«w
BEOWULF
CAMBRIDGE UNIVERSITY PRESS
C. F. CLAY, Manager
LONDON : FETTER LANE, E.C.4
NEW YORK : THE MACMILLAN CO.
BOMBAY \
CALCUTTA } MACMILLAN AND CO., Ltd.
MADRAS J
TORONTO: THE MACMILLAN CO. OF
CANADA, Ltd. TOKYO : MARUZEN-KABUSHIKI-KAISflA
ALL BIGHTS KESERVED
PLATE I
BEOWULF
AN INTRODUCTION^ TO THE STUDY OF
THE POEM WITH A DISCUSSION OF
THE STORIES OF OFFA AND FINN
BY
K. W. CHAMBEES
Dey mout er bin two deloojes: en den agin dey moutent.
Uncle Remus, The Story of the Deluge.
CAMBRIDGE
AT THE UNIVERSITY PRESS
1921
^
lil^
9-77-
TO PROF. WILLIAM WITHERLE LAWRENCE
Dear Prof. Lawrence,
WheD, more than four years ago, I asked you to allow me to dedicate this volume to you, it was as a purely personal token of gratitude for the help I had received from what you have printed, and from what you have written to me privately.
Since then much has happened: the debt is greater, and no longer purely personal. We in this country can never forget what we owe to your people. And the self-denial which led them voluntarily to stint themselves of food, that we in Europe might be fed, is one of many things about which it is not easy to speak. Our heart must indeed have been hardened if we had not considered the miracle of those loaves. But I fear that to refer to that great debt in the dedication to this little book may draw on me the ridicule incurred by the poor man who dedicated his book to the Universe.
Nevertheless, as a fellow of that College which has just received from an American donor the greatest benefaction for medical research which has ever been made in this country of ours, I may rejoice that the co-operation between our -nations is being continued in that warfare against ignorance and disease which sojne day will become the only warfare waged among men. -£jt U * ^ - Sceal hring-naca ofer heafu bringan lac ond luf-t9,cen. Ic ])a leode wat ge wis feond ge wiS freond fseste geworhte, seghwses untsele ealde wisan.
R. W. C,
PREFACE
I HAVE to ttank various colleagues who have read proofs of this book, in whole or in part: first and foremost my old teacher, W. P. Ker; also Robert Priebsch, J. H. G. Grattan, Ernest Classen and two old students. Miss E. V. Hitchcock and Mrs Blackman. I have also to thank Prof. W. W. Lawrence of Columbia; and though there are details where we do not agree, I think there is no difference upon any important issues. If in these details I am in the right, this is largely due to the helpful criticism of Prof. Lawrence, which has often led me to reconsider my conclusions, and to re-state them more cautiously, and, I hope, more correctly. If, on the other hand, I am in the wrong, then it is thanks to Prof. Lawrence that I am not still more in the wrong.
From Axel Olrik, though my debt to him is heavy, I find myself differing on several questions. I had hoped that what I had to urge on some of these might have convinced him, or, better still, might have drawn from him a reply which would have convinced me. But the death of that great scholar has put an end to many hopes, and deprived many of us of a warm personal friend. It would be impossible to modify now these passages expressing dissent, for the early pages of this book were printed off some years ago. I can only repeat that it is just because of my intense respect for the work of Dr Olrik that, where I cannot agree with his conclusions, I feel bound to go into the matter at length. Names Hke those of Olrik, Bradley, Chadwick and Sievers carry rightly such authority as to make it the duty of those who differ, if only on minor details, to justify that difference if they can.
From Dr Bradley especially I have had help in discussing various of these problems : also from Mr Wharton of the British Museum, Prof. ColUn of Christiania, Mr Ritchie Girvan of Glasgow, and Mr Teddy. To Prof. Brogger, the Norwegian state-antiquary, I am indebted for permission to reproduce photographs of the
viii Preface
Viking ships : to Prof. Finnur Jonsson for permission to quote from his most useful edition of the Hrolfs Saga and the BjarJca Rimur, and, above all, to Mr Sigf lis Blondal, of the Koyal Library of Copenhagen, for his labour in collating with the manuscript the passages quoted from the Grettis Saga.
Finally, I have to thank the Syndics of the University Press for undertaking the pubKcation of the book, and the staff for the efficient way in which they have carried out the work, in spite of the long interruption caused by the war.
K. W. C.
April 6, 1921.
CONTENTS
PAOS
GENEALOGICAL TABLES xii
PART I
CHAPTER I. THE HISTORICAL ELEMENTS
Section I. The Problem 1
Section II. The Geatas — their Kings and their Wars . . 2
Section III. Heorot and the Danish Kings .... 13
Section IV. Leire and Heorot 16
Section V. The Heathobeardan 20
Section VT. HrothuK 26
Section Vn. King Offa 31
CHAPTER II. THE NON-HISTORICAL ELEMENTS
Section I. The Grendel Fight 41
Section II. The Scandinavian Parallels — ^Grettir and Orm . . 48
Section III. Bothvar Bjarki 54
Section IV. Parallels from Folklore 62
Section V. Scef and Scyld .68
Section VI. Beow 87
Section VII. The house of Scyld and Danish parallels — Heremod-
Lotherus and Beowulf-Frotho .... 89
^- CHAPTER III. THEORIES AS TO THE ORIGIN, DATE AND STRUCTURE OF THE POEM
r^ Section I. Is Beowulf translated from a Scandinavian
^ ., / original? 98
(jv^ Section II. The dialect, syntax and metre of Beowulf as
^ evidence of its literary history .... 104
Section IIL Theories as to the structure of Beowulf . , . 112 Section IV. Are the Christian elements incompatible with the rest
of the poem? ....... 121
Contents
PART II
DOCUMENTS ILLUSTKATING THE STOEIES IN BEOWULF, AND THE OFFA-BAGA
PAGE
A. The early Kings of the Danes, according to Saxo Grammaticus: Dan, Humblus, Lotherus and Scioldus; Frotho's dragon fight; Haldanus, Roe and Helgo; Roluo (Rolf Kraki) and Biarco (Bjarki); the death of Rolf 129
B. Extract from Hrdlfs Saga Kraka, with translation (cap. 23) . 138
C. Extracts from Grettis Saga, with translation: (a) Glam episode (caps. 32-35); (6) Sandhaugar episode (caps. 64-66) . .146
D. Extracts from Bjarka Rimur, with translation . . .182
E. Extract from pdttr Orms Stordlfssonar, with translation . . 186 E. A Danish Dragon-slaying of the Beowulf- type, with translation 192 G. The Old EngUsh Genealogies. I. The Mercian Genealogy. II. The
stages above Woden: Woden to Geat and Woden to Sceaf . 195
H. Extract from the Chronicle Roll 201
I. Extract from the Little Chronicle of the Kings of Leire . . 204 K. The Story of Offa in Saxo Grammaticus .... 206
L. Erom Skiold to Offa in Sweyn Aageson . . . .211
M. Note on the Danish Chronicles . . . . . .215
N. The Life of Offa 7, with extracts from the Life of Offa 11. Edited
from two Mss in the Cottonian Collection . . . .217 O. Extract from Widsith, 11. 18, 24-49 243
|
Section |
I. |
|
Section |
II. |
|
Section |
III. |
|
Section |
IV. |
|
Section |
V. |
|
Section |
VI. |
|
Section |
VII. |
|
Section VIH. |
|
|
Section |
IX. |
PART III
*^THE FIGHT AT FINNSBUKG
The Finnshurg Fragment ..... 245
The Episode in Beowulf 248
MoUer's Theory 254
Bugge's Theory 257
Some Difficulties in Bugge's Theory . . . 260
Recent Elucidations. Prof. Ayres' Comments . 266
Problems stiU outstanding . . : . . 268 TheWeightof Proof: the Eotens. . . .272
Ethics of the Blood Feud 276
/
Contents xi
PAGE
Section X. An Attempt at Reconstruction .... 283
Section XI. Gefwulf, Prince of the Jutes. .... 286
Section XII. Conclusion ....... 287
Note. Frisia in the heroic age .... 288
PART IV
APPENDIX
A. A Postscript on Mythology in Beowvlf. (1) Beowulf the Scylding
and Beowulf son of Ecgtheow. (2) Beow . . . .291
B. Grendel . . 304
C. The Stages above Woden in the West-Saxon Genealogy . .311 / D. Grammatical and literary evidence for the 4ate of Beowvlf. The
^ relation of Beowulf to the Classical Epic ^-""^ . . . 322
E. The "Jute- question" reopened 333
F. Beowulf and the Archaeologists . . . . . . 345 (
G. Leire before Rolf Kraki 365
H. Bee-wolf and Bear's son 365
I. The date of the death of Hygelac ...... 381
BIBLIOGEAPHY OF BEOWULF AND FINNSBURG 383
INDEX 414
PLATES
PLATE
L Drida (Thryth) reproached for her Evil Deeds frontispiece
II. Leire in the Seventeenth Century . . .to face 16
III. Ofifa, miraculously restored, vindicates his Right.
At the side, Offa is represented in Prayer . „ „ 34
IV. Drida (Thryth) arrives in the land of King Offa,
"in nauicula armamentis carente" . . • „ „ 36
V. Riganus (or Aliel) comes before King Warmundus to claim that he should be made King in place of the incompetent Offa . . . . . „ „ 218
VI. Drida (Thryth) entraps Albertus (^Ethelberht) of
East Anglia, and causes him to be slain . . „ „ 242
VII. The Gokstad Ship. The Oseberg Ship . . „ „ 362
Vni. Southern Scandinavia in the Sixth Century.
English Boar-Helmet and Ring-Swords . . » „ At end
i
Xll
GENEALOGICAL TABLES
The names of the corresponding characters in Scandinavian legend are added in italics; first the Icelandic forms, then the Latinized names as recorded by Saxo Grammaticus.
(1) THE DANISH ROYAL FAMILY
Scyld Scefing [SkjoldVy Skyoldi^s]
Beowulf [not the hero of the poem] Healfdene [Halfdan, HaJdanus]
s;
Heorogar [no Scandinavian parallel]
Heoroweard HretSric [Hjgrvard'r, Hiar- [Hraerekr, warns: bvt not Bjiricus : not recognized as he- recognized longing to this as a son of family] Hroarr]
HroSgar {Hroarr^, Roe]y mar. Wealh}?eow
Halga [Hdgi, Hdgo]
a daughter iSigny]
HrotSmund
Freawaru mar. Ingeld
HroSulf [Hrdlfr Kraki, Roliko]
(2) THE GEAT ROYAL FAMILY HretSel W^gmund
Herebeald HseScyn Hygelac, mar. Hygd a daughter, mar. Ecgjjcow Weohstan
Beowulf Wiglaf
a daughter, mxir. Eofor
Heardred
(3) THE SWEDISH ROYAL FAMILY Ongenheow
Onela
[An, not recognized
as belonging to this
family]
Ohthere [6ttarr]
Eanmund
Eadgils [A&ils^, Athislus]
1 The exact equivalent to Hroffgar is found in O.N., in theioTmHroffgeirr. The by-form Hroarr, which is used of the famous Danish king, is due to a number of rather irregular changes, which can however be paralleled. The Primitive Germanic form of the name would have been *Hrdpugaisaz : for the loss of the g at the beginning of the second element we may compare A&ils with Eadgils (Noreen, Altisldndische Grammatik, 1903, § 223); for the loss of ff before w com- pare Hrdlfr with Hroffwulf (Noreen, § 222); for the absence of R- umlaut in the second syllable, combined with loss of the g, compare O.N, nafarr with O.E. nafugdr (Noreen, § 69).
2 Corresponding to O.N. A&ils we should expect O.E. Mffgils, jE&gisl. The form Eadgils may be due to confusion with the famous Eadgils, king of the Myrgingas, who is mentioned in Widsith. The name comes only once in Beowulf (1. 2392) and may owe its form there to a corruption of the scribe. That the O.E. form is corrupt seems more likely than that the O.N. Ad'ils, so well known and so frequently recorded, is a corruption of AuSgisl.
PART I
CHAPTER I
THE HISTORICAL ELEMENTS
Section I. The Problem.
The unique ms of Beowulf may be, and if possible should be, seen by the student in the British Museum. It is a good specimen of the elegant script of Anglo-Saxon times : " a book got up with some care," as if intended for the library of a nobleman or of a monastery. Yet this ms is removed from the date when the poem was composed and from the events which it narrates (so far as these events are historic at all) by periods of time approximately equal to those which separate us frojii the time when Shakespeare's Henry V was written, and when the battle of Agincourt was fought.
To try to penetrate the darkness of the five centuries which lie behind the extant ms by fitting together such fragments of illustrative information as can be obtained, and by using the imagination to bridge the gaps, has been the business of three generations of scholars distributed among the ten nations of Germanic speech. A whole hbrary has been written around our poem, and the result is that this book cannot be as simple as either writer or reader might have wished.
The story which the MS tells us may be summarized thus: Beowulf, a prince of the Geatas, voyages to Heorot, the hall of Hrothgar, king of the Danes; there he destroys a monster Grendel, who for twelve years has haunted the hall by night and slain all he found therein. When Grendel's mother in revenge makes an attack on the hall, Beowulf seeks her out and kills her also in her home beneath the waters. He then
C. B. 1
2 Tue Problem [CH. i
returns to his land with honour and is rewarded by his king Hygelac. Ultimately he himself becomes king of the Geatas, and fifty years later slays a dragon and is slain by it. The poem closes with an account of the funeral rites.
Fantastic as these stories are, they are depicted against a background of what appears to be fact. Incidentally, and in a number of digressions, we receive much information about the Geatas, Swedes and Danes : all which information has an appearance of historic accuracy, and in some cases can be proved, from external evidence, to be historically accurate.
Section II. The Geatas — their Kings and their Wars.
Beowulf's people have been, identified with many tribes : but there is strong evidence that the Geatas are the Gotar (O.N. Gautar), the inhabitants of what is now a portion of Southern Sweden, immediately to the south of the great lakes Wener and Wetter. The names Geatas and Gautar correspond exactly^, according to the rules of O.E. and O.N. phonetic development, and all we can ascertain of the Geatas and of the Gautar harmonizes well with the identification^.
We know of one occasion only when the Geatas came into violent contact with the world outside Scandinavia. Putting together the accounts which we receive from Gregory of Tours and from two other (anonymous) writers, we learn that a piratical raid was made upon the country of the Atuarii (the O.E. netware) who dwelt between the lower Rhine and what is now the Zuyder Zee, by a king whose name is spelt in a variety of ways, all of which readily admit of identification with that of the Hygelac of our poem^. From the land of the Atuarii this king carried much spoil to his ships; but, remaining on shore, he was overwhelmed and slain by the army which the
^ It must be remembered that the sound changes of the Germanic dialects have been worked out so minutely that it is nearly always possible to decide quite definitely whether two names do or do not exactly correspond. Only occasionally is dispute possible [e.g. whether Hrothgar is or is not phonetically the exact equivalent of Hroarr].
2 See below, pp. 8-10.
* Chochilaicus, which appears to be the correct form, corresponds to Hygelac ' (in the primitive form Hugilaikaz) as Chlodovechus to Hludovicus.
SECT.ii] Tlie Geatas — their Kings and their Wars 3
Frankish king Theodoric had sent under liis son to the rescue of these outlying provinces; the plunderers' fleet was routed and the booty restored to the country. The bones of this gigantic king of the "Getae" [presumably = Geatas] were long preserved, it was said, on an island near the mouth of the Rhine;
Such is the story of the raid, so far as we can reconstruct it from monkish Latin sources. The precise date is not given, but it must have been between a.d. 512 and 520.
Now this disastrous raid of Hygelac is referred to constantly in Beowulf: and the mention there of Hetware, Franks and the Merovingian king as the foes confirms an identification which would be satisfactory even without these additional data^.
Our authorities are:
(1) Gregory of Tours (d. §94):
His ita gestis, Dani cum rege suo nomine Chlochilaico evectu nuvale per mare Gallias appetuni. Egressique ad' terras, pagum unum de regno Theudorici devastant atque captivant, oneratisque navibiis tam de captivis quam de reliquis spoliis, reverti ad patriam cupiunt ; sed rex eorum in litus resedebat donee naves alto mare conpraehenderent, ipse deinceps secuturus. Quod cum Theudorico nuntiatum fuisset, quod scilicet regio ejus fuerit ab extraneis devastata, Theudobertum, filium suum, in illis partibus cum valido exercitu et magno armorum apparatu direxit. Qui, interfecto rege, hostibu^ navali proelio superatis opprimit, omnemque rapinam terrae restituit.
The name of the vanquished king is spelt in a variety of ways: Chlochilaichum, Chrochilaicho, Chlodilaichum, Hrodolaicum.
See Gfegorii episcopi Turonensis Historia Francorum, p. 110, in Monumenta Oermaniae Historica {Scriptores rerum merovingicarum, I).
(2) The Liber Historiae Francorum (commonly called the Gesta Francorum) :
In illo tempore Dani cum rege suo nomine Chochilaico cum navale hoste per alto mare Gallias appetent, Theuderico paygo [i.e. pagum} Atioarios vel alios devastantes atque captivantes plenas naves de captivis alto mare intr antes rex eorum ad litus maris resedens. Quod cum Theuderico nuntiatum fuisset, Theudobertum filium suum cum magno exercitu in illis partibus dirigens. Qui consequens eos, pugnavit cum eis caede magna atque prostravit, regem eorum interficit, preda tullit, et in terra sua restituit.
The Liber Historiae Francorum was written in 727, but although so much later than Gregory, it preserves features which are wanting ^ in the earlier historian, such as the mention of the Hetware (Attoarii). ■ Note too that the name of the invading king is given- in a form which
1 The passaj^es in JSeot^wZ/ referring to this expedition are:
1202 etc\ Frisians (adjoining the Hetware) and Franks mentioned as
the oes. 2354 etc. Hetware mentioned.
2501 etc. Hugas (= Franks) and the Frisian king mentioned. 2914 etc. Franks, Frisians, Hugas, Hetware and "the Merovingian"
mentioned.
•^v 1—2
4 The Geatas— [ch. i
approximates more closely to Hygelac than that of any of the mss of Gregory: variants are Chrochilaico, Chohilaico, Chochilago, etc.
See Monumenta Germaniae Historica {Scriptores rerum merovingi- caruniy II, 274).
(3) An anonymous work On monsters and strange beasts, appended to two MSS of Phaedrus.
Et sunt {monstral mirae magnitudinis : ut rex Huiglaucus qui imperavit Oetis et a Francis occisus est. Quern equus a duodecimo anno portare non potuit. Cujus ossa in Reni fiuminis insula, uhi in Oceanum prorumpit, reservata sunt et de longinquo venientibus pro miraculo ostenduntur.
This treatise was first printed (from a MS of the tenth century, in private possession) by J. Berger de Xivrey [Traditions teratologiques, Paris, 1836, p. 12). It was again published from a second MS at Wolfenbiittel by Haupt (see his Opuscula ii, 223, 1876). This MS is in some respects less accurate, reading Huncglacus for Huiglaucus,. and gentes for Oetis. The treatise is assigned by Berger de Xivrey to the sixth century, on grounds which are hardly conclusive (p. xxxiv). Haupt would date it not later than the eighth century (n, 220).
The importance of this reference lies in its describing Hygelac as king of the Getae, and in its fixing the spot where his bones were preserved as near the mouth of the Rhine ^.
(^ But if Beowulf is supported in this matter by what is almost contemporary evidence (for Gregory of Tours was born only some twenty years after the raid he narrates) we shall probably be right in arguing that the other stories from the history of the Geatas, their Danish friends, and their Swedish foes, told with what seems to be such historic sincerity in the different digressions of our poem, are equally based on fact. True, we have no evidence outside Beowulf for Hygelac's father, king Hrethel, nor for Hygelac's elder brothers, Herebeald and Haethcyn; and very Kttle for Hsethcyn's deadly, foe, the Swedish king Ongentheow^.
And in the last case, at any rate, such evidence might
^ The identification of Chochilaicus with Hygelac is the most important discovery ever made in the study of Beoumlf, and the foimdation of our belief in the historic character of its episodes. It is sometimes attributed to Grundt- vig, sometimes to Outzen. It was first vaguely suggested by Grundtvig (Nyeste Skilderie af Kjtj^benhavn, 1815, col. 1030) : the importance of the identification was worked out by him fully, two years later {Danne-Virke, n, 285). In the meantime the passage from Gregory had been quoted by Outzen in his review of Thorkelin's Beovmlf (Kieler Blatter, m, 312). Outzen's reference was ob- viously made independently, but he failed to detect the real bearing of thv> passage upon Beoumlf. Credit for the find accordingly belongs solely to Grundtvig.
2 Ongentheow is mentioned in Widsith (1. 31) as a famous king of the Swedes. Many of the kings mentioned in the same list can be proved to be historical, and the reference in Widsith therefore supports Ongentheow's |iistoric character, but is far, in itself, from proving it. I
SECT, ii] their Kings and their Wars 6
fairly have been expected. For there are extant a very early Norse poem, the Ynglinga tal, and a much later prose account, the Ynglinga saga, enumerating the kings of Sweden The Ynglinga tal traces back these kings of Sweden for some thirty reigns. Therefore, though it was not composed till some four centuries after the date to which we must assign Ongentheow, it should deal with events even earlier than the reign of that king: for, unless the rate of mortality among early Swedish kings was abnormally high, thirty reigns should occupy a period of more than 400 years. Nothing is, however, told us in the Ynglinga tal concerning the deeds of any king Angantyr — ^^ which is the name we might expect to correspond to Ongen- theow^.
But on the other hand, the son and grandson of Ongentheow, as recorded in Beowulf, do meet us both in the Ynglinga tal and in the Ynglinga saga.
According to Beowulf, Ongentheow had two sons, Onela and . . Ohthere: Onela became king of Sweden and is spoken of in terms of highest praise^. Yet to judge from the account given in Beowulf, the Geatas had little reason to love him. He had followed up the defeat of Hygelac by deaUng their nation a \ second deadly blow. For Onela's nephews, Eadgils and Ean- mund (the sons of Ohthere), had rebelled against him, and had / taken refuge at the court of the Geatas, where Heardred, son of V Hygelac, was now reigning, supported by Beowulf. Thither Onela pursued them, and slew the young king Heardred. ' Eanmund also was slain ^, then or later, but Eadgils escaped. '/
It is not clear from the poem what part Beowulf is supposed to have taken in this struggle, or why he failed to ward off disaster from his lord and his country. It is not even made clear whether or no he had to make formal submission to the hated Swede: but we are told that when Onela withdrew he succeeded to the vacant throne. In later days he took his revenge upon Onela. "He became a friend to Eadgils invhis distress; he supported the son of Ohthere across the broad water with men, with warriors and arms: he wreaked his
^ Strictly Anganpdr. See Heusler, Heldennamen in mehrfacher LautgeataU, Z.f.d.A. Lii, 101.
2 U. 2382-4. » U. 2612-9.
6 The Geatas— [CH. i
vengeance in a chill journey fraught with woe : he deprived the king [Onela] of his life."
This story bears in its general outline every impression of true history : the struggle for the throne between the nephew and the uncle, the support given to the unsuccessful candidate by a rival state, these are events which recur frequently in the wild history of the Germanic tribes during the dark ages, following inevitably from the looseness of the law of succession to the throne.
