The Road to Middle-earth: How J. R. R. Tolkien created a new mythology

Something like these two awarenesses, of continuing history and continuing linguistic change, can be inferred (admittedly with the aid of vast quantities of hindsight) from the first thing Tolkien ever published, bar a few lines in school and college magazines: the poem ‘Goblin Feet’ in Oxford Poetry 1915.3 This begins:

I am off down the road

Where the fairy lanterns glowed

And the little pretty flittermice are flying:

A slender band of grey

It runs creepily away

And the hedges and the grasses are a-sighing.

The air is full of wings,

And of blundering beetle-things

That warn you with their whirring and their humming.

O! I hear the tiny horns

Of enchanted leprechauns

And the padding feet of many gnomes a-coming.

This is, admittedly, not very good. Indeed one can imagine the response to it of the literary ‘side’, full of armèd vision, not to mention critical temper. ‘Why’, it might ask, ‘do we have the past tense in line 2 and the present everywhere else? Does this mean the “fairy lanterns” have gone out and the “I-narrator” is pursuing them? Or could it be that the author is stuck for a rhyme to “road”? As for “a-sighing” and “a-coming”, these look like scansion devices, mere padding. But in any case there is nothing in nature to suggest that the hedges and the grasses were “sighing” at all, while the “creepiness” of the road is just something the poet has decided to project on to the landscape from himself. That’s why we don’t believe the “I-narrator” when he says he hears “tiny horns”! And what about “enchanted leprechauns”? Does that mean they’ve been enchanted by someone else; or that they’re enchanting; or are all leprechauns enchanted, i.e. magic, i.e. not-real? The poet gives himself away. This is an evasive poem, a self-indulgent one. “Off down the road” indeed! Road to nowhere!’

So the critical indictment might run, and it is hard to counter. Readers of The Lord of the Rings will have noted further the as yet undiscriminating use of ‘fairy’, ‘gnome’ and later ‘goblin’, not to mention the quite cross-cultural use of ‘leprechauns’ and the insistence (later to be most strongly abjured) on the little, the tiny, the insect-like. Still, there are hints of hope in the poem after all, and better questions to be asked than those which have been.

What is this ‘road’, for instance, the ‘slender band of grey’, the ‘crooked fairy lane’? It clearly is not a tarmac one; on another level it is to be a recurring Tolkienian image:

The Road goes ever on and on

Down from the door where it began …

And oddly, G. B. Smith – Tolkien’s school and college friend, killed the following year in Flanders, to have his poems posthumously published with a foreword by Tolkien – had addressed himself to the same theme in a poem four pages earlier in the Oxford Poetry collection:

This is the road the Romans made,

This track half lost in the green hills,

Or fading in a forest-glade

’Mid violets and daffodils.

The years have fallen like dead leaves,

Unwept, uncounted, and unstayed

(Such as the autumn tempest thieves),

Since first this road the Romans made.4

Now this theme of time is intensely Tolkienian (if one may be permitted to put it that way round). The last sight of Lórien in The Fellowship of the Ring published thirty-nine years later, is of Galadriel singing Ai! laurië lantar lassi súrinen! ‘Ah! like gold fall the leaves in the wind, long years numberless as the wings of trees …’ (LOTR p. 368, and see also Fangorn’s song, p. 458). In this case the hope which G. B. Smith expressed in his final letter to Tolkien before death – ‘May God bless you, my dear John Ronald, and may you say the things I have tried to say long after I am not there to say them’ – appears against all probability to have been fulfilled (Biography, p. 121). However the clue to follow, for the moment, is ‘the road the Romans made’.

