BIBLIOTEKA Instyfutu Bałtyckiego
w Sopocie
6 0 Î b5" ÜI
SKRIFTER
UDGIVNE AF
VIDENSKABS-SELSKABET I CHRISTIANIA
1910
\ \ fC!„
II. HISTORISK-FILOSOFISK KLASSE
CHRISTIANIA
I K O M M IS S I O N H O S J A C O B D Y B W A D A . W . B R0GGERS B O G T R Y K K E R I
CRITICAL CONTRIBUTIONS
T O
EARLY ENGLISH SYNTAX
Dr. A. TRAMPE B0DTKER
S E C O N D S E R I E S :
IV . P E R S O N A L P R O N O U N S . V . D E M O N S T R A T IV E P R O N O U N S V I . R E L A T I V E P R O N O U N S . V I I . W H A T
( Vi d e n s k a b s-Se l s k a b e t s Sk r i f t e r. II. Hi s t.-Fil o s. Kl a s s e 1910. No. 3)
U D G IV E T FO R H . A. BEN N E C H E S FO ND
C H R IS T IA N IA
O N C O M M IS S IO N B Y J A C O B D Y B W A D
A . W . B R 0 G G E R S B O G T R Y K K E R I.
I he R elative Pronouns in E a rly E n g lish have been studied b y E. A . K o c k in his d isse rtatio n: The E nglish Relative Pronouns, Lun d 1897, and m ore especially b y H . G r o s s m a n n : Das angelsächsische R elativ, Diss. B e rlin 1906, and E. A n k l a m : Das englische R elativ im
11
. und12
. Jahrhund ert, Diss. B e rlin 1908. T o the b ib lio g ra p h y contained in these m onographs should be added: L. R. W i l s o n , Chaucer's Relative Constructions, Studies in P h ilo log y, Phil. C lub o f the U n iv e rs ity o f N o rth C arolina 1906, and H . E n g e l, Spenser's Relativsatz, Diss. B erlin
1908.
J. E. W ü l f i n g gives in his S yntax in den Werken A lfre d s des Grossen, Pt. I, pp. 419— 421, a survey o f the p ub lica tion s on the omission o f the R elative Pronoun in English. I m ention E i n e n k e l ’s articles in the A n g lia X I I I , pp. 348 ff., X I V , pp. 122 if., and X X IX , pp. 121 ff. A s regards the question in general it suffices to refer to B. D e l b r ü c k , Ver
gleichende S yntax der indogermanischen Sprachen, I I I , pp. 381 ff., and to H . P a u l, P rin z ip ie n der Sprachgeschichte Ir ■ pp. 140 f. I have n o t y e t had access to B. D e l b r ü c k , Z u den germanischen Relativsätzen, A b h . d. K g l. Sachs. Gesellsch. d. W iss. H is t.-P h il. K l. 1909.
T he Indefinite Pronouns have been treated exha ustive ly b y E i n e n k e l in the A nglia. I have added o n ly a few rem arks on ivhat. So far as I know , n o th in g has been w ritte n o f late on the e a rly h is to ry o f the othe r pronouns.
In the M odern Language Review I V , pp. 433 ff., N a p i e r has dem o
lished H e u s e r ’s th e o ry o f the e a rly date o f the Ancren R ude. N a p i e r states th a t »all the available evidence does n o t take the Ancren R iw le back fu rthe r than the b e g in nin g o f the th irte en th century«, and thus lays a solid basis fo r syn tactica l researches in th a t otherw ise im p o rta n t w o rk .
I have had the pleasure o f seeing the F irs t Series o f these C ritic a l C ontributions reviewed b y W ü l f i n g in Deutsche L ite ra tu rze itu n g , 1909, col. 2283, and b y E i n e n k e l in B eiblatt z u r A n glia, 1909, pp 267— 273.
A ddenda to the chapters already dealt w ith w ill be em bodied in some later Series. Some rem arks on a frie n d o f mine have been published recently in B eiblatt z u r A n g lia , 19x0, pp. 56 ff.
I Ç I O . No. 3. C R IT IC A L C O N T R IB U T IO N S T O E A R L Y E N G L IS H S Y N T A X . 5
IV . Personal Pronouns.
40
. A c c o rd in g to E. W in d is c h 1, French has borrow ed from Celtic the use o f the intensive phrase c’est que, as in c’est a vous que je madresse; and from F rench again A . T o b le r derives 2 the use o f corresponding phrases in m any other languages. T h e French construction can, how ever, be traced back to L a tin . Vos estis continues th ro ug h cestes vos in c’est vous, and s im ila rly vos estis qui gives the m odern cest vous qui, as seen in M at. X , 20 non enim vos estis qui loquim ini, sed S p iri
tus P a tris vestri qui lo q u itu r in vobis. T h is runs in the O E . Gospels ne synt ge na pe peer sprecad ac eotvres feeder gast pe spry cd on eow3.
I fu rthe r m ention O rm . 8465 patt ivass pe land o ff Galileo patt himm ivass bedenn sekenn. In spite o f French influence, this »um ständlichere Ausdrucksw eise« 4 is far from being so cu rren t in m odern E n g lish as i t is in Scandinavian, where the usual w a y o f p u ttin g ‘w h o comes?’ is:
hvem er det som kommer? — F rench qui est-ce. qui vient?
41
. E inenkel m entions in G ru n d r. § 173 v the O E . use o f h it in tro d ucing a tem poral statement, b u t »wo das Me. comen gebraucht, haben wie afrz. Einfluss, quant ce v in t le dym anche. . . > ivhan h yt cam on the m orn > i t drew to pe night«. A n e arly instance occurs, however, in S. Chron. E 1087, p. 223, sona swa h it com to pam Eastron, pa ferdon h i % hergodon, etc. L a te r instances are : O rm 8917 ^ 3edenn heore iv e ^e fo rb t ill patt i t t comm t i l l efenn, C ott. Vesp. H orn. p. 231 pa h it per to com pat se lila fo rd in to par halle come, ib. p. 235 1 vat h it com to pa time.In late M id E . i t drew is o f frequent occurrence: Ip o m . A 7056 h it drew to the nyghte faste, etc., com pare Scand. det drog ud paa natten.
