CORNELL UNIVERSITY LIBRARY ENGLISH COLLECTION THE GIFT OF JAMES MORGAN HART PROFESSOR OF KNGUSH PE 1165.B78" ""'™"'** '■"'"'^ "'^'SHiiiStMSy "' "•* e-vowel in accen 3 1924 026 632 723 The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://archive.org/details/cu31924026632723 AN Historical Study of the e- Vowel IN ACCENTED SYLLABLES IN ENGLISH A DISSERTATION PRESENTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. BY EDWIN W. BO WEN, Ph. D. BALTIMORE: JOHN MURPHY & CO. 1893. 9 CONTENTS. PAGE. IIJTRODUCTION, - - 5 ANGLO-SAXON PERIOD. I. — 1. Oeigin of A.S. m, e, 7 2. Obigin of A.S. EA, ED, ^ - 11 II- — Development of the e-Vo'wel in Anglo-Saxon, - 14 1. By Lengthening, - . 15 2. By Contraction, - - 18 3. By Umlaut, 22 4. By Palatalization, 29 5. By Latin Importations, 30 6. By the Influence of w, 30 MIDDLE-ENGLISH PERIOD. ni. — The e- Vowel in the Seveiull M.E. Dialects, - 32 1. General Remarks, 32 2. East-Midland Dialect, 36 3. West-Midland Dialect, 42 4. Southern Dialect, - 44 5. Kentish Dialect, 46 6. Northern Dialect, - 47 7. Dialect which Gave Rise to Standard English, 48 MODERN ENGLISH PERIOD. IV. — The ee- Vowel in the Middle Tbansition and Fikst Modebn Peeiod, 1500-1600, - - - - 54 V. — The ee-Vowel in the Second Modebn Pebiod, 1600-1700, - 60 2. Medial ie as a Graphic Representative of close ee, 64 3. Development of ee from aa in Modem English, 65 4. Development of ee from ai, ei in Modern English, 69 VI. — The ee-Vowel in the Thied Modebn Period and in Living English, ... 71 2. The Diphthongs ai and ei, - - 74 3. Development of close ee from Latin x, oe, - 76 VII.— Results, - - - - - - 77 INTRODUCTION. This study is an attempt to sketch the development and history of the long e-vowel from Anglo-Saxon to Living English. It has been found convenient in the treatment of the present subject to make three principal divisions which we term the Anglo-Saxon, the Middle and Modern English periods. It is, of course, im- possible to draw any hard and fast line between these several periods, but we can determine their approximate limits, which, for our purpose, is sufficient. Following Sweet we may say that the Anglo-Saxon period extends to 11 50, though the documents between 1050 and 1150 show the language in a state of transition from Anglo-Saxon to Middle English. Fully developed Middle English has for its inferior and superior limits 1150 and 1450, the period between 1200 and 1400 being especially well marked. The docu- ments written between 1450 and 1500 show the language assuming a, somewhat modern aspect and this period is appropriately desig- nated the Middle Transition in contradistinction to the Old Transi- tion extending from 1050 to 1150. Modern English dates from about 1500, and admits of several subdivisions according to cen- turies, which are followed in the treatment of the modern period. After brief reference to the origin of the e-vowel in the Germanic languages, its sources and development in Anglo-Saxon are inves- tigated. Copious examples with their cognates in the Old High Oerman, Old Saxon, Old Norse, and Gothic are given. The West- Saxon dialect, embracing as it does most of the literature of the early period of the language, receives especial attention and is made the basis of the investigation in Anglo-Saxon. But for the sake of completeness the e-sound in the non-West-Saxon dialects is also investigated. These dialects, however, the Kentish, Mercian, and 5 6 Introduction,. Northumbrian, are considered only in so far as they exhibit varia- tions of the e-sound from the West-Saxon norm. It will be observed that the section on Anglo-Saxon is not limited to the symbol e, but includes also the symbol » and the diphthongs ea and eo. The reason for this is very obvious. All of these vowels and diphthongs yield, in Middle English, long e, and are, therefore^ justly entitled to consideration in this study. So, also, the vowel a and the diphthongs ai and ei are considered in the section dealing with Modern English, because these symbols came at one time in the modern period to have the value of the vowel e. The treatment of the Middle English period is in a manner similar to that of the Anglo-Saxon. A representative text of each of the dialects is selected and made the basis of investigation with regard to the e-sound. The Ormulum as representing the East- Midland dialect and being more in the line of the subsequent de- velopment of the language occupies, for our purpose, the most prominent position among the Middle English texts, and is there- fore considered at length. After the examination of the several Middle English dialects attention is drawn to late Middle English,, especially to the dialect that gave rise to standard English, as represented in the Chaucerian texts and the London State and Parliamentary documents. In this section it is shown how there were two distinct long e's and what their respective sources in Anglo-Saxon were, and how the domain of long e in Middle English was extended from other sources than English. A brief review of the Middle Transitional period is given as forming a connecting link between Middle and Modern English,, and then the modern period is considered. Here it is shown what the two e-sounds, open and close, of Middle English developed into during the several centuries of the modern period, and what other vowels and diphthongs, as a, and ai, ei, came to have the value of long e, and, further, how even these later came to have another sound than the pure e-sound. So then it appears that the pure phonetic e-sound, open and close, from being quite a common sound in Anglo-Saxon and far more so in Middle English, came in Early Modern English to be less common, and that this sound in present Living English is of very rare occurrence. For the material of the modern period I have relied almost entirely on Ellis's monumental work, Early English Pronunciation. An Historical Study of the e -Vowel. THE ANGLO-SAXON PERIOD. 1. Origin op A.S. m, e. In the Primitive Germanic vowel system there existed two sounds of long e, one of which corresponds to the Indo-European e, while the other canngt be traced back, with entire satisfaction, to any definite I.E. sonant. Cf. Brugmann I, § .75, trans. The first of these sounds, viz., Germanic e corresponding to I.E. e is assumed to be an open, the second a close sound. To the open Germ, e, which must have been a low-front-wide vowel, correspond Gothic e, Old Norse a. Old Frisian e, Old Saxon a, Old High German a, Anglo-Saxon 8e(e). To the close Germ, e, which must have been a mid-front-narrow vowel, correspond Goth, e, O.N. e, O.Fris. e, O. S. e, O.H.G. e (later ea, ia, ie^), A.S. e. Gothic levelled both the Germanic e's under the close e-sound. The real Germ, close e which is a rare sound appears in Gothic in only four words, viz., her, Kreks, fera, mes. The West-Saxon representative of the Germ, open e-sound is ae for which Sievers assumes a West-Ger- manic a. Cf. Paul and Braune's Beitrage VIII, p. 88 ; Sievers's Grammar of Old English, trans, by Cook, § 57 and § 68. Bremer (cf. P.B.B. XI, p. 1, seq.) and Siebs maintaining the contrary view, hold that W.S. » and O.Fris. e represent the retention of the Primitive Germ, e-quality, which, like the I.E. e, is assumed to be an open sound. The non-W.S. dialects represent the Germ. > Cf. Braune, Ahd. Grammatik, | 35, 48. 8 An Historical Study of the e- Vowel. open e by e, and the non-W.S. dialects and W.S. alike represent the Germ, close e by e. 1. The following examples will show the representation in Anglo-Saxon of the Germ, open e (=Goth. e, O.H.G. a, etc.) which appears in West-Saxon as £e. This S is a low-front-wide vowel, but is represented in the non-W.S. dialects by a mid-front- narrow sound, e. W.S. Sfen, Merc, efen (cf. V.P.' 29, 6 ; 54, 18), O.H.G. aband, O.S. aband. W.S. bsel, non-W.S. bel (cf. Leiden, Erfurt Gloss.), O.N. bal. W.S. bjer, non-W.S. ber (cf. Epinal Gloss. 137, Erfurt 13), O.H.G. bara, O.S. bara. W.S. brie^, smell, breath. O.H.G. bradan. W.S. daed, non-W.S. ded (cf. Merc, V.P. 63, 10), Goth. gade);s, O.H.G. tat. W.S. fser, non-W.S. fer (cf. Corp. Gloss. 419), O.H.G. fara, O.S. far, O.N. far. W.S. heer, non-W.S. her (cf. V.P. 34, 13; Corpus 1594), O.H.G. har, O.S. har, O.Fris. her. W.S. lletan, non-W.S. letan (cf. V.P. 9, 11 ; 15, 10), Goth, letan, O.H.G. lazan, O.S. latan. W.S. IbstS, landed property, cf. Goth. unle)?s. W.S. mal, Goth, mel, O.H.G. mal, O.S. mal, O.N. mal. W.S. msete, poor, moderate, O.H.G. maza, O.N. mat. W.S. meg's, mowing, O.H.G. mad, Gr. a/ti^ro?. See Kluge : mahd. W.S. DEedl, non-W.S. nedl (cf. Corpus Gloss. 66), Goth. nej;la, O.H.G. nadala, O.S. nadla, O.Fris. nedle. W.S. ried, non-W.S. red (cf. Kent. Chart. 37, 26 ; 42, 22, etc.), O.H.G. rat, O.S. rad, O.Fris. red, O.N. ra'S. W.S. rsedan, Goth, redan, O.H.G. ratan, O.S. radan, O.N. ra«a. W.S. ssed, non-W.S. sed (cf. V.P. 20, 11 ; 21, 24, etc.), O.H.G. sat, O.S. sad, O.Fris. sed. W.S. ssil, non-W.S. sel (cf. V.P. 13, 3 : ungeselignis), Goth. sels, O.H.G. salig, O.S. salig blessed, fortunate. W.S. slsepan, non-W.S. slepan (cf. V.P. 3, 6 ; 12, 4, etc.), Goth, slepan, O.H.G. slafan, O.S. slapan, O.Fris. slepa. ' V.P. is the abbreviation for the Vespasian Psalter :— Sweet's O.E. Texts. An Sistcyfieal Study of the e-Vmoel. 9 W.S. strSl, non-W.S. strel (cf. V.P. 90, 6; 56, 5; 10, 3, etc.), O.H.G. strala. W.S. swses, non-W.S. swes (cf. Kent. Ch. 37, 25 ; 42, 16 ; swe- send), Goth, swes (cf. Feist, 559), O.H.G. swas, O.S. swas, O.Fris. swes. W.S. wffig, Goth, wegs, O.H.G. wag, O.S. wag, O.Fris. weg. W.S. wsepn, non-W.S. wepn (cf. V.P. 34, 2; 45, 10, etc.), O.H.G. wafan, O.S. wapan, O.Fris. wopin. W.S. wier, non-W.S. wer (cf. V.P. 118, 158), O.H.G. war, O.S. war, Lat. verus. W.S. ffit, food, O.H.G. az, Goth, uzeta, afetja (cf. Feist, 141), O.S. at. W.S. lerende, non-W.S. erendwreca (cf. V.P. 67, 32), O.H.G. arunti, O.S. arundi. W.S. mm, non-W.S. e-Sm (cf. Epinal Gloss. 89), O.H.G. adum, O.S. adorn. W.S. sis, carrion, O.H.G. awesi, M.H.G. as < awasel (cf. Schade, p. 24). W.S. blSd, blast, non-W.S. bled (cf. Epinal and Corp. Gloss.), O.H.G. blat. W.S. grSdig, non-W.S. gredig (cf. Epinal Gloss. 500), Goth, gredags, O.H.G. gratag, O.S. gradag, O.N. gradhugr. See Schade. W.S. graeg, non-W.S. greg (cf. Epinal 473, Corp. 967 ; grei), O.H.G. grat, O.N. grar. W.S. hwSr, non-W.S. hwer (cf. V.P. 41, 4, 10, 11, 78, etc.), O.H.G. hwar, O.S. hwar. W.S. mseg, non-W.S. meg (cf. Epinal 164, Kent Ch. 39, 6 ; 44, 6), O.H.G. mag, O.S. mag, O.Fris. mech. W.S. sprffic, non-W.S. spree, O.H.G. spraka, sprahta, O.Fris. spreka. W.S. );£er, non-W.S. )>er (cf, V.P. 22, 2; 35, 13, etc.), O.H.G. dar, O.S. thar. Here belong further the Pret, Pis. of classes 4 and 5 of Ablant Verbs, as : W.S. seton, non-W.S. eton, Goth, etum, O.H.G. azum. W.S. bffidon, non-W.S. bedon, Goth, bedum, O.H.G. batum, O.S. badun. 10 An Historical Study of the e -Vowel. W.S. bseron, non-W.S. beron;Goth. berum, O.H.G. barum, O.S. barun. So with the rest, as cwSlon, hselon, brScon, etc. It is noteworthy that the West Germanic a is retained in Anglo- Saxon when it occurs before w and, in some cases, when it occurs in an open syllable followed by a guttural vowel. A.S. sawon, Pret. PL, beside Sffigon, cf. Goth, sehwum, O.S. sawun. So the 2nd sing. Indie, sawe, and the Pret. Subj. sawe, sawen. A.S. tawian, getawe, cf. Goth, tewa, gatewjan. A.S. clawu beside clea < * klawu, klau, cf, O.H.G. klawa. A.S. ^awan. See Sievers, § 57, cf. Gr. rrfKos. See Kluge : tauen, A.S. lagon, Pret. PI., beside Isegon.cf. Goth, legun, O.H.G. lagun, O.S. lagun. A.S. wagon, Pret. PI., beside wsegon. See Sievers, § 57, 2, note 3. So belong here slapan beside slsepan, swar beside swser, wat beside wffit (cf. Goth, wato, O.H.G. wazzar, O.S. water), lacnian, beside Ifficnian (cf. Goth, lekeis and A.S, Isece with i-umlaut), magas be- side miegas, Fem. mage beside msege, wag beside wseg, etc. If, however, i or j follows, i-umlaut results producing ie, as Iffiwan (cf. Goth, lewjan < lew, O.H.G. laen), seltsewe (cf. Goth, tewa). 