Now the Ynglinga tal contains allusions to these events, and the Ynglinga saga a brief account of them, though dim and distorted^. We are told how Athils (= Eadgils) king of Sweden, son of Ottar (= Ohthere), made war upon Ali (= Onela). By the time the Ynglinga tal was written it had been forgotten that Ali was Athils' uncle, and that the war was a civil war. But the issue, as reported in the Ynglinga tal and Ynglinga saga, is the same as in Beowulf:
"King Athils had great quarrels with the king called Ali of Upp- A land; he was from Norway. They had a battle on the ice of Lake
^ (/ Wener; there King Ali fell, and Athils had the victory. Concerning y this battle there is much said in the SJcjoldunga saga.''
From the Ynglinga saga we learn more concerning King Athils : not always to his credit. He was, as the Swedes had been from of old, a great horse-breeder. Authorities differed as to whether horses or drink were the death of him 2. Ac- cording to one account he brought on his end by celebrating, with immoderate drinking, the death of his enemy Rolf (the Hrothulf of Beowulf). According to another:
"King Athils was at a sacrifice of the goddesses, and rode his horse through the hall of the goddesses: the horse tripped under him and fell and threw the king ; and his head smote a stone so that the skull broke and the brains lay on the stones, and that was his death. He died at Uppsala, and there was laid in mound, and the Swedes called him a mighty king."
1 Whether it be accuracy or accident, these names Ottar and Athils come just at that place in the list of the Ynglinga tal which, when we reckon back the generations, we find to correspond to the beginning of the sixth century. And this is the date when we know from Beoumlf that they should have been reigning.
2 But the accounts are quite inconsistent. Saxo (ed. Holder, pp. 66-7) implies a version in which Athils was deposed, if not slain, by Bothvar Bjarki, which is quite at variance with other information given by Saxo.
SECT, ii] their Kings and their Wars 7
There can, then, hardly be a doubt that there actually was such a king as Eadgils : and some of the charred bones which still lie within the gigantic "King's mounds" at Old Uppsala may well be his^. And, though they are not quite so well ^ authenticated, there can also be little doubt as to the historici t existence of Onela, Ohthere, and even of Ongentheow.
The. Swedish Kings.
The account in the Ynglinga saga of the fight between Onela and Eadgils is as follows:
AiHls konungr dtti deilur miklar viS^ konung JjanUy tr Ah hit inn uvplenzki : hann var 6r Noregi. peir dttu orrostu a Vaenis isi ; par fell An konungr en Acfils hafSi sigr ; fro, pessarri orrostu er langt sagt i Skjgldunga sqqu. {Ynglinga saga in Heimskringla, ed. J6nsson, Kj0benhavn, 1893, i, 56.)
The Skjoldunga saga here mentioned is an account of the kings of Denmark. It is preserved only in a Latin abstract.
Post haec ortis inter Adilsum ilium Sveciae regent et Alonem Op- plandorum regem in Norvegia, inimicitiis, praelium utrinque indicitur: loco pugnae statuto in stagno Waener, glade jam obducto. Ad illud igitur se virihus inferiorem agnoscens Eolphonis privigyii sui opem implorat, hoc proposito praemio, ut ipse Rolpho tres praeciosissimas res quascunque optaret ex universo regno Sveciae praemii loco auferret : duodecim autem pugilum ipsius quilibet 3 libras auri puri, quilibet reliquorum bellatorum tres marcas argenti defecati. Rolpho domi ipse reses pugilos suos duodecim Adilso in subsidium mittit, quorum etiam opera is alioqui vincendus, victoriam obtinuit. Illi sibi et regi proposiium praemium exposcunt, negat Adilsus, Rolphoni absenti ullum deberi praemium, quare et Dani pugiles sibi oblatum respuebant, cum regem suum eo frustrari intelligerent, reversique rem^ ut gesta est, exponunt. (See Skjoldungasaga i Arngrim Jonssons Udtog, udgiven af Axel Olrik, Kj0benbavn, 1894, p. 34 [116].)
There is also a reference to thi9 battle on the ice in the Kdlfsvisa, a mnemonic list of famous heroes and their horses, it is noteworthy that in this list mention is made of Vestein, wlio is perhaps the Wihstan of our poem, and of Biar, who has been thought (very doubtfully) to correspond to the O.E. Beaw.
Dagr reip Drgsle en Dvalenn Mdpne...
Ale Hrafne es til iss ripo,
enn annarr austr und Apilse
grdr hvarfape geire undapr.
Bjgrn reip Blakke en Biarr Kerte,
Atle Glaume en Apils Slungne.., Lieder der Edda, ed. Symons and Gering, i, 221-2. "Ale was on Hrafn when they rode to the ice: but another horse, a grey one, with Athils on his back, fell eastward, wounded by the spear." This, as Olrik points out, appears to refer to a version of the story in which Athils had his fall from his horse, not at a ceremony at Uppsala, but after the battle with Ali. {HeUedigtning,!, 203-4.)
^ Unless they are among the fragments carried off to the Stockholm Museum. Little of interest was found in these moiuids when they were opened : everything had been too thoroughly burnt.
8 The Geatas — [CH. i
For various theories as to the early history of the Swedish royal house, as recorded in Beowulf, see Weyhe, Konig Ongentheows Fall, in Engl. Stvd. xxxix, 14-39 :Schuck, Studier i Ynglingatal (1905-7): Stjerna, Vendel och Vendelkrdka, in A.f.n.F. xxi, 71, etc.
The Geatas.
The identification of Geatas and Gotar has been accepted by the great majority of scholars, although Kemble wished to locate the Geatas in Schleswig, Grundtvig in Gotland, and Haigh in England. Leo was the first to suggest the Jutes: but the "Jute-hypothesis" owes its currency to the arguments of Fahlbeck (Beovulfsgvddet sasom kdlla Jor nordisk fornhistoria in the Antigvarisk Tidskrift for Sverige, vm, 2, 1 ). Fahlbeck's very inconclusive reasons were contested at the time by Sarrazin (23 etc.) and ten Brink (194 etc.) and the argu- ments against them have lately been marshalled by H. Schiick {Folknamnet Geatas i den fornengelska dikten Beowulf, Upsala, 1907). It is indeed difficult to understand how Fahlbeck's theory came to receive the support it has had from several scholars (e.g. Bugge, P.B.B. XII, I etc. ; Weyhe, En^l. Stud, xxxix, 38 etc. ; Gering). For his con- clusions do not arise naturally from the O.E. data : his whole argument is a piece of learned pleading, undertaken to support his rather revo- lutionary speculations as to early Swedish history. These speculations would have been rendered less probable had the natural interpretation of Geatas as Gotar been accepted. The Jute-hypothesis has recently been revived, with the greatest skill and learning, by Gudmund Schiitte [Journal of English and Germanic Philology, xi, 574 etc.). /But here again I cannot help suspecting that the wish is father to the t/ thought, and that the fact that that eminent scholar is a Dane living ^ in Jutland, Jhas something to do with his attempt to locate the Geatas .^y^^- there. tNoamo^i^* of learnin^will eradicatepatriotismj / ^ The foIlowi&^'^nsidCTa^ions'"TTecd^^ '
(1) Geatas etymologically corresponds exactly with O.N. Gautar, the modern Gotar. The O.E. word corresponding to Jutes (the lutae of Bede) should be, not Geatas, but in the Anglian dialect Eote, lote, in the West Saxon lete, Yte.
Now it is true that in one passage in the O.E. translation of Bede (i, 15) the word "lutarum" is rendered Geata: but in the other (IV, 16) "lutorum " is rendered Eota, Ytena. And this latter rendering is supported (a) by the Anglo-Saxon Chronicle (lotuin, lutna) and (6) by the fact that the current O.E. word for Jutes was Yte, Ytan, which survived till after the Norman conquest. For the name Ytena land was used for that portion of Hampshire which had been settled by the Jutes: William Rufus was slain, according to Florence of Worcester, in Ytene (which Florence explains as prouincia Jutarum).
From the purely etymological point of view the Gotar-hypothesis,
then, is unimpeachable: but the Jute-hypothesis'ls"urisatlSfactory,
since it is based upon one passage in the O.E. Bede, where Jutarum
is incorrectly rendered Geata, whilst it is invalidated by the other
passage in the O.E. Bede, by the Chronicle and by Florence of
Worcester, where Jutorum is correctly translated by Ytena, or its
AngHan or Kentish equivalent Eota, lotna.
J (2) It is obvious that the Geatas of Beoumlf were a strong and
) independent power — a match for the Swedes. Now we learn from
' Procopius that in the sixth century the Gotar were an independent
SECT, ii] tlieir Kings and their Wars 9
and numerous nation. But we have no equal evidence for any similar preponderant Jutish power in the sixth century. The lutae are indeed a rather puzzling tribe, and scholars have not even been able to agree where they dwelt.
The Gotar on the other hand are located among the great nations of Scandinavia both by Ptolemy {Geog. ti, 11, 16) in the secogd century and by Procogius '{Bell, upti, n, .15) in the §i^th. When we ^fiexiT^t clear ii3ormatioii (through the Christian missionaries) both Gotar and Swedes have been united under one king. But the Gotar retained their separate laws, traditions, and voice in the selection of the king, and they were constantly asserting themselves during the Middle Ages. The title of the king of Sweden, rex Sveorum Gothor- umque, commemorates the old distinction.
From the historical point of view, then, the Gotar comply with what we are told in Beowulf of the power of the Geatas much better i;han do the Jutes.
(3) Advocates of the Jute-hypothesis have claimed much support from the geographical argument that the Swedes and Geatas fight ofer sie (e.g. when BeowuK and Eadgils attack Onela, 2394). But the term see is just as appropriate to the great lakes Wener and Wetter, which separated the Swedes from the Gotar, as it is to the Cattegatt. And we have the evidence of Scandinavian sources that the battle between Eadgils and Onela actually did take place on the ice of lake Wener (see above, p. 6). Moreover the absence of any mention of ships in the fighting narrated in 11. 2922-2945 would be remarkable if the contending nations were Jutes and Swedes, but suits Gotar xind Swedes admirably: since they could attack each other by land as well as by water.
(4) There is reason to think that the old land of the Gotar in- cluded a great deal of what is now the south-west coast of Sweden^ Hygelac's capital was probably not far from the modern Goteborg. The descriptions in Beowulf would suit the cliffs of southern Sweden well, but they are quite inapplicable to the sandy dunes of Jutland^ u - Vj
Little weight can, however, be attached to this last argument, as ,s**^< the cliffs of the land of the Geatas are in any case probably drawn M from the poet's imagination.
(5) If we accept the identification Beowulf = Bjarki (see below, pp. 60-1) a further argument for the equation of Geatas and Gotar will be found in the fact that Bjarki travels to Denmark from Gautland just as Beowulf from the land of the Geatas; Bjarki is the brother of the king of the Gautar, Beowulf the nephew of the king of the Geatas.
(6) No argument as to -the meaning of Geatas can be drawn from ^ the fact that Gregory calls Chlochilaicus (Hygelac) a Dane. For it
is clear from Beowulf that, whatever else they may have been, the Geatas were not Danes. Either, then, Gregory must be misinformed, or he must be using the word Dane vaguely, to cover any kind of Scandinavian pirate.
(7) Probably what has weighed most heavily (often perhaps not consciously) in gaining converts to the "Jute-hypothesis" has been the conviction that "in ancient times each nation celebrated in song its own heroes alone." Hence one set of scholars, accepting the identification of the Geatas with the Scandinavian Gotar, have argued that Beowulf is therefore simply a translation from a Scandinavian Ootish original. Others, accepting Beowulf as an English poem, have
^ See Schiick, Folknamnet Geatas, 22 etc.
10 The Geatas — [CH. i
argued that the Geatas who are celebrated in it must therefore be one of the tribes that settled in England, and have therefore favoured the "Jute theory." But the a priori assumption that each Germanic tribe celebrated in song its own national heroes only is demonstrably incorrect^.
But in none of the accounts of the warfare of these Scandi- navian kings, whether written in Norse or monkish Latin, is there mention of any name corresponding to that of Beowulf, as king of the Geatas. Whether he is as historic as the other kings with whom in our poem he is brought into contact, we cannot say.
It has been generally held that the Beowulf of our poem is compounded out of two elements : that an historic Beowulf, king of the Geatas, has been combined with a mythological figure Beowa^, a god of the ancient Angles : that the historical achievements against Frisians and Swedes belong to the king, the mythological adventures with giants and dragons to the god. But there is no conclusive evidence for either of these presumed component parts of our hero. To the god Beowa we shall have to return later: here it is enough to note that — the current assumption that there was a king Beowulf of the i Geatas lacks confirmation from Scandinavian sources.
And one piece of evidence there is, which tends to show that Beowulf is not an historic king at all, but that his adventures have been violently inserted amid the historic names of the kings of the Geatas. Members of the families in Beowulf which we have reason to think historic bear names which alliterate the one with the other. The inference seems to be that it was customary, when a Scandinavian prince was named in the Sixth Century, to give him a name which had an initial letter similar to that of his father : care was thus taken that metrical difficulties should not prevent the names of father and son being linked together in song^. In the case of Beowulf himself, J^y however, this rule breaks down. Beowulf seems an intruder
1 See below, p. 98 and Appendix (E); The "Jute-Question."
2 See below, pp. 45 etc.
3 Olnk {Heltedigtnirg, J, 22 etc.). The Danish house — Healfdene, Heorogar,. Hrothgar, Halga, Heoroweard, Hrethric, Hrothmund, Hrothulf : the Swedish — Ongentheow, Onela, Ohthere, Eanmund, Eadgils: the Geatic — Hrethel, Here- beald, Hsetiicjm, Hygelac, Heardred. The same principle is strongly marked in the Old English pedigrees.
SECT, ii] their Kings and their Wars 11
into the house of Hrethel. It may be answered that since he was only the offspring of a daughter of that house, and since that daughter had three brothers, there would have been no prospect of his becoming king, when he was named. But neither does his name fit in with that of the other great house with which he is supposed to be connected. Wiglaf, son of V Wihstan of the Wsegmundingas, was named according to the famihar rules : but Beowulf, son of Ecgtheow, seems an intruder in that family as well.
This failure to fall in with the alliterative scheme, and the absence of confirmation from external evidence, are, of course, not in themselves enough to prove that the reign of Beowulf over the Geatas is a poetic figment. And indeed our poem may quite possibly be true to historic fact in representing him as the last of the great kings of the Geatas; after whose death •his people have nothing but national disaster to expect^. It would be strange that this last and most mighty and mag- nanimous of the kings of the Geatas should have been forgotten in Scandinavian lands : that outside Beowulf nothing should be known of his reign. But when we consider how Httle, outside Beowulf, we know of the Geatic kingdom at all, we cannot pronounce such oblivion impossible.
What tells much more against Beowulf as a historic Geatic king is that there is always apt to be something extravagant and unreal about what the poem tells us of his deeds, con- trasting with the sober and- historic way in which other kings, like Hrothgar or Hygelac or Eadgils, are referred to. True, we must not disqualify Beowulf forthwith because he slew a dragon 2. Several unimpeachably historical persons have done this: so sober an authority as the Anglo-Saxon Chronicle assures us that fiery dragons were flying in Northumbria as late as a.d. 7933.
1 11. 3018 etc.
2 As is done, e.g., by Schiick {Studier i Beoumlf-sagan, 27).
^ "Dragon fights are more frequent, not less frequent, the nearer we come to historic times" : Olrik, Heltedigtning, i, 313, The dragpn survived much later in Europe than has been generally recognized. He was flying from Mount. Pilatus in 1649. (See J. J, Scheuchzer, Itinera per Helvetiae Alpinas regiones, 1723, m, p. 385.) The same authority quotes accounts of dragons authenti- cated by priests, his own contemporaries, and supplies many bloodcurdling engravings of the same.
12 j The Geatas— [CH. i
But (and this is the serious difficulty) even when Beowulf is depicted in quite historic circumstances, there is still some- thing unsubstantial about his actions. When, in the midst of the strictly historical account of Hygelac's overthrow, we are told that Beowulf swam home bearing thirty suits of armour, this is as fantastic as the account of his swimming home from Orendel's lair with Grendel's head and the magic swordhilt. We may well doubt whether there is any more kernel of historic fact in the one feat than in the other ^. Again, we are told how Beowulf defended the young prince Heardred, Hygelac's son. Where was he, then, when Heardred was defeated and slain? To protect and if necessary avenge his lord upon the battle- field was the essential duty of the Germanic retainer. Yet Beowulf has no part to play in the episode of the death of Heardred. He is simply ignored till it is over. True, we are told that in later days he did take vengeance, by sup- porting the claims of Eadgils, the pretender, against Onela, the slayer of Heardred. But here again diffiqpjties meet us: for the Scandinavian authorities, whilst they agree that Eadgils overthrew Onela by the use of foreign auxiUaries, represent these auxiliaries as Danish retainers, dispatched by the Danish king Hrothulf . The chief of these Danish retainers is Bothvar Bjarki, who, as we shall see later, has been thought to stand in some relation to Beowulf. But Bothvar is never regarded as king of the Geatas : and the fact remains that Beowulf is at variance with our other authorities in representing Eadgils as having been placed on the throne by a Geatic rather than by a Danish force. Yet this Geatic expedition against Onela is,
^ with the exception of the dragon episode, the only event which our poem has to narrate concerning Beowulf's long reign of fifty years. And in other respects the reign is shadowy. Beowulf, we are told, came to the throne at a time of utter national distress; he had a long and prosperous reign, and became so powerful that he was able to dethrone the mighty^
V Swedish king Onela, and place in his stead the miserable fugitive^ Eadgils. Yet, after this half century of success, the
^ Cf. on this point Klaeber in Anglia, xxxvi (1912) p. 190. 2 1. 2382. 3 1. 2393.
SECT, ii] their Kings and their Wars fl^y
kingdom is depicted upon Beowulf's death as being in the same tottering condition in which it stood at the time when he is represented as having come to the throne, after the fall of Heardred.
The destruction one after the other of the descendants of Hrethel sounds historic : at any rate it possesses verisimihtude. But the picture of the chfldless Beowulf, dying, after a glorious reign, in extreme old age, having apparently made no previous arrangements for the succession, so that Wiglaf, a youth hitherto quite untried in war, steps at once into the place of command on account of his valour in slaying the dragon — thi& is a picture which lacks all historic probability.
I cannot avoid a suspicion that the fifty years' reign of Beowulf over the Geatas may quite conceivably be a poetic fiction^; that the downfall of the Geatic kingdom and its absorption in Sweden were very possibly brought about by the destruction of Hygelac and all his warriors at the mouth of >- the Rhine.
Such an event would have given the Swedes their op- portunity for vengeance : they may have swooped down, de- stroyed Heardred, and utterly crushed the independent kingdom of the Geatas before the younger generation had time to grow up into fighting men.
To the fabulous achievements of Beowulf, his fight with Grendel, Grendel's dam, and the dragon, it will be necessary to return later. As to his other feats, all we can say is that the common assumption that they rest upon an historic founda- tion does not seem to be capable of proof. But that they have \i an historic background is indisputable.
Section III. Heorot and the Danish Kings.
a
Of the Danish kings mentioned in Beowulf, we have first^\ Scyld Scefing, the foundling, an ancient and probably a mythi- cal figure, then Beowulf, son of Scyld, who seems an intruder among the Danish kings, since the Danish records know nothing
^ Of course, even if BeowuK's reign over the Geafas is not historic, this does not exclude the possibility of his having some historic foundation.
14 Heorot and the Danish Kings [CH. i
of him, and since his name does not aUiterate with those of either his reputed father or his reputed son. Then comes the
\/ "high" Healfdene, to whom four children were born : Heorogar, Hrothgar, Halga "the good," and a daughter who was wedded to the Swedish king. Since Hrothgar is represented as an elder contemporary of Hygelac, we must date^ Healfdene and his sons,
y should they be historic characters, between a.d. 430 and 520. Now it is noteworthy that just after a.d. 500 the Danes first become widely known, and the name "Danes" first meets us in Latin and Greek authors. And this cannot be explained on the ground that the North has become more famihar to dwellers in the classical lands: on the contrary far less is known concerning the geography of the North Sea and the Baltic than had been the case four or five centuries before. Tacitus and Ptolemy knew of many tribes inhabiting what is now Denmark, but not of the Danes : the writers in Ravenna and Constantinople in the sixth century, though much less well informed on the geography of the North, know of the
v/ Danes as amongst the most powerful nations there. Beowulf is, then, supported by the Latin and Greek records when it depicts these rulers of Denmark as a house of mighty kings, the fame of whose realm spread far and wide. We cannot tell to what extent this realm was made by the driving forth of alien nations from Denmark, to what extent by the coming together (under the common name of Danes) of many tribes which had hitherto been known by other distinct names.
The pedigree of the house of Healfdene can be constructed from the references in Beowulf. Healfdene's three sons,
y Heorogar, Hrothgar, Halga, are presumably enumerated in order of age, since Hrothgar mentions Heorogar, but not Halga, as his senior 2. Heorogar left a son Heoroweard^, but it is in accordance with Teutonic custom that HTOthgar should have succeeded to the throne if, as we may well suppose, Heoroweard was too young to be trusted with the kingship.
1 Attempts at working out the chronology of Beowulf have been made by Gering (in his translation) and by Heusler (Archiv, cxxiv, 9-14). On the whole the chronology of Beowulf is self -consistent, but there are one or two discrepancies which do not admit of solution.
2 1. 468. 3 1, 2161.
SECT. Ill] Heorot and the Danish Kings 15
The younger brother Halga is never mentioned during Beowulf's visit to Heorot, and the presumption is that he is already dead.
The Hrothulf who, both in Beowulf and Widsith, is linked • / with King Hrothgar, almost as his equal, is clearly the son of Halga : for he is Hrothgar's nephew^, and yet he is not the son of Heorogar^. The mention of how Hrothgar shielded this Hrothulf when he was a child confirms us in the belief that his father Halga had died early. Yet, though he thus belongs to the youngest branch of the family, Hrothulf is clearly older than Hrethric and Hrothmund, the two sons of Hrothgar, whose youth, in spite of the age of their father, is striking. The seat of honour occupied by Hrothulf^ is contrasted with the undistinguished place of his two young cousins, sitting ^ among the giogoth^. Nevertheless Hrothgar and his wife ex- pect their son, not their nephew, to succeed to the throne^. Very small acquaintance with the history of royal houses in^'^ these lawless Teutonic times is enough to show us that trouble is Hkely to be in store.
So much can be made out from the English sources, Beowulf and Widsith. Turning now to the Scandinavian records, we find much confusion as to details, and as to the characters of the heroes: but the relationships are the same as in the Old English poem.
Heorogar is, it is true, forgotten; and though a name Hiarwarus is found in Saxo corresponding to that of Heoroweard, the son of Heorogar, in Beowulf, this Hiarwarus is cut off from the family, now that his father is no longer remembered. Accordingly the Halfdan of Danish tradition (Haldanus in Saxo's Latin: = O.E. Healfdene) has only two sons, Hroar
1 Widsith, 1. 46.
2 Beoioulf, 1. 2160. Had Hrothulf been a son of Heorogar he could not have been passed over in silence here. Neither can Hrothulf be Hrothgar's sister's son: for since the sister married the Swedish king, Hrothulf would in that case be a Swedish prince, and presumably would be living at the Swedish court, and bearing a name connected by alliteration with those of the Swedish, not the Danish house. Besides, had he been a Swedish prince, he must have been heard of in connection with the dynastic quarrels of the Swedish house.
