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

by Tom Shippey

  1) ‘The Tale of Tinúviel’ (BLT 2. pp. 4–41)

  2) ‘The Lay of Leithian’, incomplete (Lays, pp. 154–363)

  3) The Silmarillion, ch. 19

  4) ‘The Earliest Silmarillion’, ch. 10 (Shaping, pp. 24–5)

  5) ‘The Quenta Silmarillion’, ch.10 (Shaping, pp. 109–15)

  6) ‘The Earliest Annals of Beleriand’ (Shaping, pp. 300–301, 307)

  7) ‘The Later Annals of Beleriand’ (Lost Road, pp. 134–5)

  8) (a) Aragorn’s song in Fellowship of the Ring, Book I, ch.11, together with (b) its earliest version, published as ‘Light as Leaf on Lindentree’ in Leeds, 1925, and (c) its medial version, accompanied by a further paraphrase of the whole story, in Shadow, pp. 179–84.

  9) ‘The Grey Annals’ (Jewels, pp. 61–71)

  These versions vary very considerably in length (two pages to over two hundred), in completeness (the longest version is not the most complete), in intrinsic interest (the three ‘Annals’ versions are naturally annalistic), in literary merit and (not the same thing) in importance for understanding the development of the whole story. Yet the existence of all the versions together does more than merely provide one with more ‘ox-bones’ for study. It also radically alters the flavour of the soup, creating something of the ‘flavour of deep-rootedness’ which Tolkien so often detected and admired.

  As with Völsunga saga, inconsistencies are a vital part of the new reading experience. Consider for instance the major versions of the critical event when Thingol, Lúthien’s father, first meets Beren, hears Beren’s demand for his daughter, and imposes on him the task of bringing a Silmaril as bride-price.9 If one had only the Silmarillion version of this scene, its logic and development would seem perfectly clear. One irreducible fact about Beren is that he becomes Ermabwed, or Elmavoitë, but anyway ‘the One-Handed’: he loses his hand to the wolf. Since this is an irreducible fact, surely it must all along have been part of the story that Beren, in the scene with Thingol, should find himself swearing an unknowingly ironic oath: in the words of the Silmarillion version, ‘when we meet again my hand shall hold a Silmaril’ – because of course when he and Thingol meet again his hand will be holding a Silmaril, but both will be in the belly of the wolf. With that established it would seem to be only plain sense for Thingol to have provoked the oath by setting up a hand for hand, jewel for jewel exchange, as again he so clearly does in the Silmarillion: bring me a jewel (the Silmaril) in your hand, and I will put in your hand a compensating jewel (Lúthien’s hand). All this seems, I repeat, to be virually dictated by the essential core of the story: Beren’s one-handedness, Thingol’s imposition of a quest, the motif of the Rash Promise.

  Yet a glance at the BLT 2 version shows that in the beginning these connections were simply not there. Beren does say, in his second meeting with Thingol (there Tinwelint), ‘I have a Silmaril in my hand even now’, thus creating a kind of irony, but in the first meeting does not make the corresponding promise. His exact words are only ‘I … will fulfil thy small desire’: which, of course, at the time of their second meeting he has still not done. The tone of the first scene is also entirely different, almost that of a joke which goes too far, without the edge of murder which creeps in later and the edge of hidden greed to be added later still. Meanwhile in the medial version of ‘The Lay of Leithian’, the idea of ambiguous or ironic oaths has been brought very much into the foreground: Thingol swears to leave Beren free of ‘blade or chain’ but then tries unsuccessfully to make out that this oath need not apply to the mazes of Melian (Lays, pp. 188, 191). Yet even so there is still no sign of what one might have thought to be the critical phrase, ‘my hand shall hold a Silmaril’. Version 8(c) above – the paraphrase accompanying ‘Trotter’s’ poem in an early draft of Lord of the Rings – seems to have realised the potential of the ‘hand’ theme, for there Thingol at their second meeting reminds Beren that ‘he had vowed not to return save with a Silmaril in his hand’. But the first meeting in that version is hardly described at all, so that the ‘clear’, the ‘natural’ version of this scene is still so to speak in ‘imaginative space’, like the true account of Brynhild’s defloration.

