mmsiiy) 
 
THE GIFT OK 
 
 JAMES MORGAN HART 
 
 PROFESSOR OF ENGUSH 
 
Cornell University 
 Library 
 
 The original of tliis bool< 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/cu31924027323827 
 
THE 
 
 PHONOLOGICAL INVESTIGATION OF 
 OLD ENGLISH 
 
 ILLUSTRATED BY A SERIES OF FIFTY 
 PROBLEMS. 
 
 BY 
 
 ALBERT S. COOK, Ph.D. (Jena), 
 
 FBOIUBSOB of the ENQLIBH LASSnAOB AND LlTEKATUBB IN THB 
 
 Unitebsitt of Califobnia. 
 
 BOSTON: 
 
 PUBLISHED BY GINN & COMPANY. 
 1888. 
 
THE 
 
 PHONOLOGICAL INVESTIGATION OF 
 OLD ENGLISH 
 
 ILLUSTRATED BY A SERIES OF FIFTY 
 PROBLEMS. 
 
 BY 
 
 ALBERT S. COOK, Ph.D. (Jena), 
 
 Frofessob op the English Lansuaoe and Literature in the 
 Uniyebsity or California. 
 
 BOSTON: 
 
 PUBLISHED BY GENN & COMPANY. 
 
 1888. 
 & 
 E.Y. 
 
Entered, according to Act of CoDgreBB, in the year 1888, by 
 
 ALBERT S. COOK, 
 in the Office of the Librarian of CongreBS, at Washington. 
 
 Ttpogbapht by J. S. Gushing & Co., Boston. 
 Pbesswobk by Ctihn & Co., Boston. 
 
PREFACE. 
 
 f I "'HE study of Old English Phonology, to be truly 
 valuable, must be comparative. Only thus does it 
 become a study of relations, that is, of law. Only thus 
 does the language take its place in a great system, out- 
 lined by the founders of Comparative Philology, and 
 elaborated by the patience of their successors. And in 
 no other way can the foundations of English etymology 
 be solidly laid. 
 
 That part of the system most nearly connected with 
 Old English has, as is natural, been wrought out by the 
 Germans, in conjunction with the Scandinavians and 
 the Dutch. Their zeal in the historical investigation of 
 their own languages has rendered us their debtors for 
 the understanding of our own. The most important 
 of their results are accessible in German, but only ex- 
 ceptionally in English. Hence a knowledge of modern 
 German is indispensable to the Old English investigator, 
 and will be presupposed in the following pages. 
 
 The words chosen for illustration are fifty of the 
 principal nouns and verbs found in a single paragraph 
 of Sweet's Anglo-Saxon Reader (fourth edition), that 
 
4 PBEE'ACB. 
 
 under the year 894 of the English Chronicle, beginning 
 on page 36. I have supposed a student desirous of 
 working out the phonology of such a passage in a 
 manner at once independent and systematic. Should 
 he not succeed in every instance, he would not be 
 justified in concluding that the method was necessarily 
 wrong. There are problems which, in the present state 
 of our knowledge, baffle all attempts at solution. But 
 it is advantageous that this class of problems be dis- 
 • criminated from the rest, in order that effort may be 
 concentrated upon them, and they may be forced to 
 yield, as equally forbidding ones have already done. 
 
 ALBERT S. COOK. 
 Berkelet, Cai., March 19, 1888. 
 
INTRODUCTORY. 
 
 1. The phonology of a language is a systematic 
 account of its speech-sounds, their mutual relations, 
 and the modifications to which they are subject. 
 
 2. A competent knowledge of Old English phonology 
 is the indispensable foundation of all exact scholarship 
 in that tongue. Such a knowledge of Old English pho- 
 nology is most conveniently obtained by investigating 
 the phonology of a single text, or of a limited number 
 of words. No method known is so certain to imprint 
 the facts upon the memory, or to put the student in 
 command of the principles, as that of actual investiga- 
 tion. Besides, the facts thus obtained are serviceable 
 to other students, if placed at their disposal. 
 
 3. The method employed in the phonological inves- 
 tigation of Old English is the comparative ; that is, it 
 involves a consideration of related words and speech- 
 sounds in the kindred Germanic tongues. These are 
 a) Old High German, Old Saxon, Old Frisian ; h) Old 
 Norse (Icelandic), Gothic. 
 
 4. In practice, it is more convenient to postulate a 
 single Germanic language, from which all the historic 
 dialects are descended, and to make all phonological 
 comparisons with the forms of that language, or of one 
 
6 PHONOLOGICAL INVESTIGATION OP OLD ENGLISH. 
 
 of its two grand divisions (§ 5). This hypothetical 
 language is known simply as Germanic, though some- 
 times as Primitive Germanic (Urgermanisch), Com- 
 mon Germanic, General Teutonic, or Original Teutonic 
 (Philological Society's Dictionary). 
 
 5. Linguistic science at present recognizes two main 
 divisions of Germanic, known as West Germanic and 
 East Germanic; the former includes the dialects enu- 
 merated in § 3. a, together with Old English; the later 
 comprises Old Norse and Gothic. Scholars are not 
 quite agreed as to the propriety of the term East Ger- 
 manic; but the constitution of the West Germanic 
 group admits of but little doubt. 
 
 6. It is assumed that, in the lapse of time, the primi- 
 tive Germanic tongue was divided, in consequence of 
 migrations and new settlements, into the two branches. 
 East and West Germanic : and that, subsequent to this 
 period, these branches were again differentiated into the 
 several dialects through natural variation of speech, 
 confirmed by partitions of territory and greater per- 
 manency of residence. 
 
 7. We have, therefore, the following scheme of de- 
 scent, or genealogical table ; — 
 
 Germanic. 
 
 , \ 
 
 I 
 East Germanic. West Germanic. _ 
 
 Gothic. Old Norse. Old English. Old Saxon. Old High Old Frisian. 
 
 German. 
 
