SOURCE: “The Dating of Beowulf,” in The Heroic Poetry of Dark-Age Britain: An Introduction to Its Dating, Composition, and Use as a Historical Source, University Press of America, Inc., 1997, pp. 41-63.

[In the following essay, Evans examines the debate concerning the date of composition of Beowulf and argues that an oral version (probably composed between 685 and 725) of the poem preceded the earliest written version.]

It is with no little trepidation that the present study enters the battle that has raged, and continues to rage, over the dating of Beowulf. It is clear that a wide array of knowledge and expertise from a variety of disciplines has been brought to bear on the question at hand. That the question remains a subject of often intense debate certainly is not attributable to any lack of interest, nor to a dearth of publications. To begin this particular contribution to the long-running debate,¹ it is best to start with the manuscript itself.

Beowulf is found in the Nowell Codex,² the second of two codices that comprise BL MS. Cotton Vitellius A.XV. The Nowell Codex, on folios 94-209 of the manuscript, contains five works in Old English: a homily on St. Christopher (Folios 94-98r), Wonders of the East (98v-106), a “letter” from Alexander the Great to Aristotle (107-131), Beowulf (132-201), and Judith (202-209). The beginning of Christopher and a majority of Judith already was lost by 1563, when the manuscript came into Lawrence Nowell's possession, for whom the codex is named. A few decades later, it became part of Sir Robert Cotton's library. Shortly before moving to the British Museum in 1753, the codex was damaged during the Cottonian Library fire in 1731, with the tops and outer parts of many of the vellum folios being burnt, along with much of the physical evidence regarding the bindings, and the formatting and composition of the quires. In 1786-87, the two Thorkelin transcripts of the poem were made. Transcript A was made by a copyist who knew no Old English and who was employed by Thorkelin, while Transcript B was made by Thorkelin himself. In a few instances, because of the continued deterioration of the original text that occured after the 1731 fire, these transcripts can provide a clearer reading than that found in the original manuscript.

The first question that needs to be asked is: What can the palaeography of the codex itself tell us about the poem's dating and origins? Neil R. Ker has dated the Nowell Codex to somewhere within the half century surrounding the year 1000 (975 × 1025).³ Beyond the fact that the manuscript had been in a fire, and that Beowulf's Geatish lord Hygelac is indeed the Chlochilaichus mentioned by Gregory of Tours as having died in a raid against the Frisians in 521, Ker's dating of the extant manuscript is about the only statement on this issue upon which all scholars can agree. Given this state of affairs, it should come as no surprise that recent scholars have continued to hold widely divergent views and reach diametrically opposing conclusions concerning the manuscript's codicology. Of immediate note are the opposing views and conclusions held by Leonard Boyle and Kevin Kiernan.⁴

Using Ker's dating scheme as a foundation, Kiernan has proposed a historical argument to support his claim that Beowulf was a product of Cnut's reign in England, and subsequently assigned a terminus post quem for the poem at 1016. Kiernan claimed that, due to the Danish genealogy found in the poem's opening lines, the copying of the poem would have been possible only after 1016, “when the genealogical panegyric was a compliment, rather than an insult, to the reigning king”.⁵ To bolster this argument, he proposed an unusual interpretation of the codicological data contained in the manuscript. Essentially, Kiernan declared that the poem in its present form is actually a composite that had been joined together by a new transitional section (folios 171-178 in Kiernan = 174-181 BL foliation). This transitional section, which includes the poem's section known as “Beowulf's Homecoming” (1888-2199), was composed by two scribes in an effort to join together two other poems they had in their possession: one about Beowulf's fights against Grendel and his mother, and another whose story centered on the battle between the Dragon and an aged Beowulf. According to Kiernan, Scribe B returned to the text at some later date, and washed down and subsequently composed the lines found on folio 179 (BL 182) in a further effort to smooth the transition between the newer segment and the final poem. This and evidence of other scribal revisions and emendations led Kiernan to conclude that the “eleventh gathering, containing Beowulf's homecoming, is palaeographically and codicologically as well as textually transitional”.⁶ In fact, folio 179, regarded by Kiernan as a palimpsest, and the B scribe's later emendations throughout the text led him to advance the possibility that:

the Beowulf manuscript amounts to an unfinished draft of the poem. As incredible as an extant draft of Beowulf may seem to some readers, there is considerable palaeographical and codicological support for the view that the Beowulf manuscript in fact preserves for us the last formative stages in the creation of an epic.⁷

For Kiernan, the date of the poem's composition—in the form that is contained in the extant text—and the date of the poem's manuscript are one and the same; the sole manuscript represents the unfinished draft of, quite literally, an epic in the making.

Kiernan's arguments can be contested on several palaeographical and codicological grounds.⁸ First, Kiernan's interpretation of the codicological evidence has been countered forcefully and effectively by Leonard Boyle. In his interpretation of the text, Boyle agreed that the A and B scribes did collaborate, but only to the extent of deciding how much of a “copy-text” each would be responsible for copying. According to Boyle, this copy-text contained the same five works in the same sequence as they now are found in the Nowell Codex, and was written in the same format—uniform quires of eight folios each; text frames ruled for twenty lines on each page—as that first selected by both scribes. This was indeed the format used by both scribes during their initial copyings, which were Beowulf's beginning quires for Scribe A and Judith for Scribe B.⁹ Though several scribal miscalculations on their part threw them “off course” during the course of their copying, both scribes knew exactly the type and amount of modifications that they would be required to make in their textual format in order to get them back in step with the text which lay before them in their “copy-quires”.¹⁰ Contrary to Kiernan's claims, the cumulative evidence and reasoning found in Boyle's article make it clear that both scribes were copying a manuscript that contained a version of Beowulf exactly like the one that is contained in the Nowell Codex, with the poem beginning on the seventh folio of the fifth quire—a quire of normal four-sheet length—in both the extant manuscript and its antecedent copy.¹¹ Further, Kiernan's fundamental argument that folio 179 (182) represents a palimpsest, whose purpose was to join together two disparate tales, has been shown to be the least likely of several explanations that would serve to explain the poor condition of that particular folio.¹²

Katherine O'Brien O'Keefe also has objected to Kiernan's findings and methodology, and has dismissed Kiernan's claim that Beowulf is a composite poem compiled during Cnut's reign,¹³ having found that the amount and style of pointing found in the Nowell Codex is inconsistent with eleventh-century practices. Instead, she discovered that the pointing practices used by both Scribes A and B are consistent with those found in works of the third quarter of the tenth century.¹⁴ These tenth-century practices extend to Quire 11 of the Nowell Codex, the very same quire that constituted the key section of the poem for Kiernan's entire dating scheme. Consequently, she concluded that:

the pointing of the manuscript does not support Kiernan's contentions either that the manuscript represents work done in the latter part of the first quarter of the eleventh century or that the work of the eleventh quire and the so-called palimpsest of folio 179 (fol. 182) represent contemporary editing. The evidence of the pointing simply cannot sustain such a hypothesis.¹⁵

