Dialect, in popular language, a local form of speech which differs from the general or received speech of the country, but is manifestly related to it. Still more popularly, dialect is considered to be a 'corruption' of received speech, due to ignorance or carelessness, and is stigmatised as boorish. Both definitions are untenable. The last is so extremely wrong that it would be more correct to term received speech an artificial (as distinct from a natural or organic) 'corruption' of some local form of speech which political or literary circumstances have imposed upon the whole country. The first definition labours under the disadvantage of supposing that there was from the first in any country a received form of speech which 'broke up' into dialects, just as the Latin language actually broke up into the Romance languages of to-day. And this leads to the question, How does a dialect differ from a language? Are the so-called Romance languages not rather dialects of Latin? Is not English collectively a dialect of Low German? Was not Latin itself a dialect of that original Aryan speech which we only know in such dialectal forms as Sanskrit, Greek, Latin, Gothic, &c.? This is a question merely of the meaning we assign to words. It is well fitted for arguing, as Gil Blas argued (chap. i.), but it is of no practical value.
Without going into the very vexed question of the origin of language, we find that wherever two human beings are thrown together a spoken means of communication or language grows up; that wherever a family is isolated a family language is generated; that when the family separates into many distinct families they carry with them some general and various particular forms of the original family speech; that where these families again separate the same process goes on, till the one original speech is broken up into a variety of local forms which differ but slightly from each other, and are usually, though strange, mutually intelligible. Wars, especially civil wars, conquests, defeats, changes of environment due to immigration and emigration, contact with others deriving from a different original family, produce harsher changes, till the groups become mutually unintelligible. It is at this point that the old dialects may be considered to become languages. We have not to go far to seek examples. The slang of each nursery, school, college, profession, and handicraft illustrates the first; the history of Europe furnishes abundant examples of the second. We may therefore have dialects or diversities of related speech without having a received speech at all. In the flourishing time of Greece, Herodotus wrote in Ionic and Thucydides in Attic. There was not even a word for Greek, which, as separate from its dialectal forms, was unknown till long afterwards. In England the author of the first English-Latin dictionary, the Promptorium Parvulorum, apologises for writing in East Anglian, as being the only form of speech he knew. It was not till political domination called forth a court speech that the seed of a received speech was sown; but centuries of writers, each touching up the artificial changeling in his own way while using it to express vigorous thought, were required to build up our glorious English speech.
Dialect, however, is by no means yet extinct, even in England, though railways and school-boards threaten its speedy annihilation. The writer, who has been led to investigate the subject during the last twenty years, finds that we have still six great forms of local speech—Southern, Western, Eastern, Midland, Northern, and Lowland Scotch—further divisible into forty-two districts, each of which has generally numerous varieties. His object was to determine the pronunciation as at present existing all over that part of England and Scotland which is not still Celtic; hence his distinctions are mainly based on pronunciation as referred to the oldest English form of literary speech, the West Saxon or Wessex, which, however, never prevailed over the whole country, and is not even the foundation of modern literary speech. Of this pronunciation he was obliged to omit one striking and characteristic part—viz. the intonation or singsong of speech, which generally strikes a stranger first and most strongly, because he could not invent any practical means of expressing it, and it was very seldom indeed that he could hear it from native uneducated peasants. The two other points to which we must look for distinctions are vocabulary and construction. Vocabulary is the point most generally considered, but it is often delusive as a means of separation, because it is so difficult to trace the area over which certain words prevail as distinct from those where they are not known. The nature of the investigations and their difficulty is well shown in Mr Thomas Hallam's 'Four Dialect Words, Clem, Lake, Nesh, and Oss [= to starve, to play, tender, to offer], their modern dialectal range, meaning, pronunciation, etymology, and early or literary use' (English Dialect Society, 1885). With unusual words, however, or words with unusual significations, most glossarists consider that they are chiefly concerned, while at the same time the difficulty the writers feel in expressing the sounds of the words—i.e. properly speaking, in actually conveying the words—renders their identification uncertain. This is different for literary languages where there is a conventional orthography and pronunciation. The last characteristic mark of dialects is their grammatical construction. It is surprising how little this point is attended to in most glossaries, so that the reader is generally left to discover the construction from the examples, and these in many cases are not sufficiently trustworthy for that purpose. It is the more satisfactory, then, to refer to such glossaries as Elworthy's West Somerset, with its separate grammar; Darlington's Folk-speech of South Cheshire; Miss Jackson's Shropshire Word-book; and Robinson's Mid-Yorkshire Glossary, in which pronunciation and grammar are well attended to, and examples are numerous and trustworthy.