1 Grundriss der romanischen Philologie I, p. 402.
2 Vermischte Beiträge zur französischen G ram m atik I, p. 160. A n account o f the French con structio n is also give n b y P. Jochim sen, Beiträge zur Geschichte der deiktischen Hervorhebung eines einzelnen Satzteiles, bezw. eines Satzes m ittelst c’est ( . . . ) que (qui), Diss. K ie l 1907.
3 O th e r exam ples are cited b y L . K e lln e r, H is to ric a l Outlines o f Bnqlish Syntax, p. 179.
4 H . Paul, P rin zip ie n der Sprachgeschichte IV, p. 285.
42.
»Die em phatische W ie d e rh o lu n g des Fron, am Schlüsse des Satzes is t w ahrscheinlich frem d, obgleich frz. je le veux, m o i! bis je tz t n ic h t so früh belegt is t w ie me. B u t he can cherles daunten, he, . . . .> I wot not, I ! « (G n m d r. § 173 0). I wot not I (corresponding to c o llo qu ia l Scand. jeg ved ikke jeg) belongs, however, to the same ca teg ory as Bede 473,12 (S m ith) h it is welig dis E alond (see W ü lfin g , S yntax I, pp. 343 f.), K e n t. Serm. (Mise. p. 27.) pet ne seide he nocht herodes, L a m b . H orn. p. 27 he penched pe deofel, etc. T he pronoun, like the noun, is added to define e m p h a tica lly the atonic subject preceding the stressed verb h
43.
L ik e F rench comme cil qui, Scand. som den der (som en der), etc., E n g lish as he that (as one that) is liab le to take a causal shade.E inenkel (G ru n d r. § 173 ip) quotes as he that o n ly as late as Chaucer, and th in ks th a t an earlier as pe pat is m odelled on French. In re a lity, he and pe, as m ig h t n a tu ra lly be expected, appear n ea rly a t the same tim e : V ic . a. V . p. 119,25 pus us aliesde ure aliesend and sedden aros o f deade, alswo he de was sod lif-, a little later A ne. R. p. 388 [h e ] scheawede h ire his fe ire neb ase pe pet ivas o f alle men ueirest to biholden 2.
44.
»Das von einer adverb. B estim m ung begleitete Pers. in demon- strat. Sinne is t dem A frz . nach ge bilde t: C hil de la cité v in re n t contre lu i, — chiaus de dedens et chiaus de dehors Froiss > me. th a i o f Scotland Barb. > Sehe passed hem o f Y p ris . . . Ch.« (G ru n d r. § 173 ßß).E a rly instances, besides G o th ic parallels 3 4, are how ever given b y Mätzner, Gram. I l l , p. 351: S. C hron. E 1129 (p. 260) da o f Home, L a 3. I p. 252 heo o f Borne, etc., to w hich I add S. Chron. E 1140, p. 267, h i o f N o r- m andi wenden alle f r a pe king i , T rim H orn. p. 115 hie pe per-inne tuer en pus andswerden. . . % po widuten seiden . . , V ic . a. V . p. 147,5 swide nied- fu lle to dan inede. T h e interchange o f the personal and the dem on
strative pronouns needs no com m ent. T h e h is to ry o f the personal p ro noun in E n g lish as w e ll as other G erm anic languages offers abundant examples o f its in tim a te connexion w ith the dem onstrative. In sim ilar
1 See F a lk og T o rp , DansJc-Norskens S yntax, pp. 275 ff. ; H . W u n d e rlic h , D e r deutsche Satzbau I, p. 2 3 1 ; S w e e t N . E . Gr. II, p. 5 and p. 72 ; H . Paul, P rin z ip ie n Iv p. 127 (w ith references).
2 A te m p o ra l shade is pe rce p tib le in O r. o. o. Lo u e rd e (O. E. H orn. I, p. 185) Jliesu a l fe ir a^ein hwam pe sunne nis boten a schadwe ase peo pet leosep here liht.
3 See also G rim m , Deutsche Grammatik I V 2 p. 464 and p. 528.
4 S. C hro n. D 915, p. 99 pa gemytton hy o f Hereforda ? o f Gleaweceastre has n o th in g to do w ith the above con structio n as sh o w n b y A pa gemetton pa men hie o f Here
forda ï o f Gleaweceastre, in w h ic h o f doubtless belongs to pa men, and no t hie.
phrases Scandinavian uses the personal p ro n o u n : de f r a byen, co llo q u ia lly han, h im f r a byen. A kin d re d construction o f the dem onstrative pronoun w ill be dealt w ith in Section
46
.45
. »Das reflexive V e rh ä ltn is kann auch ausgedrückt w erden m it H ilfe des Passivs, nam entlich bei den Verben des Setzens, Legens etc.und zwar nach dem M uster des A frz . . . . > m e . they were sette — I was leyde — lie was clad Ch., gelegentlich in K reuzung m it R eflex. P ro n .:
These rio tto n rs . . . Were sette hem in a tavern id.« (G ru n d r. § 175 y).
K e lln e r (Engl. Stud. X V I I I , p. 289) also takes be as an a u x ilia ry to form the passive: he was leyd. sett; they were mette, and compares such O E . instances as Caedm. Gen. 120 pa lews gast o f er holm boren.
Wees boren does n o t count as the passage s im p ly renders Gen. I, 2 et sp iritu s dei ferebatur super aquas, in H ilf r ic ’s translation godes gast wees geferod o f er weeteru. Meten takes be after the analogy o f intra nsi
tiv e verbs o f m otion, w hich in Chaucer are n ea rly alw ays constructed w ith be. Be clad m ay be com pared to the m odern expression be dressed, Scand. veere klcedt, in w h ic h the p a rtic ip le has the function o f an adjective.
In be set the old p articip les seted and ¿eten have fallen together.