More importantly, the findings of Ashley Crandell Amos in her Linguistic Means of Determining the Dates of Old English Literary Texts are especially damaging, not only to Kiernan, but to all those scholars who had assigned the poem's composition to either the tenth or eleventh centuries. In that study, Amos conducted an extensive analysis and evaluation of the various linguistic tests that had been developed by Lichenheld, Morsbach, Sarazin, Richter, and others. Because Amos believed that linguistic tests should be applied in the same objective, mechanical manner as in the natural sciences,¹⁶ it is not surprising that her research led her to conclude that “most of the linguistic tests … are so limited by qualifications that they do not provide clear, unambiguous, objective evidence with respect to date”.¹⁷ However, even given the parameters of her overly-stringent methodology, Amos declared that:

a poet's metrical practice with respect to alliteration of palatal and velar g is probably a valid chronological criterion, as long as it is recognized that metrical practice is in part a function of personal poetic style, and that there is no a priori reason that a late poet might not adopt an earlier, traditional style.¹⁸

Keeping in mind the possibility that a late poet might adopt an earlier style, it is nonetheless true that alliteration of palatal and velar g suffers a marked decline as the period progresses, a decline that is evidenced in the poems to which dates readily can be assigned, such as Maldon and Brunanburh. Thus, while it is true in theory that later poets were free to follow an earlier style that alliterated palatal and velar g, it seems clear that they did not avail themselves of this option in practice. That later poets would have avoided earlier practices in this regard is to be expected, given the importance of alliterative techniques in Anglo-Saxon poetry. As Amos explains, “it is difficult to imagine a late poet depending for functional alliteration on sounds that no longer alliterated in most of his hearers ears”.¹⁹ Consequently, the fact remains that it is likely that poems in which velar and palatal g alliterate are older than poems in which they do not, a change that is seen in poems of the tenth century and later.²⁰ Using this criterion, one can claim that, on the basis of the consistent alliteration of velar and palatal g in Beowulf, the poem can be placed with some confidence prior to the beginning of the tenth century. Thus, based on these patterns of alliteration, dating schemes that had assigned the poem's composition to either the tenth or eleventh century are not likely to be valid.

Despite some problems that it shares with other linguistic tests, also important to the discussion at hand is the contention that a useful dating criterion would be where the contraction in words with original hiatus between vowels, or hiatus caused by loss of intervocalic h, j, or w affects the number of syllables in a verse, especially those verses that have only three syllables without the decontracted form of the word assumed to have contracted.²¹ Though in fact only the loss of intervocalic h should be seen as an valid chronological indicator,²² this remains a fundamental point for the dating of Beowulf because:

It has been deemed possible to infer the presence of archaic forms in the text on the basis of metrical considerations, in that certain lines appear not to scan unless they are assumed to have contained older, uncontracted spellings. Theoretically, for example, if the poem was composed before loss of intervocalic h in Old English, a sound-change datable to around the end of the seventh century, the diphthongs produced by the resulting contractions of the affected forms in the course of transmission would require disyllabic scansion. Such disyllabic diphthongs would thus not necessarily be explained by the use of the diction of the Old English poetic tradition. As there does appear to be a higher proportion of such disyllabic diphthongs in Beowulf than in other Old English verse, we might have a hint of evidence pointing to early composition, perhaps even within a generation or two of the end of the seventh century.²³

However, this test is not without its problems,²⁴ the most notable of which is the claim that “‘uncontracted’ forms like heahan and doan were re-formed in late Old English by analogy to the standard inflectional systems, so an ‘uncontracted’ form could as easily be late as early”.²⁵ This point, first advanced by Randolph Quirk over four decades ago,²⁶ recently has met with strenuous opposition by Fulk and others.²⁷ Most damaging to the case for late analogical re-formation is the regularity of spelling of these forms. In other words, to adhere to the assumption that “contractions could be reformed a century or a century and a half later in ways that exactly recapitulate the pre-history of þeon, fleon, teon, feon, and slean is less probable than to assume that these disyllabic forms are early”.²⁸

Even if Quirk's contention were valid, since Beowulf can be shown to be early on other grounds, then its higher degree of uncontracted forms—a characteristic it shares only with other early texts such as Guthlac A—must be seen as retained archaisms and therefore can be attributed directly to its early composition, and not to any late analogical re-formation. At the very least, one could plead that the poem's higher proportion of disyllabic diphthongs should “count for something in placing Beowulf earlier than most other poems”.²⁹

Another piece of evidence clearly demonstrates Beowulf to be the earliest of the longer Old English poems. This evidence is the Beowulf poet's utter and unequivocal adherence to Kaluza's law,³⁰ in sharp contrast to all other Old English poems. Briefly, Kaluza's law states that “some metrical positions resolution is governed in part by etymological considerations” that are based on a phonological distinction between long and short inflectional endings in Proto-Germanic.³¹ Fulk's analysis of the longer Anglo-Saxon poems in regard to their adherence to Kaluza's law led him to conclude that:

Beowulf is unique in respect to the great ease and regularity of the poet's ability to distinguish long and short endings. The facts about Kaluza's law in Beowulf are impressive, as Amos remarks (p.99): out of 108 unambiguous instances there are just 2 exceptions. Such a proportion is unquestionably outside the range of coincidence, and the exceptions are few enough that they are within the statistical range of being due possibly to scribal corruption. No other poem approaches Beowulf in this regard.³²

In fact, the probable explanation as to why later poets avoided using verses of the type in question is that “the metrical value of the vocalic endings was no longer obvious when they were composed”; this in stark contrast to Beowulf, whose poet undoubtedly was aware of the distinction in vowel quantities.³³ The difference between Beowulf and the other poems in respect to Kaluza's law, both in the regularity with which it is applied and in the frequency of relevant verse types “is so profound that it can hardly be dissociated from the elimination of the phonological distinction between long and short vocalic endings”, which would mean that “Beowulf, at least, must have been composed before the shortening of the long endings”.³⁴ In other words, the argument's rationale states that:

Observance of Kaluza's law … depends upon the maintainence of the original distinctions in vowel quantities. The fidelity of Beowulf to the law demonstrates that the poem can only have been composed before the quantitative distinction was lost. The law thus provides a fairly precise terminus ad quem for the dating of the poem.³⁵

Thus, if Beowulf was composed somewhere south of the Humber River—that is, in Mercia or one of the Anglo-Saxon kingdoms in southern England—it could have been composed as late as the first quarter of the eighth century; if composed in the north, its terminus ad quem could be a century later. In sum, “Beowulf almost certainly was not composed after ca.725 if Mercian in origin, or after ca.825 if Northumbrian”.³⁶ Since it is evident that a Mercian origin is more likely than a Northumbrian one,³⁷ based upon the poet's use of nemne as well as other indicators, it is likely that Beowulf was composed prior to ca.725.