 8. Since it is not always convenient to investigate 
 the phonology of a word as a whole, various elements 
 
INTEODUCTOEY. 7 
 
 or component parts of the word are often considered 
 separately. Phonology is thus divided into two 
 branches : — 
 
 1. Phonology of Vowels. 
 
 2. Phonology of Consonants. 
 
 And since the same original sounds, but especially 
 the vowels, vary according to their position, we have 
 another division, according to this principle, into — 
 
 1. Vowels of Stressed Syllables. 
 
 2. Vowels of Unstressed Syllables. 
 
 9. The investigation of Old English phonology, in the 
 case of any particular word, resolves itself into the re- 
 construction of the Germanic or West Germanic form 
 of the same word. The laws which must be observed 
 in the reconstruction of the Germanic form are the same 
 which determine the phonology of the Old English 
 word. Hence the ability to reconstruct the Germanic 
 form implies the comprehension of those laws. 
 
 10. The reconstruction of Germanic forms requires 
 access to the best grammars and dictionaries of the 
 various dialects. For our purpose, Frisian may be 
 neglected, as not absolutely necessary. The following 
 books may then be regarded as practically indispensa- 
 ble, unless there is a note to the contrary : — 
 
 a) Gothic : — 
 
 Braune's Gotische Grammatik. Halle, 1882. (There is an 
 English translation by Balg, New York, 1883.) 
 
 Gabelentz und Loebe's Ulfilas. 3 vols. (Text, Grammar, and 
 Glossary.) Leipzig, 1843-46. 
 
 Sohulze's Gotisches Glossar. Magdeburg, 1847. (Will par- 
 tially replace the last, if that cannot be obtained.) 
 
8 PHONOLOGICAL INVESTIGATION OP OLD ENGLISH. 
 
 6) Old Norse: — 
 
 Noreen's Altnordische Grammatik. Halle, 1884. 
 
 Cleasby and Vigfusson's Icelandic Dictionary. Oxford, 1874. 
 
 c) Old High German : — 
 
 Braune's Althochdeutsche Grammatik. Halle, 1886. 
 
 Graff's Althochdeutscher Sprachschatz. 6 vols. Berlin, 1834-42. 
 
 d} Old Saxon : — 
 
 Heyne's Kleine Altsachsische und Altniederfrankische Gram- 
 matik. Paderborn, 1873. (May be dispensed with.) 
 
 Heyne's edition of the Heliand. Paderborn, 1873. 
 
 Schmeller's Glossarium Saxonicum. Munich, 1840. (Either 
 this or the preceding should be accessible.) 
 
 Heyne's Kleinere Altniederdeutsche Denkmaler. Paderborn, 
 1867. (Not indispensable.) 
 
 e) [Frisian: — 
 
 Von Richthofen's Altfriesisches Worterbuch. Gbttingen, 1840.] 
 
 /) Old English: — 
 
 Here a larger number of books is desirable. In addition to my 
 translation of Sievers' Old English Grammar (2d edition, Boston, 
 1887), the following may be named, approximately in the order of 
 their importance : — 
 
 Cosijn's Altwestsachische Grammatik. Haag, 1883-86. 
 
 Zeuner's Die Sprache des Kentischen Psalters. Halle, 1881. 
 
 Dieter's Ueber Sprache und Mundart der Aeltesten Englischen 
 Denkmaler. Gbttingen, 1885. 
 
 Bosworth-Toller's Anglo-Saxon Dictionary. Oxford, 1882-87. 
 (Three parts out of four issued, to beginning of S.) 
 
 Grein's Sprachschatz der Angelsachsischen Dichter. Cassel, 
 1861-64. 
 
 Ettmiiller's Lexicon Anglo-Saxonicum. Leipzig, 1851. 
 
 Sweet's Oldest English Texts. London, 1885. 
 
 Sweet's Anglo-Saxon Reader. Oxford, 1884. 
 
 Stratmann's Dictionary of the Old English (i.e. Middle Eng- 
 lish) Language. Krefeld, 1878. (Supplement, 1881.) 
 
ZNTRODTJCTOEY. V 
 
 Matzner's Altenglische Sprachproben (Middle English), Bd. II. 
 (Worterbuch). Berlin, 1878-. (Only as far as J.) 
 
 Lye's Dictionarium Saxonico- et Gothico-Latinum. 2 vols. 
 London, 1772. (Not indispensable.) 
 
 Leo's Angelsachsisches Glossar. Halle, 1877. (Of very little 
 value.) 
 
 gy General: — 
 
 Paul and Braune's Beitrage zur Geschichte der Deutschen 
 Sprache und Literatur. 13 vols. Halle, 1874-. 
 
 Kluge's Etymologisches Worterbuch, 4th edition (now appearing). 
 Strassburg, 1888. 
 
 Kluge's Nominale Stammbildungslehre. Halle, 1886. 
 
 The Philological Society's New English Dictionary. London, 
 1884r-. (These last are important, but not indispensable.) 
 
 11. For convenience of reference, tables of the vowels 
 and consonants, with their principal correspondences, 
 are here given. In the table of consonant correspon- 
 dences, the parentheses contain an occasional form, or 
 a query, and the colon separates geminates from simple 
 consonants, or initial from medial or final consonants. 
 The dash indicates that the sound is lost (as in the final 
 position of r). The details must be learned from the 
 three grammars, the Old English, Old High German, 
 and Old Norse, upon a virtual consensus of which the 
 tables rely. 
 
 In the consonant table, the Germanic surd spirants, 
 p, f, h, s, should be carefully distinguished from the 
 sonant spirants, if, 15, y (borrowed from the Greek, and 
 appropriated to mark this distinction), and z. It should 
 also be noted that j is equivalent to English y. 
 
10 PHONOLOGICAL INVESTIGATION OF OLD ENGLISH. 
 VOWEL COEEESPONDENCBS. 
 
 Oebm. 
 
 West Gbem. 
 