T h e confusion began already in O E . (see Sievers Ags. Gram. $ 406 a. 7), and w e n t on increasing in M id E . O rig in a lly , be set does n o t con
tain settan b u t sittan, as shown b y S. Chron. A , etc., 807, p. 90 fa w urdon eac sivide unedelice dseten. Preo asoeton on da healfe pass deopes de da Deniscan scipu aseten weeron (B geseten w aran), La;-;. 18532 pa heo weoren alle iseten (Mätzner, Gram. II, p. 81), H a v e lo k 2291 he weien alle dun set. A confusion w ith the passive was not, how ever, far o ff:
H a v e lo k 1722 panne [h e ] were set, and bord leyd, A n d pe beneysun was seyd.
■ A n o th e r source o f confusion was the ethic dative in O E . him sittan
> H a v e lo k 633 he sat him up. T he dative, w h ic h m ig h t be m istaken for an accusative, was o f course k e p t in the com pound tenses. these rio tto u rs . . . were sette hem. F rench s’etre assis b ro u g h t in a th ird element.
Be [ h im ] set p ro b a b ly acted on be him laid. T he earliest instance I have m et w ith is K ent. Serm. (Misc. p. 32) and ure lord was i-leid him don to slepe ine po ssipe er pane pis tempeste a-rbos.
1 9 1 0. No. 3. C R IT IC A L C O N T R IB U T IO N S T O E A R L Y E N G L IS H S Y N T A X . 7
V . Demonstrative Pronouns.
46
. In his Gram. I l l , pp. 254 f. (cf. also p. 351) M átzner treats o f that refe rring back to a preceding noun and qualified b y an adjunct, as in lie hathe the spere heed . . . hut i t is grettere than thai at P aris, M aund., [th is croune] was o f jonkes o f the see . . . fo r I have seen and beholden many tymes that o f P a ris and that o f Costantynoble, Maund.N o earlier exam ples are quoted. M atzner compares M id H G . D ie Gunthéres man unde ouch die D ietriches b u t does n o t find a n y th in g in O E . th a t comes nearer to this construction than M a rk X I I, 17 agyfad pam casere pa ding pe pees caseres synd, and Gode pa de Godes synd 1 2.
K e lln e r [H ist. Outl. p. 199) repeating M atzner’s examples, says th a t the use o f that in connexion w ith the g en itive is scarcely to be traced back to O E .
Einenkel [G ru n d r. § 179 adds o n ly a F rench parallel.
These statements are rath er b e w ild e rin g ; fo r i t can be easily proved th a t the co nstructio n existed already in O E . \V e need n o t discuss an exam ple lik e S. Chron. A 921, p. 101, se here . . . worhton pcet geiveorc ast Tcemeseforda, 7 h it budon 7 bytledon 7 forleton pcet oper cet H im ta n - dune; i t suffices to q u o te : Cart. Sax. I I, p. 57— 58 sellad . . . det land cet Sempigaham . . . bruce . . . des landes cet S lioforda 7 . . . pes on Sempi- gaham, L an d. C hart, p. 361 pet land cet bradan wcetere. 7 pet cet Nievan túne, ib. p. 365 ic gean m inum hlaforde pes landes cet lam burnan 7 pees cet ceolsige 7 cet readingan, D ip lo m . A n g l. p. 596 ic gean HJlfgare minum suna pees landes cet Hwipstede 7 pees cet Wealtune, S. Chron. C 912, p. 96 peer da burh getimbrede. 7 pees ilcan geares pá cet Bricge, and s im ila rly in the annals fo r 913—
9
J5
» S. Chron. E 1038, p. 161, pa feng Eadsige biscop to pam arcebiscoprice 7 G rym cytel to dam on Sud Sexum, ib. 1094, p. 229, se eorl . . . gewánn pone castel cet Argentses . . . 7. syddan pone cet Hulme, ib. 1066, p. 198, se cyng g e a f. . . him f> abbotrice on B y rtu n e .% se o f Couentre f> se eorl B eofric pe woes his earn é r heafde macod. 7 se o f C rulande. 7 se o f Porneie, ib. 1086, p. 222, se abbod o f Badon and pe o f Perscoran, ib. 1090, p. 225 he begeat pone castel cet sánete W aleri
7 pa hcefenan. 7 swa he begeat pone cet Albemare, ib. 1096, p. 232, m id him se eorl o f F la n d ra n 7 se o f B unan, ib. 1100, p. 235, f bisceoprice on Winceastre 7 f on Searbyrig, ib. 1108, p. 242, ivurdon syddon manege gew inn betwux pam cynge o f France 7 pam o f Englelande, ib. 1120,
1 See fu rth e r G rim m , Deutsche G ram m atik I V 2, p. 528, and com pare S ection 4 4. 2 In M a t. X X I I , 21 and L u ke X X , 5 da ping de Godes sgnd.
I9IO. No. 3. C R IT IC A L C O N T R IB U T IO N S T O E A R L Y E N G L IS H S Y N T A X . 9
p. 249, w urdon sehte seo cyng o f Englelande % se o f France . . . acor- dedan . . . Heanriges agene mcen . . . % se eorl o f F la n d ra n . % se o f P u n - tiw. Cf. also an instance from L ib e r V ita e, W inchester, in Section
50
. In the same w a y Scandinavian uses the dem onstrative pronoun to avoid the rep etitio n o f the n o u n : huset i byen og det paa landet — et hus starve end det i byen — vcerelserne ovenpaa og de nedenunder — dette lo rd og det der (borte). T h e last tw o instances, in w hich the p ro noun is follow ed b y an adverb, correspond to M atzner’s e xam ple: the upper p a rt o f him the blow had s lit as sure as that beloiv.47
. M ost E n g lish dialects use this liere, that there as dem onstrativ e adjectives: this here child, that there child. T he N . E . D . and Ein- enkel [G ru n d r. § 1 7 9 # ) derive this use from F re n c h : ce livre-ci, ce livre-ld. T o this Logem an rig h tly objects th a t the w o rd -o rd e r is a lto gether d iffe re n t in the tw o languages; he suggests influence from Scand.
denne her m a n d fe n ] 1. I am inclined to believe th a t a parallel develop
m ent to o k place in Scand. and E nglish.
