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

by Tom Shippey

  ‘Þis is þat lond þat he wole: zuyt are þe worldes ende

  his dernelinges an erþe zyue: & hyder heo schulle wende.’

  ‘This is the country that [Christ] will give, before the end of the world, to his secret favourites on earth, and this is where they will come.’12 That land would be both sanctified and earthly, an ideal ‘mediation’ for Tolkien. It must have increased his hope and longing to observe that the last line, about the ‘dernelinges’, is not in the text but (like ‘Frodos Dreme’) has been added in the margin by a later hand –as if some early but forgotten scribe had received a mysterious promise of his own. The promise lay in the philological detail; and the philology was true, even if the promise could not be expected to ‘come true’.

  * As often, I am amazed that I did not at first recognise this myself, for at the time of my first comments on Smith I was still holding Tolkien’s former position at the University of Leeds, and was in charge of the B-scheme, still in existence (though now no more). The B = birch equation, however, was no longer current.

  CHAPTER 9

  ‘THE COURSE OF ACTUAL COMPOSITION’

  The bones of the ox

  In the introduction to his 1851 translation of Asbjörnsen and Moe’s collection of Norse fairy-tales, Sir George Dasent wrote that the reader ‘must be satisfied with the soup that is set before him, and not desire to see the bones of the ox out of which it has been boiled’.1 Dasent’s introduction was in fact one of the 19th-century classics of popularising philology, a highly revealing response to the situation described above; it is full of laudatory references to Jacob and Wilhelm Grimm, of cross-connections between Norse and Scottish, or Norse and Sanskrit, and it makes a determined attempt to press on from comparative philology to comparative mythology. In this setting, what Dasent meant by his image was that he wanted his reader to accept his conclusions, and not demand to see the philological ‘workings’ on which they were based. Tolkien did not approve. Nevertheless, he was struck by the image, and repeated it in his essay ‘On Fairy-Stories’. Only what he meant by it, he said, was this:

  By ‘the soup’ I mean the story as it is served up by the author or teller, and by ‘the bones’ its sources or material – even when (by rare luck) those can be with certainty discovered. (‘OFS’, in Tree, p. 20)

  In other words, critics should study stories in their final forms, as ‘served up’ or published, not in their intermediate stages. At the time Tolkien wrote these words,2 much of ‘the Silmarillion’ at least had been written several times over, as The Lord of the Rings eventually would be. If Tolkien had had foresight into the future, one may wonder, would he have felt that his ban on wanting to see ‘the bones of the ox’ should have been extended from fairy-tale collections (which of course may well have had an especially complex history) to his own fictional works?

  There are reasons why he might. A major danger must be that too much study of ‘the bones’ makes ‘the soup’ lose its savour. In other words again, it could destroy the appeal, or charm, or ‘glamour’ of a finished work to know that some particularly cherished feature of it was in fact only an authorial accident; while too much awareness of wrong turnings the author might have taken could blur one’s final sense of the right turning he did take. The risks are manifold. In the case of The Lord of the Rings, one might fear that too much looking at intermediate stages (in this case volumes VI–IX of ‘The History of Middle-earth’) could blur the edges of one’s perception of the final stage, or of the work as published (for even after publication Tolkien continued to have afterthoughts, as one sees from both the Unfinished Tales and volume XII of ‘The History of Middle-earth’); while in the case of The Silmarillion –which in a real sense never reached a final stage at all – over-careful picking over of volumes I–V and X–XI of ‘The History’ could easily lead to the loss of any sense of structure whatsoever.

  If that were all that could happen, this chapter would have remained unwritten. Nevertheless one has to face the fact that much of ‘The History of Middle-earth’ demands to be taken as ‘ox-bones’ – though a proportion of it is unpublished original work and some at least of the ‘bones’, like the Book of Lost Tales, are easy to read in their own right – and furthermore that the kind of reaction I have just suggested is at least a possibility, or, some of the time, a certainty. Yet one may reflect that much of the trouble (and here one loses contact with Dasent’s image and with Tolkien’s application of it) lies with the reader, and not with the author at all.