 Goth. 
 
 a 
 
 g 
 
 a 
 
 e 1 
 
 03 
 QQ 
 
 l,ai 
 
 ii 
 
 03 
 
 i, ai 
 
 o 
 
 
 u, au 
 
 u 
 
 CO 
 
 u, au 
 
 & (only before h) 
 
 
 & 
 
 
 & 
 
 6 
 
 4 
 
 u 
 
 6 
 
 i 
 
 1 
 
 ei 
 
 6 
 
 
 6 
 
 6 
 
 2 
 
 6 
 
 ai 
 
 
 ai 
 
 au 
 
 CD 
 
 an 
 
 eu 
 
 iu 
 
 CONSONANT COEEESPONDENCES. 
 Germ. WG. OHG. OB. 
 
 W: WW 
 
 13 
 
 r : rr a E 
 
 1:U I O I 
 
 t-l (O <U 
 
 m : mm 'ji a O 
 
 n : nn m □ ee 
 
 t:tt ^ z:zz a: 
 
 p:pp pf:ff 
 
 k : Itk li (c, ch) : hh c : cc 
 
 ],:>J»0) >:]»> d(th):tt(dd) ^ 
 
 f:flf(?) f:flf f:v(=f) ^ 
 
 h:hh (?) h:hh h : Uh g 
 
 s : ss s : ss s : ss a 
 
 i 
 
 9(d):dd(?) d : dd t:tt =0 
 
 l9(b):bb(?) b:f:bb b (p) : bb b:f:bb 
 
 7(g):gg(-) g:gg g(li):gg gicg 
 
 z r: — r: — r: 
 
THE PROBLEMS. 
 
 12. Supposing the Old English words given, the first 
 step is to find the corresponding words in each of the 
 other dialects. This may be done by help of the Old 
 and Middle English Dictionaries, or by direct search in 
 the dictionaries of those languages. The former method 
 is generally easier, but the results thus obtained must 
 be supplemented and verified by reference to the vari- 
 ous lexicons. A general acquaintance with the Gothic, 
 Old High German, Old Norse, and Old Enghsh Gram- 
 mars is necessary. The more intimate the knowledge 
 of these, the easier is the specific task, and the more 
 certain the result. 
 
 In the discussion of the following problems, where 
 no references are given in proof of a statement, the 
 information will be found by a glance at the Vowel or 
 Consonant Table. The following abbreviations are 
 used: — 
 
 Braune' . . . . Gothic Grammar. 
 
 Braune^ Old High German Grammar. 
 
 Gr Old English Gram'hiar. 
 
 Noreen ... Old Norse Gramknar. 
 
 Germ Germanic. 
 
 Goth Gothic. 
 
 OE Old English. 
 
 0H6 Old High German. 
 
 ON Old Norse. 
 
 OS. . . . . Old Saxon. 
 
 WG West Germanic. 
 
12 PHONOLOGICAL INVESTIGATION OF OLD ENGLISH. 
 
 1. Aiar. 
 
 OHG. eld; OS. 69; ON. elffr; Goth. alj>s. 
 
 Let us begin with the Gothic form, aips. According 
 to the table of vowels, ai might represent either Germ. 
 e, 1, or ai. The two former are short, while the OS. 
 form of the word shows that it is here long. Hence 
 we are justified in assuming that Goth, ai here = Germ. 
 ai. The d of the OHG. eid might stand either for 
 Germ, p oi & (Braune^ 174); the Goth, form is con- 
 clusive, and requires us to postulate Germ. p. The 
 final s of Goth alps represents Germ, z (Bratme^ 74. e). 
 Apparently, therefore, the Germ, form of the whole 
 word would be aipz. But dtS is an o-stem (Gr. 239. a), 
 which is a better designation of a^tem (Braune^ 192). 
 Hence Goth, aips stands for Germ, aipoz, of which the 
 stem is aipo-. 
 
 2. burg. 
 OHG. burg, Tjurug; OS. burg; Goth, baurgs. 
 
 Evidently this word can give but little trouble. WG. 
 b (because OE., OHG., and OS.) must correspond to 
 Germ. 16; u remains; r remains; g is Germ, y; Goth, s 
 is Germ, z (Braune^ 74. e). Hence Germ. f)ni<yz, since 
 it is a consonant stem (Gr. 284), and therefore has no 
 vowel before the z. The WG. form, burgz, is more 
 commonly postulated (v. borough in Phil. Soc. Diet., 
 which has burgs instead of burgz). The second u of 
 OHG. burug is accounted for by Braune^ 69. b. 
 
 3. cynlng. 
 OHG. chuiil(n)g; OS. kuning; ON. konnngr. 
 The OHG. eh points to Germ. k. We should at 
 once postulate kuning-, then, were it not for the ON. 
 
PKOBLEMS. 13 
 
 form, since OE. y is from u (Gr. 95). But the ON. o 
 can only spring from u (in the Germanic period?) 
 through the influence of a or o in the next syllable 
 (Noreen 56). Hence we should assume, judging from 
 the ON., kuuang-. But Germ, kunang- ought to yield 
 OHG. clionang- (Braune^ 32), or perhaps eventually 
 choiii(ii)g-. Such a form is indeed found in MHG., 
 as konig-, whence Mod.HG. konig. Hence we are 
 driven to postulate two Germ, forms, kunangoz and 
 kuningoz (see 1, above), late Germ, konangoz and 
 kuningoz, perhaps originally both kunangoz. 
 
 4. daeg. 
 
 OHG. tag, tac; OS. dag; ON. dagr; Goth. dags. 
 According to the consonant table, OHG. t = Germ. tf. 
 The c (= k) of tac might possibly indicate Germ, k, 
 were there not other evidence in favor of 7. Hence 
 Germ. tSaryoz (cf. 1, above), for which the books usually 
 have WG. dagoz or dagaz. 
 
 5. €ast. 
 
 OHG. OS. 6st-; ON. austr. 
 The ON. au would seem to be the Germ, diphthong ; 
 this is confirmed by OHG. 6 (Braune^ 45). The st is 
 Germ. (Braune^ 161), and the radical syllable accord- 
 ingly aust-, which, if the word be a masc. o-stem, would 
 yield the Germ, form austoz. 
 
