Dating Beowulf Part III: Late Beowulf
(Do I contradict myself? Well then I contadict mys— er... sorry about that.)

Previously, on "Dating Beowulf": we discussed the range of possible dates (514-1025) and then examined the arguments for an "early" date in the 500's, 600's or 700's.

Now we turn to the arguments for a "late" Beowulf, those which support a date for the composition of the poem in the 800's, 900's and early 1000's. Note that many of the arguments for dating Beowulf late are the same as the "problems" for dating it early, and vice versa. So, for example, the observance of Kaluza's law is both evidence for an early date and against a late date, but I am not going to repeat the arguments from Part II in their contrapositive form in Part III, so it may seem like Part III is less fully argued than Part II. Although the contrasting length of the two posts may make it seem so, I am not trying to give short shrift to the arguments for the late dating.

The Viking Era (800-900)

Support: Scandinavian raiders began attacking England in the very late eighth century (Lindisfarne was sacked in 793) and the size and frequency of the attacks increased into the ninth century. Eventually, Viking armies began staying in England over the winter rather than returning to Scandinavia, and large portions of England came under Viking rule. By the time of King Alfred's successful defense of Wessex at the end of the ninth century, much of England had been ravaged at one time or another, but Danes had also settled into peaceful and prosperous living alongside (and intermarried with) their English neighbors. Stories celebrating Scandinavian heroes could thus have become part of the cultural background of an English-speaking poet, explaining both why Beowulf was in English and why it celebrates (in its own way) Danes and Geatas. King Alfred's program of vernacular literacy, Alfred's interest in Anglo-Saxon poetry, and the peace treaties he negotiated with the Danes have suggested to some that the Alfredian period might be a good home for Beowulf. At least one scholar attributes the poem to Alfred's priest Æthelstan, and another suggests locating Beowulf at the court of King Alfred or of the Mercian king Wiglaf who came to the throne in 827.

Problems: Although I do not agree that the Danes are the unqualified heroes of the poem, they aren't the bad guys, either. Given that the Danes (and other Scandinavians) had spent this century ravaging England, murdering its people, plundering its resources and taking over a significant portion of its lands, we might be surprised to find a sympathetic portrayal of Scandinavian people in a poem written in English. Dorothy Whitelock argued that a poet would not want to have recited the opening lines of Beowulf, which proclaim the greatness of the Danes, in England after the beginning of the ninth century and thus the poem should be dated earlier.

The Reform Period (900-1000)

Support: The Benedictine Reform really begins with the coronation of Æthelstan in 926 and it reaches its high point in the 970's during the rule of King Edgar. During this time period, Alfred's program of vernacular literacy came into its own. A rebirth of Latin learning was led by Dunstan and Æthelwold, but texts in English were produced and copied in significant numbers. All of the major poetic manuscripts (Junius Manuscript, Exeter Book, Beowulf Manuscript, Vercelli Book) were copied during this time period.

We also have a charter from Wiltshire dating to 931 which includes the places "Beowan hamm" (Beowa's home) and "Grendeles mere" (Grendels' mere) in the boundary clause, suggesting that stories of Beowulf might be circulating in England during this period. There is also a much-discussed parallel between Blickling Homily XVI and the description of Grendel's mere that many scholars have concluded shows that the Beowulf poet knew the homily. The Blickling Homily manuscript dates from around 971, thus indicating that Beowulf would have to have been written after this point.

The combination of vernacular literacy, cultural revival, English national pride (seen in the Benedictine reformers' emphasis on Pope Gregory being particularly concerned about England, for example) combined with the amalgamation of Scandinavian and English people into a single nation under King Edgar, a major patron of the Reform, and the "imperial" practices under Edgar could be a possible environment for Beowulf. We also know that the poem was copied and thus was at least received if not created during the tenth century (though see below for dissent).
In addition, there are at least some passage of the poem, most famously lines 175-188, which read very much like later (10th century) Christian, homiletic material (Hrothgar's "sermon" is another passage, and I would add that line 1864 "feels" more like Latin-inspired, balanced rhetoric to me). Tolkien solved the problem by arguing that 181-88 were probably a later interpolation on top of an original poem. But other scholarship has found word pairs or formulae throughout Beowulf that are only found elsewhere in later, Christian, homiletic texts.

