The Road to Middle-Earth: How J. R. R. Tolkien Created a New Mythology

by Tom Shippey

  There are many logical difficulties with this idea – where, for instance, could one fit in the Roman occupation of Britain? – and Tolkien did not try to follow it through. Indeed he showed his dissatisfaction with it before very long by converting Eriol to Ælfwine (with evident connection to the themes discussed above), and by setting the whole tale in a distinctly later period, not the fifth century of Hengest and Horsa, but at least four hundred years later. What advantage did Tolkien think he might gain from this? Arguably, the move was one of slight desperation. As has been said, almost nothing is known of native English tradition, especially pre-Christian tradition. There has accordingly long been an impulse among comparative mythologists, like Dasent, to seize on the cognate Norse tradition, and either to say that the English are really half-Norse, or else that they were really rather like the Norse, so that you can argue back from the one to the other. Sometimes Tolkien took the latter route: Christopher Tolkien notes in BLT 1, p. 245, that his father took an Old Norse mythological name, Askr, and ‘anglicised’ it philologically to Æsc. And while Tolkien senior was not much taken by the former route, he retained at least an awareness of the Norse contribution to England. Ælfwine is thus, in Tolkien’s second self-authenticating story, actually a slave of the Forodwaith, the Men of the North who have invaded England as they did in fact from the ninth century on, ‘and his boyhood knew evil days’ (BLT 2, 314). But after he has escaped, been wrecked, and been rescued by the strange, ancient, stone-shoed Man of the Sea, a further storm casts up on his island a wrecked dragon-ship, with in it the corpse of Ælfwine’s former master. Ælfwine says ‘he slew my father; and long was I his thrall, and Orm men called him, and little did I love him’. The Man of the Sea, with his greater knowledge, does not contradict Ælfwine, but he puts a different view: ‘And his ship shall it be that bears you from this Harbourless Isle … and a gallant ship it was of a brave man, for few folk have now so great a heart for the adventures of the sea as have these Forodwaith’. Norsemen, it seems, are at least ambiguous: enemies, but worthy of respect. Tolkien never lost this ambiguity about the Old Norse heathen tradition, as one can see from his manoeuvrings between English and Norse ascriptions in ‘The Homecoming of Beorhtnoth’, discussed above. Yet the post-Viking setting of the Ælfwine story may have been felt by him as an excuse for bringing that tradition in; and indeed much of The Book of Lost Tales consists of evident borrowings, more or less ‘anglicised’, from Norse mythology. The chaining of Melko (BLT 2, pp. 100–104) recalls the chaining of Loki by the gods of Ásgarthr; in the same passage Tulkas is teased very much as Thórr is in the Norse poem Þrymskviða; the three weavers on p. 217, though labelled with words for time in Old English, strongly resemble the Norse Norns with their names Urthr, Verthandi and Skuld (or ‘Past, Present, Future’); the dragon’s heart, dwarf’s curse and dwarvish necklace of BLT 2 all have evident analogues from the Eddas. Yet once again what all this shows most clearly is how difficult it has become to create a ‘mythology for England’ out of pure English material! Tolkien tried a pre-English story with Eriol and a part-English story with Ælfwine, and saw a prospect of repair or liberation in both: yet neither was entirely adequate for the claim he would so much have liked to make, that the Engle, the English, after all ‘have the true tradition of the fairies, of whom the Iras and the Wéalas (the Irish and Welsh) tell garbled things’ (BLT 2, p. 290).

  Tolkien, it can be felt, was jealous of the much better-preserved Welsh and Irish folk-traditions, as of the Norse. He did his best with scraps of native lore that survived the post-Conquest ‘defoliation’. Who is the powerfully described ‘Man of the Sea’ in the passage discussed above? Clearly one answer is Ulmo, the sea-god of Tolkien’s mythology, as is hinted in BLT 2, 319–20. But another answer must be that he is ‘Wade’, the mythical sea-giant dimly mentioned in the furthest reaches of Old English tradition and still remembered by Chaucer, but otherwise entirely forgotten.6 Elsewhere Tolkien toyed with a brief scrap of Old English verse about ‘Ing’, quoted and translated by Christopher Tolkien in BLT 2, p. 305. His aim seems to have been to see Ing, like Eriol, as an eponymous founder of the English, who was ‘first seen by men among the East-Danes’ (i.e. near where the English originated), but then went away ‘eastwards over the waves’ (Tolkien would probably have preferred this to be ‘westwards’); but to make the semi-divine Ing, unlike Eriol, an elf and a lord of Valinor. Again, though Tolkien kept on flirting with elvish names like Ingwë, Ingil (lord of Tol Eressëa, BLT 1, p. 16), or Ingolondë (later to be Beleriand, Shaping, p. 174), he could not quite make a satisfactory connection. Yet it is clear enough what he was looking for, or groping for: a mighty patron for his country, a foundation-myth more far-reaching than Hengest and Horsa, one on to which he could graft his own stories.