 6. efes (yfes). 
 OHG. opasa, obasa ; ON. ups ; Goth, ublzwa. 
 The OE. f is Germ. =6 (Gr. 192. 2) ; with this corre- 
 spond the Goth, and OHG. The ON. p is no exception 
 
14 PHONOLOGICAI, INVESTIGATION OP OLD ENGLISH. 
 
 (cf. Noreen 230. 2; 185 Anm. 2; but 193 Anm. 2), 
 though at first sight apparently so. The disappearance 
 of the Goth, w in WG. is regular (Braune^ 109 Anm. 
 2), which accounts for the OHG. form, except for the 
 s instead of z. The latter should regularly give r when 
 " medial (Braune^ 82. h. 1), though Goth, azgd, OHG. 
 asca is a similar exception. Goth, u might be either 
 o or u ; the OHG. o points to Germ, o (Braune^ 18. a), 
 from a remoter Germ, u ; the ON. to u (Noreen 152. 1 ; 
 cf. 172) ; OE. e to o (Gr. 93. 1), but OE. y to u (Gr. 
 95). The loss of both middle and final vowel in ON. 
 ups is unusual (Noreen 134 Anm. 1). The noun is 
 originally a w6-stem. Hence we may assume two 
 forms in Germ. : uT5asw<} and u16isw«i. The former 
 would account for OHG., ON., and Prim. OE. *ofas; 
 the latter for Goth, (except the z) and Prim. OE. *ufls. 
 Prim. OE. *ufls would then yield yfes, while Prim. OE. 
 *ofas might become *ofls by analogy, and hence ^fes. 
 
 7. faran. 
 OHG. OS. Goth, faran ; ON. fara. 
 
 These various forms are identical, except for ON., 
 
 which is easily accounted for (Noreen 220. 3). Hence 
 
 the Germ, form is faran (or faron, Gr. 45. 4 ; 109. a ; 
 
 160. 2). 
 
 8. feld. 
 
 OHG. OS. feld. 
 OHG. d points to p. This would yield felf-. Noth- 
 ing forbids this (Gr. 202. 2; Braune^ 162 Anm. 1). 
 The oldest English actually has felth (cf. OET. p. 521). 
 The noun was originally a long u-stem (Gr. 272) ; hence 
 the Germ, form would be fel]>uz. 
 
PROBLEMS. 15 
 
 9. ferian. 
 
 OHG. OS. ferian; ON. ferja; Goth, farjan. 
 
 The Goth, word immediately suggests a Germ, farjan 
 (farjon, see 7, above). This would account for all the 
 forms (Braune2 26. b; Noreen 61. 1; Gr. 89. 1). 
 
 10. fierd. 
 
 OHG. vart; OS. fard; ON. ferff; Goth. far}.. 
 
 The ON. ar points to Germ. J? (Noreen 184. 2). Hence 
 we might assume Germ, farf-, which is confirmed by 
 the Gothic. On the other hand, OHG. t = Germ tf, 
 and this is also indicated by the OS. and OE. There 
 are therefore two Germ, roots, far]?- and fariif-, a differ- 
 ence probably due to the place of the Indo-European 
 accent, far]?- indicating a stressed radical syllable, and 
 farJT- a stressed suffix (cf. Noreen 175; Kluge, Nomi- 
 nale Stammbildungslehre 127; 128). The noun is an 
 i-stem (Gr. 269; Braune^ 219). Hence Germ, farfiz 
 or farSriz. 
 
 11. flfeman. 
 
 OE. ie springs from 6a or 6o (Gr. 42. 4, 5). The 
 noun fl€am (Gr. 222. 2) decides between the two, and 
 enables us to postulate a Germ, flauhniion (Gr. 398. 1), 
 in which the original li shall appear. Suppose it is 
 required to deduce the OHG. form of this word. From 
 Braune^ 45 we learn that Germ, au = OHG. d, when 
 in this position. The consonants of the stem remain. 
 The ending -ion (which here = Braune's -jan) becomes 
 -en (Braune^ 314; 58 Anm. 2). Hence OHG. *fl6hmen, 
 or possibly *floilimen (cf. Braune^ 45 Anm. 4; 51). 
 
16 PHONOLUUIUAL rNVBSTIGATION OF OLD ENGLISH. 
 
 Suppose the ON. form be required, h is lost (Noreen 
 217) ; au undergoes i-umlaut to 0y, ey (Noreen 61. 7) ; 
 n is lost (Noreen 220. 3) ; i is lost (Noreen 58; 62. 1 ; 
 cf. Gr. 177). Hence ON. *fljz<yma or *fleyina. 
 
 12. flocc. 
 
 ON. flobkr. 
 
 Noreen 235 shows that ON. kk = Germ. kk. The 
 ON. o = Germ, o (Noreen 147. 1). f and 1 are con- 
 stant. Hence Germ, flokk-. The word being a masc. 
 o-stem, we obtain as the full Germ, form, flokkoz. 
 
 13. fdn. 
 
 OHG. OS. Goth, fdhan; ON. t&. 
 
 Germ, has an (^) for OE. 6 (Gr. 67; cf. Braunei 60 
 Anm. 2 ; Braune^ 83 ; Noreen 138. 4). Hence Germ. 
 fanhon, perhaps better written fa"hon or i&lnoji 
 (Braune2 128 Anm. 1). 
 
 14. g^ar. 
 
 OHG. OS. j&r; ON. dp ; Goth. j6r. 
 
 Goth. 6, OHG. OS. d indicate Germ. sfe. Let us see 
 if this will answer. Jser would give ON. dr, since 
 Germ, sfe = ON. & (Noreen 75), and j is lost (Noreen 
 210). Jefer hkewise gives OE. g^ar, since sfe after this 
 palatal becomes ^a (Gr. 74 ; 75. 2), and g represents j 
 (Gr. 175. 2). The word, being a neuter o-stem, is Germ. 
 jsferom (for the terminal m, see the Phil. Soc. Eng. 
 Diet, under such words as hairn, bone'). 
 
PKOBLEMS. 17 
 
 15. glsl. 
 
 OHG. gisaX; ON. gisl. 
 
