The Third Class of Weak Verbs in Primitive Teutonic, with Special Reference to Its Development in Anglo-Saxon 9781607245926, 1607245922

In this paper Marguerite Sweet offers a tentative explanation by showing parallels not in Latin (as was previously sugge

256 90 3MB

German Pages 47 [51] Year 2019

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
AMERICAN JOURNAL OF PHILOLOGY
Recommend Papers

The Third Class of Weak Verbs in Primitive Teutonic, with Special Reference to Its Development in Anglo-Saxon
 9781607245926, 1607245922

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

The Third Class of Weak Verbs in Primitive Teutonic, with Special Reference to Its

A n a l e c t a Gorgiana

338 Series Editor George Anton Kiraz

Analecta Gorgiana is a collection of long essays and

short

monographs which are consistently cited by modern scholars but previously difficult to find because of their original appearance in obscure publications. Carefully selected by a team of scholars based on their relevance to modern scholarship, these essays can now be fully utili2ed by scholars and proudly owned by libraries.

The Third Class of Weak Verbs in Primitive Teutonic, with Special Reference to Its

Marguerite Sweet

l gorgias press 2009

Gorgias Press LLC, 180 Centennial Ave., Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2009 by Gorgias Press LLC Originally published in All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. 2009

1

ISBN 978-1-60724-592-6

ISSN 1935-6854

Extract from The ^American Journal of Philology 14 (1893)

Printed in the LTnited States of America

THIRD

CLASS OHG.

Goth, pahan ]>arban

OF

TEUTONIC O. Sax.

dagèn

thagon

darbèn

tharbon

WEAK

VERBS.

AgS.

455 O. Norse. pegja

pearfian

parfa

diemuotèn dicchén péowian

piwan donèn

prówian

druoèn l>ulan

dolèn

tholon, tholian polian

wachèn

wakon

*wanan

wanón

wacian

\ola ugga vaka

wanian

warnèn, -6n weihan

wartèn

wardon

wihjan

wian

weardian vigja

waron

vara

welkèn warón wernèn, -ón witan

wesanén 1

weosnian

wizzèn

witian

visna

wókan wriskan wulan *wunan

wonèn,-ón

wonon

wunian

una

MARGUERITE SWEET. 1

Probably a Teutonic weak verb in -n-an.

AMERICAN

J O U R N A L OF V O L . X I V , 4.

PHILOLOGY WHOLE

NO. 56.

I . — T H E T H I R D C L A S S O F W E A K V E R B S IN P R I M I TIVE TEUTONIC, WITH SPECIAL REFERENCE T O I T S D E V E L O P M E N T IN A N G L O - S A X O N . From the time of Grimm until the present day, scholars have been seeking for an explanation of the origin and structure of the Teutonic third weak conjugation. T o account for the diphthong of the stem alone forms one of the most perplexing problems of Teutonic grammar. The Latin conjugation in -e- is so closelyallied to the Teutonic az-conjugation as to suggest at once identity of origin; still, the Teutonic diphthong cannot be the direct equivalent of the Latin vowel. It seems, moreover, doubtful whether ai held in the Primitive Teutonic third class the position of 6 in the second, of j in the first weak conjugation. In AngloSaxon, in Old Saxon, and in Old Norse ai does not appear in the preterit, while in Gothic, Anglo-Saxon, and Old Saxon, ai in the present indicative is confined to the second and third persons singular and the second person plural. T h e condition of the inflection, too, is no less perplexing than that of the stem-vowel. Gothic shows in the present an interchange of strong forms with forms in ai-, the Saxon dialects have a corresponding interchange of j- with («-forms. What is the significance for Primitive Teutonic of this mixture of forms, and which, Gothic or AngloSaxon, is nearer the primitive condition ? In Bopp's Comparative Grammar (I2 226) is found the first attempt to explain the origin of the «z-verb; for Grimm, in his discussion of the Gothic az'-class, had done little more than state

4IO

AMERICAN

JOURNAL

OF

PHILOLOGY.

