Wednesday, 23 March 2016

Tolkien and the aesthetics of philology

A talk given to the Oxford Tolkien Society, Taruithorn, on 6 March 2015 

1. Beauty in language and pleasure in language. 

I gave you a slightly fancy title because academic audiences expect fancy titles. I didn’t put a colon in the middle, which most academic titles have, because I really dislike them. An alternative title would have been ‘Tolkien and the delight of language’. But I did have a reason for using the word ‘philology’. ‘Language’ is too vague. Lots of people talk about pleasure or delight in language, for example:
Stylistics may be defined as the study of language as art rather than as basic communication. It has to do with the aesthetics of language rather than its basic function, with that ‘expressiveness which transcends the purely referential and communicative side of language’, with emotive overtones, with special emphasis, with calculated effects; concerned, as Stephen Ullmann has said, not with the elements of language as such, but with their expressive potential. [S. Ullmann, Language and Style, 1964, pp. 101, 111, quoted, and Peter Rickard, 1992. The French Language in the Seventeenth Century. Cambridge: D. S. Brewer. Introduction, p. 43.] 
This is ‘beauty in language’ in the sense not intended. Tolkien of course did write ‘with expressiveness’, ‘with emotive overtones, with special emphasis, with calculated effects’, none better, but that is not the side of aesthetics of language which I want to talk about.

I intended to concentrate on that notable piece ‘A Secret Vice’, and indeed I will. But looking at the volume in which it is published, The Monsters and the Critics and Other Essays, I realized how much relevant material there is in the two essays which flank that piece, ‘English and Welsh’ and ‘Valedictory Address to the University of Oxford’. A certain similarity between these three pieces is to be found in the fact that Tolkien, for varying reasons, ends each lecture with a series of illustrative snippets from other languages: Welsh, Old English, Elvish. These lectures tell us what Tolkien found delightful in other languages and provide a glimpse—a rather enigmatic one, in some ways—into his views on language. We find further clues in the features of the invented languages themselves.

When you open a book on linguistics or a grammar or dictionary of a particular language, the thing you won’t find is an enthusiastic encomium of language or of the language described therein. You may find something in a grammar of English about how it has become a valuable medium of communication worldwide, but this of course has little to do with its linguistic characteristics and everything to do with social, political, and economic factors. Sometimes you wouldn’t think, reading a book by a grammarian or linguist, that they had any actual enthusiasm for the language. Scientific objectivity rules out such expressions of personal partiality. But Tolkien’s lectures are peppered with expressions of warmth about languages. In ‘English and Welsh’ he speaks about:
A personal and if you will subjective expression of strong aesthetic pleasure in contact with Welsh, heard or read… [‘English and Welsh’ in The Monsters and the Critics, p. 190] 
In the course of two pages (190–1) he uses ‘pleasure’ in this context eight times. When we move to ‘A Secret Vice’, we find the same number of occurrences of this word, eight, over five pages (206–210). And he tells us that the highest pleasure is to be found in the art of inventing languages:
And it [sc. the linguistic faculty] is allied to a higher art of which I am speaking, and which perhaps I had better now define. An art for which life is not long enough, indeed: the construction of imaginary languages in full or outline for amusement, for the pleasure of the constructor or even conceivably of any critic that might occur. [‘A Secret Vice’ in The Monsters and the Critics, p. 202] 
Please note the astonishing claim. The construction of imaginary languages is a high art.

It’s important to notice how according to Tolkien the pleasure is associated with the mere sound of a language’s words—the phonetic elements, even without knowledge of their meaning:
The basic pleasure in the phonetic elements of a language and in the style of their patterns, and then in a higher dimension, pleasure in the association of these word-forms with meanings… [‘English and Welsh’ in The Monsters and the Critics, p. 190] 
The very word-form itself, of course, even unassociated with notions, is capable of giving pleasure—a perception of beauty, which if of a minor sort is not more foolish and irrational than being sensitive to the line of a hill, light and shade, or colour. [‘A Secret Vice’ in The Monsters and the Critics, p. 207] 


2. Fitness between sound and meaning. 

Moving on from the simple sound of language, as we have seen in one of the quotes above, Tolkien closely connected with it the idea of an association of word-forms with meanings, of fitness between sound and meaning. So, you start with a language whose choice and arrangement of sounds is pleasing to you in itself. But you pass directly from there to the connection between the arrangements of sequences of these sounds, which we call words, and the notions or ideas that they stand for in that language.