 OHG. g is Germ. 7, that is, spirant g. OHG. al here 
 = Germ. 1 (Braune^ 65). s, 1, and f are constant. Since 
 the noun is an o-stem, the Germ, form is ^fsloz. 
 
 16. h^m. 
 
 OHG. belm ; OS. h6m ; ON. taelmr ; Goth, halms. 
 
 h and m are constant. Goth, s here represents z 
 (Braunei 74. 0). OHG. ei = Germ, ai (Braune^ 15. 6), 
 as does ON. ei (Noreen 156). This Germ, ai = OE. & 
 (Gr. 62). Hence Germ, haimoz (usually o-stem). 
 
 17. healdan. 
 
 OHG. haltan; OS. Goth, haldan; ON. halda. 
 
 OHG. t, OS. ON. Goth, d = Germ. tf. h and 1 are 
 constant. OE. ea = Germ, a (Gr. 80). Final n is 
 regularly lost in ON. (Noreen 220. 3). Final an in 
 infinitives = Germ, -on (Gr. 45. 4; 160). Hence Germ. 
 haliaron. 
 
 18. healf. ^ 
 
 OHG. Goth, halba; OS. balfia; ON. halfa. 
 
 OHG. h = Germ. ft. This Germ. 15 also = ON. f 
 (Noreen 248. 1) and Goth, b (= 16, Braune^ 82. b. 3). 
 The Germ. f> is represented by OE. f (Gr. 192. 2; cf. 
 191). OE. ea = Germ, a (Gr. 80). The final a would 
 be lost in OE. (Gr. 134. 5). healf being an d-(dr)stem, 
 the Germ, form is haI1&6. 
 
18 PHONOLOGICAL INVESTIGATION OF OLD ENGLISH. 
 
 19. here. 
 
 OHG. heri, hart; ON. herr; Goth, haijls. 
 
 here is a jo-stem (Gr. 246); so in Goth. (Braune^ 
 80) ; h and r are constant. Comparing the e of OE. 
 with the a of Goth., we see that the former denotes 
 umlaut (Gr. 89. 1). If this be really a jo-stem, we can- 
 not therefore escape the conclusion that the Germ, 
 form is harjoz. ON. j is lost before 1 (in this case 
 Goth. I, Noreen 218); and the vowel of the second 
 syllable would also be lost in ON. (Noreen 135), while 
 final z would here become R (Noreen 180), which with 
 the stem-final would yield rr (Noreen 200. 2. and b. a). 
 Final z is lost in WG., and consequently in OHG. 
 (Braune^ 82. b. 1). The o of the second syllable would 
 not appear, any more than in such pure o-stems as eld. 
 Hence the OHG. form would be *harj, which would 
 become hari (Braune^ 119)» and, by umlaut, heri 
 (Braune2 26. 6). 
 
 20. hi6ar. 
 
 h and 1 are constant ; tS represents Germ, p (Gr. 201) ; 
 6 is Germ. 6 (Gr. 60). The noun is feminine, and an 
 d-stem. Hence Germ. hl6p6. 
 
 21. hreddan. 
 OHG. retten. 
 
 OE. hr is usually OHG. r (Braune^ 153). OHG. tt 
 = WG. (OS. OE.) dd before j (Braune^ 188; 164). 
 OE. dd represents Germ, tfj (usually designated by dj 
 in the books; cf. Gr. 228). OE. OHG. e is Germ, a 
 (Gr. 89. 1 ; Braune^ 26. 5). Hence Germ. hraSrjon. 
 
PROBLEMS. 19 
 
 22. hwfl. 
 
 OHG. wQa; OS. ON. hwUa; Goth, hwella. 
 
 If allowance be made for the loss of initial h in OHG. 
 (Braune^ 153), the radical syllable will reduce itself to 
 hwfl-, the Germ. form. The final a represents Germ. 6 
 (Braune^ 207), which is lost in OE. (Gr. 134. 6). Hence 
 Germ, hwfld. 
 
 23. h^ar. 
 
 OHG. -hunda ; Goth. huii]>8. 
 
 bftS is classified as a jdrstem, that is, as a jd-stem (Gr. 
 258. 2). OHG. hunda might apparently be a jd-stem 
 (Braune^ 210). The decision with respect to hunda is 
 complicated, however, by OE. liAtS, with the same 
 meaning, which, since it lacks umlaut, can only be an 
 «J-stem. Since the form does not forbid, OHG. hunda 
 might also be an d-stem. Goth. hun]7s being only pos- 
 sibly a feminine (Gabelentz-Loebe, Glossary) cannot 
 help us much. Hence we are in somewhat of a quan- 
 dary. Howeyer, hj?af can only be a jd-stem. The 
 word must have originally possessed n (Gr. 185. 2), 
 and the vowel u (Gr. 96. &). The surd spirant p is 
 indicated by the Gothic and confirmed by the OHG. 
 (Braune^ 166). We may therefore postulate Germ. 
 hun]>jd. 
 
 24. lond. 
 
 OHG. lant; OS. ON. Goth. land. 
 
 The OHG. t is here Germ.ff (Braune^ 82. b. 2; 163). 
 1 and n .are constant. OE. o is here Germ, a (Gr. 
 65). Thfe noun is a neuter o-stem. Hence Germ. 
 lanSTom. 
 
20 PHONOLOGICAIi INVESTIGATION OF OLD ENGLISH. 
 
 25. mete. 
 
 OHG. maz ; OS. meti ; ON. matr ; Goth. mats. 
 
 OHG. maz is neuter, and must therefore be excluded. 
 OE. mete is masc. l-stem (Gr. 263) ; so are Goth, mats 
 (Gabelentz-Loebe, Gram. 75) and ON. matr (Noreen 
 302. 1). There is evidently i-umlaut in the OE. OS. 
 forms, but not in ON. (cf. Noreen 306). m and t are 
 constant. Hence Germ, matiz. 
 
 26. mon. 
 
 OHG. OS. man ; ON. maiffr ; Goth, manna. 
 
