Dating Beowulf, Part IV
(or, no really, it's very comfortable up here on this fence)
In previous posts I noted the possible range of dates for Beowulf (515-1025) and gave the arguments, pro and con, for a date of composition in each of the centuries of the Anglo-Saxon period: Migration Period, Conversion Era, "Golden Age," Viking Raids, Reform, Anglo-Danish Rule. Every time-frame has problems and every time-period has something to recommend it. In this last post I will bring up the more complicated "composite-Beowulf theories and then try to come to some tentative conclusions.
Although this approach is not popular now, the idea that Beowulf as we have it is not a unitary poem from a single author has a long and distinguished pedigree. German scholars in the nineteenth century believed that it was possible to dissect a long poem into the component lays (this approach was called Liedertheorie) that had, they assumed, been gathered together to make that poem. The foremost of these scholars was Karl Müllenhoff, who separated Beowulf into various parts, some made by the original poet, others by a later poet, and still others (and the stitching of the whole together) by an interpolator.
Other scholars were less ambitious than Müllenhoff and divided Beowulf differently, but they too saw the poem as a composite structure made up of shorter poems put together. Others saw Beowulf as having the core of a pagan, Germanic poem onto which had been grafted some unsightly Christian excresences (usually lines 175-188 and Hrothgar's "sermon" later in the poem).
A composite structure for Beowulf solves -- by defining out of existence -- some of the problems that we have discussed above. For example, most of the examples of consistency with "Kaluza's Law" occur in passages about war and battle which to some scholars seem to have a very "traditional" feel and could thus be old or traditional passages that were drawn on by a later Beowulf poet. Or, the Christian references could be grafted on later to an older poem.
After the excesses of Müllenhoff and others, there was a reaction against the "dissectors" -- Tolkien, although he thought that a few lines might be a later composition -- argued very strongly for the unity of the poem, and most criticism since 1936 has assumed a unity of authorship and poem (albeit with a wide variety of different structures). Thus even though there are some appealing aspects of a composite authorship, and although medieval authors and scribes had very different ideas of literary "ownership" and authorship than we do, theories of Beowulfbeing written in more than one century have not found much favor with recent critics.
The German word "zussamenhang" means "hanging together," and is to my mind the test of nearly any theory in the "historical sciences," (like paleontology), history and literature. The theory that accounts for the greatest number of significant facts should be the right one. But there is a major weasel-word in that last sentence: which facts are the most "significant" ones? One school of thought would say that "objective" tests are important. Other schools are quite comfortable with "subjective" judgments in matters of literature.
The linguists and metricists point out that their tests are repeatable and objective: scan the lines, tabulate the spellings, etc., and the data either support one theory or another. I have a lot of sympathy for this approach, and I also accept the critique that the linguists make that at least a fair number of the "lit" people are simply not capable (because they haven't learned the linguistics) of judging the technical arguments. However, there is at least some disagreement among the metricists and linguists (although I would say that, on balance, they tend to support an early date -- the key words not to be skipped in that sentence are "on balance" and "tend.")
It is also important not to confuse things that are completely objective with those that are only partially objective or subjective. For instance, when linguistic tests rely on many emendations to the poem, we end up with layers of subjective (to one degree or another) judgment underlying the apparently objective test. In addition, matters of opinion and interpretations can be stealthily converted into seemingly objective interpretations. Take, for example, the matter of the poem's audience. Dorothy Whitelock, one of the greatest historians of Anglo-Saxon England, makes the argument that an English audience would not tolerate praise for Danes in any period after the sack of Lindisfarne. Her immense authority has led others to take the audience claim as historical (rather than literary) and thus at least somewhat more objective than judgments of style or degree of Christianity. But let us imagine a scholar from 1000 years in the future trying to figure out some enigmatic piece of American fiction from somewhere in the 20th century (say, 1983 or 1954 or 1908). That future scholar could argue "the poem speaks of Germans as friends, and praises them, but we know that America fought two horrible wars against Germany, so the poem must come from the 21st, not the 20th century." Obviously that scholar would be in error (replace "Germany" with "Russia" or put both in place to expand the thought experiment). And individual writers do not always reflect the consciousness of the nation, even in early times.
The manuscript date is another quasi-objective fact. The manuscript does indeed date from late in the 10th or very early 11th. But what does this fact mean? A late manuscript does not necessarily tell us about the text it contains (although this is less true in manuscript culture than it is in print culture, where accurate reproduction is a matter of course). And because the manuscript must be a copy of something, we are left with at least some doubt about how much we want to let the date of the copy influence our thinking about the underlying text. The literary-theoretical problem is very significant and has not, to my knowledge, ever been solved: what are the differences in the ways we interpret texts based where they were created or where they were received or where they were modified?
I was for a while a "late" dater, in large part, I'm afraid, because such a date was congenial to some of my other hypotheses about the poem. In the past few years, however, many arguments for the early date have come to seem much more reasonable to me. The linguistic and metrical tests do, I think, have a different evidentiary status than literary interpretation or speculation about audience, but so does the manuscript. It is easy to say (and I've said it) that "we know that Beowulf was received in the tenth century, and that's enough for me," but much harder to separate out what that means for interpretation: for instance, how much weight can we put, for example, on the depiction of Heremod in the poem (which one scholar whom I respect a lot argues is a key to understanding the characterization of Beowulf), if this is just something that a tenth-century poet inherited without knowing how the story fit in the whole web of Germanic myth and literature?
Let me give an example closer to home: in Nathaniel Hawthorne's "Rappaccini's Daughter," we read of Giovanni's feelings about Beatrice: "Least of all, while avoiding her sight, ought Giovanni to have remained so near this extraordinary being that the proximity and possibility even of intercourse should give a kind of substance and reality to the wild vagaries which his imagination ran riot continually in producing" The date of the production of the text is very significant for our interpretation of the meaning of "intercourse" in this sentence, and we might interpret that passage very differently if we thought that a 20th-century reviser/editor/copyist would have felt free to change Hawthorne's text for one purpose or another.
In 2008 I'll have been studying Beowulf for fifteen years, probably not nearly enough time to have a real opinion on all the evidence. But the more I read and study, the more I move from "definitely late" to "maybe early." And so (and this is where I get in trouble with 90% of my colleagues), I'm very tempted to say "both": an originally early poem, written down when the tradition was still alive, that is re-worked by another artist later on. Am not advocating Liedertheories, but rather a discussion of the modification of wholes: a whole poem being written in the 8th century (or earlier) and then a whole poem being copied and revised/refreshed at other times. The poem is a whole now, and it was a whole in the tenth century, and whatever the tenth-century poet inherited was a whole, each in its own way. I think there are both dissection planes and adhesions, and it may not be possible to separate them absolutely clearly, but the underlying structure is there, and that Kaluza's law, common word pairings, influences of Latin rhetoric, Germanic history, onomastics, etc., will help in figuring it out. But very likely my opinion will change as I read more scholarship, re-read the poem, teach the poem in both translation and OE (next semester) and listen to my colleagues.
Why do people care so much?
When was Beowulf written? is a straightforward question that has a correct answer. Even if the poem was reworked over multiple centuries, it has one true history. But that history is lost (for now; maybe sheep DNA will provide an answer). And there is so much evidence that it can be assembled and re-assembled in new forms, supporting different conclusions. I think the combination of a right answer (somewhere) with the conflicting and confusing evidence of the poem generates the strong emotions. A scholar starts to learn about the evidence and thinks "Hey, I can sort this all out." You come in with an open mind (you think), try to sort out conflicting claims, and all of a sudden you are a "late dater" or an "early dater" with a theory and an opinion. Then you get to enjoy fighting your corner. That is probably reason enough to explain the vast enterprise of Beowulf scholarship.
But Beowulf is also a great poem, an important literary monument and a part of cultural history. When we don't have some kind of historical context in which to put the poem, we lose out on many opportunities for understanding. Simply to thow up our hands and say "Too complicated!" or "TLDNR" (which is what I'm sure people are thinking about this post) is to take a short cut. An a-historical Beowulf is a deliberate choice to ignore important information (the problems is, we don't know which information is important) and I think an abdication of scholarship. And more importantly, if you do this, you miss out on a lot of scholarly fun.
And the body of technical Beowulf scholarship is a beautiful thing, a monument of learning. Reading through the papers collected by Tom Shippey and Andreas Haarder in Beowulf: The Critical Heritage is a genuine pleasure (and that only covers criticism before Tolkien). Tolkien and others have on occasion mocked this tradition, but it is, I think, a great human accomplishment, in some ways as great an accomplishment as the poem itself. To contribute even a little to that long tradition is a great privilege. Just to study it is a great joy.