Silent Thunder

Did the original text of Beowulf contain the name of Thor?

I don’t usually touch the question of whether some parts of Beowulf were interpolated, or others excised from the text, between the writing of the original poem and the copying of its sole surviving manuscript. There’s no way to back up any such speculations with conclusive evidence, and the whole approach is associated with nineteenth-century efforts to cut all the Christian homilies and digressions out of the poem on the grounds that they were ‘monkish’ additions. This in turn had to do with Liedertheorie (‘lays theory’), the idea that long epics were stitched together from many shorter ballad-style lays, and to my knowledge no-one in Beowulvian scholarship takes that sort of thing seriously anymore.

Having said that, however, the older lays-theory scholars like Karl Müllenhoff were anything but stupid, and the idea of looking for interpolations in the more overtly religious parts of the poem is not fundamentally wrong. If they exist, this is precisely where we would expect to find them.

Although Beowulf may not be the pure pagan lay that modern antiquarians wanted it to be, nor is it everything that Christians of its own time¹ might have wanted it to be. Its unconverted fourth-century Danish king, Hrothgar, goes about hailing the One True God like an Old Testament patriarch; and its hero, who does the same, apparently makes it into heaven (or at least soðfæstra dom, the ‘doom of the true’ or ‘judgement of the just’) without having so much as heard the name of Christ. In all likelihood each homily and digression in the poem is the work of its original poet, who seems to have been devout enough in terms of belief; but it stands to reason that later copyists would have felt the need to omit a questionable line here, or add a more ‘orthodox’ one there, lest some overzealous abbot be moved to chuck the whole thing on the fire long before its infamous stay at Ashburnham House.

In my few years of self-studying this poem (no, I don’t have any academic qualifications in it), I have found only two places in which the aftertraces of such chopping and changing may or may not be present. They occur within a short space of each other in a controversial part of the second chapter, in which the monotheistic Danes begin to backslide into idol-worship under the persecutions of the cannibal troll Grendel. The specific lines in the original text are 175-183a; the accompanying translation (which becomes more literal in the relevant parts) is my own.

Hwilum hie geheton æt hærgtrafum

(Sometimes they made sacrifice at sheltered fanes)

wigweorþunga, wordum bædon

(of idol-worship, uttering prayers)

þæt him gastbona geoce gefremman

(that a slayer-of-spirits might succour do them)

wið þeodþreaum. Swylc wæs þeaw hyra,

(‘gainst the plights of their people. Their practice was thus,)

hæþenra hyht; helle gemundon

(their heathen hope; hell they cherished)

in modsefan; Metod hie ne cuþon,

(in their minds and hearts; the Measurer knew not,)

dæda demend, ne wiston hie drihten God;

(the Deemer of deeds; had no idea of Lord God,)

Ne hie huru heofena helm herian ne cuþon,

(Haven of the Heavens; knew not how to praise him,)

wuldres waldend.

(the Wielder-of-Glory.)

(From line 183b onward there follows a little homily about the dangers of thrusting one’s soul into the fire in order to achieve an earthly change of fortune; in my view this has nothing suspicious or otherwise untoward about it, hence it is omitted here.)

What first drew my attention to these lines is the fact that J.R.R. Tolkien – who famously defended the artistic unity of Beowulf in opposition to the lays theory – was prepared to raise doubts as to their authenticity. His reasoning, laid out in the scholarly notes to his posthumously-published translation, was that they constitute a “flat contradiction of one of the leading ideas of the poem”:

The ‘leading idea’ is that noble pagans of the past who had not heard the Gospel, knew of the existence of Almighty God, recognized him as ‘good’ and the giver of all good things; but were (by the Fall) still cut off from Him, so that in time of woe they became filled with despair and doubt – that was the hour when they were specially open to the snares of the Devil: they prayed to idols and false gods for help.

And the contradiction:

This resides in the words ‘nor had they heard of the Lord God’ [ne wiston hie drihten God; in my version, ‘they had no idea of Lord God’]. This is quite unlike the minor discrepancies to be observed in Beowulf (and in many other major works of literary art). What is the explanation? Can the theology of the Danes...have differed from that of their wise king? Is that what the poet means? … But a little consideration will show that this is no way out. The lines do not merely say that the witan [Danish counsellors] were obstinate heathens, who had recourse to idols in time of stress and temptation – that would be no discrepancy: they declare that they were wholly ignorant of God and his existence. But it would be quite impossible for any wita to associate with King Hrothgar for a day and remain in such a state.