 The ON. -tfr is inorganic (Noreen 220. 2) ; mann 
 would be regular (Noreen 326). OE. o is here Germ. 
 a (Gr. 65). m and n are constant. The Goth, manna 
 is a weak form (Braune^ 107. 1), and its ending is 
 therefore to be disregarded for this purpose. From 
 Kluge's Etym. Wort. s.v. mann, we learn that the Germ, 
 form is mann-, by assimilation from manw-, itself stand- 
 ing for Indo-European manu. This Germ, mann- will 
 fulfil every condition, since gemination was simphfied 
 in OE. (Gr. 225. 1) and OHG. (Braune^ 93). 
 
 27. mdnaW. 
 
 0HG.mAn6d; ON. mAna»r; Goth. men6]>B. 
 
 OE. 6 in this word = Germ, sfe (Gr. 68). This Germ. 
 e6 gives OHG. &, Goth. 6 (Braune^ 34), ON. & (Noreen 
 75). In the second syllable, ON. a = Germ. 6 (Noreen 
 113. 2), OHG. d = Germ, p (Braune^ 166), as does ON. 
 tS here (Noreen 184. 2). Hence Germ. raebn6p- (cf. Gr. 
 281; Noreen 325). 
 
PEOBLEMS. 21 
 
 28. niht. 
 
 OHG. OS. naht; ON. n6tt; Goth, nahts. 
 
 Goth, nalits is an anomalous i-stem (Braune^ 106). 
 The ON. form is peculiar, on account of its u-umlaut 
 (Noreen 72) ; tt springs regularly from ht (Noreen 
 197. 3). OHG. ht = Germ, lit (Braune^ 161). OE. i 
 comes from a by palatal umlaut (Gr. 101). 
 
 29. niman. 
 
 OHG. neman ; OS. Goth, niman ; ON. nema. 
 
 The 1 of OE. Goth, niman is WG. e (Gr. 69; Braune^ 
 10. 1). Final n is wanting in ON. (Noreen 220. 3). 
 Hence Germ, nemon. 
 
 30. notian. 
 
 OHG. niizon (nuzen) ; ON. nota. 
 
 OHG. nuzen must be disregarded, as belonging to a 
 different conjugation. ON. o must here equal Germ, o 
 (Noreen 147; cf. 76), as must likewise OE. o (Gr. 55). 
 OHG. z = Germ. t. OE. ia in the ending of weak 
 verbs = Germ. 6jo (Gr. 411). Hence Germ. not6jon. 
 
 31. rM. 
 
 ON. reiff. 
 
 ON. ei — Germ, al (Noreen 156). This ai would 
 account for OE. d (Gr. 62). ON. t! usually = Germ. 
 af (Noreen 249. 1). OE. d = Goth, d (Gr. 197), and 
 this = Germ. 9 (Braune^ 82. b. 2). rdd is an 6-stem 
 (cf. Noreen 288 Anm. 2). Hence Germ, raiiard. 
 
22 PHONOLOGICAL LNVESTIGATION OF OLD ENGLISH. 
 
 32. rs^can. 
 
 OHG. relhhen. 
 
 OHG. ei here = Germ, ai (Braune^ 15. b ; cf . 43 
 Anm. 4). OHG. lih = Germ, k (Braune^ 145). The 
 OHG. infinitive ending -en points to -ian, or, rather, 
 -ion (Braune^ 118 Anm. 2; cf. Gr. 45. 8). We should 
 thus have Germ, ralkion. From this Germ, al would 
 come OE. d. (Gr. 62), which would be converted by 
 i-umlaut into * (Gr. 90). Germ, k = OE. c. The 
 Germ, i regularly disappears (Gr. 177. &). The o of 
 -on becomes a (Gr. 108; 109). Hence OE. rsfecan. 
 
 33. rice. 
 
 OHG. rihhi; OS. ON. rikl; Goth, reiki. 
 
 OHG. hh = Germ, k (Braune^ 145). The noun is 
 a neuter jo-stem (Gr. 246 ; Gabelentz und Loebe, Gram, 
 p. 60). Hence Germ, rfkjom. 
 
 34. rfdan. 
 
 OHG. ritan ; ON. rifSa. 
 
 ON. tS = Germ, ts (Noreen 249. 1). This Germ. » = 
 OHG. t, OE. d (Braune2 89; Gr. 197). r and f are 
 constant. Final -an points to -on (Gr. 108; 109). 
 Hence Germ, rlZTon. 
 
 35. r^met. 
 
 The meaning allies it with OE. OHG. OS. ON. 
 mm, Goth, riims. OE. f here = Germ, li (Gr. 96. 1 ; 
 61). r and m are constant. The ending -et causes 
 umlaut, and therefore springs from -it. Such words are 
 
PROBLEMS. 23 
 
 neuter jo-stems (Gr. 247. e). Hence Germ, riimitjom 
 (cf. Kluge, Nominale Stammbildungslehre 144). 
 
 36. samnian. 
 
 OHG. samanon ; OS. samnon ; ON. samna. 
 
 s, a, and m are constant. The second a of the OHG. 
 form is peculiar to this dialect, and not original (Braune^ 
 65; Paul-Braune, Beitrage 5:90, 91). OE. ia here = 
 Germ, djo (Gr. 411). Hence Germ, samndjou. 
 
 37. scip. 
 
 OHG. scif, scef ; OS. ON. Goth. skip. 
 
 OHG. f = Germ, p (Braune2 132). OHG. sk = 
 Germ, sk (Braune" 146). The i is Germ.; for occa- 
 sional OHG. scef see Braune^ 31. Anm. 1. The noun is 
 neuter, and has the appearance of an o-stem (Gr. 241). 
 This would jrield Germ, skipom. There is a possibility, 
 however, of its having been a Germ, oz-, ez-stem (Gr. 
 288 N. 1; 182). In this case we should have Germ. 
 skepoz, skipez (Goth. *skipiz). The variation be- 
 tween radical e and i would thus be due to the vowel 
 of the following syllable. 
 
 38. sctr. 
 