Scand. denne and the adverb her o r der n o w form a com pound em phatic pronoun, w h ic h is consequently used also as a dem on
strative adjective. I t w as p ointed out above th a t the use o f that in connexion w ith an a djun ct has always been a cu rren t construction in English. In the course o f tim e that there m ay have come to be felt as a com pound pronoun lik e Scand. denne der; this here was form ed after the a na lo gy o f that there2. The new com pound can be com pared to the p rim itiv e form o f this its e lf and to the Romance com pounds o f ecce + ille (iste).
48
. T he ju x ta p o s itio n o f this and that, as in you that way, — we this way Shak., appeared, a ccording to G ru n d r. § 179 y, in the 14th c e n tu ry in im ita tio n o f F rench c il and cist. T he dem onstrative frequently takes an ind efin ite character. In O E . othe r pronouns were used, e. g.Boet. 190,9 Sonne lufap sum feet, sum elles hwwt. T h o u g h this and that already in O E . denoted w h a t was nearer and w h a t was fa rthe r o ff (see exam ples in W iilfin g ’s S y n ta x I, § 256 and § 262), th e y do not seem to be used in ju x ta p o s itio n lik e O H G e rm . noh th iz i noh thaz (G rim m ,
1 H e rrig ’s A rchiv 117, p. 45- . .
2 In this one reproach, any one name, and the like , one stands in ap position to this, any, and is perhaps d iffe re n t fro m one used as a substitute fo r a noun. See exam ples in J. E llin g e r, Vermischte Beiträge zur Syntax der Neueren Englischen Sprache, W ie n 1909, p. 50
Deutsche Gram. I V 2 p. 527). B u t the d is tin c tio n between the tw o pronouns was soon taken recourse to in E a rly M id E . O rm contrasts n o t o n ly piss and 3onnd, as stated b y Matzner, Gram. I I I . p. 251, but also piss and patt: 19429 w hatt Abraham, w hatt Moysces, w hatt tiss ^ ta tt profete.
V I. Relative Pronouns l.
49.
T he O E . clauses in w hich a relative pronoun is a pp aren tly w anting, belong to tw o d ifferen t categories, exem plified in :1. S. Chron. A 906 on pys geare gefor J A lfre d , wees cet Babum gerefa.
2. Sol. 182,31 [ i c ] fagnige pees pu cwyst.
T o the construction into xoivov presented in the la tte r instance also belong clauses co nta ining ponne, as Oros. 2, 3, 8 geortruwige Gode poet he us ne mcege gescyldan to beteran tidon ponne we me on synd. S im i
la rly pu stands fo r pu pe o r pe: Elene 726 feeder ure pu eart on heo- fonum 2, w h ic h m a y also be explained as a case o f parataxis.
In the O E . p arata ctica l construction gefor JE lfred, wees cet Babum gerefa, the subject o f the second clause is to be supplied from some noun, not necessarily the subject, in the first, and it is n early alw a ys follow ed b y hatan o r wesan. T he p a ra ta x is im p e rc e p tib ly changes in to a h ypo taxis, A lf r e d being felt as b elon gin g to both clauses; but this stage was scarcely attained in the O E. period. A m o n g the exam ples from e arly M id E . te xts quoted b y A n k la m , l. c. p. 9, there is re a lly o n ly one th a t m arks a step fo rw a rd : A n c. R. p. 408 fu s he speheb to Aloyses, pet monne mest him heuede. M id E . usage s till continues in sentences like here is somebody ivants to see you.
Before discussing the absence o f the pronoun in the object-relation, as in the man
1
saw, it w ill be w e ll to exam ine the earliest instances m et w ith.A n k la m , l. c. p. 8, quotes Gesetze d. A g s . p. 4 5 2_gyf he, bonne eat wet gefridab he healdan sceal, bonne bib he godes leanes f id wet wyrbe, but does not m ention L ie be rm a nn ’s note on the passage: d id the orignal read —bad f ? Ges. d. A gs. A 314 habbe ponne ylcan dom se be pcet feds worhte is s im p ly to be discarded.
1 F u ll title s o f books re fe rre d to are give n p. 3.
2 B o th quotations give n b y Grossmann, I. c. p. 8 and p. 69.
1910. No. 3. C R IT IC A L C O N T R IB U T IO N S TO E A R L Y E N G L IS H S Y N T A X . I I
I am e qu ally d iffid e n t w ith regard to an instance I have found m y self: S. Chron. F 1 995, p. 128 and sona o f s . . . ealla fa wisuste menn he awar gecneow rendering Hague prcecepit congregari omnes quos scie- bat prudentiores (Plum m er, A ppendix B, p. 285). A s m any other sm all w ords have been o m itte d in this M S., it is probable th a t pe or f was dropped ina d v e rte n tly b y the scribe; some lines b e lo w he le ft out f \ him tealdan pa swybe ealde menn . cegder ge gehadode ge lawede [ f ] heora y ld ra n heom tealdan hu h it was gelagod: tunc seniores et prudentiores n a rra u e ru n t ei, quod inde a suis patribus audissent.
A n k la m , l. c. p. 9, is perhaps rig h t in considering A n c. R. p. 188 pertec al he polede as due to a clerical error. T he c o nte xt shows th a t he polede is a subordinate clause. I t coincides w ith the exam ple quoted from the Law s in so far as in both cases the pronoun is w a n tin g after all. A n k la m fu rth e r adduces T rin . H orn. p. 93 bet Oder g ostliche shrud ic h embe spece : is mildhertnesse. I have found in Pater N oster, Lam b. H orn. p. 55:
halde we godes la^e pet ive habbed o f his sa^e.
pa bodes he beodeb per inne
Bute weo hes halden : we dob sunne.