In making the above division of English dialects, considered as new local forms of speech referred to one old local form, the Wessex, without any regard to received speech or spelling, the following are the principal distinctions relied upon, and the areas over which they extend.
(1) Southern takes in the whole south of England from Cornwall to Kent, extending northwards on the west to the north of Worcester, south of Warwick and Northampton, eastwards to the boundary of Oxford, and all south of the Thames. The marked characters are found in the western half, which die off eastwards, while westwards they exhibit signs of Celtic influence. Wilts, Dorset, Somerset, and Gloucester show this form in its prime. These were the seat of the West Saxon tribes, and best preserve the mark of their origin.
The main character of all the southern division is the pronunciation of r, especially where not before a vowel, as in there, sir! The tongue, in the principal region, is reverted, that is, turned with its underside to the hard palate, and its point directed towards the throat. This is the English form of the (different) glottal r of the Low German district. It is not only strong in Wilts, Dorset, and Somerset, but also in Devon (where French u may likewise be heard), though it dies out through Cornwall, and in west Cornwall, which has comparatively recently become English, has become that of received speech. In Hampshire and eastwards it becomes milder, but on the whole it is still well marked even in Kent. It is the parent of received English r, which is little different from an imperfect d when it does not become a simple vowel or is lost. Initial f and s in Saxon words become v and z. In construction the speech is peculiar in the general use of I be for I am, and we am, you am, for we are, you are; in the general prefix of a to the past participle—as I've a-done; the periphrastic form of the present tense, as I do go for simple I go, without any implied emphasis; the use of the accusative en, from Wessex hine, which is replaced by the dative him in literary speech; and in one part of Somerset (near to, but west of, Yeovil) the use of uteh for the first personal pronoun I. In the east of Sussex and in Kent de, dis, dat, &c. may be still heard for the, this, that, &c., but this is a comparative modernism.
(2) Western comprises most of Monmouth, Hereford, and Shropshire, and shows the influence of the Welsh, which in early times prevailed over them; and Shropshire especially preserves a trilled Welsh r which is in remarkable contrast to the southern and midland r.
(3) Eastern extends over the whole of the counties lying north of the Thames but east of Oxford and Leicester and south of Lincoln, embracing also the East Anglian peninsula of Norfolk and Suffolk, which two counties generally have a pronunciation of to, do, soon, as if the vowel were nearly French u, much the same as in Devonshire. But this does not extend to the rest of the eastern, which bears a remarkable resemblance to received speech, being indeed that part of the country from which it originated. Essex is especially remarkable for pronouncing paper like piper, a practice which of late years (subsequently to the time of Dickens and Thackeray) has crept into London, and may be heard from most newsboys. Over Kent, Essex, Suffolk, and Norfolk extends the land of Wee, where w is used for v, but not conversely, as wine for vine, wan (rhyming man) for van, and so on.