the difficulties to be met. 1 B o p p held that the T e u t o n i c «¿-class is the s a m e as the L a t i n class in and that both w e r e d e v e l o p e d , as w e r e also the T e u t o n i c 6- a n d / - c o n j u g a t i o n s , from the Sanskrit derivative class in -aja-. H e saw in the Latin 6 a contraction o f a and i, the O l d H i g h G e r m a n t b e i n g the result o f a corresponding, but independent, T e u t o n i c contraction. W i t h r e g a r d to the inflection, B o p p considered the irregularities of the G o t h i c and S a x o n a result o f dialectic d e v e l o p m e n t . T h e t h e o r y that the T e u t o n i c w e a k conjugations find their explanation in the Sanskrit tenth class was a c c e p t e d also b y S c h e r e r , as it was b y other scholars for m a n y y e a r s after the time of B o p p . B u t S c h e r e r 2 was led b y certain other facts to consider the az-conjugation a m i x t u r e of three original t y p e s , viz. a) t w o s t r o n g v e r b s in -mi (one r e d u p l i c a t i n g ) ; b) a derivative v e r b in -aja-. T h e s e facts w e r e : the relation o f G o t h i c \ulan to G r e e k j-Xijeat : that of G o t h i c munan to G r e e k /¿(-¿u^o-Kia; and that o f G o t h i c bauan to Sanskrit ba-bhü-va, bhä-vä-mi. In 1879, a y e a r after the a p p e a r a n c e of the second edition of S c h e r e r ' s Z u r G e s c h i c h t e der deutschen S p r a c h e , M a h l o w ' s treatise, D i e langen V o c a l e ä, t, 6 in den europäischen S p r a c h e n , was published. In spite of the m a n y arbitrary and m i s l e a d i n g explanations offered in it, the essay is of interest and importance, a n d m a r k s a turning-point in the discussion of the T e u t o n i c third w e a k class. M a h l o w starts with the assumption that there existed originally, in all the E u r o p e a n l a n g u a g e s , t w o classes o f w e a k verbal stems in -ej the one a class of causatives in -ej6-, the other a class of denominatives in -tj6-. In S o u t h E u r o p e a n the two classes fell t o g e t h e r , but in the T e u t o n i c dialects t h e y persisted as the / - c l a s s and the ¿-class. B u t it is evident that the G o t h i c at cannot be the direct equivalent of L a t i n e, and it is at the same time i m p r o b a b l e that O l d H i g h G e r m a n e represents the primitive condition. T h i s consideration leads t o M a h l o w ' s second assumption, viz. that Gothic £, O l d H i g h G e r m a n d, should 1 D e u t s c h e G r a m . I 2 850: " D e r ableitungsvocal lautet a«, erfährt aber ein von dem S zweiter conjugation verschiedenes schicksal, n ä m l i c h : a) vor consonantisch anhebender flexion b l e i b t er, g l e i c h j e n e m unbeeinträchtigt; ß) hebt die flexion mit i an, so verschlingt er dieses : also habais, kabaiy stehen für habai-is, Aabai-ip] y) hebt aber die flexion m i t a, au, oder selbst mit ai an, so wird das ableitende ai ausgeworfen, mithin stehen haban, haba, habam,

habau, habai für habaja, habajam, habajan, habajau, habajai." 2

Z G D S . 2 265 sq.

THIRD

CLASS

OF

TEUTONIC

WEAK

VERBS.

4II

be represented by a sign a', which, before j and vowels, was in Primitive Teutonic developed into ai} T h i s -ai-, developed from -a'j-, held in the Primitive Teutonic conjugation the same place as that held by the -j- of class I, by the -6- of class II. T h e Gothic strong forms in the present indicative were derived from the optative, where they were first developed by contraction in the first and second persons. 2 T h e original condition of the verb is not to be looked for in Gothic, nor in Old H i g h German, but in A n g l o - and Old S a x o n . *haba'ja became in W e s t Germanic *habeju, by loss of inlaut -e- *habju, finally *hebju. S o the W e s t Germanic inflection was developed from the original inflection in the following w a y : I n d . : P r i m . T e u t . *habaia Prim. W G . O. Sax. Opt.:

*habaiz

*habaifi

*habeju

*habas

*haba§

hebbju

habas

P r i m . T e u t . *kabaiau Prim. W G . O. Sax.

*habaif> *habcif>

habaft

kebbiaft

*habaiaiz

*habeja

*habaiai

*habejas

hebbia %

*habeja

hebbias

Part, pres.: Prim. Teut. habaianf>;

etc.