 Here we have only the OE. word. The vowel is 
 marked long in Alfred's Orosius 19®, and is long in 
 Modern English. It is a feminine, and according to 
 Cosijn an ^(d-)stem. All the sounds are constant. 
 Hence Germ, sklrd. 
 
24 PHONOLOGICAL INVESTIGATION OF OLD ENGLISH. 
 
 39. s6caii. 
 
 OHG. suohhen; OS. sokian; ON. s^ja; Goth, sdkjan. 
 
 OHG. hh points to Germ, k (Braune^ 145), and uo 
 to Germ. 6 (Braune^ 21. d^. s is constant. OS. -ian 
 points to Germ, -ion (Gr. 398. 1). Hence Germ, sdklon. 
 This will account for OE. s6can (Gr. 94. a; 177. 5), 
 and for the ON. form (Noreen 61. 4). 
 
 40. seUan. 
 
 OHG. seUen; OS. seUian; ON. selja; Goth, saljan. 
 
 The Goth, ending -an is Germ, -on (Gr. 108; 109). 
 Hence the Goth, form would indicate Germ, saljon. 
 This would account for OHG. (Braune^ 26. a; 96. a; 
 122; 118. Anm. 2); for ON. (Noreen 61. 1; 256; 260. 
 1; 220. 3); for OE. (Gr. 89. 1; 228; 177; 80. N. 2). 
 
 41. set. 
 
 OHG. sez; ON. set. 
 OHG. z = Germ. t. s and e are constant. The 
 noun is a neuter o-stem. Hence Germ, setom. 
 
 42. sittan. 
 OHG. sizzen ; OS. slttlan ; ON. sitja ; Goth, sitan. 
 
 A comparison of the OS. and ON. would indicate 
 either Germ, sittion or sitjon. OE., OHG., and OS. all 
 have double consonant; ON. and Goth. t. But the first 
 three all belong to WG. This accords with a general 
 principle (Gr. 228), and we may therefore assume that 
 the Germ, has t. OHG. zz is Germ, t (Braune^ 159), 
 and OHG. -en is regular (Braune^ 118 Anm. 2). In 
 OE. sittan the j is regularly lost (Gr. 177). 
 
PROBLEMS. 25 
 
 43. sitf. 
 
 OHG. slnd ; OS. siff ; ON. sinn ; Goth. sln>s. 
 
 Let us assume that Goth, sinp- is Germ, s would be 
 constant, p gives OHG. d (Braune^ 174. 1). n is lost, 
 and i lengthened, in OE. (Gr. 185. 2). ON. nn is 
 Germ, up (Noreen 199. 2). The noun is a masc. o-stem. 
 Hence Germ. sm]70z. 
 
 44. tr^ow. 
 
 OHG. trinwa; OS. treua; ON. trfia, trA ; Gotli. triggwa. 
 
 OE. tr^ow is a w6-(w£ir)stem (Gr. 259). The WG. 
 diphthong is eu (Gr. 64 ; Braune^ 17. d. 1), but tliis eu 
 springs from Germ, e before ww (Braune^ 49 Anm. 4 ; 
 113. by. Goth, gg is merely the sign of this Germ, 
 gemination (Braune^ 113). The w would be lost in 
 ON. (Noreen 223) ; we are not at present in a position 
 to account for the ON. li (Noreen 153 Anm. 1). Hence 
 Germ, trewwd. 
 
 45. wseter. 
 
 OHG. 'wazzap; OS. watar ; ON. vatn ; Goth, tiratd. 
 
 The Germ, must have had two stems, since EG. 
 (Goth, and ON.) differs from WG. We will concern 
 ourselves only with WG. OHG. zz = Germ, t (Braune^ 
 159). w is constant. The noun is a neuter o-stem 
 (Gr. 245). Hence Germ, watarom. 
 
 46. weald. 
 
 OHG. OS. wald; ON. voUr (Noreen vQllr). 
 OE. weald is a u-stem (Gr. 273 ; cf . Braune^ 229 and 
 Noreen 309). OE. ea = Germ, a (Gr. 80). ON. 11 
 might be Germ. Ip (Noreen 257. 2; 199. 1). This Ip 
 
26 PHONOLOGICAL INVESTIGATION OP OLD ENGLISH. 
 
 would account for OE. Id (Gr. 202. 2), and for OHG. 
 Id (Braune^ 174). ON. q stands for a, and is due to 
 u-umlaut (Noreen 71. 1). Hence Germ, walpuz. 
 
 47. (ge)weorc. 
 OHG. ■were, werah; OS. werk; ON. verk. 
 
 OHG. c = Germ, k (for werah cf. Braune^ 69. a). 
 w and r are constant. Assume Germ. werk-. This 
 will account for OE. weorc (Gr, 79. 1). OHG. ON. e 
 in this situation = Germ, e (Braune^ 29; Noreen 
 140. 1). The word, as neuter o^tem, is Germ, werkom. 
 
 48. wfcian. 
 
 Cf. OHG. wich ; OS. wik. 
 
 OHG. ch here — Germ. k. w and i are constant. 
 The ending -ian = Germ. 6jon (Gr. 411). Hence 
 Germ, wlkdjon. 
 
 49. wudu. 
 
 OHG. wltu ; ON. vlffr. 
 
 ' The word is a u-stem (Gr. 271 ; Braune^ 230 ; Noreen 
 309). The first u in OE. wudu is from I (Gr. 71). ON. 
 ts here = Germ. tS (Noreen 249. 1), since this would 
 give OHG. t (Braune^ 188 ; 89). w is constant. Hence 
 Germ. wiSTuz. 
 
 50. wundrian. 
 
 OHG. wuntardn ; OS. wundraian, -wundrolan, -wundrdn ; 
 ON. undra. 
 
 The ending is -djon (Gr. 411). OHG. t = Germ. 9 
 (Braune^ 188. 1). w, u, and n are constant except for 
 w in ON. (Noreen 211. 1). Hence Germ. wunarrtJjon. 
 
24 
 
 OLD AND MIDDLE ENGLISH. 
 
 Carpenter's Anglo-Saxon Grammar and Reader. 
 