(4) Midland occupies the whole middle region of England from the north of the three last named to, roughly speaking, a line drawn from Lancaster Bay on the west to the Humber on the east. The speech of the large tract of country thus included is by no means homogeneous. But the Southerner is at once surprised by the difference in the pronunciation of short o and u in come, up, and similar words, which in the greater part of this region resembles the German ö or French eu with a dash of oo in foot, being really a transitional form between the well-known southern and Lowland Scotch sound and the oo in foot. This last sound, however, prevails in Lincolnshire and the south of Yorkshire. Lincolnshire is remarkable for its fractured or divided vowels, which Lord Tennyson writes with ä, but which really sound to a Londoner, and even to a native Lincolnshire man, both of whom vocalise their r, as if an r were annexed, as sair, sairnts, deerd, doarut, for Lord Tennyson's saiy, saints, deid, doint (= say, saints, dead, don't), and so on. In the extreme north of Lincolnshire ow is pronounced oo, as noo, eoo, hoose, for now, eow, house, which is not found at all in the other midland counties, but is an old form now characteristic of northern speech. In the western midland, mid and south Lancashire, Cheshire, Derbyshire, one of the most striking pronunciations arises from the treatment of ow, which is pronounced ou in the south only, at the south-east of Lancashire and north-west of Derby, becoming the prolongation of a in bat, cat in the south-west of Lancashire, of a in father in the south of Yorkshire, and south Derby, and like eye nearly in Cheshire and north Staffordshire. The word for she is hoo in south Lancashire, north Derby, and Cheshire, and shoo in south Yorkshire. The definite article south of Cheshire, Derby, and Nottingham has the usual English form, but in Cheshire, north Derby, and Lancashire becomes the th in oath without any vowel, though often assimilated to a preceding t, and in south Yorkshire is reduced to a simple t' without any vowel. In construction the most remarkable feature is the verbal plural in -en in the present tense, often singularly contracted. Thus an yo? is have-n you? dun they (with the peculiar u) is do-en, 'do they?' they known, they know-en, 'they know,' and so on. This prevails in Lancashire, Cheshire, Derbyshire, and the parts of Yorkshire adjoining, but is lost in most of south Yorkshire. It has disappeared recently in Nottinghamshire, is rare in the south midland, and practically disappears in Leicester. I am is used in this region, in contrast to the southern I be. The r is very light and quite different from the southern reverted r and the Scotch strongly trilled r.
(5) Northern extends from the northern boundary of the midland to Scotland. Here ow becomes oo completely in most of Yorkshire, but remains in the first stage of transition from oo to ow in Cumberland and Westmoreland. The short u in but, cut, is oo in foot. The vowels are much fractured or divided throughout north Lancashire, Cumberland, Westmoreland, and Yorkshire, but differently from the forms found in Lincolnshire, the sound of ee or ea in beer, bear, being prefixed. In the south-eastern part long i becomes a in father, which is not the case in the south-western part. The definite article becomes simple vowelless t' (being quite lost in Holderness, Yorkshire) up to a sinuous line through north Cumberland and south Durham, after which the usual form the is resumed and continued to the Scotch border. Instead of I am, I is is regularly used, at least till the is restored, and then generally I am is heard. In Northumberland a transition occurs from the short oo in foot in such words as cut, up, through a sound more resembling German ö, and only slightly different from midland short u, to the regular southern and Scotch form which is established about Woler. But the great peculiarity of Northumberland is the uvular (or as it is usually called, guttural) pronunciation of r, almost as in the north of France and Germany. This is apparently of recent origin, and hardly overlaps the county border on the south, while it is not heard beyond Berwick on the north. The neighbouring Scots cannot pronounce it.
(6) Lowland Scotch pervades all Scotland where Gaelic is not the common language—i.e. generally the south and east, including the Orkneys and Shetlands. Historically, the English language was transplanted into Scotland from the east of Yorkshire and Northumberland, but it now differs materially from the speech there used. The speech in the south of Scotland is probably the most ancient form. (See Dr J. A. H. Murray's Dialect of the Southern Counties of Scotland (1873), which is the classical work on the subject.) Phonetically, the southern form is remarkable for retaining ow when not followed by a consonant, and using oo when a consonant follows, as cow, hoose, whereas in the rest of Scotland they have coo, hoose, as in Yorkshire. Grammatically, it is remarkable for distinguishing the present participle from the verbal noun, both confounded as -ing in modern English, as 'dansin in dansin shoes.' As the English proceeded north to Edinburgh and round to the north of the Forth, this and other distinctions were lost. It is the middle Lowland Scotch, which subsequently spread to the west and south, that is usually known as Scotch by Englishmen, through Burns and Scott in modern times. It was a highly cultivated variety, but was always considered to be English by its older writers. North Lowland Scotch lies north of the Tay, and has the well-known peculiarity of using f for wh, and in Caithness there is the reduction of the definite article to its vowel e alone. The Orkneys and Shetlands are Lowland Scotch engrafted on Norwegian, or Norn, as they term it—i.e. old Norse, which has occasioned many peculiarities both of pronunciation and construction, of which the general, but not universal, use of d, t, for the two sounds of th in the, teeth, need only be mentioned. For the evidence on which the above divisions are founded, with complete details and examples, the reader is referred to Alexander J. Ellis's Existing Phonology of English Dialects (1889), forming Part V. of his Early English Pronunciation.