  An obvious point, once again, is that authors tend not to begin with Grand Designs which they then slowly flesh out, but with scenes and visions, for which they may eventually find intellectual justification.10 In ‘The Legend of Beren and Lúthien’ (and by this I mean the collective body of texts, not any particular one) one notes also that there are scenes and images which persist regardless of their intellectual justification, or even in the absence of it. Tolkien never altered very much the dance before Melko/Morgoth, though early versions are more sexually suggestive than later ones; or the ‘Alcestis’ motif by which Lúthien rescues Beren from death and Mandos, regardless of whether he is Elf or Man. And he seems to have introduced the motif of the ring of Felagund before he knew precisely what to do with it. This does not exist at all in the BLT 2 version, where the whole Nargothrond thread has yet to appear, but is fully developed in the Silmarillion account. Insulted by Thingol, Beren holds the ring on high, and says: ‘By the ring of Felagund … my house has not earned such names from any Elf, be he king or no’. He seems here to be swearing a formal oath to the truth of his words, and swearing it on the ring – as Gollum wishes to do on the One Ring in LOTR Book IV, ch. 1, though Frodo will only allow him to swear by it, see the phrase quoted above. The corresponding scene in ‘The Lay of Leithian’ is very close to the Silmarillion one, even verbally, but it does not seem to contain the suggestion of an oath. All Beren means when he holds the ring aloft is that his inheritance and possession of it prove that he cannot be ‘baseborn’, or a spy or thrall of Morgoth. It is only a token or certificate.

  Christopher Tolkien remarks in another context (Shadow, p. 430) how in his father’s work material he had had on paper for years could nevertheless suddenly acquire ‘new resonance’ on being shifted to a new context, and this is evidently true of much of the Beren material. The ring of Felagund was there before its purpose was known – just as the ring of Sigurd keeps on appearing in all texts of the Brynhild-legend, even when its particular point seems to have vanished, or is actively denied. Similarly a web of oaths and word-twistings takes shape in the Beren/Thingol scene before the point of the central oath is realised. Even what one would have thought utterly essential bits of narrative remain unsure after multiple retellings. How, after all, did Beren lose his hand? Did he strike at the wolf with the Silmaril (an unsuitable weapon, but see BLT 2, p. 34, or Shaping, p. 113)? Did he try to daunt the wolf with the sight of it, as in Silmarillion, p. 181? Or is it just something the wolf happened to do, as in Shadow, p. 183? The fact that there is no answer is now part of the story. An effect Tolkien valued very highly is what one might call the ‘epicentric’ one: the sense that once upon a time there had been a shattering event, never fully understood, with which a whole sequence of story-tellers had tried to cope, their failures and their partial successes all alike recording the force of the central event, like the needles jumping on seismographs unguessably far from the centre of the earthquake itself.

  Tolkien wrote something to that effect in a passage of ‘The Notion Club Papers’, where Ramer says:

  ‘I don’t think you realize, I don’t think any of us realize, the force, the daimonic force that the great myths and legends have. From the profundity of the emotions and perceptions that begot them, and from the multiplication of them in many minds – and each mind, mark you, an engine of obscure but unmeasured energy.’ (SD, p. 228)

  What Tolkien could not provide, of course, was the ‘multiplication … in many minds’, an effect which genuinely has to be created by the passage of time and generations. He may well have realised, though, as time went by and as the variant versions of his stories accumulated, that he was, at first by chance and then perhaps by design, building up a corpus of texts like those he was professionally used to. The thought may well have struck him that variant
versions were nearly as good as ‘many minds’. Certainly he was attracted by the thought of deepening what he had written by presenting it from an unfamiliar or half-comprehending perspective. That is what he was doing in the very late work of ‘The Drowning of Anadûnê’ (SD, pp. 329–440), a version of the Fall of Númenor clearly conceived of as being written by a Man of late date, sceptical temperament and limited information. He created the same effect, very successfully but on a much smaller scale, by putting his old poem of ‘Looney’ into a new context with ‘new resonance’ in The Adventures of Tom Bombadil, see above. And there is a hint that Tolkien knew this was his chosen way of working in the essay on Sir Gawain quoted at the start of this section. Just after the passage quoted there, Tolkien says that his real subject in the lecture is not depth or rootedness, but ‘the movement of [the poet’s] mind, as he wrote and (I do not doubt) re-wrote the story, until it had the form that has come down to us’.

