7. Old Norse and Old Danish.

Not until the year 1,000, or the beginning of the 11th Century, do dialectal differentiations seem to be fully developed. O. N., which in general preserves best the characteristics of the old Northern speech, undergoes at this time a few changes that differentiate Dan. and Norse still more. O. Sw. remains throughout closer to O. Dan. The two together are therefore called East Scandinavian. Old Icelandic, that is, Norse on Icelandic soil, develops its own forms, remaining, however, in the main very similar to O. N. These two are then called West Scandinavian. The following are some of the chief differences between West and East Scandinavian at the time (from Noreen, P. G.² I, 527):

8. Remarks.

Assimilation of mp to pp and nk to kk appears also quite early in Danish and Swedish, e.g., kap (kapp) and drocken (see Kalkar), kapp and drokken (Sw.). U—Umlaut seems to be more limited in O. N. than in O. Ic. O. Ic. hl, hn, hr initially appear early as simple l, n, r in O. N. (see Noreen 528), e.g., O. Ic. hlaupa, O. N. loupa; O. Ic. hniga, O. N. niga; O. Ic. hringr, O. N. ringr; O. Ic. fn appears in O. N. as bn or mn, e.g., O. Ic. nafn, O. N. namn (N. Norse navn, nabn, namn). Initial hv, which was a heavy guttural spirant, became kv in Western Norway, kv and khv in Iceland (though written hv still), e.g., O. N., O. Ic. hvelva, Norse kvelva. O. N. ø became æ in Iceland, døma > dæma. O. N. æi became ei in Iceland, e.g., O. N. stæin > O. Ic. stein, O. N. bæin > O. Ic. bein (stin and bin in O. Dan.).

9. Characteristics of Old Northumbrian.

The following are some of the chief differences between O. Nhb. and W. S:

10. Remarks. Metathesis of r.

The above characteristics of O. Nhb. will not only explain a great many later Scotch forms, but also show that a number of words which have been considered loanwords are genuine English. Sco. daw, "day," need not necessarily be traced to O. N. dagr. The W. S. dæg gave Eng. day. Dæg is also the Northern form. Daw may of course be due to a in the oblique cases, but according to 2 dag may have appeared in the nominative case early in the North. This would develop to daw. Sco. daw, verb, "to dawn," is easily explained. W. S. dagian > dawn regularly, Nhb. dagia (see 17 above) > daw. The O. N. daga, "to dawn," is then out of the question. Sco. mauch, "a kinsman"; the O. E. form was mæg, which would have given may. In the North the g was probably not palatal. Furthermore a Northern form mag would regularly develop to maw, might also be mauch (cp. law and lawch, adj., "low," O. N. lagr). O. N. magr, "kinsman," may, however, be the source of mauch. Sco. hals is not from O. N. hals, but from O. Nhb. hals which corresponded to W. S. heals; Sco. hawse, "to clasp," (Ramsay, II, 257); comes from O. Nhb. halsiga, W. S. healsian. (Sco. hailse, "to greet," is a different word, see loanword list, part II.). Forms that appear later in standard English frequently are found earliest in the North (cp. § 10). No. 13 explains some differences in the later pronunciation of Sco. and Eng. No. 12 is a characteristic that is much more common in Middle and Early New Scotch. Many words in this way became identical in form with their Norse cognates, cp. broder, fad(d)er, etc. This will be discussed later. No. 14, Metathesis of r, was carried out extensively in W. S. (see Sievers, 179), e.g., beornan "burn"; iernan, "run"; burn, "a stream"; hors, "horse"; forsk, "frog"; þerscan, "to thrash"; berstan, "to burst"; fierst, "a space of time," (cp. Norse frist, Germ. Frist). This progressive metathesis of r is very common in the South. In the North, on the contrary, metathesis of r has taken place before ht in frohtian, fryhtu, etc. (Sievers, 179, 2). In addition to these a large number of words appear in Old and Middle Sco. differing from literary English with regard to metathesis, sometimes showing metathesis where Eng. does not. A list of words will illustrate this difference: thyrldom, "thraldom"; thirl, "to enthrall"; fryst, "first"; brest, "to burst"; thretty, "thirty"; thrid, "third"; thirl, "to pierce thirl"; gyrs, "grass"; krul, "curl"; drit, "dirt"; warsill, "to wrestle"; scart, "to scratch"; cruddled, "curdled"; birde, O. E. brid, "offspring." The result is that many of these words are more like the corresponding O. N. words than the Anglo-Saxon (cp. O. N. fristr, brenna, Norse tretti, tredie, etc.), hence they have in many cases been considered loanwords. Sco. braist and landbrest, "breakers," (cp. O. N. bresta, landbrest), are not from the Norse but from the corresponding O. Nhb. words. Cors which occurs in Gau may be a similar case and like Eng. cross derived from O. Fr. crois, but Gau otherwise shows considerable Danish influence and Gau's form may be due to that. Eng. curl and dirt (from O. Du. krul and O. N. drit) have undergone metathesis. The Sco. words have not.