2. The Germ, close e (= Goth, e, O.S. e, O.H.G. ea, ie, ia, etc.) appears in all the Anglo-Saxon dialects as e (mid-front-narrow). This original e is of very rare occurrence in Anglo-Saxon. A.S. cen, torch, O.H.G. kien, * ken, M.H.G. kien. See Kluge : kien. , A.S. her, Goth, her, O.H.G. hear, hiar, hier, O.S. her, O.Fris. her, O.N. her. A.S. lef, weak, O.S. lef, O.Fris. lef. The e of the Eeduplicated Preterites can not rightly be placed here, because it represents the results of contraction. So A.S. med, meed, Goth, mizdo (< Pre-Germ. mizdha), Gr. fiiar66<;, O.H.G. meta, (mieta, miata), O.S. meda, O.Fris. mede, mide, meide (Kluge : miete and Feist, 405) does not belong here. Of e (mid-front-narrow) as the representative in the Kentish and Anglian dialects of the W.S. Ie (low-front-wide) it is not necessary to multiply examples. An Historical Study of the e -Vowel, 11 ORIGESr OF THE AnGLO-SAXON DIPHTHONGS EA AND EO. 1. Origin of A.8. ea. The dipthong ea in Anglo-Saxon represents I.E. au, ou, Germ., West Germ, au which appears in Goth, as au, in Old High German as ou (6 before h and dentals), in Old Saxon as 6, in Old Norse as au (6 before h), in Old Frisian as a. The development of this diph- thong from the Germanic au is through ceu and ceo to ea, the Germ, a becoming ce in Anglo-Saxon. The first element of the A.S. ea then is a low-front-wide vowel and the second a low-back-wide vowel. The stress must have been on the first element, for it is certain that the Germ, au was accented on the first element, thus : au. The lengthening probably began with an exaggeration of the glide vowel between the two elements, as Sweet, History of Eng- lish Sounds, § 466, suggests. In some of the oldest texts seo actu- ally occurs, as genseot in the Corpus Gloss. Additional evidence that the first element of the diphthong ea is a low-front-wide vowel, and not a wide-front-narrow, is furnished (1) by the history of its subsequent development in Middle and Modern English, becoming, as it does, open e (e) in Middle English, and being represented by ea in Modern English, and (2) by the fact that in the earliest non- W.S. texts in which this diphthong is simplified before palatals, the vowel se occurs, as in baeg, hseh in the Corpus Gloss. This diphthong appears in all the dialects as ea. A.S. bean, O.H.G. bona, O.N. baun, Germ, base bauno. A.S. deaf, Goth, daufs, O.H.G. doub, cf. Gr. Tv^x66. See Kluge : doch. A.S. ^reat, troop, throng, cf.Goth. ];rutsfill, — ^|7riutan, O.H.G. thriogan, O.N. ]?raust, Lat. trudo. See Kluge : verdriessen. A.S. deaw, O.H.G. tou (touwes), O.S. dau, O.N. dogg, Germ, base dauwo — . A.S. steor, Goth, stiur, O.H.G. stior, O.N. 'Sjorr, Kluge : stier. A.S. «eo(h), thigh, O.H.G. dioh, O.N. ]>jd. 14 All, Historical Study of the e -Vowel. A.S. -Seod, Ooth. ]7iuda, O.H.G. thiota, diota, O.S. thioda, O.Fris. thiade. A.S. •Seof, O.H.G. diob, O.S. thiof, O.Fris. thiaf, cf. Goth. ^iubi. A.S. ISeote, O.H.G. diozo, cf. 'Seotan, O.N. )7J6ta. In a few words eo = Germ, eu (which is itself a contraction), as in : A.S. feower, O.H.G, fior, O.S. fiuwar, fior, O.Fris. fior. Kluge, P.G., I, 403. A.S. hweol < hweowol < * hwejwol, cf. Sk. cakra, Gr. KUKXa. Kluge, P.G., I, 331. A.S. «eo, Goth, pjus, O.H.G. deo, cf. O.N. ]?yr (with i-umlaut). II. Development of the e- Vowel in Anglo-Saxon. We have thus far considered only those sounds which were origi- nal in Anglo-Saxon. We have seen that the Germanic sounds sd and e were preserved unchanged in Anglo-Saxon, while the diph- thongs ea and eo, representing Germanic au and eu respectively, have not preserved their Germanic values unchanged. Sievei^s (Grammar, § 47) states that the transformation which the Germanic vowels have undergone in Anglo-Saxon are essentially of a twofold character. " The mutation of the vowel either takes place inde- pendently of its environment, or the latter exercises a determining influence upon it." The first kind of change is exemplified in ea and eo from Germanic au and eu which we have already considered. The second kind are phenomena such as lengthening, contraction, the various umlauts, palatalization, etc., which illustrate the devel- opment of the vowel under consideration in Anglo-Saxon. To this development let us now turn. In the earliest literature of our ancestors we observe remarkable linguistic phenomena by which the sphere of the sounds we are at present considering was very materially enlarged. The causes ■jphich led to the development of these sounds are, chiefly, the various kinds of lengthening, contraction, the various umlauts, and palat- alization, named in the order of their operation. Some of these processes, however, slightly overlap. An Historical Stvdy of the e-Vowd. 16 1. Lengthening. There are several kinds of lengthening in Anglo-Saxon, viz : lengthening in auslant, lengthening in monosyllables, lengthening by compensation for the loss of consonants, lengthening before certain consonant combinations. The extent of this process cannot be determined with entire accuracy, for the indications of quantity in the MSS. are scanty, and even these indications, few as they are, have not been fiiithfuUy reproduced by the editors of the texts. It is difficult to formulate any laws of lengthening from the fact that the phenomena in question vary with the individual dialect both with respect to extent and period of occurrence. The date of this process is also difficult to determine, but we may safely say that the beginning is Pre-Germanic. 1. A.S. e (mid-front-narrow) in the following words arises from the lengthening of final e of monosyllables, the originally final consonant being lost (cf. Sievers, § 121). The e preserves its original close quality. A.S. ge < * ge(r), O.H.G. ir, O.S. ge gi. A.S. he < * he(r), Goth, hi- in himma daga, O.S. he, O.H.G. her, he. A.S. me < * me(r), O.H.G. mir, Goth. Dat. mis, O.S. mi. A.S. se, the, that, Goth, sa, O.S. se. A.S. ■Se < *«e(r), O.H.G. dir, dih, O.S. thic beside thi. A.S. we < * we(r), O.H.G. wir, wir, Goth, weis, O.S. we. A.S. ne, nor, Goth, ni (ne), O.H.G. ni, O.S. nek. See Kluge : nein. 2. A.S. e arises from a tendency to lengthen monosyllabic words ending in a single consonant which may represent an original gem- ination (Sievers, § 122). These e's are both open and close, coming from originally open and close e. The open e is written thus : e. A.S. bet, Compar. of adv., better, beside regular bet. A.S. men, PI. of man, beside regular menu. A.S. geset < geseted gesett (PP.) beside regular geset. So A.S. weg beside regular weg, wel beside wel, wer beside wer. Note especially the ablant Prets., as bsed, br»c, sSt, etc., which are to be placed here according to Sievers. Sweet (cf. H.E.S, § 406), 16 An Hislorioal Study of the e -Vowel. on the contrary, would explain the length in these cases as due to the analogy of the Pis. bSdon, brScon, etc. 3. Syncope gives rise to lengthening by compensation in many cases. Syncope of g occurs frequently after a palatal vowel and when followed by one of the voiced consonants d, iS, n, with com- pensatory lengthening as the result. Syncope of g takes place also before originally syllabic n, likewise causing compensatory lengthening of the preceding vowel (Sievers, § 214). A.S. mseden for msegden, of. O.H.G. magatin, O.S. mage's. A.S. wffiu for wsegn. See Sievers, § 214, 2. A.S. sJede for sasgde Pret. of secgan, cf. O.H.G. sageta. A.S. beled P.P. for belegd (found in Orosius, 3, 8.). A.S. bredan beside bregdan, O.S. bregdan, O.H.G. brettan. A.S. gen, ongen beside ongegn, cf. togegnes. A.S. lede, geled for legde, gelegde. A.S. ren for regn. ren may be formed by analogy to renes. A.S. stredan beside stregdan. See Sievers, §§ 389, 214, 3. A.S. Sen for Segn, cf. O.S. thegan, O.H.G. degan. Germ, base thegna-. A.S. Senian beside Segnian, Signean, to serve. A.S. sede for sSde ( 242). Thus arise eo and ea, in Anglo-Saxon, in the following words : A.S. eoles, Gen. < eolhes,^ cf. O.H.G. elaho. A.S. feolan, to adhere, cf. Pret. fealh, Goth, filhan. See Sievers,. § 387, 2, 2. A.S. feores, Gen. < feorhes,' cf. Nom. feorh, O.H.G. ferah, A.S. seoles, Gen. < seolh + es,* cf. Nom. seolh, O.H.G. selah^ O.N. selr. 5. In the following words eo is from eo by compensation for the loss of AJSievers, § 222, 2, and note 1.) : A.S. Eomar (proper name) cored, troop < eoh, horse. 'This lengthening is not limited to tiie Gen., but appears in all the cases having vocalic endings. An Historioal Study of the e- Vowel. 17 A.S. Pleowald (proper name) < pleoh, danger. A.S. weobud, weofud, altar < * wiohbed, cf. Psalter, North, ■wibed < * wihbed. 6. Syncope of '8 before I after a short vowel also causes compen- satory lengthening of the preceding vowel (Sievers, § 202). Thus a arises in a few words. A.S. mffil beside poetical mse^l, cf. Goth. ma)>], O.S. mahal, O.H.G. mahal. A.S. mffilan beside mse^lan, ma'Selian, cf. Goth. ma)>Ijan, O.S. mahlian. A.S. stselan beside stse'Slian staiSelian < sta^al. See Sievers, § 50, n. 1. A.S. stielwierSe, stalwart, cf. A.S. sta"(Sol. See st^lan. It is late before syncope of g before 1 occurs (Sievers, § 214, 2). A.S. snffil for snsegl. 7. The last kind of lengthening to be considered is that which takes place before certain consonant combinations. But here there seems to be no uniformity, the widest discrepancy existing between one text and another (cf. Sievers, § 124). This phenomenon oc- curs before a nasal or r and I -\- a, consonant, as nd, rd, mb, Id, etc. It is to be noted that the earliest and most constant examples appear before a nasal or a liquid + a sonant, and further, that these are quite rare till late West Saxon (Sievers, § 124, notes 1, 2). Few cases are found either in the Cura Pastoralis or in the Orosius, but in the later W.S. texts, as Aelfric's Homilies, etc., and the Lindis- farne and Rushworth Gospels, and the Psalter, abundant examples occur. The beginning of this process Kluge (cf. Paul's Grundriss, I, p. 866) would place in Primitive Anglo-Saxon. Tfiis kind of lengthening has led to the development of many e's in Anglo- Saxon, both open and close, examples of which are here added. Before nd : A .S. ende beside regular ende. See Durham Gospels. A.S. sendan beside regular sendan. See Matthew's Gospel. A.S. geendod beside regular geendod. See Aelfric's Horn. Before Id : A.S. teld besides regular feld. cf M.E. feld. A.S. seldan besides regular seldan. cf M.E. selde. See Kluge, P.G. I, p. 879. 18 An Historical Study of the e -Vowel. So A.S. geldan, sceld, weldan, etc. See Kluge, P.G. I, p. 879. Before r^, rn, rd : A.S. eoriSe beside regular eorSe. A.S. leornian beside regular leornian. So A.S. meord, reord, cweorn, eornost, beside regular meord, reord, cweorn, eornost. Kluge (P.G. I, p. 879) would cite A.S, sengan, feng, teng, as examples of lengthening of e before ng, and Sweet (H.E.S. § 398) the following isolated cases : eft (cf. Chronicle), esne (cf. Durham Gosp.), stefne (cf. Luke's Gospel.) 2. Oontraction. 1. This is the second of the causes which led to the development of the vowel under consideration. We first note here as the result of contraction the e's of the Reduplicated Frets., which e's are of course close, arising as they do, from the contraction of e (mid- front-narrow) with other vowels. A.S. het,i Pret. of hatan, cf. O.H.G. hiaz; O.S., O.Fris., O.N. show contraction, het. A.S. let, Pret. IStan, cf. O.H.G. liaz ; O.S., O.Fris., O.Norse show contraction, let. So A.S. dred, red, seed, slepj lee, etc. 2. Contraction in many cases is due to a phonetic law by which in primitive Anglo-Saxon syncope of certain medial consonants, as h, and occasionally J and w, occurred. These laws in general with which we ^e concerned may be found in Sievers' Grammar, the par- ticular references to which are given below. Examples of such contraction are as follows : A.S. ten, tene < * te(h)en, cf. O.H.G. zehan, O.S. tehan. A.S. twentig < *"twejentig. See Kluge, P.G. I, 879. A.S. ece < * ejyci = * ajuki. See ibid. cf. Goth, ajukduj^s. 3. The Psalter shows e as the result of contraction in the fol- lowing words (cf. Zeuner, § 15, 3) : •These are Keduplicated Perfs. contracted thus : * hfihat ■) * h^it t * hehet ■) Anglian heht., W.S. het. So the remaining forms are the result of similar contraction. An Historical Study of the e-Vowd, 1& Gefe, Pres, Sulj. gefeon < * gefehe. So PI. gefen < * gefehen. Gese, gesen, flen also result in a similar way. So ofsle, sles, sle-S, Pres. Sing. 1st. Subj., 2nd. and 3rd. Indie, < * sleho, * slehis, etc. 4. A.S. ea results from the contraction of W.Germ. a with o or u. Cf. Sievers, §§ 111, 173, 218. A.S. ea, water, < * a(hw)u, cf. Goth, ahwa, O.H.G. ahwa, O.S. aha, etc. So the compounds ear, < * ahwur, etc. A.S. ear, ear of grain, < * a(h)ur, cf. O.H.G. ahir, Lat. acus, Kluge: ohre. A.S. gefea < * gefa(h)o, cf. Goth. fahe]?s, O.H.G. gifeho. A.S. tear < * ta(h)ur, cf. O.H.G. zaher, Goth, tagr, O.N. tar (Noreen, Altn. Gr. 110), O. Lat. dacruma, Gr. SaKpv. Here belong the contract verbs : A.S. flean < * fla(h)on. A.S; lean < * la(h)on, cf. O.H.G. lahan, O.S. lahan. A.S. slean < * la(h)on, cf. Goth, slahan, O.H.G. slahan, O.S. slab an. A.S. ^wean < * ^wa(h)on, cf. Goth, thwahan, O.H.G. dwahan, O.S. thwahan. The results of these verbs exhibit contraction also, as flea < * fla(h)u ; lea < * la(h)u, slea, 'Swea. We encounter some cases of contraction which cannot be justi- fied. Such a case is, for instance, the Pres. Subj. slea which does not admit of derivation from the basic form (=: Goth, slahai). Cf. Sievers, § 110, note 2. Such forms, therefore, must be by analogy to other forms, as slea Pres. Subj. to the Indie. Pres. 5. A.S. ea also results from the contraction of Germ, e, W.G. a, with 0, u; the a being shortened to a (Sievers, § 112). The con- traction then is the same as the preceding case. A.S. brea < * bra(w)u, cf O.H.G. brawa. See Sievers, § 118, note 2. A.S. clea < * cla(w)u/ cf. O.H.G. chlawa. A.S. hreat * hra(w)o. cf. O.H.G. ro (Gen. rawes), O.N. hrar. ' The Epinal Gloss. 