W G . *habejand;

O. Sax.

hebbiand. 1 O f this v o w e l M a h l o w says (p. 26): " E s ist sehr v e r l o c k e n d , das gotische e dem £ der anderen europäischen sprachen vollständig gleich zu stellen, um so mehr als auch die sächsischen dialekte einen «-laut ä für das gotische i haben, und ein e auch den älteren hochdeutschen dialekten nicht fremd war, v g l . S c h e r e r ' Z G D S . 126. Indessen wird sich unten zeigen, dass das germanische i, ä noch einem ganz anderen laut entspricht, und schon dies würde die gleichsetzung des gotischen e mit dem europäischen $ b e d e n k l i c h machen. E s giebt ferner noch andere gründe, die uns verhindern, als urgermanische gestalt des vocals £ anzusetzen : es muss ein langer vocal gewesen sein, der dem kurzen a näher stand als dem e. D i e s ist zu schliessen aus der entstehung von ai, die eben besprochen ist, der kürzung zu a in fadar, dauhtar, der contraction in got. frit, fretun, die aus *fra it, *fra Ptun nicht entstanden wären. Ich setze daher als urgermanischen Vertreter des europäischen i ein helles ä an, das oben mit a' b e z e i c h n e t w u r d e : dies a' wurde in allen germanischen dialekten zu a, dann theils zu theils w i e d e r zu ä." T h e argument is far from c o n v i n c i n g . Fadar and dauhtar are probably formed l i k e agent nouns in -or\ like bhrätor, not l i k e narfuj. T h a t fret has arisen from fra et by a simple and not uncommon process of elision is proved by numerous examples in Gothic and in O H G . E . g. Goth, karist = kar(a)ist, yammuh = yamm(a)uh, O H G . galtiro = ge-altiro, nein = ni-ein, etc. B r u g m a n n , Gram, der indog. Spr. I 461. 2

Prim. T e u t .

*haba'au

*haba'aiz

*haba'ai,

"

*habaiau

*hab{ai)aiz

*hab(ai)ai,

etc. etc.

412

AMERICAN

JOURNAL

OF

PHILOLOGY.

The two main points which the theory of Mahlow seeks to establish, viz. the identity of Latin -g- and Gothic -ai- and the direct development of the Saxon forms from the Prim. Teut., form as well the basis of Kögel's treatment of the subject. Kögel (PBB. I X 504 sq.) follows Mahlow very closely ; in fact, he does little more than bring together fresh material in support of the latter's view. There are, indeed, slight differences in detail. E . g . , with regard to the preterit structure, K ö g e l maintains that the contracted form of the Norse and Saxon is a remnant of the primitive condition, that Gothic habaida, O H G . habtta are late formations. Bremer (PBB. X I 46 sq.), in an essay on the Teutonic t, accepts Mahlow's opinion that the ai of haban is the equivalent of IndoEuropean ¿J, but rejects Mahlow's theory with regard to the development of the Teut. diphthong. Mahlow's argument is based upon the treatment of aim saian and waian.1 But, Bremer maintains, the ai in saian is quite different in nature and origin from the ai of haban. Saian points to no such form as *sa°jan, from *sSjan, but to *sean, and Goth, ai is here the representative of Greek at, which was, at the time of Ulfilas, the writing for open e. " D a s germanische ai kann nur aus einem vorhistorischen gj erklärt werden. Es muss also die idg. 2, 3 sg. *k'ab'e'si, *k'ab'eti, 2 plu. *k'ab'U3, im germ, zunächst zu umgeschaffen worden sein. Diese neubildung war sehr natürlich, denn sonst hatten alle verben, primäre wie abgeleitete, die endungen *-izi, -ifti." For the change of ¿J to ai, Bremer seeks to establish a law that Indo-European ¿J was transformed into ai in Prim. Teut. in unaccented syllables. The most important point in Bremer's essay is his discussion of the interchanging forms found in the present of the a«-verb. This interchange, Bremer holds, belonged to Prim. Teut., and is equivalent to the interchange of g and i in the Latin second conjugation, e. g. habgre, habitus. In Lithuanian a like duality of stem is found. The argument for a double stem is this (p. 47) : " Im germanischen entspricht i und a J. dem idg. i und g. Fraglich ist in welcher weise sich die formen ursprünglich theilten. Hätte das präsens im idg., wie im lit.-slav., überall i gehabt, so ist nicht einzusehen, weshalb got. haban nicht wie nasjan flectierte" 1 F o r literature on saian, waian see L e o M e y e r , K Z . 8, 245 sq.; P B B . V I I 152 sq., ibid. 46g sq., I X 50g sq., V I I I 210 sq., X I 51 sq., X V I I 14 sq.; Johansson, V e r b . Contr. 17g sq.; B r u g m a n n , I 128.

