Tolkien's Not-So-Secret Vice

In 1931, Tolkien wrote an essay about the somewhat peculiar hobby of devising private languages. He called it A Secret Vice. But in Tolkien's case, the "vice" can hardly be called secret anymore.

What, really, is going on inside the head of a man who all his life is toying with enormous linguistic constructions, entire languages that have never existed outside his own notes? For one thing must be perfectly clear: He made very much more of these languages than he could ever hope to include in his stories. True, there are a few Elvish poems and a swarm of exotic names in the annals of Middle-earth, but even so, this is nothing compared to all that Tolkien made. In Tyalië Tyelelliéva # 6, Lisa Star informs us that her own list of published words has twelve thousand entries. We are talking about enormous linguistic constructions. How did it start? How was it done? And why?

Young John Ronald Reuel in the Nevbosh-speaking world

One day quite exactly a hundred years ago, early teenage Tolkien was baffled to hear a couple of other kids communicating in Animalic. This was a primitive play-language that mainly consisted of English words for animals. The inventors of Animalic did not attempt to keep it a secret, and young Tolkien soon learnt some of it. In his essay A Secret Vice (published in The Monsters and the Critics p. 198-219) he gives one example of Animalic: Dog nightingale woodpecker forty, which translates as "you are an ass". (By all means: "ass" here means donkey and nothing else. In Animalic, forty meant donkey, while donkey, of course, meant forty...)

Animalic soon became a dead language, but some of the kids continued their linguistic games. They invented a language called Nevbosh (this being Nevbosh for "new nonsense" - the nonsense replacing Animalic, evidently...) Tolkien was not the originator of this language, but according to himself, he contributed to the vocabulary and helped to standardize the spelling. "I was a member of the Nevbosh-speaking world," he proudly recalls.

Nevbosh was mainly a mixture of heavily distorted English, French and Latin words. It did not represent a real breaking away from English or other normal languages. More than twenty years after it became a dead language Tolkien was still able to remember at least one connected fragment, that he calls "idiotic":

Dar fys ma vel gom co palt 'hoc
pys go iskili far maino woc?
Pro si go fys do roc de
Do cat ym maino bocte
De volt fac soc ma taimful gyróc!'

But for Tolkien, simply distorting existing words (like woc = "cow"!) was in the long run not enough. Already among the Nevbosh kids there emerged something more sophisticated: words that could not be traced to any specific source, but popped up simply because they seemed to fit their meaning - because the combination of sound and sense gave the kids pleasure. Tolkien mentions a word lint "quick, clever". Young John Ronald Reuel never forgot this word: Forty years later he had Galadriel singing how the years in Middle-earth had passed ve lintë yuldar lissë-miruvóreva, like swift draughts of the sweet mead...

Time went by, and Nevbosh joined Latin and Gothic on the long list of dead languages. But Tolkien, still a child, was already devising one of his first wholly private languages: Naffarin. He mentions one Naffarin sentence to illustrate, but there is no translation: O Naffarínos cutá vu navru cangor luttos ca vúna tiéranar, dana maga tíer ce vru encá vún' farta once ya merúta vúna maxt' amámen. Though Naffarin supposedly incorporated some of the latest stages of Nevbosh, we already sense a movement towards "Elvish" forms. Naffarin was inspired by Latin and Spanish, but Tolkien was to find two even more powerful inspirations.

Through the Welsh cellar door into the Finnish wine-cellar

One thing was important to Tolkien. Languages should be beautiful. Their sound should be pleasing. Tolkien tasted languages, and his taste was finely tuned. Latin, Spanish and Gothic were pleasing. Greek was great. Italian was wonderful. But French, often hailed as a beautiful language, gave him little pleasure.

But heaven itself was called Welsh. In his essay "English and Welsh", Tolkien recalls how he once saw the words Adeiladwyd 1887 (It was built 1887) cut on a stone-slab. It was a revelation of beauty. "It pierced my linguistic heart," he recalls. It turned out that Welsh was full of such wonderful words. Tolkien found it difficult to communicate to others what really was so great about them, but in his essay he makes an honest attempt: "Most English-speaking people...will admit that cellar door is 'beautiful', especially if dissociated from its sense (and from its spelling). More beautiful than, say, sky, and far more beautiful than beautiful. Well then, in Welsh for me cellar doors are extraordinarily frequent, and moving to the higher dimension, the words in which there is pleasure in the contemplation of the association of form and sense are abundant." He then lists concrete examples like Welsh wybren being "more pleasing" than English sky. -MC p. 190-193.

But there were more pleasures in store for young Tolkien. One day he found...a Finnish grammar!!! He soon found himself in phonaesthetic ecstasy. "It was like discovering a complete wine-cellar filled with bottles of an amazing wine of a kind and flavour never tasted before. It quite intoxicated me" (Letters:214). High on Finnish he scrapped his latest project ("make your own Germanic language"), for now he had found more powerful inspirations.

Many years later, he stated that the Elvish tongues were "intended (a) to be definitely of a European kind in style and structure (not in detail); and (b) to be specially pleasant. The former is not difficult to achieve, but the latter is more difficult, since individuals' personal predilections, especially in the phonetic structure of languages, varies [sic] widely... I have therefore pleased myself" (Letters:175-176). This in effect meant that from the point he discovered Welsh and Finnish, they were the main influences on his own linguistic constructions.

Of course, he was right in observing that individual taste varies widely. The Welsh language that he loved so much and modeled Sindarin on, was once described as "a mass of grunts and gargling sounds" by a Norwegian radio reporter. Still, many people seem to agree that the Elvish tongues are generally euphonious. Tolkien registrered positive feedback: "The names of persons and places in this story were mainly composed on patterns deliberately modelled on those of Welsh (closely similar but not identical). This element in the tale has given perhaps more pleasure to more readers than anything else in it" (MC:197).

But we are leaping ahead of things; let us return to the very beginning. While World War I was still raging, Tolkien's linguistic constructions definitely became Elvish languages. On March 2nd, 1916, 24-year-old Tolkien wrote to his beloved Edith telling her that he had been working on his "nonsense fairy language - to its improvement. I often long to work at it and don't let myself 'cause though I love it so it does seem such a mad hobby!" Mad or not, he was to give in to his longing and keep working on this hobby throughout his life. - Letters:8.

Exactly at this point, in 1916, while Tolkien was in hospital having survived the Battle of Somme, the very first parts of his "mythology for England" were written - fragments of what would one day become the Silmarillion. At the same time, or rather a little before, he wrote his first Elvish word-lists. One thing triggered the other: "The making of language and mythology are related functions," he observed in A Secret Vice. "Your language construction will breed a mythology" (MC:210-211). Or again in a letter written many years later, shortly after the publication of LotR: "The invention of languages is the foundation. The 'stories' were made rather to provide a world for the languages than the reverse. To me a name comes first and the story follows... [LotR] is to me...largely an essay in 'linguistic aesthetic', as I sometimes say to people who ask me 'what is it all about?' " (Letters:219-220) Few people took this explanation seriously. "Nobody believes me when I say that my long book is an attempt to create a world in which a form of language agreeable to my personal aesthetic might seem real," Tolkien complained. "But it is true." - Letters:264.

From the very beginning, there were two main languages in his mythology: one that sounded much like Finnish, and one that was similar to Welsh. Unlike their inspirations, they were related and were derived from a common primitive language. The Finnish-like language was called "Qenya" right from the start; a small spelling reform was all that stood between it and its final name. The other language was originally called Golgodrin or "Gnomish", it was i-Lam na-Ngoldathon or "the tongue of the Gnomes". (Its later form, so heavily revised that it was not really the "same" language, was long called Noldorin; only as Tolkien was completing LotR did he realize that its real name was Sindarin. But see below.) The first Gnomish lexicon was published a few years ago and turns out to be very comprehensive, probably the most complete "dictionary" Tolkien ever made for any Elvish language. The "Qenya" word-list was finally published in 1998 and turns out to be another very comprehensive document, as can be seen from the indices presented on this web-page (whether by English glosses or Qenya words).