29, shows the intermediate step, clawo. 20 An Historical Stvdy of the e -Vowel. A.S. pea < * pa(w)o < Ace. Lat. pavonem. of. O.H.G. pfawo. Kluge, P.G. I, 311. A.S. nean < * na(h)un. cf. O.S. nahor. So near ■< * na(h)or, A.S. streaw < * stra(w)o. cf. O.H.G. stro (Gen. strawes). A.S. "Srea < * ^ra(w)u. cf. O.H.G, dro, droa (Gen. drawa). In A.S. wea beside wawo, the ea represents earlier au which is a contraction of a (=Germ. ai) with u (o) (Sievers, § 62, § 118, note 1). cf. Goth, wai, O.H.G. wewo. 6. In the North. Dialect and the Psalter ea appears as the con- traction of W.G. e with a (from o). Cf Lindelof," § 24, 2. Zeu- ner,^ § 20, 2. Gefean to rejoice < * fe(h)an. cf. O.H.G. feho, O.S. gifehon. Gesean < * se(h)an. cf O.H.G. sehan, O.S. sehan. Feas, PI. of feh < * fe(h)as. cf W.S. feos, O.S. fehu. Flean to flee < * fle(h)an. cf Sievers, § 166, 2. Tean to draw < * te(h)an. See ibid. 7. ea ' also occurs in the Psalter, and in Mercian generally, as the contraction of "W.G. I with a. See Zeaner, § 20, 2. Frea beside freo < * fri(j)a. cf. O.H.G. fri, Goth, freis. Gefreas 2nd. Sing, gefrea (Imper.) gifreo'S < *gefrigan. 8. A.S. eo appears as the contraction of W.G. e with o, u (Sievers, § 113). The quality of this eo resulting from contraction is the same as the original A.S. eo, that is, close, the sounds which pro- duce it being close. A.S. seo, pupil < * se(h)o. cf. seon. , A.S. sweor < * swe(h)or. cf O.H.G. swehur, Goth, swaihra, OiS. swehur. A.S. tweo < * twe(h)o. cf O.H.G. zweho, O.S. tweho, twehon. Here belong the contract verbs of Ablant Class V. A.S. gefeon < * fe(h)on. So 1st. Sing, gefeo < * gefe(h)u. A.S. pleon < * ple(h)on. So 1st. Sing, pleo < * ple(h)u. A.S. seon < * se(h)(w)on. So 1st. Sing, seo < * se(hw)u. 9. A.S. eo arises from the contraction of W.G. i, i, with o, a, v, (Sievers, § 114). This eo is, of course, a close sound, arising, as it does, from close vowels. ' Die Sprache des Rituals von Durham, Helsingfors, 1890. * Die Sprache des Kentischen Psalters (Vespasian A. I.). Halle, 1881. An Historical Study of the e -Vowel. 21 A.S. feal < * f i(h)ol. cf. O.H.G. fihala, f ila, Lai. pictor, I.E. "/pik. A.S. freo < * fri(j)a, cf. Psalter frea, Goth, freis. A.S. freond < * fri(j)ond, cf. Goth, frijonds, Pres. Partcl. frijon. A.S. gleo, < * gliu(j)o-, Poetical, cf. W.S. glieg glig, Epinal Gl. gliu. A.S. heo < * hiu(j)o-, Poetical, cf. W.S. hiw. North, hiu. A.S. ■Sreo < *«ri(j)u, cf. Goth. )>rija, O.H.G. driu. A.S. beo < *bi(j)a, cf. O.H.G. bia, North, and Merc. bia. A.S. bleo < * bli(j)oj cf. O.S. bli. See Sievers, § 247, note 3. A.S. beon < * bi(j)an, cf. Kent, Ch. 37, 6, bion. See Kluge, P.G., I, p. 879. A.S. beot < * bi(h)at, be(h)at, cf. O.H.G. biheiz. See Sievers, §43, §114,2. A.S. code (Pret. gan) < * ijede (* ija + de, cf. Goth, iddja), cf. Sk. fiyan. See Feist, 309 ; Kluge, P.G., I, p. 879 ; ten Brink, Z. f. d.A. 23, 65. A.S. neol < * ni(h)old, cf. Corpus nihold ;* Epinal and Erfurt, nihol. A.S. freols < * fri(h)als, cf. Goth, freihals. See Sievers, § 43, note 4. Here belong the contract verbs of Ablant Class I. A.S. leon < * li(h)on, cf. O.H.G. lihan, Goth, leihwan, O.Fris. lia, A.S. seon < * si(h)on, cf. O.H.G. sihan, O.N. sia. A.S. teon < * li(h)on, cf. Goth, gateihan, O.H.G. zihan, O.S. lihan. A.S. wreon < * wri(h)on. So the presents : leo < * lihu, seo < * si(h)u, teo < * ti(h)u, etc. The Imperatives are probably by analogy to the above forms (Sievers, § 84, note 1). They are leoh, seoh, etc. 10. A.S. eo results from contraction in the following isolated A.S. eow, cf. O.H.G. iu. A.S. eo here represents Germ. iw. A.S. ceo < * chyae, cf. Epinal Gloss., 240, chyae.^ A.S. reo < * ryhae, cf. Epinal Gloss. 1020 ryae ; " 1080 ryhae ; * 1081 ryhae ; ^ Corpus 2128 ryee,^ etc. ^Here neither syncope nor contraction has taken place. * Here contraction has not taken place. 3 22 An Historical Study of the e -Vowel. A.S. bitweonum < * twihnaim, cf. Goth, tweihnai. See Kluge, P.G., I, 879. A.S. eo like ea, it is to be noted, absorbs the following vowel (Sievers, § 119, 2), as appears from these examples : A.S, fleon < * fleu(h)on through fleo(h)on. A.S. teon < * teo(h)on < * teu(h)on, ct. Lat. duco < douco. So the Indie, and Opt. Pres. fleo < * fleo(h)u < * fleuhu ; teo < * teo(h)u < teuhu ; fleo < fleo(h)e < * fleuhe ; teo < * teu(h)e. It is worthy of note that in the very oldest texts as the Epinal and Corpus Gloss., syncope has not uniformly taken place, and contraction consequently does not always appear. In the later texts, however, these laws are carried out with perfect regularity. 3. Umlaut. This important phenomenon in Anglo-Saxon is defined by Sie- vers, (§ 85) as follows : " Umlaut, in Germanic Grammar, denotes those mutations of an accented vowel which are caused by a vowel or semi-vowel (j, w) of the following syllable." Of the several kinds of umlaut the i- and palatal umlaut are those with which we are particularly concerned. These umlauts constitute the class which partly assimilates the basic vowel to the following sound, u- and o- umlaut the class which diphthongizes the basic vowel. This mutation is illustrated in the i- umlaut of 6 to e. The 6 is a mid -back-narrow-round vowel (long), and the i which causes the umlaut is a high-front-narrow vowel. When this i follows 6 in an A.S. word, as *d6rajan, the 6 becomes fronted to a mid-front- narrow-round (long) oe which form is not uncommon in the early non-W.S. texts. Later this 6e is unrounded, and the resulting vowel is e, the mid-front-narrow (long).^ It will be readily seen that a remarkable partial assimilation of o to i is represented in e as the result of umlaut. Umlaut e is, therefore, a close sound. The date of this umlaut has been relatively placed between the palatalization of A, g, on the one hand, and the common W.G. i- syncope on the other. Cf. Brate, P.B.B. X, 29 ; Kluge, P.G. I, 870. Pogatscher (Q. u. F. 64, 132) holds that the process was ' The steps of the process as illustrated by *d6mjan are ; * domjan ^ doeman t deman. An Histonccd Stvdy of the e - Vowel. 23 complete about the year 700. Palatal umlaut hardly dates from pre-historic times. 1. i- Umlaut. 1. A.S. se arises from the i-umlaut of a, Germ. ai. This sd is an open sound, viz. the low-front-wide, coming from a low-back- wide. Anglian as well as W.S. shows se here (of. Lindelof, § 16, 1 ; Brown, § 58). A.S. ffigen, cf. Goth, aih (< aigan), O.H.G. eigin, O.S. egan. A.S. ffig. See Sievers, § 290. n., cf. O.H.G. ei, O.S. ei, Germ, type aijas. A.S. Eenig' (< an, cf. Sievers, § 90), cf. O.H.G. einig. See Kluge : ein. A.S. clSne, cf. O.H.G. kleini. See Kluge : klein. A.S. diel (i-stem), cf Goth, dails, O.H.G. teil, O.S. del. A.S. fsege, cf O.H.G. feigi, O.N. feigr. A.S. fffimne, cf. O.S. fenmea, fehmia, O.Fris. famne. See P.B.B. 14, 245. A.S. fffitt, cf O.H.G. feitit, feizit, O.N. feitr. A.S. fllesc," cf O.H.G. fleisk, O.S. flesk. Germ, base flaiski-. A.S. hiel (i-stem), cf O.H.G. heil, O.N. heill, Goth, hails, Germ, hailaz. See Kluge : heil. A.S. hffilan « hal, cf. Sievers, § 90), cf Goth, hailjan, O.H.G. heilan. A.S. hsest, cf Goth, haifsts, O.H.G. heist, O.N. heift. A.S. hlffider,cf. O.H.G. leitara (older * hleitir). See Kluge : leiter. A.S. Isedan, cf O.H.G. leitan, O.S. ledian. See Kluge : leiten. A.S. Iffifan, cf O.H.G. leiban, O.S. far-lebian, O.N. leifa. A.S. lleran (< A.S. lar), cf Goth, laisjan, O.S. lerian. A.S. llest, cf Goth, laists, O.H.G. leists, O.S. gilesti, O.N. leistr. A.S. sffi, cf Goth, saiws, O.H.G. PI. sewi. Germ, base saiwi-. A.S. stSger, stsegl, cf O.H.G. steigal, Goth, staiga, steigau. Kluge : steig. See Feist, 535 ; Kluge, Nom. Staumb., § 189. ' In Kent, e occurs sporadically beside se as the result of this umlaut. Cf. Dieter, | 6. 2., Sievers, ^ 151. This may be due to the general tendency of Kent, to narrow the W.S. S. So are found : enig (of. Chart. 38, 11), delan (cf. Chart. 45, 31, 37, 31 ; 41, 31), flese, nest, cf. Chart. 42, 22 ; 38, 12, etc., etc. For further examples see Sweet, O.E.T., p. 593 seq. 24 An Historical Study of the e - Vowel. A.S. stSnan, cf, A.S. stan, Goth, stains, O.H.G. stein. A.S, ie (for * ai, cf. Seivers, § 173) cf. O.H.G. ewa, O.S. eo, Germ, base aiwi-. A.S. hrse (for * hrai, cf. ibid.), cf. O.H.G. hreo, O.S. hreo. Germ, base hraiwi-. A.S. rSran (< ras, Pret. of risan), cf. Goth, raisjan. So gSst, gS^, Imper. gffi beside ga < verb gan. 2. A.S. ffi arises from the i-umlaut of a Germ, e (= W.G. a) as A.S. laewan, nSgan, etc., cited above. 3. A.S. ffi arises from the i-umlaut of Lat. a. Cf. Sievers. § 57. A.S. Islden, Latin < Lat. Latinus. See Sievers, § 57, 1. A.S. radio < Lat. radicem, cf. O.H.G, rateh. Kluge, P.G., I,, p. 311. 4. In the Psalter ie ' occurs as the i-umlaut of Germ, a before I -\- a, consonant (Zeuner, § 4. 2. a). Sides, cf V.P., 88, 39 ; 77, 21 ; 70, 9 ; 91, 11, etc., cf. W.S. eald, yldo. mffiltan, cf. V.P., 21, 15; 57, 9; 74, 4; 147, 18, etc. W.S. meltan. wSlle, cf. V.P., 11, 2 ; 8, 13, etc. W.S. welle. cwSlman, cf V.P., 36, 14 ; V.Hy., 4, 12. W.S. cwelman. cffllf, cf V.P., 105, 19 ; 28, 6 J 68, 32, etc. W.S. cealf. 5. A.S. e arises from the i-umlaut of 6 representing Germ, and' W.G. 0. This holds good for all the dialects. In the very oldest texts, however, we find the unrounded form (oe). This then appears most frequently in Kentish and Northumbrian, rarely in Mercian or late W.S. (Lindelof, § 18. 4. II). This oe as well as the e is a close sound. Examples are very numerous. A.S. bee. Gen., Dat. Sing, and Nom., Ace. PI. of boc, cf. O.S. Ace. PI. bok. A.S. ben < Prim. Gen. boniz. See Sievers, § 269. A.S. brec, trousers, Gen. Dat. Sing, and Nom. Ace. PI. of broc. See Sievers, § 284. A.S. breiSer, Dat. of broiSer, cf O.S. Dat. broder. A.S. eecel, cf Epinal Gloss, coecil, O.H.G. chuohho. ' I write lie long (ft) because the sr of the Psalter probably always denotes the- long vowel (cf. Sievers, 2 151, 1, note). An Historical Study of the e -Vowel. 25 A.S. cene, cf. O.H.G. chuoni, O.N. koinn, Germ, base konja-. A.S. cepan, Germ, type * kopjan, Kluge, P.B.B., 8, 538 ; P.G., I, 897. A.S. deman < (A.S. dom), cf. O.S. adomian. A.S. efstan « ofost), cf. V.P., 69, 2, cefesta; 77, 33, oefestung. A.S. fet, Nom., Ace. PI. and Dat., Instr. Sing, of fot. Prim. Germ, fotiz. Sievers, § 281. A.S. fifflere (< flor). See Kluge, Nominale Staumbildungsl., 197. A.S. feran, gefean (< * forjan < Pret. Sing., for), cf. O.S. forian, Corpus fceran. A.S. gled (i-stem) gleed, cf. O.H.G. gluot, O.S. glodwela. A.S. gerefa, cf. O.H.G. ruova, Prim. Germ, -rofjon. A.S. hre^ (i-stem), cf. HroiSgar, HroSmund, Hro^ulf, O.N. hroiS. A.S.meder (Dat. Sing, of modor), cf. Anglican mceder. A.S. meiSe, cf. O.S. mo^i, O.H.G. muodi, O.N. modr. Prim. Germ. moiSja. A.S. ondresn (< * ondrosni < * ondrotsni). A.S. recean, beside reccean, cf. O.S. rokian, O.H.G. ruohhjan. A.S. secan, cf. Pret. sohte, O.S. sokian, Goth, sokjan. A.S. sped (i-stem), cf. Anglian speed, O.S. spod, O.H.G. spuot. A.S. steda, cf. Anglian stoeda, O.H.G. stuota. Germ, type stodjon, A.S. swete, cf. Anglian swcete, O.S. swoti, O.H.G. swuozi. Germ, base swotja-. A.S. wregan, cf. Anglian wroegan, O.S. wrogian, O.H.G. ruogen. A.S. bledan, cf. A.S. blod, German type * blodjan. See Kluge, P.G., I, 897. A.S. felan (< foljan), cf. O.H.G. foljan, O.S. gifolian. See Kluge, P.G., I, p. 897. The non-W.S. dialects, especially the Kent., generally write the unrounded ce as glred, soecan, coelan, doeman, been, ce^el, broeiSer, mceder, speed, f oet, bdec, swcete, etc. The dominant form in Mer- cian (cf. Brown, Die sprache des Rush. Gloss, u. Merc. Dialekt, § 49) is e. For further examples see Sweet, O.E.T., p. 647, seq. 6. A.S. e arises from the i-umlaut of 6 before nasals represent- ing W.Germ. a < Germ, e (Sievers, § 68). A.S. bremel, Germ, base brSmila-, cf. A.S. brom. 26 An Historical Study of the e -Vowel. A.S. cwen, cf. Goth, quens (i-stem), O.S. quan, Sk. -jani-. A.S. gecweme, cf. O.H.G. biquami. See Kluge, bequem. A.S. geteme, suitable, cf. O.H.G. gizami. A.S. weu, cf. O.H.G. wani, wan, Goth, wens, O.S. wan, wanian. In the non-W.S. dialects, beside e the unrounded oe is often found, as cwren, cwoeman, woenan, etc. 7. A.S. e arises from the i-uml. of 6 which represents older on, Germ, an (Sievers, § 94). For the secondary lengthening of this vowel, see Sievers, § 185. The older texts and the non-W.S. dia- lects, in general, quite frequently show the unrounded de beside e. The quality of this e, like the remaining umlaut e's, is close, since the open o, on being lengthened in consequence of the loss of the nasal before the surd spirants, became close 6. A.S. ehtan (< oht). M erc, re htan (V.P., 17, 38), Germ, type * anhtian, * anhto, I.E. l/angh. See Kluge : acht, A.S. e'Sian (cf. Merc. pnoeiSung, V.P., 17, 16) cf. A.S. £eSm, O.N. anda, ond (Gen. andar). A.S. fehst, fehth, 2nd, 3rd Ind. of fon. See Sievers, § 94, c. A.S. ma (cf. Corpus, 787, 840 ; V.Hy., 5, 6 : fce'Sa, Germ, base * fanjyjo. A.S. ges (< * gosi < * gonsi < * gansi), cf. O.H.G. gans. Germ, ghans-, A.S. gese'San beside geso^ian (< BO'S), cf. Goth, sunja, sunjis^ Germ, base sun(d)j6, I.E. sont-, Lat. sonticus in Leg. XII Tab. Henry's Gr. § 123. A.S. ne^an, cf. O.S. nadian, O.H.G. nendjan, Goth. *-nan)?jan. A.S. sefte (cf. Adv. softe), cf. O.S. safto, Goth, samjan, O.H.G. samfti. A.S. sme'Se (Corpus, 1610, V.Hy., 12, 7, smcethu). Germ, base, smonpja-. A.S. te« (< t6«i- < toniSi- < tan-Si-, cf. Kluge : zahne, cf. Goth. tun))us, O.H.G. zant, O.S. taud. Germ, base, tan]>-, tunj>-' < I.E. dont-, dnt-. See Kluge. A.S. esa (< 6s- God, Sievers, §§ 66, 281), cf. O.H.G. Ans-, O.N. aso,- As-. See Grimm's Teutonic Mythology ed. by Stallybrass, p. 25. 