THIRD

CLASS

OF TEUTONIC

WEAK

VERBS.

413

T h u s B r e m e r ' s a r g u m e n t leads b y a different w a y to the conclusion of M a h l o w , that the Prim. T e u t . inflection of the ai-verb is to be found in the S a x o n dialects. T h e original inflection would appear thus: Pres. ind. sg\: 1. *xabjo pi.: 1. *%abjami O p t . pres.: *x.abjaj,

2. *xaba„izi

3.

*xa6ajf>i

2. *xabaj$>i

3.

*xabJariHi

etc.

A new direction was g i v e n the discussion b y Johansson. 1 He rejected at once the explanation of M a h l o w and that of B r e m e r , and held that the peculiarities of the «¿-inflection are to be explained as the result of a mixture of t w o original inflections. T h e t w o original conjugations w e r e : 1. t h e m a t i c : a ) ( = c l . I) -¿Jd, F) ( = cl. V I ) -J6; 2. a t h e m a t i c : sg. *khabhal-i-mi, plu. *khabh{3)-i-max-s. O f neither O l d H i g h G e r m a n nor of A n g l o S a x o n m a y it be said that the one is nearer than the other to the original condition. T h e former s i m p l y follows the athematic conjugation, while the latter shows a mixture of thematic a n d athematic forms. " C o g i t a r i potest, dialectos omnino u t r a m q u e flexionem retinuisse : quae dein.de diversis causis flexioni thematic ae faverent, form as quae jam essent, inter athematicas inseruisse : ac flexionis thematicae maxime sane erant propriae 1 sg. et 1, 3plu." (p. 183). T e u t o n i c ai should be considered, a c c o r d i n g t o Johansson, a d e v e l o p m e n t from I n d o - E u r o p e a n " G e r m . £ ( g o t . i s u b acc. princ.) etiam sed m o d o tum literis ai descriptum esse, cum in g o t . q u i d e m l i n g u a non s u b accentu principali s t a r e t " (p. 187). S t r e i t b e r g 2 and B a r t h o l o m a e 3 a g r e e with Johansson in presupp o s i n g for Prim. T e u t . a d o u b l e az-conjugation. B u t both these scholars maintain that T e u t . ai must h a v e c o m e , not from the athematic, but from t h e thematic, conjugation. T h e two original conjugations were, then, thematic in -e-jd, non-thematic in -¿-mi. T o explain the c h a n g e of primitive t to ä, S t r e i t b e r g , like B r e m e r , attempts to establish a T e u t o n i c law that " I n nicht haupttoniger (flexions-)silbe ändert sich die qualität des ¿, es wird g e m e i n D e derivatis V e r b i s contractis L i n g u a e Graecae (Upsalae, 1876), cap. V . Die germ. Comparative auf -öz-. Freibourg, 1890. 3 Altindisch äsis>Lateinisch sräs. H a l l e , 1891. 1

2

414

AMERICAN

JOURNAL

OF

PHILOLOGY.

germanisch zu ä." Hence Goth, haband is the direct equivalent of prim, athematic *xaße-ndi, and habais of prim, thematic *x a ß^ z - 1 Bartholomae's position is practically the same as that of Streitberg, but he differs from the latter in the method of developing thematic *hab£izi (Goth, habais). Streitberg introduces between *habezi and *habeizi an intermediate form *habei,izi, the eij. of which "unter dem drucke des übrigen formen sehr früh einsilbig werden musste." A s a simpler explanation Barth, (p. 147) proposes "dass man das gotische habais direkt auf ein ursprachliches *kh?bhUsi—so richtiger als *khabh°—bezieht, d. i. ein aoristpraesens aus dem verbal-stamm *khabhU-, welches sich zu dem durch lat. habes vertretenen aoristpraesens *khsbheizi nicht anders verhält als lit. buvai zu buvd." Streitberg and Bartholomae agree in considering ^ of the Old High German conjugation the direct equivalent, not of Gothic ai, but of Latin, and hence of IndoEuropean L A s Johansson makes O H G . habim, habfa, habit equivalent to Prim. Teut. *khabhaimi, *khabhaisi, * khabhaifti, so Streitberg makes the same forms equivalent to Prim. Aryan *khabh£mi, etc.2 All the scholars who have as yet been mentioned have agreed in identifying the Teutonic az-class with the Latin class in L A position quite different from that of Mahlow, Kögel, Bremer, was taken by Möller 3 and supported by Sievers. 4 They identify the az-verb with the Greek verb in -oa-, Prim. Teut. -öjd-. T h e original paradigm, then, would run : *habojizi, *habojifti, etc. By contraction -öji- became -öi-, and was then regularly developed into ai. According to Möller's law of syncopation, viz. that 1