The above remarks refer to the existing forms of speech. Dr Morris in his prefaces to Hampole, Alliterative Poems, Dan Michel, &c., has endeavoured, from the MSS., to determine the characteristics of our dialects in the Early English period. He admits only three—southern, midland, and northern, but divides the midland, 'which presents us with no one typical form,' into west and east midland.
All languages present a variety of local forms, which has just been partially illustrated for English; and in many of these the discrepancies are much greater than in English. Prince Louis-Lucien Bonaparte has done much not only with the English, but with the Basque, French, and Italian forms; for the last, see also Papanti's collection, I parlari Italiani in Certaldo (1875). The dialects of Low German have been treated by J. Winkler in his Dialektikon (1874). The Swedish dialects are being especially studied now. The High German and Swiss dialects have received much attention. Schmeller's Bavarian Dialects (Mundarten Bayern, 1821) is a remarkable work. And so on.
Now to what end or purpose is this study of dialects pursued? They possess comparatively no literature having any attraction in itself. The modern forms are spoken only by the illiterate, and sound uncouth and harsh to an educated ear. They are not pursued for the ideas which they convey, but for the manner in which they convey them. They lead us to understand how a received or literary language was formed. They contain multitudes of 'missing links,' which serve to explain relations which might otherwise entirely baffle the student of language. Thus, there are two old forms, then pronounced as the present long ee and oo, which now appear as long i and ow. How did this come about? It is evident that the people could not have suddenly left off one and taken up another. Now the actual transitional forms still exist, like fossils, in dialectal speech. Modern philology obliges us to leave the merely written forms and study living usages. This is only possible by studying existent local forms of speech, in other words, dialects.
There is a considerable literature of modern English dialects, but it labours under the disadvantage of generally not being genuine—i.e. not due to dialect speakers themselves, and when so due, as in Burns's poems, it is too apt to be unduly mixed up with modern received speech. Naturally, the born dialect speaker is not a literary man, as the writer of dialect works inevitably is. Some of the best pieces of Cumberland speech (Gibson) and High Furness dialect (Piketah = Barber) have been written by men who had lived only a few years in the country, and had been attracted by the curious forms of the speech among which they came. Others seem to have been written purely 'from the fun of the thing,' such as the Essex 'Tiptree Fair' and 'John Noakes and Mary Stiles,' and the Kent 'Dick and Sal.' One of the few modern books written with a desire to show the better side of the dialect as distinguished from its ludicrous side is the late Rev. W. Barnes's Dorsetshire poems. There are also many serious Cumberland poems given by Gilpin. Nathan Hogg's (Baird's) poems and the Exmoor Seolding, as edited by Elworthy, with the pronunciation and notes, deserve study as the work of natives. The legion of Lancashire and Yorkshire books is chiefly rolling. But Tim Bobbin has become a classic. In Northumberland the Pitman's Pay takes a similar position. Both have influenced the orthography used by other writers. In Scotland, besides Burns and Scott, must be mentioned the excellent Aberdeenshire Johnny Gibb of Gushetneuk, and the same writer's Life among my ain Folk. Dennison's Orcadian Sketch Book is also first-rate.
The Bibliographical List of works relating to dialects, including both glossaries and examples, published by the English Dialect Society, which was founded in 1873, contains 200 pages; and the English Dialect Dictionary (1898 et seq.), edited by Professor Joseph Wright, is based on the Society's publications and collections.