  In this aim Tolkien was not successful, usually discarding his own explanations, whether of Eriol or Ælfwine or Ing, before they reached anything like a final shape. He was unsuccessful indeed in a further way, almost a comic way when one considers his own concern for ‘ethnic’ tone, when he eventually did submit a version of his ‘mythology for England’ for publication in 1937. We know now that Tolkien sent in to Allen and Unwin a bundle of material including his ‘Lay of Leithian’ and the ‘Quenta Silmarillion’, a close descendant of The Book of Lost Tales. But when the Allen and Unwin reader read them – or read the bits he was shown, see Christopher Tolkien’s account in Lays, pp. 364–7, and further Bibliography pp. 216–18 – he was totally perplexed, unsure whether what he was reading was ‘authentic’ or not (so far Tolkien would have felt he had succeeded), but regrettably quite clear that whatever its authenticity it certainly could not be English! His comment, ‘It [sc. the Silmarillion section he was given] has something of that mad, bright-eyed beauty that perplexes all Anglo-Saxons in the face of Celtic art’, has been much quoted. Yet its irony has not been fully perceived. Tolkien had done his best to root his Silmarillion story in what little genuine Anglo-Saxon tradition he could find. But the first time it found a reader, that reader was sure (a) that he was Anglo-Saxon, as indeed his name was (Edward Crankshaw), but (b) that the Silmarillion was not: one more sad testimony, Tolkien may have felt, to the complete deafness of modern English people, especially educated English people, to their own linguistic roots.

  What Tolkien was certainly doing through all his attempts to construct a historical frame for The Book of Lost Tales and The Silmarillion was, we would now say, trying to find a ‘space’ in which his imagination could feel free to work. In this he was in the end successful, and even his failures may have been necessary steps on the road. As for creating a ‘mythology for England’, one certain fact is that the Old English notions of elves, orcs, ents, ettens and woses have through Tolkien been re-released into the popular imagination, to join the much more familiar dwarves (stigmatised by Tolkien as a Grimms’ fairy-tale conception), trolls (a late Scandinavian import), and the wholly-invented hobbits.7 More than that could hardly be expected. And yet, one might say, it was a pity that Tolkien did not get on with telling more stories, that he was – in the material discussed both in this and the last section – so preoccupied not with what was told, but with how the telling came to be transmitted. Was he ever to gain any advantage from these professional tangles?

  Creating depth

  There is a one-word answer to that question, which is ‘depth’, the literary quality Tolkien valued most of all. But since ‘depth’ is not commonly recognised, or even noticed, in the sense that he intended, more explanation is required. Tolkien’s views on the subject have also become a good deal clearer as a result of the publications of the last ten years.

  In his essay on ‘Sir Gawain and the Green Knight’, for instance, delivered in 1953 but not published till thirty years later, Tolkien declared that the poem ‘belongs to that literary kind which has deep roots in the past, deeper even than its author was aware. It is made of tales often told before and elsewhere, and of elements that derive from remote times, be
yond the vision or awareness of the poet’ – like Beowulf, Tolkien goes on to say, or like King Lear and Hamlet. Tolkien then paused, digressing consciously from his major theme in the essay, to consider further the idea of ‘deep roots’ and how one can detect them in a work (like Sir Gawain) of whose immediate sources we know, in fact, next to nothing:

  It is an interesting question: what is this flavour, this atmosphere, this virtue that such rooted works have, and which compensates for the inevitable flaws and imperfect adjustments that must appear, when plots, motives, symbols, are rehandled and pressed into the service of the changed minds of a later time, used for the expression of ideas quite different from those which produced them. (Essays, p. 72)

  Regrettably perhaps, Tolkien then caught himself digressing, said that ‘though Sir Gawain would be a very suitable text on which to base a discussion of this question’, it was not what he meant to discuss that day – or, alas, any other day. Tolkien turned in other words from the question of ‘ox-bones’ to the flavour of the soup, and went on to consider problems only in the surviving text itself. Yet he had made the point (using in fact the word ‘flavour’) that deep roots for a text are not just something incidental, to be studied by scholars: they also affect the nature of the text itself, and can be detected by the sympathetic ear, possibly even the naive or unscholarly ear. How they do this, as he said, is an interesting question, though one virtually never studied.

