5 November 2013
The opening line of Beowulf has always posed a bit of problem for translators:
Hwæt! We Gar-Dena in geardagum,
þeodcyninga þrym gefrunon,
hu ða æþelingas ellen fremedon.
The problem is exactly what the word hwæt is doing. Hwæt is the etymological ancestor of the modern what, but the Old English word’s semantic and grammatical functions are not the same as the modern word’s. Most translators have treated it as an exclamation, along the lines “listen!” or “lo!” rendering the line as something along the lines of:
Listen! We have heard of the glory of the folk-kings of the Spear-Danes in days past, how the noble ones performed acts of courage.
In his 2000 translation, Roy Liuzza translates hwæt as “listen!”, treating it hypermetrically and giving the word its own line. Robert Fulk translates it as “yes” in his 2010 translation. In the standard edition, Klaeber’s fourth, Fulk and fellow editors Robert Bjork and John Niles gloss it as “what, lo, behold, well.” And Seamus Heaney famously translates it as “so.” The 1898 Bosworth-Toller dictionary says of the word, “used as an adv. or interj. Why, what! ah!”
But a recent article by George Walkden of the University of Manchester challenges this traditional analysis. Walkden makes a compelling, but by no means ironclad, case that hwæt is not an interjection or an adverb, rather it has no independent meaning from the clause it appears in. It combines with the remainder of the clause to produce an exclamatory effect. In this way it is similar to the modern English how, as in “how you’ve changed!” According to Walkden’s conclusion, the opening line of Beowulf would read:
How we have heard of the glory…
There are four main problems with treating hwæt as an interjection or adverb. The first is that it always appears in an unstressed position, when one would expect interjections to be stressed. Walkden also points out that hwæt, when used in this fashion, is never separated by punctuation from the rest of the clause (which is not particularly damning in and of itself because punctuation in Old English manuscripts is haphazard at best). Ælfric in his grammar, the only extant contemporary Old English grammar, does not analyze hwæt as an interjection (but this is negative evidence; we can only guess as to why Ælfric makes this omission). Finally, hwæt is not only used to initiate speech, and it is found in texts and passages that are not oral in nature. Walkden’s hypothesis does not suffer from these problems.
The core of Walkden’s argument is a syntactic analysis of the use of hwæt and its Old Saxon cognate huat. He finds that in clauses beginning with hwæt/huat, the verb tends to appear in a later position than would normally be the case for a root clause. Instead, hwæt-clauses follow the pattern expected of a dependent clause. If hwæt were an interjection independent of the clause, it should not influence word order.
Walkden’s conclusion holds for both prose works, where syntax tends to be more regular, and poetry. It also holds for original works, like Ælfric’s Lives of Saints and for translated works, like the Old English version of Bede’s Ecclesiastical History, where one would expect the original Latin syntax might have an effect on the syntax of the translation.
The chief question I have regarding Walkden’s work is a methodological one, whether he formulated his hypothesis and then tested it against the corpus, or whether he analyzed the corpus and then formulated the hypothesis to explain a pattern he found. If the first case is true, then his analysis is quite strong evidence. If the second is the case, then not so much. Random patterns will always appear in any type of “data mining” exercise, and their existence tells us little of value.
That question aside, Walkden’s analysis will have to be taken into account in any future translations of Beowulf.
The Independent provides a nice lay encapsulation of Walkden’s argument and its implications.