11. The Question of Palatalization in O. Nhb.

Just to what extent g, c, sc were palatalized in O. Nhb. is not definitely known. Until this has been ascertained the origin of a number of dialect words in the North will remain uncertain. The palatal character of g, c, sc in O. E. was frequently represented by inserting a palatal vowel, generally e, before the following guttural vowel. Kluge shows (in Litteraturblatt für germ, und rom. Philologie, 1887, 113-114) that the Middle English pronunciation of crinǧen, sinǧen, proves early palatalization, which was, however, not indicated in the writing of the O. E. words cringan, singan. And in the same way palatalization existed in a great many words where it was not graphically represented. Initial sc was always palatalized (Kluge, 114 above). In the MSS. k seems to represent a guttural, c a palatal sound of older c (Sievers, 207, 2). Palatalization of c is quite general. K became palatalized to c in primitive Eng. initially before front vowels, also before Gmc. e and eu (Kluge, P. G.² I, 991). Kluge accepts gutturalizing of a palatal c before a consonant where this position is the result of syncopation of a palatal vowel. In the South palatal c became a fricative ch. According to Kluge it never developed to ch in Northern England and Scotland, but either remained c or recurred to a guttural k. The same is true with regard to g. The exact extent of such palatalization is very difficult to determine. It is possible that the sound always remained a guttural in the North. We have seen that c or g did not cause diphthongation of the following vowel in the North as often as in the South. In view of the fact that palatalization was not always indicated, this may not prove anything, but may, however, indicate less palatalization than in the South. The fact that e or i was sometimes inserted before a following dark vowel, cp. ahefgia, "gravare," gefragia, "interrogare," proves that palatalization in these words, at least, existed.