  I considered the matter with reference to Tolkien’s own aims in writing The Silmarillion in a passage above, laying particular emphasis on a letter by Tolkien dated 20th September 1963, in which he discusses the ‘attraction’ of The Lord of the Rings, much of it created, one might say, by a skilful counterfeiting of the effect of ‘depth’. My arguments were replied to, in a thoughtful and courteous way, by Christopher Tolkien in his ‘Foreword’ to BLT 1, pp. 1–7. Some points of agreement can immediately be located, and other points now conceded. Thus both Christopher Tolkien and I agree on the critical role of the hobbits in ‘novelising’ Tolkien’s later narrative – the ‘collision’ Christopher Tolkien points to between Théoden King and Pippin and Merry being of very similar type to the one I discuss between Bard and Thorin on one side, and Bilbo on the other, in chapter 3 above. We both also agree entirely on the ill effects of too narrow a literary ‘habituation’; while our comments on Sam Gamgee as an instructor on how to read Tolkien (and a case of the naive hearer nevertheless responding immediately to the effect of ‘deep roots’) are virtually identical, see BLT 1, p. 3 and above. Meanwhile I concede freely, in note 3 to chapter 7, that I misunderstood Professor Tolkien’s letter of 1963, and I repeat the concession here. In the first two editions of this book I completed Tolkien’s sentence ‘I am doubtful myself about the undertaking’ with the clause ‘[to write The Silmarillion]’. As is now abundantly clear, it had already been written, and written several times over! I should have looked back at the antecedent sentences of the letter, and realised that what was meant was something more like: ‘I am doubtful myself about the undertaking [to make The Silmarillion consistent both internally and with the now-published Lord of the Rings, and above all to give it “some progressive shape”]’ – matters in a sense forced on Tolkien against his will. Yet with all that said, I still feel that Tolkien himself had recognised ‘the problem of depth’ and the difficulty of creating that quality (flavour, atmosphere, virtue) in The Silmarillion if published as a single book; while the solution Christopher Tolkien indicates, of providing the reader with a ‘point of vantage in the imagined time from which to look back’, while certainly right in theory, nevertheless does create striking problems of presentation and response. However we now have more than a single-book-Silmarillion. ‘The History of Middle-earth’ does make it possible to give a much more satisfying account of the nature and problems of ‘depth’.

  To see this clearly, one might begin by making a comparison with a work which Tolkien knew well, the Old Norse Völsunga saga, or Saga of the Volsungs, mentioned as a source below. This is certainly a work with deep roots; and as is not the case with Sir Gawain, some of those roots still survive and can be traced. The saga is in fact part of a complex or tradition of texts, which may be laid out as follows. (I base the diagram below on the work of Professor Theodore Andersson, in his The Legend of Brynhild (Ithaca: Cornell U.P., 1980).)8

  Among the things that this diagram means are the suggestions: (a) that the author of Völsunga had access to a text which has since been largely lost, ‘The Long Lay of Sigurthr’, though we know it existed because of a gap in one surviving manuscript, and because Snorri Sturluson seems to have read it before it was lost; (b) that both that ‘lost lay’ and three other works are nevertheless similar enough to suggest a further reconstructable or *-poem behind them; (c) that even a late and poor-grade poem like the Lied vom Hürnen Seyfrid may nevertheless have a kind of value as a witness to something greater than itself.