» 11. 1163-5. « IL 1188-91.
5 IL 1180 etc.
16 Heorot and the Danish Kings [CH. i
(Saxo's Koe, corresponding to O.E. Hrothgar) and Helgi
(Saxo's Helgo: - O.E. Halga). Helgi is the father of Kolf
^ Kraki (Saxo's Roluo : = O.E. Hrothulf), the type of the noble
(^ king, the Arthur of Denmark.
And, just as Arthur holds court at Camelot, or Charlemagne
is at home ad Ais, a sa capele, so the Scandinavian traditions
^/represent Rolf Kraki as keeping house at Leire {Lethra, Hlei(Sar
garter).
Accounts of all these kings, and above all of Rolf Kraki, meet us ' in a number of Scandinavian documents, of which three are par- ticularly important:
(1) Saxo Grammaticus (the lettered), the earlier books of whose Historia Danica are a storehouse of Scandinavian tradition and poetry, clothed in a difficult and bombastic, but always amusing, Latin. How much later than the English these Scandinavian sources are, we can realize by remembering that when Saxo was putting the finishing touches to his history, King John was ruHng in England.
There are also a number of other Danish-Latin histories and genealogies.
(2) The Icelandic Saga of Rolf KraJci, a late document belonging to the end of the middle ages, but nevertheless containing valuable matter.
(3) The Icelandic Skjoldunga saga, extant only in a Latin summary of the end of the sixteenth century.
Section IV. Leire and Heorot.
The village of Leire remains to the present day. It stands near the north coast of the island of Seeland, some five miles from Roskilde and three miles from the sea, in a gentle valley, through the midst of which flows a small stream. The village itself consists of a tiny cluster of cottages: the outstanding feature of the place is formed by the huge grave mounds scattered around in all directions. . The tourist, walking amid these cottages and mounds, may
feel fairly confident that he is standing on the site of Heorot.
There are two distinct stages in this identification : it must be proved (a) that the modern Leire occupies the site of the Leire (Lethra) where Rolf Kraki ruled, and (6) that the Leire of Rolf Kraki was built on the site of Heorot.
{a) That the modern Leire occupies the site of the ancient Leire has indeed been disputed^, but seems hardly open to
* Doubts are expressed, for example, in Trap's monumental topographical work (Kongeriket Danmark, ii, 328, 1898).
PLATE II
InLibrum II. HiSTORi.fl Danica- Saxonis Grammatics.
ANTIQLHSSIM^ IN DANIA ARCIS ET OPPIDI
LETHR/E
TOPOGRAPHIA
A. Scpulchriim Haraldi riyMetarL..
B. Sella Regular , 5)ronnin5/ltncn vulgo.
C. Locus, iibi Regia olirn crat_^»
D. S)r}lh<^f\) I foiTan ibi homagu Rcgibus
praftita-.
H. £)(iif«ilj/59 A Regis Olairepulchriira.
1. Por^s major, 5i)?'39l<brcf vulgo,
K. Equilc plim rcgiiim, ^<(l<bi<rg.
L. Stabulumpullisdcputatum9l:m,Soff6;JB
LEIRE IN THE SEVENTEENTH CENTURY
From Saxo Grammaticus, ed. Stephanius, 1644.
SECT, iv] Leire and Heorot 17
doubt, in view of tlie express words of the Danish chroniclers^. It is true that the mounds, which these early chroniclers probably imagined as covering the ashes of ' Haldanus ' or ' Roe,' and which later antiquaries dubbed with the names of other kings, are now thought to belong, not to the time of Hrothgar, but to the Stone or Bronze Ages. But this evidence that Leire was a place of importance thousands of years before Hrothgar or Hrothulf were born, in no wise invalidates the overwhelming evidence that it was their residence also.
The equation of the modem Leire with the Leire of Rolf Kraki we may then accept. We cannot be quite so sure of our thesis (6) : that the ancient Leire was identical with the*, site where Hrothgar built Heorot. But it is highly probable : for although Leire is more particularly connected with the memory of Rolf Kraki himself, we are assured, in one of the mediaeval Danish chronicles, that Leire was the royal seat of '^ Rolf's predecessors as well: of Ro (Hrothgar) and of Ro's father: and that Ro "enriched it with great magnificence 2." Ro also, according to this chronicler, heaped a mound at Leire over the grave of his father, and was himself buried at Leire under another mound.
Now since the Danish tradition represents Hrothgar as enriching his royal town of Leire, whilst Enghsh tradition commemorates him as a builder king, constructing a royal hall "greater than the sons of men had ever heard speak of" — ^it becomes very probable that the two traditions are reflections of the same fact, and that the site of that hall was Leire. That Heorot, the picturesque name of the hall itself, should, in English tradition, have been remembered, whilst that of the town where it was built had been forgotten, is natural^. For
^ For example Sweyn Aageson (e. 1200) had no doubt that the little village of Leire near Roskilde was identical with the Leire of story : Rolf Kraki, occisus in Leihra, qvae tunc famosissima Regis extitit curia, nunc autem Roskildensi vicina civitati, inter abjectissima ferme vix colitur oppida. Svenonis Aggonis Historia Regum Daniae, in Langebek, i, 45.
^ Ro...patrem vero suum Dan colle apud Lethram tumulavit Sialandie ubi sedem regni pro eo pater constituit, qvam ipse post eum divitiis multiplicibus ditavit. In the so-called Annates Esromenses, in Langebek, i, 224. Cf. Olrik, Heltedigtning, i, 188, 194. For further evidence, see Appendix (G) below.
' We must not think of Heorot as an isolated country seat. The Royal Hall would stand in the middle of the Royal Village, as in the case of the halls of Attila (Priscus in MoUer's Fragmenta, iv, 85) or Cynewulf {A.S. Chronicle, Anno 756).
O.B. 2
(
18 Leire and Heorot [ch. i
though the names of heroes survived in such numbers, after the settlement of the Angles in England, it was very rarely indeed, so far as we can judge, that the Angles and Saxons continued to have any clear idea concerning the places which had been familiar to their forefathers, but which they them- selves had never seen.
Further, the names of both Hrothgar and Hrothulf are linked with Heorot in English tradition in the same way as those of V Roe and Rolf are with Leire in Danish chronicles.
Yet there is some httle doubt, though not such as need
seriously trouble us, as to this identification of the site of
/ Heorot with Leire. Two causes especially have led students to
doubt the connection of Roe (Hrothgar) with Leire, and to place
elsewhere the great hall Heorot which he built.
In the first place, Rolf Kraki came to be so intimately as- sociated with Leire that his connection overshadowed that of Roe, and Saxo even goes so far in one place as to represent Leire as having been founded by Rolf^. In that case Leire clearly could not be the place where Rolf's predecessor built his royal hall. But that Saxo is in error here seems clear, for elsewhere he himself speaks of Leire as being a Danish strong- hold when Roy was a child^.
In the second place, Roe is credited with having founded the neighbouring town of Roskilde (Roe's spring)^ so that some have" wished to locate Heorot there, rather than at Leire, five miles to the west. But against this identification of Heorot with Roskilde it must be noted that Roe is said to have built Roskilde, not as a capital for himself, but as a market-place for the merchants: there is no suggestion that it was his royal town, though in time it became the capital, and its cathedral is still the Westminster Abbey of Denmark.
What at first sight looks so much in favour of our equating
1 Lethram pergitur, quod oppidum, a Roluone constructum eximiisque regni opihus illustratum, ceteris confinium prouinciarum urbibus regie fundacionis et sedis auctoritate prestabat. Saxo, Book n (ed. Holder, p. 58).
2 His cognitis Helgo filium Roluonem Lethrica arce conclusit, heredis saluti consulturus (p. 52).
3 A Roe Roskildia condita memoratur. Saxo, Book n (ed. Holder, p. 51). Roe's spring, after being a feature of the town throughout the ages, is now (owing perhaps to its sources having been tapped by a neighbouring mineral- water factory) represented only by a pump in a market-garden.
SECT, iv] Leire and Heorot 19
Roskilde with Heorot — the presence in its name of the element Ro (Hrothgar) — is in reaUty the most suspicious thing about the identification. There are other names in Denmark with the element Ro, in places where it is quite impossible to suppose that the king's name is commemorated. Some other ex- planation of the name has therefore to be sought, and it is very probable that Roskilde meant originally not "Hrothgar's spring," but "the horses' spring," and that the connection with King Ro is simply one of those inevitable pieces of popular etymology which take place so soon as the true origin of a name is forgotten^.
Leire has, then, a much better claim than Roskilde to being the site of Heorot: and geographical considerations confirm this. For Heorot is clearly imagined by the poet of Beowulf as being some distance inland; and this, whilst it suits ad- mirably the position of Leire, is quite inappHcable to Roskilde, which is situated on the sea at the head of the Roskilde fjord^. Of course we must not expect to find the poet of Beowulf, or indeed any epic poet, minutely exact in his geography. At the same time it is clear that at the time Beowulf was written there were traditions extant, dealing with the attack made upon Heorot by the ancestral foes of the Danes, a tribe called the Heathobeardan. These accounts of the fighting around Heorot must have preserved the general impression of its situation, precisely as from the Iliad we know that Troy is neither on the sea nor yet very remote from it. A poet would draw on his imagination for details, but would hardly alter a feature like this.
In these matters absolute certainty cannot be reached: but we may be fairly sure that the spot where Hrothgar built his "Hart-Hall" and where Hrothulf held that court to which the North ever after looked for its pattern of chivalry was
1 I owe this paragraph to mformation kindly supplied me by Dr Sofus Larsen, librarian of the University Library, Copenhagen.
2 It was once believed that, in prehistoric times, the sea came up to Leire also (Forchhammer. Steenstrup and Worsaae: Under s^gelser i geologisk-anti- qvarisk Reining, Kjjgbenhavn, 1851). A most exact scrutiny of the geology of the coast-Une has proved this to be erroneous. (Danmarks geologiske Unders^gelse T.R. 6. Beskrivelse til Kaaribladene Kj^henhavn og Roskilde, af K. Rje^rdam, Kjj^benhavn, 1899.)
2—2
20 Leire and Heorot [CH. i
1 Leire, where the grave mounds rise out of the waving corn-
Section V. The Heathobeaedan.
Now, as Beowulf is the one long Old EngHsh poem which happens to have been preserved, we, drawing our ideas of Old English stor;y^ almost exclusively from it, naturally think of Heorot as the scene of the fight with Grendel.
But in the short poem of Widsith, almost certainly older than Beowulf, we have a catalogue of the characters of the Old English heroic poetry. This catalogue is dry in itself, but is of the greatest interest for the light it throws upon Old Germanic heroic legends and the history behind them. And from Widsith it is clear that the rule of Hrothgar and Hrothulf at Heorot and the attack of the Heathobeardan upon them, rather than any story of monster- quelling, was what the old / poets more particularly associated with the name of Heorot. The passage in Widsith runs:
"For a very long time did Hrothgar and Hrothwulf, uncle and nephew, hold the peace together, after they had driven away the race of the Vikings and humbled the array of Ingeld, had hewed down at Heorot the host of the Heathobeardan."
The details of this war can be reconstructed, partly from the allusions in Beowulf, partly from the Scandinavian accounts. The Scandinavian versions are less primitive and historic. They have forgotten all about the Heathobeardan as an in- dependent tribe, and, whilst remembering the names of the leading chieftains on both sides, they see in them members of two rival branches of the Danish royal house.
We gather from Beowulf that for generations a blood feud has raged between the Danes and the Heathobeardan. Nothing is told us in Beowulf about the king Healfdelie, except that he
^ The presence at Leire of early remains makes it tempting to suppose that it may have been from very primitive times a stronghold or sacred place. It is impossible here to examine these conjectures, which would connect Heorot ultimately with the "sacred place on the isle of the ocean" mentioned by Tacitus. The curious may be referred to Much in P.B.B. xvn, 196-8 ; Mogk in PauU Grdr. (2) ra, 367 ; Kock in the Swedish Historisk Tidskrift, 1895, 162 etc. ; and particularly to the articles by Sarrazin : Die Hirsch Halle in Anglia, xix,. 368-91, Neue Beovmlfstudien {Der Grendelsee) in Engl. Stud, xui, 6-15.
SECT, v] The Heathoheardan 21
was fierce in war and that he lived to be old. From the Scan- dinavian stories it seems clear that he was concerned in the ^ Heathobard feud. According to some later Scandinavian ^ accounts he was slain by Frothi (= Froda, whom we know from Beowulf to have been king of the Heathoheardan) and this may well have been the historic fact^. How Hroar and Helgi (Hrothgar and Halga), the sons of Half dan (Healfdene), evaded the pursuit of Frothi, we learn from the Scandinavian tales; whether the Old EngHsh story knew anything of their hair-breadth escapes we cannot tell. Ultimately, the saga tells ' us, Hroar and Helgi, in revenge for their father's death, burnt tlie hall over the head of his slajer, Frothi 2. To judge from the hints in Beowulf, it would rather seem that the Old English tradition represented this vengeance upon Froda as having been inflicted in a pitched battle. The eldest brother Heorogar — known only to the English story — perhaps took his share in this feat. But, after his brothers Heorogar and Halga were dead, Hrothgar, left alone, and fearing vengeance in his turn, strove to compose the feud by wedding his daughter Freawaru *^^ to Ingeld, the son of Froda. So much we learn from the report which Beowulf gives, on his return home, to Hygelac, as to the state of things at the Danish court. .
Beowulf is depicted as carrying a very sage head upon his young shoulders, and he gives evidence of his astuteness by predicting^ that the peace which Hrothgar has purchased will not be lasting. Some Heathobard survivor of the fight in which Froda fell, will, he thinks, see a young Dane in the retinue of Freawaru proudly pacing the hall, wearing the treasures which his father had won from the Heathoheardan. Then the old warrior will urge on his younger comrade "Canst thou, my lord, tell the sword, the dear iron, which thy father carried to the fight when he bore helm for the last time, when the Danes slew him and had the victory? And now the son
* This seems to me much more probable than, as Obik supposes, that Froda fell in battle against Healfdene {Skjoldungasaga, 162 [80]).
2 Saga of Rolf Kraki, cap. iv.
' Olrik wishes to read the whole of this account, not as a prediction in the present future tense, but as a narrative of past events in the historic present. {Heltedigtning, i, 16: n, 38.) Considering the rarity of the historic present idiom in Old English poetry, this seems exceedingly unhkely.
22 The Heathoheardan [CH. i
of one of these slayers paces the hall, proud of his arms, boasts of the slaughter and wears the precious sword which thou by right shouldst wield^."
Such a reminder as this no Germanic warrior could long resist. So, Beowulf thinks, the young Dane will be slain: Ingeld will cease to take joy in his bride; and the old feud will break out afresh.
That it did so we know from Widsith, and from the same source we know that this Heathobard attack was repulsed by the combined strength of Hrothgar and his nephew Hrothulf .
But the tragic figure of Ingeld, hesitating between love for his father and love for his wife, between the duty of vengeance and his plighted word, was one which was sure to attract the interest of the old heroic poets more even than those of the victorious uncle and nephew. In the eighth century Alcuin, the Northumbrian, quotes Ingeld as the typical hero of song. Writing to a bishop of Lindisfarne, he reproves the monks for their fondness for the old stories about heathen kings, who are now lamenting their sins in Hell: "in the Kefectory," he says, " the Bible should be read : the lector heard, not the harper : patristic sermons rather than pagan songs. For what has Ingeld to do with Christ^?" This protest testifies eloquently to the popularity of the Ingeld story, and further evidence is possibly afforded by the fact that few heroes of story seem to have had so many namesakes in Eighth Century England.
What is emphasized in Beowulf is not so much the struggle in the mind of Ingeld as the stern, unforgiving temper of the grim old warrior who will not let the feud die down ; and this is the case also with the Danish versions, preserved to us in the Latin of Saxo Grammaticus. In two songs (translated by Saxo into "deUghtful sapphics") the old warrior Starcatherus stirs up Ingellus to his revenge:
"Why, Ingeld, buried in vice, dost thou delay to avenge thy father ? Wilt thou endure patiently the slaughter of thy righteous sire?...
1 U. 2047-2056.
2 Verba dei legantur in sacerdotali convivio ; ihi decet lectorem audiri, non citharistam, sermones patrunif non carmina gentilium. Quid Hinieldus cum Christo? See Jaffe's Monumenta Alcuiniana {Bibliotheca Rer. Oerm. mi), Berlin, 1873, p. 357; Epistolae, 81.
SECT, v] The Heathoheardan 23
Whilst thou takest pleasure in honouring thy bride, laden with gems, and bright with golden vestments, grief torments us, coupled with shame, as we bewail thine infamies.
>^* Whilst headlong lust urges thee, our troubled mind recalls the fashion of an earlier day, and admonishes us to grieve over many things.
For we reckon otherwise than thou the crime of the foes, whom now thou boldest in honour ; wherefore the face of this age is a burden to me, who have known the old ways.
By nought more would I desire to be blessed, if, Froda, I might see those^ guilty of thy murder paying the due penalty of such a crime*."
Starkath came to be one of the best-known figures in Scandinavian legend, the type of the fierce, unrelenting warrior. Even in death his severed head bit the earth : or according to another version " the trunk fought on when the head was gone^." Nor did the Northern imagination leave him there. It loved to follow him below, and to indulge in conjectures as to his bearing in the pit of Hell^.
Who the Heathobeardan were is uncertain. It is frequently argued that they are identical with the Longobardi; that the words Heatho-Bard and Long-Bard correspond, just as we get sometimes Gar-Dene, sometimes Hring-Dene. (So Heyne; Bremer in Pauls Ordr. (2) in, 949 etc.) The evidence for this is however unsatisfactory (see Chambers, Widsitli, 205). Since the year 186 a.d. onwards the Longobardi were dwelling far inland, and were certainly never in a position from which an attack upon the Danes would have been practicable. If, therefore, we accept the identification of Heatho- Bard and Long-Bard, we must suppose the Heathobeardan of Beowulf ,• to have been not the Longobardi of history, but a separate portion of the ^ ^^eople, which had been left behind on the shores of the Baltic, when the main body went south. But as we have no evidence for any such offshoot from the main tribe, it is misleading to speak of the Heatho- beardan as identical with the Longobardi : and although the similarity of one element in the name suggests some primitive relationship, that relationship may well have been exceedingly remote*.
/ » Saxo, Book vi (ed. Holder, 205, 212-13).
The contrast between this Ijn'ical outburst, and the matter-of-fact speech in which the old warrior in Beowulf eggs on the younger man, is thoroughly characteristic of the difference between Old English and Old Scandinavian heroic poetry. This difference is very noticeable whenever we have occasion to compare a passage in Beowulf with any parallel passage in a Scandinavian /poem, and should be carefully pondered by those who still beheve that Beoumlf is, in its present form, a translation from the Scandinavian.
=* Saxo, Book vin (ed. Holder, p. 274); Helga hvipa Hundingsbana-, ii, 19. See also Bugge, Helge-digtene, 157.
3 J)dttr porsteins Skelks in Flateyarhoh (ed. Vigfusson and Unger), i, 416.
rf^ * Similarly, there is certainly a primitive connection between the names
p of the Geatas (Gautar) and of the Goths : but they are quite distinct peoples :
we should not be justified in speaking of the Geatas as identical with the Goths.
24 The Heathoheardan [CH. i
It has further been proposed to identify the Heathoheardan with th^ HeruH^. The Heruli came from the Scandinavian district, overran Europe, and became famous for their valour, savagery, and value as light-armed troops. If the Heathobeardan are identical with the ^, Heruli, and if what we are told of the customs of the HeruU is true, Freawaru was certainly to be pitied. The Heruli were accustomed to put to death their sick and aged : and to compel widows to commit suicide.
The supposed identity of the Heruli with the Heathobeardan is however very doubtful. It rests solely upon the statement of Jordanes
ythat they had been driven from their homes by the Danes {Dani... Herulos propriis sedibus expulerunt). This is inconclusive, since the growth of the Danish power is likely enough to have led to colHsions with more than one tribe. In fact Beowulf tells us that Scyld "tore away the mead benches from many a people." On the other hand the dissimilarity of names is not conclusive evidence against the identification, for the word Heruli is pretty certainly the same as the Old English Eorlas, and is a complimentary nick-name appHed by the tribe to themselves, rather than their original racial designation. (/ v/ Nothing, then, is really known of the Heathobeardan, except that
evidence points to their having dwelt somewhere on the Baltic^.
The Scandinavian sources which have preserved the memory of this feud have transformed it in an extraordinary way. The Heatho- beardan came to be quite forgotten, although maybe some trace of their name remains in Hothbrodd, who is represented as the foe of Roe (Hrothgar) and Rolf (Hrothulf ). When the Heathobeardan were / forgotten, Froda and Ingeld were left without any subjects, and naturally came to be regarded, like Healfdene and the other kings with whom they were associated in story, as Danish kings. Ac- cordingly the tale developed in Scandinavian lands in two ways.
, Some documents, and especially the Icelandic ones^, represent the
/ struggle as a feud between two branches of the Danish royal house. Even here there is no agreement who is the usurper and who the victim, so that sometimes it is Froda and sometimes Healfdene who is represented as the traitor and murderer.
But another version* — the Danish — ^whilst making Froda and Ingeld into Danish kings, separates their story altogether from that of Healfdene and his house : in this version the quarrel is still thought .of as being between two nations, not as between the rightful heir to
V the throne and a treacherous and relentless usurper. Accordingly the feud is such as may be, at any rate temporarily, laid aside: peace between the contending parties is not out of the question. This version therefore preserves much more of the original character of the story, for it remains the tale of a young prince who, willing to marry into the house of his ancestral foes and to forgive and forget the old feud, is stirred by his more uurelenting nenchman into taking vengeance for his father. But, owing to the prince having come to be represented as a Dane, patriotic reasons have suggested to the
1 MiiUenhoff {Beovulf, 29-32) followed by Much {P.B.B. xvir, 201) and Heinzel {A.f.d.A. xvi, 271). The best account of the Heruli is in Procopius {Bell Gott. n, 14, 15).
2 See also Olrik, Heltedigtning, i, 21, 22: Sarrazin in Ev^l. Stud. XLn, li: Bugge, Helgi-digtene, 151-63; 181: Chambers, Widsith, p. 82 (note), pp. 205-6.
* Saga of Rolf KraTci : Skjoldungasaga.
* Best represented in Saxo.
SECT, v] The Heathoheardan i^i>
Danish poets and historians a quite different conclusion to the story. Instead of being routed, Ingeld, in Saxo, is successful in his revenge. See Neckel, Studien iiber Froffi in Z.j.d.A. XLvm, 182 : Heusler, Zur SkioldungendicUung in Z.j.d.A. XLvnr, 57 : Olrik, Skjoldungasaga, 1894, 112 [30]; Okik, Heltedigtning, ii, 11 etc.: Okik, Sakses Oldhistorie, 222-6: Chambers, Widsith, pp. 79-81.