It may seem perverse to seek to identify this road, but on the other hand it isn’t very hard. There are only two Roman roads near Oxford, and the better-preserved is the old highway from Bath to Towcester, still visible as a straight line across the map but dwindled along much of its Oxford stretch to a footpath. It is now called ‘Akeman Street’, like ‘Wayland’s Smithy’ a name of some fascination for philologists. It implies for one thing an old and massive population change. No town in Roman Britain had a more simply descriptive name than Aquae Sulis, ‘the waters of Sul’, and so prominent were its mineral springs with the Roman spa around them that even the Anglo-Saxons began to call it æt baðum, ‘at the baths’, and later Bath. One of them wrote a poem about the site, now called The Ruin. However they also called the town Acemannesceaster, ‘Akeman’s chester’ or ‘Akeman’s (fortified) town’. That is why the Bath–Towcester road acquired the name ‘Akeman Street’; the people who called it that knew it went to Bath, but had forgotten that Bath was ever Aquae Sulis; they were invaders, of a lower cultural level than the Romans, and soon they ceased to use the road for anything like the traffic it had once carried. Its name and its decline in status from highway to footpath bear witness to the oblivion that can fall on a civilisation. But what was the reaction of these invaders to the historical monuments they could hardly help seeing in their new land, the stone roads, the villas, the great ruins which they (as in The Ruin) called vaguely the eald enta geweorc, ‘the old work of giants’? Place-names again give suggestive clues.

About nine miles north-west of Oxford and half a mile from Akeman Street across the river Evenlode stands a villa, excavated in 1865 and once the property of some Romano-British noble. It is distinguished by the remains of a fourth-century tessellated pavement in different colours. The village nearby is called Fawler. To most people, including its inhabitants, this name now means nothing. But once it was Fauflor, a spelling recorded in 1205, and before that, in Old English, fág flór ‘the coloured floor, the painted floor’. There can be little doubt that the village was called after the pavement; so the pavement was still visible when the invaders came. Why, then, did they not occupy the villa, but chose to live instead on an undeveloped site a few furlongs off? No one can tell, but perhaps they were afraid. A further twist in the story is that there is another fág flór in Anglo-Saxon record, in the great hall of Beowulf, haunted by Grendel the maneater:

on fágne flór féond treddode,

éode yrremód; him of éagum stód

ligge gelicost léoht unfæger.

‘The fiend stepped on to the painted floor, angrily he paced; from his eyes there stood an ugly light, like fire.’

So wrote the poet, in one of his classic passages of ‘Gothic’ suggestion. Could Beowulf have been sung in Fawler? What would its inhabitants have thought? Tolkien knew Beowulf, of course, virtually by heart, and he knew what ‘Fawler’ meant, for he hailed the etymology with delight in his 1926 review of the Introduction to the Survey of Place-Names; such work, he pointed out, is fired by ‘love of the land of England’, by ‘the allurement of the riddle of the past’, it leads to ‘the recapturing of fitful and tantalising glimpses in the dark’ (YWES 5, p. 64). He was interested in the names of roads, too, for he had argued the year before that ‘Watling Street’ was an old name for the Milky Way, ‘an old mythological term that was first applied to the eald enta geweorc [i.e. the Roman road from Dover to Chester] after the English invasion’ (YWES 4, p. 21). Nor did he forget Bath and The Ruin. Legolas’s ‘lament of the stones’ on page 276 of The Fellowship of the Ring is an adaptation of part of the poem. At some stage of his life Tolkien must certainly have noted all the strange implications and suggestions of ‘Akeman Street’.

Did he know them in 1915, and share them with his friend G. B. Smith? Is the quest for Fairyland in ‘Goblin Feet’ a kind of translation of the quest for the romantic realities of history? Probably the answer to both questions is ‘No’. However, disentangling fact from inference as carefully as possible, one can say first that Tolkien and Smith evidently shared a feeling for the ancient roads, the ‘old straight tracks’ and ‘crooked lanes’ of England; second that Smith even in 1915 appreciated the sadness of the relationship between what these are and what they were; third that before many years were out it would be certain that Tolkien appreciated the same thing much more fully, with a wealth of reference to history and poetry and present-day reality. Even in 1915, one might say, a road, a real road, could possess a ‘creepiness’ for him which was based on some factual knowledge, not entirely self-generated. Philology would reinforce this. But already one image in his poem drew on some historical force.

Further, Tolkien was already thinking of words as ‘stalactites’. ‘Flittermice’ in line 3 is not normal English. According to the OED it was introduced in the sixteenth century by analogy with German Fledermaus, for ‘bat’. However ‘bat’ is not recorded in Old English, and it is possible that some ancestor of ‘flittermouse’, e.g. *fleðer-mús, was natural to English all along, but never got written down. There is an apparently similar puzzle over ‘rabbit’ (for which see below), which Tolkien at least signals awareness of in the second stanza by using the odd term ‘coney-rabbits’. Finally ‘honey-flies’ in line 30 is elsewhere unrecorded. From context one would think he meant ‘butterflies’. Perhaps he was aware, though, of the unexpected scatological sense of that innocent-looking word in Old English – a language which has had many rudenesses pruned by educated usage. He could have found out by looking ‘butterfly’ up in the OED, and at least it had occurred to him to wonder why butterflies were always and for no apparent reason so called. These verbal creations admittedly do not add much to the overall effect of ‘Goblin Feet’, but they exemplify an attempt to combine philological insight with poetry. Both roads and words hint at the early complexity of Tolkien’s inner life, its unusual combination of emotion with inquiry.

Survivals in the West

Such hints, of course, fizzle out immediately. The Silmarillion had begun its sixty-year gestation by 1914,5 but in 1915 Tolkien went off to the war in which G. B. Smith was to die. On demobilisation he was preoccupied with the problem of earning a living, first in Oxford with the OED, then in the English Department at Leeds University, finally, with secure status and no lure of further advancement, back in Oxford again in 1925. He published nothing (bar the note to Smith’s posthumous collection of poems) for five years after ‘Goblin Feet’, and a good deal of his subsequent work was written for simple motives – money, or to keep his name in front of the people who counted, who made appointments ‘with tenure’. Much of his inner life did find its way into the twenty or thirty poems contributed to various periodicals or collections between 1920 and 1937; Tolkien’s habit of thriftily rewriting them and using them in The Hobbit or The Lord of the Rings or The Adventures of Tom Bombadil shows how important some of them were to him.6 Still, it is fair to say that these remain by themselves thin, or uncertain. The brew that was to become his fiction needed a good deal of thickening yet; and this could only come from the interaction of poetry with philology.

From this point of view one of Tolkien’s most revealing pre-Hobbit pieces is his almost unread comment on ‘The Name “Nodens”’ for the Society of Antiquaries in 1932.7 This virtually repeats the story of ‘Fawler’. In 1928 excavations on a site near Lydney in the west of England had revealed a temple devoted to some kind of mystery cult and still flourishing in the fourth century, i.e. well after the introduction of Christianity to England. The temple was eventually abandoned as a result of the barbarian and also non-Christian English, who however had their own cults. As with the villa at Fawler the Lydney temple fell into disuse – but not completely into oblivion. The iron-mines not far away were remembered: and whether because of them or from a continuing superstitious respect for the site, it was given a new Anglo-Saxon name, persisting to modern times — Dwarf’s Hill. The Society of Antiquaries made no comment on all this, but in the story and the place-name one can hear the echo of a hopeless resistance from the Darkest of Dark Ages, pagan to Christian, pagan to pagan, Welsh to English, all ending in forgetfulness with even the memory of the resisters blurred, till recovered by archaeology – and by philology. For Tolkien’s job was to comment on the name ‘Nodens’ found in an inscription on the site, and he did it with immense thoroughness.

His conclusion was that the name meant ‘snarer’ or ‘hunter’, from an Indo-European root surviving in English only in the archaic phrase ‘good neat’s leather’. More interesting was his tracing of the descent of Nodens from god to Irish hero (Núadu Argat-lam, ‘Silverhand’), then to Welsh hero (Lludd Llaw Ereint, also ‘Silverhand’), finally to Shakespearean hero – King Lear. Even Cordelia, Tolkien noted, was derived from the semi divine Creiddy-lad, of whom was told a version of the story of ‘the Everlasting Battle’, which interested Tolkien in other ways. Shakespeare can naturally have known nothing about ‘Nodens’, or about Beowulf (a poem in which some have seen the first dim stirrings of ‘Hamlet the Dane’). That did not mean that the old stories were not in some way working through him, present even in his much-altered version. Like ‘Akeman Street’ and ‘Wayland’s Smithy’, Tolkien might have concluded, even King Lear could bear witness to a sort of English, or British, continuity.

And one could say the same of Old King Cole. Tolkien never actually rewrote his saga in epic verse (though one can now see why he remarked casually of Milton, ‘Monsters’, p. 254, that he ‘might have done worse’ than recount ‘the story of Jack and the Beanstalk in noble verse’ – it would have been a monster-poem, like the lost ‘Rescue of Theodoric’). Still, he would certainly have recognised the ‘merry old soul’ as a figure similar in ultimate origin and final ‘vulgarisation’ to King Arthur or King Lear.8 This interest in the descent of fables probably explains why Tolkien did try his hand at two ‘Man in the Moon’ poems, ‘Why The Man in the Moon Came Down Too Soon’ (which appeared first in 1923 and was collected in The Adventures of Tom Bombadil thirty-nine years after), and ‘The Cat and the Fiddle: A Nursery Rhyme Undone and its Scandalous Secret Unlocked’ (also out first in 1923 but to achieve far wider circulation as sung by Frodo in Book I, chapter 9 of The Fellowship of the Ring, ‘At the Sign of the Prancing Pony’). No one would call either of these serious poems. But what they do is to provide a narrative and semi-rational frame for the string of totally irrational non-sequiturs which we now call ‘nursery rhymes’. How could ‘the cow jump over the moon’? Well, it might if the Moon were a kind of vehicle parked on the village green while its driver had a drink. How could the Man in the Moon have ‘come by the south And burnt his mouth With eating cold plum porridge’? Well, it doesn’t seem very likely, but perhaps it points to an ancient story of earthly disillusionment. If one assumes a long tradition of ‘idle children’ repeating ‘thoughtless tales’ in increasing confusion, one might think that poems like Tolkien’s were the remote ancestors of the modern rhymes. They are ‘asterisk-poems’, reconstructed like the attributes of Nodens. They also contain, at least in their early versions, hints of mythological significance – the Man in the Moon who fails to drive his chariot while mortals panic and his white horses champ their silver bits and the Sun comes up to overtake him is not totally unlike the Greek myth of Phaethon, who drove the horses of the Sun too close to Earth and scorched it. Finally, the reason why Tolkien picked ‘the Man in the Moon’ for treatment rather than ‘Old King Cole’ or ‘Little Bo-Peep’ is, no doubt, that he knew of the existence of a similar ‘Man in the Moon’ poem, in Middle English and from a time and place in which he took particular interest.

This is the lyric from Harley Manuscript no. 2253, now known generally as ‘The Man in the Moon’.9 It is perhaps the best medieval English lyric surviving, and certainly one of the hardest, prompting many learned articles and interpretations. However three points about it are clear, and all gave it especial charm for Tolkien. In the first place it is extremely bizarre; it is presented as a speech by an English villager about the Man in the Moon, asking why he doesn’t come down or move. It also has a very sharp and professional eye for English landscape; the villager concludes that the Man in the Moon is so stiff because he has been caught stealing thorns and carrying them home to mend his hedges with (an old image of the Moon’s markings is of a man with a lamp, a dog, and a thorn bush, see Starveling in A Midsummer Night’s Dream V i). Finally, for all the poem’s thick dialect and involvement with peasant life, it is full of self-confidence. ‘Never mind if the hayward has caught you pinching thorns’, calls the narrator to the Man in the Moon, ‘we’ll deal with that. We’ll ask him home’:

‘Drynke to hym deorly of fol god bous,

Ant oure dame douse shal sitten hym by.

When that he is dronke ase a dreynt mous,

Thenne we schule borewe the wed ate bayly.’

‘We’ll drink to him like friends in excellent booze,

and our sweet lady will sit right next to him.

When he’s as drunk as a drowned mouse,

we’ll go to the bailiff and redeem your pledge.’

And without this evidence, clearly, the indictment will be quashed! It all sounds a most plausible way to work, and one which casts an unexpected light on the downtrodden serfs of medieval England – not as downtrodden as all that, obviously. Their good-natured resourcefulness seems to be an element in the make-up of Tolkien’s hobbits. More significantly, the poem makes one wonder about the unofficial elements of early literary culture. Were there other ‘Man in the Moon’ poems? Was there a whole genre of sophisticated play on folk-belief? There could have been. Tolkien’s 1923 poems attempt to revive it, or invent it, fitting into the gaps between modern doggerel and medieval lyric, creating something that might have existed and would, if it had, account for the jumble and litter of later periods – very like Gothic and ‘i-mutation’.

One sees that the thing which attracted Tolkien most was darkness: the blank spaces, much bigger than most people realise, on the literary and historical map, especially those after the Romans left in AD 419, or after Harold died at Hastings in 1066. The post-Roman era produced ‘King Arthur’, to whose cycle King Lear and King Cole and the rest became eventual tributaries. Tolkien knew this tradition well and used it for Farmer Giles of Ham (published 1949, but written much earlier), the opening paragraphs of which play jokingly with the first few lines of Sir Gawain. However he also knew that whatever the author of Sir Gawain thought, the Arthurian tradition was originally non-English, indeed dedicated to the overthrow of England; its commemoration in English verse was merely a final consequence of the stamping-out of native culture after Hastings, a literary ‘defoliation’ which had also led to the meaninglessness of English names like ‘Fawler’ and the near-total loss of all Old English heroic tradition, apart from Beowulf. What, then, had happened to England and the English during those ‘Norman centuries’ when, it might be said, ‘language’ and ‘literature’ had first and lastingly separated?

Tolkien had been interested in that question for some time. Not much was known about Early Middle English, and indeed several of its major texts remain without satisfactory editions today. However, one important work was evidently the Ancrene Wisse, a ‘guide for anchoresses (or female hermits)’, existing in several manuscripts from different times and places, but one of few Middle English works to be translated into French rather than out of it. With this were associated several other texts with a ‘feminist’ bias, the tract on virginity Hali Meiðhad, the saints’ lives Seinte Juliene, Seinte Marherete, Seinte Katherine, the little allegory Sawles Warde. All looked similar in dialect, and in sophistication of phrase; on the other hand their subject-matter meant they were unlikely ever to take the ‘literature’ side by storm. What could be said about them?

Tolkien began with a review of F. J. Furnivall’s edition of Hali Meiðhad, in 1923; he went on to make ‘Some Contributions to Middle English Lexicography’ in Review of English Studies (1925), most of them drawn from Ancrene Wisse, and some of them incidentally interesting, like the remark that medi wið wicchen must mean, not ‘meddle with witches’ but ‘bribe, purchase the service of witches’, apparently a known practice to the author of the ‘Rule’. In ‘The Devil’s Coach-Horses’ in the same periodical that year he spent enormous effort on the single word eaueres from Hali Meiðhad, arguing that it did not mean ‘boars’ as the OED had said, but ‘heavy horses, draft horses’. Philologically this was interesting as showing a Germanic root *abra-z, meaning ‘work’ and connected with Latin opus. Mythologically it was interesting too as showing an image of the devil galloping away not on fire-breathing steeds, but on ‘heavy old dobbins’ – a contemptuous barnyard image of evil. All very well, but still, some would have said, distinctly peripheral.

The breakthrough came with Tolkien’s article for Essays and Studies (1929), ‘Ancrene Wisse and Hali Meiðhad’, the most perfect though not the best-known of his academic pieces. This rested in classic philological style on an observation of the utmost tininess. In Old English a distinction was regularly made between verbs like hé hiereð, híe híerað, ‘he hears, they hear’, and hé lócað, híe lóciað ‘he looks, they look’. An -að ending could be singular or plural, depending on what sort of a verb it was attached to. This clear but to outsiders utterly unmemorable distinction was, after Hastings, rapidly dropped. Two manuscripts, however, one of Ancrene Wisse, the other of its five associated texts, not only preserved the distinction but went on to make another new one, between verbs within the lócian class: they distinguished e.g. between ha polieð, ‘they endure’, O.E. híe poliað, and ha fondið, ‘they inquire’, O.E. híe fondiað. The distinction had a sound phonological basis and was not the result of mere whim. Furthermore the two manuscripts could not have been by the same man for they were in different handwriting. Evidently – I summarise the chain of logic – they were the product of a ‘school’; so were the works themselves, composed in the same dialect by another man or men; and this ‘school’ was one that operated in English, and in an English descended without interruption from Old English, owing words certainly to the Norse and the French but not affected by the confusion their invasions had caused. To put it Tolkien’s way:

There is an English older than Dan Michel’s and richer, as regular in spelling as Orm’s [these are two other relatively consistent writers of Middle English] but less queer; one that has preserved something of its former cultivation. It is not a language long relegated to the ‘uplands’ struggling once more for expression in apologetic emulation of its betters or out of compassion for the lewd, but rather one that has never fallen back into ‘lewdness’, and has contrived in troublous times to maintain the air of a gentleman, if a country gentleman. It has traditions and some acquaintance with books and the pen, but it is also in close touch with a good living speech – a soil somewhere in England. (‘AW’, p. 106)