  Other authors than Tolkien have for example created amazement in their readers by their seeming utter inability to understand the logic of what they were doing. Charles Dickens was upset and alarmed when it was called to his attention that many of his heroes or hero/villain pairs had names beginning with his own initials, C.D.; while his surviving worksheets for David Copperfield (this time D.C.) show that he got even the totally transparent name Murdstone – for the murderous stepfather who replaces the dead father under the gravestone – simply by writing a string of names across a page till he got one that felt right: Hasden Murdle Murden Murdstone.3 Dickens never asked himself, seemingly, why ‘Murdstone’ felt right. In exactly the same way Tolkien dealt with several important queries by writing out a string of names, like ‘Marhad Marhath Marhelm Marhun Marhyse Marulf’ (Treason, p. 390), or – these are for Aragorn – ‘Elfstone … Elfstan, Eledon, Aragorn, Eldakar, Eldamir, Qendemir’ (Treason, p. 276), or – these are for Shadowfax – ‘Narothal, Fairfax, Snowfax, Firefoot [,] Arod? Aragorn?’ (Shadow, p. 351). It is a surprise to learn that Aragorn could ever have been a name for a horse; even more surprising, given what is said about the meaning of the name, above, that Saruman could have been the meaningless ‘Saramond’ (Treason, p. 70). I do not think there is any doubt that Murdstone in David Copperfield does ‘mean’ what is said above, and there is even less doubt about Saruman, whose name is a philological crux. But neither Dickens nor Tolkien seems to have started off with meaning; rather with sound.

  All this comes as a shock. It may also prove an irritation. At one point (Lost Road, p. 217) Christopher Tolkien remarks of a passage in the carefully prepared 1937 ‘Quenta Silmarillion’, ‘Elwë here, confusingly, is nor Thingol’, with a paragraph of explanation to follow. ‘Confusingly’ is putting it rather mildly. It seems unlikely that anyone at all could ever keep in mind all the variations and permutations which Tolkien carried out on his elvish characters for The Silmarillion. Finrod becomes Finarfin; Inglor becomes Finrod; besides the Ellu/Elwë/Olwë alternatives, one finds Elwë ‘Thingol’ (more accurately Elu Thingol in Sindarin, Elwë Singollo in Quenya) at different times as Ellon, Tinthellon, Tinto’ellon, Tinwelint, Tintaglin. Some of these changes are there to show the processes of language-change which were a major part of Tolkien’s creativity from the beginning (see BLT 1, p. 48, a passage written c.1919). But there are also signs of a continuous and seemingly-random fiddling, which generates for instance diagram after diagram of the relations between the various tribes, groupings or languages of the elves, see BLT 1, p. 50, Shaping, p. 44, Lost Road pp. 181–3, etc. Just as with Finrod or Thingol, it is at best confusing, at worst irritating, to discover that the Teleri were at one time the senior, at another the junior branch of the ‘Light-elves’; and that the change really does not seem to make much difference! I have used the term ‘fiddling’. But Tolkien commented more accurately on this tendency in himself in an interesting passage in Part 2 of ‘The Notion Club Papers’, in Sauron Defeated, pp. 239–40. There the character Lowdham criticises the very activity of inventing languages, that ‘secret vice’ of which Tolkien accused himself in Essays, pp. 198–223. Lowdham says:

  ‘Anyone who has ever spent (or wasted) any time on composing a language will understand me. Others perhaps won’t. But in making up a language you are free: too free … When you’re just inventing, the pleasure or fun is in the moment of invention; but as you are the master your whim is law, and you may want to have the fun all over again, fre
sh. You’re liable to be for ever niggling, altering, refining, wavering, according to your linguistic mood and to your changes of taste.’

  Lowdham goes on to say that the languages he finds coming to him are not like that, and I have also omitted a section in which Lowdham says there are constraints on any conscientious inventor. Yet the word italicised above – the italics are mine – is a significant one. Tolkien used it elsewhere as the name of the character in his self-descriptive allegory ‘Leaf by Niggle’ (see above). He knew that one of his temptations was ‘to niggle’ i.e. (OED) ‘to spend work or time unnecessarily on petty details; to be over-elaborate in minor points’. He could not do this (so much) with real philology, because there the data were available to others. But where ‘his whim was law’, in inventing his own languages (geographies, genealogies), he was likely to give in to temptation. Of course we should never have known it if we did not have, in this case, ‘the bones of the ox’. But the revelation could create unease.

  There are other surprising criticisms of Tolkien latent in ‘The History of Middle-earth’. Sometimes, and in contrast to the ‘niggling’ just discussed, he was stubborn to the point of pig-headedness about sticking to names, apparently in total incomprehension of their likely effect on contemporary readers. He kept using the term ‘Gnomes’ for the Noldor till at least 1937, in confidence that ‘to some “Gnome” will still suggest knowledge’, through its connection with Greek gnome, ‘intelligence’ (see BLT 1, pp. 43–4). To some, possibly. However to all but a vanishingly small proportion of English speakers, ‘gnome’ has lost all connection with its Greek root, and means instead a small, vulgar, garden ornament, very hard to take seriously. Similarly, as remarked above, Tolkien stuck to the name ‘Trotter’ while the character who bore it changed from a wandering hobbit to a hobbit-Ranger to a human Ranger to the last descendant of the kings of old. Very late in the construction of The Lord of the Rings Aragorn, or ‘Strider’ as he eventually became, is still declaring (War, p. 390), ‘But Trotter shall be the name of my house, if ever that be established; yet perhaps in the same high tongue it shall not sound so ill …’ Wrong! For ‘trot’, as the OED rightly says, implies ‘short, quick motion in a limited area’, and is quite inconsonant with dignity when applied to a tall Man. Tolkien (we can see with hindsight) should have dropped the idea much earlier, along with much else: his preference for ‘hobbit-talk’ over action (see Shadow, p. 108), his strangely hostile picture of Farmer Maggot (Shadow, p. 291), his inhibiting confusions over the number and names of the hobbits with the Ring-bearer, over Gandalf’s letter via Butterbur, and the general ‘spider’s web’ of argumentation near the start of Lord of the Rings (see Treason, p. 52).

  Meanwhile and conversely, it is almost dismaying – at least to the critic – to see what seem to be absolutely essential elements both of The Silmarillion and The Lord of the Rings excluded sometimes till virtually the last moment. The Ring is ‘Not very dangerous, when used for good purpose’, says a naïve note in Shadow, p. 42, and cp. above; ‘it is indeed a remarkable feature of the original mythology’, says Christopher Tolkien, ‘that though the Silmarils were present they were of such relatively small importance’ (BLT 1, p. 156, and cp. above). A harmless Ring, meaningless Silmarils: as one reads through ‘The History of Middle-earth’ it is possible to feel – and this applies especially to a reader who knows the finished works well – that Tolkien did not know what he was doing. Tolkien himself once imagined summoning the scribes of the Ancrene Wisse from the dead, to indicate silently to them minor errors of grammar (see above). If we were to do this to the shade of Tolkien, it would be hard not to put one’s finger on Aragorn’s ‘Trotter’ sentence just quoted without, perhaps, a look of quizzical reproach.

  Yet having said all this – and it has been said with deliberately unmitigated bluntness – one has to consider in the end exactly what one’s criticism may be. It seems hardly fair to criticise an author for not writing the book one would have liked; even less fair to complain that he did write the book one would have liked, but failed to manage it on the first try! Perhaps the real danger in picking over ‘the bones of the ox’ is no more than this: it comes as a threat to our general notion of creativity. In our often dimly-perceived ‘model’ of the author at work, there is a tendency to think of him or her as following a Grand Design to which only the author is privy, and which is both central inspiration and guiding star. Critics often search for this – certainly that is what I was doing in 1970 when Tolkien wrote me the letter referred to in my ‘Preface’ to this volume, and from which the title for this chapter is taken. Discovering that the author does not have a guiding star, and is trying things out at random, can be a disillusionment; as can the realisation that the Grand Design (the Silmarils, the nature of the Ring) was in fact one of the last things to be noticed. Yet such disillusionment is in a sense only in the reader’s head, nothing to do with the work or indeed the author. And one thing that following the progress of a work through ‘the course of actual composition’ can do is provide one with a more truthful model of the way that authors work – Tolkien being in this case (one may well suspect) more representative of authors in general than one might suppose, except in two respects: the very long gestation period of all his works, and his deep reluctance ever to discard a draft.

  Nor need one abandon absolutely the notions of guiding star or Grand Design. For all the many surprises, false roads or spiders’ webs of ‘The History of Middle-earth’, it still demonstrates very conclusively that Tolkien did have an overpowering urge towards expressing something, something which kept on pulling him even if he had lost (or not yet gained) clear sight of it. Do we now have a better image of the something? And can we find a better ‘model’ of the way Tolkien’s creativity worked? These questions are considered in the rest of this chapter.