S t r e i t b e r g ' s j u s t c r i t i c i s m of M a h l o w ' s v i e w w i t h r e g a r d to the c o n t r a c t i o n

o f Si to 6 ( D i e g e r m . C o m p . , 6 sq.) is quite as a p p l i c a b l e to S.'s o w n

argument:

" W e r sie d e n n o c h d a z u v e r w e n d e n w o l l t e , setzt sich d e m Vorwurf e i n e s Zirkels c h l u s s e s aus, i n d e m sein

ganzer

b e w e i s in d e r b e h a u p t u n g

besteht:

'die

e r k l ä r u n g d e r v e r b a a u f -ai- v e r l a n g t , dass ei zu ai, l zu ä, in n i c h t h a u p t t o n i g e r s i l b e w i r d , ' u n d 'ei w i r d zu ai, l zu ä, in n i c h t h a u p t t o n i g e r s i l b e , w e i l es die erklärung der verba auf ai v e r l a n g t . ' "

O u t s i d e the c o n d i t i o n of the az'-verb

itself, t h e o n l y e v i d e n c e to b e a d d u c e d in support of this T e u t o n i c f o u n d in fadar=

*pater a n d anstai = *anstei.

B u t fadar

e n d i n g of the a g e n t n o u n s , w h i l e anstai is p o s s i b l y d e v e l o p e d , not from b u t f r o m *atistöi. 2

l a w is

has w i t h o u t d o u b t the *anstei,

S e e U r s p r . a u s l a u t . ait B e z z . B e i t r . X V I I , p. 27.

T h e c h a n g e of ei to ai, of e to ä, is, a c c o r d i n g to S t r e i t b e r g a n d

o m a e , a common Teutonic

Barthol-

law ; s t i l l , this c o m m o n l a w is i n a c t i v e in o n e of the

dialects, O l d H i g h German.

I t is h a r d l y e v i d e n t w h y o n e d i a l e c t s h o u l d b e

e x e m p t from a p r i m i t i v e l a w . 3

P B B . V I I 472 sq.

4

P B B . V I I I 90-92.

THIRD

CLASS

OF

TEUTONIC

WEAK

VERBS.

415

after a short syllable Prim. Teut. 0 and ä w e r e lost, 1 there arose the forms *habjd, *habaizi, *habaihi, etc. There remains for consideration the recently published theory of Prof. Herman Collitz with regard to the origin of the az-class.2 Prof. Collitz's essay on the auslaut ai in Gothic, Old H i g h German, and Old Saxon has not received as yet wide recognition among scholars, but its importance must be eventually acknowledged. Prof. Collitz has succeeded in placing in a new light the whole question of the nature and origin of the «¿-conjugation, and proposes a solution of the problem radically different from any that has yet been offered. T o begin with the relation between Teutonic ai- and Latin : " Ich weiche," says Prof. Collitz, " von allen neueren Untersuchungen darin ab, dass ich einen unmittelbaren Zusammenhang dieser beiden flexionsklassen nicht annehme. V o n der bisherigen ansieht sich frei zu machen wird manchem zunächst schwer fallen. Man hat sich gewöhnt vergleichungen wie haban habere, silan silSre, \ahan tacire, witan videre, als vollgültigen beweis f ü r die ursprüngliche identität der beiden verbalklassen anzusehen. A b e r den lat. verben auf -£re entsprechen im Germanischen auch starke v e r b a ; aukan augtre, sitan sidere, ga-\airsan lorrere, wakan vegere, u. a. Dass mehrere germanische ai-verba zusammentreffen, erklärt sich zur genüge daraus, dass jene im Germanischen, diese im Laieinischen, die eigentlich intransitiv- und durativ-klasse bilden. Diese ihre gemeinsame function beruht nicht auf einem directen, sondern auf einem indirecten genetischen zusammenhange. Die lateinischen ¿-verba berühren sich nach form und bedeutung mit dem griechischen starken passivaoriste. Nimmt man nun mit Johansson ( K Z . 30, 553, anm.) an, dass der griech. 17-aorisl des passivs auf einer Verallgemeinerung des £ beruht, welches in der arischen Ursprache auf den auslaut des themas vor gewissen personalendungen beschränkt war, so ist die folgerung unabweislich, dass auch die lateinischen ¿-verba (und ebenso die entsprechenden bildungen im Griechischen und Letto-slavischen) aus einer eigenheit der arischen medialflexion erwachsen sind. Die lateinische • J o h a n s s o n , D e d e i i v . V e r b , contr. 182, 183, justly opposes Möller's theory of the T e u t . syncope, for, he says, " n u l l a est causa, cur syncopam eiusmodi generis tempore linguae germ, communis probemus, praesertim cum v o c a l e s ipsae accentum ¿JS prae se ferant, qui non subito mutatum sit." Hence Johansson proposes, not *khabhajd, but *khabhid, for the primitive form. 3