12. Sk as a Scandinavian Sign. Certain Words in sk. Palatalization in Norse.

Wall argues that non-palatalization cannot be regarded as a sign of Scand. influence and cites a number of words in support of this conclusion (see Wall, § 30). With regard to dick, "ditch," and sag, "sedge," Wall is probably right. Those in sk are, however, not so easily disposed of. The presence of certain words with sk in the South or those cited in sh in the North does not prove the case. While the presence of a word in South Eng. diall. is in favor of its genuine Eng. origin, it does not prove it, for certain words, undoubtedly Scand., are found in the Southern dialects. Shag, "rough hair," Skeat regards as Norse rather than Eng. Scaggy, "shaggy," with initial sk, I would regard as Norse from O. N. skegg, not from O. E. sceagga. Shriek Skeat regards as Scand. Bradley derives it from O. L. G. scricon which is found once in the Heliand. Eng. dial. skrike. Wall on the other hand derives it from O. E. scricon, since scric is found. Scric occurs in O. E. as the name of the shriekbird. The vb. is not found. Whether we regard "shriek" native or not, scrike is to be derived from O. N. skrika. Skeer is from O. N. skera; sheer from O. E. sceran. In form if not in meaning, we have an exact parallel in the M. E. skir, "bright," from O. N. skir, and schir from O. E. scir. In a few cases words that seem Scand. appear with sh, not sk. The etymology of such words, however, becomes rather doubtful. This is especially the case where in the Norse word a guttural vowel followed the sk. Where, however, the Norse or Dan. word had a palatal vowel after the sk the change to sh is not at all impossible, and here arises the question of palatalization in O. N. O. N. skiól, pron. sk-iól, with sk, = Norse skjūl (pron. shūl). Ski thus becomes sh in O. N. skilinn, Norse shil, O. N. skilja, Norse shilja (or skille), O. N. skipta, Norse shifta. West Norse also shows change of k to ch before i where the k has been kept in East Scand., e.g., O. Ic. ekki = W. Norse (dial.) ikkje or intje, pron. ittje, intje, Dan. ikke (igge). I between sk and a dark vowel early became j in Norse, which then gave the preceding sk something of a palatal nature. The development of O. N. skiól into shiel in Scotland and England may be explained in this way, as skiól > shul in Norway. This is, however, to be understood in this way, that if an i or e followed the sk, this was in condition to become palatalized, not that it was at all palatal at the time of borrowing. The sound was then distinctly guttural, and the guttural character of sk has in nearly every case been kept in Scand. loanwords in English, for palatalization of O. E. sc was completed before the period of borrowing. This palatalization of sk was general in Scotland as well as in England, and such words in sk must be regarded as Scand. loanwords.

13. Conclusion as to the Test of Non-palatalization.

As initial sk, corresponding to O. N. sk, O. E. sc, is due to Scand. influence, so, in general, medial and final sk may be also so regarded: cp. here Sco. harsk, "harsh," bask (adj.), mensk, forjeskit, etc. The guttural character of g and k in Sco. is not to be regarded as due to Scand. influence. Thus mirk, reek, steek, streek, breek, dik, rike, sark, kirn, lig, brig, rig, etc., are to be derived from the corresponding O. Nhb. words, not from O. N. There is something of uncertainty in these words, however, as they all could come from the O. N. O. N. hryggr, for instance, would become rig in Sco., just as would O. Nhb. rycg (rygg). O. N. bryggia would become brig, just as well as O. Nhb. brycg (brygg). The i after g in bryggia does not hinder this, since, as we know, the O. N. word was pronounced brygg-ia, not bryddja, as a later form would be.

14. Old and Middle Scotch.

After Chaucer, Northumbrian English became a mere popular dialect no longer represented in literature. But the form of Northumbrian spoken north of the Tweed, Lowland Scotch, has during the next three hundred years quite a different history. From the Scottish war of Independence to the Union of the Crowns, Scotland had its own literary language. It is customary to speak of three periods of Scottish language and literature as Old, Middle and New: Old Scotch extending down to about 1450; Middle Scotch to the Union of the Crowns; and New Scotch covering the period after the Union. This is, of course, simply a Northern and later form of the Northumbrian we have discussed above.

15. Some Characteristics of Scotch. O. E. ă, ā.

There are no monuments in O. Sco. dating back to the 13th or first half of the 14th Century. The first of any importance that we have is "The Bruce" of 1375. By this time the language of Scotland had already undergone many changes that made its general character quite different from literary or Midland English. None of these changes tended so much to differentiate the two as the very different development of O. E. long and short a. In the south O. E. a > ē (name > nę̄m > nēm); but O. E. ā > ǭ, later ō (stān > stǭn > stōne, hām > hǭm > hōme). The change of ā to ǭ (probably about 1200) took place before that of ă to ā, else they would have coincided and both developed to ō or ē. The last is precisely what took place in Scotland. O. Nhb. ă > ā and early coincided with original ā, and along with it developed to later ē, as only short a did in the south. The two appear together in rhyme in Barbour. Their graphic representation is a, ai, ay. The sound in Barbour is probably ǣ or ę̄. In "Wallace" Fr. entré is also written entray, entra. Fr. a and ei and Eng. diphthong ai (< æg) rhyme regularly with Sco. a, ay, ai, from O. E. ā. On O. E. and O. N. ā- and M. Sco. ē-sounds in general see Curtis, §§ 1-165.

16. Curtis's Table.

The following (see Curtis §§ 144-145) illustrates the development of O. E. ă, and ā, in England and Scotland:

The intermediate stage of this development, however, is explained in two ways. According to Curtis it was (in 2) ā > ę̄ > ē > ī > iə. Luik (§ 244) shows that

das Vorrücken zum Vocalextrem ist an die Abstumpfung gebunden; wir finden es nur dort, wo auch Abstumpfung zu constatieren ist, wäbrend diese selbst ein weiteres Gebiet hat. Schon daraus folgt, dass die Abstumpfung das Primäre ist, dass also ihre Basis e war, nicht i. Dies wird bestätigt durch eine einfache Erwägung. Hätte die Abstumpfung die Lautstufe i ergriffen, so hätte sie auch das e treffen müssen, das ja schon seit Beginn der neuenglischen Zeit in allen Dialekten durch i vertreten ist. Endlich bieten die frühesten Zeugnisse nur e, nicht i, auch für solche Striche, die heute i haben.

According to this, then, the development is more probably ā̆ > ę̄ > ēə > iə, or, as Luik thinks, ā̆ > æ > æə, or ę̄ə > ēə > iə

17. O. E. ō.—A List of Illustrative Words from the Aberdeen Dialect.

Another Northern peculiarity relates to O. E. ō. While in the south O. E. ō developed to an ū-vowel or an ū-fracture, in Scotland it became ee (ui, ee, i). The process involved here does not yet seem to be fully understood. The modern dialect of Aberdeen is most pronounced in this respect, older i also frequently becoming u, o. The following examples taken from "Johnnie Gibb" (Aberdeen. 1871) will illustrate:

18. Inorganic y in Scotch.

Many words have developed a y where originally there was none. This phenomenon is, however, closely connected with e-i-fracture from original ā̆. Y we find appears often before a (from original ā̆). It is, then, simply the development of the e-i-fracture into a consonant + a, and may be represented thus: O. E. āc ("oak") > ę̄c > ēc > ēəc > iəc > yak. (See also Murray D.S.C.S., 105). Cp. yance and yence, "once"; yell, "ale"; yak, "ache." This also appears in connection with fracture other than that from O. E. ā: cp. yirth, yird, for "earth."

19. D for the Spirant th.

This appears in a number of words: e.g., ledder, "leather"; fader (in Gau),fadder, "father"; moder, mudder, "mother"; broder, brudder, "brother"; lidder (A.S. liðre); de (Gau), "the" (article); widdie (O. E. wiðig), "withy"; dead, "death"; ferde, "fourth"; etc. In some works this tendency is quite general. Norse loanwords as a rule keep the spirant, but in the following loanwords ð has become d: cleed, cleeding, "clothe, clothing," from O. N. klæða; red, "to clear up," O. N. ryðja; bodin, O. N. boðinn (? See E.D.D.); bud, "bribe," O. N. boð; heid, "brightness," O. N. hæið; eident, "busy," O. N. iðinn (ythand is, however, the more common Sco. form); bledder, "to prate," O. N. blaðra (more commonly blether in Sco.); byrd, "ought," O. N. burði; stiddy, O. N. steði. I do not think ryde, "severe," can be derived from O. N. reiðr; and frody, "wise," is rather O. E. frod than O. N. fróðr. Waith, O. N. væiðr, has kept the spirant, but faid, a "company of hunters," has changed it to d. Faid probably comes in from Gaelic. I have called attention to this change of ð to d in Sco., since many words affected by it have become almost identical in form with their Scand. cognates and have consequently been considered loan-words. See § 23.