  Lost road, waste land

  Tolkien had a theory, at least, about the second of the two questions above, which he expressed in a repeated fiction, or fictional debate – one hesitates to call it a story. This exists in two main forms, ‘The Lost Road’, from c. 1937, printed in The Lost Road pp. 36–104; and, written some eight or nine years later, at a time when he might have been expected to be fully occupied with The Lord of the Rings, ‘The Notion Club Papers’, printed in Sauron Defeated, pp. 145–327.4 The fictions are close enough to each other almost to be described in Christopher Tolkien’s term (see Shadow, p. 3) as ‘phases’. They have at least strong common elements, if not a common root.

  The most obvious of these, not at all surprisingly, is continuous playing with names. In ‘The Lost Road’ the key names are two from an ancient Germanic legend, written down by Paul the Deacon in the eighth century, but dealing with events of the sixth. This germinal story tells of a king of the Lombards – for their importance to Tolkien see below – called Audoin. He refused as it were to ‘knight’ his son Alboin after a battle, because Alboin had not yet received arms from a neighbouring king, as was the custom of this people, evidently and rightly designed to avoid favouritism. Alboin accordingly went to the king he had just defeated, and whose son he had just killed, and asked him to grant him arms, with a kind of noble or quixotic confidence in King Thurisind’s magnanimity. His confidence was not misplaced; Alboin received his arms and his ‘knighthood’; though he repaid the favour only with a series of brutalities leading to his own later murder, for forcing his wife to drink a toast from her father’s skull, accompanied by what one can only call the orcish pleasantry of inviting her ‘to drink merrily with her father’.

  What caught Tolkien’s eye in this was evidently not the story but the names: Alb-oin = Old English Ælf-wine = ‘elf-friend’, Aud-oin = Old English Ead-wine = ‘friend of prosperity, bliss-friend’. ‘Elf-friend’: why should people be given names like that, consistently, over many centuries from the sixth to the eleventh, and from countries as far apart as Italy and England, if there had not been some original conception behind it? Audoin meanwhile had survived even into
modern times, via Ead-wine, as the modern English Edwin. Did this not suggest that there was still some form of living tradition in the names and their meanings? Another element was the Old English name Os-wine, or ‘god-friend’, also surviving, if not very often, in the name Oswin (cp. Oswald). From these survivals and indications of continuity Tolkien began to sketch out a story about progression: from a modern day three-generation family tree (which ran Oswin – Alboin – Audoin, all of its members philologically conscious of the forms and meanings of their own names), back to the Lombardic son and father, and then back further to the mythic Germanic past, to Irish legend, to the unrecorded men of the Ice Age, and through them to Númenor. But in Númenor the names would be of different form, though identical meaning: Elenchi = ‘elf-friend’ = Alboin, Herendil = ‘bliss-friend’ = Audoin, Valandil = ‘god-friend, friend of the Valar’ = Oswin. Moreover in Númenor the meanings of the names would be much more pointed, even incipiently antithetic: for in Númenor just before its fall to be a friend of the elves, or even worse a friend of the Valar, was to risk death by sacrifice to Morgoth. Elendil and his son Herendil are indeed in Tolkien’s story of ‘The Lost Road’ almost on the brink of separation, for the son, less wise than his father, seeks ‘bliss’ rather than truth, and bliss seems to his generation to be best achieved by obedience to their rulers and rebellion against the gods.

  Yet Tolkien never achieved a full story on this theme. ‘The Lost Road’ is, even in outline, only a sequence of oppositions; plus a thesis about how events of the past might come to be known, through dreams and through a sort of linguistic vision. A great part of ‘The Lost Road’ in fact consists of detailing how Alboin, the modern English son, later to become a professor, finds coming to him from outside – not via his own invention – snatches of languages, including the ‘Elf-Latin’ Quenya, as well as Old English and even Old Germanic (the *-ancestor of Old English, Gothic, and Lombardic as well). This trail soon petered out, with Tolkien sending his narrativeless ‘Lost Road’ fragment to Stanley Unwin in 1937, after the success of The Hobbit, having it rejected (no doubt with utter incomprehension), and dropping the idea for eight years, during which he was writing the bulk of The Lord of the Rings. Yet when he revived it, as ‘The Notion Club Papers’, with an apology to Sir Stanley for ever having troubled him with ‘The Lost Road’, what he did, with remarkable stubbornness, was to persist in not inventing a story, and instead to expand on what one might now call his obsessive playing with names and brooding on the question of transmission.