  ‘Mere “escapism” in literature’

  It would be possible, even tempting, to repeat the same exercise of comparative reading with several aspects of the Tolkien legendarium: to examine, for instance, the development of draconitas from Glorund through Glaurung to Smaug; to consider the developing but never determined theme of the ‘dragon-helm’ and its corruptions through the many versions of ‘The Tale of Túrin’ (more complex even than ‘The Legend of Beren and Lúthien’); to see how the Silmarils and the Oath of the Sons of Fëanor develop from their early relative insignificance. And there are other topics which lead outside the Silmarillion corpus: the presentation of the nature of evil in the orcs (the ironies of which have never been pointed out, though the orcs have, interestingly, a concept of virtue);11 Tolkien’s development of his own poetic technique to something approaching Old English rigour, especially in his alliterative verse; or (a major theme and a major reason for his success) his surely in the end deliberate creation of a continuum of heroic figures ranging from the fierce and quasi-pagan (Helm Hammerhand, Túrin, Dáin) to the near-saintly, the almost-Christian (Tuor, Faramir, Aragorn). All these exercises would have their point, and could make interesting single studies. Yet I do not think any of them would alter anyone’s overall view of Tolkien. And since this book hopes to do more than merely ‘preach to the converted’, it seems more important to return to the two questions asked at the end of the first section of this chapter, and see how adequately they have been answered.

  How did Tolkien’s creativity work? A good deal has been said above about self-reflection, ‘sleepwalking’, and creating ‘imaginative space’. Yet there is one further thought generated powerfully by reading Tolkien’s early drafts, though to elaborate it seems to concede advantage to some of his fiercest critics. This is – I put it candidly in the hope of an answering candour – that the drafts suggest his critics sometimes had the right idea; they detected in the finished work tendencies much more obvious in the medial stages, as also, on occasion and even more suggestively, motifs which remained forever buried to author and readers alike.

  Thus Edwin Muir (see above) said that the non-adult nature of The Lord of the Rings was proven by its lack of genuine casualties. Théoden, Denethor, Boromir – these are the kind of characters who can be picked out in every Western as to-be-dispensed-with before the end. I have replied to Muir above. Yet in all candour one has to say that the early ‘phases’ of Lord of the Rings show Tolkien struggling hard to prove Muir right. He really did not like scenes of pain. So, in The Treason of Isengard, we find Frodo laboriously explaining to Sam that though the orc hit him with a whip, he was still wearing his mithril-coat and didn’t feel it (p. 336, but cp. LOTR, p. 889). Much more seriously, the same volume shows a long and thorough attempt to pardon even Saruman and bring him back into the fold, thrusting all responsibility for the pollution of the Shire on to a mere walk-on ‘baddie’, and in the process eliminating (or rather aborting) the eerie death and ghostly rejection of Saruman in Book VI, chapter 8. Earlier on, in The War of the Ring, it is strange to see Tolkien toning down Denethor, trying very hard not to write the scene in which the father rejects the dutiful son in love and admiration for the absent prodigal (see War, pp. 327–34, and note Christopher Tolkien’s comments on p. 332). And these last two cases are not just the kind-heartedness over minor matters which I conceded to Muir in the passage just mentioned, one written before ‘The History of Middle-earth’ began to appear. If persisted in, they would have led to major differences in the plot; to a story of much narrower emotional range, with far less sense of irrevocable loss; to a situation much closer to what Muir detected. And yet, of course, Tolkien did not persist with them. He wrote them in, and then he wrote them out. It may well have gone against his own personal grain: I note elsewhere (above) that as soon as Tolkien did reach a hard solution he was liable to begin to soften it, and we can see now that reaching it was for him a laborious business in the first place. Still, grain or no grain, labour or no labour, he did it. Comparison of The Lord of the Rings with its drafts shows that Muir detected a tendency; his criticism of the entire and finished work remains false.

  In similar style, a more recent book, Christine Brooke-Rose’s A Rhetoric of the Unreal (Cambridge U.P, 1981), strangely wraps a true perception in error. It has to be said that most of the time Dr. Brooke-Rose merely continues the ‘criticism of denial’ already amply illustrated in the first few pages of this book; like so many professional critics, she resents her subject too much to read it fully. Thus on p. 247 she declares:

  Clearly LR is overcoded in this way [sc. as ‘semiological compensation’], since the megatext, being wholly invented and unfamiliar, has to be constantly explained. Apart from the ‘hypertrophic’ redundancy in the text itself, the recapitulations and repetitions, there are long appendices, not only on the history and genealogy, but on the languages of elves, dwarves, wizards and other powers, together with their philological development, appendices which, though ostensibly given to create belief in the ‘reality’ of these societies, in fact and even frankly, playfully reflect the author’s private professional interest in this particular slice of knowledge, rather than narrative necessity, since all the examples of runic and other messages inside the narrative are both given in the ‘original’ and ‘translated’. Nor are the histories and genealogies in the least necessary to the narrative, but they have given much infantile happiness to the Tolkien clubs and societies, whose members apparently write to each other in Elvish, (op. cit., p. 247)