As Tolkien notes, however, there is a distinction to be drawn between this phrase and the much less serious connotations of the one above it, Metod hie ne cuþon (‘they knew not the Measurer’, i.e. God):

Cunnan and witan are properly distinct like Latin cognosco and scio… cunnan is properly ‘know (all) about, understand (the nature of)’, in the sense that you know persons and places; witan ‘know facts’. So that strictly Metod híe ne cúþon (*180…) might mean no more than ‘they had little or no knowledge about Metod (the power that orders and governs the world, God or Providence)’. But ne wiston híe Drihten God (*181…) can only mean ‘they did not know of the existence of God at all, did not know of His being even.’

It would seem, then, that although Tolkien insisted on regarding the whole passage as an interpolation, the contradiction that led him to do so is wholly located in line 181b ne wiston hie drihten God – the meaning of which is reinforced by line 182, which adds that the Danes ‘knew not [how] to praise [or honour]’ God (ne hie huru heofena helm herian ne cuþon). This latter may or may not be taken as contradicting the leading idea of the poem, depending on whether herian ‘to praise’ is taken to mean formal worship, or the sort of praise that King Hrothgar and other Danes constantly offer up to God.

Might these two lines, then (181 and 182), be regarded as possible interpolations by some later copyist who had more narrow religious ideas than the original poet? Let us look at them again, more closely:

dæda demend / ne wiston hie drihten God

(the Deemer of deeds; had no idea of Lord God,)

Ne hie huru heofena helm / herian ne cuþon

(Haven of the Heavens; knew not how to praise him,)

No scholar to my knowledge has raised objections to these lines on metrical grounds, since a lot of leeway is permitted for the insertion of unstressed syllables at the beginning of a half-line. But from a purely aesthetic point of view, they have an obvious ‘gabbling’ quality created by these unstressed syllables, as if the ideas represented in them had to be squeezed into a tight space and not unfolded at the poet’s normal leisure. Another interesting thing about them is that they can be excised from the text without damaging the sense of the passage at all (and arguably, while improving it, by turning the focus from the supposed natural ignorance of the Danes to their conscious decision to ignore God).

Here’s how this part of the passage looks without them:

Swylc wæs þeaw hyra,

(…Their practice was thus,)

hæþenra hyht; helle gemundon

(their heathen hope; hell they cherished)

in modsefan; Metod hie ne cuþon,

(in their minds and hearts; the Measurer knew not,)

wuldres waldend.

(the Wielder-of-Glory.)

Emended thus, the passage would (we might guess) be accepted by Tolkien as wholly in keeping with the religious ideas of the poem. Metod hie ne cuþon (‘the Measurer they knew not’) in this context clearly refers to the dark and ignorant spiritual state of the Danish idolaters at the time of their wilful decision to ignore God; it is no more a general statement on the condition of all pre-conversion Germanic people than is helle gemundon in modsefan (‘hell they cherished in their minds and hearts’).

Now let us move to the main argument of this post. Common sense ought to tell us that any other interpolations from this same source are likely to be found in the same part of the poem, on other delicate religious points, just as the handprints of a thief are more likely to be found in the room from which he has stolen something and which contains the objects of his desire than in a separate wing of the house. Might there, then, be any more imperfectly-concealed ‘monkey business’ in this passage? Let us take a second look at the preceding lines:

Hwilum hie geheton æt hærgtrafum

(Sometimes they made sacrifice at sheltered fanes)

wigweorþunga, wordum bædon

(of idol-worship, uttering prayers)

þæt him gastbona geoce gefremman

(that a slayer-of-spirits might succour do them)

wið þeodþreaum. Swylc wæs þeaw hyra,

(‘gainst the plights of their people. Their practice was thus…)

This word gastbona, ‘ghost-bane’ or ‘slayer of spirits’, is a typically double-edged Beowulvian expression (like modþryð or byrgan þenceð). By turning to heathenism in the hope of slaying Grendel, an evil ‘spirit’ (gæst), the Danes are condemning their own spirits or souls (also gæstas) to eternal death. Here we must vouch another piece of common psychological sense: that when we find one such ambiguous word in a text like Beowulf, we are more likely to find others nigh to it than far away, just as we are more likely to find a second item of buried treasure in the same place that contained a first than by combing a much wider area outside it; and that the reason for this is that the poet must have slowed down his writing process in order to choose his words carefully, just as a man who buries treasure walks a long way before stopping and digging in one spot. This would be reason enough for us to examine this passage more carefully even if there were no question of secondary alterations.²

