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

Yet what this obsessive playing with words shows, better than anything, is that beneath the fog and fury of academic politics, Tolkien realised that all discussions of ‘language’ and ‘literature’ were irretrievably poisoned by the very terms they were bound to use. When he was not simply playing for his side, he accepted that ‘lang.’ was just as foolish a rallying-cry as ‘lit.’. In his manifesto of 1930, ‘The Oxford English School’ he even suggested that both terms should be scrapped in favour of ‘A’ and ‘B’ – thus attempting, with something very close to lèse majesté, to introduce the curriculum of a ‘redbrick’ university, Leeds, to the ivory towers of Oxford, with sad if entirely predictable lack of success.5 The same article makes it clear that he thought both ‘linguistic’ and ‘literary’ approaches too narrow for a full response to works of art, especially early works of art, and that furthermore what was needed was not some tame compromise between them (which is all most Schools of English usually manage to provide), but something as it were at right angles to both. This third dimension was the ‘philological’ one: it was from this that he trained himself to see things, from this too that he wrote his works of fiction. ‘Philology’ is indeed the only proper guide to a view of Middle-earth ‘of the sort which its author may be supposed to have desired’. It is not Tolkien’s fault that over the last hundred years ‘philology’ as a term and as a discipline, has been getting itself into even worse tangles than ‘English literature’.

Dictionary definitions are, symptomatically, unhelpful. The OED, though conceived and created by philologists and borne along by the subject’s nineteenth-century prestige, has almost nothing useful to offer. ‘Philology’ it suggests, is: ‘I. Love of learning and literature; the study of literature in a wide sense, including grammar, literary criticism and interpretation … polite learning. Now rare in general sense.’ Under 2 it offers ‘love of talk, speech or argument’ (this is an offensive sense in which philology is mere logic-chopping, the opposite of true philosophy); while 3 recovers any ground abandoned in 1 by saying it is ‘The study of the structure and development of language; the science of language; linguistics. (Really one branch of sense 1.)’ So ‘philology’ is ‘lang.’ and ‘lit.’ too, all very charitable but too vague to be any use. The Deutsches Wörterbuch set in motion by Jacob Grimm (himself perhaps the greatest of all philologists and responsible in true philological style for both ‘Grimm’s Law of Consonants’ and Grimms’ Fairy Tales) could do little better, defining philologie with similar inclusiveness as ‘the learned study of the (especially Classical) languages and literatures’. The illustrative quotation from Grimm’s own work is more interesting in its declaration that ‘none among all the sciences is prouder, nobler, more disputatious than philology, or less merciful to error’; this at least indicates the expectations the study had aroused. Still, if you didn’t know what ‘philology’ was already, the Grimm definition would not enlighten you.

The matter is not cleared up by Holger Pedersen’s assertion of 1924 that philology is ‘a study whose task is the interpretation of the literary monuments in which the spiritual life of a given period has found expression’6 (for this leaves you wondering why ‘spiritual’ has been put in and ‘language’ for once left out); nor by Leonard Bloomfield’s aside a year later, when, proposing the foundation of a Linguistic Society for America, he explicitly rejected the term ‘philological’ and noted that while British scholars tended to use it to mean ‘linguistic’, Americans would prefer to keep the latter term and to revere philology rather more from a distance as ‘that noblest of sciences … the study of national culture … something much greater than a misfit combination of language plus literature’.7 Anyway some Britons were very far removed from his position. John Churton Collins, nineteenth-century man of letters and candidate for an Oxford Chair, had written in 1891 (it was part of his campaign to keep men like Joseph Wright, Tolkien’s tutor, out of any prospective English School at Oxford):

it [i.e. philology] too often induces or confirms that peculiar woodenness and opacity, that singular coarseness of feeling and purblindness of moral and intellectual vision, which has in all ages been the characteristic of mere philologists … [it] too often resembles that rustic who, after listening for several hours to Cicero’s most brilliant conversation, noticed nothing and remembered nothing but the wart on the great orator’s nose.8

Opinions such as this clung on a long time in England. Tolkien wrote in 1924 ‘“Philology” is in some quarters treated as though it were one of the things that the late war was fought to end’ (YWES 4, p. 37). When I first read this I took it to be a joke. However just three years before the British Board of Education had printed a Report on The Teaching of English in England which declared, among much else, that philology ought not to be taught to undergraduates, that it was a ‘German-made’ science, and (this comes in a footnote on p. 286) that by contributing to German arrogance it had led in a direct way to the outbreak of World War I.