Section VI. Hrothulf. \j
Yet, although the Icelandic sources are wrong in repre- ^ senting Froda and Ingeld as Danes, they are not altogether wrong in representing the Danish royal house as divided •^ against itself. Only they fail to place the blame where it really lay. For none of the Scandinavian sources attribute y any act of injustice or usurpation to Rolf Kraki. He is the ideal king, and his title to the throne is not supposed to be doubtful.
Yet we saw that, in Beowulf, the position of Hrothulf is - represented as an ambiguous one^, he is the king's too powerful y nephew, whose claims may prejudice those of his less dis- tinguished young cousins, the king's sons, and the speech of queen Wealhtheow is heavy with foreboding. "I know," she says, "that my gracious Hrothulf will support the young princes in honour, if thou, King of the Scyldings, shouldst leave the world sooner than he. I ween that he will requite our children, if he remembers all which we two have done for his pleasure and honour, being yet a child^." Whilst Hrethric and Hroth- mund, the sons of King Hrothgar, have to sit with the juniors, the giogoth^, Hrothulf is a man of tried valour, who sits side by side with the king: "where the two good ones sat, uncle and nephew: as yet was there peace between them, and each was true to the other*."
Again we have mention of "Hrothgar and Hrothulf. Heorot was filled full of friends : at that time the mighty Scylding^ folk in no wise worked treachery^." Similarly in Widsith the mention of Hrothgar and Hrothulf together seems to stir the / poet to dark sayings. " For a very long time did JIiothgaT a.naA 7 Hrothulf, uncle and nephew, hold the peace together*."
1 See above, p. 15. ^ u. 1180-87. ^ U. 1188-91.
« U. 1163-5. 6 11^ 1017-19. « U. 45-6.
^6 Hrothulf [CH. i
The statement that "as_vet" or "for a very long time" or "at that time" there was peace within the family, neces- y sarily implies that, at last, the peace was broken, that Hrothulf quarrelled with Hrothgar, or strove to set aside his sons^.
Further evidence is hardly needed; yet further evidence we have: by rather compHcated, but quite unforced, fitting . together of various Scandinavian authorities, we find that Hrothulf deposed and slew his cousin Hrethric.
Saxo Grammaticus tells us how Eoluo (Rolf = O.N. Hrolfr, O.E. Hrothulf) slew a certain Rj^ricus (or Hrserek = O.E. Hrethric) and gave to his own followers all the plunder which he found in the city of R^ricus. Saxo is here translating an older authority, the Bjarkamdl (now lost), and he did not know who R^ricus was: he certainly did not regard him as a son or successor of Roe (Hrothgar) or as a cousin of Roluo (Hrothulf). "Roluo, who laid low R^ricus the son of the covetous B^kus'^ is Saxo's phrase (qui natum B^ki R^ricum stravit avari). This would be a translation of some such phrase in the Bjarkamdl as Hra3reks hani hn^ggvanbauga, "the slayer of Hraerek Hnoggvanbaugi^."
But, when we turn to the genealogy of the Danish kings ^, we actually find a Hrmrekr Hnauggvanbaugi given as a king of Denmark about the time of Roluo. This R^ricus or Hrasrekr who was slain by Roluo was then, himself, a king of the Danes, and must, therefore, have preceded Roluo on the throne. But in that case R^ricus must be son of Roe, and identical with his namesake Hrethric, the son of Hrothgar, in Beowulf. For no one but a son of King Roe could have had such a claim to the throne as to rule between that king and his all powerful nephew Roluo^.
It is difficult, perhaps, to state this argument in a way which will be convincing to those who are not acquainted with Saxo's method of working. To those who realize how he treats
^ For a contrary view see Clarke, Sidelights, 100.
2 Saxo has mistaken a title hn/jggvanbaugi for a father's name, (hins) hn^ggva Bangs "(son of the) covetous Baug."
3 LangfeSgatal in Langebek, r, 5. The succession given in LangfelTgatal is Halfdan, Helgi and Hroar, Rolf, Hraerek: it should, of course, run Halfdan, Helgi and Hroar, Hraerek, Rolf, Hraerek has been moved from his proper place in order to clear Rolf of any suspicion of usurpation.
SECT, vi] Hrothtdf 27
his sources, it will be clear that R^ricus is the son of Roe, and is slain by Roluo. Translating the words into their Old^j^ English equivalents, Hrethric, son of Hrothgar, is slain by Hrothulf.
The forebodings of Wealhtheow were justified.
Hrethric is then almost certainly an actual historic prince who was thrust" from the throne by Hrothulf. Of Hrothmund^, his brother, Scandinavian authorities seem to know nothing. He is very likely a poetical fiction, a duplicate of Hrethric. For it is very natural that in story the princes whose lives are threatened by powerful usurpers should go in pairs. Hrethric and Hrothmund go together like Malcolm and Donalbain. Their helplessness is thus emphasized over against the one mighty figure, Rolf or Macbeth, threatening them^.
Yet this does not prove Hrothmund unhistoric. On the contrary it may well happen that the facts of history will coincide with the demands of well-ordered narrative, as was the case when Richard of Gloucester murdered two young princes in the Tower.
Two other characters, who meet us in Beowulf, seem to have some part to play in this tragedy.
It was a maxim of the old Teutonic poetry, as it is of the British Constitution, that the king could do no wrong: the real fault lay with the adviser. If Ermanaric the Goth slew his wife and his son, or if Irminfrid the Thuringian unwisely challenged Theodoric the Frank to battle, this was never supposed to be due solely to the recklessness of the monarch himself — it was the work of an evil counsellor — a Bikki or an Iring. Now we have seen that there is mischief brewing in Heorot — and we are introduced to a counsellor Unferth, the i thyle or official spokesman and adviser of King Hrothgar. And Unferth is evil. His jealous temper is shown by the hostile and inhospitable reception which he gives to Beowulf. And Beowulf's reply gives us a hint of some darker stain : " though
1 1. 1189.
^ See Olrik, Episke Love in Daiiske Studier, 1908, p. 79. Compare the remark of Goethe in Wilhelm Meister, as to the necessity of there being both a Rosencrantz arid a Guildenstem {Apprenticeship^ Book V, chap. v).
28 Hrothidf [CH. i
thou hast been the slayer of thine own brethren — thy flesh and blood : for that thou shalt suffer damnation in hell, good though thy wit may be^." One might perhaps think that Beowulf in these words was only giving the "countercheck quarrelsome," and indulging in mere reckless abuse, just as Sinfjotli (the Fitela of Beowulf) in the First Helgi Lay hurls at his foes all kinds of outrageous charges assuredly not meant to be taken literally. But, as we learn from the Helgi Lay itself, the uttering of such unfounded taunts was not considered good form ; whilst it seems pretty clear that the speech of Beowulf to Unferth is intended as an example of justifiable and spirited self-defence, not, like the speech of SinfjotH, as a storehouse of things which a well-mannered warrior should not say.
Besides, the taunt of Beowulf is confirmed, although but darkly, by the poet himself, in the same passage in which he has recorded the fears of Wealhtheow lest perhaps Hrothulf should not be loyal to Hrothgar and his issue: "Likewise there Unferth the counsellor sat at the foot of the lord of the Scyldingas: each of them [i.e. both Hrothgar and Hrothulf] ^ ^ trusted to his spirit : that his courage was great, though he had not done his duty by his kinsmen at the sword-play ^.^^
But, granting that Unferth has really been the cause of the death of his kinsmen, some scholars have doubted whether we are to suppose that he literally slew them himself. For, had that been the case, they urge, he could not be occupying a place of trust with the almost ideal king Hrothgar. But the record of the historians makes it quite clear that murder of kin did happen, and that constantly^. Amid the tragic complexities of heroic life it often could not be avoided. The comitatus- system, by which a man was expected to give unflinching support to any chief whose service he had entered, must often / have resulted in slaughter between men united by very close bonds of kin or friendship. Turning from history to saga, we find some of the greatest heroes not free from the stain. Sigmund,
1 11. 587-9. 2 u^ 1165-8.
' Perhaps such murder of kin was more common among the aristocratic houses than among the bulk of the population (Chadwick, H.A. 348). In some great families it almost becomes the rule, producing a state of things similar to that in present day Afghanistan, where it has become a proverb that a man is "as great an enemy as a cousin" (Pennell, Afghan Frontier, 30).
SECT, vi] Hrothidf ') 29
Gunnar, Hogni, Atli, Hrothulf, Heoroweard, Hnaef, Eadgils, Haethcyn, Ermanaric and Hildebrand were all marred with this taint, and indeed were, in many cases, rather to be pitied than blamed. I doubt, therefore, whether we need try and save Unferth's character by suggesting that the stern words of the poet mean only that he had indirectly caused the death of his brethren by failing them, in battle, at some critical moment^. I suspect that this, involving cowardice or incom- petence, would have been held the more unpardonable offence, and would have resulted in Unferth's disgrace. But a man might well have slain his kin under circumstances which, while leaving a blot on his record, did not necessitate his banishment from good society. All the same, the poet evi- dently thinks it a weakness on the part of Hrothgar and Hrothulf that, after what has happened, they still put their trust in Unferth.
Here then is the situation. The king has a counsellor : I v that counsellor is evil. Both the king and his nephew trust v the evil counsellor. A bitter feud springs up between the king and his nephew. That the feud was due to the machinations of the evil adviser can hardly be doubted by those who have studied the ways of the old Germanic heroic story. But it is only an inference: positive proof we have none.^
Lastly, there is Heoroweard. Of him we are told in
Beowulf very little. He is son of Heorogar (or Heregar),
Hrothgar' s elder brother, who was apparently king before him,
but died young^. It is quite natural, as we have seen, that,
if Heoroweard was too young for the responsibility when his
father died, he should not have succeeded to the throne. What
is not so natural is that he does not inherit his father's arms,
which one might reasonably have supposed Hrothgar would
have preserved, to give to him when he came of age. Instead,
Hrothgar gives them to Beowulf^. Does Hrothgar deliberately
avoid doing honour to Heoroweard, because he fears that
any distinction conferred upon him would strengthen a rival
^ This is proposed by Cosijn {Aanteekeningen, 21) and again independently by Lawrence in M.L.N, xxv, 167.
2 11. 467-9. 3 u. 2155-62.
s ; , . y V , / . , ^ -I . . .ft 1 7
Y
30 Hrotkulf [CH. I
whose claims to the throne might endanger those of his own sons? However this may be, in any future struggle for the throne Heoroweard may reasonably be expected to play some part. V Turning now to Saxo, and to the Saga of Rolf Kraki, we find that Rolf owed his death to the treachery of one whose
^name corresponds exactly to that of Heoroweard — Hiarwarus (Saxo), Hj^rvarthr {Saga). Neither Saxo nor the Saga thinks of Hiarwarus as the cousin of Rolf Kraki : they do not make it really clear what the cause of his enmity was. But they tell us that, after a banquet, he and his men treacherously rose upon Rolf and his warriors. The defence which Rolf and his men put up in their burning hall : the loyalty and defiance of Rolf's champions, invincible in death — these were amongst the most famous things of the North; they were told in the BjarJcamdl, now unfortunately extant in Saxo's paraphrase only.
But the triumph of Hiarwarus was brief. Rolf's .men all
^fell around him, save the young Wiggo, who had previously, in the confidence of youth, boasted that, should Rolf fall, he would avenge him. Astonished at the loyalty of Rolf's cham- pions, Hiarwarus expressed regret that none had taken quarter, declaring that he would gladly accept the service of such men. Whereupon Wiggo came from the hiding-place where he had taken refuge, and offered to do homage to Hiarwarus, by placing his hand on the hilt of his new lord's sword: but in doing so he drove the point through Hiarwarus, and rejoiced as he received his death from the attendants of the foe he had slain. It shows how entirely the duty of vengeance w^as felt to outweigh all other considerations, that this treacherous act
Joi Wiggo is always spoken of with the highest praise.
For the story of the fall of Rolf and his men see Saxo, Book ii (ed. Holder, pp. 55-68) : Saga of Rolf Kraki, caps. 32-34: Skjoldunga Saga (ed. Okik, 1894, 36-7 [118-9]).
How the feud between the different members of the Danish family forms the background to Beowulf was first explained in full detail 'by Ludvig Schr0der {Om Bjovulfs-drapen. Efter en raskke foredrag pa folke-hojskolen i Askov, kj0benhavn, 1875). Schr0der showed how the bad character of Unferth has its part to play: "It is a weakness in Hrothgar that he entrusts important office to such a man — a
SEC\ viil Khig Ofa S3
weakness which will carry its punishment." Independently the domestic feud was demonstrated again by Sarrazin (Bolf Krake und sein vetter im Beowulf liede'. Engl. Stud, xxiv, 144-5). The story has been fully worked out by Olrik {Heltedigtning, 1903, i, 11-18 etc.).
These views have been disputed by Miss Clarke (Sidelights, 102), who seems to regard as "hypotheses" of Olrik data which have been ascertained facts for more than a generation. Miss Clarke's contentions, however, appear to me to be based upon a misunderstanding of Olrik.
Section VII. King Offa.
1 The poem, then, is mainly concerned with the deeds of J Geatic and Danish kings : only once is reference made to a king of Anglian stock — Offa.
The Anglo-Saxon Chronicle tells us of several kings named y' Offa, but two only concern us here. Still remembered is the historic tyrant-king who reigned over Mercia during the latter half of the eighth century, and who was celebrated through the Middle Ages chiefly as the founder of the great abbey of '^ St Albans. This Ofla is sometimes referred to as Offa the Second, because he had a remote ancestor, Offa I, who, if the Mercian pedigree can be trusted, lived twelve generations earlier, and therefore presumably in the latter half of the fourth century. Offa I, then, must have ruled over the Angles >^ whilst they were still dwelling in Angel, their continental home, in or near the modern Schleswig.
Now the Offa mentioned in Beowulf is spoken of as related ^ to Garmund and Eomer (ms geomor). This, apart from the abundant further evidence, is sufficient to identify him with Offa I, who was, according to the pedigree, the son of Waermund ^ and the grandfather of Eomer.
This Offa I, king of Angel, is referred to in Widsiih. Widsith is a composite poem : the passage concerning Offa, though not the most obviously primitive portion of it, is, nevertheless, early: it may well be earUer than Beowulf. After a list of famous chieftains we are told:
Offa ruled Angel, Alewih the Danes ; he was the boldest of all these men, yet did he not in his deeds of valour surpass Offa. But Offa gained, first of men, by arms the greatest of kingdoms whilst yet a boy ; no one of equal age ever did greater deeds of valour in battle with his single sword: he drew the boundary against the Myrgingas at Fifeldor. The boundaries were held afterwards by the Angles and the Swaefe as Offa struck it out.
30 Hrothulf [cji. i
Mucli is obscure here: more particularly our ignorance as to the Myrgingas is to be regretted: but there is reason for thinking that they were a people dwelling to the south of the old continental home of the Angles.
After the lapse of some five centuries, we get abundant further information concerning Offa. The legends about him, though carried to England by the Anglian conquerors, must also have survived in the neighbourhood of his old kingdom of
\/ Angel : for as Angel was incorporated into the Danish kingdom, so these stories became part of the stock of Danish national legend. OfEa came to be regarded as a Danish king, and his
>^ story is told at length by the two earliest historians of Denmark, Sweyn Aageson and Saxo Grammaticus. In Saxo the story runs thus:
Wermund, king of Denmark, had a son Uifo [Offa], tall beyond the measure of his age, but dull and speechless. When Wermund grew blind, his southern neighbour, the king of Saxony, laid claim to Denmark on the ground that he was no longer fit to rule, and, relying upon Ufio's incapacity, suggested that the quarrel should be decided by their two sons in single combat. Wermund, in despair, offered himself to fight, in spite of his blindness : this offer the envoys of the Saxon king refused with insult, and the Danes knew not what to say. Thereupon Ufio, who happened to be present, suddenly asked leave to speak. Wermund could not believe that it was really his son who had spoken, but when they all assured him that it was, he gave the permission. "In vain," then said Ufio, "does the king of Saxony covet the land of Denmark, which trusts to its true king and its brave nobles: neither is a son wanting to the king nor a successor to the kingdom." And he offered to fight not only the Saxon prince, but any chosen champion the prince might bring with him.
The Saxon envoys accepted the offer and departed. The blind king was at last convinced, by passing his hands over him, that the speaker had been in truth his son. But it was found difficult to arm him; for his broad chest split the rings of every coat of mail: the largest, his father's, had to be cleft down the side and fastened with a clasp. Likewise no sword
SECT, vii] King Offa 33
was so well tempered that he did not shatter it by merely brandishing it, till the old king directed his men how they might find his ancient sword, Skrep (= ? stedfast) which he had buried, in despair, thinking his son unworthy of it. The sword, when found, was so frail from age that Uffo did not test it : for Wermund told him that, if he broke it, there was no other left strong enough for him.
So Uffo and his two antagonists were taken to the place of combat, an island in the river Eider. Crowds lined either bank, and Wermund stood prepared to throw himself into the river should his son be slain. UfEo held back at first, till he had discovered which of his antagonists was the more dangerous, since he feared the sword would only be good for one blow. Then, having by his taunts induced the champion to come to close quarters, he clove him asunder with one stroke. Wermund cried out that he had heard the sound of his son's sword, and asked where the blow had fallen: his attendants assured him that it had pierced, not any particular part, but the man's whole structure.
So Wermund drew back from the edge, desiring Hfe now as keenly as before he had longed for death. Finally Uffo smote his second antagonist through, thus opening a career which after such a beginning we may well believe to have been glorious.
The story is told again by Sweyn Aageson in a slightly ^ varying form. Sweyn's story has some good traits of its own — as when it makes Uffo enter the Usts girt with two swords, intending to use his father's only in an emergency. The worthless sword breaks, and all the Danes quake for fear: whereupon Uffo draws the old sword and achieves the victory. But above all Sweyn Aageson tells us the reason of Uffo's dumbness and incapacity, which Saxo leaves obscure: it was the result of shame over the deeds of two Danes who had combined to avenge their father upon a single foe. What is the incident referred to we can gather from Saxo. Two Danes, Keto and Wigo, whose father Frowinus had been slain by a hostile king Athislus, attacked Athislus together, two to one, thus breaking the laws of the duel. Uifo had wedded the sister of
O. B. 3
34 King Offa [CH. i
Keto and Wigo, and it was in order to wipe out the stain left
^upon his family and his nation by their breach of duelUng etiquette that he insisted upon fighting single-handed against two opponents.
That this incident was also known in England is rendered probable by the fact that Freawine and Wig, who correspond to Saxo's Frowinus and Wiggo, are found in the genealogy of English kings, and that an Eadgils, king of the Myrgingas, who is almost certainly the Athislus of Saxo^, also appears in Old English heroic poetry. It is probable then that the two tales were connected in Old Enghsh story : the two brethren shame- fully combine to avenge their father: in due time the family of the slain foe take up the feud: Ofia saves his country and his country's honour by voluntarily undertaking to fight one against two.
About the same time that the Danish ecclesiastics were ^ at work, a monk of St Albans was committing to Latin the
^ English stories which were still current concerning OfEa. The
object of the English writer was, however, local rather than
national. He wrote the Vitae duorum Offarum to celebrate
the historic Ofia, king of Mercia, the founder of his abbey, and
/that founder's ancestor, Offa I : popular tradition had confused
i/ the two, and much is told concerning the Mercian Offa that seems to belong more rightly to his forefather. The St Albans writer drew upon contemporary tradition, and it is evident that in certain cases, as when he gives two sets of names to some of the chief actors in the story, he is trying to harmonize two distinct versions : he makes at least one error which seems to point to a written source^. In one of the mss the story is illustrated by a series of very artistic drawings, which might possibly be from the pen of Matthew Paris himself^. These drawings depict a version of the story which in some respects differs from the Latin text which they accompany.
I The story is located in England. Warmundus is repre- sented as a king of the Western Angles, ruling at Warwick.
^ See Widsith, ed. Chambers, pp. 92-4.
2 See Rickert, " The Old English Offa Saga" in Mod. Phil n, esp. p. 75. * The common ascription of the Lives of the Offas to Matthew Paris is erroneous: they are somewhat earlier.
» ■» » « » * * »
^« • • •
PLATE III
|
H |
|
|
ffl |
|
|
O |
|
|
p:5 Ph |
|
|
C/i >^ |
|
|
s^ |
|
|
rn ^ |
|
|
|s |
rfj |
|
^ Q |
<M |
|
3g |
4 1«S |
|
Ci |
|
|
^ |
|
|
g« |
1 |
|
C/2 !/) |
o |
|
W *^ |
O |
|
M < |
2Q |
|
^^ |
^ |
|
f^° |
|
|
fi< |
|
|
t:) c/3 |
|
|
<S |
|
|
^ s |
|
|
a ^ |
|
|
^H |
|
|
<^ |
|
|
fe |
|
|
fe |
|
|
O |
SECT, vii] King Offa 35
Offa, his only son, was blind till his seventh, dumb till his thirtieth year. Accordingly an ambitious noble, Riganus, otherwise called Aliel, claims to be recognized heir, in hope of gaining the throne for his son, Hildebrand (Brutus). OfEa gains the gift of speech in answer to prayer; to the joy of his father and the councillors he vindicates his right, much as in the Danish story. He is knighted with a chosen body of companions, armed, and leads the host to meet the foe. He dashes across the river which separates the two armies, although his followers hang back. This act of cowardice on their part is not explained: it is apparently a reminiscence of an older version in which Ofia fights his duel single handed by the river, and his host look on. The armies join battle, but after a long struggle draw away from each other with the victory undecided. Offa remaining in front of his men is attacked by Brutus (or Hildebrand) and Sueno, the sons of the usurper, and slays them both (a second reminiscence of the duel-scene). He then hurls himself again upon the foe, and wins the victory.
Widsiih shows us that the Danish account has kept closer to the primitive story than has later English tradition. Widsith confirms the Danish view that the quarrel was with a foreign, not with a domestic foe, and the combat a duel, not a pitched battle: above all, Widsith confirms Saxo in repre- senting the fight as taking place on the Eider — hi Fifeldore^, whilst the account recorded by the monk of St Albans had localised the story in England.
^ The identification of Fifeldor with the Eider has been doubted, notably by Holthausen, though he seems less doubtful in his latest edition (third edit. n, 178). The reasons for the identification appear to me the following. Place names ending in dor are exceedingly rare. When, therefore, two independent authorities teU us that Offa fought at a place named Fifd-dor or Egi-dor^ it appears unlikely that this can be a mere coincidence: it seems more natural to assume that the names are corruptions of one original. But further, the connection is not limited to the second element in the name. For the Eider {Egidora, Mgiadyr) would in O.E. be Egor-dor: and Egor-dor stands to Fifd-dor precisely as egor-stream (Boethius, Metra, xx, 118) does to fifd-stream {Metra, XXVI, 26), ^'egor" and ^'fifeV being interchangexible synonyms. See note to Widsith f 1. 43 (p. 204). It is objected that the interchange of fifd and egor, though frequent in common nouns, would be unusual in the name of a place. The reply is that the Old English scop may not have regarded it as a place- name. He may have substituted fif el-dor for the synonymous egor-dor, "the monster gate," without realizing that it was the name of a definite place, just as he would have substituted ^/eZ-5^ream for egor-stream, "the monster stream, the sea," if alliteration demanded the change.