8. It is to be borne in mind that e occurs sporadically in W.S. as the i-umlaut of ea (< Germ. au). Cf. Cosijn, Alt. W.S. Gram., An Histmieal Study of the e -Vowel. 27 p. Ill, § 97. The regular W.S. form is ie, later i, y. This spor- adic e for regular ie, i, y, must be due to Kentish ' influence, in which dialect the regular umlaut of ea is e. This class of words is of special importance, for some of them continue and become fixed in Middle English, and are preserved in Modern English. Such, for instance, is heran. W.S. degal beside regular W.S. diegal, digal, dygal. W.S. geman, hehst, henan, heran, hersumian, herman, lefan, gele- fan, nedan, gefleman, alesan, etc., beside regular W.S. gieman, etc., hihst, etc., hiran, etc., lifan, etc., etc. 9. In the Kent, dialect e appears as the regular i-umlaut of ea (= Germ, an) and eo (= Germ. eu). The Kent. Glosses sometimes show y. eo, however, is generally free from i-umlaut in the non- W.S. dialects. Mercian regularly exhibits e beside sd and ei, which are of rare occurrence. In North, e appears beside ei as the i-um- laut of ea. The quality of this umlaut is doubtless close, since the first element of the diphthong ea from which it arises goes back to an original open sound, which, as we have seen, is represented in the non-W.S. dialects by the close e. henan (cf. W.S. hynan, etc.). Cf. Corpus Gloss, gehende, Sweet, O.E.T. heran (cf. W.S. hieran, etc.). Cf. Chart, 48, 3. Sweet, O.E.T., p. 174. cegan (cf. W.S. cigan). See V.P., 17, 4 ■ 14, 2,4, etc. gelefan (cf. W.S. gelyfan) ; hlet (cf. W.S. hliet) ; ned (cf. W.S. nied). So leg flame over against W.S. lieg, lig, etc. 10. A.S. eow arises in a few words from the i-umlaut of au, older awi (Sievers, § 73). A.S. eowde, cf. Goth. awe);i, O.H.G. ewit. See Feist, 68. A.S. eowestre, cf. Goth, awistr (< * awi + wistr. Feist, 68), O.H.G. ewist. A.S. eowu, cf. above. Germ, base awi-, cf. Lat. ovis. A.S. meowle, cf. Goth, mawilo, diminutive of mawi. A.S. streowede, cf. O.H.G. strewita, Goth, strawida, Pret. of straujan. 1 Or to Mercian, in which dialect likewise the umlaut of ea is e. 28 An Historical Study of the e- Vowel. 2. Palatal Umlaut. 1. This umlaut, though of secondary importance in W.S., is quite extensive in its operation in the non-W.S. dialects. Its operation in W.S. is not till late. In late W.S. e appears as the palatal umlaut of ea. Here the following palatal in prox- imity to the open diphthong ea causes a simplification and narrowing, so that, with a slight raising, a long mid-front- narrow vowel (e) results. Before h (x) : L.W.S. hehsta beside regular W.S. heahst, hiehst, hyhst. L.W.S. neh beside regular W.S. neah. See Sievers, § 101. So teh, 'Seh beside regular W.S. teah, 'Seah. See ibid. Before g and c : L.W.S. beg, ege beside regular W.S. beag, beah, eage. L.W.S. been, cec, gelec, to ecan beside regular W.S. beacm, ceac, galeae, eacan. 2. Likewise in late W.S. e appears after the palatals g, c, so, for the regular ea (Sievers, § 102). It is to be noted that this um- laut rarely occurs in an open syllable. L.W.S. ces beside regular W.S. ceas, Pret. of ceosan. L.W.S. ger beside regular W.S. gear, cf. Gen., Dat. PI. geara, gearum. So L.W.S. gesced, get, ongen, seep, scet, geton for regular W.S. gescead, geat, Pret. of geotan, ongean, sceap, sceat, Pret. of sceotan, geaton, Pret. PI. of geotan. 3. In the non-W.S. dialects e' appears as the palatal umlaut of ea and eo. It is hardly necessary to state that this e is a close sound. Cf. Zeuner, § 19, III, Dieter, § 22, II; 26, II, Lindelofj § 22, III. The Kentish Glosses, however, generally retain the ea and eo before palatals (Sievers, §§ 163, 165). * In the very oldest documents, as the Corpus, Erfurt Gloss., etc., and in the Durham Bitual and R', x occurs sporadically for this e before palatals, hseh, cf. Corp., 1960, tSg, bseg, cf. Corpus, 1339 ; randbseg, cf. Erfurt, 1156 ; So, cf. Kent Chart., 39, 9, 14, 27, etc. ; Durham Ritual, 7, 6 ; E', 24, 27 ; tSseh, cf. Durham Ritual, 48, 9, etc., R', 15, 20. So gsec, belsec, etc. For further examples see Sweet, O.E.T., p. 613, ff. An Historical Stvdy of the e - Vowel. 29 N011-W.S. beg, cf. W.S. beag ; non-W.S. legan, cf. W.S. leogan ; non-W.S. ege, cf. W.S. eage; non-W.S. flegan, cf. W.S. fleogan; non-W.S. ec, cf. W.S. eac ; non-W.S. neh, cf. W.S. neah. So non-W.S. heh, ]>eh, leht, etc., for W.S. heah, ]>eah, leoht, etc.* 4. Palatalization. 1. The palatals g, c, s6, when initial, have a singular effect on the primary palatal vowels se, S, e, converting them into ea, ea, ie, respectively. In a similar position the palatal semi-vowel j unites with the vowels a (se) and o to form gea, geo (Sievers, § 74). Thus arises in W.S. ea from the union of the palatals g, 6, sc with se (= Germ. e). It is to be noted that the non-W.S. dialects do not diphthongize the vowel after these palatals, g, 6, sc, but retain the a, which these dialects represent by e (Sievers, § 157, 2; 391, 2), The diphthong results in W.S. from the development of the glide when the tongue passes from the position of the palatal to that of the vowel, from high to low.^ W.S. geafon (for * gsefon), cf non-W.S. gefun. W.S. geaton (for * gseton), cf. non-W.S. getun. W.S. scearon (for *scieron). W.S. sceap (for * scSp), non-W.S. seep. So ciese, cyse, umlauted form of * ceasi for * cffisi < Lat. caseus. 2. So ea results from the combination j (Germ, j) with ie (Sievers, § 74) in the following words : W.S. gea for * g», cf. O.H.G. ja. See Kluge : ja. W.S. gear for * gser, cf. non-W.S. ger, cf Goth, jer, O.H.G., O.S. jar. We have now completed the survey of the chief causes which led to the development in Anglo-Saxon of the vowel under considera- tion. But there remain to be considered two less important sources from which a few e's were developed in Anglo-Saxon. The first of these is Latin from which words were imported into Anglo- Saxon ; the second is the influence of w. *R' often shows eo unchanged in these words (Brown, § 63). *The sporadic sceoc, sceod, sceot, etc., for scoc, scod, etc., may be placed here (Sievers, | 76, 1). 30 An Historical Study of the e-Vowd. 5. Importations from Latin. 1. si arises in A.S. from Lat. a (Sievers, § 57). A.S. nap, turnip, < Lat. napus, cf. O.N. niepa. See Kluge, P.G., I, p. 310. A.S. striet, street, < Lat. strata (via), cf. O.H.G. strazza. See ibid. A.S. strlegl, strael, corpet, < Lat. stragulum. See ibid. A.S. sseternesdseg, Saturday, < Lat. saturni dies, cf. Fris. sater- dei. See ibid. 2. ea ^ arises in A.S. from Latin au in the following words : A.S. eeac, basin < Lat. caucus, cf. O.Irish cuach. See Kluge, P.G., I, 309. A.S. ceas, strife < Lat. causa, cf. O.H.G. choso. See ibid. 3. eo arises in A.S. from Lat. in the following words : A.S. deofol < Lat. diabolus, cf. O.S. diubul, O.H.G. tioval. Kluge : teufel. A.S. preost cannot be placed here since it is < Lat. presbyter, or is a mutilated form as * prester, cf. O.H.G. priester, O.S. prester, O.N. prestr. See Kluge, priester ; Pogatscher, Q. u. F., 64. 6. Influence of w. 1. The combinations aw and ew originally generated a u between the vowel and the w ; the auw and the euw thus formed passed regu- larly into eaw, eow (Sievers, § 73, 1). A.S. feawe PI. cf. Goth, fawi, O.N. far, Lat. paucus. A.S. sceawian, cf Goth, usskawjan, O.S. skawon. A.S. cneowes, Gen. of cneo, cf. O.H.G. knewes, Gen. of chniu, Goth. kniu. A.S. treowes. Gen. of treo, cf. Goth, triu, Gen. triwis. 2. In late W.S. eo arises by the transformation of final ew into eow (Sievers, § 113, note ; § 371, note). So result L.W.S. fleowS for regular flew^, spreowS for regular W.S. spew^. 'ea in A.S. Greocas, Creocas (< Lat. Grseci, Greci) beside A.S. Grecas, Grecas appears to be the palatalization of Lat. se, late Lat. e, before c. See Kluge, P.G., I, p. 310; Pogatscher, Q. u. F., 64, 5. An Historical Study of the e - Vowel. 31 It may be noted that in the Northumbrian, W.S. eow was often reduced to ew (Sievers, § 156, 6 ; Lindelof, § 22, IV). So result fewer, gitrewia, cf. L.W.S. feower, gitreowia. We have now seen what was the original stock of the e-vowel (both open and close;, and of the diphthongs ea and eo. "We have further seen how this original stock was very largely increased by the development of e (open and close) and ea and eo, on native soil ; and we have seen that this development was brought about by four chief causes, viz., various' kinds of lengthening, contraction, and umlaut, and the influence of palatals, and by two subordinate causes, viz., the importation of Latin words and the influence of w. With this survey then we are prepared to consider the problem before us at another and more advanced stage of the language. But before turning to the Middle English period, we desire to draw attention more particularly to some of the phenomena hitherto merely men- tioned which late W.S. exhibits. Such are the reduction to the monothong e of the diphthongs ea and eo under the influence of the following palatals h (x), g, c, and, also, under the influence of the preceding palatals, g, c, sc ; and e as the i-umlaut form of ea, as degol, geman, heran, etc. These phenomena are found in all the late W.S. texts. Others are found only in special texts.' Some of these exceptional forms of which many might be cited are : ded for died, forbseran for forberan, efre for sefre, trewe for treowe, wsel- hrewan for wselhreowan, blseda for bleda, secere for ecere, twsegen for twegan, sece for ece, hele for hlele, helend for hSlend, ^era for iSffira, ondredende for ondrffidende, sec for eac, 'Ses for ^ees, reddon for rseddon, feringa for fseringa, weron for wSron, hselle for helle, ewe's for cwse'S, etc., etc. These forms, which are, of course, of sporadic occurrence, do not seem to be governed by any fixed or definite law. They appear rather to be the first indications in late W.S. of the setting in of a tendency which reached its climax in Middle English, where order and method are found. In view of the changes which in Middle English have taken place in the vowel-sounds, special significance ' The texts from which these exceptional forms are taken are the Homilies of Wulfstan, herausg. von Arthur Napier, Berlin, 1883 ; and the Harrowing of Hell from the A.S. version of the apocryphal Gospel of Nicodemus as represented in MS. Cotton Vitellins, A, XV. (C, the Beowulf MS.). 32 An Historical Study of the e - Vowd. seems to attach to these exceptional late W.S. occurrences as show- ing the tendency of the language at that period of its history. MIDDLE ENGLISH PERIOD. IIL The e- Vowel in the Several M.E. Dlalects. 