The years passed by and the stories of the Silmarillion evolved, but it seems that the relevance of the original dictionaries soon dwindled: Frequent revisions inevitably rendered them obsolete. In the second half of the thirties, however, Tolkien made a list of some seven hundred Primitive Elvish "stems" and some of their derivatives in later languages. It was apparently this list, the so-called Etymologies, he was referring to when he started to write The Lord of the Rings (he added to the list some words and names from this work, e.g. mith "grey" and rhandir "pilgrim", that together give Mithrandir). The Etymologies was published in its entirety by Christopher Tolkien in The Lost Road p. 347-400. A fairly typical entry goes like this:

MBUD- project. *mbundu: Q mundo snout, nose, cape; N bund, bunn. Cf. *andambundâ long-snouted, Q andamunda elephant, N andabon, annabon.

Tolkien's technique

Christopher Tolkien describes his father's strategy as a language-maker in one formidable sentence: "He did not, after all, 'invent' new words and names arbitrarily: in principle, he devised from within the historical structure, proceeding from the 'bases' or primitive stems, adding suffix or prefix or forming compounds, deciding (or, as he would have said, 'finding out') when the word came into the language, following it through the regular changes in form that it would thus have undergone, and observing the possibilities of formal or semantic influence from other words in the course of its history." The result: "Such a word would then exist for him, and he would know it." (LR:342)

As an example of this process we may use the Elvish numerals. Consider the primitive stems for the words for the numbers 1-10, plus words derived from these stems as they appear in Quenya and Sindarin:

1: MINI: Q minë, S min
2: AT(AT): Q atta, S tad
3: NEL(ED): Q neldë, S neledh
4: KÁNAT: Q canta, S canad
5: LEPEN: Q lempë, S leben
6: ÉNEK: Q enquë, S eneg
7: OTOS/OTOK: Q otso, S odog
8: TOL-OTH/OT: Q tolto, S toloth
9: NÉTER: Q nertë, S neder
10: KAYAN/KAYAR: Q cainen, S caer

Christopher Tolkien notes how his father took into consideration "the possibilities of formal or semantic influence from other words in the course of its history". The numerals provide us with an example of that, too. According to the Etymologies, the Sindarin word for "three" was originally neledh as in the list above. But later it became neled because it was "influenced" by canad "four". (One imagines the Elf counting min, tad, neledh, canad; one day he says neled, canad instead!)

But no matter how much Tolkien was toying with sound-changes and did not just invent new words and names arbitrarily, the words would still have to come from somewhere. Were they, after all, arbitrary anyway? Often not. When Tolkien was interviewed by the Daily Telegraph in 1968 and got to read a preliminary version of the interview before it was printed, he was horrified to discover that he had said this: "When you invent a language, you more or less catch it out of the air. You say boo-hoo and that means something." This was not at all what he really meant; he wasn't sure that he had said this at all. He carefully explained that he made words based on personal predilections, his guide being what he thought was phonetically fitting (Letters:375). It may be discussed how "personal" these associations were. Many would probably agree that quite a few Elvish words in a strange way seem to fit their meaning: elen "star", menel "sky", vanya "beautiful", wen or wendë "maiden", lótë "flower", masta "bread". (Of course one may disagree, too: The present writer thinks MOR, the well-known stem for "black", sounds brown instead - and how could Tolkien think that carnë means "red"? To me, the word sounds green!)

Tolkien has explained the basis of some of his predilections: "The element (n)dor 'land', probably owes something to say such names as Labrador (a name that might as far as style and structure goes be Sindarin)" (Letters:383-4). He also tells us how GON(O), GOND(O) got to be the Elvish root for "rock, stone" (as in Gondor "stone-land", Gondolin "Stone-song"): When he was eight years old, Tolkien read a book stating that nothing was known of the language of the pre-Celtic and pre-Roman tribes, except possibly ond "stone". Young John Ronald Reuel thought this word "fitted the meaning", so he remembered it and used it in his home-made languages many years later: Sindarin gond or gonn, Quenya ondo. (Letters:410. The book that provided Tolkien with the word ond was finally identified in Vinyar Tengwar #30: Celtic Britain by Professor John Rhys, that according to Carl F. Hostetter and Patrick Wynne "consists of over 300 densely-set pages and eschews neither etymological discussion, untranslated Latin passages, nor untransliterated Greek words". This was Tolkien's preferred reading at the age of eight.)

Many "Elvish" words turn out to be picked from a wide variety of sources: Pé "mouth" is Hebrew, lá "no, not" is Arabic, nér "man" from the reconstructed Indo-European language, ken- "see" is similar to Chinese kan, and roch "horse" is reminiscent of the Hebrew verb râkháv "ride". The stem ÑGAR(A)M "wolf" yields (besides Quenya narmo and Sindarin garaf) the Doriathrin word garm, Garm being one of the names of the monstrous Fenris-wolf haunting Norse mythology. Not just Old Norse, but also modern Scandinavian languages seem to be represented: Quenya varya "protect" is suspiciously similar to Norwegian verge, verje; "arrow" is pil in Scandinavian and pilin in Quenya, and while Quenya mat-/Sindarin medi means "eat", Norwegian/Swedish mat, Danish mad means "food"! Given the fact that one of the prime influences on Tolkien's languages was Finnish, we might also wonder if Quendi as a name of the Elves has something to do with kvener, an old Scandinavian name of the Finns. If there is any internal factor showing that Tolkien's languages are fictious, it must be the fact that some "plagiarism" can be detected in the vocabulary. But Tolkien willingly admitted that he did not try do avoid the influence of real-world languages. After all, he made languages for his own pleasure, not to fool others and make them think that they were "real".

When devising the few fragments of non-Elvish languages, such as Sauron's Black Speech and also the Adûnaic tongue (elaborated in structure but not in vocabulary), Tolkien was probably less reluctant to simply make words arbitrarily. Or so he thought. The Black Speech word nazg "ring" (as in Nazgûl, Ringwraith) seems to be an unconscious borrowing from Gaelic nasc of the same meaning (Letters p. 385). For a change the Black Speech was constructed to be as ugly as it possibly could be, and Tolkien did not like Gaelic (yet another example of his fine-tuned linguistic taste - except for native speakers, how many people are able to tell Gaelic and Welsh apart?)

Tolkien insisted that "all the names in the book, and the languages, are of course constructed, and not at random" (Letters:219). Yet there are a few "random" names. A note reproduced in The War of the Jewels p. 318 suggests that Tolkien did not know what the names Amloth and Ecthelion meant when he first used them, but since they "are well-sounding and have been in print", he took the time to find out what they meant. But the name Eöl proved too tough: "It isn't really absolutely necessary that names should be significant"! (The War of the Jewels p. 320.)

The questionable question of stability

However, Tolkien's languages changed in other ways than just the simulated changes within the imagined history. In The Monsters and the Critics p. 218-19, Tolkien observes that "if you construct your art-language on chosen principles", you can write poetry in that language - "in so far as you fix it, and courageously abide by your own rules, resisting the temptation of the supreme despot to alter them".

Tolkien did not courageously abide by his own rules. Tolkien did not resist the temptation of the supreme despot.

He never got his languages really "finished". The one thing that finally ensured total stability was his demise in 1973. In Sauron Defeated p. 240, Tolkien's character Lowdham speaks for Tolkien himself: "In making up a language you are free: too free. It is difficult to fit meaning to any given sound-pattern, and even more difficult to fit a sound-pattern to any given meaning. I say fit. I don't mean that you can't assign forms or meanings arbitrarily, as you will. Say, you want a word for sky. Well, call it jibberjabber, or anything else that comes into your head without the exercise of any linguistic taste or art. But that's code-making, not language-building. It is quite another matter to find a relationship, sound plus sense, that satisfies, that is when made durable. 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, fresh. You're liable to be for ever niggling, altering, refining, wavering, according to your linguistic mood and to your changes of taste."