3—2
36 Kifig Offa [CH. i
In Beowulf too we hear of Offa as a mighty king, "the best of all mankind betwixt the seas." But, although his wars ar& referred to, we are given no details of them. The episode in V Beowulf relates rather to his wife Thryth, and his deaUngs with her. The passage is the most obscure in the whole poem, but this at least is clear : Thryth had an evil reputation for cruelty and murder: she wedded Offa, and he put a stop to her evil deeds: she became to him a good and loyal wife.
Now in the Lives of the two Off as quite a long space is devoted
to the matrimonial entanglements of both kings. Concerning
Offa I, a tale is told of how he succoured a daughter of the
king of York, who had been turned adrift by her father ; how
when his years were advancing his subjects pressed him to
marry : and how his mind went back to the damsel whom he
had saved, and he chose her for his wife. Whilst the king
was absent on his wars, a messenger whom he had sent with
a letter to report his victories passed through York, where the
\ wicked father of Offa's queen Hved. A false letter was sub-
^^ jstituted, commanding that the queen and her children should be
mutilated and left to die in the woods, because she was a witch
/and had brought defeat upon the king's arms. The order was
1 carried out, but a hermit rescued and healed the queen and her
I children, and ultimately united them to the king.
This is a popular folk-tale which is scattered all over Europe,
and which has many times been clothed in literary form: in
France in the romance of the Manekine, in English in the
r^ metrical romance of Emare, and in Chaucer's Man of Lawes
Tale. From the name of the heroine in the last of these
J versions, the tale is often known as the Constance-stoiy. But
it is clear that this tale is not identical with the obscure
story of the wife of Offa, which is indicated in Beowulf.
. When, however, we turn to the Life of Offa II, we do find
a very close parallel to the Thryth story.
This tells how in the days of Charles the Great a certain beautiful but wicked girl, related to that king, was condemned to death on account of her crimes, but, from respect for her birth, was exposed instead in a boat without sails or tackle, and driven ashore on the coast of King Offa's land. Drida, as
PLATE IV
DRIDA (THRYTH) ARRIVES IN THE LAND OF KING OFFA, *'IN NAUICULA ARMAMENTIS CARENTE"
From MS Cotton Nero D. I, fol U a.
^
SECT, vii] King Offa 37
she said her name was, deceived the king by a tale of injured innocence, and he committed her to the safe keeping of his mother, the Countess MarcelHna. Later, Offa fell in love with Drida, and married her, after which she became known as Quendrida. But Drida continued her evil courses and com- passed the death of St ^Ethelbert, the vassal king of East Anglia. In the end she was murdered by robbers — a just' punishment for her crimes — and her widowed husband built the Abbey of St Albans as a thank-offering for her death.
The parallel here is too striking to be denied : for Drida is but another way of spelHng Thryth, and the character of the murderous queen is the same in both stories. There are, however, striking differences : for whereas Thryth ceases from " her evil deeds and becomes a model wife to Offa, Drida con- tinues on her course of crime, and is cut off by violence in the midst of her evil career. How are we to account for the parallels and for the discrepancies?
As a matter of historical fact, the wife of Offa, king of Mercia, was named (not indeed Cwoenthryth, which is the form which should correspond to Quendrida, but) Cynethryth. The most obvious and facile way of accounting for the Hkeness between what we are told in Beowulf of the queen of Offa I, and what we are elsewhere told of the queen of Offa II, is to suppose that Thryth in Beowulf is a mere fiction evolved from / v the historic Cynethryth, wife of Offa II, and by poetic licence ' represented as the wife of his ancestor, Offa I. It was in this way she was explained by Professor Earle:
The name [Thrytho] was suggested by that of Cynethryth, Offa's queen.... The vindictive character here given to Thrytho is a poetic and veiled admonition addressed to Cynethryth^.
Unfortunately this, like many another facile theory, is open to fatal objections. In the first place the poem of Beowulf can, nX with fair certainty, be attributed to a date earlier than that at which the historic Offa and his spouse lived. Of course, it may be said that the Offa episode in Beowulf is an interpolation of a later date. But this needs proof. I There are metrical and above all syntactical grounds
* The Deeds of Beovmlf, lxxxv.
38 King Offa [CH. i
which have led most scholars to place Beowulf Yeiy early i. If we wish to regard the Offa-Thryth-ei^isode as a later interpolation, we ought first to prove that it is later in its syntax and metre. We have no right to assume that the episode is an interpolation merely because such an assumption may suit our theory of the development of Beowulf. So until reasons are forthcoming for supposing the episode of Thryth to be later than the rest . of the poem, we can but note that what we know of the date
J of Beowulf forbids us to accept Earle's theory that Thryth is a reflection of, or upon, the historic Cynethryth.
But there are diiB&culties in the way of Earle's theory even more serious than the chronological one. We know nothing very definitely about the wife of Ofia II, except her name, but
Y from a reference in a letter of Alcuin it seems clear that she was a woman of marked piety : it is not likely that she could have been guilty of deUberate murder of the kind represented in the Life of Offa II. The St Albans Life depends, so far as we know, upon the traditions which were current four centuries after her death. There may be, there doubtless are, some historic facts concerning Offa preserved in it : but we have no reason to think that the bad character of Offa's queen is one of them. Indeed, on purely intrinsic grounds we might well suppose the reverse. As a matter of history we know that Offa did put to death ^Ethelberht, the vassal king of East Anglia. When in the Life we find Offa completely exonerated, and the deed represented as an assassination brought about by the mahce and cruelty of his queen, it seems intrinsically likely that we are deahng with an attempt of the monks to clear their founder by transferring his cruel deeds to the account of his wife. So far, then, from Thryth being a reflection of an historic cruel queen Cynethryth, it is more probable that the influence has been in the reverse direction; that the pious Cynethryth has been represented as a monster of cruelty because she has not unnaturally been confused with a mythical Thryth, the wife of Offa I.
To this it may be objected that we have no right to assume remarkable coincidences, and that such a coincidence is in-
^ See below, pp. 105-12, and Appendix (D) below. ^U
I
SECT, vii] King Offa 39
volved by the assumption that there was a story of a mythical Thryth, the wife of Offa I, and that this existed prior to, and independently of, the actual wedding of Offa II to a C5me- thryth. But the exceeding frequency of the element thryth in the names of women robs this objection of all its point. Such a coincidence, far from being remarkable, would be the most natural in the world. If we look at the Mercian pedigree we find that almost half the ladies connected with it have that element thryth in their nancies. The founder of the house, Wihtlaeg, according to Saxo Grammaticus^, wedded Hermu- thruda, the old English form of which would be Eormenthryth.
It is to this lady Hermuthruda that we must now devote our attention. She belongs to a type which is common in folk- tale down to the time of Hans Andersen — the cruel princess who puts her lovers to death unless they can vanquish her in some way, worsting her in a contest of wits, such as the guessing of riddles, or a contest of strength, such as running, jumping, or wrestling. The stock example of this perilous maiden is, of course, for classical story Atalanta, for Germanic tradition the Brunhilt of the Nibelungen Lied, who demands from her wooer that he shall surpass her in all three feats ; if he fails in one, his head is forfeit^.
Of this type was Hermuthruda: "in the cruelty of her arrogance she had always loathed her wooers, ^ and inflicted upon them the supreme punishment, so that out of many there was not one but paid for his boldness with his head^," words which remind us strongly of what our poet says of Thryth.
Hamlet (Amlethus) is sent by the king of Britain to woo this maiden for him: but she causes Hamlet's shield and the commission to be stolen while he sleeps: she learns from the shield that the messenger is the famous and valiant Hamlet, and alters the commission so that her hand is requested, not for the king of Britain, but for Hamlet himself. With this request she complies, and the wedding is celebrated. But when Wihtlaeg (Vigletus) conquers and slays Hamlet, she weds the conqueror, thus becoming ancestress of Offa.
^ Wihtlaeg appears in Saxo as Vigletus (Book iv, ed. Holder, p. 106). 2 Nibelungen Lied, ed. Piper, 328. » Book iv (ed. Holder, p. 102).
40 King Offa [CH. i sect, vii
It may well be that there is some connection between the Thryth of Beowulf and the Hermuthruda who in Saxo weds Offa's ancestor — that they are both types of the wild maiden who becomes a submissive though not always happy wife. If so, the continued wickedness of Drida in the Life of Offa II would be an alteration of the original story, made in order to exonerate OfEa II from the deeds of murder which, as a matter of history, did characterize his reign.
^
w'
CHAPTER II__
THE NON-HISTORICAL ELEMENTS
Section I. The Geendel Fight.
When we come to the story of Beowulf's struggle with Grendel, with Grendel's mother, and with the dragon, we are faced by difficulties much greater than those which meet us when considering that background of Danish or Geatic history in which these stories are framed.
J In the first place, it is both surprising and confusing that, in the prologue, before the main story begins, another Beowulf is introduced, the son of Scyld Scefing. Much emphasis is laid upon the upbringing and youthful fame of this prince, and the glory of his father. Any reader would suppose that the poet is going on to tell of his adventures, when suddenly the story is switched off, and, after brief mention of this Beowulf's son, Healfdene, we come to Hrothgar, the building of Heorot, Grendel's attack, and the voyage of Beowulf the Geat to the rescue. /
Now " Beowulf" is an exceedingly rare name. The presence of the earlier Beowulf, Scyld's son, seems then to demand explanation, and many critics, working on quite different lines, have arrived independently at the conclusion that either the story of Grendel and his mother, or the story of the dragon, or both stories, were originally told of the son of Scyld, and only afterwards transferred to the Geatic hero. This has indeed been generally accepted, almost from the beginning of
42 The Grendel Fight [CH. ii
Beowulf criticism^. Yet, though possible enough, it does not 1 admit of any demonstration.
Now Beowulf, son of Scyld, clearly corresponds to a Beow or Beaw in the West Saxon genealogy. In this genealogy Beow is always connected with Scyld and Scef, and in some
/ versions the relations are identical with those given in Beowulf i Beow, son of Scyld, son of Scef, in the genealogies^, corre- sponding to Beowulf, son of Scyld Scefing, in our poem. Hence arose the further speculation of many scholars that the hero who slays the monsters was originally called, not Beowulf, but Beow, and that he was identical with the hero in the West Saxon pedigree ; in other words, that the original story was of a hero Beow (son of Scyld) who slew a monster and^a dragon : ^nd that this adventure was only subsequently transferred to Beowulf, prince of the Geatas.
This is a theory based upon a theory, and some confirmation may reasonably be asked, before it is entertained. As to the dragon-slaying, the confirmatory evidence is open to extreme doubt. It is dealt with in Section vii (Beowulf-Frotho), below. As to Grendel, one such piece of confirmation there is. The conquering Angles and Saxons seem to have given the names- of their heroes to the lands they won in England : some such names — 'Wade's causeway,' 'Weyland's smithy' — have sur- vived to modern times. The evidence of the Anglo-Saxon charters shows that very many which have now been lost ^ existed in England prior to the Conquest. Now in a Wiltshire
< charter of the year 931, we have Beowan hammes hecgan men- tioned not far from a Grendles mere. This has been claimed as evidence that the story of Grendel, with Beow as his adversary, was locaUzed in Wiltshire in the reign of Athelstan, and perhaps had been locaHzed there since the settlement four centuries previously. Until recently this was accepted as definitely
^ Kemble, Beovmlf, Postscript ix; followed by MuUenhojfif, etc. So, lately, Chadwick {H.A. 126): cf. also Sievers ('Beowulf und Saxo' in the Berichte d. k. sacks. Gesell. d. Wissenschaften, 1895, pp. 180-88); Bradley in Ericyc. Brit, m, 761; Boer, Beovmlf, 135. See also Olrik, Danmarks Heltedigtning, I, 246. For further discussion see below, Appendix (A).
2 Beo — Scyld — Scef in Ethelwerd : Beowius — Scddius — Sceaf in William of Malmesbury. But in the Anglo-Saxon Chronicle five generations intervene between Sceaf and his descendant Scyldwa, father of Beaw.
SECT, i] The Grendel Fight 43
y proving that the Beowulf-Grendel story was derived from an ancient Beow-myth. Yet one such instance of name-associa- tion is not conclusive. We cannot leave out of consideration the possibiUty of its being a mere chance coincidence, especially considering how large is the number of place names recorded in Old EngHsh charters. Of late, people have become more sceptical in drawing inferences from proper names, and quite recently there has been a tendency entirely to overlook the evidence of the charter, by way of making compensation for having hitherto overrated it.
All that can be said with certainty is that it is remarkable
, that a place named after Beowa should be found in the im- mediate proximity of a "Grendel's lake," and that this fact supports the possibihty, though it assuredly does not prove, that in the oldest versions of the tale the monster queller was
''^ named Beow, not Beowulf. But it is only a possibility : it is not grounded upon any real evidence.
These crucial references occur in a charter given by Athelstan at Luton, concerning a grant of land at Ham in Wiltshire to his thane Wulfgar. [See Birch, Cartularium Saxonicum, 1887, vol. n, p. 363.]
...Ego iESelstanus, rex Anglorum...quandam telluris particulam meo fideb* ministro Wulfgaro...in loco quern solicolae CBt Hamme vocitant tribuo...Praedicta siquidem tellus his terminis circumcincta clarescit....
tSonne norS ofer diine on meos-hlinc westeweardne ; tJonne adune on tSa yfre on beowan hammes hecgan, on bremeles sceagan easteweardne ; tJonne on Sa blacan grsef an ; Sonne nor© be Sem ondheaf dan to Ssere scortan die biitan anan aecre ; tSonne to fugelmere to San wege ; ondlong weges to ottes forda; Sonon to wudumere; Sonne to Ssere riiwan hecgan ; Saet on langan hangran ; Sonne on grendles mere ; Sonon on dyrnan geat....
Ambiguous as this evidence is, I do not think it can be dismissed as it is by Lawrence {Puh. Mod. Lang. Assoc. Amer. xxiv, 252) and Panzer {Beowulf, 397), who both say "How do we know that it is not the merest chance ? " It may of course be chance : but this does not justify us in basing an argument upon the assumption that it is the merest chance. Lawrence continues: "Suppose one were to set up a theory that there was a saga-relation between Scyld and Bikki, and offered as proof the passage in the charter for the year 917 in which there are mentioned, as in the same district, scyldes treow and bican sell.... How much weight would this carry?"
The answer surely is that the occurrence of the two names together in the charter would, by itself, give no basis whatever for starting such a theory : but if, on other grounds, the theory were likely, then the occurrence of the two names together would certainly have some corroborative value. Exactly how much, it is impossible to say, because we cannot estimate the element of chance, and we cannot
44 The Grendel Fight [CH. ii
be certain that the grendel and the beowa mentioned are identical with our Grendel and our Beowulf.
Miller has argued [Academy, May 1894, p. 396] that grendles is not a proper name here, but a common noun signifying "drain," and that grendles mere therefore means "cesspool."
Now "grindle" is found in modern dialect and even in Middle English^ in the sense of "a narrow ditch" or "gutter," but I doubt if it can be proved to be an Old EngUsh word. Evidence would rather point to its being an East Anglian corruption of the much more widely spread drindle, or dringle, used both as a verb "to go slowly, to trickle," and as "a snmll trickling stream." And even if an O.E. grendel as a common noun meaning "gutter" were authenticated, it seems unlikely to me that places were named "the fen," "the mere," "the pit," "the brook" — "of the gutter." There is no ground what- ever for supposing the existence of an O.E. gfrerwieZ= "sewer," or anything which would lead us to suppose grendles mere or gryndeles sylle to mean "cesspool^." Surely it is probable, knowing what we do of the way in which the English settlers gave epic names to the localities around their settlements, that these places were named after Grendel because they seemed the sort of place where his story might be localized — like " Weyland's smithy" or " Wade's causeway" : and that the meaning is "Grendel's fen," "mere," "pit" or "brook."
Again, both Panzer and Lawrence suggest that the Beowa who gave his name to the ham may have been, not the hero, but "an ordinary mortal called after him "..."some individual who lived in this locah'ty." But, among the numerous English proper names recorded, can any instance be found of any individual named Beowa ?
1 "Item there is vii acres lend lying by the high weye toward the grendyU" : Bury Wills, ed. S. Tymms (Camden See. xlix, 1850, p. 31).
2 I shoiid hardly have thought it worth while to revive this old "cesspool" theory, were it not for the statement of Dr Lawrence that "Miller's argument that the word grendel here is not a proper name at all, that it means 'drain,' has never, to my knowledge, been refuted." {Pvh. Mod. Lang. Assoc. Amer. XXIV, 253.)
Miller was a scholar whose memory should be reverenced, but the letter to the Academy was evidently written in haste. The only evidence which Miller produced for grendel standing alone as a common noun in Old English was a charter of 963 (Birch, 1103 : vol. m, p. 336) : J?anon for& eft on grendel: Jyanon on clyst: grendel here, he asserted, meant "drain": and consequently gryndeles sylle and grendles mere in the other charters must mean "cesspool." But the locaUty of this charter of 963 is known (Clyst St Mary, a few miles east of Exeter), and the two words exist there as names of streams to this day — "thence again along the Greendale brook, thence along the river Clyst." The Grindle or Greendale brook is no sewer, but a stream some half dozen miles in length which "winds tranquilly through a rich tract of alluvial soil" {Journal of the Archaeol. Assoc, xxxix, 273), past three villages which bear the same name, Greendale, Greendale Barton and Higher Greendale, under Greendale Bridge and over the ford by Greendale Lane, to its junction with the Clyst. Why the existence of this charming stream should be held to justify the interpretation of Grendel or Gryndel as "drain" and grendles mere as "cess- pool" has always puzzled me. Were a new Drayton to arise he might, in a new Polyolbion, introduce the nymph complaining of her hard lot at the hands of scholars in the Hesperides. I hope, when he next visits England, to conduct Dr Lawrence to make his apologies to the lady. Meantime a glance at the "six inch" ordnance map of Devon suffices to refute Miller's curious hypo- thesis.
SECT, i] The Grendel Fight 45
And was it in accordance with the rules of Old English nomenclature to give to mortals the names of these heroes of the genealogies^?
Kecent scepticism as to the "Beow-myth" has been largely due to the fact that speculation as to Beow had been carried too far. For example, because Beow appeared in the West Saxon genealogy, it had been assumed that the Beow-myth belonged essentially to the Angles and Saxons. Yet Beow would seem to have been also known among Scandinavians. For in somewhat later days Scandinavian genealogists, when they had made the acquaintance of the Anglo-Saxon pedigrees,, noted that Beow had a Scandinavian counterpart in a hero whom they called Bjar^. That something was known in the north of this Bjar is proved by the Kdlfsvisa, that same cata- logue of famous heroes and their horses which we have already found giving us the counterparts of Onela and Eadgils. Yet this dry reference serves to show that Bjar must once have been sufficiently famous to have a horse specially his own^. Whether the fourteenth century Scandinavian who made Bjar the Northern equivalent of Beow was merely guessing, -^e un- fortunately cannot tell. Most probably he was, for there is reason to think that the hero corresponding to Beow was named, not Bjar, but Byggvir^: a correspondence intelhgible to modern philologists as in agreement with phonetic law, but naturally not obvious to an Icelandic genealogist. But however this \ may be, the assumption that Beow was peculiarly the hero of ' Angles and Saxons seems hardly justified.
^ It is often asserted that the same Beowa appears as a witness to a charter (Miillenhoff, Beovulf, p. 8: Haak, Zeugnisse zur altenglischen Heldensage, 53). But this rests upon a misprint of Kemble {CD. 8. v, 44). The name is really Beoha (Birch, Cart. Sax. i, 212).
2 Beaf er ver hollum Biar, in the descent of Harold Fairhair from Adam, in Flateyarbdk, ed. Vigfiisson and Unger, Christiania, 1859, i, 27. [The genealogy contains many names obviously taken from a MS of the O.E. royal pedigrees, not from oral tradition, as is shown by the mis writings, e.g., Beaf for Beaw, owing to mistaking the O.E. w for/.] "This is no proof," Dr Lawrence urges, "of popular acquaintance with Bjar as a Scandinavian figure." (Pw6. Mod. Lang. Assoc. Amer. xxiv, 246.) But how are we to account for the presence of his name among a mnemonic Hst of some of the most famous warriors and their horses — mention along with heroes like Sigurd, Gimnar, Atli, Athils and Ali, unless Bjar was a well-known figure?
* en Bjdrr [rei(f] Kerti. Kortr, "short" (Germ. Kurz), if indeed we are so to interpret it, is hardly an Icelandic word, and seems strange as the name of a horse, Egilsson {Lex. Poet. 1860) suggests kertr, " erect," " with head high " (cf. Kahle in LF. XIV, 164). * See Appendix (A) below.
46 The Grendel Fight [CH. ii
/,V Again, since Beow is an ancestor of Woden, it was furtlier assumed that he was an ancient god, and that in the story of his adventures we had to deal with a nature-myth of a divine deUverer who saved the people from Grendel and his mother, the personified powers of the stormy sea. It is with the name of Miillenhoff, its most enthusiastic and ablest advocate, that this "mythological theory" is particularly associated. That Grendel is fictitious no one, of course, would deny. But MuUenhofi and his school, in applying the term "mythical" to those portions of the Beowulf story for which no historical explanation could be found, meant that they enshrined nature- myths. They thought that those elements in heroic poetry which could not be referred back to actual fact must be traced to ancient stories in which were recorded the nation's behef about the sun and the gods: about storms and seasons.
The different mythological explanations of Beowulf-Beowa and Grendel have depended mainly upon hazardous etymo- logical explanations of the hero's name. The most popular is ■ Miillenhoff's interpretation. Beaw is the divine helper of man in his struggle with the elements. Grendel represents the stormy North Sea of early spring, flooding and destroying the habitations of men, till the god rescues them : Grendel's mother represents the depths of the ocean. But in the autumn the power of the god wanes : the dragon personifies the coming of the wild weather: the god sinks in his final struggle to safe- guard the treasures of the earth for his people^. Others, remembering that Grendel dwells in the fen, see in him rather a demon of the sea-marsh than of the sea itself: he is the pestilential swamp^, and the hero a wind which drives him away^. Or, whilst Grendel still represents the storms, his antagonist is a " Blitzheros*." Others, whilst hardly ranking Beowulf as
^ Mullenhoff derived Beaw from the root 6M, "to be, dwell, grow" : Beaw therefore represented settled dwelling and culture. Miillenhoff's mythological explanation {Z.f.d.A. vn, 419, etc., Beovulf. 1, etc.) has been largely followed by subsequent scholars, e.g., ten Brink {Pauls Grdr. n, 533: Beowulf, 184), Symons {Pauls Ordr. (2), m, 645-6) and, in general outline, E. H. Meyer {Mythol, der Germanen, 1903, 242). 2 Uhland in Germania, 11, 349.
3 Laistner {Nebelsagen, 88, etc., 264, etc.), Kogel {Z.f.d.A. xxxvn, 274: Geschichte d. deut. Litt. 1, 1, 109), and Golther {Handhuch der germ. Mythologie, 1895, 173) see in Grendel the demon of combined storm and pestilence.
« E. H. Meyer {Germ. Mythol. 1891, 299).
SECT, i] The Grendel Fight
47
a god, still see an allegory in his adventures, and Grendel mast be a personification either of an inundation^, or of the terror of the long winter nights^, or possibly of grinding at the mill, the work of the enslaved foe^.