Beitr. zur K u n d e der indogerm. Spr. X V I I , p. I sqq.

4l6

AMERICAN

JOURNAL

OF

PHILOLOGY.

¿-conjugation hängt dann mit der german. az-conjugation ebenso nahe zusammen, ohne jedoch mit ihr identisch zu sein, wie in dem formensystem der ursprache die verbalen ¿-Stämme mit den verbalen az'-stämmen: das band, welches beide ursprünglich verknüpfte, ist die arische flexion des mediums." Prof. Collitz lays special stress upon the medial character of the conjugation, holding that in that fact is found the key to the whole problem. He points out that, of the thirty or forty verbs that may be ascribed to Prim. Teut., no more than four or five are derivatives, and these may be proved to be of late development. The az'-conjugation should not be treated as a weak conjugation, but as an old middle, which is connected with the weak conjugation only by the fact of their common adoption of an originally medial preterit. T h e present of the az'-verb has active endings, just as the original medial plural endings of the weak pret. were replaced by those of the active. " W i r dürfen aber erwarten, spuren des urspr. mediums noch in den besonderheiten zu finden, welche die praesensflexion der az'-verba charakterisieren. Hält man nun zusammen, dass das ai- im Urgermanischen sich auf die 2. und 3. person des praesens beschränkte, und dass in der arischen ursprache die 2. und 3. person des duals im praesens medii der 'thematischen conjugation' vor dem dental der endung (nach ausweis der ind. 2. du. -e-ihe, 3. du. -e-ti= av. -öi-\e) den ausgang -oi- hatten, so liegt der schluss nahe, dass das germanische -ai- nicht anders als die fortsetzung des thematischen -oider 2. und 3. person des duals ist." It is evident from the preceding historical review of the subject that the discussion from Bopp to Bartholomae has been guided by two assumptions, viz. (1) that the Latin ¿-verb is the same as the Teutonic az'-verb; (2) that the -j- of the Anglo-Saxon and Old Saxon is to be ascribed to Prim. Teut. The attempt to account for a -j- in the primitive conjugation led to Mahlow's complicated and arbitrary theory; it gave rise to the still more arbitrary, if more logical, view of Bremer. T o the same attempt is due the supposition of Johansson, Streitberg, and Bartholomae, that the az'-verb resulted from a mixture of two original conjugations. Prof. Collitz alone proposes a solution of the problem which is based upon the Gothic as representative of the original inflection. On this point he says: " Man sieht nicht recht, weshalb z. b. urgerm. *habjan (inf.) und *habjd (1. sing.) im Gotischen sollte durch haban haba ersetzt sein. Es ist jedenfalls ebensowohl

THIRD

CLASS

OF

TEUTONIC

WEAK

VERBS.

417

möglich von urgerm. *haban *kabS auszugehen und die nordwestgerm. *habjan *habjd als neuerungen zu fassen." 1 T h e object of the present paper is not to enter into the discussion concerning the origin of the az-conjugation, but to give an historical treatment of the «z-class and its development. There is much needed at this point, it seems to me, a careful consideration of the conjugation from the Teutonic standpoint. Hitherto, although the az'-problem has received its share of attention, no independent study of the verb has appeared. All the various theories referred to above have found expression either incidentally in the consideration of another subject, or in studies of a single aspect of the question. In every case, too, the primary object of study has been to ascertain the origin of the conjugation. It is my purpose, therefore, to confine my attention to the third weak class as it exists in Teutonic, hoping by that means to reconstruct the Primitive Teutonic az-class and the Primitive Teutonic «¿-inflection. T h e reconstruction of the az-class will involve (a) a collection of all the az'-verbs that may be ascribed to Primitive Teutonic, (b) a study of the dialectic development of the class. P A R T I.