L a te r M id E . instances, from Gen. and E x ., Brunne, etc. have been collected b y E in e n k e l2.
E inenkel distinguishes between tw o types o f relative clauses in w hich the relative pronoun is w a n tin g : The o ld er and more n orthern ty p e is seen in that swerd he wan o f S ire Cesar B y hym in grave they leide h it ih a r (Brunne). T h is construction in w hich the connecting lin k is p u t first, is unknow n in P'rench. In the other type, w hich appears some hundred years la te r and is independent o f the first, the connecting lin k is placed between b oth predicates, as in he had a sone, men cald E cto r (Brunne). H ere F rench influence has been at w o rk .
A s show n b y Gen. and E x . 2167 nu, bi be feib ic og to king pha- raon, sule ge nogt alle eben gon, a fro n t-p o s itio n m ay easily change in to a m id -p o s itio n ; be feib belongs ju s t as w ell to the p re p o sitio n bi as to the fo llo w in g og. It is scarcely safe to urge the difference o f the tw o types, either as regards lo c a lity o r an earlier appearance o f the one.
I do n o t th in k the p ro ble m can be solved a ccording to one fo r
mula. In F rench the pronoun is ra re ly o m itte d unless the p rin c ip a l
1 F ro m about A . D . n o o , see P lum m er, P t. II, p. X X X V I and p. X L IV . 2 See Artgl. X I I I , pp. 348 ff., X IV , pp . 122 ff., X X I X , pp. 121 ff.
sentence contains a negation, nor does the e a rly appearance o f the fron t- p osition favour the th e o ry o f French influence. Jespersen holds the v ik in g settlem ents responsible fo r the loss o f the p ronoun L B u t to this Einenkel objects th a t the relative is never o m itte d b y O rm , w h o h o w ever a few tim es leaves out the co njun ction that (see Series I, p. 48).
I th in k foreign influence m et w ith a congenial tendency in the language itself.
Gefor jE lfr e d wees gerefa had developed in to a construction tmo v.oivov, perhaps to w ards the end o f the tw e lfth century. A sim ilar but independent process m ay have taken place in the man I saw.
The ¿-w ords in O E . and s till m ore the b old O H G erm an clauses show th a t a construction ano v.oivov agrees v e ry w ell w ith fu ll case- endings. O n the other hand i t cannot be denied th a t the loss o f the O E . in fle x io n s offered new facilities to the language. A noun placed at the head o r in the m id dle o f tw o clauses could m ore easily be felt as b e lon gin g to both.
A p rim itiv e structure such as the man you saw — the man is my brother could scarcely develop on its o w n accord the man you saw [h e ] is my brother. T he w ord-o rd er w ould be too unusual. T he same a rg u m ent applies to I ate the apple — [the apple] John gave me. B u t the w ord-o rd er established in relative clauses m ig h t lead to a com prom ise resu ltin g in the man you saw . . . , the apple John gave me. I t is o f course im possible to decide how far Scandinavian o r French influence was concerned in this developm ent.
5 0
. A n o th e r feature w hich E n glish and Scandinavian have in com mon, is the place o f the p re po sition afte r the verb in a relative clause, as in O rm 3470 ff.Acc f r a pait Kalldewisshe la n d, PaU teffi pa comenn off'e, Wass m ikell w e ^ e t ill patt land Patt C rist wass borenn inne.
Jespersen 1 and Onions 2 th in k th a t the E n g lis h w o rd -o rd e r is due to Scandinavian. A few O E . instances have, how ever, been found b y Grossmann, l. c. p. 16, in the M etres o f Boethius and v E lfric ’s G ram m ar.
I have m et w ith an instance in L ib e r V itae, N ew M in ste r and H y d e A b b e y , W inchester, ed. de G ray Birch, p. 161 pcet is . . . o f dam stáne
1 Growth and Structure o f the English Language, p. 82.
2 A n Advanced English S yntax, L o n d o n 1905, p. 104.
1910. No. 3. C R IT IC A L C O N T R IB U T IO N S T O E A R L Y E N G L IS H S Y N T A X . T3
pe seo rod stod on uppan de ure dryhten ondrowode % o f pcere binnan be lire dryhten onlceg1, and another in S. C hron. F 992, p. 126 an o f pam pa se cyng hcefde mcest truw e to.
T h is s h iftin g o f w ord-o rd er is n o t restricted to relative clauses. T he question does not, in fact, belong to the S y n ta x o f the Pronoun, b u t to the S y n ta x o f the V erb. T he same sh ift takes place ju s t as w ell w ith an in fin itiv e : S. C hron. E 1140, p. 266 me lihtede candles to ceten h i 2 3. There is on the w hole a tendency to assume the n orm a l w o rd -o rd e r o f the p rin c ip a l clauses: S. C hron. E 1140, p. 267 pat he alle his castles sculde iiu e n up . . . Sume he ia f up . . . t i l h i aiauen up here castles.
T h is tendency, w hich w ill be studied in some subsequent chapter, is also seen in such subordinate clauses in w hich the p re p o sitio n is placed after the noun (‘sem i-com pounds’), as S. Chron. A 918, p. 98 se cyng hcefde fu n d e n poet him mon scet ivip on suphealfe Seefern m il pan, ib. 921, p. 102 op pam burgiuarmn com m ara fu ltu m to, H o ly R ood 5,34 poer wolde anbidi^en od dauid him come to (com pare S. C hron. 921, p. 101 op him m ara fu ltu m to com).
51
. E. A . K o c k , l. c. p. 67, and Anklamr, l. c. p. 57, give o n ly a few late O E . instances o f hwa swa, etc. fo r swa hwa, swd. I add from m y own collections: S. Chron. E 675, p. 37 1 hwa swa h it tobreceb pa ivurbe he amansumed, Lie be rm a nn in H e rrig ’s A rc h iv 111, p .2 8 3 , hua swa braves hisses, braue C rist hine pisses Hues hele (N o rth u m b ria n documents 1099— 1128), D ip lo m . A n g l. p. 369 ^ hwa se pas celmesse holdlice healde, healde hine God . . . 1 hwa swa hio awende sieo he aivcen- ded fra m Gode, ib. 567 and wo so pis guides bereuen w ylle bereue hym God heueriche, ib. 593 and wo so pis awende God awende his ansene fro m him.T h e second swa is ra re ly o m itte d (see E. A , K o c k , l. c. p. 66, p. 71).