Such explanations were till recently universally current: the instances given above might be increased considerably.
Sufficient allowance was not made for the influence upon heroic poetry of the simple popular folk-tale, a tale of wonder \ with no mythological or allegorical meaning. Now, of late ' years, there has been a tendency not only to recognize but even to exaggerate this influence: to regard the hero of the folk- tale as the original and essential element in heroic poetry*. Though this is assuredly to go too far, it is but reasonable to recognize the fairy tale element in the O.E. epic.
We have in Beowulf a story of giant-kilhng and dragon- slaying. Why should we construct a legend of the gods or a nature-myth to account for these tales ? Why must Grendel or his mother represent the tempest, or the malaria, or the drear long winter nights ? We know that tales of giant-killers and dragon-slayers have been current among the people of Europe for thousands of years. Is it not far more easy to regard the story of the fight between Beowulf and Grendel merely as a fairy tale, glorified into an epic^?
Those students who of late years have tried thus to elucidate the story of Beowulf and Grendel, by comparison with folk- tales, have one great advantage over MiillenhofE and the "mythological" school. The weak point of Miillenhofi's view was that the nature-myth of Beow, which was called in to explain the origin of the Beowulf story as we have it, was itself only an assumption, a conjectural reconstruction. But the various popular tales in which scholars have more recently tried to find parallels to Beowulf have this great merit, that
^ Mogk {Pauls Ordr. (2), in, 302) regards Grendel as a "water-spirit."
2 Boer {Ark. f. nord. Filol. xix, 19).
' This suggestion is made (very tentatively) by Brandl, in Paula Qrdr. (2), Ti, i, 992.
* This view has been enunciated by Wundt in his Volkerpsychologie, n, i, 326, etc. J 382. For a discussion see A. Heusler in Berliner Sitzungsberichte, xxxvn, 1909, pp. 939-945.
^ Cf. Lawrence in Pub. Mod. Lang. Assoc. Amer. xxiv, 266, etc., and Panzer's "Beowulf" throughout.
48 The Grendel Fight [CH. ii
they do indubitably exist. And as to the first step — the A parallel between Beowulf and the Grettis saga — there can, fortunately, be but little hesitation.
Section II. The Scandinavian Parallels — Grettir and Orm.
The Grettis saga tells the adventures of the most famous of all Icelandic outlaws, Grettir the strong. As to the historic existence of Grettir there is no doubt: we can even date the main events of his life, in spite of chronological inconsistencies, with some precision. But between the year 1031, when he was killed, and the latter half of the thirteenth century, when his saga took form, many fictitious episodes, derived from folk-lore, had woven themselves around his name. Of these, one bears a great, if possibly accidental, likeness to the Grendel story: the second is emphatically and unmistakably the same story as that of Grendel and his mother. In the first, Grettir stops at a farm house which is haunted by Glam, a ghost of monstrous stature. Grettir awaits his attack alone, but, like Beowulf, lying down. Glam's entry and onset resemble those of Grendel : when Grettir closes with him he tries to get out. They wrestle the length of the hall, and break all before them. Grettir supports himself against anything that will give him foothold, but for all his efforts he is dragged as far as the door. There he suddenly changes his tactics, and throws his whole weight upon his adversary. The monster falls, undermost, so that Grettir is able to draw, and strike off his head ; though not till Glam has laid upon Grettir a curse which drags him to his- doom.
The second story — the adventure of Grettir at Sandhaugar (Sandheaps) — ^begins in much the same way as that of Grettir and Glam. Grettir is staying in a haunted farm, from which first the farmer himself and then a house-carl have, on two suc- cessive Yuletides, been spirited away. As before, a light burns in the room all night, and Grettir awaits the attack alone, lying down, without having put off his clothes. As before,. Grettir and his assailant wrestle down the room, breaking all
SECT, ii] The Scandinavian Parallels — Grettir and Orm 49
in their way. But this time Grettir is pulled out of the hall, and dragged to the brink of the neighbouring gorge. Here, by a final effort, he wrenches a hand free, draws, and hews off the ,, arm of the ogress, who falls into the torrent below.
Grettir conjectures that the two missing men must have been pulled by the ogress into the gulf. This, after his ex- perience, is surely a reasonable inference : but Stein, the priest, is unconvinced. So they go together to the river, and find the side of the ravine a sheer precipice: it is ten fathom down to the water below the fall. Grettir lets down a rope: the priest is to watch it. Then Grettir dives in: "the priest saw the soles of his feet, and then knew no more what had become of him." Grettir swims under the fall and gets into \. the cave, where he sees a giant sitting by a fire : the giant '' aims a blow at him with a weapon with a wooden handle (" such a weapon men then called a hefti-sax "). Grettir hews it asunder. The giant then grasps at another sword hanging on the wall of the cave, but before he can use it Grettir wounds him. Stein, the priest, seeing the water stained with blood from this wound, concludes that Grettir is dead, and departs )* home, lamenting the loss of such a man. "But Grettir let little space come between his blows till the giant lay dead." Grettir finds the bones of the two dead men in the cave, and X bears them away with him to convince the priest: but when he reaches the rope and shakes it, there is no reply, and he has to climb up, unaided. He leaves the bones in the church porch, for the confusion of the priest, who has to admit that he has failed to do his part faithfully.
Now if we compare this with Beowulf, we see that in the Icelandic story much is different: for example, in the Grettis saga it is the female monster who raids the habitation of men, the male who stays at home in his den. In this the Grettis saga probably represents a corrupt tradition: for, that the female should remain at home whilst the male searches for his prey, is a rule which holds good for devils as well as for men^.
^ The tradition of "the devil and his dam" resembles that of Grendel and his mother in its coupling together the home-keeping female and the roving male. See E. Lehmann, "Fandens Oldemor" in Dania, vm, 179-194; a paper which has been mideservedly neglected in the Beotoulf bibliographies. But the
O. B. 4
50 The Scandinavian Parallels — [CH. ii
The change was presumably made in order to avoid the difl&culty — which the Beowulf poet seems also to have realized — that after the male has been slain, the rout of the female is felt to be a deed of less note — something of an anti-climax^.
The sword on the wall, also, which in the Beowulf-stoij is ^ used by the hero, is, in the Grettir-stoij, used by the giant in his attack on the hero.
But that the two stories are somehow connected cannot be disputed. Apart from the general likeness, we have details such as the escape of the monster after the loss of an arm, the J^ fire burning in the cave, the hefti-sax, a word which, like its old English equivalent {hseft-mece, Beowulf, 1457), is found in this story only, and the strange reasoning of the watchers that the y blood-stained water must necessarily be due to the hero's death2.
Now obviously such a series of resemblances cannot be the result of an accident. Either the Grettir-stoiy is derived directly or indirectly from the Beowulf epic, more or less as we have it, or both stories are derived from one common earlier source. The scholars who first discovered the resemblance believed that both stories were independently derived from one original^. This view has generally been endorsed by later investigators, but not universally^. And this is one of the questions which the student cannot leave open, because our view of the origin of the Grendel-stoiy will have to depend largely upon the view we take as to its connection with the episode in the Grettis saga.
If this episode be derived from Beowulf, then we have an interesting literary curiosity, but nothing further. But if it is
devil beats his dam (cf. Piers Plowman, C-text, xxi, 284) : conduct of which one cannot imagine Grendel guilty. See too Lehmann in Arch. f. Religionstoiss. vin, 411-30: Panzer, Bacwulf, 130, 137, etc.: Klaeber in Anglia, xxxvi, 188.
1 Cf. Beovmlf, U. 1282-7.
2 There are other coincidences which may be the result of mere chance. In each case, before the adventure with the giants, the hero proves his strength by a feat of endurance in the ice-cold water. And, at the end of the story, the hero in each case produces, as evidence of his victory, a trophy with a runic inscription : in Beowulf an engraved sword-hilt ; in the Grettis saga bones and a "rune-staJBf."
3 Vigfiisson, Corp. Poet. Boreale, n, 502: Bugge, P.B.B. xii, 58.
* Boer, for example, believes that Beowulf influenced the Grettis saga {Grettis saga, Introduction, xUii); so, tentatively, Olrik {Hdtedigtning, i, 248).
SECT, ii] GretMr and Orm 51
independently derived from a common source, then the episode in the saga, although so much later, may nevertheless contain features which have been obhterated or confused or forgotten in the Beowulf version. In that case the story, as given in the Grettis saga, would be of great weight in any attempt to re- \ construct the presumed original form of the Grendel-atoiy. ]
The evidence seems to me to support strongly the view of the majority of scholars — that the (rre^^tV-episode is not de- rived from Beowulf in the form in which that poem has come / down to us, but that both come from one common source. ^
It is certain that the story of the monster invading a dwelling of men and rendering it uninhabitable, till the ad- venturous deUverer arrives, did not originate with Hrothgar and Heorot. It is an ancient and widespread type of story, of | which one version is localized at the Danish court. When therefore we find it existing, independently of its Danish setting, the presumption is in favour of this being a survival of the old independent story. Of course it is conceivable that the Hrothgar-Heorot setting might have been first added, and subsequently stripped off again so clean that no trace of it remains. But it seems going out of our way to assume this, unless we are forced to do so^.
Again, it is certain that these stories — like all the subject matter of the Old Enghsh epic — did not originate in England, v but were brought across the North Sea from the old home, j /l And that old home was in the closest connection, so far as the passage to and fro of story went, with Scandinavian lands. Nothing could be intrinsically more probable than that a story, current in ancient Angel and carried thence to England, should also have been current in Scandinavia, and thence have been carried to Iceland.
Other stories which were current in England in the eighth century were also current in Scandinavia in the thirteenth. Yet this does not mean that the tales of Hroar and Rolf, or of Athils and Ali, were borrowed from English epic accounts of Hrothgar and Hrothulf , or Eadgils and Onela. They were part of the common inheritance — as much so as the strong verbs
1 For this argument and the following, cf. Schiick, Studier i BeowuJfssagan, 21.
4—2
y
>
i
52 The Scandinavian Parallels — [CH. ii
or the alliterative line. Why then, contrary to all analogy, should we assume a literary borrowing in the case of the Beowulf-Grettir-atoTj ? The compiler of the Greltis saga could not possibly have drawn his material from a ms of Beowulp^ : he could not have made sense of a single passage. He con- ceivably might have drawn from traditions derived from the Old English epic. But it is difficult to see how. Long before his time these traditions had for the most part been forgotten in England itself. One of the longest lived of all, that of Offa, is heard of for the last time in England at the beginning of the thirteenth century. That a Scandinavian sagaman at the end of the century could have been in touch, in any way, with Anglo-Saxon epic tradition seems on the whole unlikely. The Scandinavian tradition of Offa, scholars are now agreed^, was not borrowed from England, and there is no reason why we should assume such borrowing in the case of Grettir.
The probability is, then, considerable, that the Beowulf- story and the Grettir-atoij are independently derived from one common original.
And this probability would be confirmed to a certainty if we should find that features which have been confused and half obliterated in the O.E. story become clear when we turn ' to the Icelandic. This argument has lately been brought forward by Dr Lawrence in his essay on "The Haunted Mere in Beowulpy Impressive as the account of this mere is, it does not convey any very clear picture. Grendel's home seems sometimes to be in the sea: and again it seems to be amid marshes, moors and fens, and again it is "where the mountain torrent goes down under the darkness of the cliffs — the water below the ground (i.e. beneath overhanging rocks)."
This last account agrees admirably with the landscape depicted in the Grettis saga, and the gorge many fathoms deep through which the stream rushes, after it has fallen over the precipice; not so the other accounts. These descriptions are
1 Even assuming that a ms of Beotvulf had found its way to Iceland, it would have been unintelligible. This is shown by the absurd blunders made when Icelanders borrowed names from the O.E. genealogies.
2 Cf. Ob-ik, A. f. n. F., vm (N.F. iv), 368-75; and Chadwick, Origin, 125-6.
3 Pub. Mod. Lang. Assoc. Amer. xxvn, 208 etc.
SECT, ii] Grettir and Orm 63
best harmonized if we imagine an original version in which the monsters live, as in the Grettis saga, in a hole under the waterfall. This story, natural enough in a Scandinavian country, would be less intelUgible as it travelled South. The '^ Angles and Saxons, both in their old home on the Continent and their new one in England, were accustomed to a somewhat flat country, and would be more inchned to place the dwelhng of outcast spirits in moor and fen than under waterfalls, of which they probably had only an elementary conception. "The giant must dwell in the fen, alone in the land^."
Now it is in the highest degree improbable that, after the landscape had been blurred as it is in Beowulf, it could have been brought out again with the distinctness it has in the Grettis saga. To preserve the features so clearly the Grettir- X story can hardly be derived from Beowulf: it must have come down independently.
But if so, it becomes at once of prime importance. For by a comparison of Beowulf and Grettir we must form an idea of what the original story was, from which both were derived
Another parallel, though a less striking one, has been -found in the story of Orm Storolfsson, which is extant in a short saga about contemporary with that of Grettir, Ormspdttr Storolfssonar^, in two ballads from the Faroe Islands^ and two /( from Sweden*.
It is generally asserted that the Orm-story affords a close parallel to the episodes of Grendel and his mother. I cannot find close resemblance, and I strongly suspect that the re- petition of the assertion is due to the fact that the Orm-story has not been very easily accessible, and has often been taken as read by the critics.
But, in any case, it has been proved that the Orm-tale borrows largely from other sagas, and notably from the Grettis saga itself^. Before arguing, therefore, from any parallel, it must first be shown that the feature in which Orm resembles
^ Cotton. Gnomic Verses, 11. 42-3. ^ Farnmannaaqgur, in, 204-228.
* Hammershaimb, Fseroiske Kvoeder, n, 1855, Nos. 11 and 12.
* A. I. Arwidsson, Svenska Fornaanger, 1834-42, Nos. 8 and 9, 5 Boer, Beowulf, 177-180.
54 The Scandinavian Parallels — Grettir and Orm [CH. ii
Beowulf is not derived at second hand from the Grettis saga. One such feature there is, namely Orm's piety, which he cer- tainly does not derive from Grettir. In this he with equal certainty resembles Beowulf. According to modem ideas, indeed, there is more of the Christian hero in Beowulf than in Orm.
Now Orm owes his victory to the fact, among other things, that, at the critical moment, he vows to God and the holy apostle St Peter to make a pilgrimage to Rome should he be successful. In this a parallel is seen to the fact that Beowulf is saved, not only by his coat of mail, but also by the divine interposition^. But is this really a parallel? Beowulf is too much of a sportsman to buy victory by making a vow when in a tight place. G^d' a wyrd swd hio sceP is the exact antithesis of Orm's pledge.
However, I have given in the Second Part the text of the Orm- episode, so that readers may judge for themselves the closeness or remoteness of the parallel.
The parallel between Grettir and Beowulf was noted by the Icelander Gudbrand VigMsson upon his first reading Beowulf (see Prolegomena to Sturlunga saga^ 1878, p. xUx: Corpus Poeticum BorealCf n, 501: Icelandic Reader, 1879, 404). It was elaborately- worked out by Gering in Anglia, ni, 74-87, and it is of course noticed in almost every discussion of Beowulf. The parallel with Orm was first noted by Schiick {Svensk Literaturhistoria, Stockholm, 1886, etc., I, 62) and independently by Bugge {P.B.B. xn, 58-68).
The best edition of the Grettis saga is the excellent one of Boer (Halle, 1900), but the opinions there expressed as to the relationship of the episodes to each other and to the Grendel story have not re- ceived the general support of scholars.
Section III. Bothvar Bjarki.
We have seen that there are in Beowulf two distinct elements, which never seem quite harmonized : firstly the historic back- ground of the Danish and Geatic courts, with their chieftains, Hrothgar and Hrothulf , or Hrethel and Hygelac : and secondly the old wives' fables of struggles with ogres and dragons. In the story of Grettir, the ogre fable appears — unmistakably connected with the similar story as given in Beowulf, but with
1 11. 1563-6. 2 1. 455_
Ii
I
SECT. Ill] Bothvar Bjarhi 66
no faintest trace of having ever possessed any Danish heroic setting.
Turning back to the Saga of Rolf Kraki, we do find against that Danish setting a figure, that of the hero Bothvar Bjarki, bearing a very remarkable resemblance to Beowulf.
Bjarki, bent on adventure, leaves the land of the Gautar (Gotar), where his brother is king, and reaches Leire, where Rolf, the king of the Danes, holds his court; [just as Beowulf, bent on adventure, leaves the land of the Geatas (Gotar) where his uncle is king, and reaches Heorot, where Hrothgar and Hrothulf (Rolf) hold court].
Arrived at Leire, Bjarki takes under his protection the despised coward Hott, whom Rolfs retainers have been wont to bully. The champions at the Danish court [in Beowulf one of them only — Unferth] prove quarrelsome, and they assail the hero during the feast, in the Saga by throwing bones at him, in Beowulf only by bitter words. The hero in each case replies, in kind, with such effect that the enemy is silenced.
But despite the fame and splendour of the Danish court, it has long been subject to the attacks of a strange monster^ — a winged beast whom no iron will bite [just as Grendel is immune from swords ^J. Bjarki [Uke Beowulf 3] is scornful at the inabihty of the Danes to defend their own home : " if one beast can lay waste the kingdom and the cattle of the king." He goes out to fight with the monster by night, accompanied only by Hott. He tries to draw his sword, but the sword is fast in its sheath : he tugs, the sword comes out, and he slays the beast with it. This seems a most pointless incident: taken in connection with the supposed invulnerability of the foe, it looks like the survival of some episode in which the hero was unwilling [as in Beowulf's fight with Grendel*] or unable [as in Beowulf's fight with Grendel's mother^] to slay the foe
^ The attacks have taken place at Yule for two successive years, exactly as in the Qrettia saga. [In Beowlf it is, of course, "twelve winters" (1. 147).] Is this mere accident, or does the Orettis saga here preserve the original time limit, which has been exaggerated in Bemmlft If so, we have another point of resemblance between the Saga of Rolf Kraki and the earliest version of the Beoumlf&toTy,
2 Beovmlf, U. 801-6. 3 Cf. Beovmlf, 11. 690-606.
< Beovmlf, 1. 679. » Beovmlf, U. 1608-9, 1524.
66 Bothvar BjarJd [CH. ii
with his sword. Bjarki then compels the terrified coward Hott to drink the monster's blood. Hott forthwith becomes a vahant champion, second only to Bjarki himself. The beast is then propped up as if still alive : when it is seen next morning the king calls upon his retainers to play the man, and Bjarki tells Hott that now is the time to clear his reputation. Hott demands first the sword, Gullinhjalti, from Kolf, and with this he slays the dead beast a second time. King Rolf is not deceived by this trick ; yet he rejoices that Bjarki has not only himself slain the monster, but changed the cowardly Hott into a champion; he commands that Hott shall be called Hjalti, after the sword which has been given him. We are hardly justified in demanding logic in a wild tale like this, or one might ask how Rolf was convinced of Hott's valour by what he knew to be a piece of stage management on the part of Bjarki. But, however that may be, it is remarkable that in Beowulf also the monster Grendel, though proof against all ordinary weapons, is smitten when dead by a magic sword of which the golden hilt^ is specially mentioned.
In addition to the undeniable similarity of the stories of these heroes, a certain similarity of name has been claimed. That Bjarki is not etymologically connected with Beowulf or Beow is clear: but if we are to accept the identification of Beowulf and Beow, remembering that the Scandinavian equi- valent of the latter is said to be Bjdr, the resemblance to Bjarki is\)bvious. Similarity of sound might have caused one name to be substituted for another^. This argument obviously depends upon the identification Beow = Bjdr, which is ex- tremely doubtful : it will be argued below that it is more likely that Beow = By ggvir^.
But force remains in the argument that the name Bjarki (little bear) is very appropriate to a hero like the Beowulf of
^ It is only in this adventure that Rolf carries the sword Gullinhjalti. His usual sword, as well known as Arthur's ExcaHbur, was Skofnungr. For Oyldenhiltf whether descriptive, or proper noun, see Beovmlf, 1677.
* Cf. Symons in Pauls Grdr. (2), m, 649 : Ziige aus dem anglischen Mythus von Beaw-Biar (Biarr oder Bjar?; s. Symons Lieder der Edda, i, 222) wurden auf den danischen Sagenhelden (BoSvarr) Bjarki durch Ahnlichkeit der Namen veranlasst, iibertragen. Cf. too, Heusler in A.f.d.A. xxx, 32.
* See p. 87 and Appendix (A) below.
I
SECT. Ill] Bothvar BjarTci 57
our epic, who crushes or hugs his foe to death instead of using his sword ; even if we do not accept explanations which would . interpret the name "Beowulf" itself as a synonym for "Bear." '
It is scarcely to be wondered at, then, that most critics have seen in Bjarki a Scandinavian parallel to Beowulf. But serious difficulties remain. There is in the Scandinavian story a mass of detail quite unparallelled in Beowulf, which over- shadows the resemblances, Bjarki's friendship, for example, with the coward Hott or Hjalti has no counterpart in Beowulf. And Bjarki becomes a retainer of King Rolf and dies in his service, whilst Beowulf never comes into direct contact with Hrothulf at all ; the poet seems to avoid naming them together. Still, it is quite intelhgible that the story should have developed on different lines in Scandinavia from those which it followed in England, till the new growths overshadowed the original resemblance, without obliterating it. After nearly a thousand years of independent development discrepancies must be ex- pected. It would not be a reasonable objection to the identity of Gullinhjalti with Gyldenhilt, that the word hilt had grown to have a rather different meaning in Norse and in English; subsequent developments do not invalidate an original re- semblance if the points of contact are really there.
But, allowing for this independent growth in Scandinavia, we should naturally expect that the further back we traced the story the greater the resemblance would become.
This brings us to the second, serious difficulty : that, when y^e turn from the Saga of Rolf Kraki — belonging in its present form perhaps to the early fifteenth century — to the pages of Saxo * Orammaticus, who tells the same tale more than two centuries '^ earlier, the resemblance, instead of becoming stronger, almost vanishes. Nothing is said of Bjarki coming from Gautland, or indeed of his being a stranger at the Danish court : nothing is said of the monster having paid previous visits, visits repeated till king Rolf, hke Hrothgar, has to give up all attempt at resistance, and submit to its depredations. The monster, instead of being a troll, like Grendel, becomes a commonplace bear. All Saxo tells us is that "He [Biarco, i.e. Bjarki] met a, great bear in a thicket and slew it with a spear, and bade his
58 Bothvar Bjarhi [CH. ii
comrade lalto [i.e. Hjalti] place his lips to the beast and drink its blood as it flowed, that he might become stronger."