  Much of this is so familiar as to be formulaic, the product of a small closed society of critics whose members too readily reach agreement with each other, not least by way of the ‘automatic snigger’ (to use Orwell’s phrase). Thus, if happiness is conceded, it has to be ‘infantile’; ‘professional interest’ in philology must ipso facto be playful (cp. above); the ‘megatext’, we are told, is ‘wholly invented and unfamiliar’ (a save-all footnote declares that even if there are sources in ‘Old Norse and other materials’, these have nothing to do with ‘ultimate “truth”’, a concept apparently securely in the critic’s possession, see again above). Much of the rest is just plain wrong, with the usual inference that the critic has been too angry or self-confident to read the book: wizards don’t have a language, as it happens; and though it must have seemed a pretty safe bet to complain about all the ‘runic and other messages’12 given in the narrative being translated – for ninety-nine authors out of a hundred would have felt obliged to do just that – Tolkien, as it happens, was the hundredth (see here and here, and notes 5 and 10 to chapter 6 below). As for ‘Nor are the histories and genealogies in the least necessary to the narrative’, that ignores the whole question of ‘depth’ – the one literary quality, to say no more, which most certainly distinguishes Tolkien from his many imitators.

  And yet, and in spite of all this, Dr. Brooke-Rose has a point. She feels that The Lord of the Rings, viewed as fantasy, is weighed down by ‘hypertrophic’ realism, by ‘naive and gratuitous intrusions from the realis
tic novel’. Must genres always practice apartheid? Evidently not. Still, reading the drafts of Lord of the Rings does make it clear what a temptation it was for Tolkien to fall back on the familiar clichés of the realistic novel. Rayner Unwin, the young son of Tolkien’s publisher, noticed this at a very early stage indeed; his father wrote to Tolkien that he had said chapters 2 and 3 had ‘a little too much conversation and “hobbit-talk” which tends to make it lag a little’ (War, p. 108). Tolkien replied, ‘I must curb this severely’, and he did: not totally, but a great deal more than in his first intention. In general, one may say that especially in the earliest ‘phases’, whenever the hobbits become the central figures of the narration – the hobbits being obviously the most modernistic and novelistic characters in the book – Tolkien found himself getting bogged down in sometimes strikingly unnecessary webs of minor causation. How many hobbits set out with Frodo (originally Bingo), and what were their names and families? Why did Farmer Maggot dislike Frodo/Bingo? How did ‘Trotter’ authenticate himself, was he an eavesdropper, and how many letters did Gandalf write? Most of these questions would now appear to be easily soluble, but they were not easily solved. What Tolkien’s sometimes maddening hesitations show is exactly how difficult he found that blend of ancient and modern, realistic and fantastic, which in the end he developed so successfully, and so much to his critics’ disapproval. I repeat that Dr. Brooke-Rose’s comments on Tolkien mostly strike me as prejudiced to the point of wilful blindness. Like Muir, she is a guide often only to what Tolkien was not. Yet like Muir, she does see, with a certain insight, what he was tempted to be. The final point to make, obviously, is that while Tolkien might not have eradicated every trace of soft-heartedness (Muir) or ‘realistic hypertrophy’ (Brooke-Rose), he did nevertheless in the end and painfully fight off most of both temptations. Indeed one could go further and say this: it seems an inherent temptation in romance to produce what is now called a ‘cop-out’ ending, an ending which defies the narrative logic of the story in the interests of popular sentiment or intellectual rationalisation. So Dickens gave Great Expectations a last-minute reconciliation; while the author of Sir Gawain tried to pass off all the events of his story as a totally unsuccessful practical joke on Guenevere. Tolkien too felt this temptation. He even wrote the ending: i.e., the ‘Epilogue’ (SD, pp. 114–35), with its strange similarity to the ending of George Lucas’s Star Wars (medals, triumph, the gratifying elevation of the humble). But having written it, he rejected it. The rejection makes one realise that creativity involved for Tolkien not only invention, not only philological brooding and the discovery of self-licensing fictions: it also demanded self-knowledge and self-restraint.