and e.Yks. (7) e Cai. Bnff. In all other dialects it has had the same development as in lit. Eng., viz. Ði?, Ð?. In those dialects which have both t and Þ, the former is used before consonants (tman, &c.), and the latter before vowels (Þapl, &c.), and when the sentence begins with the definite article.

(6) The Sibilants. s. Initial voiceless s has become z in those dialects where f in the same position has become v, cp. (4)(c) above. There is in the dialects a large number of words beginning with s plus a consonant where in most cases the s is not original. It occurs most frequently in the combinations sk and sq. In fact nearly all the sq words occurring in the dialects have forms with and without initial s. No rule can be laid down about the geographical distribution of the words belonging to this category. Examples are: sclasp beside clasp, sclimb beside climb, scrawl beside crawl, scroodle beside croodle, to crouch, skist beside kist, a chest, snotch beside notch, squench beside quench, strample beside trample, &c., &c. Dr. Johnson was familiar with scraunch beside craunch, cp. ‘To Craunch. v.a. [schrantsen, Dutch; whence the vulgar say more properly to scraunch.] To crush in the mouth. The word is used by Swift.’ In Glo. and the s. and sw. counties sp has generally become ps by metathesis, as aps, asp, klaps, clasp, lipsy, to lisp; wÆps and wÆsp existed in O.E., so in the modern dialects there are double forms.

(7) The Gutturals. (a) k. Initial k, generally written c in O.E., has remained before n in such words as knave, knead, knit, knock, &c., in ne.Sc. In the remaining parts of Scotland it has disappeared in the dialect of the younger generation. In the early part of the last century it was preserved in all Sc. dialects. tn from older kn is still used by old people in w.Frf. and e.Per. A generation ago this tn was also common in the dialects of Cum. and Wm., but it is now obsolete. Initial cl has become tl in many of the dialects of Eng. especially in Yks. Lan. the Midlands, and the s. and sw. dialects, in such words as clap, claw, cliff, climb, cloak, cloud. No Sc. or Ir. dialect has changed initial cl to tl. In other respects initial c has generally had the same development in the dialects as in the standard language. Initial sc has become, ? [the initial sound in she] in native Eng. words just as in the lit. language, as shade, shell, ship, &c.; whereas in words of foreign origin it has remained in the dialects just as in the lit. language, as scaffold, scale, scatter, school, skin, &c. Excluding all sc- words which are of various origins and which are common both to the lit. language and the dialects—such as the words in the above list: scaffold, scale, &c.—it is a remarkable fact that the English Dialect Dictionary contains no less than 1,154 simple sc- words. This points to one of two things: either the dialects contain a far larger number of Norse words than is generally supposed, or else it is not certain that initial sc has under all circumstances become ? in native words in the dialects. Words where a final k has become t? in the lit. lang. generally have t? also in the dialects, as bleach, flitch, reach, stitch, &c. But in the dialects of Sc. Irel. n.Cy. and parts of the n.Midlands assibilation has not taken place to the same extent as in the lit. language, hence such forms as skrik, sik, Þak, &c., lit. Eng. screech, such, thatch, &c.

(b) g. Initial g has remained before n in gnat, gnaw in ne. and s.n.Sc., but it has disappeared in the remaining parts of Sc. Irel. and Eng. Initial gl has become dl in many dialects of Eng., especially in Yks. Lan. the Midlands, and the s. and sw. dialects, parallel to the change of cl to tl.

Final g. O.E. geminated g, written cg, has generally become d? [the final sound in sedge] in the dialects in such words as bridge, edge, ridge, &c., but as in the case of the change of final k to t?, in Sc. and the northern parts of Eng. assibilation has not taken place to the same extent as in the lit. language, hence such forms as brig, rig, seg, &c., lit. Eng. bridge, ridge, sedge, &c.

(c) h. Initial h has remained before vowels in Sc. Irel. Nhb. and perhaps also in portions of n.Dur. and n.Cum. In the remaining parts of Eng. it has disappeared, but words originally beginning with a vowel or h often have an h prefixed when the dialect speaker wishes to express a strong emphasis. The emphatic form of it has retained the h in Sc. and Irel. The emphatic form of us is h?z in Sc. and Nhb., the only word in the Sc. dialects containing an inorganic h. Medial and final ? [the final sound in Sc. loch] has generally become f in the dialects of Eng. in those words which have f in the lit. language, as cough, laugh, rough, tough, but f also occurs in many dialects in certain other words besides, as daft?r, slaft?r, Þoft, Þruf, &c., lit. Eng. daughter, slaughter, thought, through, &c.

To turn now from phonology to accidence, we shall find that here, too, system and rule prevail to a surprising extent.