Hence the Danish scholar, Axel Olrik, in the best and most elaborate discussion of Bjarki and all about him, has roundly- denied any connection between his hero and Beowulf. He is astonished at the slenderness of the evidence upon which previous students have argued for relationship. "Neither Beowulf's wrestling match in the hall, nor in the fen, nor his struggle with the firedrake has any real identity, but when we take a little of them all we can get a kind of similarity with the latest and worst form of the Bjarki saga^." The develop- ment of Saxo's bear into a winged monster, "the worst of trolls," Olrik regards as simply in accordance with the usual heightening, in later Icelandic, of these early stories of struggles with beasts, and of this he gives a parallel instance. ^
Some Icelandic ballads on Bjarki (the BjarJca rimur), which were first printed in 1904, were claimed by Olrik as supporting his contention. These ballads belong to about the year 1400. Yet, though they are thus in date and dialect closely allied to the Saga of Rolf Kraki and remote from Saxo Grammaticus, they are so far from supporting the tradition of the Saga with regard to the monster slain, that they represent the foe first as a man-eating she- wolf, which is slain by Bjarki, then as a grey bear [as in Saxo], which is slain by Hjalti after he has been compelled to drink the blood of the she-wolf. We must there- fore give up the winged beast as mere later elaboration; for if the Bjarki ballads in a point like this support Saxo, as against the Saga which is so closely connected with them by its date and Icelandic tongue, we must admit Saxo's version here to represent, beyond dispute, the genuine tradition.
Accordingly the attempt which has been made to connect Bjarki's winged monster with Beowulf's winged dragon goes overboard at once. But such an attempt ought never to have been made at all. The parallel is between Bjarki and the Beowulf-Grendel episode, not between Bjarki and the Beowulf- dragon episode, which ought to be left out of consideration. And the monstrous bear and the wolf of the Rimur are not so
Heltedigtning, i, 1903, 135-6.
I
SECT. Ill] Bothvar Bjarhi 69
dissimilar from Grendel, with his bear-hke hug, and Grendel's mother, the * sea- wolf i.'
The likeness between Beowulf and Bjarki lies, not in the , wingedness or otherwise of the monsters they overthrow, but 1 1 in the similarity of the position — in the situation which places the most famous court of the North, and its illustrious king, at the mercy of a ravaging foe, till a chance stranger from Gautland brings deliverance. And here the Rimur support, not Saxo, but the Saga, though in an outworn and faded way. In the Rimur Bjarki is a stranger come from abroad: the bear has made previous attacks upon the king's folds.
Thus, whilst we grant the wings of the beast to be a later elaboration, it does not in the least follow that other features in which the Saga differs from Saxo — the advent of Bjarki from Gautland, for instance — are also later elaboration.
And we must be careful not to attach too much weight to the account of Saxo merely because it is earlier in date than that of the Saga. The presumption is, of course, that the earlier form will be the more original : but just as a late manu- script will often preserve, amidst its corruptions, features which are lost in much earlier manuscripts, so will a tradition. Saxo's accounts are often imperfect^. And in this particular instance, there is a want of coherency and intelligibility in Saxo's account, which in itself affords a strong presumption that it is imperfect.
What Saxo tells us is this:
At which banquet, when the champions were rioting with every kind of wantonness, and flinging knuckle- bones at a certain lalto [Hjalti] from all sides, it happened that his messmate Biarco [Bjarki] through the bad aim of the thrower received a severe blow on the head. But Biarco, equally annoyed by the injury and the insult, sent the bone back to the thrower, so that he twisted the front of his head to the back and the back to the front, punishing the cross-grain of the man's temper by turning his face round about.
But who were this "certain Hjalti" and Bjarki? There seems to be something missing in the story. The explanation [which Saxo does not give us, but the Saga does] that Bjarki has come from afar and taken the despised Hott-Hjalti under his
1 Beovmlf, 1618.
2 See Heusler in Z.f.d.A. XLVin, 62.
60 Bothvar BjarU [CH. ii
protection, seems to be necessary. Why was Hjalti chosen as the victim, at whom missiles were to be discharged? Ob- viously [though Saxo does not tell us so], because he was the butt of the mess. And if Bjarki had been one of the mess for many hours, his messmates would have known him too well to throw knuckle-bones either at him or his friend. This is largely a matter of personal feeling, but Saxo's account seems to me pointless, till it is supplemented from the Saga^.
And there is one further piece of evidence which seems to clinch the whole matter finally, though its importance has been curiously overlooked, by Panzer and Lawrence in their argu- ments for the identification, and by Olrik in his arguments to the contrary.
We have seen above how Beowulf "became a friend" to Eadgils, helping him in his expedition against King Onela of Sweden, and avenging, in *' chill raids fraught with woe,'* cealdum cearsi^um, the wrongs which Onela had inflicted upon the Geatas. We saw, too, that this expedition was remembered in Scandinavian tradition. "They had a battle on the ice of Lake Wener; there King Ali fell, and Athils had the victory. Concerning this battle there is much said in the Skjoldunga sagay The Skjoldunga saga is lost, but the Latin extracts from it give some information about this battle^. Further, an account of it is preserved in the Bjarka rimur, probably derived from the lost Skjoldunga saga. And the Bjarka rimur expressly mention Bjarki as helping Athils in this battle against Ali on the ice of Lake Wener^.
' Olrik does not seem to allow for this at all, though of course aware of it. The other parallels between Bjarki and Beowulf he believes to be mere coincidence. But is this likely?
To recapitulate: In old English tradition a hero comes from the land of the Geatas to the royal court of Denmark, where Hrothgar and Hrothulf hold sway. This hero is re- ceived in none too friendly wise by one of the retainers, but
* Cf. on this Heusler, Z.f.d.A. XLVin, 64-5.
* Cf. Skjoldunga saga, cap. xii; and see Olrik, Heltedigtning, i, 201-5 Bjarka rimur, vm.
* Similarly Skdldskaparmdl, 41 (44).
i I
SECT. Ill] Bothvar Bjarhi 61
puts his foe to shame, is warmly welcomed by the king, and slays by night a monster which has been attacking the Danish capital and against which the warriors of that court have been helpless. The monster is proof against all swords, yet its dead body is mutilated by a sword with a golden hilt. Sub- sequently this same hero helps King Eadgils of Sweden to ^ overthrow Onela.
We find precisely the same situation in Icelandic tradition some seven centuries later, except that not Hrothgar and Hrothulf, but Hrothulf (Rolf) alone is represented as ruling the sj Danes, and the sword with the golden hilt has become a sword named " Golden-hilt." It is conceivable for a situation to have been reconstructed in this way by mere accident, just as it is conceivable that one player may have the eight or nine best trumps dealt him. But it does not seem advisable to base one's calculations, as Olrik does, upon such an accident happening.
The parallel of Bjarki and Beowulf seems to have been first noted by Gisli BrynjuKsson {Antiquarisk Tidsskrift, 1852-3, p. 130). It has been often discussed by Sarrazin [Beowulf Stvdien, 13 etc., 47 : Anglia, IX, 195 etc. Engl. Stud, xvi, 79 etc., xxni, 242 etc., xxxv, 19 etc.), Sarrazin' s over-elaborated parallels form a broad- target for doubters: it must be remembered that a case, though it may be discredited, is not invalidated by exaggeration. The problem is of course noted in the Beowulf studies of Miillenhoff (55), Bugge [P.B.B. xii, 55) and Boer [Die Beowulfsage, ii, in Arkiv f. nord. filol. xix, 44 etc.) and discussed at length and convincingly by Panzer (364-386) and Law- rence {Pub. Mod. Lang. Assoc. Amer. xxiv, 1909, 222 etc.). The usual view which accepts some relationship is endorsed by all these scholars, as it is by Finnur J6nsson in his edition of the Hrdlfs Saga Kraka og Bjarkarimur (K0benhavn, 1904, p. xxii).
Ten Brink (185 etc.) denied any original connection, on the ground of the dissimilarity between Beowulf and the story given by Saxo. Any resemblances between Beowulf and the Hrdlfs Saga he attributed to the influence of the EngHsh Beowulf-stoTj upon the Saga.
For Okik's emphatic denial of any connection at all, see Danmarks Heltedigtning, i, 134 etc. (This seems to have influenced Brandl, wha expresses some doubt in Pauls Grdr. (2) n. 1. 993.) For arguments to the contrary, see Heusler in A.f.d.A. xxx, 32, and especially Panzer and Lawrence as above.
The parallel of Gullinhjalti and gyldenhilt was first noted tentatively by Kluge {En^l. Stud, xxn, 145).
I
62 Parallels from FolUore [CH. ii
Section IV. Parallels from Folklore.
Hitherto we have been deahng with parallels to the Grendel story in written literature: but a further series of parallels, although much more remote, is to be found in that vast store of old wives' tales which no one till the nineteenth century took the trouble to write down systematically, but which certainly go back to a very ancient period. One particular tale, that of the Bear's Son^ (extant in many forms), has been instanced as showing a resemblance to the Beowulf -stoiy . In this tale the hero, a young man of extraordinary strength, (1) sets out on his adventures, associating with himself various companions ; (2) makes resistance in a house against a supernatural being, which his fellows have in vain striven to withstand, and succeeds in mishandUng or mutilating him. (3) By the blood-stained track of this creature, or guided by him in some other manner, the hero finds his way to a spring, or hole in the earth, (4) is lowered down by a cord and (5) overcomes in the underworld different supernatural foes, amongst whom is often included his former foe, or very rarely the mother of that foe : victory can often only be gained by the use of a magic sword which the hero finds below. (6) The hero is left treacherously in the lurch by his companions, whose duty it was to have drawn him up...
Now it may be objected, with truth, that this is not like the Beowulf-stoiy, or even particularly like the Grettir-stoTj. But the question is not merely whether it resembles these stories as we possess them, but whether it resembles the story which must have been the common origin of both. And we have only to try to reconstruct from Beowulf and from the Grettis saga a tale which can have been the common original of both, to see that it must be something extraordinarily lik; the folk-tale outUned above.
1 Barensohn. Jean I'Ours. The name is given to the group because tb hero is frequently (though by no means always) represented as having been brought up in a bear's den. The story summarized above is a portion of Panzer's "Type A." See Appendix (H), below.
I
SECT, iv] Parallels from FolUore 63
For example, it is true that the departure of the Danes homeward because they beheve that Beowulf has met his death in the water below, bears only the remotest resemblance to the deliberate treachery which the companions in the folk- tale mete out to the hero. But when we compare the Grettir- story, we see there that a real breach of trust is involved, for there the priest Stein leaves the hero in the lurch, and abandons the rope by which he should have drawn Grettir up. This can hardly be an innovation on the part of the composer of the Grettis saga, for he is quite well disposed towards Stein, and has no motive for wantonly attributing treachery to him. The innovation presumably Hes in the Beowulf -stoiy, where Hrothgar and his court are depicted in such a friendly spirit that no dis- reputable act can be attributed to them, and consequently Hrothgar's departure home must not be allowed in any way to imperil or inconvenience the hero. A comparison of the Beowulf-stoxy with the Grettir-stoiy leads then to the con- clusion that in the oldest version those who remained above when the hero plunged below were guilty of some measure of disloyalty , in ceasing to watch for him. In other words wa^ see that the further we track the Beowulf-stoij back, the) more it comes to resemble the folk-tale. ^
And our belief that there is some connection between the A folk-tale and the original of Beowulf must be strengthened when we find that, by a comparison of the folk-tale, we are able to explain features in Beowulf which strike us as difficult and even absurd: precisely as when we turn to a study of Shakespeare's sources we often find the explanation of things , that puzzle us : we see that the poet is dealing with an un- manageable source, which he cannot make quite plausible. V For instance: when Grendel enters Heorot he kills and eats the first of Beowulf's retinue whom he finds: no one tries to prevent him. The only explanation which the poet has to offer is that the retinue are all asleep^ — strange somnolence on the part of men who are awaiting a hostile attack, which they expect will be fatal to them all 2. And Beowulf at any rate is not asleep. Yet he calmly watches whilst his henchman is 1 U. 704, 729. 2 U. 691-6.
64 Parallels from Folklore [CH. ii
both killed and eaten: and apparently, but for the accident that the monster next tackles Beowulf himself, he would have allowed his whole bodyguard to be devoured one after another.
But if we suppose the story to be derived from the folk-tale, we have an explanation. For in the folk-tale, the companions and the hero await the foe singly, in succession: the turn of the hero comes last, after all his companions have been put to shame. But Beowulf, who is represented as having specially voyaged to Heorot in order to purge it, cannot leave the defence of the hall for the first night to one of his comrades. Hence the discomfiture of the comrade and the single-handed success of the hero have to be represented as simultaneous. The result is incongruous : Beowulf has to look on whilst his comrade is killed.
Again, both Beowulf and Grettir plunge in the water with a sword, and with the deliberate object of shedding the monster's blood. Why then should the watchers on the cliff above assume that the blood-stained water must necessarily signify the hero's death, and depart home? Why did it never occur to them that this deluge of blood might much more suitably proceed from the monster?
But we can understand this unreason if we suppose that the story-teller had to start from the deHberate and treacherous departure of the companions, whilst at the same time it was not to his purpose to represent the companions as treacherous. In that case some excuse must be found for them: and the blood-stained water was the nearest at hand^.
Again, quite independently of the folk-tale, many Beowulf scholars have come to the conclusion that in the original version of the story the hero did not wait for a second attack from the mother of the monster he had slain, but rather, from a natural and laudable desire to complete bis task, followed the monster's tracks to the mere, and finished him and his mother below. Many traits have survived which may conceivably point to an original version of the story in which Beowulf (or the figure corresponding to him) at once plunged down
^ In the Beoumlf it was even desirable, as explained above, to go further, and completely to exculpate the Danish watchers, t/
SECT, iv] Parallels from Folklore 65
in order to combat the foe corresponding to Grendel. There are unsatisfactory features in the story as it stands. For why, it might be urged, should the wrenching ofi of an arm have been fatal to so tough a monster? And why, it has often been asked, is the adversary under the water sometimes male, some- times female ? And why is it apparently the blood of Grendel, not of his mother, which discolours the water and burns up the sword, and the head of Grendel, not of his mother, which is brought home in triumph? These arguments may not carry much weight, but at any rate when we turn to the folk-tale we find that the adventure beneath the earth is the natural following up of the adventure in the house, not the result of any renewed attack.
In addition, there are many striking coincidences between individual versions or groups of the folk-tale on the one hand and the Beowulf-Grettir story on the other: yet it is very difficult to know what value should be attached to these parallels, since there are many features of popular story which float around and attach themselves to this or that tale without any original connection, so that it is easy for the same trait to recur in Beowulf and in a group of folk-tales, without this proving that the stories as a whole are connected^.
The hero of the Bear's son folk-tale is often in his youth unmanageable or lazy. This is also emphasized in the stories both of Grettir and of Orm: and though such a feature was uncongenial to the courtly tone of Beowulf, which sought to depict the hero as a model prince, yet it is there^, even though only alluded to incidentally, and elsewhere ignored or even denied^.
Again, the hero of the folk-tale is very frequently (but not necessarily) either descended from a bear, nourished by a bear, or has some ursine characteristic. We see this recurring in certain traits of Beowulf such as his bear-like method of hugging
^ From the controversial point of view Panzer has no doubt weakened his case by drawing attention to so many of these, probably accidental, coincidences. It gives the critic material for attack (cf, Boer, Beowulf, 14)
2 U. 2183 etc.
3 U. 408-9.
C. B. 5
66 Parallels from Folklore [CH. ii
his adversary to death. Here again the courtly poet has not emphasized his hero's wildness^.
Again, there are some extraordinary coincidences in names, between the Beowulf-Grettir story and the folk- tale. These are not found in Beowulf itself, but only in the stories of Grettir and Orm. Yet, as the Grettir-eipisodG is presumably derived from the same original as the Beowulf- eipisode, any original connection between it and the folk- tale involves such connection for Beowulf also. We have seen that in Grettis saga the priest Stein, as the unfaithful guardian of the rope which is to draw^ up the hero, seems to represent the faithless companions of the folktale. There is really no other way of accounting for him, for except on this supposition he is quite otiose and unnecessary to the Grettir-stoij : the saga-man has no use for him. And his name confirms this explanation, for in the folk- tale one of the three faithless companions of the hero is called the Stone-cleaver, Steinhauer, Stenhlj>ver, or even, in one Scandinavian version, simply Stein^.
Again, the struggle in the Grettis saga is localized at Sand- haugar in Barthardal in Northern Iceland. Yet it is difficult to say why the saga-teller located the story there. The scenery, with the neighbouring river and mighty waterfall, is fully described : but students of Icelandic topography assert that the neighbourhood does not at all lend itself to this description^. When we turn to the story of Orm we find it locaUzed on the island Sandey. We are forced to the conclusion that the name belongs to the story, and that in some early version this was localized at a place called Sandhaug, perhaps at one of the numerous places in Norway of that name. Now turning to one of the Scandinavian versions of the folk-tale, we find that the descent into the earth and the consequent struggle is localized in en stor sandhaug^.
^ It comes out strongly in the Bjarki-story.
2 It can hardly be argued that Stein is mentioned because he was an historic character who in some way came into contact with the historic Grettir: for in this case his descent would have been given, according to the usual custom in the sagas. (Cf. note to Boer's edition of Grettis saga, p. 233.)
^ P. E. K. Kaalund, Bidrag til en historisk-topografisk Beskrivelse af Island, Kj^benhavn, 1877, n, 151.
* The locaUzation in en stor sandhaug is found in a version of the story to which Panzer was unable to get access (see p. 7 of his Beoivulf, Note 2). A copy
SECT, iv] Parallels from Folklore ^ Q7
On the other hand, it must be remembered that if a collection is made of some two hundred folk-tales, it is bound to contain, in addition to the essential kernel of common tradition, a vast amount of that floating material which tends to associate itself with this or that hero of story. Individual versions or groups of versions of the tale may contain features which occur also in the Grendel-stoiy, without that being any evidence for primitive connection. Thus we are told how Grendel forces open the door of Heorot. In a Sicilian version of the folk-tale the doors spring open of themselves as the foe appears. This has been claimed as a parallel. But, as a sceptic has observed, the extraordinary thing is that of so slight a similarity (if it is entitled to be called a similarity) we should find only one example out of two hundred, and have to go to Sicily for that^.
The parallel between the Beonmlf-story and the "Bear's son'* folk-tale had been noted by Laistner (Das Rdtsel der Sphinx, Berlin, 1889, n, 22 etc.): but the prevalent behef that the Beoumlf-storj was a nature-myth seems to have prevented further investigation on these Unes till Panzer independently (p. 254) undertook his monumental work.
Yet there are other features in the folk-tale which are entirely unrepresented in the Beowulf-Grettir story. The hero of the folk-tale rescues captive princesses in the underworld (it is because they wish to rob him of this prize that his com- panions leave him below); he is saved by some miraculous helper, and finally, after adopting a disguise, puts his treacherous comrades to shame and weds the youngest princess. None of these elements^ are to be found in the stories of Beowulf, Grettir, Orm or Bjarki, yet they are essential to the fairy tale^.
is to be found in the University Library of Christiania, in a small book entitled Nor, en Billedbog for den norske Ungdom. Christiania, 1865. {Norske Folke- Evenly r...fortalte af P. C. Asbj0rnsen, pp. 65-128.)
The sandhaug is an extraordinary coincidence, if it is a mere coincidence. It cannot have been imported into the modem folk-tale from the Orettia saga, for there is no superficial resemblance between the two tales.
1 Cf. Boer, Beowulf, 14.
2 Yet both Beowulf and Orm are saved by divine help.
^ Panzer exaggerates the case against his own theory when he quotes only six versions as omitting the princesses (p. 122). Such unanimity as this is hardly to be looked for in a collection of 202 khidred folk-tales. In addition to these six, the princesses are altogether missing, for example, in the versions which Panzer numbers 68, 69, 77 : they are only faintly represented in other versions (e.g. 76). Nevertheless the rescue of the princesses may be regarded as the most essential element in the tale.
6—2
68 Parallels from FolUore [CH. ii
So that to speak of Beowulf as a version of the fairy tale is undoubtedly going too far. All we can say is that some early story-teller took, from folk-tale, those elements which suited his purpose, and that a tale, containing many leading features found in the "Bear's son" story, but omitting many of the leading motives of that story, came to be told of Beowulf and of Grettiri.
Section V. Scef and Scyld.
Our poem begins with an account of the might, and of the funeral, of Scyld Scefing, the ancestor of that Danish royal house which is to play so large a part in the story. After Scyld's death his retainers, following the command he had given them, placed their beloved prince in the bosom of a ship, surrounded by many treasures brought from distant lands, by weapons of battle and weeds of war, swords and byrnies. Also they placed a golden banner high over his head, and let the sea bear him away, with soul sorrowful and downcast. Men could not say for a truth, not the wisest of councillors, who received that burden.
Now there is much in this that can be paralleled both from the literature and from the archaeological remains of the North. Abundant traces have been found, either of the burial or of the burning of a chief within a ship. And we are told by different authorities of two ancient Swedish kings who, sorely wounded, and unwilling to die in their beds, had themselves placed upon ships, surrounded by weapons and the bodies of the slain. The funeral pyre was then lighted on the vessel,, and the ship sent blazing out to sea. Similarly the dead body of Baldr was put upon his ship, and burnt.
Haki konungr fekk sv^ stor sar, at hann si, at bans lifdagar mundu eigi langir verSa ; ])k let hann taka skeiS, er hann atti, ok let hlatJa dauSum mgnnum, ok vapnum, 16t ]>a flytja lit til hafs ok leggja styri
^ I cannot agree with Panzer when (p. 319) he suggests the possibility of the Beovmlf and the Grettir-story having been derived independently from the folk-tale. For the two stories have many features in common which do not belong to the folk-tale : apart from the absence of the princesses we have the hspft-mece and the strange conclusion drawn by the watchers from the blood- stained water.
SECT, vj See/ and Scyld 69
i lag ok draga upp segl, en leggja eld i tyrviS ok gera bal k skipinu ; veSr st63 af landi ; Haki var ])6. at kominn dautJa eSa dauSr, er hann var lagiSr 4 bdlit; sigltJi skipit siSan loganda ut i haf, ok var t)etta allfrsegt lengi sfSan.
(King Haki was so sore wounded that he saw that his days could not be long. Then he had a warship of his taken, and loaded with dead men and weapons, had it carried out to sea, the rudder shipped, the sail drawn up, the fir-tree wood set alight, and a bale-fire made on the ship. The wind blew from the land. Haki was dead or nearly dead, when he was placed on the pyre. Then the ship sailed blazing out to sea; and that was widely famous for a long time after.)
Ynglinga Saga, Kap. 23, in Heimskringla, udg. af Finnur J6nsson, K0benhavn, 1893, vol. i, p. 43.
The Skjoldunga Saga gives a story which is obviously connected with this. King Sigurd Ring in his old age asked in marriage the lady Alfsola; but her brothers scorned to give her to an aged man. War followed; and the brothers, knowing that they could not withstand the hosts of Sigm-d, poisoned their sister before marching against him. In the battle the brothers were slain, and Sigurd badly wounded.
Qui, Alfsola funere allato, magnam navim mortuorum cadaveribus oneratam solus vivorum conscendit, seque et mortuam Alfsolam in puppi collocans navim pice, bitumine et sulphure incendi jubet: atque sublatis velis in altum, validis a continente impellentibus ventis, proram dirigit, simulque manus sibi violentas intuHt; sesc.more majorum suorum regali pompa Odinum regem (id est inferos) invisere malle, quam inertis senectutis infirmitatem perpeti
Skjoldungasaga i Arngrim Jonssons udtog, udgiven af Axel Olrik, Kj0benhavn, 1894, Cap. xxvn, p. 50 [132].
So with the death of Baldr.
En sesirnir toku lik Baldrs ok fluttu til sasvar. Hringhorni het skip Baldrs ; hann var allra skipa mestr, hann vildu gotSin framm setja ok gera I^ar a balfgr Baldrs... p4 var borit tit 4 skipit lik Baldrs,... OSinn lagSi 4 b41it gullhring ]?ann, er Draupnir heitir...hestr Baldrs var leiddr a b4Ut meS qIIu reiSi.
(But the gods took the body of Baldr and carried it to the sea-shore. Baldr' s ship was named Hringhorni: it was the greatest of all ships and the gods sought to launch it, and to build the pyre of Baldr on it.... Then was the body of Baldr borne out on to the ship.... Odin laid on the pyre the gold ring named Draupnir... and Baldr's horse with all his trappings was placed on the pyre.)
Snorra Edda : Oylfaginning, 48 ; udg. af Finnur J6nsson, K0ben- havn, 1900.
We are justified in rendering setja skip f ram by "launch": Olrik (HeUedigfning, I, 250) regards Baldr's funeral as a case of the burning of a body in a ship on land. But it seems to me, as to Mr Chadwick (Origin, 287), that the natural meaning is that the ship was launched in the sea.
But the case of Scyld is not exactly parallel to these. The ship which conveyed Scyld out to sea was not set alight. And the words of the poet, though dark, seem to imply that it was intended to come to land somewhere: "None could say who / received that freight."
70 8cef and Scyld [ch. ii
Further, Scyld not merely departed over the waves — he had in the first instance come over them : " Not with less treasure did they adorn him," says the poet, speaking of the funeral rites, "than did those who at the beginning sent him forth alone over the waves, being yet a child."
Scyld Scefing then, like Tennyson's Arthur, comes from the unknown and departs back to it.
The story of the mysterious coming over the water was not confined to Scyld. It meets us in connection with King Scef, who was regarded, at any rate from the time of Alfred, and possibly much earlier, as the remotest ancestor of the Wessex \ kings. Ethelwerd, a member of the West Saxon royal house, who compiled a bombastic Latin chronicle towards the end of the tenth century, traces back the pedigree of the kings of Wessex to Scyld and his father Scef. " This Scef," he says, " came to land on a swift boat, surrounded by arms, in an island of the ocean called Scani, when a very young child. He was unknown to the people of that land, but was adopted by them as if of their kin, well cared for, and afterwards elected king^." Note here, firstly, that the story is told, not of Scyld Scefing, but of Scef, father of Scyld. Secondly, that although Ethelwerd is speaking of the ancestor of the West Saxon royal house, he makes him come to land and rule, not in the ancient homeland of continental Angeln, but in the "island of Scani," which signifies what is now the south of Sweden, and perhaps also the Danish islands^ — that same land of Scedenig which is men- tioned in Beowulf as the realm of Scyld. The tone of the ^narrative is, so far as we can judge from Ethelwerd's dry summary, entirely warHke : Scef is surrounded by weapons.
In the twelfth century the story is again told by William of Malmesbury . " Sceldius was the son of Sceaf . He, they say, was carried as a small boy in a boat without any oarsman to a certain isle of Germany called Scandza, concerning which
1 Ipse Scef cum uno dromone advectus est in insula Oceani, quae dicitur Scani, armis circundatus, eratque valde recens puer, & ab mcolis illius terrae ignotus; attamen ab eis suscipitur, & ut familiarem diligenti animo eum custodierunt, & post in regem eligunt.
Ethelwerdus, in, 3, in Savile's Rerum Anglicarum Scriptores post Bedarrif Francofurti, 1601, p. 842.
a See Chadwick, Origin, 259-60
SECT, v] Scef and Scyld 71
Jordanes, the liistorian of the Goths, speaks. He was sleeping, and a handful of corn was placed at his head, from which he was called * Sheaf.' He was regarded as a wonder by the folk of that country and carefully nurtured; when grown up he ruled in a town then called Slaswic, and now Haithebi — that region is called ancient Angha^."
William of Malmesbury was, of course, aware of Ethelwerd's account, and may have been influenced by it. Some of his variations may be his own invention. The substitution of the classical form Scandza for Ethelwerd's Scani is simply a change from popular to learned nomenclature, and enables the historian to show that he has read something of Jordanes. The altera- tion by which Malmesbury makes Sceaf, when grown up, rule at Schleswig in ancient Angel, may again be his own work — a variant added in order to make Sceaf look more at home in an Anglo-Saxon pedigree.
But WilUam of Malmesbury was, as we shall see later, prone to incorporate current ballads into his history, and after allowing for what he may have derived from Ethelwerd, and what he may have invented, there can be no doubt that many of the additional details which he gives are genuine popular poetry. Indeed, whilst the story of Scyld's funeral is very impressive in Beowulf, it is in WilHam's narrative that the story of the child coming over the sea first becomes poetic.
Now since even the English historians connected this tale with the Danish territory of Scani, Scandza, we should expect to find it again on turning to the records of the Danish royal house. And we do find there, generally at the head of the pedigree^, a hero — Skjold — whose name corresponds, and whose relationship to the later Danish kings shows him to be the same as the Scyld Scefing of Beowulf. But neither Saxo Gram- maticus, nor any other Danish historian, knows anything of
^ Sceldius [fuit films] Sceaf. Iste, ut ferunt, in quandam insulam Germaniae Scandzam, de qua Jordanes, historiographus Gothorum, loquitur, appulsua navi sine remige, puerulus, posito ad caput frumenti manipulo, dormiens, ideoque Sceaf nuncupatus, ab hominibus regionis illius pro miraculo exceptus et sedulo nutritus: adulta aetate regnavit in oppido quod tunc Slaswic, nunc vero Haithebi appellatur. Est autem regio ilia Anglia vetus dicta....
William of Malmesbury, De Oestis Regum Anglorum. Lib. ii, § 116, vol. i, p. 121, ed. Stubbs, 1887.
2 Although Saxo Grammaticus has provided some even earUer kings.
72 Scef and Scyld [ch. ii
Skjold having come in his youth or returned in his death over the ocean.
How are we to harmonize these accounts?
Beowulf and Ethelwerd agree in representing the hero as " surrounded by arms " ; Wilham of Malmesbury mentions only the sheaf ; the difference is weighty, for presumably the spoils which the hero brings with him from the unknown, or takes back thither, are in harmony with his career. Beowulf and Ethelwerd seem to show the warrior king, William of Malmes- bury seems rather to be telling the story of a semi-divine foundhng, who introduces the tillage of the earth^.
In Beowulf the child is Scyld Scefing, in Ethelwerd and William of Malmesbury he is Sceaf, father of Scyld.
Beowulf, Ethelwerd and Wilham of Malmesbury agree in connecting the story with Scedenig, Scant or Scandza, yet the two historians and the Anglo-Saxon Chronicle all make Sceaf the ancestor of the West Saxon house. Yet we have no evidence that the Enghsh were regarded as having come from Scandinavia.
The last problem admits of easy solution. In heathen times the English traced the pedigree of most of their kings to Woden, and stopped there. For higher than that they could not go. But a Christian poet or genealogist, who had no belief in Woden as a god, would regard the All Father as a man — a mere man who, by magic powers, had made the heathen believe he was a god. To such a Christian pedigree- maker Woden would convey no idea of finality; he would feel no difficulty in giving this human Woden any number of ancestors. Wishing to glorify the pedigree of his king, he would add any other distinguished and authentic genealogies, and the obvious place for these would be at the end of the Hue, i.e., above Woden. Hence we have in some quite early (not West Saxon) pedigrees, five names given as ancestors of Woden. These five names end in Geat or Geata, who was apparently regarded as a god, and was possibly Woden under another name^. Somewhat later, in the Anglo-Saxon Chronicle, under
1 Cf. MuUenhoff in Z.f.d.A. vii, 413.
2 In Orimnismdl, 54, Odin gives Oautr as one of his names.
4
SECT, v] Scef and Scyld 73
the year 855, we have a long version of the West Saxon pedigree with yet nine further names above Geat, ending in Sceaf. Sceaf is described as a son of Noah, and so the pedigree is carried back to Adam, 25 generations in all beyond Woden^. But it is rash to assume with Miillenhoff that, because Sceaf comes at the head^ of this English pedigree, Sceaf was therefore essentially an Enghsh hero. All these later stages above Woden look like the ornate additions of a later compiler. Some of the figures, Finn, Sceldwa, Heremod, Sceaf himself, we have reason to identify with the primitive heroes of other nations.
The genealogist who finally made Sceaf into a son born to Noah in the ark, and then carried the pedigree nine stages further back through Noah to Adam, merely made the last of a series of accretions. It does not follow that, because he made them ancestors of the English king, this compiler regarded Noah, Enoch and Adam as Enghshmen. Neither need he have so regarded Sceaf or Scyld ^ or Beaw. In fact — and this has constantly been overlooked — the authority for Sceaf, Scyld and Beaw as Anglo-Saxon heroes is but little stronger than the authority for Noah and Adam in that capacity. No manuscript exists which stops at Scyld or Sceaf. There is no version which goes beyond Geat except that which goes up to Adam. Scyld, Beaw, Sceaf, Noah and Adam as heroes of English mythology are all aHke doubtful.
We must be careful, however, to define what we mean when we regard these stages of the pedigree as doubtful. They are doubtful in ^o far as they are represented as standing above Woden in the Anglo-Saxon pedigree, because it is in- credible that, in primitive and heathen times, Woden was credited with a dozen or more forefathers. The position of these names in the pedigree is therefore doubtful. But it is only their connection with the West Saxon house that is un- authentic. It does not follow that the names are, fer se, unauthentic. On the contrary, it is because the genealogist had such implicit belief in the authenticity of the generations
^ See below. 2 Excluding, of course, the Hebrew names.
* Scyld appears as Scyldwa, Sce{a)ldwa in the Chronicle. The forms correspond.
74 Scef and Scyld [ch. ii
from Noali to Adam that he could not rest satisfied with his West Saxon pedigree till he had incorporated thfese names. They are not West Saxon, but they are part of a tradition much more ancient than any pedigree of the West Saxon kings. And the argument which applies to the layer of Hebrew names between Noah and Adam applies equally to the layer of Ger- manic names between Woden and Sceaf. From whatever branch of the Germanic race the genealogist may have taken them, the fact that he placed them where he did in the pedigree is a proof of his veneration for them. But we must not without evidence claim them as West Saxon or Anglo-Saxon : we must not be surprised if evidence points to some of them being con- nected with other nations — as Heremod, for example, with the Danes^.
More difficult are the other problems. William of Malmes- bury tells the story of Sceaf, with the attributes of a culture- hero : Beowulf, four centuries earlier, tells it of Scyld, a warrior hero: Ethelwerd tells it of Sceaf, but gives him the warrior attributes of Scyld^ instead of the sheaf of corn.
The earlier scholars mostly agreed^ in regarding Malmes- bury's attribution of the story to Sceaf as the original and correct version of the story, in spite of its late date. As a representative of these early scholars we may take Miillenhoff*. MiillenhofE's love of mythological interpretation found ample scope in the story of the child with the sheaf, which he, with considerable reason, regarded as a " culture-myth." Miillenhoff beUeved the carrying over of the attributes of a god to a line of his supposed descendants to be a common feature of myth — the descendants representing the god under another name. In accordance with this view, Scyld could be explained as an "hypostasis" of his father or forefather Sceaf, as a figure further explaining him and representing him, so that in the end the tale of the boat arrival came to be told, in Beowulf, of Scyld instead of Sceaf.
^ See Part II. ^ armis circundatus. ^
3 For a list of the scholars who have dealt with the subject, see Widsith p. 119.
* Beovulf, p. Q etc.
*
SECT, v] Seef and Scijld 7o
Kecent years have seen a revolt against most of Miillenhoiffi's theories. The view that the story originally belonged to Sceaf has come to be regarded with a certain amount of impatience as "out of date." Even so fine a scholar as Dr Lawrence has expressed this impatience :
"That the graceful story of the boy sailing in an open boat to the land of his future people was told originally of Sceaf. . .needs no detailed refutation at the present day.
"The attachment of the motive to Sceaf must be, as an examination of the sources shows, a later development^."
Accordingly the view of recent scholars has been this: That the story belongs essentially to Scyld. That, as the hero of the boat story is obviously of unknown parentage, we must interpret Scefing not as "son of Sceaf" but as "with the sheaf" (in itself a quite possible explanation). That this stage of the story is preserved in Beowulf. That subsequently Scyld Scefing, standing at the head of the pedigree, came to be mis- understood as "Scyld, son of Sceaf." That consequently the story, which must be told of the earlier ancestor, was thus transferred from Scyld to his supposed father Sceaf — the version which is found in Ethelwerd and William of Malmesbury.
One apparent advantage of this theory is that the oldest version, that of Beowulf, is accepted as the correct and original one, and the much later versions of the historians Ethelwerd and William of Malmesbury are regarded as subsequent cor- ruptions. This on the surface seems eminently reasonable. But let us look closer. Scyld Scefing in Beowulf is to be in- terpreted ''Scyld with the Sheaf." But Beowulf nowhere mentions the sheaf as part of Scyld's equipment. On the contrary, we gather that the hero is connected rather with prowess in war. It is the same in Ethelwerd. It is not till WiUiam of Malmesbury that the sheaf comes into the story. So that the interpretation of Scefing as "with the sheaf" assumes the accuracy of William of Malmesbury's story even in a point where it receives no support from the Beowulf version. In other words this theory does the very thing to avoid doing which it was called into being^,
^ Pub. Mod. Lang. Assoc. Amer. xxiv, 259 etc.
^ This objection to the Scyld-theory has been excellently expressed by Olrik — at a time, too, when Olrik himself accepted the story as belonging to Scyld
76 See/ and Scyld [ch. ii
Besides this, there are two fundamental objections to the theory that Sceaf is a late creation, a figure formed from the misunderstanding of the epithet Scefing applied to Scyld, One portion of the poem of Widsith consists of a catalogue of ancient kings, and among these occurs Sceaf a, ruling the Lango- bards. Now portions of Widsith are very ancient, and this catalogue in which Sceafa occurs is almost certainly appreciably older than Beowulf itself.
Secondly, the story of the wonderful foundling who comes over the sea from the unknown and founds a royal line, must ex hypothesi be told of the first in the line, and we have seen that it is Sceaf, not Scyld, who comes at the head of the Teutonic names in the genealogy in the Anglo-Saxon Chronicle.
Now we can date this genealogy fairly exactly. It occurs under the year 855, and seems to have been drawn up at the court of King ^Ethelwulf . In any case it cannot be later than the latter part of Alfred's reign. This takes us back to a period when the old English epic was still widely popular. A genealogist at Alfred's court must have known much about Old English story .
These facts are simply not consistent with the belief that Sceaf is a late creation, a figure formed from a misunderstanding of the epithet Scefing, applied to Scyld^.
rather than Sceaf. "Binz," says Olrik, "rejects William of Malmesbury as a source for the Scyld story. But he has not noticed that in doing so he saws across the branch upon which he himself and the other investigators are sitting. For if William is not a reliable authority, and even a more reliable authority than the others, then 'Scyld with the sheaf is left in the air." Heltedigtning, I, 238-9, note.
^ The discussion of Skjold by Olrik {Danmarks Heltedigtning, i, 223-271) is perhaps the most helpful of any yet made, especially in emphasizing the necessity of differentiating the stages in the story. But it must be taken in connection with the very essential modifications made by Dr Olrik in his second volume (pp. 249-65, especially pp. 264-5). Dr Olrik's earlier interpretation made Scyld the original hero of the story : Scefing Olrik interpreted, not as "with the sheaf," but as "son of Scef." To the objection that any knowledge of Scyld' s parentage would be inconsistent with his unknown origin, Olrik replied by supposing that Scyld was a foundling whose origin, though unknown to the people of the land to which he came, was well known to the poet. The poet, Dr Olrik thought, regarded him as a son of the fcangobardic king, Sceafa, a connection which we are to attribute to the Anglo-Saxon love of framing genealogies. But this explanation of Scyld Scefing as a human foundling doei not seem to me to be borne out by the text of Beowulf. "The child is a pool foundling," says Dr Olrik, "Ae suffered distress from the time lohen he was firs found as a helpless child. Only as a grown man did he get compensation foi his childhood's adversity" (p. 228). But this is certainly not the meaning o: egsode eorl[as]. It is ''He inspired the earl[s] with awe."
SECT, v] Scef and Scyld 77
To arrive at any definite conclusion is difficult. But the following may be hazarded.
It may be taken as proved that the Scyld or Sceldwa of the genealogists is identical with the Scyld Scefing of Beowulf. For Sceldwa according to the genealogy is also ultimately a Sceafing, and is the father of Beow ; Scyld is Scefing and is father of Beowulf^.
It is equally clear that the Scyld Scefing of Beowulf is identical with the Skjold of the Danish genealogists and historians. For Scyld and Skjold are both represented as the founder and head of the Danish royal house of Scyldingas or Skjoldungar, and as reigning in the same district. Here, however, the resemblance ceases. Beowulf tells us of Scyld's marvellous coming and departure. The only Danish authority who tells us much of Skjold is Saxo Grammaticus, who records how as a boy Skjold wrestled successfully with a bear and over- came champions, and how later he annulled unrighteous laws, and distinguished himself by generosity to his court. But the Danish and EngKsh accounts have nothing specifically in common, though the type they portray is the same — that of a king from his youth beloved by his retainers and feared by neighbouring peoples, whom he subdues and makes tributary. It looks rather as if the oldest traditions had had little to say about this hero beyond the typical things which might be said of any great king ; so that Danes and English had each supplied the deficiency in their own way.
Now this is exactly what we should expect. For Scyld- I Skjold is hardly a personality: he is a figure evolved out of the name Scyldingas, Skjoldungar, which is an old epic title for the Danes. Of this we may be fairly certain : the Scyldingas did not get their name because they were really descended from Scyld, but Scyld was created in order to provide an eponymous father to the Scyldingas^. In just the same way
^ See below (App. C) for instances of ancestral names extant both in weak and strong forms, like Scyld, Sceldwa (the identity of which no one doubts) or Sceaf, Sceaja (the identity of which has been doubted).
^ "As for the name Scyldungas-Skjoldungar, we need not hesitate to believe that this originally meant 'the people' or 'Mnsmen of the shield.' Similar appellations are not uncommon, e.g., Rondingas, Helmingas, Brondingas...
78 Beef and Scyld [ch. ii
tradition also evolved a hero Dan, from whom the Danes were supposed to have their name. Saxo Grammaticus has com- bined both pedigrees, making Skjold a descendant of Dan; but usually it was agreed that nothing came before Skjold, that he was the beginning of the Skjoldung line^. At first a mere name, we should expect that he would have no character- istic save that, like every respectable Germanic king, he took tribute from his foes and gave it to his friends. He differs therefore from those heroic figures like Hygelac or Guthhere (Gunnar) which, being derived from actual historic characters, have, from the beginning of their story, certain definite features attached to them. Scyld is, in the beginning, merely a name, the ancestor of the Scyldings. Tradition collects round him gradually.
Hence it will be rash to attach much weight to any feature which is found in one account of him only. Anything we are V)ld of Scyld in English sources alone is not to be construed as evidence as to his original story, but only as to the form that story assumed in England. When, for example, Beoivulf tells us that Scyld is Scefing, or that he is father of Beowulf, it will be very rash of us to assume that these relationships existed in the Danish, but have been forgotten. This is, I think, univer- sally admitted^. Yet the very scholars who emphasize this, have assumed that the marvellous arrival as a child, in a boat, surrounded by weapons, is an essential feature of Scyld's story. Yet the evidence for this is no better and no worse than the evidence for his relationship to Sceaf or Beow — it rests solely on the English documents. Accordingly it only shows what was told about Scyld in England.
Of course the boat arrival might be an original part of the story of Scyld-SJcjold, which has been forgotten in his native
probably these names meant either 'the people of the shield, the helmet,' etc., or else the people who used shields, helmets, etc., in some special way. In the former case we may compare the Ancile of the Romans and the Palladion of the Greeks; in either case we may note that occasionally shields have been fomid in the North which can never have been used except for ceremonial] purposes." Chad wick, Origin, p. 284: cf. Olrik, Heltedigtning, i, 274.
^ Sweyn Aageson, Shiold Danis primum didici praefuisse, in Langebek, 8.B.D. I, 44.
2 Olrik, Heltedigtning, i, 246; Lawrence, Pub. Mod. Lang. Assoc, xxiv, 254.
SECT, v] Beef and Scyld 79
country, but remembered in England. But I cannot see that we have any right to assert this, without proof.
What we can assert to have been the original feature of Scyld is this — that he was the eponymous hero king of the Danes. Both Beowulf and the Scandinavian authorities agree upon that. The fact that his name (in the form Sceldwa) appears in the genealogy of the kings of Wessex is not evidence against a Danish origin. The name appears in close connection with that of Heremod, another Danish king, and is merely evidence of a desire on the part of the genealogist of the Wessex kings to connect his royal house with the most distinguished family he knew : that of the Scyldingas, about whom so much is said in the prologue to Beowulf.
Neither do the instances of place-names in England, such as Scyldes treow, Scildes well, prove Scyld to have been an English hero. They merely prove him to have been a hero I who was celebrated in England — which the Prologue to Beowulf alone is sufficient to show to have been the case. For place- names commemorating heroes of alien tribes are common enough^ on English ground.
So much at least is gained. Whatever Miillenhoff^ and his followers constructed upon the assumption that Scyld was an essentially Anglo-Saxon hero goes overboard. Scyld is the ancestor king of the Danish house — more than this we can hardly with safety assert.
Now let us turn to the figure of Sceaf. This was not necessarily connected with Scyld from the first.
The story of Sceaf first meets us in its completeness in the pages of William of Malmesbury. And William of Malmesbury is a twelfth century authority; by his time the Old English courtly epics had died out — for they could not have long survived the Norman Conquest and the overthrow of Old Enghsh court life. But the popular tradition^ remained, and
^ It is odd that Binz, who has recorded so many of these, should have argued on the strength of these place-names that the Scyld story is not Danish, but an ancient possession of the tribes of the North Sea coast (p. 150). For Binz also records an immense number of names of heroes of alien stock — Danish, Gothic or Burgundian — as occurring in England {P.B.B. xx, 202 etc.).
2 Beovulf, p. 7. ' Chad wick. Origin, p. 278,
80 Scef and Scyld [CH. ii
a good many of the old stories, banished from the hall, must have lingered on at the cross-roads — tales of Wade and Wey- land, of Offa and Sceaf. For songs, sung by minstrels at the cross roads, William of Malmesbury is good evidence, and he owns to having drawn information from similar popular sources^. William's story, then, is evidence that in his own day there was a tradition of a mythical king Sheaf who came as a child sleeping in a ship with a sheaf of corn at his head. How old this tradition may be, we cannot say. Ethelwerd knew the story, though he has nothing to say of the sheaf. But we have seen that when we get back to the ninth century, and the formation of the Anglo-Saxon Chronicle, at a court where we may be sure the old English heroic stories were still popular, it is Sceaf and not Sceldwa who is regarded as the beginning of things — the king whose origin is so remote that he is the oldest Germanic ancestor one can get back to^ : " he was born in Noah's ark."
Whether or no Noah's ark was chosen as Sceaf 's birthplace because legend represented him as coming in