1. General Remarks. In Middle English there are four distinct dialects which repre- sent, in the main, the continuation of the Anglo-Saxon dialects. These dialects are designated in Middle English the Northern, Midland, Southern and Kentish. Of these the Midland admits of further division into East Midland, which is quite closely related to the Southern, and the West Midland, which is related to the Northern. Whenever these dialects were geographically connected, they were not so sharply differentiated, and so admixture of forms resulted. This confusion of dialects is especially noticeable in late Middle English. If we compare Middle English texts with Anglo-Saxon texts, we observe marked differences. Now these changes did not all take place simultaneously, but gradually. Soine of the dialects changed much more rapidly than others. The Northern, for example, changed far in advance of the others. The Southern, on the con- trary, was very conservative. The earliest standard M.E. text which shows the changes which Anglo-Saxon vowels have under- gone is the Ormulum, written probably before 1200, in the East Midland dialect. But even in this text the changes are not fully developed. We shall have occasion later to recur to this text when we shall consider it at length. There are two principles in Middle English which must not be overlooked, and these we now consider at the outset. They are lengthening and shortening before certain consonant combinations. The process of lengthening before certain consonant combinations was recognized, as we have seen, in Anglo-Saxon. But in Middle An Historical Study of the e - Vowd. 33 English this process is far more extensive in its operation. As in Anglo-Saxon, so in Middle English, the consonant combinations that cause secondary lengthening in the preceding vowel consist of vowel-likes -|- a consonant which must be voiced, as Id, rd, nd, etc. (Kluge, P.G., I, p. 8€6, § 83; Sweet, H.E.S., § 636). Orm does not lengthen the vowel before such combinations as nc, nt, etc., as the writing drannc, drinnkenn, sallt, stunnt, hellpenn, etc., shows. The lengthening of the Infinitive and Participle of many verbs is by analogy to the monosyllabic Pret. Compare fand whose Infin., if formed regularly, should be finndenn ; whereas by analogy to the Pret. it is findenn. So with the Partcl. fundenn. Those verbs not having a monosyllabic Pret. retain their short vowel, as senndenn Pret. sennde, Ptcl. sennd ; brinngenn, brohhte, etc. This secondary lengthening is regularly prevented by back- shortening. So, Orm writes clene, but clennsenn. Another kind of lengthening peculiar to Middle English is that which takes place in open syllables. But it must be borne in mind that only certain vowels, viz., a, e, o, lengthen in open syllables, the high vowels i and u not being subject to this lengthening in open syllables. So* we find in M,E. faren, etan, hopen, over against A.S. f^ran, 6tan, hSpen. It is difficult to determine the date of the beginning of this process. At the time of the Ormulum (1200) the process has certainly begun. Something quite like it appears sporadically in late A.S. texts. We meet in Aelfric's Homilies such forms as herian, hafenleast, hSfene, for the regular herian, hafenleast, haefene; and in the Gospels etan eta's (ic) ete, etc., for etan, etc. (cf. Mark), all of which Sweet (H.E.S., § 392) has pointed out. Perhaps too much stress is not to be laid on these sporadic occurrences in late Anglo-Saxon, but they appear to be significant as showing the beginning of a tendency which reached its climax in Middle English. The Ormulum seems to represent the trans- itional stage, for Orm writes the same words frequently both long and short, as hete beside h6te, mete beside m6te, mele beside mSle, etc. The symbol (") is frequently written over a vowel in an open syllable, as late, h6te, ISdeun, tSkeun, sp6keun, etc., which would seem to indicate a recognized tendency to lengthen the vowel in open syllables. At all events it appears then that the beginning of this process must be placed at least as early as the Ormulum (1200). 34 An Historical Study of the e -Vowel. As there is lengthening in Middle English before certain con- sonant combinations, so there is shortening also before certain consonant combinations. Long vowels in Middle English are shortened before certain combinations of consonants, as st, ht, tt, dd, etc., etc. This process is especially exemplified in the domain of the weak Prets., and of compounds, as fedde, ledde, Pret. of feden, leden, laffdig < A.S. hliefdige, etc. It is noteworthy that Orm often preserves the original length before st, as brest, prest, sest, laest, etc. This process, too, according to Kluge, P.G., I, p. 868, seems to date from A.S. times. M.E. samdead points back to A.S. sam (with short a) and not to sam, since A.S. a would give M.E. o, except in the Northern dialect. So A.S. seo)j)jan, sy);]>an beside si)>J» points back to short i, sij^J^an, and not to long i, sijjjjan, because only short i is subject to breaking. The exam- ples of this process of shortening in Anglo-Saxon are, of course, of very rare occurrence. We may assume the date of this process to be the 11th and 12th centuries. These two processes affect materially the problem before us, the one largely increasing, the other somewhat diminishing, the stock of e's in Middle English. The most marked of the M.E. vowel changes is the simplification of the A.S. diphthongs ea, ea, eo, eo. This change was generally carried out in all the dialects except the Kentish, which preserved the old diphthongs. This must have been a gradual change. It is found in the most important old transitional document, the Peterborough Chronicle (Laud MS.), written between 1132 and 1164. Traces of it, as we have seen, appear in late A.S. documents. The process is still incomplete in the Ormulum, the diphthongs ea and eo and the monothongs e and e existing side by side. By the time of Chaucer the diphthongs have all been simplified in the dialect in which he wrote. There existed in M.E., as in A.S., two distinct long e's, one of which was an open sound, the other a close sound. These two sounds are continued from A.S. and are referable to different sources, as was shown. They are sharply differentiated in M.E., but are sometimes found combined in rime. The following is a tabular view of the more important sources of the M.E. close and open e : An Histofrical Study of the e -Vowel. 35 e appears in M.E. from the following A.S. sources : 1. A.S. e, i-umlaut of Germ, a : cwen (< * cwani), wen (< * wani). 2. A.S. e, i-umlaut of 6, Germ. 6 : secan, swete. 3. A.S. e, i-umlaut of ea, Germ, au : non-W.S. heran, lefan. 4. A.S. eo corresponding to Germ, eu : deor, freosan. 5. A.S. e before Id, etc. : feld, eld. e appears in M.E. from the following A.S. sources : 1. A.S. ea corresponding to Germ, au : leas (Goth, laus), eage, Goth, augo, dream. 2. A.S. S corresponding to Germ, se, W.G. a : sad, dsed, msel. 3. A.S. se, i-uml. of a: laran, ISdan, clsene. 4. A.S. e in open syllables : mete, etan, specan. But it must be noted that both 2 and 3 gave in M.E. not e ex- clusively, but also e. The reason for this apparent confusion in M.E. is obvious. It will be remembered that these sounds had different values even in A.S. The W.S., for example, preserved the Germ, sd (low-front- wide), while the non-W.S. dialects changed it to e (mid-front-narrow). Again, the i-umlaut of a is in W.S. ie, but in the Kentish dialect, on the contrary, it is e (mid-front-nar- row). So then we see that these A.S. sounds are preserved in M.E., thus causing an apparent confusion between the open and close sounds of long e. As the foregoing table shows, A .8. ae and ea regularly became e (low-front-wide) in M.E. The ea after the simplification had taken place was preserved in M.E., for, as we saw, this sound in A.S. though written ea was really an open sound ieo. Orm repre- sents this, by the symbol se with perfect regularity, as does, also, Layamon frequently. Of course in these earlier M.E. texts ea is also fou!id beside se and e. This ie (low-front-wide) was probably later in M.E. raised to the mid position. The A.S. e (mid-front- narrow) is retained in M.E., and with this M.E. e, coincides the e which arises from A.S. e by lengthening before Id, etc. A.S. eo is simplified, and the resulting e (mid-front-narrow) is pre- served in M.E. In the earliest M.E. texts eo occurs beside the simplified e. A.S. e of whatever origin became M.E. e in open 36 An Historical Study of the e - Vowel. syllables. See Ten Brink, Chaucer's Spr. u. Versk, § 11. See further, however, Kluge, Geschichte d. Eng. Spr., § 96 (Paul's Grundriss, I., p. 878). As to the graphic representation of the two sounds of e in M.E. there is no uniform distinction. Orm, however, distinguishes be- tween them, writing for the open sound uniformly se, and for the close sound, e. In the earliest Southern dialect (cf. Ancren Riwle) ea is used, but not regularly, to represent the open sound, and in the West Midland sb, but not exclusively nor with perfect regu- larity, e being also employed to represent the same sound. In the other dialects no graphic distinction is made between the open and the close sound, the symbol e being used for both. In late M.E. the same is true of all the dialects, with this difference that ee is also employed. Having taken this preliminary survey and dealt in a general way with the present problem in M.E., we are now prepared to enter upon the detailed proof of the sources of M.E. e and e. The best method for accomplishing this end appears to be that furnished by a careful examination of the texts of each of the M.E. dialects. We have accordingly chosen representative texts of each of the dialects and examined them, noting from what sources M.E. e and e arose. The texts selected for this investigation are as follows : Kent. : Dan Michel's Ayenbite of Inwyt, ed. by Morris (Early Eng, Text Soc, 23) ; South. : Ancren Riwle, ed. by Morton (Cam- den Soc.) ; East Mid. : Ormulum (reed, by Holt from White's ed.) ; West Mid. : Layamon's Brut (ed. by Madden) ; North. : Richard Rolle de Hampole's Pricke of Conscience, (ed. by Morris, Phil. Soc. of London, 1863). Chaucer of course is examined as representing the dialect which gave rise to standard English. , 2. The East-Midland Dialect. One of the most important M.E. texts is written in this dialect and deserves, for obvious reasons, especial investigation. It is, of course, the Ormulum. The best edition of this text yet issued is that by Holt (Oxford, 1878), which is based on White's (Oxford, 1852). It is unnecessary to make any strictures on the unsatisfac- An Wstorieal Study of the e - Vowel, 37 tory condition of Holt's text on which, with but slight reference to White's edition and Kolbing's collation (Englische Studien, 1, 1), the following investigation is based. se (low-front-wide vowel). 1. A.S. « (low-front-wide) corresponding to Germ, ss is retained in the Orm. as ss or is represented by e, which probably is the close sound. The e's are doubtless due to Mercian influence, in which dialect the W.S. open se sound was represented by the close e. sed seed, cf Merc, sed, v. 5070, 15905. ssed does not occur, dede deed, cf. ded, 2267, 4723, 7817, 7918, etc. dsed does not occur, nedle needle, cf. Merc, netl, 6341. needle does not occur, bsere bier, cf. W.S. bser, 8167, 8172. here does not occur, tselenn to blame, cf. W.S. tielan, 2033, 2039, 2040. hser hair, cf. W.S. hier, 3208, 9223. here does not occur. €fenn evening, cf. Merc, efen, 4353, 8910. sefen does not occur, rsed, cf. W.S. rsd, 12464, 12502, 18719. rede does not occur, rsedenn, cf. W.S. rjedan, 17286, 18336. reden does not occur, slsepenn, cf. W.S. sliepan, 7483. slepenn does not occur, slsepe, cf. W.S. slsep; slep, cf. Merc, slep, 3143, 3152. drsedenn, cf. W.S. drsedan; dredan, cf. Merc, dredan, 1218, 5907, 11493. Iffitenn, cf W.S. Isetan; letenn < Merc, letan, 2017, 9058, 10157. Here belong the Pret. Pis. and Subjs. of Ablaut classes IV. and v., as bffirenn, 14046, bsere, 2029, 2047. 2. A.S. £6, i-umlaut of a, is retained in the Orm., e being some- times found which is perhaps to be referred to Merc. e. Iserenn, cf. W.S. lleran, 1781, 6129, 6215. leren does not occur, dffilenn, cf. W.S. dselan, 6175, 8326. delenn does not occur, del, cf. Merc, del (?), 1722; dale, 8270, 8273. hffile, cf W.S. hffil, 5378, 15503. hele does not occur, hselenn, cf W.S. hSlan, 2528, 3047. helen does not occur, flash, cf. W.S. flsesc, 3524, 3652. flesh does not occur, clsennesse, cf. W.S. clien, 1194, 4594, 4598. clennesse does not occur. 4 38 An Historical Study of the e -Vowel. dene, cf. Merc, clene, 1592, 2447, 2932. clsene does not occur, ledenn, cf. Merc, ledan (?), 938, 1612. Isedenn does not occur, gredisnesse, cf. Merc, gredig (?), 10210, 10213. So grediglis, 12280. sel, cf. Merc, sel (?), 14304. ssel does not occur. 3. A.S. ss (= Latin a and its i-umlaut) is preserved in the Ormulum, as in strsete from A.S. strset. The list of examples given under 1 and 2 exhibits some variation as to the employment of se and e. For W.S. as (= Germ, le) the Ormulum shows usually se, which of course is an open sound. Sometimes e is found which is probably to be referred to the non- W.S., or to be more specific, to the Merc, e, a close sound. So se representing A.S. se (== i-umlaut of a) is usually preserved in the Ormulum, but e is occasionally found. It will be remembered that in the Merc, dialect this sound is sometimes represented by e, though generally by se, as in the other dialects. See Brown, p. 69. These e's found in the Ormulum then are probably the survivals of the exceptional Merc, e's, and represent a close sound. In general, it may be said that se in Orm. is long (Klu'ge, P.G., I, 882) and indicates an open sound, while e may be either long or short and indicates a close sound. ea. Open Diphthong. 1. A.S. ea of whatever origin becomes se in Orm, which of course has the same value as the se already discussed, since after simplifi- cation it represents the identical open e-sound (S) of A.S. sere, cf. A.S. eare. Dedication, 133, 309. hsep, cf. A.S. heap, 4230; Ises, cf. A.S. leas, 14208. dsef, cf. A.S. deaf, 9887, 15500 ; Isefess, cf. A.S. leaf, 13737. drsem, cf. A.S. dream, 923 ; dsew, cf. A.S. deaw, 9883. Vsew, cf. «eaw, 3169, 6754. hsewenn, cf. A.S. heawan, 9285, 9965. fiewe, cf. A.S. feawe, fea, 424, 13013. shsewenn, cf. A.S. sceawian, 251, 958, 1041. se, river, cf. A.S. ea, 7091. tsem, oflFspring, cf. A.S. team, 2415. tseress, cf. A.S. tear, 13849. An Historical Study of the e- Vowd. 39 Exceptions. Two forms ec and eshe occur which must be from Merc, ec, ege. Otherwise, A.S. ea before palatals must have already become close e. Shep and ger beside gser occur which must be from Merc, e forms, as seep, ger, and not from W.S. sceap, gear, which would give shsep, gser. e (mid-front-narrow vowel). 1. A.S. e of whatever origin is retained in the Ormulum, and of course represents the long-mid-front-narrow sound, her, here, here, <[ A.S. her (of frequent occurrence). So belong here the Prets. of the reduplicating verbs which show e as let -< A.S. let, etc.,' etc., and the Prons. he, me, we, etc. " Here fall further the umlaut e's, as kene < A.S. cene, 16139, 19962; kepenn < A.S. cepan, 17978; cwen, fet, smej>e, sekenn, gledess, etc., etc. herenn,^ 923, 5377 ; gemenn, lefenn, etc. eo, Close Diphthong. 1. A.S. eo of whatever origin is simplified to e (mid-front-nar- row) in the Ormulum, or remains unchanged. See, however, Hale's Art, Mod. Lang. Notes, VIII., p. 21 . deor beside der < A.S. deor, 1177, 1312. j7eode beside pede < A.S. -Seod, 1762, 3436. seoc beside sec < A.S. seoc, 6165, 8073. wheol beside whel < A.S. hweol, 3639, 5897. fleon beside flen < A.S. fleon, 2934, 4144. seon beside sen < A.S. seon, 318, 2449. freo beside fre < A.S. freo, 2968, 8015. freond beside frend < A.S. freond, 1609, 2367. fend < A.S. feond, 5554, 12354. sede < A.S. eode, 119, 136, 413. deofell beside defell < A.S. deofol, 671, 3832. preost beside prest < A.S. preost, 219, 338, 461. ' slsepenn has gone over to the weak Conj. in the Pret. and shortened the e, so that it appears as sleppte, slepptenn ; drsedenn and rsedenn have foUowed the same course. * These must be from the Merc, e, i-umlaut of ea. 40 An Historical Study of the e-Vowd. Lengthening. 1. Lengthening before certain consonant combinations produces the following cases of S and e in the Ormulum : serd^ < A.S. card, 1394, 1416, 2153. fljerd < A.S. gefleard, 7334, 10027. gserd •< A.S. geard or O.N. garSr, 15254. swerd < A.S. sweord, 6639, 8146. werdenn < A.S. werden, 5185, 5189, 6255. ferd, brerd, etc. < A.S. brerd, 14040, 14792. So sern^ < A.S. earn, bsern < A.S. beam, ernenn < A.S. eornan, lernenn < A.S. leornian, cherl < A.S. ceorl, eorless < A.S. eorl, erpe eorpe < A.S. eoriSe, feld < A.S. feld, weldenn < A.S. weal- dan, geldenn < A.S. geldan, elde < A.S. eld, beldenn < A.S. beldan, well < A.S. well, genge < A.S. geng, ende < A.S. ende, wendenn < A.S. wendan, sendenn < A.S. sendan, etc., etc. 2. In open syllables. In the Ormulum e may of course repre- sent either the long or short sound. See Anglia, VII., Anz., 95, ibid, VIII., 196. brekenn < A.S. brecan, 1547; swerenn < A.S. swerian, 4480; spekenn < A.S.specan, 2733, 3389 ; wrekenn < A.S. wrecan, 914, 19900; fele < A.S. fela; tredenn< A.S. tredan; etenn < A.S. etan ; hefenn < A.S. hebban, etc. Shortening. The process of shortening explains 6^ in the following words where we should expect e : fell Pret. of fallen < A.S. feol, 620, 897, 2505. blettcenn < A.S. bletsian, 2298, 4826. clennseun < A.S. Adj. clene, 1126, 4028, 4463. So fedde < A.S. fedde, Pret. of fedan ; keppte < A.S. cepte ; demmde < A.S. demde ; wesste < A.S. westen. • Lengthening before rn, rd seems to have taken place before ea was simplified to a, as sem, serd, etc., show. 'The regular shortening of » is a, as in wraj>>e < A.S. wrte«, lasse (A,S. Isessa), lasstenn, laffdig, etc. The same is true of ea, as in shadde, shadd, etc. The steps are probably these : STseia, eaieaia. That is to say, se and ea were short- ened to se and ea, respectively, which coincided with all A.S. k's and ea's, and these of course regularly became M.E. a. An lEstorieal Study of the e-Vowd, 41 This process is especially exemplified in the comparatives, as derre, etc., and in the Prets. of weak verbs. In a few verbs the Prat, is formed on the Pres. that exists in Orm. and does not rep- resent the A.S. Pret., as weppte, ledde, etc. Old Norse Influence in the Ormulum. The Old Norse element* in Orm. is quite important and cannot be ignored since it contributed quite a large number of words con- taining e, e, some of which still survive in Living English. 1. O.N. e appears in O. as e (mid-front-narrow) : he]7enn, hence, c£ O.N. he&n hsej^an, 15570, 16092. hej>ennwarrd. ser, separate, cf. Icel. ser, 18653, seq., 18678; serlepess, 513, 519, 573. 2. O.N. SB apart from i-umlaut appears in O. as e : )>e]jeun, thence, cf. Icel. J>a];an, O.N. ]78e];an, 1098, 7492. So l7e}>ennforth, thenceforth, 10786, 11180, 12930. whe]?enhwarrd, whence, cf. Icel. hva'San, 16668, 17292. leshe, hire, 6234 ; leshemenn, 6222 (cf. Icel. leiga), show the later redaction of the O.N. ei to e. — Brate. 3. O.N. ffi, i-uml. of 6 appears in O. as e : epe];]?, cries v. < O.N. oepa (< * vcepa, cf. A.S. wepan), 9198, 9562. sleh, cunning < O.N. sloegr. Icel. slaegr, 13498. fere, sufficiency, cf. O.N. unfcerr, Icel. foer, 1251, 4429. seme]j]>, seems, cf. Icel. sama, somme, A.S. seman, Ded. 66. 4. O.N. iu, io become e in O. mec beside meoc < O.N. mjukr, 667, 1258. So the compounds meoclike, meoclessc, and the verb mekenn sket < O.N. skjotr, soon, 1266, 2297. 5. O.N. e, i- and R-umlaut of a appears in O. as S, which must represent the open sound. gsetenn, to direct, preserve, cf. Icel. gseta, 2079, 3765. gsetelses, careless. See above, 6190. ssete, seat, sitting, cf. Icel. sseti, 11959, 11972, 14007. 1 Cf. Brate, P.B.B., X, 1 seq. ; Kluge, P.O., I, p. 791. 42 An Historioal Study of the e-Vowd. hsej^enn, v. to scorn, cf. Icel. hae'Sa, to scoff at, 13682. hffij>elis, scornfully, cf. Icel. hsej^iligr, 7408, 11593. hsepinng, scorn, cf. Icel. hsej^ing, 240, 4876, 7391. ■wselinng < * wselen, cf. O.N. vela, to deceive, 2192. wengess, wings, cf. Icel. vsengr, 8024, 16433. These e's and e's from O.N. coalesce with the e's and e's from A.S. and share their subsequent fate. It will be seen then that se which is always long in Orm. regu- larly represents the open sound, and e the close sound of long e. eo is also used for the latter sound, but does not differ at all from e. In the later East-Midland texts the open sound is sometimes graphically represented by e. The examples of lengthening and shortening given for East- Midland may serve for Middle English in general, so that it will not be necessary to multiply examples of these phenomena in each of the remaining dialects. So much then, for Orm. and the East-Midland dialects. We pass on now to Layamon as the next author. 3. The West-Midlamd Dialect. In this dialect, especially in the text we consider, the distinction between the open and the close e is not carried out with so perfect regularity as in the Ormulum. In general, we may say that se and ea in Layamon represent the open e-sound, but not always, since the same sound is sometimes represented by e simply. The close sound is generally represented by e. In the later MS.* of Layamon e is frequently used to represent the se of the earlier MS. se. se in Layamon represents A.S. ie generally, and A.S. ea, e, and a, occasionally. 1. se in Lay. for A.S. se appears in the following : seven, A., 1. 1116 ; ser. A., 1. 372 ; selc. A., 1. 4260 ; clsene. A., 1. 6290 ; fl»sc, A., 1. 25986 ; Iseche, A., 1. 17679 ; Isefdi, A., 1. 1256 ; dseden, A., 1. 10440 ; rsed. A., 1. 1648 ; ssele. A., 1. 10527 ; mseste, A., 1. 'A is used to indicate the older MS. and Bthe younger. An BHaUyrical Study of the e - Vowd. 43 27482 ; wsepne, A., 1. 1702 ; sse, A., 1. 1093. In Eedupl. verbs : slsepe, A., 1. 733 ; Isete, A., 1. 8662 ; and in the Pret. PI. of Abl. olass IV. and V: spseken, A., 1. 13444; »ten, A., 1. 13444, etc. Of course e also occurs here. 2. se in Lay. represents A.S. ea : daeth, A., 1. 20676 ; drsem, A., 1. 8649 ; hseh, A., 1. 7094, beside heh; dsed, A., 1. 2569, beside dede; hsep. A., 1. 816, beside hep; Isesinge, A., 1. 3068, beside lesinge ; nseh, J;aeh, etc. So the Pret. Sing, of Abl. class II. : bsed, A., 1. 2454; chses, A., 1. 12175 ; Ises, A., 1. 637, etc., etc. 3. se in Lay. represents A.S. a : seiS beside a«. A., 1. 704, 4340 ; brseiS beside brad. A., 1. 14, 219 ; clsetS beside cla«. A., 1. 22878 ; gsere beside gare. A., 1. 15225 ; hsel beside hal, A., 1. 1252, etc. Conversely, Lay. sporadically represents A.S. se by a : flasc be- side flsesc, clane beside clsene, aven beside seven, rad beside rsed, A., 1..4411 ; baren beside bseron, B., 1. 26483, etc., etc. 4. se in Lay. in the older MS. occasionally represents A.S. e : hset for A.S. het ; wseste. A., 1. 9932, for weste ; hser for her. A., 1. 3164, etc., etc. ea in Lay. represents A.S. ea, and occasionally A.S. se, e, and eo. 1. ea in Lay. represents A.S. ea : dead, A.B., 1. 3911 ; dream, A., 1. 1010; deathe, A.B., 1. 1733 ; seare. A., 1. 2671 ; cheap. A., 1. 385; heap, bead, B., 1. 9023; leas. A., 1. 6931, B., 1. 21251, ■etc., etc. 2. ea in Lay. represents A.S. S : cleane, B., 1. 6290 ; deal, A., B., 1. 2995; deade, A.,B., 1. 6895; heale, B., 1. 23072; read, A.,B., 1. «91 ; sea. A., 1. 123, eaven. A., 1. 19570, etc. 3. ea in Lay. represents A.S. e rarely : deam, B., I. 24250, for deme ; cweane, B., 1. 24555, for cwene, etc. 4. ea in Lay. in MS. B. rarely represents A.S. eo : deap, B., 1. €47, for deop ; weap, B., 1. 25533, for weop ; leame for leome, etc. e in Lay. represents A.S. e, ea, eo, and occasionally A.S. as. 1. e in Lay. represents A.S. e of whatever origin : demen. A., 1. 11074, 19979 ; queue. A., I. 43, 182 ; weste, A., 1. 1124, 25662 ; blessinge, A., 1. 13261 ; kene. A., 1. 11499 ; grene, A., 1. 24652 ; mete. A., 1. 18127 ; her. A., 1. 21 ; wene. A., 1. 18752 ; swete. A., I. 19797, etc., etc. 2. e in Lay, represents A.S. se occasionally, or perhaps non- W.S. e : er. A., 1. 1581 ; del. A., 1. 2951 ; clene, A., 1. 32078 ; 44 An Historical Study of the e - Vowel. dede, A., 1. 21072; leche, A., 1. 17769, B., 1. 19746; slepe, B., L 610, etc. 3. e in Lay. represents A.S. ea : chepmon, A., 1. 30681, B., 1. 13313 ; bred, A., 1. 31800 ; ded, A., 1. 28317 ; deth, A.B., 1. 285 ; drem. A., 1. 24554, B., 1. 1009 ; heh, A., 1. 559 ; hepe, lesinge, neh, etc., etc. So the Pret. Sing, of Abl. class II, bed, beh, fleh, les, etc, 4. e in Lay. represents A.S. eo, though eo is frequently retained : breste, A., 1. 6497; der, A., 1. 313; depre. A., 1. 15901 ; fre. A., 1. 5225 ; frend. A., 1. 5531, B., 1. 3705 ; lef. A., 1. 30361, B., 1. 344 ; thede. A., 1. 25464 ;— Pres. of Ablant class II : crepe. A., 1, 29313 ; fleghe, etc. ; Prets. of Redupl. verbs ; cnew, B., 1. 3501, fellen^ held, hewe, etc. eo in Lay. rep. A.S. eo : deop. A., B., 1. 21997 ; deor. A., B., 1. 6438 ; feond. A., B., 1. 237 ; freond. A., B., 1. 708 ; leof, A., B., 1. 585 ; freo, A., B., 1. 361, etc. So belong here Pres. of Ablant class II : beode, cheose, leose, etc. ; Redupl. Prets. : greowen, wex, weoxen, etc. In a few words eo and ea are confused in Lay. : heop for heap^ B., 1. 16308; leos for leas, B., 1. 