H ere are some instances: A n g l. Sax. M in. {A n g lia X I, p. 119) geedniwa
% geniwa arfcestusta feeder swa hwwt on deede, swa hiveet on worde swa hwcet on poere spcece m id deoflicum facne ys gewemmed, rendering renova . . . quicquid actione quicquid verbo quicquid ipsa locutione diabolica fraude violatum est. S. Chron. F 994, p. 128 $ mceste y fe l worhton pe cefre ceni here mihte on eallon bingan swa hw ar h i ferdon, com pare ib.
1 seo rod is the re ading o f the MS. (Stowe M S . 944, B r it. Mus., f. 58), b u t i t is m is
p rin te d se orbd in de G ra y B irc h 's e d itio n. D r. W a rn e r has k in d ly in fo rm e d me th a t the passage in question is w ritte n in a late eleventh c e n tu ry hand.
3 Q uoted b y O nions, l. c. p. 105.
995, p. 130 het h i fa ra n loc w har l i i woldon, S. Chron. E, etc. 1009, p. 139 h i f er don loc hu h i woldon (C om its
7
m), etc.A n early instance o f hw ilc from S. Chron. E w ill be found in the fo llo w in g Section.
5 2
. A n k la m , l. c. pp. 59 f., r ig h tly observes th a t be in V ic . a. V . sometimes interchanges w ith se from O E. siva after a noun preceded b y hwilch, e. g. 113,18 hwilche dai be and 51,23 hwilche dai^e se. A n k la m supposes th a t be, w h ic h also occurs in par be and hw ar be, is due to a m istake on the p a rt o f the scribe. T he scribe kn o w in g th a t the O E . relative pronoun se had been replaced b y pe, w ro n g ly p u t pe fo r se — O E . sica. T o this I object th a t the scribe v e ry w e ll knew how to use swa in 71,34 hicilche kennes swa h it hie, and th a t the construction occurs m uch earlier: S. Chron. E 675, p. 36 [ I c w ille . . . ] % hw ilc abbot pe hep peer coren o f pe munecan f he beo gebletsad . . . I c w ille . . . j> while man swa 1 haued behaten to fa re n to Rome, and he ne muge h it forbian, ouber fo r untrum nisse . . . ouber fo r hwilces cinnes ober neod he ne muge peer cumon, beo he o f Englelande ouber o f hw ilc ober igland beo he, cume to f mynstre on Medeshamstede.W e set aside hwilces (cinnes, etc.) th a t belongs to the preceding Section, and compare o n ly hw ilc abbot pe and while man siva. The om ission o f the firs t correlative in swa hw ilc + noun siva im paired the feeling o f correlation. The subordinate clause w as then connected w ith the noun, and the in tro d u c to ry w ord swa replaced b y the relative pronoun pe a ccordingly.
5 3
. If, in the above connexion, who [which) is substituted fo r which + noun, that m ay s till be k e p t: La3_ B 5207 and peos fo u r hinges h ii leide to pan gründe bote wo pat par fleh (the A - te x t reads ivhulc riche mon per at-fleh), W o h . o. u. Lau. (O E . H orn. p. 275) penne ayaines kinde Gab hwa pat swuche kinsemon ne luueb (cf. some lines below penne hwa se pe ne luues, he is mon unwreastest). T h is e a rly use o f that after an indefinite relative m a y account fo r the later insertion o f that after any relative p ro no un : the hye God on whom that we bileve, Chauc. The developm ent was supported b y a sim ilar inse rtio n o f that afte r an in te rrog ative p ro n o u n : W o h . o. u. Lau. (O E. H orn. I, p. 283) A hu pat ha nu d riu e n irnene neiles purh pine fe ire hondes in to hard rode (here hu
1 while man swa renders quicumque in the spurious ch a rte r on w h ic h th is P eterboroug h in se rtio n is based (C art. Sax. No. 48).
pat in re a lity introduces an dependent question o r e xclam ation); H ave- lo k 2211 [Y e w iten w e l] how pat he . . . bitauhte hise children pre. F o r a sim ilar use o f at and relative particles in O N orse and later in Scandi
navian, see F a lk og T o rp , Dansk-norskens Syntax, pp. 232 f., p. 241, p. 247. H o w far Scand. usage and influence from F rench que are con
cerned in this developm ent, is scarcely possible to decide. T he insertion o f that afte r pronouns runs p arallel to the use o f that after conjunctions, see Series I, p. 45.
54
. V e r y e a rly ever was attached to who that, which that, etc. E ver o rig in a lly belonged to the verb o r the w ho le sentence, b u t was g ra d u a lly attracted b y the pronoun. In m any cases i t is im possible to te ll where its real place was m eant to be: S. C hron. E 1048, p. 174 het se cyniny bannan u t here, cegder ge be sudan Temese ge be nordan eall f cefre betst wees, M or. O. 130, L a m b . H . p. 167 a hwilke time se enre mon o f pinchp his mid-dede, O rm 18757 f o rr PaM wf re C rist se llf shop a ll iss it t god ivipp alle, ib. 5074 % a ll patt cefre iss god in n me . . . a ll b irrp pe lu fe n n i t t in n me, V ic . a. V . 67,5 luue dine nexte. al siva be seluen, hwat manne swo he ceure bie. Ne bie he nceure suM swide fo rfe it, ceure he is bin nexte, ib. 37,31 hwat swo ceure o f him betide, L am b. H orn. 47 hwa efre penne ilokie wel pene sunne dei . . . beo heo dal neominde o f heofene riches blisse, Sawles W a rde , O E H . I, p. 265 opened fo r hiver se eauer pe gast wide, U r. o f ure Lou., O E H . I, p. 185 hwa se euer haueb longe wone o f gastliche elne, ib. p. 187 hwa se euer wule habbe lot w ip p e . Gen. E x . 270 wisdom be made ilc ding o f nogt quuat-so-euere on heuone o r her is wrogt. I t is n o t at a ll necessary to recur to F rench q u i que onques (cf. G ru n d r. § 172 y).5 5
. F o r hwon re fe rrin g to a w hole clause occurs already e arly in the tw e lfth ce ntu ry (see examples A n k la m , l. c. p. 63). F o r why referring to a preceding idea, and corresponding to quam ob rem, qua re, m ay be seen in O rm ’s co nclud ing rem arks 12690 fo rrw h i wass patt ta tt Sannt Johan amang pe leode se^de off C rist ta tt he ne meow himm nohht.C om pare H ilfr ic ’s tran slatio n, Gram . 231,10, o f the interro ga tive p ro nouns in L a tin : quare : fo r hwi, quam ob rem : fo r hw i obbe fo r bi.