A . — The Primitive

Teutonic

ai-verbs.

T w o difficulties stand in the way of determining with certainty which of the ai-verbs may be ascribed to Primitive Teutonic. T h e first of these difficulties arises from the fact that the distinction between the three weak classes is in none of the dialects kept with absolute integrity. Even Gothic, which preserves the purity of its forms with much greater consistency than do any of the other dialects, shows beside hausjan hausjdn, beside beistjan beisijdn; and in the az-class hatjan appears beside hatan, with no apparent distinction of use or of meaning. In the other dialects, so uncertain are the lines of demarcation between the classes that, without the most cautious comparative study, it is impossible to determine the original condition of any given verb. When we find, for example, in Old High German sagjan, sagen ; fr tig en, fragdn, what shall be said about the relative antiquity of the forms ? In general it may be assumed that, where we find in Old High 1

D i e B e h a n d l . des urspr. auslaut. -at, 43, n o t e .

418

AMERICAN

JOURNAL

OF

PHILOLOGY.

German variants in and -6-, or in -6-, -6- and -j-, the -¿-form is probably the oldest. But this assumption can become a certainty only in case the verb exists in Gothic in the «¿-inflection, or is an underived verb. That in the case of derivative verbs, where variants are found, the /-form is not necessarily primitive might be evidenced by numerous examples. T h e following instances are the result of a study of the ¿-conjugation in O H G . and Gothic: Gothic kardn-, O H G . charSn, chartn ( A g S . cearian, -ode; O S . karon). Gothic la\dn; O H G . ladon, ladin ( A g S . Ionian, -ode-, O S . laftoian). Gothic smi\dn; O H G . smiddn, smiden (AgS. smiHian, -ode). Gothic tilon; O H G . zildn, zilhi ( A g S . tilian, -ode-, O S . tilori). Here, without doubt, the ¿-forms are comparatively late, of purely O H G . development. In doubtful cases A g S . and O S a x . are of no practical assistance; for in those dialects the az-conjugation has no longer an independent existence, the original az-verbs still existing there have passed into the ¿-class. Norse, too, is unreliable, for although it has kept the «z-conjugation, the class is a very small one of mixed character; and, moreover, many of the verbs which belonged without doubt to the Prim. Teut. az'-class appear in Norse among the verbs of ¿-inflection or of the -j-. Another difficulty in deciding which were the Prim. Teut. aiverbs lies in the fact that the classes by which the «¿-conjugation is represented in O H G . , A g S . and O S a x . are in those dialects the main classes of new formation. Consequently there will be found many denominatives common to the three dialects, having the treatment of at-verbs, and still of late origin. Furthermore, verbs that belonged in Prim. Teut. to other classes may in the dialects be treated as verbs of the az'-class. Since in O H G . the third class received special stress as a class of new formation, it is not unnatural that verbs originally of other classes should be drawn into the prevailing class. T h e same may be said of the second class in A g S . and in O S a x . T h e following verbs may be ascribed, without hesitation, to Prim. Teut.: 1. Goth. ctisian\ ON. ¿esia (pret. czsta). 2. Goth, arman; O H G . arm£n-, O S . armon-, A g S . earmian.

THIRD

CLASS

OF

TEUTONIC

WEAK

VERBS.

419

3. OHG. biben1-, AgS. bifian.beofian-, OS. bib on; ON .bifask? 4. Goth.Ji(j")an; OHG. fien; KgS.feon,fe'ogean\ O N . f j h . 5. OHG.folgen*\ AgS.folgian,folgode*; OS.folgon. 6. O H G . f r & g b i l \ OS.fr Agon. 7. O H G . full en; O S .fullon\ AgS. fullian; O N . / m & ( - « S Z ) . 8. OHG.giSn,ginin*\ AgS.ginian,geonian; ON.gina(giri6i. Perhaps late formation and not connected with OHG. giin). 9. Goth. haban\ OHG. habin\ OS. hebbian-, AgS. habban\ ON. hafa. 10. Goth, hahan; OHG. hangen ; OS,hangon\ AgS. hangian ; ON. hanga (only in pret. hangf>i and in present). 11. Goth, hatan ; OHG. hazzen (rarely hazzSn) ; AgS. hatian; OS. haton\ ON. hata (-¿iSz). 12. OHG. hlinen7; OS. hlinon ; AgS. hlinian, hleonian. 13. OHG. hoghi*\ OS. huggian; AgS. hycg(e)an (late hogian). 14. OHG. kleben ; AgS. clijian, cleofian] OS. clibon ; ON. klifa (-aSz). ' O n c e a form in is f o u n d ; pret. ir-bibtiten, Graff, I I I 22. 2 In the oldest Norse writers Ufa is found as a deponent verb. Later the pret. bifafti appears in place of the older bifSi; and with the change in pret. the verb assumes an active meaning. Cf. Cleasby, Vigfusson, 62. ' T w i c e ¿-forms are quoted, both times from W i l l i r a m , Graff, I I I 512. Note also A g S . fylg(e)an, O N . fylgja. I am inclined to presuppose for Prim. Teut. two v e r b s : *folgan, represented in A g S . folgian, O H G . folgen, OS. folgon ; *fulgjan, represented in A g S . fy!g(e)an, O N . fylgja. Sievers ( O E . Gram. 416, n. 5) reckons A g S . folgian among those original «¿-verbs which have in A g S . " a more or less perfect double formation." Such double formations are not infrequent in Prim. T e u t . 4

