Introductions to the Old English poem called Beowulf often begin with something of the sort: 'Beowulf is written in West Saxon and recorded only in the Cotton Vitellius A. xv manuscript…' One may wonder why such a work would be introduced in this rather dry and relatively uninformative manner. Unfortunately, very little can be said definitively as regards the poem's authorship, date or location of composition, purpose, theme, &c.

On definitive ground, we can describe Beowulf as the longest surviving poem in Old English and one of the earliest European epics written in the vernacular (rather than in Latin). Written in unrhymed, four-beat alliterative metre of Old English poetry, it tells of the exploits of the hero Beowulf. The first part of the tale narrates Beowulf's youthful adventures in Denmark battling the monstrous creature Grendel on behalf of the King Hrothgar of the Danes, and the second part narrates his later life, including his fight with a fire-dragon, during his reign as the King of Geatland (traditionally located somewhere in southern Sweden or one of the Baltic island on the east coast of Sweden).

We may say that Beowulf was composed somewhere in England between about 521 AD (the approximate date of the death of the historical model for the character Hygelac) and 1026 AD (more or less the latest possible date of the manuscript itself). We do not know for sure where in England the poem was composed. Nor do we know if the poem was composed by a single author, or whether it is the result of the merging together of ballads by different authors, nor whether the poem was significantly altered subsequent to its first written form. The poem's purpose is also unclear - arguments have been made for a naturalistic mythic allegory, a Christian allegory, a criticism of heroic culture, a mourning for the loss of heroic culture, a Germanic 'Old Testament', an allegory concerning contemporary politics in one or other of the Saxon kingdoms - just to mention a few. The title Beowulf itself is an editorial convenience -- the manuscript copy of the poem is untitled. We also know almost nothing about Beowulf's place in English literature in the Anglo-Saxon period - we do not know what popularity, if any, the poem enjoyed. Certainly, awareness of the poem seems to have disappeared entirely by the early Middle English period, and the poem does not re-enter the canon of English literature until the late 18th and early 19th centuries - which places Beowulf in an odd ancient/modern position within the history of English literature.

Let it suffice that very little can be said with certainty on almost any aspect of the poem. This introduction is an attempt to provide an overview of what is known about the work and also to present some of the claims which have been made on the less certain aspects. I shall endeavour to provide a fair account of some of the major theories, though I reserve the editorial prerogative to suggest what seems most plausible to me.

---

I. The Beowulf-manuscript [13.04.02/last modified 16.12.2003]

  a. Provenance of Beowulf ms.
The author of Beowulf is unknown, as is the exact date of the composition of the poem. In its present form, Beowulf was possibly composed as early as the seventh century or as late as 1025. Beowulf survives in a single manuscript codex [British Library [Museum] MS Cotton Vitellius A.xv (Gneuss 399)]. This codex is a composite codex assembled in the first half of the 17th century, itself consisting of two Old English codices: the first, the Southwick Codex (so named in the 20th century by Kemp Malone); the second, containing the Beowulf MS, the Nowell Codex (so named, again by Malone, after its earliest known owner, Laurence Nowell [1520-1576], Dean of Lichfield and an early 'Saxonist' who compiled the first Anglo-Saxon dictionary and who wrote his name and the date (1563) at the top of the first folio of the codex. The Southwick Codex (again, itself almost definitely a composite codex) contains the four (prose) items: (1) The Soliloquia of St. Augustine (ascribed to King Alfred); (2) The Gospel of Nicodemus; (3) The Debate of Solomon and Saturn; (4) eleven lines of a St. Quintin Homily (fragment). The Nowell Codex contains the five items:
  (1) The Life of St. Christopher (prose fragment);
  (2) The Wonders of the East (prose, illuminated);
  (3) Alexander's Letter to Aristotle (prose);
  (4) Beowulf (verse);
  (5) Judith (verse fragment).

The Nowell Codex is also a composite book; and Kevin Kiernan offers convincing evidence that Beowulf itself originated as a separate codex (see below).

The Beowulf MS was written down circa 1000CE by two scribes in late West Saxon (the literary and posh dialect of the period). The first scribe, who writes in an Anglo-Saxon rounded insular minuscule hand with some carolignian features, copied the first three prose pieces of the Nowell Codex and a little over the first 85 pages of Beowulf (up to the word moste on line 1942 in this edition, on folio 172v). The second scribe, who writes a more conservative Anglo-Saxon square minuscule hand, copied the remainder of Beowulf (roughly 1238 lines) and the poem Judith which follows. Both scribes proof-read their writing, and the second scribe also proof-read the writing of the first scribe (see Kiernan for further details), as well as possibly making some later alterations to Beowulf (see II. below). There are currently 116 leaves in the Nowell Codex, Beowulf filling 70 of them.

The beginning of St. Christopher was already lost when Nowell put his name on the first leaf (and probably the bulk of Judith as well). Nowell probably preserved the manuscript after King Henry VIII's dissolution of the monasteries after the break with Rome. Its history prior to coming into Nowell's possession remains a blank.

During the reign of Queen Elizabeth I, the Nowell Codex passed into the library of Sir Robert Cotton [1571-1631], an antiquarian; whence its British Library designation, for Cotton catalogued his books according to busts of the Roman emperors that surmounted each of his bookcases, and in the case of the codex in question, it was presided over by Vitellius. The Cotton collection was presented to the British people in 1700 by Sir John Cotton, grandson of Sir Robert. A committee of the House of Commons became trustees of the bequest and appointed Humfrey Wanley as one of the members of a commission to report on the state of the library. Wanley reported that the location of the library, in the Cotton House, was unfit and eventually, in 1722, the collection was moved to Essex House, Strand and there it remained for seven years. Though an Act of Parliament (6 Anne, c. 30) had endorsed the construction of a new building to house the collection in 1706, the Cotton Library still had no permanent home. So in 1729, after Essex House was deemed unfit, it was moved to another interim residence: the Ashburnham House in Westminster. The ill-omened name did not fail to disappoint, and in October 1731, the house caught fire. Much of the library was damaged and some volumes completely destroyed--damage resulted not only from the flames themselves, but also from the water used to extinguish the fire; the Beowulf MS was presumably saved by being thrown from the window, though it is badly burnt along its outer edges (see Syd Allan's Ashburnham House Fire page and Andrew Prescott's online article '"Their Present Miserable State of Cremation": the Restoration of the Cotton Library' for more details on the fire). The collection finally made its way in 1757 into the care of the British Museum, where it was finally provided a much-needed rebinding in August 1845, as well as having its leaves inlaid within heavy sheets of paper--the latter of which, for the most part, halted the loss of letters and words from the crumbling edges of the MS, which had previously suffered this sort of damage continually from the late 1700s until 1845.

It is unclear exactly when Beowulf began to be studied in post-Saxon times. Nowell and Cotton may or may not have paid it any attention. The first modern reference to the poem is a brief description by Humfrey Wanley (see above) who mistakenly described it as a story about Beowulf 'the Dane' who feuded with Swedish princes. This inaccurate description drew the attention of the Icelandic scholar Grímur Jónsson Thorkelin, who worked in Denmark as an archivist, eventually becoming the Royal Danish Archivist. In 1786 Thorkelin went to England to research Danish heroes in British libraries; coming across Wanley's description, he hired a professional scribe to copy the manuscript (the document now known as Thorkelin A) and later himself made a copy (Thorkelin B). These transciptions are invaluable in that they preserve almost 2000 letters which are now lost from the edges of the scorched vellum. The poem was not printed in its entirety until Thorkelin's edition (with facing Latin translation) appeared in 1815; the first complete vernacular translation (in Danish) of Nicholas Grundtvig appeared in 1820; and not until 1837 did the first complete English translation of John Mitchell Kemble (from a family of prominent actors, himself a student of the German linguist, folklorist and political activist Jacob Grimm [of the 'Brothers Grimm']). Innumerable subsequent editions and translations (into many languages) have appeared since.

   b. Construction of the manuscript: foliation
We use Kiernan's foliation in referring to the folios of the Beowulf codex, the following table provides the equivalences between the 'old' foliation, the 'new' 1884 foliation and Kiernan's 1981 foliation [essentially the 'old' foliation, with the two misplaced folios, 131 & 197, put in their proper ordering and renumbered as 147A(131) and 197A(192) respectively]:

Beowulf foliation (numbering)

As mentioned earlier, the British Museum MS Cotton Vitellius A.xv is a composite codex, with the Nowell and Southwick codices originating as separate volumes. Kiernan suggests that both of these may themselves be composite codices, but here we are concerned with the Nowell codex (which contains Beowulf) only. This is no place to expound upon the construction of mediaeval manuscripts in detail, but a few basic facts may prove expedient: the codices are composed of sheets of vellum (specially prepared sheep-skin) which are generally folded in half and stitched together into gatherings, which are then themselves stitched together. The gatherings are constructed by placing a number of folded vellum sheets inside one another and stitching them together along the middle-fold (similar to the construction of many modern magazines & journals). Single (non-folded) sheets were sometimes also stitched into these quires (gatherings). As vellum is actually sheep skin, its two sides are not uniform in appearance, with the 'hair' side generally somewhat darker than the 'flesh' side. Unfortunately, the Ashburnham fire destroyed the majority of the original binding, making the determination of the composition of the codex uncertain.

Following Ker, the composition of the Nowell Codex has been assumed to be made up of twelve 4-sheet (bifolia, i.e. folded) quires, and two 5-sheet quires, as follows (using to Kiernan's foliation numbering):

This means that Beowulf begins in the middle of quire 5, suggesting that it was copied along with (and at the same time as) the three prose texts preceding it. Kiernan has proposed that Beowulf actually begins a new quire and thus that it may well have originally formed its own codex. Kiernan's proposal is as follows ('leaves', in the following table, refer to non-folded single sheets):

Thus, Kiernan's conclusion is that

'...Beowulf was initially a separate codex[: ] the intelligent, conscientious proofreading of Beowulf alone [i.e., and not the prose texts preceding it in the Nowell codex--BMS], by both scribes and in several stages, and the scribes' manifest efforts to copy accurately in the first place, are the best confirmation of all that Beowulf was a special poem in the early 11th century, and that it was copied in our extant MS as a separate codex'. (Kiernan 146)

Other codicological studies of the Beowulf-ms. arrive at different foliations, see further: Boyle, Clement, Gerritsen (1988, 1989, 1991), Malone.


   c. the Manuscript's Date
Additionally, there are some questions concerning the dating of the manuscript itself. The late palaeographic expert Neil Ker dates it to ca. 1000, usually interpreted as 975-1025 AD. Kiernan and Dumville both argue that in fact it is actually unlikely to have been written before 1000AD. However, Kiernan--for reasons based on codicology, palaeography, and his thesis that Beowulf itself is nearly contemporaneous with the extant MS.--places the most likely date to post-1016AD, after the ascension of Cnut to the throne of England.

David Dumville disagrees, stating:

   'Stylistically, the last observable phases of Square minuscule are to be found in manuscripts of the works of Abbot Ælfric (fl.990--ca1010), but otherwise in manuscripts which are rarely datable. Some folios containing additions, of 990 x 1010, to the Sherborne Pontifical are in a fine Square minuscule. No book (or charter) certainly datable by its contents to after A.D. 1000 is a specimen of Square minuscule....On the other hand, the new form of vernacular minuscule [e.g., like the hand of the 1st scribe of the Beo. MS.--BMS] is first found as a bookhand in the period 1001 x 1013....The clearest demonstration is provided by the script of annal 1001 in the Parker Chronicle, which must date from 1001 x 1013'. (Dumville 61-2, 54)
   'The new minuscule was not being employed as a bookhand before the first decade of the eleventh century. Square minuscule is likely to have been in use for only a very few years after A.D. 1000. The few manuscripts, like that containing Beowulf, which display contemporaneous writing in these two successive styles of Insular minuscule must therefore have been written very early in the eleventh century. There is neither evidence nor need to attribute a lingering death to Square minuscule. It is in the highest degree unlikely that the Beowulf-manuscript was written later than the death of Æthelred the Unready (1016) or earlier than the mid-point of his reign (which fell in A.D. 997)'. (Dumville 63)

Kiernan points out '[a] closely datable example of Square insular script survives in a chirograph of Bishop Byrhteh of Worcester (1033-38) leasing land to his cniht Wulfmær' (Kiernan xvii-xviii), though Greg Rose (136-9) argues that the hand is not the same type of Square minuscule as in the Beo.-ms., or that meant by Dumville.
 Kiernan also notes a number of interesting similarities between the Beo.-ms. and the ms. containing the Blickling Homilies (Princeton, Scheide Library MS 71):

   'Another case is the Blickling Homilies manuscript....which Ker assigns the same date [990-1040] for the same reasons. It is clear that neither manuscript comes from a scriptorium where uniformity of script was enforced or even encouraged, for in both manuscripts scribes with old-fashioned, Square scripts are paired with scribes with more up-to-date, Caroline tendencies. Scholars have known for over a century about the case of literary borrowing [or, at least, sharing--BMS] between the description of Grendel's mere in Beowulf and St. Paul's vision of Hell in Homily 16.... [These common features] would be most easily explained if the two manuscripts derive from the same scriptorium'. (Kiernan xix)
   'There are paleographic and codicological reasons to suspect that they do have a common provenance. ...Max Förster observed that "The hand of the second Beowulf scribe displays in overall appearance a striking resemblance to the first scribe of the Blickling Homilies, so that both must belong to approximately the same period." Förster undoubtedly recognized the many significant differences in specific letterforms, and Ker rightly omits the two manuscripts from his tally listing closely similar hands (lvii). Another difference is that the Blickling scribe was calligraphic, paying attention to details of his letterforms, whereas the Beowulf scribe was almost crudely utilitarian. ...[but] the similarities in general aspect are more immediately evident than the specific differences'. (Kiernan xix-xx)
   'Greatly contributing to the broad paleographical resemblance between the first hand of the Blickling Homilies and the second hand of Beowulf is the virtually identical size of the writing grids.....no other manuscript in the more than 400 described in Ker's Catalogue comes as close to Beowulf as the Blickling Homilies manuscript in the combination of line numbers and grid size. If both manuscripts derive from the same scriptorium, its one telling uniformity was that it produced books with text faces of the same relative size, even when the rulings vary from 20-22 lines per page'. (Kiernan xx-xxi)

Fulk, however, comments on this point that the variation in number of rulings is not as uncommon as Kiernan's discussion might suggest, for

   '...in the four Old English poetic codices, variation in the number of rulings from quire to quire is the rule rather than the exception. In the Exeter Book on hand wrote all the poetic quires, which have from twenty-one to twenty-three rulings (one time the number changes in the middle of a quire). The Vercelli Book is also in one hand, and the number of rulings varies from twenty-four to thirty-three (usually each quire has the same number of rulings throughout). Even Scribe B of the Beowulf manuscript is inconsistent, using twenty-one lines for quires twelve and thirteen, and twenty lines for Judith. Only Junius 11 shows any consistency: the main hand uses twenty-six lines throughout, in all sixteen quires plus one leaf.....in other words, the variation is unusual for this scribe [=first scribe of Beo.-MS], as far as his eleven Nowell quires go' (Fulk 348; emphasis is Fulk's)

Kiernan points to yet another interesting common feature of these two mss.-- that they are both

   '...quite inconsistent in the way they arrange the hair and flesh sides of the vellum sheets within gatherings... Hair and flesh sides often contrast noticeably in color and texture, and the normal insular practice for manuscripts of this period was to place hair sides against hair sides, and flesh sides against flesh sides, to obscure the contrast on facing pages. Our scribes [of the Beo.-ms.] ignore this trend. The same odd mixture of format occurs in the Blickling Homilies as in the Beowulf-manuscript: sometimes the contrast of hair and flesh sides is carefully observed, while sometimes it is deliberately displayed, with flesh on the outside of all leaves in a gathering in the Homilies, and hair on the outside of all leaves in Beowulf (second scribe, and first scribe, first gathering). This shared lack of uniformity, also evident in the ill-matched script, begins to look like a distinctive feature of a particularly provincial scriptorium'. (Kiernan xxi)

To conclude, the composition of the MS. itself we can probably safely place between 1000-1025 AD; the disagreement of professional palaeographers (as above) makes it difficult to feel confident in assigning a narrower band of dates.

   d. the Palimpsest? (further aspects of the structure of the Beo.-ms.)
One of the key issues of the MS, is the status of fol. 179, which is badly damaged and quite difficult to read, creating several uncertainties in the text. Julius Zupitza, in his facsimile edition of the manuscript, notes that '[a]ll that is distinct in the [facsimile] in fol. 179 has been freshened up by a later hand in the MS'. In the 1960s, Westphalen, on the other hand, identified the 'freshening up' hand as that of the second scribe, but about twenty years later than the original, and the folio itself as a palimpsest. A palimpsest is made, generally, to eradicate an old text in order to provide parchment for a new one. Kiernan, in the early 1980s, proposed that the palimpsest was made by the second scribe himself, in order that he might revise the text. Kiernan attributes the spots of illegibility of the folio to the fact that the vellum was still damp in places when the revision was made, so that the ink did not properly adhere where the parchment was still wet. The more standard position (e.g. Boyle) is that fol. 179 sustained some sort of accidental damage and someone (not the second scribe) attempted to 'touch up' all of the text that he could.

The 'water-damage' theory seems rather unlikely - one would expect damage to other folios as well if that were the case.The suggestion of Tilman Westphalen and Kiernan that fol. 179 is a palimpsest seems the most likely explanation. However, Robert Fulk (1982) offers a very plausible alternative scenario for the creation of the palimpsest, which is worth quoting at length:

    '....the person who altered the fitt numbers XXV-XXVIIII did not alter any after that [presumably] because the manuscript unit comprising quires twelve and thirteen was unavailable to him. Likewise [Leonard Boyle (1981)] states that folio 178v, the last page of the eleventh quire, is smudged, as if it had been at one time an outside cover--and it is true, in the facsimiles the script all over 178v looks quite worn in comparison to the preceding pages. Since the last folio of Beowulf of the wear and soiling it received, if folio 179 was the other cover of the unit, doubtless it was also in bad condition. Perhaps it was grimy as well as worn, and that is why the scribe decided to immerse and scour it. Of course he did not have to scour the verso side, but perhaps he simply preferred writing the text anew to tracing over the old letters faded by the immersion--a preference he did not later indulge on the last folio when he saw what a ruin this practice had made of 179. Before he washed away the text he must have preserved it somewhere: either he copied it elsewhere or committed it to memory. I suggest he chose the latter course--a natural course, I think, for a short text in Anglo-Saxon times--one is reminded of King Alfred's famous feat of poetic memorization. This choice explains a great many of the peculiarities of the palimpsest. An experienced scribe would know better than to write on wet vellum, and even an inexperienced scribe would have stopped when he saw the ink begin to run. But a scribe who had committed his irrecoverable exemplar to memory might write on wet vellum, being afraid he would forget the verses before the parchment dried. The same haste would also account for the carelessness that produced the long dittograph. It is now also possible to say why the scribe did not fill in his erasures when he wrote the single thorn [þ] at 179r10--he had in fact forgotten the verses by the time the leaf was dry' (355).

Kiernan's argument for the palimpsest as a result of active revision nearly contemporary with the MS. itself, however, is appealling in certain respects, and is supported by various aspects of the MS and its text:

    'The signs of revision on fol. 179 receive repeated support elsewhere in the MS. To begin with, it is not the only folio in the Beowulf MS from which the original text has been deliberately deleted. The first three lines of text on fol. 180v have been carefully rubbed off with the aid of some liquid solution, which has left the vellum napped and discolored'. (Kiernan 1981:245)
    'It is virtually certain...that the second scribe copied the eleventh gathering....after he had already copied the twelfth and thirteenth gatherings....The evidence that the second scribe copied the eleventh gathering last is that he had to squeeze in four extra lines of text, in disregard of the original rulings, on fols. 174v-176r. The compression would not have been necessary if the scribe knew he had at least two extra gatherings of unused vellum ahead of him...Unless he was compelled to fit the extra material into the eleventh gathering, it is far too early in the copying for the scribe to be worrying about running out of vellum...The twelfth and thirteenth gatherings must have been copied already'. (258-9)

Fulk argues that the revision is not actually editorial 'smoothing' of the textual transition, but rather simply the somewhat careless result of trying to fix a dittograph:

    'If Kiernan and Boyle are right, Scribe B had to go to some extraordinary lengths to fit his text into the eleventh quire. Likewise he had to resort to some extreme abbreviations on the last page of Beowulf, and actually wrote part of the last word on an extra line, in order to make the end of the poem coincide with the end of the quire. If Scribe A copied the tenth quire after he had begun the eleventh, it is extraordinary that he did not need to resort to such unusual methods to make the text fit'. (Fulk 1982:347)
    'Syntactically lines 2251-52 [at the transition of f.179v & fol.180r] are corrupt....Metrically, too, the transition between the two folios is defective....we have the testimony of Scribe B that gesawon seledream(as) is not one verse, since he has inserted a point between the two words, according to Kiernan. We can be sure, then, the scribe himself did not think the end of 179v and the beginning of 180r fit together....And so certainly some material seems to be missing between the two folios, the gap being the result of the dittograph. This ultimately explains the erasure at the top of folio 180v. Faced with the realization that the text did not fit, the scribe decided to erase the following page 180r, and recopy it writing twenty-three lines of text in order to fit in the extra material. But a better opportunity offered itself in the observation that a mere three lines further, on 180v, the end of a manuscript line coincided not only with the end of a sentence, but also with the end of an off-verse--a natural break in both the manuscript and the poem. ... He began to erase the three lines but never finished that job, and so never started erasing 180r'. (356)

Some sort of revision seems to have taken place round folios 179v & 180r -- whether this revision was repairing accidental damage, incomplete correction of a dittograph or other scribal error, or deliberate revision of the text, is more difficult to establish, as the above discussion of various explanations shows. These arguments are also bound up with different theories about the composition and its time period. Kiernan's argument that there is textual revision contemporary with the MS itself is complementary to his thesis of a late dating of the composition of the text of Beowulf itself, as discussed in the following section.

---

II. Date and Authorship of the Poem [14.04.02/last modified 21.12.03]

   a. Historical content and historical contexts
As stated above, the dates ranging from the sixth century to the early eleventh century have been proposed for the composition of Beowulf. The terminus a quo (earliest date) has been established by Grundtvig's identification of Hygelac of the poem with the historical figure Chochilaicus, mentioned by Gregory of Tours (d. 594) in his History of the Franks as falling during his raid on Frisia (though Gregory identifies Hygelac as a Dane, not a Geat), which probably dates somewhere between 515 and 530 CE. Obviously, the latest possible date for the composition of the poem would be contemporary with the manuscript itself, ca. 1000 CE, probably in the reign of King Cnut. Two quotes are illustrative of the arguments for dates at the extreme ends of this span; the first from Clark Hall's introduction to his translation of Stjerna's Essays on Questions connected with the Old English poem of Beowulf :

'…suppose we take A.D. 504 as the date of the birth of the Geatic prince Heardred, 515 as that of the death of his father, King Hygelac, in the historic raid against the Frisians, and 520 as that of the death of (King) Heardred and the accession of Beowulf to the throne. Let us assume also that the statement that Beowulf reigned fifty years is a poetical exaggeration…and give him twenty. This brings us to A.D. 540. Very soon after that the Geatic kingdom is conquered and annexed by the Swedes. On its downfall a Geatic scóp [oral poet] journeys to Denmark to escape from the unsympathetic and unremunerative society of the conquerors. Here he would be able to sing freely about the last hero of his race, giving the first place to an adventure in Denmark, for some details of which he may have had recourse to local tradition, and, speaking of the Danes, his paymasters for the time being, in flattering terms. After singing his lays threadbare at the Danish court, he moves on to the territory of the Angels, and finally migrates with members of that tribe to the new El Dorado beyond the sea, which he reaches about A.D. 550, when the last Anglian invasion of England is generally supposed to have taken place. A later date, say A.D. 560, or even 570, is not impossible…' (pg. xxi-i)

We may note that Clark Hall's theory supposes not only Hygelac, but other characters from Beowulf to be historical as well--which has not been proven to be true. He also supposes a 'Geatic' origin (which he supposes to be in West Göteland, in southwest Sweden) of a poem narrating the adventures of a Geatish prince, which picks up its Danish setting on the poet's way through Denmark before he ends up in England--where he is seemingly able to pick up not only English, but the verse-techniques of the Anglo-Saxons--where the poem reaches its final form (more or less). A foreign origin of Beowulf has also been espoused by Thorkelin, who in his 1815 edition claimed that the poem was a translation from Danish. Sarrazin took this a step further, and not only identified the translator as Cynewulf (whose 'runic signature' is embedded in some Old English poem, but not Beowulf), but identified the original (Danish) author as the skald [oral poet] Starkathr, who he placed at the Danish court of King Ingeld at Lejre ca. 700 CE. The famous German metrist Eduard Sievers (1886) refuted the notion of a Danish origin on linguistic grounds.

In sum, it is blindingly improbable that Beowulf is either a translation of a non-English poem or the work of a non-English poet. That much, at least, seems clear.

Arguing largely from paleographical and codicological grounds, Kevin Kiernan proposes that the extant Beowulf MS was actually the working copy of the poem, and attributes authorship to the two scribes themselves (implying that the second scribe may have been the primary director of the project):

'…the obvious conclusion is that the Danish and Geatish exploits of Beowulf were first brought together in the extant MS by the second scribe. The aesthetic fusion of these parts does not reflect a dim, romantic view of a non-Anglo-Saxon past, but rather a vivid imaginative response to chilling contemporary events. The fall of a great and noble hero, and the imminent extinction of the race he ruled, was well understood by this 11th-century Anglo-Saxon who had recently seen the fall of [King] Alfred's house and the subsumption of his homeland in the Danish empire. The second scribe begins to look like "the last survivor of a noble race", while the Beowulf MS, the treasure he continued to polish after the death of his old lord, no longer looks like a reproduction' (12).

Kiernan locates the composition of the poem within the reign of King Cnut (r. 1016-35), who accompanied his father, the Danish King Svein Forkbeard, on his invasion of England which forced King Æthelred II (the 'Unready') into exile in Normandy. Cnut became king on the death of his father and married Æthelred's widow, Emma of Normandy, to help legitimise his reign. Cnut's reign was peaceful, especially when compared against the harsh Danish raids which preceded his rule. Kiernan suggests that like Cnut, Beowulf created a synthesis of Anglo-Saxon and Danish culture, though forestalling Whitelock's objections that 'if the poem is later than the time when the Viking invasions began in earnest, about 835, it can hardly be placed before the tenth century, and even then it would have to be put, as Schücking puts it, in the court of an Anglo-Danish king in the Danelaw' (25). Kiernan remarks that 'Cnut brought together Danish and Anglo-Saxon culture in the way no petty king of the Danelaw ever could have done' (22).

We shall return to examining the different possibilities for the date of the origin of Beowulf, but it is here expedient to observe the import of assigning a date to the poem, as it affects not only our possible interpretations of the 'meaning' or 'purpose' of the poem, but also our view of the relationship between the extant Beowulf MS and the poem itself. If, for instance, Beowulf was first written down in the 6th century and our extant MS is the result of a long line of copies, made by speakers of different dialects of Old English, then we have much reason for suspecting 'scribal corruption' and thus will suppose more licence to make emendations to the extant text. If, on the other hand, we take the composition of the poem to be contemporary with our extant MS, we will assign more authority to the MS text and feel more wary of emendations. Similarly, both the date and location we posit for the composition will colour our reading of the poem based on what we know of Anglo-Saxon history. For instance, if we assume a post-Danelaw date, we might suspect a political purpose to the poem, i.e. a literary synthesis of English and Danish cultures, which emphasises the commonality between the races. On the Danish elements in Beowulf as clues for dating, see further below, Section c.

Early scholarship of Beowulf tended to favour placing the composition of the poem in late 7th or early 8th century Northumbria, in the time of the Venerable Bede (673-735), most likely in the court of the scholar-king Aldfrith (d. 705), considering that age the height of Anglo-Saxon learning, and cultural pre-eminence of Northumbria in that time. The Northumbrian connexion of Cædmon, often thought of as one of the pioneers of Anglo-Saxon poetic literature has also bolstered that notion. More recently, Schneider suggested, based on the inherent paganism in the poem, as well as the linguistic features, a composition in Mercia between 640-50 in the reign of King Penda.

Arguments made for a late date (9th - 11th century) are usually based at least in part on the Danish focus of the poem (e.g. Schücking 1917, Frank 1981, Kiernan 1981, &c.), which does have a possible political motivation in providing a notion of the kinship of the Anglo-Saxons and the Danes, which from the start of the Danish invasions in 835 until the 11th century was a central issue in English politics: the synthesis of the two cultures. Kiernan's argument for an 11th century origin rests largely on his claims for evidence of active participation of the scribes in the production of the poem - the core of which is his analysis of fol. 179 (see section I:MS above). Since the identity of the apparent third hand in this folio is argued variously as being that of the second scribe himself several years after his original writing of the MS or that of an unknown third person the status of fol. 179 is not yet proved decisive in determining the date of the origin of Beowulf's written source.

Various scholars have attempted to date the poem by possible allusions to historical persons: Earle suggests that the seemingly incongruous lines about Offa I, the legendary late 4th century King of the Old Angles, as an allusion to the 8th century King Offa of Mercia, taking the supposed name 'Thryth' (l. 1934b) to be a form of Cynethryth - the name of Offa II of Mercia's queen (see n. to l. 1934 for a counterargument on this latter point); Cook, again placing the poem in 7th century Northumbria, sees King Aldfrith concealed beneath the name Offa; Brandl, because of the character Wiglaf, dates the poem to the reign of Wiglaf of Mercia (r. 827-38)--seeing in the character of the evil king Heremod a representation of the 'pagan' king Penda of Mercia (r. ca. 632-55); Kiernan (xxi) to early 11th-century Lincoln (in part because of the possible connexions between the Beo.-ms. & the Blickling Homilies ms. [see above]). I favour, with Bruce-Mitford & Sam Newton, a location somewhere in East Anglia.

In sum, the historical allusions in the poem, with the exception of that of Hygelac, seem largely insufficient as a basis for any sound dating of the text.

   b. Language of the poem: archaism, dialect & metre
Turning to the linguistic and metrical evidence puts us on somewhat firmer ground. The language of the poem is predominantly (late) West Saxon, with a significant amount of Northumbrian and Mercian characteristics and forms, with some signs of Kentish influence as well--fairly well covering the entire Anglo-Saxon dialectic map. The overwhelming late West Saxon nature of the language could, in theory, indicate an equally late date for the composition of the poem itself, as Kiernan suggests. However, we know that scribes tend to adjust forms from older exemplars in accordance with their own dialects, so it is unsurprising to find that an 11th century MS is written in predominantly late West Saxon. Nor does the use of the West Saxon dialect provide any clear indicator of place of composition, as West Saxon became the predominant literary dialect in the late Saxon era.

Scholars have generally assumed that the poem was originally composed in an Anglian (Northumbrian or Mercian) dialect. Kenneth Sisam (1953:119-39) has questioned the evidence usually cited in support of an Anglian origin, pointing out that so-called Anglianisms like waldend (for late West-Saxon wealdend 'ruler') are simply features of the poetic koine in which almost all Old Englsh poetry is preserved, though the origin of such forms is ultimately Anglian. Fulk's important work (1992), however, shows that there is a large number of Anglian morphological characteristics, which are mainly absent from poetry known to origin in the south, which is found in poems usually thought to be Anglian--and in Beowulf. Important features include: the use of sægon for Southern sawon ('saw'); cwóm(on) for Southern cóm(on) (the preterite of cuman 'to come'); in- as a verbal prefix in place of Southern on-, in its inceptive uses such as inhaétan ('inflame'); the preposition mid ('with') may use the accusative case, whereas it is only found with the dative and instrumental cases in Southern texts; and many others (see Fulk 1992:309-25).

Whatever the date of the poem, the language favours an Anglian provenance, and the evidence slightly favours an origin in the Midlands--in Mercia or East Anglia (see Sam Newton 1993 on the latter)--than Northumbria, despite an early academic preference for a Northumbrian origin.

Important linguistic clues to the dating of the poem include the apparent lack of any Scandinavian loan words (as noted by Frederick Klaeber (cxvii)), which are found, for instance, in The Battle of Brunanburh, which has cnearr for 'ship', derived from Old Norse *knarru, or The Battle of Maldon which contains the probably Scandinavian loans dreng 'warrior', grið 'truce' - the lack of such items is particularly suggestive in poem so intimately concerned with Scandinavia and suggests a pre-Viking date (<835). A related cue is the form of proper names in Beowulf : the character Hrothgar, who appears in Nordic sources as well, is named there as Hroárr or Roe;likewise the character Hroþulf appears in the Nordic as Hrolf and other forms showing syncope of the medial consonant. Fulk (1982) points out that 'we have evidence of extensive changes [in the forms of the names] long before the Vikings' arrival in England, e.g., the name rhóltR on the Vatn stone (Norway, ca. 700). This derives from *HróþuwaldaR, giving OE Hroðweald (hroðuald, early ninth century, Liber Vitae Dunelmensis), which in turn suggests *HróþuwulfaR should already have been nearly identical with ON Hrólfr by the time the Beowulf poet used the name Hroþulf. And so the proper names in the poem must be exactly what they appear to be--authentic inheritances from a much earlier time'. As Fulk goes on to point out this proves neither earliness nor lateness. However it does appear to establish the authentic Englishness of the name-forms in the poem--i.e. the appearance of these names in the poem does not seem to require Viking-age transmission.

Fulk's 1992 book, A History of Old English Meter, examining both linguistic and metrical aspects of Beowulf, provides some very important and solid data and arguments for determining the most likely date of composition. Fulk outlines several possible means of establishing a terminus a quo [earliest possible date of composition] for Beowulf and other longer OE poems. One of these is loss of intervocalic h, which appears sporadically in the early Latin-OE glossaries (Fulk 368-74). The glossaries also show occasional lack of smoothing (smoothing is an Anglian phenomenon in which diphthongs ( ea, eo, io) were monophongised to æ, e, i, respectively, before the back consonants (c [k], g [g], h[x])), though this is more likely to reflect dialectal and scribal variation than chronological variation (Fulk 374-6). Another possible terminus a quo cue is the apocope of -i,-u (in final unaccented syllables after a long accented syllable, or a short accented syllable and another syllable -- see Campbell 144-7), which seems a promising possibility, since, as Fulk says, 'none of the longer poems contain verses that will not scan correctly without the final vowel, though all contain examples of verses that would be spoilt metrically by the addition of the vowel' (378). Fulk (378-81) is sceptical of the reliably of the vowel apocope for dating, and Amos (18-29) even more so. However, Fulk concludes, based largely on the apparent chronographic alternations of intervocalic h, that '[n]oncontraction [of intervocalic h] is clear frequently enough in the early glossaries that a measurable difference can be observed between the glossaries and the verse of Cynewulf in this regard. In earlier verse the distinction is less pronounced. Thus, Beowulf... [is] unlikely to have been composed much before ca. 685, though [it] may be considerably later than that' (381).

Perhaps more interesting is the remarkable argument Fulk presents for determining a terminus ad quem [latest possible date of composition] for Beowulf by applying 'Kaluza's law' to the poem. Kaluza's law (see Max Kaluza 1896, 1909) is the important observation that in certain metrical positions resolution is governed in part by etymological considerations no longer represented in the phonology (briefly, in OE metre, stress usually only falls on a heavy syllable [a syllable with a long vowel or one which is 'closed' by a consonant], however, two syllables of which the first is short can count as the equivalent of one long syllable and thus 'share' the stress, this latter phenomenon is resolution of the stress - see Bliss 1962, Stockwell & Minkova 1997 for good introductions to resolution and OE metre in general). Kaluza's law is predicated on a distinction between 'short' and 'long' inflectional endings. The latter are those which end in a closed syllable, e.g. the masc. a -stem gen. sg. -es; OR in a vowel which carried the Schleifton, 'circumflex', in Proto-Germanic, e.g. the ó-stem nom. pl. -a < *ô(z) < PIE *-âs. The short endings are everything else. Kaluza's law essentially states that if a light stressed syllable is immediately preceded by another stressed syllable, AND if the syllable which follows it is heavy or was historically heavy, then resolution does not take place. Kaluza's law is not consistently followed in OE verse -- and it is this which provides an important dating criterion: poems which conform to Kaluza's law are likely to be earlier than those which do not. Based on this evidence, Beowulf would seem to date from between 685-825, probably pre-725 if more Mercian than Northumbrian - see further:

   c. Epithets & genealogies - Danes and Wuffings
The pre-occupation of Beowulf with Scandinavia--beginning with its opening line hwæt we gardena in geardagum... '...of the Spear-Danes in times of old...', its 'Geatish' (Gautish?) hero--and absence of any direct reference to England or Anglo-Saxons, leads one to wonder about the context in which the poem was composed. 19th-century critics usually speculated that Beowulf's setting in the 'Old Countries' (Scandinavia) pointed to its antiquity, to its having been produced when the memory of migration was still clear and bright in the minds of the Angles and Saxons. Yet, neither Angles nor Saxons make any obvious appearance in the poem, nor does Britain figure in its geography. The only people traditionally connected to the Anglo-Saxon settlement of England mentioned in the poem are ostensibly the Jutes (OE Eotan), though even here there is some confusion as to whether the form eotena which occurs in the poem means "Jutes" or "giants" (or both or something else..). The prominence of the Danes at the outset (which lead some early scholars to claim it was an English translation of an originally Danish poem, see above in section a.), calls to mind Anglo-Saxon England's Viking Age, and the settlement of the Danelaw, and the Scandinavian 'emperor' Cnut, King of England, Denmark and Norway. However, despite, or rather, because of, the centrality of Danes in the poem, until the later 20th-century it was usually assumed to have had to have preceded the violent Viking incursions into England; Dorothy Whitelock's statement to this effect is well-known:

'The poem is surely pre-Viking Age. It may be true that we should not attach an exaggerated importance to the high terms of praise and respect with which the poet speaks of the Dane and their rulers....Yet, I doubt whether he would have spoken in these terms during the Viking Age, or whether his audience would have given him a patient hearing if he had. It is not how men like to hear the people described who are burning their homes, pillaging their churches, ravaging their cattle and crops, killing their countrymen or carrying them off into slavery. So, if the poem is later than the time when the Viking invasions began in earnest, about 835, it can hardly be placed before the tenth century, and even then it would have to be put, as Schücking puts it, the court of an Anglo-Danish king in the Danelaw. It could hardly be located in English England until reign of Cnut, and that is later than our surviving manuscript'. (Whitelock, 24-5)

More recently, a number of scholars have questioned the truth of this view, that a poem concerning Danes would be anathema to any Englishman during the Viking Age, that all Danes would have be viewed alike (see Page for evidence against undiscerning, blanket anti-Danish sentiment in Viking Age Britain). Furthermore, unsurprisingly perhaps, as the marauding Vikings decided to remain in England and began to settle what was later known as the Danelaw in northern England, Scandinavia elements and fashions began to permeate Anglo-Saxon culture and art. One revealing letter from Alcuin written in 793 after the Viking raid on Lindisfarne to Æthelred, king of Northumbria, rebukes the king and his court for having adopted Danish manners and fashions:

   'Consider the dress, the way of wearing the hair, the luxurious habits of the princes and the people. Look at your trimming of beard and hair, in which you have wished to resemble the pagans. Are you not menaced by terror of them whose fashion you wished to follow?' (Alcuin, cited in Clemoes 184)

Not only Viking haircuts, but Viking ancestors also seem to have been adopted, at the highest levels, presumably, at least in part for purposes political imagery:

'The West-Saxon regnal table that was compiled or reworked after Alfred's succession in 871 did not go back beyond Woden, the old stopping point for royal genealogies. The name of Scyld, the eponymous ancestor of the Scyldings, is first found, along with other Scandinavian names like Beaw, Geat, and Heremod, in the remotest part of the official royal pedigree, the genealogy of Alfred's father Æthulwulf, that was added to the Anglo-Saxon Chronicle around 892 and copied by Asser around 893'. (Frank 128)

'...during [the] period [between the 830s and the reign of Knut] supporters and members of the West-Saxon dynasty were claiming descent for the kings of England from the "Danish" Scyld and Scef....Æthelwulf's pedigree, extending beyond the old termination of Geat into a new line of Danish kings and heroes, reflects the political realities of late ninth-century England. It also reflects the aspirations of the West-Saxon kings. A Danish background was intended to give them prestige and leverage among the petty and disunited Scandinavian kings and earls of northern England and to support the claim of West-Saxon suzerainty over the north....the Danish invasions and settlements did not lead Englishmen to disassociate themselves from the Scandinavian heroic age. On the contrary, the heroic North attracted them, and their interest in it permitted them to establish a common background for contemporary political and ethnic relations'. (Murray 105)

Roberta Frank also points to the appearance (or lack thereof) of the dynastic title 'Scylding', used frequently throughout part one (Beowulf's Danish adventures) of Beowulf:

'....[no-one mentions Scyldings in England until Beowulf]... Names like Healfdene, Hrothgar, and Heorogar, if they are found in England at all, are not documented until the tenth century. Legends of the Scyldings in [Scandinavian] skaldic verse get off to an equally slow start. The first reference by a skald to an event associated with one of the Scyldings of Beowulf occurs around 965 when Eyvindr skáldaspillir calls gold 'the seed corn of Fýrisplains' (74, 64, 39), alluding to the story of how Hrólfr kraki (OE Hrothulf) sowed those plains with gold to delay the Swedish horsemen pursuing him'. (Frank 126)
    '...in the early eleventh century, Norse skalds appear to bestow the title in a specifically political way. In the quarter century between 1014 and 1040, they use Scylding to designate three kings: Knútr the Great, ruler of an Anglo-Danish-Norwegian empire; his contemporary and defeated rival Óláfr Haraldsson, king of Norway; and the latter's son, Magnús the Good, king of Denmark and Norway and later claimant to the English throne of Hortha-Knútr. There seems to be a connection between a poet's use of the title and the English adventures of these kings'. (Frank 126)

Further noting a political advantage in the Anglo-Saxons 'acquiring' a Danish progenitor:

    'By acquiring a founder namer Scyld, the West-Saxons strengthened their position in the Danelaw. By calling a Norse king Scylding, a skald confirmed his patron's English heritage. The Beowulf poet's incentive for composing an epic about sixth-century Scyldings may have had something to do with the fact that, by the 890s at least, Heremod, Scyld, Healfdene, and the rest, were taken to be the common ancestors both of the Anglo-Saxon royal family and of the Danish immigrants'. (Frank 129)

Frank also argues that the use of the epithet þoedcyning (literally, 'people-king') suggests an 'emperor'-king ruling over multiple peoples (clans, tribes), which only occurred somewhat late both in Anglo-Saxon England, and Denmark:

'In tenth-century skaldic verse, the title [ON þjóðkonungr, cognate with OE þeodcyning, as in Beo. l.2a--BMS] is used of only two kings, the Vestfold Halfdanr and his descendant, Haraldr hárfagr (23, 21, 13), the king who united all Norway before 900 and who apparently had diplomatic relations with Athelstan. The term þjóðkonungr does not occur again in skaldic verse until the next viking age of England, the quarter century between 1014 and 1040, when the skalds restricted its use to the same three kings they called Scyldings: Knútr, Óláfr, and Magnús. Þjóðkonungr has, if not imperial overtones, at least a suggestion of overwhelming power; in Beowulf, it is explicitly associated with regional overlordship of surrounding tribes. The Beowulf poet depicts the Danish nation's former glory in a time when powerful þeodcyningas had been able to unite the various peoples of the land, something that did not occur in Denmark, as far as we know, until the late tenth century'. (Frank 130)

Certainly the conclusion reached by Jacobs, one of the scholars working to dispell the notion that Beowulf would have been impossible after the Viking invasions, seems correct:

'To sum up: a review of Anglo-Danish relations suggests that any periods in which political considerations may have discouraged the composition of Beowulf are likely to have been brief and at most intermittent'. (Jacobs 24-5)

For even if one assumes an early written form of Beowulf, its survival and transmission--if such rabid anti-Danish sentiment existed after, say, 825AD--would be rather unexpected.

However, those these facts concerning the later refashioning of Anglo-Saxon royal genealogies and the appearance of 'Scylding' and 'þeodcyning' in Beowulf may be suggestive, they do not necessitate that the usages in Beowulf are guided by (late) Danish and Anglo-Danish employment of the same or similar terms. Old English Þeodcyning and its Norse cognate may have developed a meaning similar to that of 'emperor', without this necessarily being the earlier sense of the compound. Likewise, a meaning akin to 'emperor' could have existed earlier, whether or not it reflected the existing political situation of Anglo-Saxon England (or Denmark). Concepts can exist without immediate material correlates, and a concept of 'emperor' may still have been useful to early Germanics in their relations with Rome. Similarly, the late appearance of 'Scylding' outside of Beowulf still does not eliminate the possibility of an early usage of the title, though it may have been more geographically restricted in pre-Viking age England.

These aspects, and the overall Scandinavia-centred setting of the poem may reflect a different, and earlier, situation in which there existed strong Saxon-Norse contact/community. One such situation which offers itself is that of early East Anglia. A number of features of Beowulf, in fact, have strong East Anglian associations, as pointed out by the well-known British archaeologist Rupert Bruce-Mitford, and, in greater detail, by his student, Sam Newton. Newton points out that in the East Anglian royal genealogy:

'Hróðmund is listed as the tenth name on Ælfwald's ancestral tally, a position well beyond the pedigree's horizon of historical credibility. The name occurs there in the exact form as that borne by a Scylding prince in Beowulf. ...apart from the place-name Rodmundes Dæn, ...in the description of a Hampshire estate boundary in a suspicious tenth-century charter, the only other known Old English instance of this royal compound-name is that belonging to the son of Hroðgar in Beowulf'. (Newton 79)

The pedigree of King Ælfwald of East Anglia (r. ca.713-749) in British Library MS Cotton Vespasian B.vi (early ninth century):
Not only does the East Anglian pedigree contain Hroðmund, but based on Bede's records, the founder of the East Anglian royal line was Wuffa, 'a quo reges Orientalium Anglorum Uuffingas appellant' ("from whom the kings of East Anglia are called Wuffings"). Further, as Newton argues, 'the name Wuffa appears to be a diminutive variant of Wulf, and can thus be translated as 'Little Wolf'. The patronymic form Wuffingas seems similarly best explained as a variant of Wulfingas, "the kin of the wolf", a folk-name which is mentioned in both Beowulf and Widsith' (Newton 105-6; see also O'Loughlin 4).
   Newton also shows that the Wulfingas of Beowulf are more central to the poem then might appear at first blush:

   '[Queen Wealhþeow] is described...in the poem as ides Helminga, 'the Helming lady' (l.620b). Helmingas must be her family name, the eponym of which is listed in Widsith: "Helm (wéold) Wulfingum", '"Helm (ruled) the Wulfings"' (l.29b)...the implication arising from Widsith is that Helmingas was an alternative name for Wulfingas in Old English poetic tradition. The epithet ides Helminga thus can be interpreted as being synonymous with ides Wulfinga and Wealhþeow thus can be identified as a Wulfing princess'. (Newton 122, 124)

Thus Hroðmund is a Wulfing through his maternal line, through Queen Wealhtheow, who, as Newton says, 'may herself have been regarded as an East Anglian dynastic ancestor' (104). In the Scandinavian tradition, Hrólfr (=OE Hroþulf) seems to be the slayer of Rørik (=OE Hreðric) in an internecine struggle for the Danish throne, after Hroðgar's death (in Saxo, Gesta Danorum, Book 2.62--see further discussion in: Malone, 'Hrethric', pp.275-282; Chambers, Introduction, pp.25-7; Newton, Origins, pp.77-104). As well, from dark hints within Beowulf itself, we can feel confident that the poem's audience knew a story of later strife amongst the Scyldings, after Hrothgar's death, when Hrothulf seized the throne from his nephews, Hrethric and Hrothmund:
(e.g., Beo.-ll.81b-85: Sele hlifade / heah ond horngeap heaðowylma bad / laðan liges ne wæs hit lenge þa gen / þæt se ecghete aþumsweoran / æfter wælniðe wæcnan scolde 'The hall towered, / high and horn-gabled; it awaited the cruel surges / of hateful flames; nor was the time yet nigh / that the furious edge-malice of son-in-law and father-in-law, / arising from deadly enmity would inevitably awaken'; cf. ll.1010-1018 (esp. 1016b-18: Heorot innan wæs / freondum afylled; nalles facenstafas / Þeod-Scyldingas þenden fremedon ("Heorot was filled within by friends; at that time, in no way did the Theod-Scyldings (the Danes, Hrothgar & Hrothulf in part.) perform ?treachery (=facenstafas, lit. 'baleful runes', cp. ON. feiknstafir)); 1162b-1165a; also Widsith, ll.45-49: Hroþwulf ond Hroðgar heoldon lengest / sibbe ætsomme suhtorfæderan / syþþan hy forwræcon Wicinga cynn / ond Ingeldes ord forbigdan / forheowan æt Heorote Heaðobeardna þrym ("Hrothwulf and Hrothgar, uncle and nephew, had kept kinship together for a very long while, when they drove off the ?Viking folk and humbled Ingeld's vanguard, hewing down at Heorot the Heathobard force"))
   In the Scandinavian tradition, Hrethric is killed, but his brother Hrothmund has no Norse counterpart figure. Newton suggests that in an English tradition, Hrothmund survived, but in exile: '...on the one hand, we have Hroðmund the Scylding prince, fate unknown, though perhaps in exile; and on the other, we have Hroðmund the East Anglian [early] dynastic ancestor, perhaps associated with a royal origin-myth' (Newton, pg.104). Thus we find some facts pointing towards an East Anglian royal interest in some of the actors of Beowulf.


The Sutton Hoo ship burial in Suffolk (East Anglia) also provides material evidence for many objects mentioned in Beowulf. This is discussed at length in many basic introductions to Beowulf, and I leave the more general aspects in favour of citing the most substantial and suggestive connexions between Sutton Hoo and Beowulf (and Sutton Hoo and Vendel, Sweden):

The sword (but not the scabbard or its bosses or the pyramids from the sword knot); the shield, the helmet and the loose 'ring'...were probably all either made in Sweden itself, or by armourers fresh from Sweden working in Suffolk, exclusively in their traditional Swedish manner, and with Swedish dies, moulds and other equipment....we also find that the great gold buckle, not a Swedish piece but locally-made, is decorated in what seems to be an unmistakably Swedish style; and the naturalistic subjects on the spectacular purse-lid, particularly the man between the beasts, which is intimately related to the same scene as depicted on one of the Torlunda, Öland, dies for impressing foil sheets used in embellishing helmets, seem to indicate an incursion of Swedish Vendel-age native traditional figural art into the locally made Sutton Hoo reglia....Thus the Swedish element appears as a powerful leaven, in a milieu which is that of the East Anglian court...the custom of elaborately furnished boat-inhumation, which becomes widespread later in the Viking age, is at present only recognized in two places in Europe, the Uppland province of Sweden and south-east Suffolk (the two boat-inhumations at Sutton Hoo and the one at Snape)...The most plausible explanation of the hard fact of the Swedish connection seen at Sutton Hoo is that it is dynastic. The evident antiquity of some of the Swedish pieces at Sutton Hoo, especially the shield, suggests that the connection goes back into a period earlier than the burial. The most likely explanation seems to be that the dynasty of the Wuffingas was Swedish in its origins, and that probably Wehha, said to be the first of the family to rule over the Angles in Britain, was a Swede. (Bruce-Mitford 69-71)

Thus, there are strong arguments for associating the creation of Beowulf with East Anglia, even aside from the general evidence of its early Scandinavia elements. Though there are also many aspects of the later Viking-Age Danelaw, as well as of the still later Northern Empire of Cnut, which are also possible sources or reasons for the centrality of Scandinavia in Beowulf. Even if these latter ages played little role in the initial creation of Beowulf (which is by no means a foregone conclusion in any case), they certainly suggest reasons why Beowulf may have remained relevant (and popular?) for at least some Anglo-Saxons (and Anglo-Danes), at least until the beginning of the eleventh century.

   d. The Return Home: lays, ballads and transitions
The last 150 years or so have seen a variety of views on the structural and thematic unity of Beowulf. Many of the early (esp. German) critics analysed within Liedertheorie, considering the poem to be made up of a variety of originally separate pagan Germanic lays or ballads, loosely woven together by Christian scribes. Karl Müllenhoff's well-known 'Die innere Geschichte des Beovulfs' is perhaps the most comprehensive analysis of the poem in the Liedertheorie 'dissectionist' tradition, suggesting that Beowulf's Danish adventure against Grendel and the fight with the dragon were originally two separate lays pieced together later--the first section dealing with Denmark having already augmented separately with a 'second episode' featuring Grendel's dam and a third episode concerning his return to Geatland (so-called 'Beowulf's Homecoming'). From the early 20th century, most critics have argued strongly against this sort of 'dissectionist' view of Beowulf, stressing its thematic and stylistic unity: e.g. Klaeber (1911-2 & ed.), Chambers, Tolkien, Chance (Nitzsche), & many others.

Although the divisions made by Müllenhoff do indeed correspond to natural breaks in the plot of the poem, I am in agreement that we find an essential unity in Beowulf, which tells against the Liedertheorie, or, at least, a simplistic version of it. However, this does not exhaust the possibility of something resembling composite authorship. Within the oral-formulaic theoretic framework (as developed by Parry, Lord, et al.), some critics, notably Francis P. Magoun, ressurected the idea of something like a composite poem, and thus some began think again about the possible pre-history of our poem, to though not in the same way as in the older Liedertheorie, as it does not necessitate a heaping up of unrelated 'lays' or 'ballads', but can conceive of composite authorship within the framework of skilled verbal-artistics located within, and manipulating, a structured communal tradition. Oral-formulaic theory is discussed in more depth below at section e..

Beowulf displays a pervasive thematic and stylistic unity, and thus is certainly not the result of a haphazard heaping together of unconnected pieces. However, as noted above, we do observe natural episodic breaks in the plot of Beowulf, like any longer work of verbal art. The major ones, with approximate line numbers, are:

  (1)Beowulf's adventures in Denmark against Grendel & his dam (lines: 1-1890);
     [with a more minor episodic break occurs inbetween:]
        (a)fight with Grendel (lines: 1-1252)
        (b)fight with Grendel's mother (lines: 1253-1890)
  (2)Beowulf's return to Geatland and report to King Hygelac (lines: 1891-2222);
  (3)King Beowulf's fight against the dragon (lines: 2223-3184).

Additionally, in terms of plot sequence, a large break occurs inbetween parts (2) and (3): parts (1) & (2) concern Beowulf's youthful adventures in Denmark, and his triumphant return to Geatland; part (3) rapidly skips 50 years or so forward, Beowulf appearing as a greatly aged King of the Geats. Numerous commentators have noted the subtle change in tone between the sections telling of young Beowulf, and that recounting Beowulf's final days on earth (of course, that we should observe a change in tone, and some thematic development between these two parts is not unexpected). Interesting discussions of these alterations in tone include: Tolkien, Osborne, and Irving (1984), amongst many others.

However, while admitting the overall unity of the poem as it stands, one could imagine that the adventures in Denmark and the fight against the dragon were separate, though not unrelated, stories. Clearly, if they were indeed separate, they have been joined in a skilful manner; the essential unity of the poem is not at issue. An interesting scenario put forward at the beginning of the 20th century by Levin Schücking in his Beowulfs Rückkehr suggests just this, that a skilful poet may have linked these two stories, with the section concerning Beowulf's return to Geatland being a more newly created transitional 'weld' between Beowulf se geogoð and Beowulf cyning. As Schücking puts it, the difference between the two parts of the poem is somewhat like the difference between the Iliad and the Odyssey:

'Sie erblickt den großen Riß durch die Beowulfkomposition nur zwischen Grendelabenteuern und Drachenkampf. Hier ist ein weit größerer Unterschied als zwischen den beiden Teilen des Nibelungenliedes, ja fast ein ähnlicher Unterschied wie zwischen Ilias und Odyssee'. (Schücking 9)

Both Schücking and Magoun (1958, 1963) have pointed out a number of minor plot inconsistencies between the action of part (1) and the account given by Beowulf to Hygelac in part (2), including: Hrothgar's sons are mentioned prominently in part (1), but in part (2) they are not mentioned, and Beowulf discusses only the impending marriage of Hrothgar's daughter Freawaru (not mentioned in part (1)); in part (2) we learn that Grendel apparently had some sort of 'bag' (glof) to stuff his victims into, but this is not mentioned at all in part (1), nor is the name of his Geatish victim, Hondscio. Other possible discrepancies exist, for which see Schücking (esp. 12-13), Magoun (1958, 1963); for convincing counter-explanations of many of these as deliberate artistic variations, see esp. Brodeur (1970).

However, Schücking does point out a number of syntactic constructions which occur only in 'Beowulf's Homecoming' (=part 2) (except for a few which, interestingly, also occur in the 'Scylding' prologue, ll.1-52):

  (a) the form siððan ærest ('after earlier') occurs only at l. 1950b (in the 'Thryðo-Episode') and in the prologue at l.6b [Schücking 53]
  (b) the conjunction þylæs ('lest') occurs only at l.1921b, elsewhere we find þæt ne, e.g. l.1448, 1445, etc. [ibid.]
  (c) the formula hyrde ic occurs (twice) only at l.2175a, 2166a, and in the prologue at l.62a [ibid.]
  (d) the use of ac as an interrogative particle occurs only at l.1993b [ibid.] (elsewhere occurs as the conjunction 'but')
  (e) the placement of ða so 'deep' within a sentence only at l.2195b [ibid.] (i.e., næs mid Geatum ða; contrast with normal construction like l.2592b: næs ða long to ðon, or l.53a: ða wæs on burgum)
  (f) the only certain use of forþam, rather than forþon or forþan, as the conjunction meaning 'since, because' [Schücking cites Sievers (Beitr. 29, page 313) to the effect that the use of forþam in this sense is a characteristic of later OE]
  (g) the transitional phrases ic sceal forð sprecan (l.2072b) and to lang is to reccenne (l.2096a) are also unique to part (2) [Schücking 54]

Schücking (53-4) also mentions a number of other features which are characteristic of part (2), though not as distinguishing as those above. Kiernan also argues that there is evidence from variation in spellings used through the Nowell codex that the two scribes tended to faithfully reproduce at least some features of their exemplar texts, and, further, that there are suggestive spelling differences between what are roughly parts (1) & (3) which would suggest that Beowulf was copied from two different exemplar texts, rather than a single text: '...though the spellings in the prose codex and in the Judith fragment are not appreciably different from the spellings in Beowulf, there are enough minor differences to suggest that both scribes faithfully copied the spellings of their exemplars. Hence a possible explanation for the orthographical differences between the first and second scribe in Beowulf is that their respective parts of the MS are based on two different exemplars' (pg.275).

Kiernan (see esp. 219-78) also argues that the MS here too 'provides compelling codicological evidence that [the second scribe] copied a new episode, designed to fuse Beowulf's Danish and Geatish exploits into a unified epic' (pg. 262). The codicological evidence is Kiernan's analysis of f.179 as a palimpsest (see above, section I.d.), and thus the reason, on this theory, for the creation of the palimpsest is that: '[a]pparently, many years later, [the scribe] was still working toward a better fusion of the parts, on the revised text of fol. 179' (ibid.).

This idea, unsurprisingly, remains controversial, and is not widely accepted. But it is attractive for reasons, such as those cited above: the subtle, but suggestive differences in phrasing and syntactic usage in part (2), 'Beowulf's Homecoming', some apparent minor plot inconsistencies, and a shift in tone between the Danish and Geatish sections (parts 1 & 3, respectively). However, this idea is at odds with the evidence for great antiquity provided by an examining of the conformity of Beowulf's verses to Kaluza's law. Of course, as I note in my article on Kaluza's law, there remains some uncertainty about the exact interpretation one should give for Beowulf's strict adherence to the law.

  e. The Scop and the Scribe: formulae in speech and ink
Milman Parry's fieldwork in Yugoslavia in the early-mid 20th-century discovered a living tradition of illiterate singers (both Muslim & Christian) of traditional folk-poetry. Parry argued that this tradition of the Serbo-Croatian oral-formulaic guslar singers is part of a larger Indo-European tradition of oral-formulaic poets, initially with the argument that the Greek Iliad and Odyssey were created in a manner not dissimilar to the method he observed in the songs of the Serbs and Croatian oral poets. It was soon recognised that the repeated phrases which occur in Germanic traditional poetry were evidence of a similar tradition amongst Germanic scopas and skalds.

Indeed, in Beowulf we observe a pervasive use of identical or nearly-identical verses, an examination of but eight lines reveals the following re-occurrences (taken from Orchard 1997:105):

also occurs at Beo.	a-verse	b-verse	line cited from:
5a	manigre mægþe	geond þisne middangeard	Beo. 75
	manigum mægþa	geond þysne middangeard	Beo. 1774
6b	inwitniða	syððan ærest wearð	Beo. 1950
8a	weox þa under wolcnum	and wriðade	Genesis A 1702
	wod under wolcnum	to þæs þe he winreced	Beo. 713
	weold under wolcnum	ond his wigge beleac	Beo. 1773
8b	werodes wise	wurðmyndum spræc	Exodus 258
9b	ymbesittendra	ænig ðara	Beo. 2735
10a	geond hranrade	inc hyrað eall	Genesis A 205
	ofer swanrade	secean wolde	Beo. 200
10b	þære þe þam hæðenan	hyran scolde	Daniel 135
	hu ge heofoncyninge	hyran sceoldon	Elene 367
	heaðorinca gehwilc	heran sceolde	Metres of Boethius 9.45
11a	gomban gieldan	ond gafol sellan	Genesis A 1978
11b	glædne Hroðgar	ac þæt wæs god cyning	Beo. 862
	Geatum wealdan	þæt wæs god cyning	Beo. 2391

As Orchard observes, we find that 'ten out of the sixteen half-lines here exhibit some some kind of formulaic phrasing, and that nine of the seventeen lines of supporting evidence cited...derive from Beowulf itself..' (Orchard 1997:105)

Magoun (1953[1991]:46) describes an abstract idea of the Indo-European oral-formulaic singer:

'The oral singer does not memorize either the songs of singers from whom he learns nor late does he memorize in our sense of the word songs of his own making. His apprenticeship involves the learning of thematic material, plots, proper names, and formulas with which he will gradually become able to compose in regular verse of his own. A good singer is one able to make better use of the common fund of formulas than the indifferent or poor singer, though all will be drawing on essentially the same body of material..'

This somewhat restrictive picture of juggling of formulae has been expanded somewhat in later work following in the oral-formulaic theory. In particular, that of course not each verse must be or is a learned formula, obviously poets can and must invent new formulae; and, further, that Old English formulae are quite malleable. The use, invention and adaptation of formulae in Old English verse reveals a deliberate and complex artistry. And the idea of the requirement of a formal 'apprenticeship' seems incorrect as well (though in some cases, especially in the case of professional bards, there may be something like an apprenticeship), as these formulaic techniques of construction are not restricted to verse, as we find them pervasively in OE prose as well, such as the sermons of Archbishop Wulfstan of York (see Orchard 1997:109f.). So, to a certain extent, OE speakers and writers seem to have 'passively' acquired some of their formulaic diction, simply by virtue of being part of an oral-formulaic culture. In fact, formulaic diction is not limited to English of over a millenium ago, nor to Serbocroatian, but in fact can be observed to still be pervasive in structure of speech of many public speakers and others. For instance, Orchard (1997:111-4) notes similar technique and use of repeated formulae not only in the OE homilies of Wulfstan, but also in the that of speeches of the American civil reformer Dr. Martin Luther King, Jr. (e.g., 'Free at last! Free at last! Thank God Almighty, we are free at last!'). Likewise, Alta Cools Halama notes similarities between the structure and use of common and/or repeated lyrics in a variety of modern "gansta' rap" music, and in particular contexts in Old English narrative poetry, most particularly in the flyting contest of words between Beowulf and Unferth.

Further, Geoffrey Russom in an important assessment of the oral-formulaic nature of Old English verse, notes that the OE formula differs, for instance, from that employed in ancient Greek verse:

'Homer is most dependent on the repeated phrase just where oral theory would predict...[while the Anglo-Saxon] scop shows unmistakably his ability to express a given idea in as many ways as he likes' (Russom 1978[1991]:207).
   '[Homer's] adherence to a single verse-making strategy can be observed most clearly in his whole-line formulas and systems....[he] says of the Cretans [in the Iliad] in 650 that "spear-famed Idomeneus led them" (των μεν αρ' 'Iδομενευς δονρικλυτος `ηγεμονενε). Seven lines later, the leader of the Rhodians appears in an identical type of line (των μεν Tληπολεμος δονρικλυτος `ηγεμονενεν). The two formula frames differ only with respect to 'αρα, a meaningless verse-filler necessary for the first name, but not the second....such details testify to a concern with overall structure and at the same time to a complete lack of interest in elegant variation' (Russom 208).
   '...in Anglo-Saxon verse the number of light syllables is not strictly regulated [as is the Greek verse--BMS], a poet can change unstressed function words and pronouns with great facility. A given formulaic pattern can therefore be adapted to a much wider variety of uses than can Homeric formulas. Fully stressed words can also be exchanged where their position makes alliteration unnecessary..' (Russom 209).

So one should not understand the oral-formulaic construction of poetry to be confining in a limiting sense, for skilled poets can manipulate it in many ways, but rather as forming the medium within which Old English verse is cast (as well as that of other traditional oral poetry in certain other languages/cultures).

One of the important aspects of oral-formulaic theory is that it necessitates a shift in our thinking about the 'author'. Within oral-formulaic poetry, we find a heavy reliance on tradition, both in the metrical structure and use of common formulae, as well as in the choice of subject-matter and story (see also part c. above). Likewise, in oral traditions, poets usually and frequently re-use material from other poems or songs, freely borrowing and adapting material from other poets.

So, at least in theory, Beowulf may not have been composed all at one go. Certainly, even if it did originate with a single author, this author relied on and worked within a very traditional cultural framework of vernacular poetry. It could even be the case that Beowulf was continually reworked by a number of different poets, and further that different versions and episodes ('sequels'?) of the poem may have arisen. It should be noted, however, that there is little direct evidence for these sorts of possible pre-histories of Beowulf, other than the common occurrence of this phenomenom within oral/oral-formulaic 'texts' in general. Kiernan's theory that folio 179 is evidence for revision (by the second scribe) of the transition between two originally separate tales about Beowulf (or a Beowulfian hero) should be considered with the structures and traditions of oral-formulaic verse in mind (though Beowulf may be a 'conjoined' poem in this sense whether or not folio 179 is the result of later editorial revision).

  f. some concluding remarks
Orality continues to affect even the written transmission of poems as well. Most OE poems are extant only in a single MS, but we do have limited evidence from a handful of poems which exist in more than one copy. The poems known as Cædmon's Hymn and Bede's Death Song are two notably instances of poems recorded in a number of mss. appearing over a sizeable span of time. For our purposes, two interesting observations can be made from OE poems appearing in multiple copies: (1) we find evidence for high levels of scribal accuracy over long periods of time; and, (2), that when we do find variation, it is often variation of formulas (as Kiernan (174-5) says: '..aside from dialectal transliteration, the only significant differences between the 10th-century West Saxon MS (Tanner 10) [of Cædmon's Hymn], and the early 8th-century Northumbrian MS (Kk. v. 16), known as the "Moore MS", is the aelda versus eorðdan variant'). Though this is not to say that we do not also sometimes observe significant scribal corruption in these and other poems. But in the cases of 'formulaic' variation, what we see is that a scribe will substitute a semantically equivalent, or near-equivalent, word or formula: '...a scribe [may] consciously or unconsciously "improv[e]" a received text from his own knowledge of the formulaic tradition...which has been described as the result of what is termed [by O'Keeffe] as "residual orality"' (Orchard 1997:117). The variation in Caedmon's Hymn of aeldu barnum (biblical 'sons of men') for eorðan bearnum ('sons of the earth'), or vice-versa, is conceivably an example of this sort of formulaic-revision; as too is the variation between the highly formulaic compound eorðbugendum ('earth-dwellers') for iegbuendum ('island-dwellers') at l.3a in Alfred's Metrical Preface to Gregory's Pastoral Care (see again Orchard 1997; and see esp. O'Keeffe, O'Donnell for further examples and discussion of the transmission/evolution of these and other OE poems). In Beowulf, a plausible instance of formulaic-substitution occurs at l.964, þæt he for handgripe minum scolde ('..that he because of the hand-grip of mine must...'). Usually modern editors emend handgripe to the semantically-equivalent mundgripe (which occurs elsewhere in Beo. at ll. 380, 752, 1536, 1941), to provide standard linking-alliteration with minum ('(of) my'). The most straight-forward explanation for the appearance of handgripe here is the inadvertant substitution by a scribe (not necessarily the first scribe of our MS., though this is quite possible as well) of a formulaically-equivalent compound for mundgripe.

These sorts of occurences of formulaic substitutes (which may be somewhat frequent in Beo., though largely undetectable except when they 'violate' the normal alliteration or metre) raise the question of the value of Kaluza's law (see section b. above) for purposes of dating Beowulf. However, even a cursory glance at the verses to which Kaluza's law applies (at least to what I call, following Fulk, types I & II) shows that most instances involve double-alliteration in the -verse, e.g. l.76: folcstede frætwan. him on fyrste gelomp (Kaluza's law applies to underlined a-verse). In most such verses double-alliteration is compulsory (in Bliss-type 2A3a verses (Sievers-type A2l) with secondary stress in the a-verse, see Bliss §53; Bliss-type D3 (Siever-type D2) also require double alliteration, see Bliss §52), and therefore formulaic substitution (which would have great potential to cause violations of Kaluza's law), intentional or inadvertent, would be generally blocked by the difficulty in substituting a semantic/formulaic equivalent without disturbing the compulsory double-alliteration. Furthermore, these types of verses not only are found violating Kaluza's law in late poems when they occur, but they become increasingly infrequent (regardless of conformity to K's law) in late poetry. Therefore, not only are these types of verses more likely to be preserved undisturbed in the transmission process due to the general inavailability of ready formulaic substitutes, but later poets (and 'editing' scribes) who may 'revise' a poem in various ways, are also less likely to add in verses which would violate Kaluza's law, since specific verse structures to which it applies become very infrequent on the whole in later OE verse.
   The upshot of this is that Beowulf could constitute a relatively early poem (8th-century), or poems--as the evidence of Kaluza's law would seem to require, and yet have be revised in certain (limited) ways at a later point, along the lines suggested by Kiernan. This sort of revision would presumably not include major addition or alteration, but could include minor adjustments, and certainly the joining of two poems with a revised transitional textual bridge. This sort of analysis may be one way to reconcile conflicting claims of 'early' and 'late' datings of Beowulf, though it must remain rather speculative at this point.

Whatever may be the exact timeline of its construction may be, we may feel confident, and fortunate, that our sole extant version of Beowulf as found in British Library MS Cotton Vitellius A. xv is a complete and relatively reliable text of matters concerning a character known to us as 'Beowulf'. It is certainly one of the finest of Old English poems, and in my opinion one of the most splendid works of literary art, in the class of the Iliad, Mahabharata, or Shakespeare.

---

III. The Theme(s) of Beowulf [21.12.2003]

The question of the authorial intent of Beowulf is perhaps more vexed even than the questions concerning the structure and origins of the poem (as dealt with, briefly, above). I shall add further to this section as time permits. For the nonce, I leave the reader with a passage from another Anglo-Saxon text, which seems to me to be relevant for modern readers (and listeners) seeking to understand the import of Beowulf:

  "Wella, wisan men, wel; gað ealle on þone weg ðe eow lærað þa foremæran bisna þara godena gumena & þara weorðgeornena wera þe ær eow wæron. Eala, eargan & idelgeornan; hwy ge swa unnytte sien & swa aswundne? Hwy ge nyllen ascian æfter þæm wisum monnum & æfter þæm weorðgeornum, hwylce hi wæron þa þe ær eow wæron? & hwy ge þonne nyllen, siððan ge hiora þeawas geascod hæbban, him onhirian, swa ge swiðost mægen? forðæm hi wunnon æfter weorðscipe on þisse worulde, & tiolodon goodes hlisan mid goodum weorcum, & worhton goode bisne þæm þe æfter him wæron. Forðæm hi wuniað nu ofer ðæm tunglum on ecre eadignesse for hiora godum weorcum."

  "Ah, you wise men! you all walk in the paths which were shown to you by the famous patterns of those good men and those who were eager for honour--who came before you. Alas! you wretches and lovers of idleness; why are you so useless and so inactive? Why won't you ask about those wise men and about those who were eager for honour, to find out what sort of men they were, who came before you? And why, having found out their qualities, won't you emulate them, as much as you can? For they strove for honour in this world, and obtained great glory by their good actions, and made a good example for those who came after them--thus, they now dwell above the stars in eternal happiness, because of their good deeds."

---