6^31, etc. Such then is the result of the investigation in Layamon. It is very clear that, while the open sound is represented generally by se and ea, and the close sound by e and eo, there is yet much con- fusion. In the later West-Midland texts se and ea are represented by e simply, so that no graphic distinction is made between the open and the close sound of long e. 4. The Southern Dialect. In the Ancren Riwle we meet only three graphic representations of the long e-sound, viz., ea, e, eo. Of these ea represents the open sound and eo the close sound, while e represents both the open and the close sound. 1. ea in A.E. represents A.S. ea of whatever origin : ear, pp. 90, 100, 104, etc, ; eadi, pp. 28, 142, 146, 154, etc. ; lease, p. 268 ; leasunge, pp. 82, 424; dream, pp. 210, 214; leave, p. 230; bileave, p. 280; dea];, p. 6; dead, p. 112; adeaden, p. 112; scheawen, pp. 90, 98, 154, 292, 344 ; reaven, pp. 68, 96, 286, 300, 396 ; team, pp. 216, 288, 336 ; ];eaw, pp. 88, 158, 200, 240, 276, An Historical Study of the e - Vowel. 4& etc.; wean, pp. 80, 108, 114, 156, etc.; strea, p. 295; lo]>\eas, p. 188 ; weamode, pp. 118, 134, etc., etc. 2. ea in A.E. represents A.S. se : eaver, p. 4 ; leafdi, p. 4 ; cleane, pp. 4, 8 ; dealen, p. 224" 248, 400 ; leareti, p. 64 ; healen, p. 330 j tear, p. 64 ; read, pp. 6, 198, 224, etc. e in A.E. represents : 1. A.S. e of whatever origin : her, pp. 236, 240, 246, 262, 264^ etc.; cwene, pp. 88, 170, 296, etc.; cwemen, pp. 138, 192, 238, etc.; fere, pp. 114, 152, 206, 284, etc.; sechen, pp. 164, 274; gleden, pp. 122, 406 ; kene, pp. 130, 140, 190, 272 ; kepen, pp. 156, 190, 286, 332, etc.; vet, fet, pp. 122, 166, 274, etc. ; wenen^ pp. 62, 64, 106, 128, etc. ; smepe, pp. 4, 184, etc., etc. 2. A.S. S, though ea also occurs. This e is doubtless an open sound. Examples are: leten, pp. 6, 8, 12, 102, 268, etc.; dreden beside dreaden, pp. 222, 428 ; slepen, pp. 238, 270, 272, 212, etc. ;. dedbote, pp. 206, 348, 372, etc. ; her, pp. 10, 130, 382, etc. ; leren^ pp. 108, 114, etc. ; del, p. 276 ; delen, pp. 28, 38, 248, etc. ; hele, pp. 300, 312 ; dene, pp. 154, 164 ; leche, pp. 178, 182, etc., etc. 3. A.S. ea, though ea is generally retained (unsimplified). This e representing A.S. ea must also be an open sound. Examples are not numerous, ec, pp. 168, 236, 240; chepman, p. 418; temen^ pp. 220, 288, 308 ; bed, Pret., p. 270, etc. 4. A.S. eo. This is a close sound. Examples are rare since A.S. eo is retained in A.R. in the unsimplified form, sec, seke, pp. 176, 370 ; comp. seccure, p. 46 ; derewur]; beside deorwur];,. p. 98. ' eo in A.E,. represents : 1. A.S. eo of whatever origin, and of coijrse is a close sound, deore, pp. 190, 329, 408; deorewur);, pp. 98, 102, 106, 112; deorling, p. 336 ; freo, p. 370 ; fleon, vleon, pp. 132, 134 ; eode, pp. 52, 152, 128, 220, 260, 318; deofle, pp. 84, 232; fleotende, fleose, deop, feoUe, leof, dreor, breost, ];eode, vreond, heold, weop, etc., etc. There was a tendency in the Southern dialect, as well as in the Kentish, to preserve the old diphthongs, which, of course, retained their original values. 46 An JSistorioal Study of the e -Vowel. 5. The Kentish Dialed,. In the Ayenbite of Inwyt and in the Kentish dialect in general the preservation of the old diphthongs ea and eo * is remarkable. Here the open e is represented by ea and e, the close e by io and e. 1. ea in the Ayenbite of Inwyt represents A.S. ea. Occasion- ally there is produced before the ea, y, which is a kind of glide, dyea]?, pp. 72, 86, 87, 129, 130, etc. ; year, p. 71 ; stream, pp. 72, 98 ; beleave, pp. 72, 101, 106, 123 ; tyeares, pp. 74, 83, 96, etc.; great, pp. 75, 76, 77, etc.; cheas, p. 77; leasinges, p. 77, dyead, p. 79 ; veawe, pp. 162, 254 ; ]>reapneth, pp. 79, 162, etc. ; beat, beag, eare, heaved, heap, etc. 1. e in the Ayenbite of Inwyt represents A.S. e of whatever origin. wepej>, pp. 71, 73, 83, 93, etc. ; benes, pp. 74, 90, 98, etc.; deme, pp. 74, 76, 82, etc.; wenej;, pp. 21, 22, 82, etc.; zvete, pp. 82, 83, 92, etc. ; made, pp. 42, 90, etc. ; zeche, pp. 94, 98, etc. ; herye];, pp. 59, 92, 177, etc. ; her, p. 88 ; quene, queme, bedes, gret, het, te}>, vet, weri, etc, 2. e in the Ayenbite of Inwyt represents A.S. se of whatever origin. But it must be remembered that in A.S. the W.S. se was represented in the Old Kentish dialect by e. So then this e in the Kentish is really a close sound, since it represents the retention of the old Kentish e (= W.S. m). mest, pp. 72, 89, 90, 92. etc. ; teche, pp. 73, 78, 96, etc. ; clene, p. 73 ; clenliche, pp. 76, 88, 92, etc. ; yclensed, p. 73 ; drede)?, pp. 74, 75, 84, etc.; lede pp. 76, 44, 96, 116, etc; dele, p. 90; del, pp. 17, 86, etc. ; lessa, pp. 76, 90, etc ; speche, pp. 89, 90, 91 ; zed, pp. 95, 143, 216, etc. ; dedes, blest, bleve, brede, e)7en, hele, here, let, rede, etc. 3. e represents the old Kentish e before z, where W.S. exhibits ea. This e, standing for old Kent, e (mid-front-narrow), is, there- fore, the close vowel. 656, pp. 19, 75, 81 ; hesnesse, pp. 89, 97, 102, etc., etc. 4. e (ie, ye in Dan Michel) represents the old Kentish io. mede, myede, pp. 63, 73, 95, 142, etc.; tyene, pp. 31, 66, 124, etc.; Vyef, ^yeve, pp. 37, 79, 263, etc. ; ]?yester, pp. 45, 264, 266, ' A.S. eo appears in Old Kentish almost exclusively as io. An Historical Study of the e - Vowd. 47 276, etc.; bryest, pp. 175, 247; cryepe, p. 107; dyepe, pp. 105, 211, 264; lief, lyef, pp. 117, 213; lyerne, lierne, pp. 72, 73, 99, etc. ; chiese, chyese, p. 86, etc., dyevels, dievels, etc., pp. 77, 82, 83, 86 ; viend, veyend, pp. 75, 79, 117, etc. ; vryend, vriend, vrend, pp. 30, 67, 79, etc. It will be seen that most of the e's in the Kentish dialect are close, and that the open sound is generally represented by ea. We take up next Richard Rolle de Hampole's Pricke of Con- science, the text on which the investigation in the Northern dialect is based. 6. The Northern Dialed. In the Northern dialect, especially as represented in the Pricke of Conscience, the most important document, we find only one symbol used to represent both the open and the close e. This dialect, always in advance of the other M.E. dialects in exhibiting vowel-changes, has reduced all the A.S. diphthongs to monothongs. The majority of the e's we have to deal with are close, coming as they do, from the old Northumbrian dialect which generally repre- sented the W.S. open S by e (mid-front-narrow). e of the North, dialect arises from various sources and represents : 1. A.S. se of whatever origin : drede, 11. 291, 294, 328, etc. ; dede, 1. 2484; red, rede, 11. 2014, 4303, 6085; red (verb), 11. 1677, . 3953, 6286 ; brejje, 1. 613 ; fere, 1. 2291 ; teche, reche, ete, 1. 4848 ; lere, 11. 155, 174, 175, 186, etc. ; dele, 1. 3460 ; hele, 11. 757, 1326, etc. ; lef, clene, lede, flesh e, etc., etc. 2. A.S. ea of whatever origin : ches, 1. 2132 ; grete, 11. 742, 8499, 8511, etc. ; eke, 11. 3256, 6239 ; eghe, II. 574, 2234, etc. ; ere, I. 782 ; neghe, 1. 1208 ; dede, 11. 112, 807, 815, etc. ; reve, 11. 251, 308 ; lesyng, 1. 4274 ; heghe, 11. 1872, 1887 ; deef, heed, J^ewes, etc., etc. 3. A.S. e of whatever origin : here, 11. 220, 1019, 1024, etc. ; spede, 1. 2882 ; kene, 11. 1228, 4383, etc. ; reck, rekles, 11. 5546, 5792, etc. ; spede, 11. 5, 2682, etc. ; smej^e, 1. 6349 ; mede, 1. 96, etc. ; medeful, wene, reke, queue, kepe, hede, here, etc., etc. 4. A.S. eo of whatever origin : der, dere, 11. 2978, 3019 ; J;ef, II. 1237, 5210 ; «eke, 11. 672, 798, etc. ; fer)>e, 11. 356, 1246, etc. ; flegh, 1. 4394 ; fende, 11. 36, 1253, 4160 ; frend, 11. 1116, 3941, etc. ; 48 An Historical Study of the e -Vowel. hew, 1. 4889 ; lefe, 1. 2978 ; grete, 11. 502, 5392, etc. ; dregh, drery, devel, chese, cleve, crepe, bete, etc., etc. Having completed the survey of each of the M.E. dialects we are now prepared to consider the dialects that gave rise to Standard English, especially as represented in Chaucer and in the London State and Parliamentary documents ' during the period embraced between the dates 1384 and 1430. 7. The Dialed which Gave Rise to Standard English. 1. In this dialect, as in all the M.E. dialects proper, we have the two sounds of long e. As to the graphic representation of the two sounds in this dialect, there is no distinction made, both being written e and e6, indifferently. The best MSS. of the Canterbury Tales, ac- cording to Ten Brink, usually write ee in closed syllables. In open syllables e is generally written. Exceptions, however, are numerous. e in this dialect (as represented in Chaucer and the London State and Parliamentary documents) corresponds to : 1. A.S. e of whatever origin : chese (cheese), fete, demen, seken, wenen, grete (to greet), mete, weep, wepe, slepe, swete, heng (< A.S. heng), reeve « A.S. gerefa), teeth, lete, fele, etc., etc. ; here ^ (< non-W.S. heran, cf. W.S. hieran), leeve, bileeve^ (< non-W.S. lefan, cf. W.S. liefan), etc. 2. A.S. e before consonant combinations which cause lengthen- ing : feld, feeld, sheeld, selde, seelde, etc. 3. A.S. eo of whatever origin : theef,' leef, feend, freend, seek, knee, tree, breest, deer, depe, deep, cheesen, reesen (< A.S. hreosan), ' The language of these documents has been investigated by Morsbach in his Neuenglische Schriftspraohe (Heilbronn, 1888), of whose results I have availed myself. * These forms go back to non-W.S. forms in which the i-umlaut of ea is e. 'Chaucerian texts exhibit ie occasionally for A.S. eo, io, where we should expect e, ee. Especially does this ie occur before f, as in thief, lief, etc. The value of this ie is doubtless close e, as the rime shows. The explanation of this phenomenon is possibly to be sought, not in the Southern or Midland dialect, but in the Kentish where, as was pointed out, ie, ye, regularly corresponds to A.S. eo, io. Cf. Ten Brink, Ch. Spr. u. Versk., § 39. French influence, of course, can- not be ignored in this case. See, however, Behrens, Beitrage zur der Geschichte der Franzosischen Sprache in England, p. 146 seq. — Frauzosische Studien, V» Band, 2 Heft. • An Historical Study of the e- Vowd, 4& se (< A.S. seon), be (< A.S. beon). EeduplicatiDg Prets. : heeld, leep (< A.S. hleop), etc., etc. 4. M.E. e in final position. When e in M.E. occurs in final position, it becomes close, as slee beside' sleth, slest, etc. See Ten Brink, Ch. sp. u. Versk., § 23, d. 5. A.S. », e (=Germ. m, and i-umlaut of a) occasionally : deed, drede, sleep, sleepen (verb), leche, eel, sheep, meet (suitable), mete (measure), clene, leeren, etc. Note. — These e-forms can be explained as coming from the non- W.S. dialects. e corresponds to: 1. A.S. ie of whatever origin: breeth, deel (< A.S. dsel), eer (< A.S. ffir), feer « A.S. f»r), meste (< A.S. mast), shethe (< A.S. scffiiS), threed « A.S. frSd), even « A.S. Jefen), hele, speche, mede, reden, there, where, streete, eet, etc. In final position this M.E. e becomes close. Some of the above list of words sometimes appear with e (mid-front-narrow). See above under 4. 