 By Stephen H. Cakpenter, late Professor of Elietoric and English 
 Literature in the University of Wisconsin. 12ino. Cloth. 212 pages. 
 Mailing Price, 70 cents; Introduction, 60 cents. 
 
 Carpenter's English of the XIV. Century. 
 
 By Stephen H. Cabpenteb. 12nio. Cloth. 313 pages. Mailing Price, 
 $1.00; Introduction, 90 cents. 
 
 TLLUSTRATED by Notes, Grammatical and Philological, on 
 Chaucer's Prologue and Knight's Tale, and so forming an excel- 
 lent introduction to that author. 
 
 Beowulf, and The Fight at Finnsburh. 
 
 Translated by James M. Garnett, M.A., LL.D., Professor of the 
 English Language and Literature in the University of Virginia. With 
 Facsimile of the Unique Manuscript in the British Museum, Cotton. 
 VitelUus A XV. Second Edition, revised. 12mo. Cloth. 146 pages. 
 Mailing Price, Sl.lO; Introduction, $1.00. 
 
 Francis A. March, Prof, of Coin- 
 parative Philology, Lafayette Col- 
 lege : This is the best translation so 
 far in our language, and will do 
 honor to American scholarship. 
 
 J. Earle, Prof, of Anglo-Saxon in 
 the University of Oxford, Eng. : It is 
 a very complete piece of work, bring- 
 ing the whole subject up to the very 
 front line of its progress. 
 
 An Old English Grammar. 
 
 By Eduabd Sievers, Ph.D., Professor of Germanic Philology in the 
 University of Tubingen ; translated and edited by Albert S. Cook, 
 Ph.D. (Jena), Professor of the English Language and Literature in the 
 University of California. Second edition, revised and enlarged. 12mo. 
 Cloth. XX + 273 pages. Mailing Price, $1.25 ; for Introduction, $1.12. 
 
 TT is hoped that this version will be found not only to present 
 
 in English the most approved text-book on the subiect, but to 
 
 present it in a form better adapted for the use of students, and in 
 
 some respects more in accord with the views of the best authorities. 
 
 F. J. Child, Prof, of Eng., Harvard 
 Univ. : It is an absolutely masterly 
 book, as would be expected of tliose 
 who have made it. (Feb. 4, 1888.) 
 
 C. J. Eichardson, Prof, of Eng., 
 
 Dartmouth College : No more impor- 
 tant work is now accessible to the 
 student of the early grammatical 
 forms of our twelve-hundred-year-old 
 English language. (Feb. 28, 1888.) 
 
OLD AND MIDDLE ENGLISH. 
 
 25 
 
 OPrNIOITS AND REVIEWS OF THE FIRST EDITION. 
 
 American Journal of Philology: 
 
 The well-recognized merit of Sievers' 
 unrivalled Angelsachsische Gram- 
 matik created an imperative demand 
 for an English version. That this 
 demand would be met was inevita- 
 ble; how it would be met was doubt- 
 ful. Delay and apprehension are 
 now happily removed by the appear- 
 ance of this volume. 
 
 The characteristic feature of Sie- 
 vers' work is its independence of 
 preceding Anglo-Saxon grammars, 
 and its dependence upon the most 
 authentic documents of the earliest 
 periods of the language, discrimi- 
 nated as to age and dialect. Prof. 
 Cook's leading purpose has been to 
 supply to our institutions of learning 
 a serviceable text-book. To this end 
 the requirements of a perspicuous 
 and idiomatic style, and of a clear 
 and consistent terminology, have 
 been well met. 
 
 In the distribution of the matter 
 and the variety of the type, such an 
 improvement of the original is ef- 
 fected as might in itself almost jus- 
 tify the new version. 
 
 Deserving of notice, moreover, is 
 the care with which Prof. Cook has 
 corrected many erroneous cross-ref- 
 erences, his amplification of the In- 
 dex, and his excellent modification 
 of the system of accentuation. We 
 need hardly add that the work is 
 worthy of general acceptance. 
 
 The Independent: It maybe said 
 to mark an era in the history of Old 
 or First English, somewhat as did 
 Thorpe's first edition of Kask's Gram- 
 mar (1830) . Upon no scholar among 
 us could the work of presenting tliis 
 grammar in modern English form 
 have more appropriately fallen than 
 upon Prof. Cook. 
 
 The volume is a scholarly edition 
 of a scholarly original, and is not 
 only translated into English, but is 
 made as English in spirit and method 
 as such a work can be. 
 
 The SchoolmaBter, London : Both 
 as a work of reference and as a text- 
 book, this volume \» a valuable addi- 
 tion to the resources of the English 
 student. 
 
 Prof. Henry Johnson (Bowdoin 
 College), in the " Andover Xeview " ; 
 Every student of English owes a 
 great debt of thankfulness to Prof. 
 Sievers for making his Angelsach- 
 sische Grammatik. It was published 
 In 1882, and was recognized at once 
 as far superior in fulness and accu- 
 racy of treatment of the oldest Eng- 
 lish dialects to all preceding gram- 
 mars of the language. It was founded 
 on the prose literature. Since 1882 
 the most important contribution, to 
 Old English grammar is the publica- 
 tion of emendations of his grammar. 
 Besides these materials. Prof. Cook 
 has incorporated in the American 
 edition the results of the work of 
 other scholars up to the present, and 
 has produced what is by far the best 
 Old English grammar to be had. 
 
 Charles F. Bichardson, Prof, of 
 Anglo-Saxon and English, Dart- 
 mouth College : Anglo-Saxon stu- 
 dents have been waiting with interest 
 lor the work, and it is fortunate that 
 it has been translated by a competent; 
 American scholar, himseU a pupil of 
 the author. 
 
 W. D. McClintock, Professor of 
 English and Registrar, Chautauqua 
 University: I am greatly pleased 
 with Prof. Cook's translation of Sie- 
 vers' Old English Grammar, and 
 shall use it in our classes. 
 
Cornell University Library 
 PE 140.C77 
 Phonological investigation of 0^^^^^ 
 
 3 1924 027 323 827