6 ^-forms are not infrequent in O H G . , but the form of the verb shows that fr&g£n could not have been originally of the »-class. W e a k verbs showing in their stem the third ablaut of a strong verb are found only in the «¿-class or in the /z-class. Cf. Goth, fraihnan, frah, frchum^ fraihans. 6 T h e n- must have belonged in Prim. T e u t . to the present alone. Cf. also L a t . hi-are, K s l . zijati and zinati, F i c k , 434. T h e relation between glen and ginen is the same as that between Gothic keian and keinan. 7 C f . Gr. KTLCVU, L a t . in-clino, -are. T e u t . *hlinan and causative *hlainjan ( O H G . hleinjan, A g S . hlanan, O N . hleina) point to a lost strong verb *hleinan, *hlain, *hlinum. In T e u t . , as in L a t i n , the -re- originally characteristic of the pres. has extended to the whole conjugation. Cf. gien, gin£n above. 8 In O H G . hogln is found only occasionally with the commoner inflectional forms of huggan. Evidently in O H G . the «¿-verb became confused with the original /-verb, *hugjan (Goth, hugjan, O N . hyggja, O H G . huggan). Prim. T e u t . *hogan, *hugjan are parallel with Prim. T e u t . *folgan, *fulgjan noticed above. T h e /-forms of A g S . and O S . are not to be assigned to the prim, / - v e r b , but to the A g S . development of the (¡¿-inflection.

420

AMERICAN

15. Goth .kunnan 16. Goth. liban\ Ufa

O N . lika

PHILOLOGY.

; OHG. kunnin ; OS. kunnon ; AgS. kunnian. OHG. leben ; OS. libbjan ; AgS. libban ; ON.

OHG.

leikan;

OS.

lieehen;

AgS.

likon;

lician-,

OSz).

18. Goth.

O N . lofa.

19. Goth. 20. Goth, 21. Goth. ON.

(in

*luban

C f . L a t . libere

O N . skamma

OS.

lob£n\

lubere)

; Skr.

AgS.

lobon\ lubhyati.

OS.

mornon\

AgS.

runen;

AgS.

riinian\

A g S . secg(e)an;

ON.

morncn\

reiran,2

(in

* rim an

OHG.

riinains);

O S . seggian-,

segja.

si/an.3 silaji.

Cf. Latin

silere.

OHG.

seamen;

slavan.3 skaman\

AgS.

scamian,

sceamian-,

(-¡282').

27. Goth, saurgan1-, 28. O H G . swighi;

29. Goth,

OHG.

(=

morna.

2 2 . O H G . sagen;

Goth, Goth, Goth, Goth.

OHG.

lubains)-,

maurnan\

murndeON.

23. 24. 25. 26.

OF

(-afti).

17. Goth,

lofian\

JOURNAL

staurran;

30. O H G .

OHG. OHG. OS.

OS.

sorgim;

A g S . swigian\

AgS.sorgian.

swigon.

ON.

stiira.

A g S . talian\

ON.

storren\

talon-,

sorgon;

OS.

tala.

31. Goth. trauan\ OHG. trtiin, trfiwin-, OS. truon; AgS. trilwian; ON. /"rz/a. 32. Goth, \ahan; OHG. dagen; OS. thagon\ ON. \egja. 33. Goth, \ivan; A g S . \eoivian. 34. Goth. \arban\ OHG. darbin\ OS. iharbon\ A g S . J>