2. A.S. ea* of whatever origin: breed (< A.S. bread), deed (< A.S. dead), deef, dreem, ere, eere (< A.S. eare), beem, heep, greet (< A.S. great), leef, cheep, chepe, deth, teer (< tear), steep (< A.S. steap), lepen (< A.S. hleapan), streem, heed, hed (< A.S. heafod), red (< A.S. read). So Pret. Sings, of Ablant class II., as creep, etc., etc. 3. A.S. e, ea in open syllables : here (< A.S. bera), beren (v. < A.S. beran), breken, dere (< A.S. deran), eren, mete (< A.S. mete), sp(r)eken, gere (< A.S. gearu), sweren, teren, weeren (< A.S. weran), etc., etc. In this dialect there are some words that appear with both e and e. It will be seen on examination that these words go back to se. A.S. ffi corresponding to Germ, se, as was shown in the section on Anglo-Saxon, is represented in W.S. by si and in non-W.S. dialects bye (mid-front-narrow). These two sounds were retained side by side in M.E., as we have seen, the W.S. se giving e, the non-W.S. ' In the London State documents e appears before palatals, from A.S. ea where we should expect e. But it will be remembered that even in A.S. this ea was narrowed to e before palatals in Anglian and in late W.S., and it is quite natural that some of these e's (mid-front-narrow) should find their way into M.E. 50 An Historical Study of the e-Vowd. e giving e. In the dialect we are now considering both these sounds appear in such words as deed, drede, reden, sleep, etc. The natural inference, then, is that the e-forms are from the W.S. sd, and that the e-forms are from the non-W.S. (Merc.) e. Such words as sheep, etc., which show exclusively e are clearly from non-W.S. forms only. Such words, on the other hand, as show e exclusively, as most of these words do, are, of course, from W.S. se. A few words from A.S. se, i-umlaut of a vary between e and e. As a rule this dialect represents this A.S. as by e, but there are a few words which show both e and e, as clene, leden, leren, etc. The above explanation holds good for these words also. The e-forms are from W.S. ae, while the e-forms are from non-W.S. (Kent.) e. This explanation is of especial importance, for it serves to clear up the difficulty attaching to a class of words in Modern English repre- sented by seed, deed, speech, sleep, etc., where we should expect the spelling ea. 2. Romance Element in Middle English. It remains now to consider the Romance Element in M.E. which furnishes quite a large number of both sounds of long e. Since this element is especially prominent in late M.E., particularly in the dialect under present consideration, it seems proper that the examples cited should be taken from Chaucer, the greatest of the late M.E. authors, and that these should be regarded as representa- tive of the entire M.E. period. The French words introduced into English at this period came from the Anglo-Norman dialect. These importations were gradu- ally changed more and more under French influence, and later were made to conform to a French model. This statement pre-supposes the well-established fact that the Anglo-Norman dialect, represent- ing the original Franco-Norman transplanted on English soil, had a development peculiar to itself and independent of that of the various other Old French dialects. These importations show both e and e. e ' represents : 1. O.Fr. e from Latin a, except when followed by 1. It is worthy of note that Latin a, whether long or short, is usually repre- ' I follow Ten Brink in this classification. An Historical Study of the e -Vowel. 51 sented in O.Fr. by an e-sound, written e or ai. The Anglo-Norman prefers simple vowels, and so we find here generally e. peer, comper (< O.Fr. per < Lat. par), sopeer, soper (O.Fr. soper), frere, freer « O.Fr. frere < Lat. fratrem), clere (O.Fr. cler < Lat. clarum), entree (O.Fr. entree), auctoritee (O.Fr. id. < Lat. auctori- tatem), degree (O.Fr. degre < Lat. de + gradum), pitee (< O.Fr. pite See Ellis, E.E.P., IV, p. 1040 ff. .74 An Historical Study of the erVowd. 2. Ee which, during the 17th century, became established as (ii), retained this sound. So then in the 18th century both ea and ee coincided, having the common sound of (ii) (high-front-narrow), except in a few words such as, bear, swear, etc. In many words ea and ee also became short under the back-shortening influence of th, d, t, etc. These sounds continued to exist till about the middle of the present century when, under the general tendency of Living Eng- lish toward the diphthonging of long vowels, they became diph- thongs. The diphthonging .of the (ii)-sound of ee and ea began with a widening of the first half of the vowel, which, by still farther widening, would develop into i (high-front-wide) with the glide slightly raised. This diphthong is written ij. It is not generally heard in American English, but may be heard in such words as, he, me, etc. The open (ee)-sound of ea in such words as swear, etc., was diphthonged into (ea), the first symbol of which indicates a low-front-narrow sound, the second a mid-mixed-wide sound, as the e in eye or better. The close (ee)-sound of ea in break, great, which, being preserved by the preceding r, did not become (ii), was diphthonged into (ei) in Living English, thus coinciding with the diphthong (ei) which the open (ee)-sound of aa, as in, name, tame, etc., etc., became. 2. The Diphthongs ai, ei. It was shown that the diphthongs ai and ei, at the close of the 17th century, had the sound of open (ee), the i having vanished. In the 18th century they have the same sound, but ei in some words became (ii). The expert Orthographist (1704) says that " ai, ei, ay, eyj are much the same sound in many words as pail, hay, eight, they." He next gives a list of ten words in which " the sound of e is lengthened by ei," by which he means (ii), as Ellis remarks. These words are as follows : conceit (s. and v.), conceive, deceit, deceive, either, inveigle, receipt, receive, weild (= wield). But he gives the open (ee)-sound to these : " con, de re, ceipt, or ceive, heir, leisure, neither, rein, reign, their, vein, height, inveigh, neighbour, weight." An Historioal Study of the e - Vowel. 75 It appears then that in the early part of the 18th century ei in some few words, especially of French origin, as conceive, deceive, etc., had been narrowed from its former open (ee)-sound and raised to (ii). Elsewhere ei and ai retained their open (ee)-sound. This (ii)-sound of ei, is also mentioned by Lediard (1725), who says that ei is sounded as German ih or ie, in conceit, conceive, deceit, deceive, inveigle, leisure, perceive, receive, seize, etc. Lediard also gives the open (ee)-sound to ei, and to ai as well, in deign, eight, feign, heir, reign, rein, air, complain, etc. Buchanan in 1766 and Franklin in 1768 agree with Lediard and the Orthographist. So does Sheridan (1780), who notes that "the Irish in attempting to pronounce like the English " and to convert ei, ey, into (ii), often overstrained the rule and pronounced prey, convey, prii, convii. This shows clearly the tendency of ei to become (ii). To sum up the changes which ai, ei, underwent in the 18th century, we may say that ai and ei retained their open (ee)-sound of the 17th century, but that ei in some words, especially of French origin, as seize, conceive, etc., was narrowed and raised to (ii). So then in the 18th century both ai and ei, except in a few words, coincided with aa, and were confused under the common open (ee)- sound. This open (ee)-sound may have been narrowed toward the close of the 18th century, but this is not certain. In Living English, about the middle of the present century, this open (ee) was diphthonged, under the tendency of Living English toward diphthonging,^ so that the old diphthongs ai, ei, which became a monothong in the 17th century have been again diphthonged. , This diphthong is ei which is distinctly heard, as in day, they, hay, clay, rain, pail, fail, eight, freight, inveigh, etc. Before r the diphthong is (ea) as in hair, their, fair, air, pair, which is the same diphthong as that of aa before r ; as, rare, fare, care, etc., and as that of ea before r, as swear, bear, wear, etc. These sounds then coincide in present Living English. The ei in such words as conceive, receipt, seize, etc., which had the (ii)-sound coincided with ee and ea and developed the same diphthong in Living English, (ij), as in seed, he, glee, seize, etc. But this diphthong is not fully developed in American English. ' This applies, of course, to long vowels only. 76 An Historical Study of the e - Vowel. A tabular view is here subjoined which shows the development of the (ee)-sound in Modern English. In the latter part of the 17th century M.E. e before r and M.E. a as in name, etc., and M.E. ai, ei, coincided, all having the common open (ee)-sound which they all retained during the 18tb century, with a possible narrowing toward the close of this period. In the 18th century ea (< M.E. e), ee (< M.E. e) and ei in such words as receive, seize, etc., coincided, all having the (ii)-sound. In the 19th century these were all diphthonged into (ij). In present English ai, ei (except before r), and aa (except before r), and ea after r (as in great) coincide, all having been diphthonged into (ei) ; and these same vowels before r were diphthonged into (ea). 16th Cent. 17th Cent. 18th Cent. 1. 2. 3. 4. 5. M.E. e > ee' (spelt ea) M.E. e > ee,' ii (spelt ee) M.E. a > aa, sese M.E. ai > ai sei M.E. ei > ei > ee (except 1 > ii > sete, «e > sese, ee > ee 19th Cent. before r) > ee (except before r), ii. > ii. > ee (possibly ee). >■ ee (possibly ee). > ee (possibly ee). 1. > Diph. ij (steal, weal) ; ea (before r, wear, swear), ia (near, rear), ei (after r, great). 2. > Diph. ij (he, me, see, sweet, seed, deed, etc.). 3. > Diph. ei (name, lame, tale, dale, etc.) ; ea (before r, ware, care, hare, etc.). 4. > Diph. ei (fain, nail, main, tail, day, etc.) ; ea (before r, fair, hair, etc.). 5. > Diph. ei (rein, eight, freight, etc.) ; ea (before r, their, heir, etc.), ij (receive, seize, etc.). 3. Development of Close (ee) from Latin M, CE. There remains to be considered one more source of close (ee) which, though in chronological order not entitled to be placed here, may yet for the sake of convenience and completeness be treated at this point. It is se and oe from the Latin which came into our language in the Modern period. During the Middle Ages Latin se was considered an e-sound and was often transcribed e. Pals- grave is the first of our phonetic authorities to mention sb and oe. He says that se and oe " be written in Latin and not sounded," by ' ee indicates the long open e. 'ee indicates the long close e. An Historical Study of the e-Vmod. 77 which he probably means not sounded as diphthongs, as Ellis observes. BuUokar, in 1580, uses se as a sign for ee, the symbol of the close sound, which fact points clearly to the assumption that SB was considered equal to close (ee). Lediard, in 1725, confirms tliis hypothesis when he says se is pronounced as German ih or ie m : sera, Csesar, Cseres, perinseus, etc., for this was the sound of close ee at that time. The facts in the case seem then to warrant the inference that Latin se and oe, whenever they were introduced into the language, were regarded equivalent to the close ee-sound with which they have up to the present been identical. VII. Eesults. The following is a summary of what, in my judgment, are the results of the foregoing investigation : 1. The Germanic open e-sound (low-front- wide) « was retained in Anglo-Saxon, especially in West Saxon, in which dialect it had a wide range, but it was much restricted in the non-W.S. dialects, being here represented by the close e (mid-front-narrow). 2. The Germanic close e was a rare sound in Anglo-Saxon, but the secondary close e which was developed on Anglo-Saxon soil by lengthening, contraction, umlaut, and palatalization, had a very wide range in all the dialects. 3. In Middle English, in the Southern and Midland dialects, open e is of far more frequent occurrence than in the Kentish and Northern, in which two dialects, in A.S. times, it was of very rare occurrence. 4. In the Ormulum and in the dialect which gave rise to Stand- ard English, close e occurs frequently where W.S. exhibits se (open). Three classes of these words are exemplified in sed, heh, heran. Such words are probably from the Mercian which exhibits close e for the W.S. se, ea, ie, respectively, in the given classes of words. 6. The sources of open and close e in Middle English are those found on page 53. 6. M.E. e is written ea in Modern English which is a tra- ditional Southern spelling ; M.E. e is written ee. Great confusion 78 A7i Historical Study of the e-Vowd. prevailed between these symbols till the latter part of the 16th century. Hence ea crept into some words, as hear, for example, where it had no right, ie was revived in the 17th century as a graphic representative of the close (ee)-sound. 7. In the 17th century M.E. e became close ee, and in the 18th this became (ii) (high-front-narrow) except in a few words, as bear, swear, there, etc., in which the open ee, on account of the conserving influence of the following r was preserved. From this it follows as a corollary that the original open ee no longer exists as such in English. 8. M.E. e in the latter part of the 15th century became ii in a few words, as he, bee, etc., and in the early part of the 17th century it became (ii) in all words. 9. Latin se and ce, whenever introduced into English had the value of close ee at the period of their introduction and in their subsequent history have shared the development of this sound. 10. Early Modern English aa became sese and in the latter part of the 17th century developed into open ee which was possibly narrowed in the latter part of the 18th. 11. Early Modern English ai became ssi and in the latter part of the 17th century developed into open ee, thus coinciding with ei, the i vanishing. This sound of ai, ei continued and had the same development as aa. But in some words ei, in the 18th century, became narrow and was raised to ii (cf. seize, receipt, etc., especially words of French origin). 12. Open and close ee in many words from M.E. times to the present, have been undergoing the process of shortening before d, th, t, etc. 'M# i«^. ;^P '^^dA M