Two other arguments based on the text's orthography also serve to confirm the existence of a written exemplar at a date that is far earlier than that of Beowulf's extant manuscript. The first of these was advanced long ago by Klaeber in his edition of Beowulf. Essentially, Klaeber made a claim for a written exemplar for the poem, whose existence he dated ca.750, based on the orthographic evidence provided by various stray language forms—especially those in an East Anglian dialect—that are found scattered throughout Beowulf, a poem written primarily in West Saxon.³⁸ According to Klaeber, the presence of these stray language forms in the extant manuscript was the result of a long process of copying and recopying a text whose original text had been written in an East Anglian dialect. During this long process of transmission, certain forms were updated while others, whether by simple omission or due to metrical considerations, were not. It was the existence of this medley of spellings in the text that led Klaeber to declare that “the significant coexistence in the manuscript of different forms of one and the same word, without any inherent principle of distribution being recognizable, points plainly to a checkered history of the written text”.³⁹

However, Klaeber's hypothesis has come under attack recently, notably by Kiernan and J.F. Tuso, who claimed that the various language forms in Beowulf were the result of a traditional, conservative poetical diction that employed older language forms with little regard for their specific dialect.⁴⁰ While these arguments have managed to weaken Klaeber's claim for an early East Anglian genesis for the poem, they have by no means discredited the hypothesis entirely. In fact, Klaeber's claim has been taken up recently by Sam Newton in a wide-ranging and persuasive work.⁴¹

A second argument, formulated by C.L. Wrenn, addressed the evident scribal confusion of d for ð (eth) and d for þ (thorn). Wrenn argued that a number of textual corruptions could be explained as scribal in the process of modernizing a text which used d for sounds that would later be represented by ð or þ.⁴² Prior to the eighth century, ð evidently could be employed either as a voiced spirant or as a stop.⁴³ Based on this evidence, Wrenn concluded that a written exemplar of Beowulf must have existed already by the late eighth century, when ð was introduced as a letter.⁴⁴ While some of these textual corruptions might be explained by simple mechanical error—a scribe forgetting to cross an eth—given the number of instances that this occurs in our text, it seems unlikely that this explanation is entirely satisfactory. In any event, it is almost certain that scribal confusion over d and ð or þ is evidenced at 1.1278 in the poem (which has þeod for deod/dead), a confusion that cannot be attributed to mechanical error (since a thorn, which is not crossed, is involved).⁴⁵ If this view is correct, Wrenn's argument for an eighth-century exemplar for the Beowulf manuscript can be seen to be valid. In fact, Wrenn appears to have been overly cautious in his dating for the introduction for the letter ð. Recent scholarship on the Epinal Glossary has dated this development much earlier than Wrenn had supposed, perhaps even to the seventh century.⁴⁶

In any event, Wrenn's call for an eighth-century exemplar for Beowulf has been bolstered recently by Anita Riedinger, who has advanced a compelling argument for an eighth-century date for a written copy of Beowulf. In that article,⁴⁷ Riedinger sets forth new evidence that demonstrates unequivocally that the author of Andreas had before him a copy of Beowulf from which he borrowed, often and methodically, during the course of writing Andreas. She concludes that it was this methodical, and often verbatim, borrowing of formulaic expressions from Beowulf on the part of the Andreas-poet that accounts for the striking verbal similarities of the two poems, and not literary convention or a shared oral-formulaic tradition.⁴⁸ In fact, Riedinger concluded that “the formulaic relationship between Beowulf and Andreas is stunning”, stating that “I believe that the later poet needed no other technique of composition here than a Beowulf-manuscript at his elbow (with his Latin source nearby) and that this explains the very special relationship that scholars long believed existed between these two poems alone”.⁴⁹

Scholars have assigned dates to Andreas' composition that range from the late-eighth to the late-ninth century, with most dating schemes centering on the early to mid-ninth century.⁵⁰ Thus, based on Riedinger's analysis, a written copy of Beowulf can be placed in the early ninth century at the latest, or even somewhere in the eighth.

Finally, while not offering absolute and irrefutable proof on their own, there are a few other arguments, uniformly pointing to an early date for the composition of Beowulf, that reinforce the conclusions provided by the other types of data. In particular, one argument shows that, in sharp contrast to Anglo-Saxon poems of the tenth century—most notably Brunanburh and Maldon—there is a marked absence of identifiable Scandinavian loan-words from Viking-Age Britain in Beowulf.⁵¹ It is thus unlikely that the material used by the Beowulf poet was derived from information obtained by Anglo-Saxon interaction with Vikings of this period. This point is reinforced further by the forms of the names of Scandinavian dynastic figures that are found in Beowulf, which, as they stand in the poem, “although etymologically related to their later Northern equivalents, none of these names show any evidence for any Scandinavian sound changes.”⁵² This point is reinforced yet again by the accuracy with which Beowulf presents the names of its Scandinavian dynastic figures, again in sharp contrast to the inept transposition of Scandinavian names found in the Anglo-Saxon Chronicle.⁵³ When taken together, these various indications uniformly suggest that Beowulf's composition is unlikely to have been as late as the first part of the ninth century, and was in fact probably earlier.

The initial decades of the eighth century then, or perhaps even a little earlier, is the approximate limit to which most of the orthographic and linguistic evidence of the extant text can carry the argument for a written exemplar. This does not mean that a written version of Beowulf could not have existed earlier than the eighth century. Theoretically, there is nothing that would preclude, at least in terms of Anglo-Saxon orthography, a written form for the poem in the seventh century. As Dumville has noted, the Old English language evidently had developed an efficient orthography no later than the 670's, which is the date of the earliest charters and the first works of Aldhelm.⁵⁴ Still, it is unlikely that a written copy of Beowulf existed prior to c.685, a terminus a quo for the composition of Beowulf and the other longer poems that is based upon the higher levels of non-contraction that is evidenced in the early glossaries.⁵⁵ In sum, while Anglo-Saxon orthography would permit the existence of a copy of Beowulf as early as c.670, the orthography of the extant manuscript itself would call for a written exemplar for the poem that is anywhere within the 685-725 period.⁵⁶

An oral form of the poem certainly preceded this 685-725 period, the period in which the poem assumed its written form. As was related in Chapter 3, because of its oral-formulaic nature, the poem could have been the product of a process that might have stretched over several centuries. However, there are several arguments that can be used to establish a more precise period in which the poem and/or its component parts were orally composed. The earliest terminus a quo for the poem is based on the one historical fact in the poem upon which scholars can agree: the death of Hygelac, Beowulf's Geatish lord, during a raid against the Frisians. Mentioned in the poem, this event is corraborated by Gregory of Tours in his Historia Francorum⁵⁷, who dated the death of Chlochilaichus (the Latin form of Hygelac) to around 521. In fact, Gregory's corraboration of this event constitutes our only indisputable external evidence for the poem's composition.

Moreover, the internal evidence of the poem also leads one to believe that the poet assumed a ready knowledge and easy familiarity on the part of his audience of various sixth-century Scandinavian heroes, a knowledge and familiarity that provides “solid internal evidence for placing the poem at a time when these stories were still vivid in the memory of the early Angles and Saxons”.⁵⁸ Further, it seems that the poet was familiar with pagan ship burials and other funeral practices, and with the material culture of both early Anglo-Saxon England and of Sweden during the seventh century. In fact, the archaeological evidence of Sutton Hoo's grave goods and artifacts—addressed below in Chapter 7—shows a remarkable resemblance to the material goods—most notably to weapons and armor—that Beowulf describes in great detail.

While it is evident that some sections of Beowulf clearly portray pagan Anglo-Saxon and Scandinavian society, it is equally clear that the poem's “author” was a Christian, both by his references to events in the Bible and by his use of Latin terms that were probably adopted after the coming of Christianity to Britain.⁵⁹ Thus, the poem as we have it must be later than 597, the date of Augustine's mission to the island. Still, the degree to which Christianity actually is in evidence in Beowulf has been a matter of no little debate. While the arguments of this debate are found elsewhere,⁶⁰ the persuasive arguments advanced by William Whallon⁶¹ are of special relevance at this point. Whallon has shown that the knowledge of Christianity exhibited in the poem is rudimentary at best, stating that “the epic knows little of Christianity besides two stories from the first nine chapters of Genesis”.⁶² Whallon also has shown that the words and phrases for God and the Devil in Beowulf need not imply a Christian background for the poem, but could be explained just as easily as deriving from traditional Germanic poetic vocabulary. In fact, “the words fæder, alwalda, and meotod are as biblical as pater, omnipotens, and fatum are in the Aeneid, and Beowulf is to this extent neither Christian nor un-Christian but pre-Christian”.⁶³ From this it should be clear that the Beowulf poet is not writing allegory; Beowulf the Hero is not to be equated with Christ the Savior, nor is Hrothgar's “sermon” to Beowulf to be regarded as some type of Germanic accolade of all things Christian. Quite simply, Christianity as a subject was not high on the list of topics that the poet thought needed to be covered. Instead, the main themes and topics presented in Beowulf are those associated with heroic society. Thus, it seems reasonable to conclude that:

The narrator and the important heroes in Beowulf speak as if they were superficially Christian. Within the confines of the poem itself, however, the Christianity appears rather elementary in quality; scriptual references are limited to perhaps two stories from Genesis, while references to theology or dogma are virtually impossible to identify with an appreciable degree of confidence. Germanic heroic ideals of conduct control the action of the poem; Christian piety is incidental to it. Without the various pious comments and Christian allusions in the poem, Beowulf would remain our finest Germanic poem.⁶⁴

It is likely that Beowulf, whose subject matter is centered on a heroic pre-Christian society, acquired whatever Christian trappings it possesses during its oral stage of existence. This acquisition of Christian themes and concepts would have occurred at some point after the conversion of the Anglo-Saxons to the new religion. It was during this specific period that the originally pagan poem incorporated Christian material during the process of being sung by one or more converted oral poets.⁶⁵ As a consequence, over the course of several generations, this Christian material became incorporated into the poem with varying degrees of intensity and success, depending on the particular portion of the narrative in question. This uneven treatment of Christian material is evident to anyone familiar with the poem. A likely explanation for this uneven incorporation, and for the poem's other discrepencies as well, can be found in the proposal that Beowulf represents a “song amalgam”, a joining together of two or more originally distinct songs.⁶⁶

Once joined, these separate tales—most notably the poem's stories regarding the fights at Finnsburg and at Ravenswood⁶⁷—would have retained their usefulness to an oral poet as distinct units of presentation that could be recited before an audience. In any case, because of the poem's general unity, as well as the essentially uniform densities of formulas and formulaic expressions found in the various parts of the poem, it seems likely that this process of amalgamation already had been completed prior to the poem being written down. Thus, the poem in its present form must be seen as a product of an oral, rather than a written tradition, and “the spectre of a ‘monkish scribe’ tampering with a pre-Christian text, altering a word here and a line there in order to make the heathen king of the Danes sound like a Christian Sunday-School teacher, may be exorcised”.⁶⁸

From the above discussion, it is apparent that the dating of Beowulf presents no easy task. Still, given the evidence presented, I would propose that some sections of the poem—the ones about Finnsburg and Ravenswood, for example—had their genesis in oral form sometime during the sixth century. This contention is supported by three arguments: first, Gregory's corroboration of Hygelac's death in Frisia; second, by the audience's knowledge of sixth-century Scandinavian heroes; and third, by the poet's familiarity with sixth-century funerary practices and material culture of both the Anglo-Saxons and the Scandinavians.

Beowulf—though more likely its component parts—underwent an extensive modification at the hands of converted oral poets as they attempted to make the poem better suited for a Christian audience. This incremental process of modification and alteration of the pagan poem must have occurred somewhere during the period ca.625x700, and eventually produced the version of Beowulf that, with few exceptions, is found in the manuscript. I would propose also that the incorporation of the poem's Christian material occurred during the earlier portion of this period, based upon a degree of sophistication regarding the new religion that only can be described as rudimentary. Further, I would propose that the incorporation of this material occurred at some point before the poem's several component parts were woven into an epic, based on the poem's uneven handling of Christian themes and concepts. This too indicates that the pagan poem's modification and alteration had been effected during the earlier portion of this 625-700 period. It also seems evident that the process of literary amalgamation, which joined together the once-independent component poems comprising the epic, occurred during the latter part of the period, and produced to a great degree the poem as we have it today. However, in light of the uniform densities of formulas and formulaic expressions throughout the epic's various parts, it appears that the amalgamation process essentially was complete before the oral poem was first written down.

The above description of the inception and subsequent evolution of the oral poem, as well as the proposed period in which this process occurred, is attractive for several reasons. First, the 625-700 period is late enough to have allowed sufficient time for Christianity to have taken hold among the Germanic peoples who lived on the island, yet early enough to have allowed for the accurate transmission of sixth-century heroes, societal values, and material culture. Such a time period also is reflected in the elementary quality of the poem's Christianity, and would allow several generations of oral poets enough time to mold Beowulf into some semblance of a Christian poem. Further, the proposed dating is late enough to allow for a period spanning several generations in which oral poets would have woven the varied fabrics of the disparate poems into an epic tapestry, though early enough to ensure that any flaws in its manufacture still would be present when eventually it was written down.

Whether the poem existed in a written form prior to the 685x725 period probably will never be resolved to anyone's total satisfaction. Indeed, it might be possible that the present form of the Beowulf poem-the epic as it is found in the extant manuscript—never existed until the oral poem was set down in writing.⁶⁹ One can state only that ca.685x725 is the earliest date that can be assigned to a written exemplar for the poem, based on the available orthographic and linguistic evidence. This period is attractive for well-known historical and literary reasons that have been presented many times before.⁷⁰ For my part, I see no reason for assuming that the poem did not achieve a written form during this 685-725 period, though the location where the poem actually was written down must remain unknown. While such a proposition cannot be verified beyond doubt, it accords well with the arguments and data that have been presented. Certainly, there is nothing that would prohibit this possibility.

In any case, it seems likely that the lines of Beowulf known as the “Offa Digression” (1931-62) were inserted into the poem during the reign of Offa of Mercia (757-96) as an obvious compliment to that powerful king and warlord.⁷¹ It is impossible to determine whether this interpolation occurred when the text first was committed to writing or whether these lines were added during the course of later copyings. In this regard, I would conclude that Hetware and Hugas, the names of two early continental tribes mentioned in Beowulf, are written interpolations inserted at a late stage of the manuscript's history. In no way should these terms lead one to place the poem's composition within the tenth century, with a terminus a quo of ca.923, as Walter Goffart has done.⁷² The same can be said for arguments regarding the term merewioingas.⁷³ They are interpolations and nothing more, and as such are useful only in providing some insight into the poem's various stages of textual history.⁷⁴ These interpolations show that the epic was the subject of continued modifications—the extent of which will never be known—even after it had achieved a written form.

Notes

1. For a useful history and overview of the debate, see Colin Chase, “Opinions on the Date of Beowulf, 1815-1980”, in The Dating of Beowulf, edited by Colin Chase (Toronto: University of Toronto Press, 1981), pp.3-8. For a summary of the dates that have been advanced for Beowulf, see pp.186-7 (fn.12) in Alain Renoir's A Key to Old Poems: The Oral-Formulaic Approach to the Interpretation of West-Germanic Verse. (University Park: Pennsylvania State University Press, 1988). The amount of scholarship that has been generated and subsequently applied to almost every conceivable aspect of Beowulf, and especially to the question at hand, is enormous. See Douglas D. Short (ed.), Beowulf Scholarship: An Annotated Bibliography, Vol.193 of the Garland Reference Library of the Humanities (New York: Garland Publishing, 1980) or Robert J. Hasenfratz (ed.), “Beowulf” Scholarship: An Annotated Bibliography, 1979-1990, Garland Medieval Bibliographies, Vol.14. (New York and London: Garland Publishing, 1993). The sheer volume and scope of this scholarship is overwhelming, and was itself a key factor in establishing the methodology allowing Chapters 4, 5, and 6 to sequence their arguments.

2. The description in the text is based on the one in Leonard E. Boyle's “The Nowell Codex and the Poem of Beowulf”, in The Dating of Beowulf, edited by Colin Chase (Toronto: University of Toronto Press, 1981), p.23. For the study's purposes, Boyle's description is sufficient. For a thorough investigation of BL MS. Cotton Vitellius A.XV, see Kevin Kiernan, Beowulf and the Beowulf Manuscript (New Brunswick, N.J.: Rutgers University Press, 1981). See also J. Gerritsen's “British Library Cotton Vitellius A.xv—A Supplementary Description”, English Studies 69 (1988), pp.293-302 or his “Have with you to Lexington! The Beowulf Manuscript and Beowulf”, in In Other Words: Transcultural Studies in Philology, Translation, and Lexicography Presented to Prof. Dr. H.H. Meier on the Occasion of his Sixty-fifth Birthday, (Dordrecht: Foris, 1989), edited by J. Lachlan Mackenzie and Richard Todd, pp.15-34. For the Nowell Codex, see Kemp Malone's The Nowell Codex: British Museum Cotton Vitellius A.XV. Second Manuscript. Early English Manuscripts in Facsimile, no.12 (Copenhagen: Rosenkilde and Bagger, 1963).

   Unless otherwise stated, all folio numbers in this chapter follow the official numbering of the British Library. The reader should be aware that Kiernan and Boyle use different numbering systems for the foliations of the Beowulf manuscript. Kiernan uses the older numbering found in the manuscript itself and the one employed in both the Zupitza facsimile (EETS 77, 1882) and Klaeber's edition of the poem. Boyle prefers the 1884 numbering accepted as official by the British Library and found in Kemp Malone's facsimile edition (EEMF 12, 1963). The difference results in Boyle's numbers being three more than the reference numbers of Kiernan's.

3. Neil R. Ker, Catalogue of Manuscripts Containing Anglo-Saxon (Oxford: Oxford University Press, 1957), MS. no.216. Ker (p.281) dates the manuscript as 's.X/XI, equivalent to 975-1025 (i.e. to within 25 years on either side of 1000).

4. See Kiernan, Beowulf Manuscript, especially pp.65-169 and 219-43, and his summary article “The Eleventh-Century Origin of Beowulf and the Beowulf Manuscript”, pp.9-21 in The Dating of Beowulf, edited by Colin Chase (Toronto: University of Toronto Press, 1981). In the latter book, see also Boyle's “The Nowell Codex and the Poem of Beowulf”, pp.23-32. For a quick summary of both sets of arguments, see O'Brien O'Keefe's Visible Song: Transitional Literacy in Old English Verse, Cambridge Studies in Anglo-Saxon England, Vol.4 (Cambridge: Cambridge University Press, 1990), at pp.172-74.

5. Kiernan, “Eleventh-Century Origin”, p.10.

6. Kiernan, Beowulf Manuscript, p.258. The argument is detailed on pp.249-58 and again at pp.271-2.

7. Kiernan, “Eleventh-Century Origin”, p.14. This possibility seems incredible. It is impossible to believe that the poem's sole extant manuscript is also its interim rough draft. Kiernan has overstated the capabilities of the two scribes, especially in regard to the poetic sensitivity and linguistic abilities that he attributes to the second one.

   In “Anglo-Saxon Scribes and Old English Verse”, Speculum 67 (1992), pp.805-28, Douglas Moffat saw that too great an emphasis was being placed on the abilities of Anglo-Saxon scribes by scholars such as Eric Stanley, who seemed “willing to equate facility with the language of the exemplar with probable accuracy as a copyist”, a position that he thought was not entirely tenable (p.809). Moffat also took aim at Kiernan's unfounded assumptions, declaring (ibid) that “far more radical than Stanley in his deference to scribal ability is Kevin Kiernan”. Quite simply, such deference to scribal abilities is likely to prove unfounded.

   Ironically, these overly optimistic views on scribal ability were, in part, in reaction to the earlier, widely-held scholarly opinion that saw these scribes as incompetent dolts who uniformly mangled the manuscripts that lay before them. For more on this, see Moffat's article (op.cit.), pp.806-07.

8. Kiernan's conclusions came under attack almost at once by Boyle in “Nowell Codex”. For other opposition, see the reviews by J.D. Niles and Ashley Crandell Amos of Beowulf and the Beowulf Manuscript, found in Speculum 58 (1983), pp.76-77, and Review 4 (1982), pp.335-45 respectively, O'Brien O'Keefe's Visible Song, pp.173-74, and Sam Newton's The Origins of Beowulf and the Pre-Viking Kingdom of East Anglia (Cambridge, England: D.S. Brewer, 1993), pp.3-5.

   Regarding Kiernan's historical argument (Beowulf Manuscript, pp.18-23), little need be said here except that West Saxon kings also employed some of the same warriors and kings listed in the poem's opening lines (1-63) as a part of their genealogy since the late ninth century. On this point, see Alexander C. Murray, “Beowulf, the Danish Invasions, and Royal Genealogy” and Roberta Frank, “Skaldic Verse and the Date of Beowulf” in The Dating of Beowulf, edited by Colin Chase (Toronto: University of Toronto Press, 1981), pp.103-05 and 128-29. In essence, it can be argued that the poem's genealogy would have been “panegyric” to either West Saxon or Danish kings. Further, I would credit the scribes with enough common sense to be able to distinguish the difference between the migration-era Danes whom the poem portrayed, and the contemporary Danes who were busy ravaging the countryside.

9. Boyle, “Nowell Codex”, p.27.

10. Ibid, pp.27-29. Especially useful is his schematic on p.29 showing the relationship between the “Copy-Beowulf” and the “Nowell-Beowulf”. Boyle's chart on p.24 clearly presents his view of the divisions and organization of the entire codex. For the differences between Kiernan's and Boyle's views of quire organization, see Newton, The Origins of Beowulf, pp.3-4.

11. Despite Kiernan's claims, Quire 5 should be seen as a normal four-sheet quire. As Newton has noted (ibid, pp.4-5), the ruling of this quire differed from that of both the preceeding and following folios, reinforcing the impression that it was a normal four-sheet quire. The same conclusion has been reached by: Boyle (“Nowell Codex”, p.23); R.D. Fulk (“Dating Beowulf to the Viking Age”, Philological Quarterly, 61 (1982), pp.352-3); and Gerritsen (“Have with you to Lexington!”, pp.15-16).

12. See Newton, ibid (pp.8-9) and the references cited.

13. Like Kiernan, O'Brien O'Keefe has been accused of having too high a regard for the abilities of Anglo-Saxon scribes, especially in regard to their alleged abilities in the Anglo-Saxon oral poetic tradition. For more on the abilities of these scribes, see Moffat, “Anglo-Saxon Scribes”, pp.810-14.

14. In Visible Song, O'Brien O'Keefe concluded that “it is quite unlikely that the pointing in Beowulf represents eleventh-century practice or was added by the scribes who copied this work” (p.174). Also, she equated the pointing practice evidenced by Beowulf with that found in Solomon and Saturn, which Ker has dated s.xmed (p.178), while noting that it also likely antedates the copying of the B-text of the Chronicle (p.179). It is useful to reiterate that both Scribes A and B used the same scribal practices in regard to pointing in their portions of the manuscript, a fact that indicates, in no uncertain terms, the existence of an earlier written exemplar of the poem's text.

15. Ibid., p.179.

16. As we saw earlier in Chapter 1, this belief really is not a valid one; the social and behavioral sciences employ inductive, probabilistic logic in establishing hypotheses, whose proofs are derived primarily from statistical data. By their very nature, such proofs are relative rather than absolute, and are to some degree always subjective. For further opposition to the application of purely objective, mechanically applied standards to the tests of historical linguistics, see Fulk's Old English Meter, especially at pp.6-24, which was written largely in direct response to Amos' strict methodology for evaluating the validity of linguistic tests.

17. Amos, Linguistic Means, p.167. The book provides a useful overview, summary, and evaluation of a wide array of linguistic tests. Despite its belief that the methodology of the natural sciences could be applied to historical linguistics, it remains a seminal and valuable work. This is not to imply that Amos felt that all linguistic tests are without value. Amos relates (p.167) that “the syncope and apocope of unaccented vowels, and in particular u-apocope, predate all surviving Old English texts, and thus provide a terminus a quo for the literature” in the seventh century or earlier. Further, according to Amos (pp.167-8), other potentially useful dating criteria would be the shortening of the vowel lengthened after the loss of post-consonantal prevocalic h and, in a very limited number of cases, where “the contraction in words with original hiatus between vowels, or hiatus caused by loss of intervocalic h, j, or w affects the number of syllables in a verse”. In addition, datable phonological changes can be used to “establish secure dates for those texts in which authorial spelling is preserved (inscriptions, early manuscripts, verses with acronyms)”. A final valid criterion would be the lifetime of certain words where it can be determined (p.167).

18. Ibid., p.102.

19. Ibid. It is this situation that led Amos to state (p.168) that “Alliterative practice with respect to palatal and velar g is by far the most reliable of the purely metrical tests”.

20. Ibid, pp.100-102 and p.167. In fact, Colin Chase noted in “Opinions on the Date of Beowulf” (p.5, fn.7), in his The Dating of Beowulf that this important finding was presented by Amos and several others at the Toronto conference from which The Dating of Beowulf was to emerge. Yet, as Chase notes, “this point failed to appear in any of the papers which follow”.

21. Amos, Linguistic Means, pp.167-8. For more regarding this test, see R.D. Fulk, “West Germanic Parasiting, Siever's Law, and the Dating of Old English Verse”, Studies in Philology 86 (1989), pp.117-38. For examples of contraction and loss of intervocalic h, j, and w, see Klaeber's Beowulf, pp.274-5. Thomas Cable, “Metrical Style as Evidence for the Date for Beowulf”, in The Dating of Beowulf, edited by Colin Chase, pp.77-82 (Toronto: University of Toronto Press, 1981), also viewed (p.81) the higher proportion of disyllabic diphthongs in Beowulf as a possible indication for an early composition for the poem.

22. For more on the validity of the loss of intervocalic h as a valid chronological indicator, as opposed to the loss of intervocalic j or w, see Chapter II in Fulk's Old English Meter.

23. Newton, The Origins of Beowulf, p.13. For Newton's references for this statement, see the works cited (ibid) in fnn. 55,56,57, and 58.

24. For a discussion, see Amos' Linguistic Means, pp.40-9.

25. Ibid., p.168.

26. Randolph Quirk, “On the Problem of Morphological Suture in Old English”, Modern Language Review 45 (1950), p.2.

27. For more regarding this opposition, refer to Fulk's Old English Meter, at pp.116-21.

28. Thomas Cable, review article of Amos' Linguistic Means of Determining the Dates of Old English Literary Texts. JEGP 85 (1986), pp.93-95.

29. Cable, “Metrical Style”, p.82.

30. This subject forms the topic of Chapter VI in Fulk's Old English Meter. The reader is encouraged to refer to this for a much fuller discussion of Kaluza's Law and its implications for the dating, in both relative and absolute terms, of Beowulf and other Old English poetry.

31. Ibid., p.153.

32. Ibid., p.164. While Fulk does admit that most poems do observe the distinction between long and short endings to some degree, he attributes this (ibid., p.165) to the formulaic character of the verses in which this observance is found; usually those involving poetic compounds.

33. Ibid., pp.164-5.

34. Ibid., p.166. No other explanation is really credible. As Fulk explains (Ibid., p.167):

    Appendix C demonstrates that the distribution of vocalic endings is irregular, and too complex to have been preserved as consciously archaic language: there must have been a phonological difference between the two types of endings at the time Beowulf was composed.

35. Ibid., p.167.

36. Ibid., p.390. The dating arguments and references for the shortening of the long endings are found on pp.381-90.

37. The arguments for this contention, as well as further references, are found in ibid., pp.390-91.

38. For a more detailed look at his arguments and examples, see the introduction to his Beowulf and the Fight at Finnsburg, 3d ed. w/supplements (Lexington, Mass.: D.C. Heath & Co., 1950), pp.lxxxi-lxxxviii. As for the poem's date of composition, Klaeber stated that “the only extant manuscript of Beowulf was written some two and a half centuries after the probable date of composition” (p.lxxxviii).

39. Ibid., p.lxxxiii.

40. Kiernan, Beowulf Manuscript, pp.48-50 and J.F. Tuso, “Beowulf's Dialectal Vocabulary and the Kiernan Theory”, South Central Review 2 (1985), p.3. For other opposition to Klaeber and for details, see Newton, The Origins of Beowulf, pp.11-12.

41. Using a wide-ranging and impressive array of evidence, Newton makes a convincing case for an East Anglian genesis for Beowulf in his The Origins of Beowulf. His book is an important work and one that, in my opinion, will be crucial in any future discussions regarding the dating of Beowulf.

42. For a listing of these possible corruptions, see Newton's The Origins of Beowulf, p.9 (at fn.35).

43. Ker, Catalogue of Manuscripts, p.xxxi.

44. C.L. Wrenn, Beowulf: With the Finnsburg Fragment, 3d ed. with revisions and preface by W. F. Bolton (London: Harrap, 1973), p.17. As noted in Newton's The Origins of Beowulf (p.9 at fn.35), Wrenn first advanced this proposal in his “The Value of Spelling as Evidence”, Transactions of the Philological Society (1943), pp.14-27.

45. Newton, The Origins of Beowulf, pp.9-10.

46. On this, see Malcolm Parkes, “Palaeographical Commentary”, in The Epinal, Erfurt, Wurden, and Corups Glossaries, edited by B. Bischoff, et al (Copenhagen: Rosenkilde and Bagger, 1988).

47. Anita R. Riedinger, “The Formulaic Relationship Between Beowulf and Andreas”, in Heroic Poetry in the Anglo-Saxon Period: Studies in Honor of Jess B. Bessinger, Jr., Studies in Medieval Culture XXXII, edited by Helen Damico and John Leyerle (Kalamazoo Michigan: Medieval Institute Publications, 1993), pp.283-312.

48. Ibid., p.299.

49. Ibid. Riedinger (ibid., pp.304-5) saw it as likely that the Andreas-poet employed a written manuscript of Beowulf, rather than working from memory. In any case, she concludes (p.305) that “whether the Andreas-poet knew Beowulf in a written or memorized form, however, it is clear that his source was very close to the version that survives in Cotton Vitellius A.xv”.

50. See ibid, pp.305-6 for details about the dating schemes advanced for Andreas.

51. Newton, The Origins of Beowulf, p.14. As Newton relates, both Klaeber (Beowulf, p.cxvii) and Roberta Frank (“Skaldic Verse”, p.123) had noted Beowulf's lack of these Viking-Age Scandinavian loan-words.

52. See ibid., pp.14-15 for the argument and its references.

53. See ibid., pp.15-16 for details and further references.

54. Dumville, “The Uses of Evidence”, p.120.

55. Fulk, Old English Meter, pp.380-81.

56. Ibid., pp.389-90.

57. Gregory of Tours, The History of the Franks, L. Thorpe, tr. (Harmondsworth: Penguin Books, Ltd., 1974), III.3. For more on Hygelac's raid, and the possible interrelationships between the poem and the Frankish sources, see Dorothy Whitelock, The Audience of Beowulf, pp.38-54. See also Walter Goffart, “Hetware and Hugas: Datable Anacronisms in Beowulf”, in The Dating of Beowulf, edited by Colin Chase (Toronto: University of Toronto Press, 1981), pp.83-100.

58. Chickering, Beowulf: A Dual-Language Edition, p.248.

59. See Newton, The Origins of Beowulf, pp.13-14, for an overview and further references pertaining to the two categories of Latin loan-words found in Beowulf, and their utility in providing a terminus ad quem for the poem.

60. While the debate usually is framed around the question of whether Beowulf is essentially a “pagan” or “Christian” poem, Cherniss has suggested that it might be more productive to view the poem's secular elements as “pre-Christian”, “Germanic”, or “heroic” rather than as either “pagan” or “heathen” (Ingeld and Christ, p.125). Cherniss provides (ibid, pp.125-34) an overview and summary of the arguments advanced by H.M. Chadwick, Fr. Klaeber, J.R.R. Tolkien, et alia regarding the Christianity (or lack thereof) of the poem. It is an issue that has been argued forcefully and cogently by both sides.

61. William Whallon, “The Christianity of Beowulf”, Modern Philology LX (1962), pp.81-94 and “The Idea of God in Beowulf” in Proceedings of the Modern Language Association LXXX (1965), pp.19-23. These points were brought out later in his Formula, Character, and Context: Studies in Homeric, Old English, and Old Testament Poetry (Washington, D.C.: The Center for Hellenic Studies, 1969). In that work, Whallon clearly and effectively showed that most, if not all, of the poem's “Christian” words and motifs could just as easily be explained within a Germanic, pre-Christian context. (See Chapter 4, pp.117-138).

62. Whallon, “The Christianity of Beowulf”, p.81.

63. Whallon, “The Idea of God in Beowulf”, p.20.

64. Cherniss, Ingeld and Christ, at pp.133 and 149.

65. Robert D. Stevick, “Christian Elements and the Genesis of Beowulf”, Modern Philology LXI (1963), p.88. Like Magoun's article cited above, this article (pp.79-89) should be viewed as seminal in the application of Oral Theory to Beowulf studies.

66. John Miles Foley, “Beowulf and Traditional Narrative Song: The Potential and Limits of Comparison”, in Old English Literature in Context: Ten Essays, p.136. As we see below in Chapter 7, these narrative inconsistences are a reflection of the oral-formulaic tradition that is represented by the poem.

67. The Fight at Finnsburg (1068-1159, 874-902) and The Battle of Ravenswood (2922-98) certainly must have had a separate existence apart from the Beowulf poem as we now have it. See Nicholas Jacobs, “The Old English Heroic Tradition in the Light of Welsh Evidence”, CMCS No.2 (Winter 1981), pp.9-20.

68. Cherniss, Ingeld and Christ, p.134. In all fairness to the monks, it should be noted, as Dumville has done (“The Uses of Evidence”, pp.109-21), that it is quite possible that Anglo-Saxon monks of the seventh and eighth centuries, trained in the Irish tradition, would have had no difficulty whatsoever in copying down any kind of heroic poem, even if the poem's subject matter alluded to a pagan past. Irish monks of the same period who were working in Ireland certainly had no qualms about committing to writing the blatantly pagan poems of the Ulster Cycle. The study, in fact, agrees that the poem's eventual existence in manuscript form is attributable to the efforts of a monastic community (though it is impossible to say precisely which one), and that such an existence may very well antedate the date provided by the orthography of the extant manuscript. The study is attempting to show that the poem as we now have it is the product of an oral tradition and a process of amalgamation that must be seen as a fait accompli before the poem was committed to writing.

69. The interaction between the written and oral traditions as the poem was committed to writing might have influenced the bard dictating the text. Elaborating on this point in “From Horseback to Monastic Cell”, Opland notes (p.43) that “one of the consequences of the meeting between an oral tradition and writing: the issue may well be a longer and fuller text, a version that could never have any oral existence among the people. Beowulf might well be such a product”.

70. For the historical arguments, see Dorothy Whitelock, The Audience of Beowulf, or Patrick Wormald, “Bede, Beowulf, and the Conversion of the Anglo-Saxon Aristocracy”, in Bede and Anglo-Saxon England: Papers in Honour of the 1300th Anniversary of the Birth of Bede, No.46 British Archaeological Reports (Oxford: BAR Publications, 1978), pp.32-95, or Eric John's “Beowulf and the Margins of Literacy”, Bulletin of the John Rylands University Library of Manchester 56 (1973-74), pp.388-422. See also Dumville's wide-ranging historical arguments in his “The Uses of Evidence”, pp.109-60. Finally, see M.J. Swanton's comments about the poem and the reign of Offa in his Crisis and Development.

71. This is not a novel idea. The digression has long been seen to be an indication for the poem's dating in dating schemes proposed by many scholars. For example, see Whitelock, Audience, pp.57-59 and 63-64. For an overview of this subject, see Chase, “Opinions on the Date of Beowulf, 1815-1980”, p.5.

72. Goffart, “Hetware and Hugas”, pp.83-100. Hetware is found on lines 2363 and 2916 of the poem; Hugas on lines 2502 and 2914. The first term's use in Britain is dated by Goffart to no earlier than the second half of the eighth century, while he assigned a terminus a quo of ca.923 for the second one (p.100). The article offers a well-written and thoughtful analysis, and one that has my wholehearted approval with the exception of his specific use of these terms to date the poem's composition to the second quarter of the tenth century. Such a date is simply at odds with the available orthographic and linguistic evidence, as well as with the oral-formulaic process. I would agree that these terms, if they are viewed as items of interpolation into a written text, can be of great use in helping to determine the textual history of the manuscript. However, neither Hetware nor Hugas can be used in any effort to date the poem's composition per se.

73. Louise E. Wright, “Merewioingas and the Dating of Beowulf: A Reconsideration”, Nottingham Medieval Studies 24 (1980), provides a helpful summary of the major arguments concerning the term on pp.1-2. She herself assigns a terminus a quo for the poem, based on merewioingas, to sometime after 751, and possibly as late as the early ninth century (p.5). Again, I would see this as a written interpolation belonging to an earlier stage of the manuscript's textual history than that evidenced by Hugas, though not necessarily Hetware.

74. One of the major contentions in Swanton's Crisis and Development is that Beowulf reflects the political conflicts that had arisen between the older (horizontal) and newer (vertical) concepts of kingship and society during the course of the eighth century. During this period, the older concept of the king as vox populi was replaced by one which viewed the king as vox Dei as one aspect of a general stratification of the political structure and the society at large. For Swanton, this conflict forms a persistent theme throughout the poem, and is an indicator of the period in which the poem was composed.

    However, if undercurrents of this eighth-century political/philosophical conflict indeed are present in Beowulf (and I am not convinced that this is the case), then the presence of such a theme should be seen as an interpolation to the text, the result of a scribe's or poet's efforts during the latter stages of the poem's history in much the same manner that the incorporation of Christian motifs and themes had been effected during the earlier period of the poem's evolution.