  But one further point one can make about Völsunga, independently of its own merits, is that framed in a context like this even its demerits can create a kind of eerie charm. There is something very strange about a central aspect of the Brynhildr-story in Völsunga. Most of the texts above agree that Brynhild the valkyrie was married to Gunnar king of the Burgundians as a result of deceit: she had sworn to marry only the man who could best her, and Gunnar could not manage it. Sigurd did it in his place and in his shape, handing Brynhild over to Gunnar only once she had been won. How did he tame her? Did he take her virginity – so, in some way or other breaking a taboo and depriving her of her magic strength, as Delilah did by cutting Samson’s hair? The German Nibelungenlied, its courtly author perhaps embarrassed by this aspect of the story, tells a confused tale. He declares that Sifrit (as he is called in that tale – one sees why Tolkien early thought it acceptable or even necessary to keep changing the forms of his character’s names) did not deflower Prünhilt, though admitting that he was present in the marriage chamber. Yet later on Sifrit’s wife, terribly jealous of Prünhilt, calls her a kebse, a paramour, and declares that it was her husband who took Prünhilt’s maidenhead. For proof she has a ring which Sifrit took from Prünhilt when he overpowered her, and gave later to his wife. The accusation leads to Sifrit’s death.

  In Völsunga the story is even more doubtful. There Sigurthr releases an enchanted Valkyrie from her sleep and her coat of mail at once (is she Brynhildr?); promises to marry her, but leaves her; two chapters later meets her again, again promises to marry her, and again leaves her; and is then given a potion of forgetfulness (which he hardly seems to need) and under its influence marries another woman. The story then goes on reasonably similarly to the Nibelungenlied: Sigurthr agrees to help Gunnarr win Brynhildr, rides through her magic wall of flame in Gunnarr’s shape, lies with her for three nights (though with a sword between them), takes a ring from her, and hands her over to Gunnarr. The ‘quarrel of the queens’ takes place, with the wife here telling Brynhildr that her husband was Brynhildr’s frumverr, ‘first man’. The ring is shown, Sigurthr is killed. Yet in this text there can hardly be any question of Brynhildr not knowing that it was really Sigurthr in disguise, and not her husband Gunnarr, who took her virginity – the obvious motive for her hatred and revenge – for Brynhildr herself declares that before she married Gunnarr she had already born a daughter to Sigurthr! It is impossible for this part of the Völsunga saga to make sense. If Brynhildr is to take offence, it can only be over the deceit, not over defloration. So how did she lose her strength? What is the significance of the ring? Why did Sigurthr put a sword between them, and if Brynhildr thought he was Gunnarr, why did she think he did it? One could imagine answers to these questions. But they lead you outside the saga, outside the text, into its complex frame of tradition.

  Now this, I would suggest, is ‘depth’ as Tolkien understood it: to repeat his words on Sir Gawain, the quality ‘which compensates for the inevitable flaws and imperfect adjustments that must appear, when plots, motives, symbols, are
rehandled and pressed into the service of the changed minds of a later time’ (my emphases added). It is a quality which may exist in one text, but is more likely to be produced by a complex of them. It is intensified by age, by loss, by reconstruction, by misunderstanding. A vital part of it is the sense that even the authors of texts like the Völsunga did not understand their own story, but were doing the best they could with it. And the charm of it, the sense of puzzlement, of a factual base, of a better and richer and truer story somewhere in the hinterland but never yet told, may in fact be created not by literary success but by literary failure. Even the ‘inevitable flaws and imperfect adjustments’ have an effect. For one thing, they may well urge later authors into retelling the story, to impose their own sense of how it should be told: sometime in the late 1920s or early 1930s Tolkien wrote a long poem called ‘The New Volsung-Lay’ (‘Volsun-gakviða En Nyja’), probably to fill in some of the gaps in the diagram above, or as he put it ‘to unify the lays about the Völsungs … to organise the Edda material dealing with Sigurd and Gunnar’ (see Letters, pp. 379, 452). It will be interesting, when it appears, to see how it deals with the problems indicated.

  What the widely variant texts of ‘The Legend of Brynhild’ do between them (and this includes even the latest, most faulty and inadequate works like Þiðreks saga or the Lied vom Hürnen Seyfrid) is create once more an imaginative space in which later authors can work, a space moreover enriched by discrepancies, arguments, the sense of different opinions and different cultures, all in a way trying to interpret the same events. Perhaps the most important result of the publication of ‘The History of Middle-earth’ is that it has created, especially as regards the Silmarillion, a corpus in many respects similar to ‘The Legend of Brynhild’ and the diagram above. How many ancient versions of the Brynhild-story are there? Eight, with another hypothesised. How many extant versions are there of ‘The Legend of Beren and Lúthien’? At least nine, as follows: