An after-dinner speech given at the Tolkien Society’s AGM, 19 April 2008, revised 2015 for the anniversary meeting of Taruithorn
Note: the conclusion of this piece is in no way intended to undermine Tolkien’s achievements, but simply to draw attention to a logical tension within the linguistic framework of his Legendarium
Mathom, eleventy, flet, wolfrider, barrow-wight, kingsfoil: not Elvish but English words skilfully devised by Tolkien to play their part in his great story. In our book The Ring of Words, Peter Gilliver, Jeremy Marshall, and I set out to investigate such words and names, denoting people, natural features, artefacts, and institutions of Middle-earth, especially in the Third Age. Words referring to hobbit custom, such as smial, shirriff, thain, and farthing; words used by others, such as elven, flet, halfling, and waybread. Whether they relate to human, elvish, entish, or even orkish life, they are English-style words because the story is told in modern English, and the protagonists—the hobbits, Aragorn the Ranger, Gandalf, and so on—are represented as speaking modern English. Alongside English there are many glimpses of other languages: sometimes Elvish ‘loanwords’ are employed alongside their English equivalent, for example lembas alongside waybread, or Haradrim alongside and much more usually than Southrons. This makes it clear to the thoughtful reader that the protagonists did not really speak English, as the narrator of The Lord of the Rings points out, but actually spoke the Common Speech or the Westron (the latter a word invented by Tolkien on the basis of the pre-existing real word Southron).
Now Tolkien has been compared with Lewis Carroll in his inventiveness, and although he himself pointed out the great differences, he was not wholly averse to the comparison. In Through the Looking-glass (1871) Carroll introduced the mock-heroic poem ‘Jabberwocky’. I’m sure you know the first verse:
Twas brillig and the slithy toves
Did gyre and gimble in the wabe;
All mimsy were the borogoves
And the mome raths outgrabe.
Carroll had written this single verse sixteen years previously. It occurred in a work called The Rectory Umbrella and Mischmasch, written in 1855 but not published till 1932, and in this context, Carroll pretended it was a fragment of Anglo-Saxon verse. Let’s consider the words wabe and outgrabe: to a reader with only a rather general idea of Old English they might appear ‘Anglo-Saxon’ enough. For example, outgrabe looks like an archaic past tense of an irregular verb, like outspake or outbrake.
But Tolkien would never have passed off words like these either as Anglo-Saxon or as modern descendants of Anglo-Saxon words. Why not? First, because they conflict with the known principles of English sound-development. In Anglo-Saxon a single b could not occur at the end of a word or between two vowels; only a double -bb- could arise between vowels, and in such a place it could only be preceded by a short vowel. Hence there are no words of Anglo-Saxon origin that end in -abe or have it in the middle; nearly all words with b in this position, such as label and gable, are taken from other languages; and in any case no Old English ‘strong’ verb could have had a b in this position, so outgrabe is an impossible past tense.
Secondly, and more importantly, because they are wholly invented; they are not based on words in real languages. All Tolkien’s invented English words are based on some previously existing word, or name, or word-forming element in modern English, Middle English, Old English, or Old Norse (the language of the sagas). There is one exception, strangely enough: the word for which Tolkien is most famous. It is a remarkable paradox that hobbit is virtually the only word of English form that Tolkien (apparently and untypically) invented from nothing. If Tolkien had inadvertently saddled himself with two such etymological monstrosities as wabe and outgrabe, he would have tried to ‘discover’ philologically convincing explanations for them. It is characteristic that Tolkien, uneasy that hobbit seemed to have no prehistory, devised an (imaginary) Anglo-Saxon compound to be its etymological ancestor, namely holbytla (hole-dweller), and put this into the mouths of the Rohirrim, whom he represented as speakers of Anglo-Saxon.