The referent of gastbona is clearly some unnamed pagan god, and the most obvious candidate would be Thunor – that is to say, Thor, whose name in Old English was equivalent to the word thunder. As the Norse tales tell us, it was Thor who went around slaying giants (jotunns or jǫtnar, known in English as ettens or eotenas, the species to which Grendel belongs) and other evil spirits on behalf of humanity, and one would imagine that the tendency of new Christians to crawl back to him in a crisis was notoriously hard to stamp out.

Thus it seems a little suspicious that once the gastbona is mentioned in line 177, the very next line that follows happens to chime on the letter-sound þ or th – as if the poet were consciously or unconsciously evoking the name of Thunor. And when we look at the first verse in that line – wið þeodþreaum, ‘against the plights of the people’ – our suspicions are whetted further. Just as the two lines we have just discussed looked a little ‘stuffed in’, this one looks a little ‘strung out’.

Although the compound þeodþrea ‘people-plight’ is apparently unique to Beowulf, a look at the corpus examples for the simplex þrea ‘threat, plight, affliction’ in the Bosworth-Toller dictionary shows that it is usually spelt þream in this grammatical form (plural in dative case, denoting indirect object), although it is also found as þreaum. The difference is that the more common spelling þream implies pronunciation as one syllable, and the less common þrea-um as two. Whatever the spelling, however, one imagines that it would have been most natural to run the word and its inflectional ending into a single syllable in speech.

The reason why this matters at all is that almost all half-lines in Beowulf and other ‘classical’ Old English poems obey the rule of Five Types, often named for their nineteenth-century discoverer Eduard Sievers. This is a taxonomy purporting to show (though not at all to explain) the patterns of stressed and unstressed syllables that were favoured and avoided by Anglo-Saxon poets. The basic forms of these types are best memorized in modern English by means of a mnemonic such as the following:³

TYPE A: Ambling adders (DUM-duh-DUM-duh)
TYPE B: With bumbling birds (duh-DUM-duh-DUM)
TYPE C: Became clustered (duh-DUM-DUM-duh)
TYPE D1: Dire death-struggle (DUM-DUM-Dah-duh)
TYPE D2: Droves death-destroyed (DUM-DUM-duh-Dah)
TYPE E: Earth-staining end (DUM-Dah-duh-DUM)

There are many more sub-rules for legitimate and proscribed type-variants, as well as plenty of academic debates over how best to categorize them, and it must be said that my own understanding of the rules is far from perfect. But what becomes clear even from a cursory look at the types is that each one has at least four metrical ‘beats’ representing stressed, semi-stressed and unstressed syllables (expressed here as DUMS, Dahs and duhs). The verse wið þeod-þreaum can be read as a legitimate Type C line (duh-DUM-DUM-duh), but the alternative *wið þeod-þream would end up too short (duh-DUM-DUM), with only three syllables.

This is by no means a watertight case for tampering, since there are other verses in Beowulf that come up short and have to be read with more syllables according to more archaic rules.⁴ But given the present context, one has to wonder if the original line was not something like *þunor wið þeodþream – ‘thunder ‘gainst tribe-plights’. According to the inventory of Sievers types laid out in Geoffrey Russom’s Syntax of Beowulf (pp.33-7), this would be a variant A-type called A6 (DUM-duh-duh-DUM-Dah), equivalent to other such examples in Beowulf like þolode ðryðswyð (‘suffered the stout-heart’, line 132a) or beorhtode bencsweg (‘rang out the bench-noise’, line 1161a).

Here’s that reconstructed line in context:

þæt him gastbona geoce gefremman

(that a slayer-of-spirits might succour do them:)

þunor wið þeodþream. Swylc wæs þeaw hyra,

(thunder ‘gainst tribe-plights. Thus was their practice…)

Had the Beowulf-poet written such a line, it would have been very much in keeping with the wry and punning style that produced gastbona in the line above it.⁵ Since there was no capitalization for names in Old English texts, the word þunor might be read as thunder or as Thunor – as an effect produced by the god in response to Danish prayers or as the name of the god himself. By no means would it constitute an uncharacteristic pagan invocation by our Christian poet; rather it would serve to subtly denigrate the old religion, and further emphasise the wretchedness of the idolators, since the idea of Grendel’s being scared off by a magic thunderclap is comically feeble. Most importantly, considering the gradually hardening English religious environment in which the poet and his work had to live, the line has plausible deniability: now you see the thundergod, now you don’t.

Thor veiled in female disguise in the Norse Lay of Thrym

But evidently some monkish copyist with metrical skills⁶ judged that this turn of phrase did not help the poem’s already tenuous claims to Christian orthodoxy, and deigned to efface it with a stroke of his pen. He is to be thanked for it, I suppose, as long as the result was that his good abbot had to warm his feet on some other great treasure of Germanic cultural heritage.

Or perhaps not – since all of this is really just cunning guesswork, and there is not the slightest bit of hard evidence that the Beowulf text was monkeyed with at all. As always when reading my unqualified speculations, take with a large pinch of salt.⁷ All I know is that from now on, as I continue to probe through the wilds of this and other Anglo-Saxon poems, I shall keep my mind’s ear open for the distant roll of thunder – and yes, perhaps even for the muted crack of wood…

1. Here I am thinking particularly of Alcuin of York, who wrote to an English bishop in 797:

Let the words of God be read at the priestly banquet: there it is fitting that a reader be heard, not a harpist; the discourses of the Fathers, not the songs of the heathens. What has Ingeld to do with Christ? The house is narrow; it cannot hold both. The King of Heaven does not wish to have communion with pagans and lost kings who are kings in name only; for the eternal King reigns in heaven, while the lost pagan wails in hell.

Alcuin was a fixture at the court of Charlemagne, who took pains to memorize traditional Frankish poems and have them written down, but whose son Louis the Pious would ban them (possibly under the early influence of Alcuin) – consigning not only Frankish poetry but also the Frankish language to the midden of history.

2. This logic of course does not apply to modern literary poems that constitute whole underground networks of hidden meanings, nor to recomposed oral poems in which there is no way to ‘slow down and bury treasure’, except by inserting prememorized passages. But it applies to a dual-faced transitional text like Beowulf, in which long oral-style passages with words apparently chosen for meter and narrative are followed by compressed and allusive digressions better resembling Anglo-Saxon riddles. I should add here that the whole notion of puns and deliberate ambiguity in Beowulf is controversial, and apparently not favoured by most scholars, though it should be noted that the application of oral-formulaic theory to Old English that would probably rule it out has been decisively refuted.

3. This is my take on a popular mnemonic that can be found at various different places on the internet, and seems to have gone through a Chinese-whispers-style evolutionary process.

4. For example, the word geþeon ‘to prosper’ in line 25 has to be stretched out to geþeohan (the form probably known to the poet) in order to produce the requisite number of beats, and there are other examples like this dotted throughout the poem. These, however, concern the forms of words and not of inflectional endings. With regards to these, a quick search in the Electronic Beowulf shows me that the word fea ‘few’ in line 1081 is spelt feaum in dative plural, but the half-line nemne feaum anum ‘except a few only’ would still have enough beats for a Type C line were this word to be treated as a single syllable.

5. More dubitably, geoce gefremman ‘to do aid’ could also be interpreted as a double meaning, provided we accept the proposition that it might could be construed alternatively as ‘to urge [or lead] to the yoke’ (short-vowelled geoc ‘yoke’ in dative case with verb (ge)fremman, with meaning along the lines of an example in Bosworth-Toller, sume ic…to geflite fremede ‘some I have urged to strife’ – note that this also comes from a poem, Juliana, and occurs in a speech made by a devil). Whether or not this is a stretch with regards to meaning, I do not think it implausible that a sight-pun might be made on two words that were distinguished by vowel length in speech but looked identical in writing. This possible pun might be expressed in modern English by translating the line as ‘that a slayer of spirits might be salver to them’ (at a flying glance, ‘might be a slaver to them’).

6. By this I mean that he was obviously not the rather ignorant scribe of our extant text, but a more skilled one who copied the work at some intermediate point in its transmission.

7. Criticism is always welcome as well. One difficulty that presents itself is Russom’s assertion that verses like the one we have proposed here ought not to chime on all three stresses (so leof landfruma is permissible, leof landleod would not be), though I have never seen this cited as a hard-and-fast rule of the meter. Of course the verse may have been edited more heavily than we have proposed, but in this case our suspicions rest wholly on the choice of þ as alliterant in the very line that follows the mention of a gastbona.