Problems: Most of the arguments against a late date have already been given as arguments for an early date, so I won't rehearse them here at length, but the standard "Anglo-Saxons, speaking English, would not have liked a poem about the greatness of marauding Danes" apply to the tenth century, particularly after Danish attacks resumed during the reign of Æthelræd. Also, finding bits of Beowulf in later, Christian materials could support an 8th-century date as well, if those phrases had antecedents in that time period in now-lost texts (this is the problem with having so many of our texts in 10th-century copies). The parallels with the Blickling Homily are interesting, but it is only one small passage and the two passages may have a common origin. The Wiltshire charter only shows that there were some tales with versions of the names in Beowulf floating around; place names are very conservative, so this could just be evidence for earlier circulation of the stories that only got recorded in the tenth century.

Anglo-Danish Rule (1000-1025)

Support: The manuscript dates from between 975 and 1025, but that date is a creation of the average date of the two hands (Scribe A and Scribe B). Scribe A, who copies the first part of Beowulf and the prose texts in the manuscript, would probably in isolation be dated to the early part of the 11th century. Scribe B, taken in isolation, would be dated to the end of the tenth century. Taking the later date would date Beowulf to early in the 11th century.
One page of the manuscript is a palimpsest: the text has been scraped off the leaf and then re-written. Kevin Kiernan thinks that in this re-writing we see the work of the poet, who is also Scribe B, very carefully and artfully combining two pre-existing poems and joining them at exactly this point.
Kiernan thinks that the layout of the manuscript (ruled for 20 lines per page and laid out so that the manuscript pages match the hair sides and flesh sides of the leaves for each open folio -- therefore the color of the underlying membrane for the pages that face each other is the same) and the fact that the manuscript is written in two different hands (like the Blickling Homily manuscript), shows that Beowulf comes from the same scriptorium (if this is true, however, it could also support merely a generic "late" date).
The reign of King Cnut, who was Danish, would be an opportune time to "publish" a poem in which the Danes and other Scandinavians (Geats) were portrayed heroically or at least sympathetically. Some of my students like this "sucking up to Cnut" thesis a great deal. Cnut's reputation is that he tried to be more English than the English, but he was of course still a Dane. So an English poem about Scandinavians seems like it might have appeal in his court.

Problems: Dating the manuscript to 1016 is definitely pushing the envelope for possible dates based on handwriting. Even scribe B, who corrects the entire manuscript and is under this theory the poet seems to have trouble with the Merovingians passage, and I think the Modthryth passage also looks like the scribe may not have entirely understood that Modthryth was a personal name until after he wrote it (wacky Drout theory; feel free to ignore). The palimpsest page is so difficult to read that much of the interpretation has to rely on conjecture, and different scholars make different conjectures.

General arguments that apply to all "late" datings:

The meter of Beowulf seems much "tighter" than that of the dateably late poems, such as The Battle of Maldon (which obviously cannot date from before the 10th century). There are also a variety of linguistic tests that have been devised to attempt to date poems, the specifics of which I am leaving out due to their technical nature. Almost all of these tests (and all of them are disputed) seem to point to an early rather than a late date (I should note that the linguistic and metrical tests are often considered to be more reliable than evidence of, say, verbal parallels, but the applicability of any specific linguistic or metrical test is hotly debated).

Additionally, a number of scholars, including J.R.R. Tolkien (whose unargued hunches are worth taking very seriously) think Beowulf is more like the poem Exodus than it is any other Old English poem (this is obviously subjective), even though Andreas actually has a number of lines or formulae that are also found in Beowulf. Exodus is usually considered, on the basis of language and meter, to be an "old" poem (even though it too exists only in a 10th-century manuscript). It also seems as if the Exodus scribe was having great difficulty understanding his exemplar. This difficulty could be explained if the exemplar for Exodus was in an older form of the language, unfamiliar to the 10th-century scribe of Exodus (on the other hand, Emily Thornbury has recently argued that Christ and Satanalso in the Junius manuscript, might be the mess it is due to a poet copying from a damaged exemplar). So even if we do not accept the linguistic-chronology tests (like Kaluza's Law or various rules about spelling) as giving us accurate information about what century a manuscript was copied, we could perhaps use general, holistic comparison. There are obvious problems with this approach (hunches are hunches and guesses are guesses), but it is also the case that we are dealing with a literary artifact, so that intuition and holistic analysis may count for something.

Finally, the strongest support for late dating always seems to revolve around the manuscript. We do know when it was copied (within 50 years), and, if we do not want to accept some kind of composite authorship (i.e., parts of Beowulf are early and parts are late), we keep coming back to the manuscript, as it is, and its obvious relationships (which are with tenth-century texts). The problem here is that the vast majority of Anglo-Saxon manuscripts were copied in the tenth century, and we know for certain that many of these are copies or translations of much earlier texts, so a tenth-century copy does not prove a tenth-century date of composition.

In the next post I will try to tie all of this together, disclose my own biases, and lay the whole issue to rest (ha!).