The reason {for) why, w hich E inenkel (G rundr. § 172 o) derives from French, corresponds to Germ, die Ursache warum, D anish grunden h vorfor, aarsagen hvi and m ay have developed th ro ug h a blending o f 1 ig io . No. 3. C R IT IC A L C O N T R IB U T IO N S T O E A R L Y E N G L IS H S Y N T A X . 15
1 F a lk og T o rp , Dansk-norskens Syntax, p. 233.
I tell you why . . . and
1
tell you the reason f o r which . . . . , compare O rm 2421 n u wile1
shcewenn $uw fo r r w h i ÿ io 3a ff sivillc sivare onn^æness, and ib, 9729 3
^ wass operrwhatt fo rrw h i ^ 3 3 wœrcnn D r ih h tin lape.56
. I jo in to the documents an early, th ou gh n o t v e ry conclusive, instance o f whose as a relative p ro n o u n : Eadvv. Cant. H y m n s 15,40 [C ris t sye peolade. . . ] to whces tocuman alle menu sculen arisan == ad cujus adventum omnes homines resurgere habent. (de hw ilc in the next verse has no importance). Some o ther e arly form s o f the in te rro g a tive pronoun used as a relative occur in W in t. Vers. Reg. Ben. e. g. 19,27 saule to gyernenne, fo r hwarn hu sceal 3ewistale upa^ifen, ib, 35,30 urne d rihte be hivam pe apostel pus cived (A -S . Vers, be h im, In te rl. Vers.be dam), ib. 47,24 on pan 3ebedehuse purh hwylcere 3yemelyste h it 3elamp (per cuius, A -S . V erb pe hit, In te rl. Vers, pees . . . pe hit), ib. 27,4 pa stowwenn Invar ive ealle pos ping ma^e sycerlice wyreasn.
A n k la m , l. c. pp. 61 f., cites some interestin g b u t n o t quite certain exam ples w hich seem to indicate th a t the in te rro g a tiv e pronoun began to be used as a relative a lrea d y about the m iddle o f the eleventh century. In W u lfs ta n 129,9 ac bid œt gode anum gelang eal hwcet we gefaran scylon the m eaning o f hwcet is rath er ‘h ow ’ o r perhaps ‘where’, compare S. Chron. E 1052, p. 177 pa eorlas ne m ihton gewitan hwet Godwine eorl gefaren hcefde. W u lfs ta n ’s hwcet does n o t refer to eal as its antecedent, b u t introduces a noun clause in a pp osition to the subject o f the preceding sentence. Hivcet is a c c o rd in g ly an in te rro g a tiv e p ro noun used as a general o r in d efin ite re la tive ; and eal, to w hich the hwcet-clause stands in an appositive relation, m ay be o m itted. T h e use o f hivcet before gefaran is som ewhat lik e th a t o f F rench ce que in je ne sais pas ce qu’i l est devenu.
57
. A . T o b le r treats in Vermischte Beiträge I, p. 203 f. o f French phrases form ed b y a noun and a relative clause o n ly : Massacre que nous avons oublié! Such e x c la m a to ry sentences belong to v iv id fa m ilia r speech, and m ay, I th in k , be form ed spontaneously in a ny language, th ou gh instances are rare in p rin t. A n exam ple like O n Oreis. o. o. Lou., (O E. H orn. p. 187) m in heoueneliche leche pet makedest us o f p i seolf se m ih ti medicine, iblesced beo pu euer shows an interm ediate stage, a p rin c ip a l clause being tagged on. B u t no verb appears in W o h . o. u. Lau.(O E. H orn. p. 283) A pat luuelike bodi pat henges siva re w li swa blodi and siva kalde.
ig io . No. 3. C R IT IC A L C O N T R IB U T IO N S T O E A R L Y E N G L IS H S Y N T A X .
17
V II. W h a t.
58.
In his dissertation on tem poral clauses, W . Böhm e explains nl what, m eaning ‘u n til’, as a further developm ent o f the O E. conju n c tio n pad 1.
Böhme firs t observes th a t the use o f pad in the tem poral sense
‘u n til’ began in consecutive clauses, as seen in the substitution o f pat fo r op in swä lange pcet: S. Chron. E 1052, p. 177 swa lange j> seo scip- fy r d eall belaf. The ultim ate stage u n til appears early, e.g. S. Chron.
E 1076. p. 213, pa B ry tta s hiñe heoldon fi se cyng com o f F ra n c la n d2.
So far there are no difficulties. T he S. C hron. and Charters offer abundant proofs th a t the sense u n til was fir m ly established in late O E . The same change from a consecutive to a te m p ora l clause is seen in French tant que : je vais tra in e r une mourante vie ta n t que p a r ta po u r- suite elle me soit ravie, Corn. (D id , Oén.), and in early D anish: rob nw th ijn kettere saa lenge tw re ffn e r ‘c ry n o w you heretic so lo n g [th a t i, e. t ill] yo u burst’ (F a lk og T o rp , Dansk-norskens Syntax, p. 229).
T he question is next h o w to com bine pcet w ith al what. Böhme calls a tte n tio n to S. Chron. F 1013, p. 144 (T horpe p. 271) and se cing sona him sylf ferde a fte r ivces par begeondan eal d Sivegen iveard dead, and continues: «eal pad ve rh ä lt sich zu pcet, w ie eal swa zu swa\ man vg l. übrigens dazu eine andere Stelle aus S. C. (F s. 143) % p(er wunode eal to his Ufes ende. Das In te rro g a tiv an Stelle des D em onstrativs hat nichts A u ffä llig e s an sich«. Böhme states th a t al what in the tra n s itio n period occurs o n ly in three passages in V ic . and V . and in one passage in K e nt. Serm. 27,8 al-wat h i kam over po huse, and concludes th a t al what is peculiar to the South-E ast o f E ngland. T he conjunction is s till used in A 3enbite (al-huet and al-huet pet).
These statements are n o t quite exact. In his excellent study Böhme has overlooked some im p o rta n t passages th a t p o in t in another direction.
A l has o n ly an intensive force, as shown b y C ott. Vesp. H orn. p. 235 per efter arerde god, pas la ye .. . and si 3eleste sume wile . . . swa lange (f) pat si alswa sivih abroad and adili^ede purh unhersamnesse wat h it com to pa, time pe god sende pe halie w iti^e. T h is passage is also quoted b y E inenkel, w h o supposes an im ita tio n o f F rench que (A nglia X X V I I p. 142).
1
D ie Temporalsätze in der Übergangszeit vom Angelsächsischen zum Altenglischen D iss. L e ip z ig 1903, pp. 51 ff.2 Cf. a lre a d y O ros. 161,31 hie peak swa ondrcedendlice gebidon pcet se ege of ergangen wees, W ü lfin g , S ynta x II, p. 120. Som etim es a fin a l shade blends w ith the tem poral e. g. P er. D id a x. 21,24 wlece h y t eall togadere pcet h y t wlcec beo.
V id .-S e lsk. S k rifte r. I I . H .-F . K l. 1910. No. 3. 2
Secondly, what is also used in connexion w ith an a dverb: K ent. Serm.
(Misc. p. 30) pu hest i-hialde pet betste w yn wath nu, corresponding to John II. 10 tu autem servasti bonurn vinum usque adhuc 1. T h ird ly , what occurs in the sense ‘as lo n g as’ : L am b. H orn. p. 11 Muchel is us penne neod leoue bredren tuet tue on pisse middelerd Ihnen sod s c rift
‘great need have w e then, dear brethren, as lo n g as we live on this earth, o f true s h rift’.
T he last passage o f Section
56
showed th a t hwcet in late O E. had acquired a vaguer sense. H o w far the interro ga tive-rela tive pronoun hwwt m ay have developed on its own accord in a tem poral d irectio n I do not dare to say. T he h is to ry o f what and that is s till obscure in m any respects. T he p rim a ry sense was perhaps ‘the tim e t h a t , ‘w hile , w h ic h before an ingressive verb was turned in to ‘u n til : wet tue liu ie n , tuut h it com. B u t as wath n u appears n early at the same tim e, there is reason to believe th a t what has som ehow o r othe r been exposed to the in flu ence o f French que, the m ore so as this p a rtic u la r use o f what seems to be restricted to the southern dialects.5 9
. In the T rin . H orn, (al) so tuhat so occurs several tim es in the sense ‘as soon as’ , e. g. 125 alse w at se pat holi meide m id worde grette pe holie spuse, po ward sod pat pe engel hadde er bi pis child seid (see Böhme l. c. p. 38). In L am b. H orn, o n ly one instance has been found. Böhme is perhaps rig h t in e x p la in in g so w hat so as form ed b y the adjective liwcet ‘q u ic k ’ on the analogy o f swa hrade swa, though there m ay also have been some confusion o r vague connexion w ith the pronoun ivhat.6 0
. Ney wat in R ob. G louc. m ay be due to French presque (see E inenkel in A n g lia X X V I I , p. 142), b u t there is an old er expression w hich is n o t so easily accounted fo r: L a m b . H orn. 137 and pesne leide) habbed niest hwet alle men, rendering et hoc a n im i morbo laborat fere omnis homo; s im ila rly in the corresponding passage T rin . H orn. 157 and bis custume haued mestiuat alle men 2.61
. M ätzner (Gram. I l l , p. 377) compares the correlative adverbial iv h a t. . . what w ith F rench que .. . que, b u t m entions at the same tim e1 In the W e s t-S a x o n V e rsio n du geheolde j> gode w in od pis, in the N o rth u m b ria n Glosses . . . uid to dises vel uid nu vel uid dageana, O rm . 14066 . . . t ill nu.
2 I take hwet to be the O E . e xcla m a to ry hwcet in C o tt. V esp. H orn. p. 231 pa sende se king his mrndraches . . . t o zeladie pis foie hwet bute icome sum cofer sum letter sum frencl sum fend.
the a n a lo g y o f sum . . . sum. Sweet N . E . G r . 2122 also explains what as the O E . indefinite pronoun. E inenkel [A n g lia X X V I I , p. 143) has found an e arly instance in C ott. Vesp. H orn. 237 wat fre n d w at fa, to w h ich I add L am b. H orn. 145 alle we beod in m onifald ivaive inne pisse wreche Hue hivat fo r ure eldere Werkes, hwat fo r ure a^ene gultes; the corre
sponding passage in T rin . H orn. 203 has o n ly fo r are eldrene giltes and ec f o r ure agene sinnen. A correlated (some-)what.. . (some-)what could easily take the adverbial m eaning p a rtim . . . p a rtim , whether in fluenced o r not b y French que . .. que.
1910. N o . 3 . C R IT IC A L C O N T R IB U T IO N S T O E A R L Y E N G L IS H S Y N T A X . 19
P rin te d Aug. 9, 1910.