This is precisely what Tolkien did himself. Throughout his life he kept revising, revising, revising. In the words of his son, "The linguistic histories were...invented by an inventor, who was free to change these histories as he was free to change the story of the world in which they took place, and he did so abundantly... Moreover, the alterations in the history were not confined to features of 'interior' linguistic development: the 'exterior' conception of the languages and their relations underwent change, even profound change" (LR:341-342).

Sindarin is a good example of changed ideas about the outer history of the languages. The scenario set out in the appendices to LotR is that this is the language of the Sindar, the Grey-elves - the Elves that came to Beleriand from Cuiviénen, but did not go over the sea to Valinor. But in Tolkien's pre-LotR notes, Sindarin is called Noldorin, and before that Gnomish, for this was the language of the Noldor or "Gnomes", the "Wise Elves". It was developed in Valinor, while Quenya in the former scenario was the language of the Lindar, the first of the three clans of the Eldar (to complicate matters even further, the Lindar were later renamed and became the Vanyar, while Lindar became a name of the third clan, the Teleri...) But then Tolkien must have realized that the Elves, immortal and all, would hardly develop radically different languages when they lived side by side in Valinor. So according to the revised scenario, both the Vanyar and the Noldor spoke Quenya with just minor dialectal differences, while the "Noldorin" language that Tolkien had already made was simply re-christened Sindarin, transferred from Valinor to Middle-earth and relocated to the mouths of the Grey-elves there. It was, of course, far more plausible that they had developed a language very different from Quenya, having been separated from their kin in Valinor for thousands of years. Christopher Tolkien comments, "So far-reaching was this reformation that the pre-existent linguistic structures themselves were moved into new historical relations and given new names" (LR:346).

But also the vocabulary, the phonology and the grammar of the languages were repeatedly revised. Consider these lines from an early "Qenya" poem, published in MC:213-14:

Man kiluva lómi sangane,
telume lungane
tollalinta ruste,
vea qalume,
mandu yáme,
aira móre ala tinwi
lante no lanta-mindon?

Man kenuva lumbor ahosta
Menel akúna
ruxal' ambonnar,
ëar amortala,
undume hákala,
enwina lúme elenillor pella
talta-taltala atalantië mindonnar?

Not just words, but even grammatical endings were subject to revision. In the Etymologies, there are quite a few examples of "Qenya" having a genitive in -n, e.g. Ar Manwen "Manwë's Day" (LR:368). But in the published LotR, -n has become the dative ending, while the genitive now ends in -o. Does the ending -o sound more "genitival" than -n does? One day, Tolkien must have decided just that.

Some words had their meaning totally reversed. We have learnt that the Avari are the Elves that refused to leave Cuiviénen and go to Valinor. But the Etymologies shows that Tolkien originally intended Avari as the name of the Elves that did go to Valinor! The name Fëanor existed at a very early stage, but it did not always mean "Spirit of Fire", as it is translated in the Silmarillion. In the Etymologies it is interpreted as "Radiant Sun", from earlier *Phay-anâro (LR:381). Before that, in the earliest word-lists, it meant "goblet-smith" (The Book of Lost Tales I, p. 253).

Even when something had appeared in print, Tolkien could not always resist the temptation to keep tampering. In the first edition of LotR, Frodo's greeting to Gildor was elen síla lúmenn' omentielmo. Later Tolkien decided that the last word should have been omentielvo instead, and this form was used in later editions. (One of the pioneers in the study of Elvish, Dick Plotz, was shocked to see the new form. He thought the American publishers, Ballantine, had made a mistake and prevailed upon them to correct it. In the next edition the publishers - necessarily incompentent in these matters - introduced the form omentilmo, that means nothing at all: Even honest efforts can have sad consequences!)

Nonetheless: the greatest changes and revisions undoubtedly happened before the mid-thirties. Concerning the original "Gnomish" language of 1915 or thereabouts, old Tolkien considered it merely a "language that ultimately became that of the type called Sindarin", and his earliest "Qenya" he now held to be "very primitive" (The Peoples of Middle-earth p. 379). But with the appearance of the Etymologies in the mid-thirties, the near-mature form of Q(u)enya and "Noldorin" = Sindarin came into place, and the remaining forty years of Tolkien's life were spent niggling over details.

Students, imitators, satirists and writers

How, then, do Tolkien's languages fare today, when a quarter of a century has passed since their maker went to the halls of Mandos? Some of us have embarked on the study of Elvish, perhaps with somewhat the same attitude as people enjoying a well-made crossword puzzle: The very fact that no real Elvish grammars written by Tolkien have been published makes it a fascinating challenge to "break the code". Or it may be pure romanticism, a special form of literary immersion: By studying the Eldarin languages, you try to get closer to - indeed into the heads of - the immortal Elves, fair and wise, the Firstborn of Eru Ilúvatar, teachers of mankind in its youth. Or, less romantically, you want to study the constructions of a talented linguist and the creative process of a genius engaged in his work of love. And many simply enjoy the Elvish languages as one might enjoy music, as elaborate and (according to the taste of many) gloriously successful experiments in euphony.

Whatever the motive of the student, the study is undoubtedly instructive: To describe Tolkien's languages properly, one has to get acquainted with much linguistic terminology. (The present writer would hardly have been intimately familiar with such terms and concepts as allative, ablative, locative, svarabhakti, assimilation, lenition and many more if I had not needed them in my study of Elvish. Once I impressed one of my lecturers with my knowledge of Welsh lenition patterns. How could she know that my examples were actually based on Sindarin?) It has also been suggested that some of Tolkien's insights as a linguist are buried in his languages, waiting to be unearthed. Modern Language Association International Bibliography deemed that the study of Elvish was sufficiently serious for them to register Vinyar Tengwar, the news letter of the Elvish Linguistic Fellowship, on their index.

Furthermore, it can easily be demonstrated that the nomenclature of The Lord of the Rings has inspired other fantasy writers - in this genre, names often have a distinctly Celtic or Welsh style. We can even find direct borrowings of morphemes. Reading such examples as Eriador, Gondor, Mordor etc. some have evidently gathered that the element -dor means "land", and in fantasy novels, you often encounter quite a few countries having names in -dor. Cf. for instance Alan Gardner's golden land of Elidor. There exists a Norwegian fantasy comic, Ridderne av Dor or "The Knights of Dor", that satirizes this phenomenon: countries have names like Kondor, Matador and Glassdor! Parodies indeed appeared already when Tolkien was alive; just consider this, ahem, version of A Elbereth Gilthoniel from Bored of the Rings. Prominent scholars of Elvish, including Arden R. Smith and Anthony Appleyard, have analyzed this text just as seriously as it does not deserve.

More serious attempts to write Elvish texts - mostly in verse - have also been published over the years. By now it would certainly be possible to put together a small anthology of such compositions. Hence, a small body of Elvish literature does exist today. Of course, there is no way of knowing what Tolkien would have thought of such newly written texts. It can hardly be doubted that if he ever returns from the dead, he will soon be busy with a red pencil.

But as Tolkien's papers are being published and our knowledge of Quenya and Sindarin becomes more complete - the gaps are still enormous - it might become possible to write long texts in Elvish. In her journal Tyalië Tyelelliéva, Lisa Star has boldly declared that "the ultimate goal is the revival of the Elvish languages for speaking, writing and art". Realistic or not, Tolkien does deserve it: A lifetime of work is left on the long road from Nevbosh to mature Quenya and Sindarin. It would be the final monument to Tolkien's efforts if his beloved languages could be brought to life - and indeed it would be the only fitting monument to a man who had to invent a whole world just to have a place where people could greet one another with the words Elen síla lúmenn' omentielvo.

Ardalambion Index