Tolkien’s English words and names are successful and satisfying because they fit properly into the sound pattern of English; they are not barbarous or outlandish or absurd. They fit into the structure of English; they use word-forming elements such as -ling and -ish or they form part of an existing set, as for example eleventy and tweens. They fit with vocabulary that is appropriate to Third-Age Middle-earth: a world based on agriculture, with a traditional social structure and customs; they eschew word elements drawn from modern scientific terminology or urban slang (the latter is reserved for orcs and Saruman’s henchmen). Above all, they are all in some sense ‘real’ words: form and meaning are rooted in historical truth. Some of them exist in current English, but with a different meaning. For example, farthing—not too long ago the name of a coin. Tolkien’s meaning ‘one of the four parts into which a territory is divided’ is not found in any English dictionary. But it is a real meaning belonging to the equivalent word in Old Norse, fjórðungr: Iceland used to be divided into four fjórðungar, north, south, east, and west. Tolkien has taken a word from Old Norse and converted its form into the equivalent form in English, because just as English farthing is made up of fourth and the ending -ing, so the Norse word is made up of the equivalent Norse elements. Mirkwood is a similar adaption of a Norse compound into English form, being based on the Old Norse name for the great central European forest, Myrkviðr. Where he has introduced an Anglo-Saxon word Tolkien has carefully applied his knowledge of the historical sound changes of English to give the word a current English form: so Anglo-Saxon máðm ‘treasure’ has got the form it would now have if it had survived, mathom, and Anglo-Saxon eorcnanstān ‘precious stone’ has likewise been updated to Arkenstone. All these words have a definite feel about them: they are strange and yet lie just on the edge of familiarity, so that they seem to come from a half-forgotten time.
Tolkien did not just pluck each word out of some handy source on the spur of the moment. Rather he brooded on a word till it yielded its secret: a special application within the story or even a whole story of its own. The Old English words máðm, orþanc, samwīs, emnet, entisc, and dēagol (and others) all occur within a few pages of each other in lists at the back of Joseph Wright’s Anglo-Saxon Grammar, a book which Tolkien would have used as an undergraduate. I think that he contemplated each word for a long time. Flet is an ancient word meaning ‘floor’ and ‘hall’; Tolkien, having explained the word as an inhabited platform in a tree, adds the aside ‘as such things were called in those days’. In fact it has never meant this in our world. But it did in Middle-earth: this is the story that Tolkien drew from a word he had thought about ever since he encountered it, probably in Beowulf. A particularly good example of this process is the word ent; in Old English literature it’s rare and mysterious: Roman ruins are described as orþanc enta geweorc, ‘the cunning works of giants’. Tolkien wrote:
I always felt that something ought to be done about the peculiar A. Saxon word ent for a ‘giant’ or mighty person of long ago—to whom all old works were ascribed.
From this tiny, obscure word he drew what many people think is his most brilliant invention: Treebeard. The way that Tolkien worked at chosen words from the past until they yielded up a new application or a new history is of a piece with his approach to narrative. As he wrote about this:
My feeling throughout, especially when stuck, [was] that I was not inventing but reporting (imperfectly) and had at times to wait until ‘what really happened’ came through.
Again, take the name Westernesse: it has an air of mystery and ancientness: its termination, as Tolkien himself pointed out, is like that of fabled lands in Arthurian romance, Logres, Lyonesse. Remarkably, this word occurs only in the little known Middle English romance King Horn. But he introduces it as if it is a perfectly familiar term that will explain the unfamiliar name Númenor: ‘It was called Númenor, that is Westernesse, and Andúnië or the Sunsetland’. The Lord of the Rings introduces it even more baldly; it just mentions: ‘The Kings of Men that came over the Sea out of Westernesse’. But it was familiar and natural to Tolkien. Tolkien’s special words have a dual nature: their links with the languages of the actual world are necessary to give them historical depth and half-familiarity, but Tolkien has infused them with a deeper meaning. If he had just made up words, as Carroll did, The Lord of the Rings and the other works would simply not have the same powerful evocativeness.
Let’s go back to the word farthing. The adaptation is a kind of learned word-play: Tolkien’s farthing, based on Old Norse, is a kind of pun on the familiar farthing, the name of the small coin, and the link is the etymology. His coinages are full of learned word-play. Consider the place-name Cracks of Doom, where the Ring is destroyed. The phrase crack of doom was an established English expression (occurring in Macbeth) in which ‘crack’ refers to sound: it means the last trump, the trumpet-call announcing the day of judgement. Tolkien has deliberately taken ‘crack’ in a different meaning, that of a physical fissure in the ground, and applied it to the fearful chasm on the mountain Orodruin. The phrase borrows its terror from the original meaning but it is a pun—a play on two different meanings of ‘crack’. Another example is the ent-name Quickbeam, whose primary meaning is ‘sudden beam (of light)’, as the Sindarin equivalent Bregalad shows. It is also a play on the English dialect word quickbeam meaning ‘rowan’ or ‘mountain ash’, the kind of tree beloved of the ent Quickbeam. Or take the name of Beorn. In Anglo-Saxon the word means ‘warrior’or ‘hero’. Its Old Norse cousin, bjǫrn, means ‘bear’. From that etymological divergence Tolkien in The Hobbit spins the story of a mighty man who changes into a bear by night—based on the northern stories of so-called ‘skin-changers’.
Also in hobbit history there is a learned pun which probably not many people notice. When the hobbits migrated from the east into the Shire in Third Age 1601, they were led across the Brandywine by the two brothers, Marcho and Blanco. A number of hobbit names end in -o, and several, like Frodo, Drogo, Odo, and Otho, are borrowed from medieval Frankish, as are some other hobbit names like Fredegar and Isengrim. Marcho and Blanco’s names have this Frankish o-form and are based on two Germanic words meaning ‘horse’, which in Old English are mearh (compare ‘mare’) and blanca (literally meaning ‘the white one’). Now why should Tolkien give the two brothers the name ‘horse’? Well, consider the two leaders of the first Anglo-Saxons to cross the Channel and settle in England, according to tradition: Hengest and Horsa. Both their names mean ‘horse’. (You will recall that Tolkien used to lecture on Hengest, literally ‘stallion’.) If you’re not convinced about Marcho and Blanco, let me point out that in an earlier draft of The Lord of the Rings the brothers’ names are Italian, Marco and Cavallo, and cavallo is the Italian for ‘horse’.
Tolkien enjoyed puns; there are many in his writings. He liked the silly sort; for example, in a letter written during the Second World War he refers to the hens he kept as the ‘fouls’ (because they were messy and uncooperative). But he also frequently used puns that had both a more serious motivation and a more learned etymological foundation. An example occurs in his 1959 Valedictory Address to the University of Oxford where he remarks ‘when I survey with eye or mind those who may be called my pupils (though rather in the sense “the apples of my eyes”)’. This plays on the etymology of the two English words pupil. The apple of one’s eye, which now means ‘the object of one’s regard’ originally meant ‘the pupil of the eye’ (which was thought to be a hard round object). English pupil (of the eye) is from Latin pupilla, with the same meaning, which originally meant ‘doll’ or ‘young girl’. Pupilla is the feminine of pupillus meaning ‘student’, and from this word we get the other word pupil. I guess not many of the 1959 audience grasped this pun. (By the way, the reason for the eye’s pupil being called ‘doll’ was of course the little image of oneself reflected in the other person’s eye (sometimes called a baby, as in looking babies).) This ‘serious punning’ is the ability and desire to read the same thing in two or more different ways and at the same time to discern an underlying, suggestive, connection.
Moreover, in the invented languages too there are words and names with the same radical duality. The first name in the Legendarium, as you know, is Eärendel. Tolkien knew, and appreciated, its real Germanic etymology (a root meaning ‘dawn’). But, as John Garth says, he wanted also to read in it the unconnected Anglo-Saxon word ēar, meaning sea. Eärendel, then, has two roots: it is an Anglo-Saxon name probably meaning ‘morning star’, but it is also a Legendarium name meaning ‘sea-friend’. Tolkien coyly ascribes the duality to coincidence, in the Notion Club Papers, when Arry Lowdham argues that ‘it is not Anglo-Saxon, or rather, it is not only Anglo-Saxon…I think it is a remarkable case of linguistic coincidence, or congruence’. 
Let’s go back to the earlier point that English in Tolkien’s narrative represents the Westron, the real Common Speech of humans and hobbits. Tolkien summarized this transposition of Middle-earth tongues into real world languages in a note of February 1942:
Language of Shire = modern English, Language of Dale = Norse (used by Dwarves of that region), Language of Rohan = Old English, “Modern English” is lingua franca spoken by all people (except a few secluded folk like Lórien) - but little and ill by orcs.
Were you taken aback, as I was, on first reading, in Appendix F to The Lord of the Rings, that the hobbits’ real word for themselves was kuduk and the Rohirrim really called them kûd-dûkan? Christopher Tolkien refers to this ‘transposition of languages’ as ‘the fiction of authenticity’. He surmises that the need to account for the Dwarves having Old Norse names (instead of Dwarvish ones) was the starting-point for the whole theory:
The conception emerged that the Dwarves had ‘outer names’ derived from the tongues of Men with whom they had dealings...And this was very evidently an important component in the theory of the ‘transposition of languages’: for the Dwarves had Norse names because they lived among Men who were represented in The Lord of the Rings as speaking Norse.
However, the doctrine of the transposition of languages collides with Tolkien’s creative use of learned word-play. Take the name of Saruman’s tower Orthanc; we are told that it
had (by design or chance) a two-fold meaning; for in the Elvish speech orthanc signifies Mount Fang, but in the language of the Mark of old the Cunning Mind.
Orthanc comes from Anglo-Saxon. As a mysterious-sounding word for ‘cunning’ it had probably appealed to Tolkien for a long time—it is in those lists in Wright’s Grammar that I mentioned, where it is a near neighbour of entisc—it was another word, perhaps, that something ought to be done about. But according to the ‘fiction of authenticity’ Old English only represents the language of the Rohirrim; therefore their actual word for ‘cunning mind’ could not have been orthanc, so it could not have coincided with an Elvish name, derived from orth ‘mount’ and anc ‘fang’. As long as the reader is ignorant of the ‘fiction of authenticity’, this name which mysteriously means both cunning mind and Mount Fang has magic, but when the transposition of languages is applied, the ingenious play on words collapses like Isengard itself. To a correspondent who tried to read significance into the resemblances between Elvish words and words in actual languages, Tolkien responded, in effect, that Eärendel is simply a beautiful and suggestive name borrowed from Anglo-Saxon, turned into Elvish, and nothing more. But when it came to Eärendel’s ship Wingalótë, his tune was different. (To understand this point you need to know that Eärendel in Germanic mythology appears to have been connected with a giant named Wade or Wada. Wade had a celebrated boat called Guingelot which Tolkien transmuted into Eärendel’s ship Wingalótë, analysed as an Elvish compound meaning ‘foam-flower’.) Tolkien said in his notes:
Wingalótë..must be retained, since..it is in intention formed to resemble and ‘explain’ the name of Wade’s ship Guingelot.
Tolkien here says that his myth requires Wingalótë to have a root in both worlds. And how about Atalantë , a name for Númenor, and Avallóne or Avallonde, a name for Tol Eressëa? These derive their atmosphere from their resemblance, in form and significance, to Atlantis the drowned land of Greek legend, and Avallon the Faerie destination to which the wounded King Arthur was taken after his last battle. Tolkien however surprisingly remarks, in a letter:
It is a curious chance that the stem √talat used in Quenya for ‘slipping, sliding, falling down’, of which atalantie is a normal (in Quenya) noun-formation, should so much resemble Atlantis.
The dilemma is that, on the imaginative level, the linguistic links between the Legendarium and our actual history and legend are vital, but on the logical level they are self-contradictory and can only be treated as coincidence. We can illustrate this from Appendix D of The Lord of the Rings. In ‘The Shire Calendar’, Tolkien sets out the hobbit names for the days of the week: Sterday, Sunday, Monday, Trewesday, Hensday, Mersday, Highday. Supposedly they are translations of the original Elvish names, reflecting the ruling elements of Arda: Stars, Sun, Moon, Two Trees, etc. Christopher Tolkien writes:
The rhyming of ‘Trewsday, Hensday, Mersday, Highday’ with our ‘Tuesday, Wednesday (Wensday), Thursday, Friday’ he naturally called an accidental likeness; but it was an astonishing coincidence! I am much inclined to think that the Hobbit calendar was the original conception, and that the names of the days were in fact devised precisely in order to provide this ‘accidental likeness’.
Probably the first step, once Tolkien had decided to retain Sunday and Monday, was to create a ‘Star-day’. This was done by turning Saturday into the punning name Sterday, which could plausibly be presented as a modern development of the (invented) Old English Steorrandæg. But according to the ‘transposition of languages’ the hobbits did not really speak English, archaic or otherwise, so why invent a whole set of archaic day names as if to suggest a mysterious connection with ours?
To conclude: many words and names of Tolkien’s mythological world, chiefly the English ones, but also some Elvish ones, are inherently rooted, by word play and word echoes, in their counterparts in real-world languages. Originally, in The Book of Lost Tales and its successors, the mythology was meant to be an extension of the ‘real’ mythology of our world, so this was only to be expected. Our human languages were even supposed to have evolved from Elvish ones. However, independently of all this, Tolkien with carefree abandon in The Hobbit applied Old English and Old Norse names and words to the inhabitants of Middle-earth. When the Legendarium ceased to be the prehistory of England and the two narratives were linked up, the linguistic material rooted in our world posed a dilemma. To solve it, the ‘fiction of authenticity’ emerged, stating that the real-world languages are only used to translate the Middle-earth languages. But this strikes at the taproot, the linguistic echoing and word play that are crucial to Tolkien’s creative imagination. As a symbol of this dilemma stands Orthanc, the cunning mind.
 Peter Gilliver, Jeremy Marshall, and Edmund Weiner, The Ring of Words: Tolkien and the Oxford English Dictionary (Oxford: Oxford University Press, 2008).
 Lewis Carroll, Through the Looking-Glass (1871) i. 21.
 See Peter Gilliver, Edmund Weiner, and Jeremy Marshall, ‘The Word as Leaf: Perspectives on Tolkien as Lexicographer and Philologist’ in Stratford Caldecott and Thomas Honegger (editors), Tolkien’s The Lord of the Rings: Sources of Inspiration (Zurich and Jena: Walking Tree Publishers, 2008), 75.
 Letters 157.
 Though Tolkien probably did not invent the name Treebeard: see Gilliver et al. (2008), 72–3.
 Letters 163.
 The Fall of Numenor, HME V. 14.
 LR, Prologue.
 LR, Prologue and Appendix C.
 HME XII. I. i. 9 and note 14.
 Letters 55.
 The Monsters and the Critics (1983), p. 240.
 John Garth, Tolkien and the Great War, (London: HarperCollins, 2003), p. 45.
 1946, in HME IX. 237.
 HME VI. xxiii. 424.
 Commentary to Appendix on Languages in HME XII. ii. 70-71.
 LR iii. viii.
 August 1967, in Letters 297.
 ‘The Problem of Ros’ in HME XII. ii. xii. 371.
16 July 1964, in Letters 257.
 ‘The Calendars’ in HME XII. i. iv. 125.
 HME I. 24.
 ‘The Lhammas’, HME V. 179.