SCYLD SCEFING AND THE DATING OF BEOWULF AGAIN*
AUDREY L. MEANEY
SCHOOL OF ENGLISH, MACQUARIE UNIVERSITY
The dates which have been suggested and accepted by reputable
scholars for the composition of Beowulf vary from the seventh century1
- an old idea and now largely discounted - right up to the early
eleventh: that is, composition by the scribes who wrote the poem into
the extant manuscript, the Nowell Codex, as late as the reign of Cnut2
- a new idea and very controversial. One of the latest publications on
the topic is Rowland L. Collins' discussion of the 'long-recognized
relationship' between the description of Grendel's mere and the
version of St Paul's vision of hell found at- the end of the sixteenth
Blickling Homily. 3 Collins demonstrated the similarity of their 'language (the particular vernacular words)' and argued that the two
passages must have been directly connected. However, the 'descriptive elements of Blickling XVI are much closer to one of the redactions
of the Visio Pauli . . . than they are to Beowulf: they share 'the dark
water, the deep pit, the trees ... on a precipice from which
unrepentant sinners are hanging; . . . indeed the only striking
divergence between Blickling XVI and the Visio is the temperature of
the sinners; in the Visio they burn, while presumably they freeze in
. . . Blickling XVI.' Moreover, the homilist cites only St Paul as his
source; and whereas in the homily the description is compacted into
* The T. Northcote Toller Memorial Lecture delivered in the John Rylands University Library
of Manchester on Monday, 14 March 1988. In preparing this paper for publication, I gratefully
acknowledge the help of Donald Scragg.
1 Ritchie Girvan, Beowulf and the Seventh Century: Language and Content (London, 1st
ed.1935; repr.1971); see also Patrick Wormald, 'Bede, Beowulf and the Conversion of the
Anglo-Saxon Aristocracy', Bede and Anglo-Saxon England, ed. Robert T. Farrell (British
Archaeological Reports, 46, Oxford, 1978), 32-95.
2 Kevin S. Kiernan, Beowulf and the Beowulf Manuscript (New Brunswick, 1981), and The
Eleventh-Century Origin of Beowulf and the Beowulf Manuscript', fn The Dating of Beowulf, ed.
Colin Chase (Toronto, 1981), 9-22. For alternative dates, see e.g. other articles in the Chase
collection and W.G. Busse and R. Holtei, "Beowulf and the Tenth Century', Bulletin of the John
Rylands University Library of Manchester, 63 (1981), 285-329. A useful, sceptical review of the
latest ideas and arguments which concludes 'Die Datierung des Beowulf bleibt eine offene Frage'
is Claus-Dieter Wetzel's 'Die Datierung des Beowulf: Bemerkungen zur jungsten Forschungsentwicklung', Anglia, 103 (1985), 371-tt)0.
3 Medieval Studies (Conference, Aachen, 1983: Language and Literature, ed. Wolf-Dietrich Bald
and Horst Weinstock (Bamberger Beitrage zur Englischen Sprachwissenschaft, 15, Frankfurtam-Mam, 1984), 61-9.
8
BULLETIN JOHN RYLANDS LIBRARY
'three lines of vivid prose' (without any sign of poetic rhythm) from
which borrowing would have been easy, in Beowulf 'phrases and
images' are scattered over more than sixty verse lines. Clearly,
therefore, the influence is towards Beowulf, not away from it.
Ker dated the manuscripts of the Blickling Homilies and of
Beowulf both to around 1000;4 but Collins accepted the date of the
Blickling Codex as 971 (which is given in the text of Homily XI),5 so
that in his view it would have been chronologically possible for this
memorable passage in Homily XVI to have influenced the extant
version of Beowulf.
However, the influence need not have been so late. 6 Blickling
XVI could have had an earlier independent existence. 7 Moreover, the
Visio passage itself could have had a former independent existence;
Collins described it as 'seemingly tacked-on' to the homily - 'just
when the homilist seems to be ending, there is a distinct diversion.'
Some of the vocabulary in the Visio passage, in particular two terms
which Collins picked out as being more securely part of the text of
Homily XVI than of Beowulf* is crucial here. The phrase hrimige
bearwas is varied by hrimige wuda elsewhere in the homily; it
corresponds in Beowulf to hrinde bearwas, found only in line 1363.
Both adjectives are rare, but hrinde is the rarer; therefore (according to
the normal rules of textual criticism), it is the more likely to be
original. Secondly, there is the compound pystrogenipu, 'dark mists'
which appears twice in Homily XVI, but only there. 9 The corresponding phrase in Beowulf is neessa genipu, 'the mists of the cliffs'. 10 Could
the compound have been the homilist's own coinage? At the least, it
would have been in his mind because of its use earlier in the homily,
and therefore both hrimige and fcystrogenipu could have been his own
4 Neil R. Ker, Catalogue of Manuscripts Containing Anglo-Saxon (London, 1957), nos.216 and
382.
5 On the grounds that some configurations 'of hair and flesh sides in the gatherings ... in the
Blickling codex . . seem to be related to . . an older insular pattern'; Rowland L. Collins,
'Blickling XVI', Medieval Studies, ed. Bald and Weinstock, 67-8. If Kiernan's date for the
Nowell Codex (see n.2) is correct, the Blickling manuscript could be as much as forty years the
elder.
6 However, Collins conceded (ibid., 68 n.4) that the Visio passage could have influenced
Beowulf 'through some now unknown source . . . Discovery of such a document could always
challenge the chronology and influence suggested.'
7 For the compilation of the Blickling Homiliary, see Milton McC. Gatch, 'Eschatology in the
Anonymous Old English Homilies', Traditio, 21 (1965), 117-65, at 117-36; Donald G. Scragg,
The Homilies of the Blickling Manuscript', Learning and Literature in Anglo-Saxon England:
Studies presented to Peter Clemoes, ed. Michael Lapidge and Helmut Gneuss (Cambridge, 1985),
299-316.
8 Collins, 'Blickling XVI', 65-6.
9 The Blickling Homilies, ed. R. Morris (EETS, OS, 58, 63 and 73, Oxford, 1874-80; 1967
repr.), 209 line 33 (also hrimige bearwas), and 203 line 8 respectively. See also Rowland L.
Collins, 'Six Words in the Blickling Homilies', Philological Essays . . . in Honour of Herbert Dean
Meritt, ed. James L. Rosier (The Hague, 1970), 137^1, at 141.
10 Line 1360; at line 2808 Beowulf has flodagenipu. The most frequently cited edition is that by
F. Klaeber, Beowulf and the Fight at Finnsburg (3rd ed., Boston, 1941).
9
DATING OF BEOWULF
replacements for other terms (such as Beowulf s hrinde and ruessa
genipu) in an earlier version of the Visio passage. The influence of the
Visio Pauli on Beowulf could therefore have antedated its incorporation into the homily, or the homily's incorporation into the
Blickling Homiliary; the date 971 as a terminus post quern is not firm.
Another passage of Beowulf which has importance in any attempt
to date the poem by means of the sources used is its so-called Prologue,
concerned with the pedigree of the Danish kings. For convenience of
reference, I quote here lines 4-19, 26-63; and (to anticipate the
following argument) have placed letters against the lines indicating
which sources (or natural extensions of those sources) I think may
have been used: A = the pre-ae Chronicle; B = knowledge of
ship-burial ritual; C = Christian comments by poet; D = Danish
traditions; G = Vita Gildae. A combination of letters means any or all
of the suggested sources may have been used; a slash between letters
indicates different sources for the two half-lines.
Oft Scyld Scefing sceabena breatum,
monegum maegbum meodosetla ofteah,
egsode eorlas, sy53an aerest wearS
feasceaft funden; he baes frofre gebad,
weox under wolcnum, weor8myndum bah,
oS baet him aeghwylc ymbsittendra
ofer hronrade hyran scolde,
gomban gyldan; bast waes god cyning!
Daem eafera waes aefter cenned
geong in geardum, bone God sende
folce to frofre; fyrenSearfe ongeat,
be hie aer drugon aldorlease
lange hwile; him baes Liffrea,
wuldres Wealdend woroldare forgeaf.
Beowulf waes breme - blaed wide sprang Scyldes eafera Scedelandum in. ...
Him 5a Scyld gewat to gescaephwile
felahror feran on Frean waere;
hi hine pa aetbaeron to brimes faroSe,
swaese gesibas, swa he selfa baed,
benden wordum weold wine Scyldinga leof landfruma lange ahte.
Paer aet hy5e stod hringedstefna
isig ond utfus, aebelinges faer;
aledon ba leofne beoden
beaga bryttan on bearm scipes,
maerne be maeste. I»aer waes madma fela
of feorwegum fraetwa gelaeded;
ne hyrde ic cymlicor ceol gegyrwan
hildewaepnum ond heaSowaedum,
billum and byrnum; him on bearme laeg
madma maenigo, ba him mid scoldon
on flodes aeht feor gewitan.
Nalaes hi hine laessan lacum teodan,
beodgtstreonum, bon ba dydon,
A/D
D
D/A
A
A
D
D
D/DA
AD
D/C
D
DC
D/C
C/D
A/D
DA
ADG
ADG/G
G
G
AGO
D
/B
?G/B
G/DA
/B
B
B
B
B
B
BBG
G
AB
AB
10
BULLETIN JOHN RYLANDS LIBRARY
be hine aet frumsceafte forS onsendon
aenne ofer y5e umborwesende.
Pa gyt hie him asetton segen gyldenne
heah ofer heafod, leton holm her an,
geafon on garsecg; him waes geomor sefa,
murnende mod. Men ne cunnon
secgan to soSe, seleraedende,
haeleS under heofenum, hwa baem hlaeste onfeng.
Da waes on burgum Beowulf Scyldinga,
leof leodcyning longe brage
folcum gefraege - feeder ellor hwearf,
aldor of earde - ob baet him eft onwoc
heah Healfdene; heold benden lifde
gamol ond guSreouw glaede Scyldingas.
Daem feower beam forSgerimed
in worold wocun, weoroda raeswan,
Heorogar ond Hrobgar ond Halga til;
hyrde ic baet [. . . waes On]elan cwen,
HeaSo- Scilfingas healsgebedda. 11
AB
A
?B
?B/G
G
G/C
C
C
AD
D
D
D
D
D
D
D
D
D
D
In this passage, then, the genealogy runs, in descending order, Scyld
Scefing - Beowulf - Healfdene - Heorogar/Hrothgar/Halga; and the
term 'Scyldingas' is used either for the Danes in general ('he ruled the
gracious Scyldings') or for their kings, the descendants of Scyld
(Beowulf Scyldinga). Presumably it originally meant 'shield-men',
referring to the Danes' character as defenders of their homeland. Scyld
is an eponym, a personification regarded as an ultimate ancestor, 12 and
11 'Scyld Scefing often took meadbenches away from many peoples, put fear into the nobles,
after he was first found destitute. He experienced consolation for that; grew under the clouds;
flourished in honours, until each of his neighbours over the sea had to obey, to pay tribute; he
was a good king.
To him an heir was afterwards born, a young one in the courts, whom God sent as a comfort to
the people, understanding the dire distress which they had suffered earlier, for a long while,
when without a lord. The Lord of life, Ruler of Glory, gave him honour in the world; in Skane
Beowulf, Scyld's heir, was famous - his glory spread widely . . . Scyld then departed from him at
the fated time, to go, very vigorous, into the Lord's protection; dear companions carried him
then to the waves of the sea as he, lord of the Scyldings, bade, while he ruled with words - the
beloved lord of the land had power for a long time. There at the harbour the prince's ring-prowed
vessel stood icy and ready to depart. Then they laid their dear illustrious lord, giver of rings,
amidships, by the mast. There were many treasures and ornaments brought from far away; nor
have I heard of a ship more splendidly adorned with weapons of war and battledress, with swords
and coats of mail; amidships lay a multitude of treasures, which should depart far away with it,
into the power of the sea. They did not at all provide him with lesser gifts, people's treasures,
than those who sent him forth at the beginning, alone over the waves when he was a child. Then
they placed a gold standard over his head, let the sea carry him, gave him to the ocean. Their
spirits were sad, their minds sorrowing. Men, dwellers in hall, heroes under the heavens, did not
know how to say truly who received that freight.
Then Beowulf of the Scyldings, the beloved king of the land, renowned among the peoples,
was in the fortress for a long time - his father the prince had gone elsewhere, out of the land until the exalted Halfdane was born to him; he ruled while he lived (old and fierce in battle) the
gracious Scyldings. To him, leader of the troops, four children (all told) were born into the
world: Heorogar and Hrothgar and the good Halga. I heard that . . . was Onela's queen, the
consort of the Heatho-Scilfing.'
12 'Eponymi erhalten den Namen und die mythische Existenz erst durch eine Abstraktion''
Ernst A. Philippson, Die Genealogie der Gotter (Illinois Studies in Language and Literature 37
no.3, Urbana, 111., 1953), 13.
DATING OF BEOWULF
11
is to be equated with Scioldus/Skjoldr in the Scandinavian sources. 13
Sven Aggesen, writing in the mid-1180s, makes Skjold's character
very clear:
Skjold Danis primum didici prefuisse. Et ut eius alludamus uocabulo, idcirco tali
functus est nomine, quia uniuersos regni terminos regie defensionis patrocino affatim
egregie tuebatur. A quo primum, modis Islandensibus, 'Skioldunger' sum reges
nuncupati. Qui regni post se reliquit heredes, Frothi uidelicet et Haldanum. 14
For Saxo Grammaticus (a younger contemporary) Scioldus was preceded by another eponym, Dan, and his sons Humblus and Lotherus,
though he gave his name to the dynasty. Saxo was probably conflating
two traditions. His Scioldus is a hero-king, defeating a huge bear
without weapons as a youth, growing to fine manhood by the age of
fifteen and fighting boldly even before he had his full strength. He won
back the realm lost by his predecessor, and gave pensions and a share
of booty to his warriors. He was a great law-giver, caring for the sick
and needy, and reforming the debauched. 15
However, in his Rerum Danicarum Fragmenta - probably based
here on the lost Skjoldunga Saga of about 1200 - Arngrimr Jonsson
rejects Saxo's account, and claims Scioldus - from whom the Danes in
olden days were called 'Skiolldunga' - as the ultimate ancestor. 16 The
compiler of the Skjoldunga Saga appears to have strung together
Icelandic oral traditions about the ancient history of Denmark on to a
13 See discussion of Scyld in Danish tradition by Axel Olrik, Danmarks Heltedigtning (2 vols.,
Copenhagen, 1903-10), i.223-77; transl. Lee M. Hollander, Heroic Legends of Denmark (New
York, 1919), 381^45.
14 'Skjold is said to have been the first to have ruled the Danes. And as we have alluded to his
name, the reason that he was known by such a name was because he admirably guarded all
boundaries of the kingdom by the protection which his defence afforded. From whom kings were
first named "Skioldunger" in the Icelandic fashion. He left heirs after him to the kingdom,
namely Frothi and Haldan.'
Svenonis Aggonis Filii Brevis Historia Regum Dacie, ch.I, ed. M.C1. Gertz in Scriptores Minores
Historic Daniue Medii &vi, i (1st ed., Copenhagen, 1917-18), 96-7. Quoted by R.W.
Chambers, Beowulf: An Introduction (3rd ed., Cambridge, 1959), 211; translated in G.N.
Garmonsway and Jacqueline Simpson, Beowulf and its Analogues (London, 1968), 120.
15 Saxonis Gesta Danorum, 1.1-3, eds. Joergen Olrik and Hans Raeder (2 vols, Copenhagen,
1931), 10-12; Chambers, Beowulf, 129-30; Saxo Grammaticus, The History of the Danes, transl.
P Fisher, notes H.E. Davidson (2 vols., Cambridge, 1979), i. 14-15, ii. 25. Dan (son of Ypper
of Uppsala) is also the eponym in the Lejre Chronicle of the Most Ancient Kings of Denmark, I-III,
ed. Gertz, Scriptores, 43-6; his son was Ro (who built Roskilde); his grandsons Helgi and
Haldan; there is no mention of Skjold. The Lejre Chronicle may date from the 1170s; see
Chambers, Beowulf, 204, 216-17; Saxo Grammaticus, The History, ii.25. Skjold also appears in
the Danish King Lists, which are dependent on Saxo. All are edited in Gertz, Scriptores, 149-51,
161, 167, 175.
16 Arngnmi Jonae Opera Latine Conscripta, ed. Jakob Benediktsson (vols.i (text) and iv
(introduction and notes), Bibliotheca Arnamagnaeana, 9 and 12, Copenhagen, 1950-57), i.
148-9, 333-4; iv. 89, 107-17, 227. The relevant passage is translated in Garmonsway and
Simpson, Beowulf, 120. Snorri Sturluson also briefly mentions Skjold - from whom came the
family of the Skjoldungar - as son of Othin. In the Prologue to Snorri's Edda, Othin gave Skjold
Denmark to govern, and his son was FriSleif (ed. Anthony Faulkes, Oxford, 1982), 5-6; and in
the Yngltnga Saga, ch. 5 Skjold was husband of the goddess Gefjon, who created the island of
Sjaelland, on which they dwelt at Lejre (Hemsknngla, ed. Bjarni ASalbjarnarson (3 vols., Islenzk
Fornrit, 26-8, Reykjavik, 1941-51), i, 14-15).
12
BULLETIN JOHN RYLANDS LIBRARY
genealogy of the Scyldings, who were claimed as the ancestors of a
prominent Icelandic family, the Oddaverjar. It appears to have been
antiquarian speculation or fabrication, 17 however, which made Scioldus a son of Othin, who assigned Denmark to him and Sweden to his
other son Ingo, from whom the Swedes were called Inglings. After
Skioldus had subdued Jutland, the first of his conquests, he established himself at Hledra in Sjaelland.
For Sven Aggesen, Scioldus's son Frotho was quickly succeeded
by (that is, eliminated by) his 'brother' Halfdan. Saxo has two
generations (Gram and Hadingus) between Scioldus and Frotho, but it
may well be that he inserted them into the genealogy himself, and that
in one of his sources Frotho was also Scioldus' son. 18 Saxo has much to
say about Frotho as a martial king (but none looks like genuine
tradition), 19 and calls Halfdan his son. In the Skjoldunga Saga,
Scioldus' son is Leif (called Fridleif from his good peace) and his
grandson Frotho, who was succeeded by several other kings, including a Dan and three more Frothos, the last of whom had Halfdan as his
son. 20
The various Scandinavian versions of the Danish royal genealogy
more or less concur, therefore, in the three generations Skjold Frotho - Halfdan, equivalent to Scyld - Beowulf - Healfdene in
Beowulf. Skjold and Scyld Scefing are alike in being eponyms who
established (or re-established) the Danish kingdom by successful
warfare against their neighbours, restoring the fortunes of the
kingdom after an inglorious period.
Sven Aggesen and the Skjoldunga Saga agree that Skjold was the
first to rule Denmark, but though in Beowulf Scyld was the first of his
line, a certain Heremod must have ruled sometime before him. 21 For
Saxo and his successors Skjold is the son of Lother, whom Chambers
equates with Beowulf s Heremod.22 Though some of the characteristics
common to Skjold and Scyld Scefing may be commonplace, Beowulf's
17 Whether by the compiler or an earlier writer; see Einar Ol. Sveinsson, Sagnaritun
Oddaverjar (Islenzk FraeSi, i, Reykjavik, 1937), 15, 39^2 (English summary, 47-51), and Jakob
Benediktsson, 'Icelandic Traditions of the Scyldings', Saga-Book of the Viking Society, 15
(1957-61), 48-66. I am grateful to Margaret Clunies-Ross of Sydney University for these
references. Benediktsson questions whether Jonsson could have derived the description of Skjold
as Othin's son from the Codex Wormianus manuscript of Snorri's Edda, but considers it more
likely that Snorri got it from the Skjoldunga Saga; see Benediktsson, Arngrimijonae Opera, iv.
227.
18 Chambers, Beowulf, 92 n.l.
19 Saxo Grammaticus, The History, i. 37-8, 40-50.
20 The versions of the Danish^ pedigree discussed above give no sign of influence from the
English sources, whereas those which include Scialdun er ver kqllum Skjqld as ancestor of Othin
are clearly dependent on the West Saxon Genealogy in its later manifestations; see Snorri, Edda,
ed. Faulkes, 4-5; R.C. Boer, 'Studien iiber die Snorra Edda', Acta Philologica Scandinavica I
(1926-27), 54-150, at 141-5.
21 Beowulf, lines 901-15, 1709-22; ed. Klaeber (1941), notes 162-5.
22 Chambers, Beowulf, 89-91. In the OE Chronicle's genealogy s.a. 855 Sceldwea's father is
Heremod. See also David Williams, Cain and Beowulf: A Study in Secular Allegory (Toronto
1982), 75.
DATING OF BEOWULF
13
Prologue may therefore reflect some genuine Scandinavian traditions
about him.
However, none of these northern sources indicates that Skjold
came over the sea as a child, nor that he had a splendid ship-funeral.
The first of these elements is found in the genealogy s.a. 855 of the
Anglo-Saxon Chronicle in a Latin translation claimed as his own work
by ^thelweard, ealdorman of the south-western provinces at the end
of the tenth century. 23 I have argued in detail elsewhere that the
version of the Chronicle translated by ^thelweard occupied a position
midway between the earliest version evidenced - that used by the St
Neots compiler - and the common archetype (ae) of all the extant
Chronicles in Old English, since on the one hand it has the chronological dislocation St Neots lacks, but on the other various features which
appear more original than the Old English versions.24 One of these
early features is, indeed, that ^Ethelweard's genealogy s.a. 855 is
shorter than that in the Old English Chronicles. As a rule, shorter
genealogies are earlier than longer ones: once a prestigiously long set
of ancestors has been claimed, none of them is likely to be discarded
deliberately - only by accidental omission.
The most remote names in ^thelweard's version run (Geat) Tetuua - Beo - Scyld - Scef, and he then adds:
Ipse Scef cum uno dromone aduectus est in insula oceani que dicitur Scani, armis
circundatis, eratque ualde recens puer, et ab incolis illius terrae ignotus. Attamen ab
eis suscipitur, et ut familiarem diligent! animo eum custodierunt, et post in regem
eligunt; de cuius prosapia ordinem trahit Athulf rex. 25
No other early source has the two names Scyld and Scef together in the
same forms, as ^thelweard and Beowulf have;26 but this may be due
only to the coincidence that their manuscripts were both written about
AD 1000 or a little later. 27 Scef is a typically late West Saxon form;28
23 Alexander Callander Murray has discussed this relationship ('Beowulf, the Danish Invasions, and Royal Genealogy', Dating, ed. Chase, 101-11), but was concerned to establish a
historical context for the composition of the poem, rather than with specific details of texts, as I
am. Whether in fact yEthelweard translated the Chronicle himself or merely supervised a
(possibly non-native-speaking, probably monastery-trained) secretary is immaterial here; see
Alistair Campbell (ed.), The Chronicle of JEihelweard (London, 1962), xxvi-xxxvii; and Cyril
Hart, The B text of the Anglo-Saxon Chronicle', Journal of Medieval History, 8 (1982), 241-99, at
285.
24 'St Neots, /Ethelweard, and the Anglo-Saxon Chronicle'1', Studies in Earlier Old English Prose,
ed. Paul E. Szarmach (Albany, New York, 1986), 193-245.
25 This Scef, surrounded by arms, came in a ship to an island of the ocean which is called
Scani; and he was a very young boy and unknown to the inhabitants of that land. However, he
was accepted by them, and they looked after him with solicitous affection as one of their own, and
afterwards elected him as king. From his line King ^thelwulf took his descent'. Campbell,
dlthelweard, 33.
26 Though /Ethelweard does not use the patronymic form Scefing.
27 Campbell, /Ethelweard, xii, citing and disagreeing with H.E. Barker, The Cottonian
Fragments of /Ethelweard's Chronicle', Bulletin of the Institute of Historical Research, 24 (1951),
46-52.
2S Klaeber, Beowulf, Ixxvii; A. Campbell, Old English Grammar i Oxford, 1959 j, 131.
14
BULLETIN JOHN RYLANDS LIBRARY
Scyld is the most frequent form for the word 'shield' in any kind of
Old English, but in the OE Chronicle s.a. 855 the name reads
Sceldwea. 29
vEthelweard's Scef and Beowulf's Scyld Scefing both came (alone)
over the sea, when very young, to Skane,30 and became king.
However, whereas ^thelweard explicitly stated that the people of the
land adopted the child as one of their own though he was unknown to
them, this is only implicit in the poetically allusive Beowulf. Yet there
is only one possibility of disagreement between them: in ^thelweard
the child was surrounded by weapons, but while the poet says that the
Danes at his funeral gave him gifts as good as those who first sent him
forth, and the objects named are all arms or armour, earlier Scyld had
been described as possessing little when he was found. There is no
necessary contradiction in describing a young child surrounded by
weapons as destitute, however.
I believe ^thelweard took the genealogy from his source Chronicle,31 but my case depends on arguments concerning the stages by
which the West Saxon genealogy acquired names above Woden, who
was regarded as the ultimate ancestor in the earliest versions of nearly
all Anglo-Saxon genealogies. 32 The first extension backwards was
probably to Geat,33 echoes of whose character as a pagan god or a
semi-divine hero seem to have still been heard in ninth-century
Britain. 34 Already in ^thelweard's version of the West Saxon
pedigree, however,35 Geat had acquired the four ancestors Tetuua Beo - Scyld - Scef. As with the group of names from Woden to Geat,
some of those added appear to have come from traditions belonging to
the Germanic peoples, either to their Heroic Age, or to their semi29 Chronicle A has Sceldwaing, Sceldwea, B Sceldweaing, Scyldwa, C Scealdwaing Scealdwa,
D Scealdwaing Scealdhwa; Benjamin Thorpe (ed.), The Anglo-Saxon Chronicle (Rolls Series, 23,
2 vols., London, 1861), i. 126-8.
30 For Skane Beozuulfhas Scedenigge (line 1686, dat. sg.), as well as Scedelandum (d.pl.) in the
prologue; however, the seaman Wulfstan, who was a contemporary of King Alfred and may have
been Mercian, has Sconeg (Sweet's Anglo-Saxon Reader, revised D. Whitelock (Oxford, 1967),
21, line 107, n.230-1); The Old English Orosius, ed. Janet Bately (EETS SS, 6, London, 1980),
16, line 24, which ^thelweard's Scani may represent.
31 Apart from the Norse flavour he gives to Woden and Baeldaeg, transforming them into
Vuothen and Balder; see A.L. Meaney, '^thelweard, /Elfric, the Norse Gods and Northumbria', Journal of Religious History, 6 (1970), 105-32.
32 Chambers, Beowulf, 319-22; Kenneth Sisam, 'Anglo-Saxon Royal Genealogies', Proceedings
of the British Academy, 39 (1953), 287-348, at 308.
33 PG*Gautaz; whom Hermann Moisl argued first appears as the Gothic eponym; 'AngloSaxon Royal Genealogies and Germanic Oral Tradition', Journal of Medieval History, 7 (1981),
215^8, at 219-23.
34 See comments about Geta in the Historia Brittonum, ch.31 (ed. and transl. John R. Morris,
Nennius: British History and the Welsh Annals (Chichester, 1980), 67); and in Aster's Life of King
Alfred, 1, ed. W.H. Stevenson (Oxford, 1904), 3, which may have been influenced here by the
Historia Brittonum; see Simon Keynes and Michael Lapidge, Alfred the Great (Harmondsworth,
Middlesex, 1983), 54, 299 n.6. In Old Norse Gautr is one of the by-names of Othin.
35 ^thelweard's is the shortest of the extant versions of the West Saxon genealogy which is
taken back beyond Woden Frealafing. The gradual expansion of the West Saxon genealogy is
discussed in my forthcoming edition of the West Saxon Regnal List and Genealogy.
DATING OF BEOWULF
15
mythical, semi-legendary origins. Sisam remarked on the simple
meanings of the names Sceaf, Scyld and Beo - 'which has suggested a
mythological interpretation'. 36 Moreover, Widsith (line 32) mentions a
Sceafa as ruling the Lombards,37 and this indicates that he might have
been celebrated in heroic poetry, now lost. 38
CONCORDANCE OF GENEALOGIES
Chronicle A
855
Noea
Sceaf(ing)b
Bedwigb
Hwalab
Hraf>ra
Itermon
Heremod
Sceldwea
Beaw
Taetwa
Geat
Godwulf
Fin
Frijsuwulf
Frealaf
FriJ>uwald
Wodenc
/Ethelweard
Beowulf
Scef
Scef(ing)
Sven
Saxo
Dan
Lotherus
Scioldus
Gram
Hadingus
Scyld
Scyld
Skjold
Beo
Tetuua
Geat
Goduulf
Fin
Frithouulf
Frealaf
Frithouuald
Vuothend
Beowulf
Healfdene
Hrofjgar
etc
Frotho \
Haldan }
Helghi
etc
Frotho
Haldan
etc
Skjoldunga Saga
Scioldus
Fridleifus
Frodo I
Herleifus
Havardus
Frodo II
Vermundus
Olava (f.)
Frodo III
Fridleifus
Frodo IV
Halfdanus
Roas
etc
a Above Noah, the names in the A manuscript of the Anglo-Saxon Chronicle read Lamach,
Matusalem, Enoh, laered, Maleel, Camon, Enos, Sed, Adam.
b Names taken from the B manuscript because of an accidental omission in A.
c Below Woden the names read Baeldaeg, Brond, Fribogar, Freawine, Wig, Giwis, Esla,
Elesa, Cerdic, Cynric, Ceawlin, Cupwine, Cupa, Ceolwald, Cenred, Ingild, Eoppa, Eafa,
Ealhmund, Ecgbryht, ^belwulf.
d Below Vuothen the names read Balder, Brond, Frithogar, Freauuine, Vuig, Geuuis, Esla,
Elesa, Cerdic, Cynric, Ceaulin, Cuthuuine, Ceoluuald, Cenred, Ingild, Eoppa, Eafa, Ealhmund,
Ecgbyrht, .Ethelwulf.
36 Sisam, 'Genealogies', 318; see also 307-14.
37 R.W. Chambers (ed.), Widsith (Cambridge, 1912), 200-1.
38 Alistair Campbell, The Use in Beowulf of Earlier Heroic Verse', England Before the
Conquest: Studies in Primary Sources presented to Dorothy Whitehck, ed. Peter Clemoes and
Kathleen Hughes (Cambridge, 1971), 283-92, at 284. Murray argues (Dating, ed. Chase, 106
16
BULLETIN JOHN RYLANDS LIBRARY
The Lombards had a tradition that they originated in Scandinavia,39 but their legendary history starts with their departure thence.
None of their historians mentions any king or duke with a name
resembling Sceaf; however, Paul the Deacon relates a foundling story
about Lamissio, their second king. He was one of septuplet boys born
to a prostitute and cast into a pond to drown. The first king,
Agelmund, passing by, stirred the boys with his spear, whereupon
Lamissio clutched it. The king had him rescued and carefully
nurtured,40 and he became a great warrior. When on an expedition the
king was barred from crossing a river by 'Amazons',41 Lamissio killed
the strongest of them in a duel in the river and so won passage. In the
lands on the other side the Lombards grew careless and the 'Vulgares'
surprised them, killing Agelmund and carrying off his daughter, and
Lamissio was made king. He exhorted the Lombards to vengeance,
winning a great victory and much booty. When he died he was
succeeded by Lethu (whose paternity is not stated).42
Ludwig Schmidt believed this story to be a 'fabulous expansion of
the original myth of Skeaf. This may be going too far, but we can
agree that 'the germ of the myth is that a hero of unknown origin came
from the water to the help of the land in time of need'. 43 Moses is
perhaps the best-known of this type of folklore hero.44 The fact that
v^Ethelweard described Sceaf as surrounded by weapons - which Olrik
takes to reflect the concept that the hero finds his life's work in the
cradle - appears to put him into this category. However, to judge from
his name, 'Sheaf, he may instead (or also?) have been a culture hero, a
corn-spirit or an agricultural deity who came over the water to help his
future people with a new crop. 45
n.24) that Scef is best seen as 'English' no matter where he is located, since 'the Lombards
themselves knew of no Scef so far as we know'. What is important here, however, is that there is
nothing outside this genealogy to connect him with the Danes.
39 Origo Gentis Langobardorum, 1, ed G. Waitz (M[onumenta] G[ermaniae] Hfistorica],
Scriptores Rerum Langobardicarum, Hanover, 1878), 2-3; W.D. Foulke (transl.), Paul the
Deacon, History of the Lombards (1st ed., Pennsylvania, 1907; repr. 1974 with intro. by E.
Peters), 315-7; Pauli Historia Langobardorum, i, chs. 11-13, ed. L. Bethmann and G. Waitz
(MGH Script. Rer. Lang., 1878), 53-4; Foulke, Paul, 53-4.
40 This motif is frequent and ancient in folklore; see discussion under SS Dyfrwyr in Sabine
Baring-Gould and John Fisher, The Lives of the British Saints (4 vols., London, 1908), ii.
398^05. For differing interpretations (e.g. prostitute = bitch and Lamissio/Laiamicho =
half-animal ancestor) see Kemp Malone, 'Agelmund and Lamicho', American Journal of
Philology, 47 (1926), 319-46; Karl Hauck, 'Lebensnormen und Kultmythen in germanischen
Stammes- und Herrschergenealogien', Saeculum, 6 (1955), 186-233, at 206-9.
41 Conjectured by Ludwig Schmidt, Zur Geschichte der Langobarden (Leipzig, 1885), 17 n.l to
be some kind of Germanic water-spirits.
42 Paul, Historia, i, chs. 15-18; Waitz, MGH Script. Rer. Lang., 54-6.
43 Foulke, Paul, 28 n.l, reporting Schmidt, Langobarden, 50.
44 See A. Aarne, Motif-Index of Folk-Literature, rev. Stith Thompson (6 vols, Copenhagen and
Bloomington, Ind., 1955), i.119 (A513.2); v.9 (Llll.2.1); Olrik, Legends, 392-3. However,
Olrik rejects the 'usual motif of the foundling* in considering 'traces of the Scyld motif outside
the Scandinavian-English myths about him', though he includes discussion of Ing (ibid.,
413-21).
45 Chambers, Beowulf, 81-6.
DATING OF BEOWULF
17
Another ultimate ancestor who may have come from overseas and
departed again - though neither a voyage in youth nor a ship-burial is
specified for him - is the Ing who has a verse devoted to him in the Old
English Rune Poem:
Ing waes aerest mid Eastdenum
gesewen secgun oj> he siSSan est [?eft or east?]
ofer waeg gewat, waen aefter ran;
Sus heardingas 5one haele nemdun
(lines 67-70)46
Ing is presumably the Swedish and Danish eponym (of the Inglingar
and the Ingwine), described in the Skjoldunga Saga as brother of
Skjold and son of Othin. Hilda Davidson equates the Heardingas with
'the Hasdingi, the royal dynasty of the Vandals, [who] may have
worshipped Ing while they were in South Sweden, and carried his cult
to Denmark and further east when they migrated from Scandinavia'. 47
Perhaps there was a general folklore tradition among the south
Scandinavian tribes of a saviour from overseas, with variant versions
attached to various names.
Since Beo's son is called Tetuua by ^thelweard (Taetwa in the
Old English Chronicles} - a name unknown among the Danes or
anywhere else in Heroic Age history or literature - and Frotho's son is
called Halfdan in the Scandinavian sources, it may have been the name
Scyld alone which was borrowed from the Danish line into the West
Saxon. By 858 - the earliest date at which the annal containing the
genealogy could have been written into the archetypal Chronicle4* - or
soon after, the Anglo-Saxons may well have learnt some Danish
traditions, including that of their eponym Skjold (whose name would
surely have been easily recognizable to them as equivalent to the OE
scyld 'shield').
The earliest text in which the Danes are called Scyldings - in the
form Scaldingi, which appears to reflect ninth- or tenth-century Norse
speech49 - is the anonymous Historia de Sancto Cuthberto (chs.ll,
46 'Ing was first seen by men among the East Danes, until he departed (east or again) over the
sea; the waggon ran after. Thus the Heardingas named the hero.' The Anglo-Saxon Minor Poems,
ed. Elliott van Kirk Dobbie (New York, 1942; 1968 repr), xlvi-1, clxii, 28-30 (text), 153-60.
Aarne-Thompson (Motif-Index, i.119, 124 (A513.2 and A563) cite no other examples of
'Divinity's departure in boat' than Scyld Scefing's.
47 Hilda R.E. Davidson, Gods and Myths of Northern Europe (Harmondsworth, Middlesex,
1964), 104.
48 This does not imply accepting that an earlier Chronicle ended with this annal; it is simply a
statement of fact.
49 Alan L. Binns, The Viking Century in East Yorkshire (East Yorkshire Local History Series,
15, 1963), 49-50; and 'The York Viking Kingdom: Relations between Old English and Old
Norse Culture', The Fourth Viking Congress, York, August 1961, ed. Alan Small (Edinburgh and
London, 1965), 179-89, at 184 and n.l. Binns makes the point that Scaldingi is not a
conversational word, and is therefore 'evidence of the recitation of formal poetry, no doubt
glorifying the Viking dynasty'. Roberta Frank ('Skaldic Verse and the Date of Beowulf, Dating,
18
BULLETIN JOHN RYLANDS LIBRARY
12),50 first compiled in Chester-le-Street about 945,51 and referring to
the attacks on Northumbria in the late 860s by the brothers Ivar and
Halfdan. According to legend they were sons of Ragnar Lothbrok,
and, while not belonging to the Scylding line, were evidently considered true Danish kings even by their contemporaries. 52 However,
the climate in Wessex may well have been unfavourable to the idea
that its kings shared an ancestor with the marauding Danes until after
Alfred had begun to convert them (Guthrum in 878, Haesten's sons by
894 according to the Chronicle^), Moreover, the Chronicle is unlikely to
have reached its final pre-ae form (that translated by jEthelweard)
before Alfred's reign. 53
Whoever was responsible for the '^Ethelweardian' extension to
the West Saxon genealogy therefore picked up names somewhat
indiscriminately - from the Danes and perhaps also from the Lombards - in order to glorify his kings; and this pedigree-maker was
presumably the chronicler who composed the annal s.a. 855 in
^thelweard's pre-ae Chronicle. For ^thelweard carefully numbers
^Ethelwulf s ancestors - Scef being the nineteenth - and he would
surely not have rejected the chance of connecting himself (through
^thelwulf) with Adam if he had been shown the way,54 since he alone
prefaces his version of the Chronicle with an account of the Six Ages of
the World, from Adam onwards. 55 So it is a reasonably safe assumption that in ^thelweard's pre-ae text of the Chronicle the West Saxon
ed. Chase, 123-39, at 126-7) traces the usage of the word scyldingas/skjoldungar/scaldingi. In
Beowulf, in the Historia de Sancto Cuthberto, and in the Danish historians the term is used for the
Danes and for their kings, though Frank claims that it was never used by the Scandinavian
writers as a 'dynastic title' in the singular as it is in Beowulf lines 1792, 2105 (Hrothgar = gamela
Scylding, 'the old Scylding'). However, since Hrothgar was both a member of the dynasty and
king, this is surely only a minor extension of usage. Early eleventh-century Norse skalds used the
term for Cnut and contemporary Norwegian kings, particularly in connection with their 'English
adventures', but this could have been an idiosyncratic usage by non-Danish poets.
50 Ed. in Thomas Arnold, Symeonis monachi opera omnia, I (Rolls Series, 75, London, 1882),
196-214, at 202.
51 Edmund Craster, 'The Patrimony of St Cuthbert', English Historical Review, 69 (1954),
177-99, at 177-8.
52 See Alfred P. Smyth, Scandinavian Kings in the British Isles, 850-880 (Oxford, 1977), 1-6,
10-15, 29-35.
53 Meaney, 'Chronicle', Studies, ed. Szarmach, 205-6.
54 Sisam, 'Genealogies', 320, argues that ^thelweard's friend -lElfric would have dissuaded
him from accepting an uncanonical link with Noah, but (apart from the fact that jEthelweard's
Chronicle was probably composed in the early part of .flithelred's reign before the ealdorman is
likely to have come under the influence of the younger man; see Campbell, /fthelweard, xiii n.2),
if jEthelweard had been worried by such scruples, he could have adopted some other kind of link
(as the post-Conquest writers did, making Scef Shem's son rather than Noah's, for example),
instead of preferring a mysterious arrival from overseas to Noah and Adam, from whom, as a
Christian, he would anyway have believed himself descended. The manuscript of Asser's Life of
Alfred has Seth as the name of Noah's son, Bedwig's father, but Sisam regards this as merely a
scribal error; 'Genealogies', 316.
55 This account cannot have been part of the early Chronicle which ^thelweard was
translating, since it is attached to no other version.
DATING OF BEOWULF
19
pedigree ended Geat Taetw(a56)ing, Taetwa Beowing, Beo Scylding,
Scyld Scefing - the final name, as in all other genealogies in Old
English, being expressed only as a patronymic.
Sisam supposed that ^thelweard took what he said about Scef
from family tradition,57 but again it can be argued that it was in his
exemplar-Chronicle. Here the revisions which were made to ae (the
common ancestor of the extant Chronicks in Old English), as
compared to the earlier version of the genealogy which ^thelweard
transcribed, must be discussed briefly. Scyld's name appears in the
weak form Sceldwea;58 and just as some of the Danish sources made
Skjold a son of Lother, so the ae Chronicle made Sceldwea a son of
Heremod, separating him and Sceaf. Therefore, some further Danish
influence may have been involved.
Three other names (Itermon, Haf>ra and Hwala)59 were added
above Heremod, and then came Bedwig60 son of Sceaf, who is surely
Beo/Beaw son of Scyld with a little misreading, and with a duplication
common in genealogies. Sceaf himself stands effectively where he did
in vEthelweard's copy of the Chronicle at the head of the genealogy, but
with his provenance strangely changed: Bedwig Sceofing, id est filius
Noe, se was geboren in pare earce Noes61 - a statement which surely
proves that the motif of the hero coming over the sea as a child must
have belonged to Sceaf, or what would have been the point of his
transmogrification? Nor do I think the statement at all ambiguous;
clearly the clause refers to Sceaf, not to Bedwig.
Finally, though it is not immediately obvious from his Latin,
jEthelweard's remarks about Scef s journey echo quite closely the
wording of the brief Chronicle accounts of the coming of the first
English invaders. 62 Therefore the sentences which ^thelweard was
56 Forms preserving the weak -a ending in the patronymic are usually held to be the result of
creating an Old English genealogy from a Latin list lacking patronymics. However, since the late
ninth-century A Chronicle omits the a, whereas the late tenth- and eleventh-century BCD
Chronicles usually preserve it, the variation may be due only to date.
57 Sisam, 'Genealogies', 317; Murray, Dating, ed. Chase, 107. Yet this 'family tradition' can
have gone back only to an artificial creation in the time of ^thelwulf or Alfred.
58 This may not be of any great significance in itself: Geat and Sceaf are also found with weak
endings, and this may be due to their being first current as patronymics, so that their
terminations would not be obvious. However, the addition of weak endings may at times have
been deliberate, in order to make common nouns look like proper nouns.
59 These names are discussed by J.M. Kemble, UeberdieStammtafelderWestsachsen (Munich,
1836), 13-14; by Sisam, 'Genealogies', 315-6; and in the Appendix to my forthcoming edition of
the West Saxon Regnal List and Genealogy.
60 In BC; in A all unbiblical names after Hrabaing are omitted; however, Asser (who used a
version of the Chronicle close to A) also has Huala and Beduuig; therefore BC's form must have
stood in ae, and D's Beowung Beowi must be erroneous; see Asser, Alfred, ed. Stevenson (1904),
3; Sisam, 'Genealogies', 315 n.l.
61 'Bedwig son of Sceaf, that is the son of Noah, who was born in Noah's Ark'.
62 Compare, for example, the beginning of jEthelweard's annal 494: . . . Cerdtc el Cinnc filius
suus cum quinque cannis aduecti Brittanniam in portum qm Cerdices man nuncupatur; Campbell,
£lhelweard, \ 1. Conversely, an attempt to render .flithelweard's Latin concerning Scef back into
20
BULLETIN JOHN RYLANDS LIBRARY
rendering had most probably been composed for the early Chronicle,
pointing up Scef s aptness as the ultimate ancestor for a royal line
which had itself come over the sea.
From Halfdan/Healfdene onwards, it appears to be purely Danish
traditions which Beowulf is following, in an earlier form than that in
the extant Scandinavian sources;63 but nothing corresponding to the
name Beo (or Beowulf as it is in the poem) is found in a comparable
position in any of these sources, nor is there to be found in them any
reason for calling Skjold Sce(a)fing. Since it seems very unlikely that
Scyld and Sceaf (who did not belong to the same tradition) could
anywhere have been brought into contact as father and son except in
an artificial genealogy such as ^thelwulf s, the obvious place for the
poet to have obtained the sequence is from ^thelweard's pre-ae
version of the Chronicle. From it he would have got a sequence ending
something like Tcetwa Beowing, Beo ScyMing, Scyld Sce(a)fing se com
mid .i. scipe . . .; and from Danish tradition the sequence (expressed
here in English name forms) Hrothgar - Healfdene - Froda64 - Scyld,
and the character of Scyld as a warlike restorer of his nation after
previous disasters.
Now, the Beowulf poet frequently used the termination -ing
meaning 'son of, and must have known that the foundling story really
belonged to Scef. 65 But in order to achieve the effect he wanted, he
may have decided to add Scef s characteristics to Scyld's: to make the
Scyldings' eponym come alone over the sea as a helpless boy, and with
a folklore reversal of fortune become the restorer of Denmark's
greatness. The cheating involved in giving Scyld attributes which
really belonged to Sce(a)f was partly offset by calling him 'son of Scef
Old English might read something like: Se Sceaf com mid. i. scipe on garsecges ealond be is genemned
Scant, mid wcepnum betined, 7 he wees swibe geong umbor, 1 landleodum uncud, ac hie him onfengon
swa swa hira agene 7 feddon mid welwillendum mode, 7 swa siddan hie hine to cyninge gecuron; barn
goeb &belwulfes cyninges rihtfaderencynn.
" For the Norse sources see Chambers, Beowulf, 16, 129-32; Garmonsway and Simpson,
Analogues, 116-211.
64 Though it does not create any real difficulties in this discussion, the greatest puzzle in this
reconstruction is why the Danes have the name Frotho(i) and the English Beo(w) in the same
position. Was Frotho, i.e. 'the wise', a title which ousted the original name, as Chambers
(Beowulf, 92 n.2) suggested? Or was Frothi a fertility god, as Davidson suggested (Gods and
Myths, 103^), whose name was replaced by something more innocuous in the Old English? Or
is the Beo of the English sources a purely arbitrary addition to the genealogy? Two of the Danish
king lists (see n. 16) have B0gi as the name of an otherwise unknown son of Lother and father of
Skjold, and this may indicate that an ancestor with a similar name to Beo belonged to some
Danish tradition. See Alfred Ebenbeauer, 'Frodi und sein Friede', Festgabe fur Otto Hofler zum
75. Geburtstag, ed. Helmut Birkhan (Vienna and Stuttgart, 1976), 128-79, for a recent discussion
of the Danish royal genealogies, with comprehensive references to earlier scholarship.
65 Hector M. Chadwick (The Origin of the English Nation (1st ed., Cambridge, 1907), 276)
supposed that Scefing was to be taken as 'child of the sheaf or 'sheaf-child'. However, Sisam
('Genealogies', 318-20) argued convincingly that William of Malmesbury's account of Sceaf
coming with a sheaf of corn at his head was William's adaptation of ^thelweard, and not due to
popular tradition. Therefore Chadwick's explanation of Scefing is untenable.
DATING OF BEOWULF
21
and by ignoring the fact that a foundling story can only be told of the
first of his line. 66
If this reconstruction is correct, Scyld Scefing and his arrival from
overseas cannot have become part of the prologue of Beowulf before
858, and almost certainly not before Alfred's reign. But it could have
been any time after that, for though ^thelwulf s genealogy reached
unprecedented and unsurpassable lengths in the ae Chronicle and all its
descendants (being taken back to Adam), two versions of the pre-ae
Chronicle survived at least to the late tenth century to be used in Latin
chronicles.67 Scyld Scefing's childhood voyage (like the details of the
description of Grendel's mere from the Visio Fault) could even have
been added to the poem at the last possible moment, by the scribe who
wrote the first part of the poem into the No well Codex.
However, the name of Scyld's son, which is Beo in ^Ethelweard,
Beaw in ^Ethelwulf s genealogy in the Anglo-Saxon Chronicle., poses a
problem. Beo has been taken to mean 'barley';68 otherwise, it could
simply mean 'bee', a creature of great importance to the AngloSaxons69 and to other early Germanic peoples, who may not always
have recognized its femininity. 70 In either case, whether his name is to
be interpreted as 'bee' or 'barley', Beo could (as could Sceaf) have
been a culture-hero. In Beowulf Scyld's son has the same name as that
of the Geatish hero of the poem, but scholars have queried that such a
rare name should be borne by two unrelated heroes in the same poem
without comment. 71 A more telling point is that the half-line Beowulf
66 The poem has comparable inconsistencies which serve the purposes of poetry: in lines
2183-9 Beowulf himself is described as sluggish and despised in youth, whereas elsewhere (e.g.
lines 408-9) he becomes a hero very young; see Adrien Bonjour, 'Jottings on Beowulf and the
Aesthetic Approach', Old English Poetry: Fifteen Essays, ed. Robert P. Creed (Providence, 1967),
179-92, at 182-3. Moreover, Beowulf told Hrothgar that leaders of his people had encouraged
him to venture against Grendel (lines 415-8); but on his return Hygelac (whose advice would
have mattered most) says he urged him for a long time to leave the Danes to their own devices
(1992-7); K. Sisam, The Structure of Beowulf (Oxford, 1965), 46-7.
67 As well as that used by ^thelweard, which was probably somewhere in the western shires,
the yet earlier version of the Chronicle excerpted by the St Neot's compiler (which presumably
also had the 855 genealogy in the earlier form) was still extant in the east; see Cyril Hart, 'The
East Anglian Chronicle', Journal of Medieval History, 7 (1981), 249-82, at 254-5, 274-7;
Meaney, 'Chronicle', Studies, ed. Szarmach, 193-245.
68 Chambers (Beowulf, 10, 42-7, 56, 87-8, 291-3, 296-304) summarized earlier scholarly
opinions about Beo; there are no certain clues as to the characteristics attributed to him.
69 Bosworth and Toller's Anglo-Saxon Dictionary (1898) comments that bee-keeping was very
important in the Anglo-Saxon economy. The charm for recovering a swarm, and a comment that
bees carry a declicious food and have a poisonous tail show the respect accorded to them. Beaw,
according to Toller's Supplement (Oxford, 1921), means 'gad-fly'. Was the bee replaced by the
glad-fly in the official, x version of the West Saxon genealogy?
70 The bee was used as a symbol in the consecration of the Visigothic king Wamba; Roger
Collins, 'Julian of Toledo and the Royal Succession in late Seventh-Century Spain', Early
MeJuiul Kingship, ed. Peter H. Sawyer and lan N. Wood (Leeds, 1977), 30-49, at 42,46-7. The
northern Germans - like J.J. Chiflet (Anastasis Childenci, 1 (Antwerp, 1655), 141 (figure labelled
Apes aurete gemmate), 155-63) - may even have interpreted the insects on the cloak of the
fifth-century Prankish king Childeric as bees, not as the southern cicadas (Peter Lasko, The
Kingdom of the Franks (London, 1971), 25).
71 Sisam, 'Genealogies', 340.
22
BULLETIN JOHN RYLANDS LIBRARY
Scyldinga (53b) is suspect metrically;72 whereas Beo(w) Scyldinga
would be quite acceptable. Therefore, the name Beowulf for the Dane
may be simply a scribal error.
I have referred to the poet's giving Scyld an origin really
belonging to Scef as cheating; but (confusing as it has been for modern
scholars) the gain to the poet was immeasurable: for all kinds of
resonances are set up in the Prologue, all kinds of themes suggested
which are developed, mostly in the person of the hero Beowulf
himself, in the course of the poem. For Beowulf too came over the sea
to rescue the Danes in time of trouble; Beowulf too was a king who
reigned long and successfully after difficulties in his youth. However,
in one respect Beowulf did not parallel Scyld, who had a son, eagerly
awaited, since the Danes (the poet tells us) had suffered a long time of
distress when without a lord, and (it is implied) a good king is twice as
good if he has heirs. Beowulf did not. 73
One other element is woven into the Prologue to the poem:
Scyld's splendid and unusual ship-funeral, which not only balances
his own mysterious arrival as a child, but also Beowulf s own funeral at
the end of the poem. It is one of the poem's most brilliant strokes.74
Therefore, the search for parallels to it has been intense, especially
since for the migration and early medieval periods more than four
hundred ships or boats have been identified as playing a part in the
burial ritual (predominantly of aristocratic males) on nearly three
hundred sites in northern Europe. 75
Some archaeologists consider these vessels as no more than 'part
of the equipment of the dead in the same sense as his horses, clothing
[and] weapons'. 76 When the boat was placed (sometimes inverted)
over a burial-chamber, or was burnt among other grave-goods (as were
slightly more than half the buried boats), it may indeed have been
provided primarily as grave furniture, with perhaps the idea that it
would be of use to the spirit in the next world. Knut Stjerna
summarized the alternative view: that the ship was primarily for
transportation to the land of the dead, which was believed to be some
physical distance from the land of the living, and so, to a seafaring
72 It is a type D with a tertiary stress, only found elsewhere in the first half-line and with
double alliteration; see Alan J. Bliss, The Metre of Beowulf (Oxford, 1958), 57-8, para.64. If
Beowulf (the Dane) were emended to Beo, 18a (the only other half-line in which the name
occurs) would then become Beo woes breme - the most basic kind of A verse.
73 This is also emphasized by John C. Pope, 'Beowulf s Old Age', Philological Essays, ed.
Rosier, 55-64, at 56.
74 See e.g. Klaeber (ed.), Beowulf (1941), 228; Adrien Bonjour, Twelve Beowulf Papers,
1940-60, with Additional Comments (Neuchatel, 1962), 122-3.
75 Michael Miiller-Wille, Bestattung im Boot: Studien zu einer nordeuropdischen Grabsitte (Offa
25/26, 1968-69); 'Boat-graves in Northern Europe', International Journal of Nautical Archaeolog/
and Underwater Exploration, 3 (1974), 187-204.
76 Knud Thorvildsen, Ladby-Skibet (Nordiske Fortidsminder, 6, Copenhagen, 1957), 95-106
(English summary 113-14); Schonback, 'Custom', Vendel Studies, ed. Lamm and Nordstrom,
129-31 (see n.81).
DATING OF BEOWULF
23
people, necessitated a sea-journey. Stjerna thought that these ideas
first appeared in the north in the fourth century AD; his evidence
being that in some Scandinavian graves a 'Charon's obol' was found
'in the dead man's mouth, after the Greek fashion'. 77 More positive
evidence for the idea of the journey of the dead is provided by the
Gotland memorial stones depicting a ship78 and by the stone-settings
in the shape of ships around cremation burials, which are particularly
common in Denmark, for example, at Lindholm H0je. 79
Though an established burial custom does not necessarily reflect
belief, one cannot assert that the burial-ship was never associated in
the minds of the early Scandinavians with the journey to 'the other
side'. Therefore, it is legitimate to consider here ship- and boat-burials
in the Scandinavian archaeological record as manifestations of at least
some of the same ideas as those underlying Scyld's literary funeral. In
the following brief survey I shall disregard cremation burials, 80 and
concentrate on the more princely inhumations, which are themselves
conveniently concentrated in time and place.
First, in Uppland in central Sweden, there are some cemeteries
called 'Vendel' from the type-site, which seem to have served
settlements of one household only. In them, most of the burials were
cremations, but there was typically one male inhumation burial per
generation, most frequently in a boat. At Valsgarde, for which the
longest sequence is evidenced, the boat-burials begin in the mid-sixth
century and continue (with some diminution in richness in the ninth
century) into the eleventh. The goods were disposed in the boats as if
for a long voyage:
In the prow the provisions [mostly joints of meat] and cooking equipment, close to
midships the more personal equipment, such as gaming-board, dice and playing
pieces, dishes, drinking vessels and small chests containing small tools etc. ...
There was a distinct concentration of weapons around and above the dead man's
bed amidships ... It is very likely that the position of the body corresponded to the
space which was usually used as sleeping-quarters. . . .
The stern of the burial-boat was often completely empty. Dogs on
leashes were also found in the boats, either in the prow or amidships,
and horses were tumbled into the sides of the burial-pit. None of the
burial-boats appears to have had a mast. Those buried in the boats
77 Knut Stjerna, Essays on Questions connected with the Old English Poem of Beowulf, transl.
John R. Clark Hall (Coventry, 1912), 97-135, at 101-2.
78 A.F Major, 'Ship-burials in Scandinavian Lands and the Beliefs that Underlie them',
Folklore, 35 (1924), 113-50; Sune Lindqvist (ed.), Gotlands Bildsteine (2 vols., Stockholm,
1941-12).
79 Thorkild Ramskou, 'Lindholm H0je', Acta Archaeologia, 24 (1953), 186-99; 26 (1955),
176-85; 28 (1957), 193-201.
80 Since cremation adds a further element of uncertainty in interpreting the ship-burial ritual.
However, some cremation burials with ships may have been royal; for example, one was
apparently among the goods burnt on the pyre beneath 'Othin's howe' at Old Uppsala; H.
Shetelig, 'Ship Burials', Saga-Book, 4 (1904-05), 326-63, at 331-2.
24
BULLETIN JOHN RYLANDS LIBRARY
MAP 1: Early boat-graves
up to c. AD 600
1
c. AD 600-800
2
Ve Vendel
Va Valsgarde
SH Sutton Hoo
DATING OF BEOWULF
25
certainly had considerable hereditary power; nevertheless, in spite of
the richness of the arms and armour, these were not royal graves: they
did not contain anything which could be interpreted as regalia, or any
fine gold-and-garnet cloisonne work. 81
When in 1939 the seventh-century ship-burial was excavated at
Sutton Hoo in Suffolk, both archaeologists and Old English scholars
quickly found parallels to Beowulf. 82 In some ways the Sutton Hoo
ritual was closer to Scyld's than that of the Vendel graves, which it
much resembles: there were no sacrificed animals in the ship; and it
was clearly royal, and abundantly supplied with weapons, regalia and
other treasures. If the ship had had a mast, it had been removed, and
amidships was a plank-built burial-chamber, inside which the body (if
there were one) and all the grave-goods were placed, with cooking
equipment to the east, drinking and eating vessels and personal
equipment including a sword and purse near but not on the body in
the centre, and other weapons and regalia to the west. 83 Moreover, as
Hilda Davidson pointed out, if 'the Sutton Hoo ship was rowed along
the Deben to the point where she was pulled up the hill to the grave
prepared for her, then it is possible that the departure of a dead chief
on his ship, rowed towards his last resting-place, was a practice
witnessed up to the seventh century by the people of East Anglia'.
Such a scene might well linger in folk-memory. 84
However, ship-burials are most widespread and frequent in the
Viking period, particularly along the Norwegian coast, some of the
most princely being around the Oslo fjord. They appear to be isolated
chieftains' burials; and Alfred Smyth has argued that the man buried
at Gokstad between 870 and 890 is to be identified with a king of
Vestfold, also known as Olaf the White, who became king of Dublin in
853, and who had been a close ally of Ivar (son of the legendary
81 Description summarized from articles in Vendel Period Studies, Transactions of the Boatgrave Symposium in Stockholm, February 2-3,1981, Museum of National Antiquities, Stockholm,
Studies 2, ed. J.P. Lamm and H.A. Nordstrom (1983); see Bjorn Ambrosiani, 'Background to
the Boat-graves of the Malaren Valley', 17-22, and 'Regalia and Symbols in the Boat-graves',
23-30; Birgit Arrhenius, The Chronology of the Vendel Graves', 39-70; Greta Anvidsson,
'Valsgarde', 71-82 (quotation from 76-7); Peter Sawyer, 'Settlement and Power among the Svear
in the Vendel Period', 117-22; Bengt Schonback, The Custom of Burial in Boats', 123-32.
Further references, e.g. to excavation reports, will be found in the bibliographies to these
articles.
82 See articles in Antiquity, 14 (1940), 6-87, especially H.M. Chadwick, 'Who was he?', 76-87;
also S. Lindqvist, 'Sutton Hoo and Beowulf, Antiquity, 22 (1948), 131-40; C.L. Wrenn in his
supplement to the 3rd ed. of Chambers, Beowulf, 508-23; Rupert L.S. Bruce-Mitford, 'Sutton
Hoo and the Background to the Poem', in Girvan, Seventh Century, 85-98.
83 R.L.S. Bruce-Mitford, The Sutton Hoo Ship-Burial, i: Excavations, Background, the Ship,
Dating and Inventory (London, 1975), 176-221, 410-2, 420-35, 488-577.
84 'Archaeology and Beowulf, in Garmonsway and Simpson, Analogues, 350-60, at 358.
Davidson also remarked that 'there is a close link between objects ... as described in Beowulf
and archaeological evidence from the sixth and seventh centuries A.D.', ibid., 359; see also
Rosemary J. Cramp, 'Beowulf and Archaeology', Medieval Archaeology, i (1957), 57-77.
However, this may be archaism, and need not preclude later composition of at least parts of the
poem.
26
BULLETIN JOHN RYLANDS LIBRARY
Ragnar Lothbrok) in England until 866. 85
The mounds covering the Viking ship-burials are often very
large; but though their arrangement resembles that in the Vendel
cemeteries, there is often the addition of a sturdy burial-chamber
(sometimes built across the boat, and even resting on stone walls). In
the Gokstad ship, right behind the mast, which was cut off short, was a
gable-roofed burial-chamber of 'great wooden beams'. In it, a richlyclothed man about fifty years old had been laid on a bed with fine
bed-clothes. Disturbed artefacts including buckles, ornaments for
straps and mounts from a box, a leather purse and a board-game were
found nearby. Before the mast were boats, beds and kitchen equipment; and also in the ship were a sleigh and the remains of a peacock,
and just outside it the bones of horses and dogs. In undisturbed Viking
ship-burials weapons were found near the body; Shetelig lists those in
a mound on the island of Karm0y as two swords, two spears, a round
quiver with two dozen arrows, as well as smithying tools. 86
Ship-burials were not usual in Denmark, though 'a very small
number . . . have been found in Denmark and Schleswig, of which
the most spectacular is that from Ladby.'87 At this site were a few
modest inhumations, one cremation, and a richly-furnished shipburial of the mid-tenth century, with no signs of a burial-chamber, but
in which a splendidly-clothed body had probably lain on a feather bed
'just aft of midships'. The excavator surmised that a local chief with
extensive overseas contacts had emulated a Norwegian burial.88
Viking boat-burials are not known from England,89 though they have
been found in the western isles of Scotland and the Isle of Man, mostly
dating from the later ninth or early tenth centuries. 90 While some may
have been chieftains' burials, their grave-goods cannot compare to
those at Valsgarde, Sutton Hoo or Gokstad.
If we compare the funeral ritual evidenced in the Vendel and
Viking ship-burials with Scyld's, there is one specific area of simi85 Smyth, Scandinavian Kings, 101-53; for Gokstad, see 109-10. Doubts have been thrown on
the possibility of identifying the Gokstad king as Olaf GeirstaSaalfr, e.g. by Torleif Sjovold, The
Viking Ships (Oslo, 1954), 7, 30-1; it appears to me that this is carrying academic caution too far,
though identification with Olaf the White is less certain.
86 Shetelig, 'Ship Burials', 353-4; see also A.W. Bregger and H. Shetelig, The Viking Ships,
transl. K. John (Oslo, 1951), 81-6.
87 Gerhard Bersu and David M. Wilson, Three Viking Graves in the Isle of Man (Society for
Medieval Archaeology Monograph Series, 1, London, 1966), 92.
88 Thorvildsen, Ladby-Skibet.
89 A boat found at Walthamstow was long regarded as a Viking burial, but carbon-14 dating
has shown that it is sixteenth-century or later; see Valeric H. Fenwick, 'Was there a Body
beneath the Walthamstow Boat?', International Journal ofNautical Archaeology, 7 (1978), 187-94.
90 For a survey of possible boat-burials in the British Isles, see H. Shetelig (ed.), VtiaM
Antiquities in Great Britain and Ireland, vi (Oslo, 1954), 73-5. The best-attested in the western
isles appears to be that from Kiloran Bay, Colonsay; see H. Shetelig, Saga-Book, 5 (1906-07),
172-4; J. Anderson, Proceedings of the Society of Antiquaries of Scotland, 41 (1907), 443-50. For
the Isle of Man, see P.M.C. Kermode, 'Ship-burial in the Isle of Man', Antiquaries Journal, 10
(1930), 126-33; Bersu and Wilson, Viking Graves, 1-44, 84-92.
DATING OF BEOWULF
27
MAP 2: Boat- and ship-burials from the Viking period (c. AD 800-1050)
1
2
G
K
L
I
cremation grave
inhumation grave
Gokstad
Karm0y
Ladby
lie de Groix
28
BULLETIN JOHN RYLANDS LIBRARY
larity: he was also laid amidships, with precious objects, arms and
armour, around him. It was probably not consonant with the poet's
purpose to mention sacrificed animals, kitchen equipment or even a
burial chamber; we cannot assume that he was unaware of the part
they played because he does not describe them. However, the fact that
Scyld's funeral ship had a mast, and, by implication, an efficient sail,91
and its attribution to a Dane, may indicate that the poet had a Viking
burial in mind. If it had been an earlier burial (Vendel or East
Anglian) which he recalled, would it not have been more appropriate
for Beowulf himself? That Viking chiefs had ship-burials may well
have been common gossip among the English, who would not always
have distinguished between Norwegians (who did) and Danes (who
did not). Indeed, Ragnar Lothbrok's 'Danish' family appears to have
originated in the ship-burial area of south-east Norway. 92
All ship-burials differ from Scyld's funeral, however, in that their
destination was known - and, indeed, if comparison is made with
archaeological remains, it could hardly be otherwise. For a ship of the
dead which sailed out to an unknown destination across the sea we
have to turn to literary sources, beginning with the Norse Sagas.93
Sigvardus Ringo (father of Ragnar Lothbrok) in the Skjoldunga
Saga,94 and King Haki in the Ynglinga Saga, ch.23,95 when mortally
wounded, ordered themselves to be placed on board ships which were
then launched, and sailed out to sea, driven by an offshore wind. King
Haki's ship was laden with weapons as well as with dead men; and
Sigvardus Ringo's funeral had 'royal pomp', so that his ship was
certainly also envisaged as laden with treasures.
91 The earliest Scandinavian evidence for a mast is late sixth century, on 'type B ship
representations from Gotland'; however, sails get bigger and masts presumably stronger on the
later picture-stones (Lindqvist, Bildsteine, i. 62-73, dating 108-23). Mast and sail are depicted on
an early seventh-century strapend from northern France; Bruce-Mitford (ed.), Sutlon Hoo Ship
Burial I, 422, 433; D.M. Goodburn, 'Do we have Evidence of a Continuing Saxon Boat-building
Tradition?', International Journal of Nautical Archaeology, 15 (1986), 39-47, at 45-6. However,
ships found in burials do not appear to have had masts before the Viking Age, and maybe it was
not until then that keels and sturdy masts became a fixed feature of warships as well as trading
vessels; see Richard Hodges, Dark Age Economics (London, 1982), 95-100. The Bayeux Tapestry
shows a warship lowering its mast, which (Detlev Ellmers comments, Fruhmittelalterlidie
Handehschiffahrt in Mittel- und Nordeuropa (Offa 28, Neumunster, 1972), 128, 143) would have
been far harder for a merchant ship at this period.
92 Smyth, Scandinavian Kings, 17-21. The classic case of confusion is the report in the
Chronicle of the first Viking raid in the south of England, with the comment 'these were the first
ships of Danish men', whereas the E Chronicle describes them as from the Hardanger district in
Norway; Earle and Plummer, Chronicles, i.54-5; ii.59.
93 Stjerna, Essays, 112. Balder's funeral ship, too, was laden with great treasures and
launched, but is not described as sailing off; Snorri Sturluson, Edda, Gylfaginning, ch. 49; ed.
Faulkes, 46; transl. in Garmonsway and Simpson, Analogues, 347-8.
94 As translated into Latin by Arngrimr Jonsson in Ad Catalogum Requm Svecia, Arngrm
Jonae Opera, i, 463; notes iv. 107-9, 260. Text and transl. in Chambers, Beowulf, 68-9; transl. in
Garmonsway and Simpson, Analogues, 345-6.
95 Snorri, Heimskringla, I, ed. ASalbjarnarson, 45; Garmonsway and Simpson, Arudogi®<
348-9.
DATING OF BEOWULF
29
Though these funeral ships had an unknown destination, they
were set alight, as Scyld's was not. Whether such fiery funerals ever
really took place is impossible to know; but it is much more difficult to
imagine that a king's body and accompanying treasures would have
been simply pushed out to sea, where they would have been in danger
of returning, or of falling into the hands of strangers or even enemies,
who might maltreat the one and plunder the other. 96 Such funerals
appear to belong strictly to the realm of legend; and Stjerna made
comparisons with Celtic stories in which dying kings were carried
away on boats attended by ladies; and, more relevantly, with the 'Lady
of Shallot, in the romance of Lancelot, [who] orders her body to be
laid on a ship, richly adorned, and that the ship shall be allowed to
drift whither the wind carries it, without any helm.'97
More recently - but now nearly twenty years ago - the Celtic
connections of Scyld's funeral seemed to be reinforced when Angus
Cameron compared it to the death-bed instructions of St Gildas, as
described in his Vita by a monk of Rhuys in Brittany (where the saint
was claimed to have spent most of his adult life). 98 The relevant
passage runs:
Deinde discipulos protestatus est dicens: Per Christum vos filios meos moneo, ne
contendatis pro corporis mei cadavere, sed mox ut spiritum exhalavero, tollite me et in
navim deponentes supponite humeris meis lapidem ilium, super quern recumbere
solitus eram: nemo autem ex vobis in navi mecum remaneat, sed impellentes earn in
mare permittite ire, quo Deus voluerit. Providebit autem Dominus sepulturae mihi
locum, ubi fuerit ei placitum . . . Et reddidit spiritum quarto kalendas Februarii
senex et plenus dierum."
These directions must first be put into their hagiographic context:
indeed, for a saint to be miraculously transported on an unsteered ship
appears to be one of those motifs which (like St Gildas' funeral ship)
tends to drift where God willeth, and come to rest wherever appropriate. 100 However, to judge from their Vitae, the funerals or transla-
96 Girvan, Seventh Century, 33-4.
97 Stjerna, Essays, 104, 107
98 'Saint Gildas and Scyld Scefing', Neuphilologische Mitteilungen, 70 (1969), 240-6.
99 Then he testified to his disciples, saying: I charge you through Christ, my sons, not to
contend for the remains of my body; but as soon as I have given up the spirit, bear me away, lay
me down in a ship, and under my shoulders place the stone on which I was wont to lie down. Let
no one of you remain with me in the ship, but push the ship into the sea, and let it drift whither
God willeth. The Lord will provide for me a place of burial, where it shall seem good unto Him
. . And he gave up the ghost on the 29th January, an old man and full of days,' Vita Gildae
Auctore Monacho Ruiensi (Vita /;, in Gildae De Excidio Britanniae, Fragmenta, Liber de
Paemtentia, ed. Hugh Williams (Cymmrodorion Record Series, 3, London, 1899), 317-89, at
366-8; also Ferdinand Lot (ed.), Melanges d'Histotre Bretonne i\'Ie-XIe Sicdc) (Paris, 1907),
207-83, 431-76; at 458-9.
100 Hippolyte Delehaye (The Legends of the Saints, transl. Donald Attwater (London, 1962),
23) claims There is no theme more hackneyed in popular hagiography than the miraculous
arrival of
the body of a saint in a derelict vessel; nor anything more commonplace than the
miraculous stopping of a ship
. in order to ... confirm a church in the lawful ownership of a
30
BULLETIN JOHN RYLANDS LIBRARY
tions of a little group of Rhenish saints - Werenfrith, Maternus,
Lubentius and Arnulf- most closely resemble Gildas', in that their
bodies (or relics thereof) were placed alone on a ship, so that God
might manifest where they should be buried.
The motif appears in its most naturalistic form in the Life of
Werenfrith, an Anglo-Saxon missionary in the Netherlands (a companion of Willibrord) who died in the mid-eighth century, according
to his Vita while preaching at Westervoort. He was revealed as a saint
by the fragrance of his corpse, possession of which was then disputed
between the inhabitants of Westervoort and those of Elst, the centre of
his mission. Certain better ones of the priests (quidam presbyterorum
meliores) cited the story of the Ark of the Covenant which had been put
on a driverless waggon to reveal God's will. But in order to discover on
which side of the Rhine the saint's body should rest, it was first put
alone on a ship, which went straight across to the other side, as if it had
an angelic oarsman. There, the corpse was put on a waggon drawn by
two cows unused to the yoke, who took it straight to Elst, where it was
buried with honour. 101
Maternus was apparently a historical early fourth-century bishop
of Cologne, about whom an 'extravagant legend' developed, in which
he was sent to Gaul by St Peter himself. 102 According to it, Maternus
died at Cologne, and people came from both Tongeren and Trier
(which he had also evangelized) to claim his body. Instead of the
'better ones of the priests', there was one angelic counsellor, and the
unmanned ship with the saint's body moved against the current, to
land at Rodenkirchen, just outside Cologne but in the territory of
Trier, where it was taken for burial. 103
A Lectionary of Saints tells us that Lubentius was martyred in his
parish at Kobern, but God prevented his burial there because of the great
sins of the inhabitants. His corpse was placed alone on a ship, which
descended the Moselle to its confluence with the Rhine, then (miracle of
miracles!) went upstream to the confluence with the Lahn, and then up
the Lahn to Dietkirchen, where he was honourably laid to rest. 104
Finally, there is Arnulf, 105 who was recognized as a saint only by
saint's relic.' He, Hermann Usener (Die Sintfluthsagen (Bonn, 1899), 136-7) and C. Grant
Loomis (White Magic (Cambridge, Mass., 1948), 90-1) have cited several examples, all of which
I have checked for similarities to Scyld and to Gildas; none appears to me so close as those
concerning the saints mentioned in the text.
101 Acta Sanctorum Augusti, vi (Antwerp, 1743), 103. The outline of this story is to be found in
Usuard's Martyrology of the mid-ninth century. For Usuard, see Rene Aigrain, L' Hagiographie
(1953), 62-3.
102 See discussion in Alban Butler's Lives of the Saints, rev. Herbert Thurston and Donald
Attwater (4 vols., 1956), iii. 552-3; The Catholic Encyclopedia, ed. Charles G. Herbermanneta/.
(16 vols., New York, 1913-14), v. 594-5 (sub Eucharius).
103 Acta Sanctorum Septembris, iv (Antwerp, 1753), 392.
104 Acta Sanctorum Octobris, vi (Antwerp, 1794), 200.
105 Feast Day 24 July; not to be confused with the St Arnulf whose Feast Day is 18 July; see
Acta Sanctorum Julii, v (Antwerp, 1727), 588-9.
DATING OF BEOWULF
31
Scale
1:2,500,000
Dietkirchen
MAP 3:
Places named in legends of post-mortem voyages of saints
32
BULLETIN JOHN RYLANDS LIBRARY
miracles at his tomb. His marvellous translation is mentioned under
970 in the Chronicle of Mosomum (Mouzon on the Meuse), which was
taken as far as the year 1033. After many vicissitudes, the saint's relics
were placed on a ship at Warcq; but while the otherwise empty vessel
was tied up, waiting for Archbishop Adalbero of Rheims to decide
who would go on it, an eagle flew down and seated itself on the prow,
and soon divine power moved the ship against the flow of the river, to
Mouzon.
It has not been possible, on the evidence that I have been able to
accumulate so far, to see plainly the chronological spread of this motif
from one saint to another; but the naturalism of the Werenfrith version
indicates that it is the earliest extant, and perhaps even the original once a miracle has been accepted for one saint, later writers have to
outdo it with exaggerated details. The motif is therefore hagiographic
rather than Celtic; given its distribution, it may well have first arisen in
the area of the lower Rhine among the Anglo-Saxon missionaries, and
only later have spread further afield. By what means and at what date
it reached Rhuys in Brittany is probably irrecoverable, but I do not
think it would have been before the ninth century at the earliest.
Gildas' version resembles Werenfrith's and Maternus' in having
the element of contention, but some features differentiate it from all
the others. The saint, awaiting his death on the island of Houat where
he is said to have lived for some years formerly as a hermit, gave
instructions for his body to be placed on the ship not so that God
should decide between rival claimants connected by the flow of a
river-system, but so that it should sail across the open sea. His
instructions were carried out, but the ship sank while monks from
Cornugallia (probably Cornuailles in Brittany106) were planning to
seize it. Three months later, in answer to their prayers, the Rhuys
monks found it in a creek, with the saint's body 'whole and unharmed,
just as it was when placed in the ship by themselves', and took it to
Rhuys. The writer of the Vita may have been attempting to combat the
claim that his body was entombed near Carnoet in Cornuailles. 107
Cameron picks out six points of similarity between Gildas' and
Scyld's funerals: (1) that the form of the funeral is due to the specific
instructions of the dying man; (2) that he was to be placed on a ship;
(3) that he was to be given treasure (Gildas' bedstone was all he had);
(4) that the ship was to be allowed to go where God willed (compare
Beowulf's in Frean w<ere, line 27); (5) that no-one would know its
destination; and (6) that it would have been a winter funeral. 108 This is
an impressive list, but the destination of Gildas' ship was surely not to
remain forever unknown - the revealing of God's will is the whole
n.2.
368 n./.
Gildas, iOK
106' Williams, Uildas,
107 Arthur de la Borderie (and B. Poquet), Histoire de la Bretagne (6 vols, Rennes and Pans,
396-1914), i.440.
1896-1914),
108 Cameron, 'Gildas', 244-5.
DATING OF BEOWULF
33
point of the hagiographic motif. Moreover, Beowulf shares all these
elements except the last (minor) one with some or other of the parallels
already cited: the second and fourth with the Rhenish saints; the first
and perhaps the fourth with the Lady of Shalott; and the first, second,
third and fifth with Sigvardus Ringo and Haki.
In fact, it is only the presence or absence of cremation, and the
change of tone consequent on the one burial being that of an ascetic
hermit-saint, and the others those of powerful pagan autocrats, which
cause the differences between Gildas' and the Saga funerals. Since the
Saga accounts are so much later, could they (and, indeed, the funeral
journey of the Lady of Shalott) have been affected by the hagiographic
motif? Or could Gildas' death-bed instructions have been affected by
Viking custom? Could the emphasis on Gildas' bedstone, for example,
have been intended to contrast with the Vikings' feather beds?
For much of the ninth and tenth centuries, local chroniclers
recorded Viking raiders who must have passed by the Rhuys peninsula. 109 In 835 the Vikings raided the monastery on the island of
Noirmoutier at the mouth of the Loire, and in 843 sixty-seven ships of
men from Vestfold (the district of the Norwegian ship-burials) sacked
Nantes. In 846 they overwintered at Noirmoutier, now abandoned by
its monks. Ten years later, the Vikings were again outside Nantes;
some went on to Brittany. They must have passed Rhuys on their way
to sack the rich abbey of Redon in 868; and in 886 they again
devastated Nantes. In 888 Alan the Great of Vannes organized a
successful resistance, but in 911 the area to the east of Brittany was
ceded to the Vikings and became Normandy. After this it is not always
easy to tell if the attacks on Brittany are by new Viking raiders or by
these settlers. In 912 the Vikings attacked the Loire again, and in 913
or 914 'Danes' destroyed the abbey of Landevennec near Brest. In 919
all Brittany was devastated, and many of the nobility fled (either then
or in 931). The monks of Rhuys took their holy relics to Berry.
Brittany remained under Norman control until 937; but it was not
until 939 that Duke Alan Barbatorta decisively defeated them; and
after his death the Viking attacks began again. The monastery of St
Gildas de Rhuys was not re-established until 1008.
Moreover, not far to the east of the Rhuys peninsula an unrecorded Viking was cremated on the lie de Groix. Scattered through a
thick layer of ashes under a cliff-top mound were more than 800
clincher rivets and 200 nails from a large Viking ship - with a mast
probably of non-local pine - and fragments of about twenty shield109 The following abbreviated account is based on Annales Engolismenses, ed. G.H. Pertz.
(MGH Scriptores, xvi, 1859), 485-7; Ex Miraculis S Filiberti auctore Ermentano, ed. O.
Holder-Egger (MGH Scriptores, xv. 1, 1887), 297-8, 302; La Chrmique de Nantes, chs.v-vii, xxi,
xxvii-xxxi, xxxviii, ed. Rene Merlet (Paris, 1896), 11-22, 66-8, 81-96, 111-12; Les Annales de
Fbdoard, ed. Ph. Lauer (Paris, 1905), 1, 50-2,63; Holger Arbman, The Vikings (London, 1961),
78-82; Gwyn Jones, A History of the Vikings (Oxford, 1968), 210-11, 215; Jean Delumeau (ed.),
Histmre de la Bretagne (Toulouse, 1969), 136-9.
34
BULLETIN JOHN RYLANDS LIBRARY
Arzon
StCildas-dc-Rhuis
<* Houat
/// Forest
%
j§
c=t>
0
Monastery
Bishop's seat
Viking raids 819-920
50
Noirmoutin>
100km
Indret A
3
StPhiiib«fl-d»-Grandli»u-0*oi
MAP 4: Viking raids on Brittany
bosses, along with cremated bones from an adult male, from a younger
person (a woman?), from dogs and 'some small birds'. Fine jewellery,
'rich cloth of gold', bronze bowls, personal effects, riding and kitchen
equipment were also recognizable. In the centre was an unburnt iron
cauldron surrounded by weapons and smith's tools. The excavators
commented on the similarity of the burial to those in the Vestfold and
Vestland districts of Norway, but Arbman conjectured that the Viking
interred here was (to judge from the continental weapons and other
material found) a sea-king, perhaps 'a second-generation Scandinavian
who had lived more or less permanently on the island' - whose
strategic position must have made it 'an excellent den for pirates.' 110
110 P. du Chatellier and L. Le Pontois, 'A Ship Burial in Brittany', Saga-Book, 6 (1908-09),
123-61; see also inventory in H. Shetelig and Anathon Bjorn, Viking Antiquities in Great Britain
and Ireland, iv (Oslo, 1940), 109-14; discussion by H. Arbman in 'En Sjokonungs Grav',
Arkeologiska Forskningar och Fynd; Studier utgivna med anledning av H.M. Konung Gustav VI
Adolfs Sjuttiodrsdag (Stockholm, 1952), 326-34, and in Vikings, 82-4. I have been unable to
consult du Chatellier and Le Pontois, 'La Sepulture Scandinave a Barque de 1'Ile de Groix,
Extrait du Bulletin de la Societe Archeologique du Finistere, 35 (1908), 3ff. For some comparable
cremation burials in Norway, see Shetelig, 'Ship Burials', 339-45.
DATING OF BEOWULF
35
The ship-cremation appears to belong somewhere in the middle of the
tenth century, 111 but if there had been a Viking settlement on the
island for more than a generation, it is quite possible that the monks of
Rhuys could have learnt something of Viking customs from it as well
as from the recorded raiders before their enforced move to Berry.
Something virtually identical to Gildas' proposed ship-funeral
must have been in the poet's mind, 112 for it neatly supplies all the
details in the Prologue not found in the English or Scandinavian
sources; and I have demonstrated this by placing against the lines of
the quotation from Beowulf (above, pages 9-10) letters indicating
sources I think are being used. Without similarity of names (as with
^thelweard's story of Scef), or of diction (as with the version of the
Visio Fault), a direct connection between Beowulf and the Vita Gildae
cannot be proved, but remains very probable. Any direct influence,
however, must have been towards Beowulf, not away from it, for, even
if affected by Viking custom, Gildas' funeral directions constitute a
version of a known hagiographical motif. In any case, influence from
an allusive poem in English on a Breton saint's life in Latin is virtually
unimaginable, whereas the reverse is as feasible as the influence of the
Visio Pauli on Beowulf. 113
Unfortunately, scholars are divided in their opinions about the
date of composition of Gildas' Vita. Ferdinand Lot believed that the
Life was compiled as a whole by Vitalis, the second abbot of the
revived monastery - about sixty years after the writing of the Beowulf
manuscript. 114 Hugh Williams, however, argued for a date in the 880s
for the account of Gildas' life and death, and considered the chapters
concerning the subsequent history of the Rhuys foundation as a later
111 Viking archaeologists are not agreed on the date of the lie de Groix burial; M. Miiller-Wille
puts it in the second half of the tenth century ('Das Schiffsgrab von der He de Groix (Bretagne) Ein Exkurs zum "Bootkammergrab" von Haithabu', Berichte iiber die Ausgrabungen in Haithabu,
12: Das archdologische Fundmaterial III der Ausgrabung Haithabu (Neumiinster, 1978), 48-84);
James Graham-Campbell in the first half (personal communication).
112 I cannot forbear to speculate that what may have called it forth was the poet's reading the
last entry in the pre-ae Chronicle: the account of the three Irishmen who came in 891 to King
Alfred in a boat without oars, because they wanted for the love of God to be in exile, they did not
care where. This speculation is unfortunately too uncertain to provide another reason for dating
the composition of the prologue after 891. jEthelweard has more detail concerning this visit, but
at least one mistranslation and some obfuscation of language; therefore this note relies more on
the Old English account, the essence of which must have been in the pre-ae Chronicle; Campbell,
lEthelweard, 48; Earle and Plummer, Chronicles, i.82.
113 Similarly slender evidence for the influence of a Breton Saint's Life on an Anglo-Saxon text
is provided by the use of the term Ormesta for Gildas' De Excidio in Cambridge, Pembroke College
MS 25, ed. James E. Cross (King's College London Medieval Studies, i, 1987), 65-6, 156 - only
paralleled in the Vita of St Pol de Leon. The possibility of another link between the Vita Gildae
and Beowulf is still more tenuous, yet needs to be mentioned. Gildas too was a dragon-slayer, but
had a much easier time of it than Beowulf, and there appears to be nothing which forces us to
connect the two. Hearing that the inhabitants of Rome were being killed by the noxious breath of
a dragon which lived in a cave in a mountain, Gildas got up early, climbed with a staff in his hand
to the mouth of the cave and there commanded the dragon, in the name of Christ, to die, which it
did; Williams, Gildas, 344-5.
114 Lot, Melanges, 230-8.
36
BULLETIN JOHN RYLANDS LIBRARY
addition. Only a detailed stylistic study can resolve the matter; but for
the present I accept Williams' arguments: chapters 1-31 appear to me
to resemble the oldest Breton saints' lives (of the late ninth century), 115
and are different in tone and preoccupations from the later chapters.
Moreover, the circumstantial account (however fictitious) of how the
Rhuys monks obtained Gildas' remains would have had far more point
if it were written before they were moved to Berry.
Though individual links between Brittany and England certainly
took place earlier, 116 it is about the time of the abandonment of Rhuys
that the clearest evidence for contact begins. About 919, a 'multitude
of Bretons' fled from the Danes to England with Mathuedoi, count of
Poel, whose son Alan (later called Barbatorta), became Athelstan's
godson, and may have been brought up at his court. Refugees appear
to have gone to England again about 931; 117 and in 936 Alan returned
with Athelstan's help. Athelstan had many Breton relics, and many
manuscripts written in Brittany came to England at this time. 118 A
great deal of Breton ecclesiastical influence (which begins earlier than,
and is independent of, the 'Benedictine Revival') is found in tenthcentury English texts, especially in Winchester liturgies and in
charters and penitentials. 119
However, there were also strong links between Brittany and the
monastery of St Benoit-sur-Loire at Fleury in the tenth and eleventh
centuries. The monks of St Pol-de-Leon took their patron, his Vita
and other manuscripts there for refuge. Gildas' Life was extant only in
a manuscript from the Fleury library, 120 which may even have had a
115 Williams, Gildas, 317-21. For the characteristics of the oldest Breton saints' lives, see
Francois Kerlouegan, 'Les Vies de Saints bretons les plus anciennes dans leur rapports avec les
lies britanniques', Insular Latin Studies, ed. Michael Herren (Pontifical Institute of Mediaeval
Studies, Papers in Mediaeval Studies, 1, Toronto, 1981), 195-213.
116 Edward the Elder is said to have entered into confraternity with St Samson's monastery at
Dol (Joseph Armitage Robinson, The Times of Saint Dunstan (Oxford, 1923), 73). A letter
survives from a Breton who went to England in time of peace, and obtained permission from
Athelstan to remain; I am grateful to Caroline Brett for allowing me to see her paper, 'A Breton
Pilgrim in England in the Reign of King Athelstan', before publication.
117 It seems necessary to postulate two separate refugee periods: in 919 Athelstan was not yet
king; but in 931 Alan Barbatorta (who led an army in 936) would have been too old for baptism;
see Dudonis Sancti Quintini De Moribus el Actis Primorum Normannite Ducum, ch.41, ed. Jules
Lair (Memoires de la Societe des Antiquaires de Normandie, 23.2, Paris, 1865), 71,185; Merlet,
Nantes, chs. xxvii, xxix, 82, 89, n.4. Some of the refugees may have been buried at St Mary's
church in Wareham; see Edmund McClure, 'The Wareham Inscriptions', English Historical
Review, 22 (1907), 728-30.
118 Helmut Gneuss, 'A Preliminary List of Manuscripts written or owned in England up to
1100', Anglo-Saxon England, 9 (1981), 1-60, nos.61, 119, 295, 444, 459, 532, 629, 794. The
subject is discussed in David Dumville's O'Donnell Lectures, England and the Celtic World in the
Ninth and Tenth Centuries (forthcoming); see his 'English Libraries before 1066: Use and Abuse
of the Manuscript Evidence', Insular Latin Studies, ed. Herren, 153-78, at 169-70, 177.
119 See, e.g., F.A. Gasquet and Edmund Bishop (ed.), The Bosworth Psalter (London, 1908),
53-7; Robinson, Times, 73-5; Donald A. Bullough, The Educational Tradition from Alfred to
yElfric: Teaching Utriusque Linguae', Settimane del Centra Italiano di Studi Sull' Alto Medwevo, 19
(1972), 453-94, at 472-6.
120 Williams, Gildas, 317.
DATING OF BEOWULF
37
copy of the earlier Life for Vitalis to augment. There was further
strong influence from Fleury on the English Benedictine Revival,
which may have begun soon after the restoration of St Benoit by Odo
of Cluny in 930, for Oda (Archbishop of Canterbury, 940-58) is said to
have made his profession as a monk there, and certainly chose to send
his nephew Oswald there for instruction in monastic discipline. 121
To sum up: the Prologue to Beowulf has one clear source which
provides one clear terminus post quern: the genealogy of ^thelwulf
from the pre-ae Chronicle translated by ^thelweard, which could not
have been fabricated before 858, and probably was not before Alfred's
reign. The Prologue may also have been influenced by the account of
Gildas' funeral directions in his Vita., written by a monk of Rhuys
probably around 900. 122 Knowledge of the Vita is most likely to have
reached England sometime after about 920. Two elements in the
Prologue which are less precisely datable are the legends of the Danish
royal house, which may have been known in England in earlier
centuries, but which certainly would have come there along with the
'Great Army' in the 860s or any time thereafter; and the knowledge of
ship-burial ritual, which was available in East Anglia in the sixth and
seventh centuries, but may have been reinforced during the Viking
Age. All in all, the reign of Athelstan (924-55) appears the most
probable for the composition of the Scyld Scefing Prologue to
Beowulf. 123
As to the audience towards whom the Prologue was aimed,
Patricia Poussa's suggestion of the Christianized inhabitants of the
Danelaw deserves attention. 124 The history of the remarkable family
whose founder (Byrhtferth tells us) was said to have come over with
Hinguar (Ivar), and which included an archbishop of Canterbury
(Oda, 940-58) and two archbishops of York (Oscytel, 956-71, and
121 Vita Oswaldi, ed. James Raine, Historians of the Church of York, I (Rolls Series, 71,
London, 1879), 399^75, at 413; see Joseph A. Robinson, St Oswald and the Church of Worcester
(British Academy Supplemental Papers, 5, London, 1919), 41-2. It is tempting to conjecture that
Oda's contact with Fleury began in 936 when (according to Richer) he is said to have gone to
Frankia to negotiate the return of Louis d'Outremer; see Robert LaTouche (ed.), Richer, Histoire
de la France (888-995), ch. ii.4 (2 vols., Paris, 1930), i. 130-1; Dorothy Whitelock, English
Historical Documents, I, c.SOO-1042 (2nd ed., London, 1979), 344 n.5.
122 The Andreas poet (often believed to have been influenced by Beowulf) may have intended
to give Andreas a funeral like Scyld Scefing's; the interpretation of the passage is difficult; see
C.M. Doherty, The Journey-Motif in the Longer Old English Poems' (Unpublished M.A.
dissertation, Macquarie University, N.S.W., 1987), 87 n.
123 R.L. Reynolds ('An Echo of Beowulf in Athelstan's Charters of 931-33 A.D.?', Medium
/Evum, 24 (1955), 101-3) has suggested that the proems of some of Athelstan's charters show
familiarity with the poem.
124 Patricia Poussa, The Date of Beowulf Reconsidered: The Tenth Century?', Neuphilologische Mitteilungen, 82 (1981), 276-88. See also Dorothy Whitelock, The Conversion of the
Eastern Danelaw', Saga-Book, 12 (1937-45), 159-76; W.S. Angus, 'Christianity as a Political
Force in Northumbria in the Danish and Norse Periods'; and Binns 'York Viking Kingdom',
both in Fourth Viking Congress, ed. Small, 142-65 and 179-89 respectively; D.M. Wilson, The
Vikings' relationship with Christianity in Northern England', Journal of the British Archaeological
Association, 3rd ser. 30 (1967), 37^6.
38
BULLETIN JOHN RYLANDS LIBRARY
Oswald, 972-95), is proof enough of the penetration of these AngloDanes into tenth-century English life. 125 The composer of the Prologue was clearly not only well-versed in Scandinavian traditions, but
also probably Latin-literate and with the resources of a (monastic?)
library to hand. To judge from his authorial comments in lines 13-17
he also believed that God was watching over the destiny of the pagan
Danes. The patron (and perhaps the poet himself), then, may have
been proud of a Danish heritage, and desirous of seeing pagan
ancestors in the best possible light. 126
Most modern critics, when discussing the date of Beowulf, seem
to regard the poem as we have it in the Nowell Codex as the creation of
one man, with one date of composition. I wonder, however, if they are
not chasing a chimaera. 127 True, it is the only version we have any
evidence for; but that may be due solely to the accidents of survival.
No substantial poem in Old English is extant in more than one
manuscript; and the brief poems which are found in two sometimes
vary greatly. Often this is due to scribal corruptions; but two poems
found in early Northumbrian as well as late West Saxon versions have
apparently deliberate changes. In the Northumbrian Mail CoatRiddk
the last two lines are supported by the Latin original, but have been
replaced in the later version by a conventional riddle ending. 128 The
text of the Dream of the Rood found carved in runes on the eighthcentury Ruthwell Cross consisted of about fifteen lines; that written
into the Vercelli Book about AD 1000 has 156 lines. This may be
partly due to selectivity on the part of the rune-carver; 129 but in some
places the Vercelli Book version appears to have been internally
expanded, 130 and all the lines after 78 have been claimed as an inferior
addition. 131
Substantial Middle English poems preserved in more than one
manuscript also vary considerably. Sometimes this is due only to an
attempt to update and clarify the text; but a later scribe also drastically
125 Robinson, Oswald, 38-51. For Byrhtferth as author of the Vita Oswaldi, see Michael
Lapidge, 'Byrhtferth of Ramsey and the Early Sections of the Historia Region attributed to
Symeon of Durham', Anglo-Saxon England, 10 (1982), 97-122, further references at 99 n.ll.
126 Fred C. Robinson has pointed out that, though the poem is full of reminders that its
characters are heathen, it only depicts those aspects of paganism which would be the least
shocking to Christians; Beowulf and the Appositive Style (Knoxville, Tennessee, 1985), ch.2,
29-59.
127 The inspiration for the following argument derives from work by Bernadette A. Masters tor
her Ph.D. Thesis for the University of Sydney, 'Le Moulin a Paroles: Revaluation de PArt de la
Manuscripture au Moyen Age' (1987).
128 A.H. Smith (ed.), Three Northumbrian Poems (London, 1933), 9, 17, 46-7; Craig
Williamson (ed.), The Old English Riddles of the Exeter Book (Chapel Hill, N.C., 1977), 88-9,
,
243-8.
129 Michael Swanton (ed.), The Dream of the Rood (Manchester, 1970), 41, points to the
defective metrical arrangement of the Ruthwell Cross lines.
130 E.g., line 39; Ruthwell ongeredee hirue god alme^ttig = Vercelli ongyrede hine pa geong ham
(\xet was god almihtig).
131 B. Dickins and A.S.C. Ross (ed.), The Dream of the Rood (3rd ed., London, 1951), 18
DATING OF BEOWULF
39
abbreviated the immense Layamon's Brut,, often combining parts of
lines to form new ones which lack the alliteration of the earlier
version. 132 The three manuscripts of Sir Orfeo all differ widely;
moreover, in the earliest of them, the Auchinleck, the prologue
attached to Sir Orfeo in the other manuscripts is found with another
poem. 133 Diane Speed, in her recent edition of six medieval romances,
remarked:
there has been an increasing recognition of the difficulty of identifying any single
'correct' form for works such as these when the concept of authorship was not as
individual-oriented as our own and when both conscious and unconscious modifications almost certainly arose with each new performance or copying. 134
Scholars are right to emphasize that we can only know the poem
as we have it in the No well Codex; it would be pointless to try to
reconstruct the shape and content of any putative earlier versions. Yet
earlier versions of at least parts of the poem there must have been.
Even the strongest supporter of early eleventh-century composition,
Kevin Kiernan, assumes the use of earlier material when he claims that
the join between two poems, one on Beowulf s youth, the other on his
old age, was only made in the extant manuscript. 135 H.L. Rogers, on
the other hand, assumed a reworking of the whole of the originally
pagan material by a Christian poet, claiming that it had been most
effective for Beowulf s fight with Grendel, and least effective for the
dragon fight. 136 It is unlikely that scholars will ever agree what forms
any 'Ur-Beowulfs' may have taken; but clearly it is not out of the
question that the poem we have in the Nowell Codex is the product of
development over a considerable period. A Prologue concerned with
the ancestors of the royal Danish house, and a description of Grendel's
mere influenced by a version of the Visio Pauli may (or may not) have
been among the latest additions or alterations before the poem was
written into the Nowell Codex somewhere about the year 1000.
I end as I began, with Rowland Collins:
A poem as rich and as reminiscent as Beowulf could only arise from important
societal traditions and narratives and it could only reach the literary form in which we
have it after long gestation and sustained social thought. The traditions which
132 La^amm: Brut, ed. G.L. Brook and R.F. Leslie (EETS, 250, 277, 2 vols., Oxford,
1963-78). As an example of thorough-going reworking, see the story of Hengist and Horsa; from
about line 6880 until about 7360 the scribe of London, British Library Cotton Otho c.xviii has
some rewording in every line, as compared to Cotton Caligula A.ix.
133 See the edition by Alan J. Bliss, Sir Orfeo (Oxford, 1954), especially ix-xvii, xlvi-xlviii.
134 Diane Speed, Medieval English Romances (2 vols., Sydney, 1987), ii. 323. David N.
Dumville has also drawn attention to the effects on a piece of literature of a period of oral
transmission, such as cannot be ruled out for at least some of the Beowulf material; 'Beowulf and
the Celtic World: The Uses of Evidence', Troditw, 37 (1981), 109-60, at 149-54.
135 In Dating, ed. Chase, 17.
136 H.L. Rogers, 'Beowulf s Three Great Fights', Review of English Studies, New ser. 6 (1955),
339-55; repr in An Anthology of Beowulf Criticism, ed. Lewis E. Nicholson (Notre Dame, 1963),
233-56.
40
BULLETIN JOHN RYLANDS LIBRARY
contributed to this masterpiece were not welded to each other mechanically or
quickly. Whether the poem as we now have it was conceived in the eighth, ninth
tenth, or even early eleventh century does not affect either the fact that the manuscript
was produced not far from the year 1000 or the certainty that the poem builds its
narrative from the ancient Germanic antecedents of the English people. This
suggested documentation of some growth in the poem . . . [underlines] the great
richness, the consummate skill, and the magnificent inventiveness which the author
and his assistants (whether original, editorial or scribal) brought from the centuries of
Germanic, Christian, and English tradition. 137
In the present state of research we can see this 'magnificent inventiveness' welding together disparate traditions more clearly in the
Prologue than anywhere else in Beowulf. In spite of all the conjecture,
what is certain is the consummate skill of the poet who wove together
references from English and Danish ancestral legends (whether traditional or fabricated), from his knowledge of ship-burial ritual and from
hagiography to provide a seamless introduction which was appropriate
at all levels for the greatest of the Old English poems.138
137 Collins, 'Blickling XVI', Medieval Studies, ed. Bald and Weinstock, 69.1 do not wish here
to go into the question of the method and date of the composition of Beowulf as a whole; one
fruitful approach is expressed in a preliminary way by Ruth Finnegan, 'Anthropologists and the
Sociology of Literature: An Example from the Study of Epic', Journal of the Anthropological
Society of Oxford, 3 (1972), 53-66.
138 The author acknowledges the generosity of Professor Michael Muller-Wille in allowing
three of his maps to be adapted for this article: Maps 1 and 2 are from 'Boat-graves in Northern
Europe', InternationalJournal of Nautical Archaeology, 3 (1974), 187-204; Map 4 from'Das
Schiffsgrab von der lie de Groix (Bretagne) - Ein Exkurs zum "Bootkammergrab" von
Haithabu', Berichte uber die Ausgrabungen im Haithabu, 12: Das archdologische FundmateriallH
der Ausgrabung Haithabu (Neumunster, 1978), 48-84.