20. O. E. ā and O. N. æi. How far we can Determine such Words to be of Native or of Norse Origin.

Certain Eng. dialect words in ē corresponding to O. E. ā have been considered Scand. loanwords. We have, however, seen that in the north O. E. ā > ē just as did O. N. æi (ei). How many of these words are genuine English and how many are loanwords becomes, then, rather uncertain. Wall argues that the Norse words were always in M. E. spelled with a diphthong, while the genuine English words were spelled with an a—thus bain, baisk from O. N. bæinn, bæiskr, but hame, stane, hale from O. E. hām, stān, hāl. If this were always the case we should have here a safe test. It is, however, a fact that in Scottish texts at least, no such consistency exists with regards to these words. The following variant spellings will show this: hame, haim, haym; stain, stane, stayne; hal, hale, hail, hayle; lak, lake, laik, layk; blake, blaik, blayk, etc., etc. There is, however, another way in which to determine which of such words are loanwords and which are not. In Southern Scotland in D. 33, and in Northwestern England (D. 31), O. N. æi and O. E. ā did not coincide, but have been kept distinct down to the present time (see Ellis's word-lists and Luik, 220, 221). In these two dialects O. E. ā developed to an i-fracture (see § 16.2), while O. N. æi never went beyond the e-stage, and remains an e-vowel in the modern dialects. Here, then, we have a perfectly safe test for a large number of words. Those that have in D. 31 and D. 33 an i-vowel or an i-fracture are genuine English, those that have an e-vowel are Scandinavian loanwords. Ellis's list offers too few examples of words of this class. We find hi'm, bi'n, hi'l, sti'n, and in Murray's D.S.C.S. heame, and heale (beside geate (O. N. gata), beath, meake, tweae, neame, etc.). This then proves that Sco. haim, bain, hail, and stain are from O. E. hām, bān, hāl, stān and not from O. N. hæim, bæinn, hæil, stæinn. Mair, in spite of its e-vowel, is not from O. N. mæir, for a following r prevented the development to i, as a rule, although in Cumberland meear is found beside mair. The word "steak" (O. N. stæik), which occurs in Ellis's list, has had an irregular development and cannot be considered here (see further Luik, 323). In the following works are found a number of words of this class:

Westmoreland and Cumberland Dialects, by J.R. Smith. London. 1839.
A Glossary of Words and Phrases of Cumberland, by William Dickinson. London. 1859.
Folk Speech of Cumberland, by Alexander Craig Gibson. London. 1873.
A Glossary of Words used in Swaledale, Yorkshire, by John Harand. E.D.S. 1873.
Whitby Glossary, by F.K. Robinson. E.D.S. 1876.

21. A List of Some Words that are Norse. Further Remarks.

These all aim at giving the phonetic value of the sounds. O. E., O. N. ā is represented by ea or eea, indicating i-fracture. For instance: heam, steean, neam, geat, beeath, leath (O. N. laði), heeal, brea (O. N. brā), breead (O. E. brād, not O. N. bræi), greeay, blea, etc. Those that have a, ai, or ay, that is an e-vowel, and must consequently be derived from the corresponding O. N. words, are the following:

In addition to these, blain, "to become white," is a Scand. loan-word, but rather from Dan. blegne than Norse blæikna, cp. blake above. Blained, adj. "half dry," said of linen hung out to dry, is, of course, simply the pp. of blain, cp. Dan. blegned. Skaif, "distant, wild, scattered abroad, or apt to be dispersed" (is the definition given), corresponds exactly to O. N. skæif in form, but not in meaning. Skæif meant "crooked." Sco. daive, "to stun, stupefy," is here regularly spelled deeave (deave in Swaledale). It must, then, be derived from O. E. deafian, not O. N. döyfa, O. Ic. deyfa. Swaledale slaiching, "sneaking," is the same as O. N. slæikja, "to lick"; a secondary meaning of O. N. slæikja is "to sneak"; keeal, "kail," could come from O. N. kál or Gael. cál. It is probably from the latter. The word slaister, "to dawdle, to waste one's time," is not clear. The sb. slaisterer, "a slink, an untidy person," is also found. The ai indicates an original diphthong. It is probably the same as Norse slöysa, sb. "an untidy person," as vb. "to be untidy, to be careless." Ster (slais + ster) would, then, be an Eng. suffix, or it may be the same as that in Sco. camstary, cp. Germ. halsstarrig. The Norse word slöysa is probably not the direct source of the Eng. dialect word. Slaister, however, for slöysa, seems to be a recent word in Norse. Skane, "to cut the shell fish out of the shell" (Wall, list B), is to be derived from O. N. skæina, rather than from O. E. scænan. Slade, "breadth of greensward in plowed land," cannot be from O. N. slettr, "plain," sletta, "a plain." Neither form nor meaning quite correspond. The Sw. slägd corresponds perfectly in form but not in meaning. It is, however, probably from O. E. slæd. This word is taken from Wall's list, not from the works named above.

22. Celtic, Lowland Scotch, and Norse.

In Gaelic and Irish, in the Western Isles and the Highlands, considerable Norse elements are found as the result of Norse occupancy that continued in the Isles, at least, for several hundred years. A number of words that have come into Gaelic and Irish from Norse are also found in Lowland Scotch. In some cases it seems that the word has not come into Lowland Scotch direct from Norse, but by way of Gaelic or Irish. Craigie has given a list of about 200 words in Gaelic that seem to come from Norse. Out of these I will take a few that have corresponding words in Scotch:

Garth and loft agree perfectly with the O. N. and are not doubtful. With the Gael. gardh cp. O. N. garðr and O. Sw. gardher. The Sco. garth has changed the original voiced spirant to a voiceless one. In Gael. lobht f has become v. Prin is rather doubtful. There is an O. E. prēon from which the Gael. word may have come. The Sco. word prin does not seem to come from either O. E. prēon or O. N. prjónn, but from the Gael. prine. There is a Northern dialectic prēon which may come from O. E. prēon. There is also a pren in Dan. dial. Stoup has the Norse diphthong which has been simplified in Gael. stop. Skep is a little doubtful because of meaning. The loanword sgeap in Gael. has the specialized meaning of "a beehive." This meaning the Sco. word has very frequently, the Norse to my knowledge never. It may be a case of borrowed meaning from Gael. Girth is from the Norse. Girt is probably simply change of th to t, which is also found elsewhere in Sco. Knap may be from either. Mull in Sco. may be native English. The word occurs in L. G. Sker is from O. N. Skarth is anomalous, showing change of f to th. In the Gael. scarbh, f is changed to v as in lobht. Ged is nearer the O. N. Scait could be from either, as also brod. Sco. mask is probably not at all a loanword, and may be from older mex by metathesis of s; cp. O. E. mexfat and Sco. maskfat cited by Skeat, Et. Dict. The Gael. masg is probably not a loanword from the Scand., but from O. E., or perhaps from O. Sco. An O. Nhb. mesk probably existed. Ransack agrees with the Norse word. The spelling runsick found once (Wallace VII, 120), probably does not represent the exact sound, and is, in any case, as ransack to be derived from the O. N. and not through the Gael. Faid, "a company of hunters," has already once been referred to. This cannot possibly come from the O. N. væiðr, for while the spirant ð sometimes becomes d, O. N. v regularly becomes w in Sco. (rarely v). We should expect the form waith, and this is the form we have in Wallace I, 326, in the sense "the spoil of the chase." There is a Gael. fiadhoig, meaning "a huntsman." The first element fiad seems to be the O. N. veiðr with regular change of ð to d (or dh, cp. gardha), and v or w to f which is considered a sign of Gael. influence in Aberdeen Sco., cp. fat for what, fen for when, etc., the development probably being wh > w > v > f. Faid in Sco. is then probably from the Gaelic.

23. Some Words that are not Scandinavian Loanwords.

We have spoken in §§ 10, 13, 20 and 22, of a number of words that are to be considered regular Sco. developments of O. E. words. The following words have also generally been derived from the Scand., but must be considered native, or from sources other than Norse: