In his famous essay on ‘The Alliterative Metre’, C. S. Lewis memorably declared: ‘A man who preaches a metre must sooner or later risk his case by showing a specimen’. The number of my public writings on the technicalities of Old English verse composition is now considerable (others would no doubt prefer to call it excessive), and so I believe it is time for me to risk my case and present The Fall of the King, my most serious attempt so far at composing neo-Old English verse. This ninety-eight-line poem in Old English retells in classical metre and idiom King Théoden’s death at the Battle of the Pelennor Fields and the encounter between Éowyn and the Witch-king of Angmar as narrated in The Return of the King. My poem is thus a tribute to J. R. R. Tolkien, who has been described as ‘one of the most prominent composers of neo-Old English’. I would like it to be read as the manuscript copy of the lay that a bard from Rohan (an individual with a learned mind and a gentle heart) composed after the War of the Ring upon realizing that the enemy that Théoden dared to confront was of a more evil nature than his lord could ever have suspected. The publication of The Fall of the King, it is to be hoped, will encourage others to try their hand at the noble art of Old English versification. Before showing my poem and discussing verse composition, however, a few words are in order about the value of this creative activity for students and scholars of Old English poetry.

Latin verse composition has traditionally been seen by many as a central component of classical education. After a decline in popularity that started in the second half of the twentieth century, it is now experiencing a resurgence among teachers of Latin, thanks in part to the notable success of the active Latin movement. That this should be the case is hardly surprising. Experience tells us (and in this receives ample support from numerous studies in the field of second-language acquisition) that learners learn best through creative interaction with their objects of study. The beneficial effects that verse-making has had on generations of students of Latin poetry are well-known and have been discussed elsewhere.

Much of A History of Old English Meter (Fulk 1992) is concerned with chronological questions, and it is matters of chronology that have elicited the most fervent responses from critics and admirers alike. What has often been overlooked in its aftermath is the critical reassessment of Old English metrical theory that A History of Old English Meter also contains. In the introduction, for example, Fulk observed that the Beowulf poet's compliance with Kaluza's law provides firm indication that Sievers’ positional analysis of Old English meter is essentially correct (1992: §§26, 65, 69). Moreover, in chapter VII, in which he endeavored to gauge the chronological significance of the variable metrical behaviour of so-called tertiary stress, Fulk detected a regularity that led him to conclude that syllable quantity is more integral to the formation of metrical ictus than phonological stress (1992: §260), thereby making a significant revision to traditional Sieversian metrics. This conclusion, in conjunction with some distributional evidence from a large corpus of Old English poetry, allowed Fulk to demonstrate that Bliss's scansional system (for which see Bliss 1962 and 1967), despite its widespread use in the profession, is in actuality incompatible with Sievers’ and therefore fundamentally erroneous.

This crucial aspect of A History of Old English Meter, however, has either passed unnoticed or been misunderstood by the majority of Old English scholars. In one of the most visible elementary essays on Beowulfian meter, for instance, Robert P. Stockwell and Donka Minkova describe Fulk's work as “a triumph of the Sievers-Bliss-Cable tradition” (1997: 58), a statement that is not quite accurate in the light of Fulk's conclusions about metrical theory. A.J. Bliss cannot be regarded as the successor to Eduard Sievers if, as Fulk demonstrated, it is precisely Bliss's departures from Sievers that constitute the main flaws of Blissian metrics. Consequently, although it is fair to say that Fulk's work is a victory of Sievers, it is paradoxical to consider it also a triumph of Bliss. One factor that underlies the scholarly community’s failure to apprehend Fulk's views on Old English metrical theory is the fact that A History of Old English Meter, in part because of its title, has been taken to be a chronological study exclusively, and hence its theoretical component has been relegated to a second place.

The correspondences between the names in the Scylding genealogy at the beginning of Beowulf and three names in the upper reaches of the genealogy of Æthelwulf in the Anglo-Saxon Chronicle, Beaw, Sceldwa and Sceaf, frequently appear in arguments for a late dating of Beowulf. But these arguments overlook many aspects of Æthelwulf's genealogy that disrupt their case for a late dating. As H. Munro Chadwick pointed out over a century ago, the forms Sceldwa and Beaw found in the Chronicle for Scyld and Beow are not West Saxon spellings, and the -wa suffix of Sceldwa and Tætwa suggests that these forms may be archaic. Thus spelling alone indicates that these names were probably copied from an older, non-West Saxon text. Furthermore, the very presence of these names in the royal pedigree is puzzling. On one level the presence of Scyld is easy to explain: Scyld and the Scyldings were famous in heroic legend, and his inclusion in Æthelwulf's pedigree provides reflected glory for the West Saxon dynasty and implies genealogical, political and cultural connections between the West Saxons and the Danes that could be useful for Alfred and his heirs to foster. But on another level his inclusion is rather surprising: according to genealogical conventions, the presence of Scyld implies that the West Saxon royal family is a cadet branch of the Scylding dynasty, and is thus potentially subordinate to Scandinavian rulers in England claiming direct descent from Scyld.

Since the date of the Beowulf manuscript is widely agreed upon, the very question which prompts this volume (and the conference it derives from, and even the 1980 conference with its 1981 proceedings volume) must assume that the date of the poem may not be the same as the date of the manuscript. It is certain that there must have been a moment of first inscription for the poem, and that the time and place of that moment remains a central point of interest for students of the poem. In this essay, I will bring new evidence to bear on this venerable question, and my argument shall be that Beowulf is metrically conservative according to a variety of independent metrical criteria. Further, I will suggest that that conservatism is so varied and consistent as to strongly indicate that the original version of Beowulf must be placed among the very earliest of the longer narrative Old English poems that survive, probably in the eighth century.

Of course, it remains true, I believe, that the moment of inscription is only one of the moments of interest which might engage modern scholars of the poem. As I argued in Authors, Audiences, and Old English Verse, our focus on authorship (and on moments of authorship) may sometimes cause us to lose sight of what can be gained by also considering audience, and I proposed there two later audiences for Beowulf, one located at Alfred's Wessex court in the late ninth century, and another, sometime around the turn of the eleventh century, perhaps in Canterbury, represented most clearly by the author of Maldon.

As the introduction to this collection makes clear, the various forms of linguistic and metrical evidence bearing on the dating of Beowulf point to a date of composition fairly early in the Anglo-Saxon period. In his article for The Dating of Beowulf in 1980, Thomas Cable proposed a rough guide to the metrical dating of poems using the incidence of type C, D, and E verses, which decline in frequency over the Anglo-Saxon period. Cable's criterion places Beowulf toward the beginning of a relative chronology. Since then, much additional metrical and linguistic evidence has been gathered that places Beowulf in the early to mid-eighth century. R.D. Fulk's A History of Old English Meter is the most substantial work of this kind, for it examines the presence of archaic metrical features through-out the corpus of Old English poetry and finds that Beowulf is by far the most archaic poem. Since that work, other scholars have written articles on individual metrical or linguistic features of the poetic corpus, which have corroborated the conclusions that Fulk so carefully reached.

Some scholars, however, remain dubious about the reliability of this type of evidence. At this point, the force of linguistic scholarship is too formidable to be undermined by the doubts raised by E.G. Stanley, who urged that the poem should not be dated by means of sundry linguistic oddities that could well be scribal error or just a few bad lines.

From the publication of the poem's editio princeps in 1815 to the emergence of the present collection two centuries later, few topics in Anglo-Saxon studies have generated as much speculation and scholarship as the dating of Beowulf. Marshaling disparate forms of evidence and argumentation, scholars have assigned dates to Beowulf that range from the seventh to the eleventh century. Various individuals have been unpersuasively identified as the author of Beowulf and dozens of kings, clerics, and contexts have been associated with the poem's genesis. Scholarship on the dating of Beowulf is markedly uneven in quality: alongside sober and thoughtful argumentation, there has been a great deal of improbable hypothesizing about the author of the poem or the milieu in which it was composed. Awareness of the qualitative differences in the scholarly literature is tacitly registered in the relative frequency with which publications are cited, but these differences have rarely received explicit discussion. This introduction to the dating of Beowulf controversy examines the changing standards of evidence, methodology, and argumentation that have attended this topic, particularly in the past thirty years. The dating of Beowulf has not been a static or monolithic subject, but has undergone considerable change in the disputes it connotes and the practices it encompasses. In the following account, emphasis will be given to the reasons for prevailing opinions rather than to the multiplicity of opinions as such.

Arguments about the date of Beowulf are more impassioned than the question seems to merit. Even so, the controversy has its uses. Beowulf is a great work, all agree, but it constitutes only a sliver of the poetic canon and is doubtless more important to Anglo-Saxon culture now than it was a thousand years ago. For all its glory, Beowulf provides no better an index to Anglo-Saxon poetry than Hamlet to Renaissance drama, which is to say that one can know both works well without knowing much about the corpus to which either belongs. It is to welcome and good effect, then, that several chapters in this volume link the date of Beowulf to the date of everything else, which, for purposes of this discussion, is the rest of Old English poetry.

At the Harvard conference, R.D. Fulk argued that the date of the poem's composition is less significant than the means used to hypothesize the date. The introduction to Fulk's Chapter 1 in this volume sums up an extended discussion regarding probability, proof, and linguistic evidence drawn from his History of Old English Meter. Fulk observes that the criteria for dating verse are not uniformly rigorous and that they have not been subjected to uniformly rigorous testing. Words can be counted and their forms analyzed, so that exceptions to linguistic and metrical criteria emerge quickly; in these cases, the relative probability of competing hypotheses can be readily gauged.