Philology was ‘the noblest of sciences’; it was literary; it was linguistic; it was German; it was Classical; it was different in America; it was about warts on noses; it was ‘the special burden of the Northern tongues’ (Tolkien speaking, ‘OES’ p. 780); also ‘the special advantage they possess as a discipline’ (Tolkien once again, in the same sentence). This begins to sound like the Babel of conflicting voices which Tolkien guyed so fiercely in his lecture on Beowulf, except that in this case the final universal chorus of all voices ‘it is worth studying!’ would clearly be somewhat ragged. If no single answer to the question ‘what is philology?’ can be found, at least few authorities would dissent from the view that the redefinition of philology – the moment when it stopped being used in the OED’s vaguest senses of ‘love of talk’ or ‘love of learning’ – came in 1786 when Sir William Jones informed the Bengal Society in Calcutta that Sanskrit resembled Greek and Latin too strongly for this to be the result of chance, but that all three, together with Germanic and Celtic, must have ‘sprung from some common source which, perhaps, no longer exists’.9

Obviously this thought must have crossed many minds before 1786, for even between English and Latin, say, there are enough similarities – one, two, three, unus, duo, tres – to make one think there may be some sort of a connection. But until the turn of the eighteenth century such speculations had foundered immediately on the great reefs of dissimilarity surrounding the occasional identical rocks. After all the main thing anyone knew about languages was that they were so different they had to be learnt one at a time. The great alteration Jones and his successors brought to the problem was the idea of looking not for chance resemblances – which had already been used to ‘prove’ relationships all over the map – but for regular change. Bad in modern Persian had the same sound and sense as ‘bad’ in English (remarked A. E. Pott in 1833), but that was just coincidence. On the other hand xvahar in Persian was originally the same word as xo in Ossetic, and both were related to English ‘sister’; furthermore the intermediate stages could be inferred and on occasion recovered.10 Like many mental revolutions, this linguistic one depended on being counterintuitive. It was also to an intense degree comparative, using many languages to explain and corroborate each other; and, since different stages of the same language could be used comparatively, by nature overwhelmingly historical. ‘Philology unfolds the genesis of those laws of speech which grammar contemplates as a finished result’ says a citation in the OED, dated 1852. Its author did not mean ‘philology’ in any of the senses quoted from the OED above; he meant comparative philology, the science inspired by Sir William and carried on through many inheritors to Professor Tolkien himself. One may remark that the confidence with which ‘genesis’ is approached was characteristic of the time.

By 1852, indeed, ‘the new philology’ had many triumphs to look back on, with several yet to come: one might pick out the prize-winning essay of Rasmus Rask in 1814, on Old Icelandic, and on the relationship of Scandinavian languages to Slavic, Celtic, Finnish and Classical ones; the enormous ‘Comparative Grammar’ or Vergleichende Grammatik of Franz Bopp in 1833–49, which covered Sanskrit, Zend, Armenian, Greek, Latin, Lithuanian, Old Slavic, Gothic and German; the Deutsche Grammatik (1819) of Jacob Grimm, and all their many successors.11 The point which all these works brandished was the intensely systematic nature of discovery, expressed as time went on increasingly by the word ‘laws’ (see OED citation above), and on the analogy of physics or chemistry by the association of laws with discoverers: Grimm’s Law, Verner’s Law, Kuhn’s Law, Thomsen’s Law, etc. There was and still is something insidiously fascinating about the relationships these laws uncover, in such detail and such profusion. Latin pisces is the same word as Old English fisc, observed Jacob Grimm, or indeed modern English ‘fish’; pes is the same as ‘foot’ and pellis as ‘fell’ (the old word for ‘skin’). What about porcus and ‘pig’ though, where the p/f alternation breaks down? Well, there is an Old English word fearh which corresponds properly, noted Grimm, its modern descendant being ‘farrow’, again an old or dialectal word for a ‘birth’ of piglets. The mill of comparisons will not work on basic or standard or literary languages alone, but demands ever-increasing grist from older or localised or sub-standard forms. The reward it offers is first an increasing sense that everything can be worked out, given time and material, second an exciting tension between the modern meanings of words – words everyone has known all their lives – and what appear as the ancient meanings. ‘Daughter’ in modern Hindustani comes out as beti; yet there is a connection between the two languages in the word dudh, ‘milk’. In ancient days, it seems, a word like Sanskrit duhitar meant ‘the little milker’; but the job was so often given to daughters that task and relationship became fused. It ‘opens before our eyes a little idyll of the poetical and pastoral life of the early Aryans’ enthused Max Müller,12 whose lectures on comparative philology bowled over not only (or not even) the learned world in the 1860s and after, but also London’s high society. Comparison was the rage: it didn’t tell you only about words, it told you about people.

But somewhere towards the end of the nineteenth century things had begun to go wrong. As is obvious from all that Tolkien ever said about literature and about philology, he felt that he had taken over (perhaps unfairly, but possibly not) a losing position in the academic game from his predecessors. Why – he could hardly have helped wondering – was that? Why had philology so ignominiously belied its promise?

Probably the short answer is that the essence of comparative philology was slog. There is something wistful in Tolkien’s astonished praise of the ‘dull stodges’ of Leeds University (Biography, p. 111), in his insistence that at Leeds anyway ‘Philology is making headway … and there is no trace of the press-gang!’ (Letters, p. 11). For matters were different elsewhere. No science, Jacob Grimm had said of philology, was ‘prouder, nobler, more disputatious, or less merciful to error’ (my italics). All its practitioners accepted, to a degree now incredible, a philosophy of rigid accuracy, total coverage, utter right and utter wrong: in 1919 the old and massively distinguished Eduard Sievers happily put his reputation on the line when he offered to dissect a text provided unseen by Hans Lietzmann, and to show from linguistic evidence how many authors had composed it (he had already done the same thing to the Epistles of Paul). He got Lietzmann’s specimen totally wrong. But no one said the idea of the test itself was unfair.13 Further down the scale, the discoveries of Grimm and his successors as far as Ferdinand de Saussure (now famous for inventing ‘structuralism’ but before that a student of Ablaut) were communicated increasingly to students as facts, systems of facts, systems divorced from the texts they had been found in. We must have philology within English Studies, wrote F. York Powell the Icelandicist in 1887, ‘or goodbye to accuracy’.14 The claim was false – you can be accurate about other things besides sound-shifts – but after seventy years of unbroken progress for the subject it was also damningly unambitious. Looking back many years later, R. W. Chambers (the man who turned down the Chair of Anglo-Saxon which eventually went to Tolkien in 1925) summed up success and failure by observing that in 1828 ‘the comparative philologist was like Ulysses’ but ‘Scoffers may say that my parallel is all too true – that students of comparative language, like [Dante’s] Ulysses, found only the mountain of Purgatory – Grimm’s Law, Verner’s Law, Grassmann’s Law – rising in successive terraces of horror – and then were overwhelmed …’ 15 Scoffers said exactly that; their viewpoint became dominant; comparative philology seen as ‘hypothetical sound-shiftings in the primeval German forests’16 went into a decline nearly as precipitate as its rise.

This is why ‘philology’ has first the old vague sense of ‘love of learning’; then the new nineteenth century one of ‘study of texts leading to comparative study of language leading to comprehension of its evolution’; and in the twentieth century the specialised meaning, within departments of English Studies, of ‘anti-literary science kept up by pedants (like Professor Tolkien) which ought to be stopped as soon as possible’. But these interesting semantic changes leave something out: the ‘spiritual life’ waved at by Holger Pedersen, the ‘national culture’ saluted by Leonard Bloomfield – or, to put it another way, the Grimms’ Fairy Tales.

Lost Romances

For philology, after the Rask-Bopp-Grimm breakthrough, had moved in other directions beside the phonological and morphological. The mill of historical comparison called increasingly for fresh material, and one natural effect, besides the study of language in general, was the study of languages in particular. Scholars became much more interested in unread texts; they also became spectacularly better at reading them, at producing dictionaries of stone-dead languages. As Tolkien noted himself (‘Preface’, p. xii), the word hós(e) in Beowulf was never found anywhere else in Old English, so that one would have to guess at its meaning from context, were it not for the fact that philology proved it was the ‘same’ word as Old High German hansa, as in ‘Hanseatic League’, with the meaning ‘retinue’ or possibly ‘band of people connected by mutual oaths’. The dead languages furnished comparative material; the comparative material illuminated dead languages. Men learnt to read Hittite, recognised as an Indo-European language in the 1920s (with marked effect on Old Testament studies), Tokharian (another Indo-European language once spoken by steppe-nomads but now represented mostly by texts preserved accidentally in an oasis in Turkestan), more recently to decipher ‘Linear B’ (an exploration of Cretan archaeology which would have been impossible in a pre-Bopp era). Much obscurer discoveries were made. A whole nation was theorised to lie behind the tiny fragment of Kottish, a language spoken when it was investigated by only five people. Holger Pedersen said of their relatives the Yenisei that they seem to be ‘the last remnants of a powerful folk who, with the Thibetan empire as their southern neighbour, ruled over a great part of Siberia, but were at length compelled to submit to the Turks’.17 Yet of their rule no traces remain other than linguistic ones. The romance of these investigations can still be felt. It is a large-scale analogue of Muller’s remarks on duhitar, of the awareness that some forms even of modern language took you back to the Stone Age (as in English ‘hammer’, cognate with Old Slavic kamy, ‘stone’). The romance became stronger, perversely, the closer it got to home.

Thus Old English itself looked very strikingly different after the philologists got hold of it – and it was they who insisted on calling it Old English instead of Anglo-Saxon to mark what they saw as an essential continuity. The story of Gothic, however, was even more dramatic. Some awareness of this language had been around from an early period. People knew that such texts as the Uppsala Codex Argenteus were in Gothic, that the Goths were an East Germanic tribe who had overrun parts of the Roman Empire from about AD 376, that they had been converted to literacy and Christianity, and become linguistically extinct some time round the eighth century. Philology shattered this picture. For one thing Gothic became suddenly more than comprehensible, it became vital: it was the earliest Germanic language recorded at any length, Germanic was the area of most philologists’ main interest (they were mostly Germans), and Gothic exhibited, in ways that Old English and Old High German did not, stages in the history of all the Germanic languages inferable from but not recorded in its cousins. So, modern English says ‘old’ but ‘elder’, Old English (in its Early West Saxon form) eald but ieldra, both say (more or less) ‘to heal’ but ‘hale (and hearty)’. For these Gothic offers respectively altheis, althiza, háiljan, háils. The common element deduced is that when an -i- or -j- followed a or ái in old Old English (this goes back to the time before Englishmen had learnt to write) speakers began to change the earlier vowel into e, ae – with similar changes affecting other vowels. Where there is a succeeding -i- in Gothic there is a change of vowel in Old (and often still in modern) English; not otherwise.

This phenomenon, known as ‘i-mutation’ became one of the most familiar horrors of university philology, but there is in it something both mysterious and satisfactory: a whole series of things which people said, and still say, without in the least knowing why, turn out to have one very old but clear, 100 per cent predictable reason. It is almost like genetics. No wonder that Grimm said Gothic was a ‘perfect’ language, Tolkien (‘EW’, p. 38) that it took him by storm. A further stage in the developing romance of ‘Gothia’ was the thought that the Goths might not be extinct. At some time in the 1560s one Ogier van Busbecq, a Fleming then acting as ambassador in Istanbul, had heard some foreigners whose speech sounded familiar. He recorded a list of words from them and printed it in 1589. They proved to be Gothic, nearly a thousand years out of place. Their interest aroused several centuries later, scholars could for a while entertain the hope that a living Gothic was still somewhere in existence, as a kind of Abominable Snowman of language. Alas, it wasn’t. But at least it became clearer how Gothic had survived, in the remote Crimea, and it became possible to piece together once again the history of a vanished people.

It is not too much to say that this language and this people haunted Tolkien all his life. As is noted by Christopher Tolkien (UT, p. 311), the names of the leaders of the Rohirrim before the dynasty of Eorl are not Old English, like everything else in the Riders’ culture, but Gothic, e.g. Vidugavia, Vidumavi, Marhwini, etc. (see LOTR, pp. 1021–2). They function there to suggest language behind language and age behind age, a phenomenon philologists so often detected. On a larger scale the Battle of the Pelennor Fields closely follows the account, in Jordanes’s Gothic History, of the Battle of the Catalaunian Plains, in which also the civilisation of the West was preserved from the ‘Easterlings’ and in which the Gothic king Theodorid was trampled by his own victorious cavalry with much the same mixture of grief and glory as Tolkien’s Théoden. Perhaps the most revealing remark, however, comes in a letter from Tolkien to his son Christopher after the latter had read a paper on the heroes of northern legend. In this he praised his son’s paper for the light it shed on men and on history, but added:

All the same, I suddenly realized that I am a pure philologist. I like history, and am moved by it, but its finest moments for me are those in which it throws light on words and names! Several people (and I agree) spoke to me of the art with which you made the beady-eyed Attila on his couch almost vividly present. Yet oddly, I find the thing that really thrills my nerves is the one you mentioned casually: atta, attila. Without those syllables the whole great drama both of history and legend loses savour for me. (Letters, p. 264)

The point is that Attila, though a Hun, an enemy of the Goths under Theodorid, and a byword for bloody ferocity, nevertheless does not appear to bear a barbarian name. ‘Attila’ is the diminutive form of the Gothic word for ‘father’, atta: it means ‘little father’ or even ‘dad’, and it suggests very strongly the presence of many Goths in Attila’s conquering armies who found loot and success much more attractive than any questions of saving the West, Rome or civilisation! As with duhitar, ‘little milker’, or kamy as a cognate for ‘hammer’, the word tells the story. Tolkien went on in his letter to say that in his mind that was exactly how The Lord of the Rings grew and worked. He had not constructed a design. Instead he had tried ‘to create a situation in which a common greeting would be elen síla lúmenn’ omentielmo’. Literary critics might not believe him, but philologists (if any were left) ought to know better.

Atta, Attila: what’s in a name? One answer is, a total revaluation of history. It is instructive to look at older and newer editions of Edward Gibbon’s Decline and Fall of the Roman Empire (first published 1776–88). Gibbon knew the Goths from many Roman and Greek historians, including Jordanes, but these were his only sources of information and he could not imagine another one. ‘The memory of past events’, he remarked with classically-educated superciliousness, ‘cannot long be preserved, in the frequent and remote emigrations of illiterate Barbarians’ (chapter 26). As for the great Gothic king of the fourth century, he said, ‘The name of Hermanric is almost buried in oblivion’. It did not stay buried. ‘Hermanric’ turned up in recognisable form in Beowulf (not printed till 1815) as Eormenric. The same name and man, with little stories attached, appeared also in the Old English poems Déor and Widsith. As Ermenrich he survived into the Middle High German romances of Dietrichs Flucht, Alpharts Tod, and many others. Most powerfully, Jörmunrekkr turned out to be a most prominent character in the Old Norse poems of the Elder Edda, which had lain unnoticed in an Icelandic farmhouse till the 1640s, and not been published in full till Rasmus Rask did the job in 1818. The ‘illiterate Barbarians’ were not as forgetful as Gibbon thought. They could at least remember names, and even if these had been affected by sound-changes in the same way as other words, no archaic poet produced anything as false as Gibbon’s ‘(H)ermanric’. From the joint evidence of old poems in English, Norse and German one could in fact deduce that the king’s name, though never recorded in Gothic, must have been *Aírmanareiks.

And, as with ‘Attila’, there is a thrill of old passion lurking in the name, buried though this may be in editors’ footnotes and the inferences of scholarly works. The tales of Ermanaric’s death vary. He committed suicide (round AD 375) for fear of the Huns, says an early Roman source. Jordanes tells a more complicated story of treachery, punishment and revenge. The Old Norse poems, more grisly and more personal, insist that Ermanaric was attacked by his brothers-in-law for murdering their sister, and was left after their death under a hail of Gothic stones – for on them no weapon would bite – to survive as a heimnár or ‘living corpse’, a trunk with both arms and legs cut off. This last tale seems totally unlikely. But it does preserve some agreement over names and incidents with Jordanes: maybe something peculiar and tragic did take place during the collapse of the Gothic Empire in the fourth century. To the philologist who compared these versions there was a further charm in guessing what strange chains of transmission and quirks of national bias had transformed king into villain. Had the defeated Goths cast him as a scapegoat? Had he been made a wife-murderer to gloss over the feelings of those Goths who changed sides and joined the ‘Easterlings’, calling the Hunnish king their ‘little father’? Had Crimean Goths sung lays of Ermanaric to Norsemen of the Varangian Guard in the courts of the Greek emperor? Tolkien followed these inquiries closely, buying for instance the volumes of Hermann Schneider’s Germanische Heldensage as they came out 1928–34,* and claiming in 1930 (‘OES’, pp. 779–80) that Gothic was being studied under his direction not only for sound-laws but ‘as a main source of the poetic inspiration of ancient England and the North’. As he said in the letter quoted above, the legends of heroes had a fascination in themselves; they were also part of ‘a rational and exacting discipline’.