Writing in the context of language invention, Tolkien says:
This idea of using the linguistic faculty for amusement is however deeply interesting to me. I may be like an opium-smoker seeking a moral or medical or artistic defence for his habit. I don’t think so. The instinct for ‘linguistic invention’—the fitting of notion to oral symbol, and pleasure in the new relation established, is rational, and not perverted. In these invented languages the pleasure is more keen than it can be even in learning a new language—keen though it is to some people in that case—because more personal and fresh, more open to experiment of trial and error. And it is capable of developing into an art, with refinement of the construction of the symbol, and with greater nicety in the choice of notional-range. Certainly it is the contemplation of the relation between sound and notion which is a main source of pleasure. [‘A Secret Vice’ in The Monsters and the Critics, p. 206] 
So Tolkien found great pleasure in contemplating the relationship between the sound of words and what they meant. He made it clear that he wanted to investigate the associated questions:
No vocal noises mean anything in themselves. Meaning has to be attributed to them by a human mind. This may be done casually, often by accidental (non-linguistic) associations; or because of a feeling for ‘phonetic fitness’ and/or because of preferences in the individual for certain phonetic elements or combinations. The latter is naturally most evident in private invented languages, since it is one of their main objects, recognized or unconscious, to give effect to these likings. [Lett. 294, 8 February 1967]
I am personally most interested perhaps in word-form in itself, and in word-form in relation to meaning (so-called phonetic fitness) than in any other department. Of great interest to me is the attempt to disentangle—if possible—among the elements in this predilection and in this association (1) the personal from (2) the traditional. The two are doubtless much interwoven—the personal being possibly (though it is not proven) linked to the traditional in normal lives by heredity, as well as by the immediate and daily pressure of the traditional upon the person from earliest childhood. The personal, too, is doubtless divisible again into (a) what is peculiar to one individual, even when all the weighty influence of his native language and of other languages he has learnt in some degree has been accounted for; and (b) what is common to human beings, or to larger or smaller groups of them—both latent in individuals and expressed and operative in his own or any language. [‘A Secret Vice’ in The Monsters and the Critics, p. 211] 


3. The concept of the ‘native language’ 

I am sure that Tolkien had a generalized love of language, in the sense that language as a phenomenon was a principal and dominant interest of his. But the love he is speaking of here is a love of particular languages. He believes, for one thing, in an idea that I have not encountered anywhere else:
Language—and more so as expression than as communication [we need to return to this proviso later]—is a natural product of our humanity. But it is therefore also a product of our individuality. We each have our own personal linguistic potential: we each have a native language. But that is not the language that we speak, our cradle-tongue, the first-learned. Linguistically we all wear ready-made clothes, and our native language comes seldom to expression, save perhaps by pulling at the ready made till it sits a little easier. But though it may be buried, it is never wholly extinguished, and contact with other languages may stir it deeply. [‘English and Welsh’ in The Monsters and the Critics, 190] 
These tastes and predilections which are revealed to us in contact with languages not learned in infancy..are certainly significant: an aspect in linguistic terms of our individual natures. And since these are largely historical products, the predilections must be so too. [‘English and Welsh’ in The Monsters and the Critics, 194] 
He says the same in a letter of 1967:
It is these preferences, reflecting an individual’s innate linguistic taste, that I called his ‘native language’; though ‘native linguistic potential’ would have been more accurate, since it seldom comes to effect, even in modifying his ‘first-learnt’ language, that of his parents and country. [Lett. 294 8 February 1967] 

Whatever one thinks of this idea, it is quite difficult to place it in relation to current theories of language acquisition. Tolkien seems to agree with most modern theoreticians that the language ability is innate. I think (and I speak as no specialist) that where he is original is in asserting that aspects of a person’s linguistic potential are special to that person. We do not just have a passive ability to acquire whatever language is spoken around us in childhood. He thinks we have some kind of inner leaning or bias towards certain linguistic modes. Would he have seen this as analogous to the genetic determination of our features, in that we do not just have a genetic propensity to develop eyes, but we develop eyes of a particular colour and shape? Anyway, Tolkien thinks that the source of the pleasure that we feel in the sounds and meanings of a language are because that language is, as it were, calling to the innate language potential within us.

4. Sound and fitness. 

We have these two concepts, pleasure in the sound of language, and, closely connected, fitness of sound and meaning: both of prime importance to Tolkien and, as far as I know, neither of them much considered by anyone else. What can we learn about these from his writings and from the languages he invented?

Dealing first with the sound of languages. Let’s look at those whose mere sound gave Tolkien this special pleasure, and how they are reflected in the invented languages.

In ‘A Secret Vice’ he mentions:
Greek, Finnish, Welsh (to name at random languages which have a very characteristic and in their different ways beautiful word-form, readily seizable by the sensitive at first sight) are capable of producing this pleasure. [‘A Secret Vice’ in The Monsters and the Critics, p. 207] 
In ‘English and Welsh’ he gives us more. Latin, he says
‘seemed so normal that pleasure or distaste was equally inapplicable’. 
‘The fluidity of Greek, punctuated by hardness, and with its surface glitter, captivated did not touch home.’ 
‘Spanish..gave me strong pleasure’ (but he thinks the preservation of so much Latin in it was an ingredient in the pleasure—‘a historical and not purely aesthetic element’). 
Gothic ‘was the first to take me by storm, to move my heart…The contemplation of the vocabulary in A Primer of the Gothic Language was enough.’ 
‘Of all save one among them the most overwhelming pleasure was provided by Finnish, and I have never quite got over it.’ 
[‘English and Welsh’ in The Monsters and the Critics, 191–2] 
And finally, Welsh, about which of course he says a great deal in the essay ‘English and Welsh’. He tells us what appeals to him:
If I were pressed to give any example of a feature of this a source of pleasure to myself, I should mention the fondness for nasal consonants, especially the much-favoured n, and the frequency with which word-patterns are made with the soft and less sonorous w and the voiced spirants f and dd contrasted with the nasals: nant [stream], meddiant [possession], afon [river], llawenydd [joy], cenfigen [envy], gwanwyn [Spring], gwenyn [bees], crafanc [claw], to set down a few at random. A very characteristic word is gogoniant “glory”. [‘English and Welsh’ in The Monsters and the Critics, 194] 
Tolkien’s love of the phonetic shape of Welsh, as he tells us, is expressed in Sindarin. He says:
The names of persons and places in this story [The Lord of the Rings] 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. [‘English and Welsh’ in The Monsters and the Critics, footnote 33, p. 197] 
But of course many of those names, or the elements of which they are composed, were already in existence in pre-Lord of the Rings Sindarin (or in earlier nomenclature Noldorin or Gnomish). I think that the resemblance to Welsh is fully borne out by this bit of early Sindarin quoted in ‘A Secret Vice’, which is much earlier than anything in The Lord of the Rings:

Dir avosaith a gwaew hinar
engluid eryd argenaid
dir Tumledin hin Nebrachar
Yrch methail maethon magradhaid
Damrod dir hanach dalath benn
ven Sirion gar meilien
gail Luithien heb Eglavar
dir avosaith han Nebrachar.
[‘A Secret Vice’ in The Monsters and the Critics, p. 217] 

That is pretty like Welsh in sound: the omnipresent n, the frequent v and w and dh sounds, in particular. There are two major differences: the complete absence of the sound /ə/ (‘schwa’) which occurs a great deal in Welsh (in non-final syllables, written with the letter y), and the position of the stress accent on the first syllable of three-syllable words instead of always on the penultimate syllable (actually it’s slightly more complicated than that in Sindarin). I’d say that the lack of /ə/ and the position of the stress accent in both Elvish tongues owe much to the influence of Latin.

What about Tolkien’s other intoxication, Finnish? This is not nearly so closely replicated in Quenya, though there are certain aspects of it that resemble Finnish. In fact the older Quenya of the specimens given at the end of ‘A Secret Vice’ seems to me closer in sound to Finnish than the developed version of LR.

Take this verse in ‘A Secret Vice’:
Norolinde pirukendea
elle tande Nielikkilis,
tanya wende nieninqea
yar i vilya anta miqilis.
I oromandin eller tande
ar wingildin wilwarindeën,
losselie telerinwa
tálin paptalasselindeën.
[‘A Secret Vice’ in The Monsters and the Critics, p. 215–6] 
Compare it with this random snippet from the Kalevala:
Kun hän oli päässyt perille,
hän asettui istumaan
kiryavalle kivelle,
auringonpaisteiselle karille.
Kivi vierähti veteen,
kari vajosi pohjaan,
neito kiven mukana,
Aino karin kulmalla.
[Kalevala suomeksi runo 4, 319-327] 
Then, when she got there, / she sits herself down upon the bright boulder: / the rock plopped in the water / the boulder sank down / the maid with the rock / Aino beside the boulder [K. Bosley, The Kalevala, p. 48]

Quenya as seen in LR is more like Latin, but it resembles Finnish in the following ways:
  • the severe restriction of the voiced stops /b/, /d/, and /g/, the first of which occurs only after /m/, the second only after /l/, /n/, and /r/, and the third not at all, 
  •  the limitation on what consonants can be word-final (only /l/, /n/, /r/, /s/, and /t/) 
  •  the combination of the semivowel y with other consonants, as in, for example, ilye, máryat, tyeller, hyarmen (these last two being particularly original, and to me, especially attractive). 
In other ways it is completely different, for example in the combination of /w/ with consonants, as in ninque, vanwa, hwesta, and in featuring initial /ŋ/ and /ŋw/ (I like ngwalme ‘torment’) and the /ç/ sound before /t/ as in Telumehtar. And /k/ is far less common in Quenya than it is in Finnish.

Moving on to ‘fitness’: what did Tolkien mean by it? Do his invented languages shed light on this idea? In this connection I want to consider three areas of his invention: vocabulary, grammar, and historical development.

4.1. Vocabulary 

Firstly, it may be true that when inventing languages Tolkien chose word forms that were to some extent ‘phonaesthetic’. The idea of ‘phonaesthesis’ is that certain sounds or sound combinations symbolize basic concepts, or at least that they have ‘recognizable semantic associations due to recurrent appearance in words of similar meaning’ (Oxford Dictionary of English Grammar). A quite well-founded example in English is sn- symbolizing nose-related unpleasantness (snarl, sneer, sneeze, etc.). A less demonstrable one is the idea that the i-sound in combination with certain consonants often conveys small size, lightness, or insubstantiality (e.g. little, slim, thin, flimsy, mini-, etc.). Many phonaesthetic associations may be derived from the actual languages which one knows, and therefore culturally conditioned. 

Perhaps whether they are universal or culturally conditioned does not matter in Tolkien’s context. He was aiming to make languages that had a fitness in the world of ancient Northern Europe; their resonances need only come from this milieu. Tolkien, in one of his mentions of Finnish and Greek, actually says that they give him ‘phonaesthetic’ pleasure (Lett. 144); in other words, he does seem to have interpreted his ‘sense of fitness’ in terms of some pre-existing association of phoneme with meaning. 

The possibility that he regarded a simple association of sound and meaning, such as between the i-sound and small size or lightness, as a linguistic universal, is possible. Take the word lintë ‘swift’ in ‘Namárië’. It came to Tolkien long before the invention of Elvish and from what he says is unlikely to have a direct model in a known language: 
‘I can also remember [from the childhood invented language Nevbosh] the word lint “quick, clever, nimble”, and it is interesting, because I know it was adopted because the relation between the sounds lint and the idea proposed for association with them gave pleasure’ [‘A Secret Vice’ in The Monsters and the Critics, p. 205]. 
This word could be said to have ‘phonaesthetic’ qualities derived from the short i-sound and the initial l conveying quickness and lightness. Other inventions share this quality, such as wilwarin ‘butterfly’.

Another illuminating remark occurs in his translation of the short poem ‘Earendel’ in ‘A Secret Vice’ [in The Monsters and the Critics, p. 216]. The Qenya line is ‘i súru laustaner’ and Tolkien translates this as 
 The wind ‘lausted’ (not ‘roared’ or ‘rushed’ but made a windy noise)’ 
We can deduce from this that for Tolkien the verb lausta captured a particular idea of the sound of the wind that no English words do. 

Thirdly, speaking of Welsh, he writes in ‘English and Welsh’ [in The Monsters and the Critics, footnote 33, p. 193]: 
It is the ordinary words for ordinary things that in Welsh I find so pleasing. Nef may be no better than heaven, but wybren is more pleasing than sky. 
And he goes on a little later: 
This pleasure is felt most immediately and acutely in the moment of association: that is in the reception (or imagination) of a word-form which is felt to have a certain style, and the attribution to it of a meaning which is not received through it. 
Then he quotes a series of Welsh words that have this quality for him: 
adar, birds
alarch, swan
eryr, eagle
tân, fire
dwfr, water
awel, breeze
gwynt, wind
niwl, mist
glaw, rain
haul, sun
lloer, moon
sêr, stars
arglwydd, lord
gwas, servant
morwyn, maiden
dyn, man
cadarn, firm (adj.)
gwan, weak
caled, hard
meddal, soft
garw, rough
llyfn, smooth
llym, severe
swrth, heavy
glas, blue
melyn, yellow
brith, grey 
In view of these observations, what do we think were Tolkien’s sources when he set out to create the vocabularies of the Elvish languages? Well, there is considerable evidence that he drew on elements of real languages. In view of what he has said about personal predilection, there must have been particular words and forms in real languages that he found so well suited to their meanings that he drew on them. 

 In early Qenya and ‘Gnomish’ there are several words that resemble forms in Germanic. Examples: 
Qenya autë ‘prosperity’ (HME II. 336), compare the Germanic root auð- ‘riches’
Qenya maiwe gull (HME V. 373) compare Old English mǽw ‘gull’
Qenya mat- eat (HME V. 371: compare Gothic mats ‘food’
Qenya qalmë ‘death’, qalin ‘dead’ (HME I. 264) compare Old English cwealm ‘death’, cwelan ‘die’ – but perhaps also suggested by Finnish kuolla?
Qenya qet- ‘speak, talk’, Gnomish cwed- (HME II. 348), compare Gothic qiþan, Old English cweðan ‘speak’. 
But Tolkien seems to have drawn vocabulary items from several other languages that he liked. In ‘Tolkien’s Invented Languages’ we itemized as many as we could identify. 

For example, a few from Finnish: 
the root KULU- gold : Finnish kulta
tie ‘path’: Finnish tie ‘way’
the root TUL- ‘come, approach’: Finnish tulla ‘to come’
the root AN- ‘give’: Finnish anta ‘to give’, annan ‘I give’ 
Several apparently from Latin: 
Qenya lambë ‘tongue’ from the root LAVA- ‘lick’: Latin lambere ‘to lick’
Qenya ‘hand’: Latin manus ‘hand’
Qenya manë ‘good’: Latin manes
Qenya niqe ‘snow’: Latin nix, genitive nivis
Qenya talas ‘sole of the foot: Latin talus ‘ankle’
Qenya ‘went’, vand- ‘way, path’: Latin vado ‘I go’
the root ORO- ‘rise’: Latin orior ‘I rise’
the root OWO- and oa ‘wool’: Latin ovis sheep
the root UR- ‘be hot’: Latin uro ‘I burn’ 
 Perhaps from Greek: 
Sindarin orod mountain: Greek oros
Qenya erumë desert: Greek eremos 
Speaking for myself, much of Tolkien’s vocabulary does seem to me to have that quality of ‘fitness’. Why is this? In ‘Tolkien’s Invented Languages’ (in From Elvish to Klingon), we speculated that some Elvish words simply recall real-world words that are vaguely relevant, for example: 
Quenya carnë (Sindarin caran) ‘red’: carmine ‘crimson pigment’, carnation, or Latin caro, carn- ‘flesh’.
Sindarin minas ‘tower’: minaret or minatory (suggesting something looming above one).
Quenya mornë (Sindarin mor-) ‘black’: Greek mauros, English Moor(ish), perhaps even morose.
The root tin- ‘sparkle’ and its derivatives, such as tintilar ‘they twinkle’: twinkle or scintillate
In another discussion Tolkien does seem to imply that invented names, at least, are often derived from such pre-existing patterns: ‘One’s mind is, of course, stored with a “leaf-mould” of memories (submerged) of names, and these rise up to the surface at times, and may provide with modification the bases of “invented” names.’ (Lett. 324). 

Tolkien admitted only to a few such influences, such as the suffix –dor ‘country’ coming from Labrador, and discouraged the investigation of them. But one reason that I think they can be traced is that we know that originally some of the key names in the Legendarium were intended to resemble real-world names, because in its early stages there was meant to be a direct link. You’ll recall that early on, since Eressëa was England, Warwick – Welsh Caergwâr – was Kôr, Gnomish Gwâr, the harbour of Eressëa was Avallóne, resembling Avalon, and so on. Although this foundational element of the Legendarium was later set aside, it could not be totally expunged; but more importantly from the creative point of view, Tolkien went on creating language this way. It’s a kind of two-track or multi-track mode of linguistic thinking. An invented word looks in several directions at once. 

Let me give you one example. The Elvish words for ‘swan’ are Quenya alqua, Sindarin alph (plural eilph). These are related: the ancestor word alkwa became /alxwa/, then /alfa/, finally alph in Sindarin. In the document called the ‘Etymologies’, Tolkien gives the ultimate root as ALAK- ‘to rush’. So much for the invented story. What, in his own imagination, is Tolkien calling up? I think two unrelated words for ‘swan’ in different languages that he liked. The first is Old Norse alpt, probably from the same root as Latin albus ‘white’. The second is Welsh alarch, plural elyrch, from a British root *alarko, very similar (in sound) to his invented Elvish root. We know Tolkien liked this Welsh word, as he tells us so in ‘English and Welsh’ (see above). No word can come from two different roots, and anyway in the developed Legendarium it can’t be related to real-world languages at all. The link is in the realm of Tolkien’s linguistic taste. 

 4.2. Grammatical arrangements 

Clearly Tolkien’s invented languages are a success because they work grammatically. It’s like the difference between a completely static model village and one in which the model trains actually run round the tracks. This means that he had to have taken a very close interest in the way that grammar operates in real life languages. A technical, informed interest, not just a language user’s one.

Remarkably, in ‘A Secret Vice’ Tolkien speaks of the language inventor having 
 the same creative experience as that of those many unnamed geniuses who have invented the skilful bits of machinery in our traditional languages, for the use (and often the misunderstanding and abuse) of their less skilful fellows. [‘A Secret Vice’ in The Monsters and the Critics, p. 212] 
A throwaway line that seems to make an extraordinary assertion: that grammatical structures in real languages have been consciously created by skilled people just as they are in invented languages. It echoes the idea, which we shall discuss below, that linguistic change in Elvish is consciously directed. 

You probably remember the story of the little man in the camp: 
The man next to me said suddenly in a dreamy voice: ‘Yes, I think I shall express the accusative case by a prefix!’ [‘A Secret Vice’ in The Monsters and the Critics, p. 199] 
I think there is an important insight into Tolkien’s approach to grammatical structures when he says in the ‘Valedictory Address’, with reference to a completely different situation: 
The wide view, the masterly survey; plans and prophecies: these have never been in my line. I would always rather try to wring the juice out of a single sentence, or explore the implications of one word than try to sum up a period in a lecture, or pot a poet in a paragraph. [‘Valedictory Address to the University of Oxford’ in The Monsters and the Critics, p. 224] 
I think we can apply this to his attitude to language. Some people’s minds would move from particular instances to general rules or principles, and they would then concentrate on the general principles, and not be especially interested in or moved by their instantiations in particular grammatical forms. I believe that for Tolkien it was the other way round. He loved the individual form which resulted from the general law. In devising his languages, he would be concerned to introduce really good morphological, inflectional, and syntactic devices because of the pleasurable forms they would produce, not for their own sake. 

In his invented languages we can trace the kinds of features he liked. For example, Tolkien appreciated the initial mutation of consonants in Welsh: 
cath ‘cat’, ei gath ‘his cat’, ei chath ‘her cat’ 
a form of which he incorporated in Sindarin: 
lasto beth lammen ‘listen to the words of my tongue’: peth ‘word’
ernil i Pheriannath ‘prince of the halflings’: perian ‘halfling’ 
He also evidently loved vowel mutation, used especially for forming plurals in Welsh and Germanic, which he also incorporated into Sindarin: 
orod mountain plural ered
amon hill plural emyn
annon door plural ennyn
Perian halfling plural periain 
And like Welsh, Sindarin favours placing the modifier after what it modifies. We read ‘Im Narvi teithant i thîw hin’ ‘I Narvi carved these signs (literally the signs these)’ – one is irresistably reminded of the Welsh construction seen in y llyfrau hyn ‘these books’. 

In Quenya on the whole he displays more exotic linguistic tastes. The dominant idea in the grammar of Quenya is that of suffixation. Nouns take a range of inflectional suffixes. In outline these are suggestive of Finnish, although when you compare them in detail they are unlike. 
Endore-nna ‘to Middle-earth’ cf. Finnish talo-lle ‘to the house’, taloon ‘into the house’
Eare-llo ‘from the Great Sea’ cf. Finnish talo-lta ‘from the house’
ya-ssen ‘which-in’ cf. Finnish käde-ssä ‘in the hand’
alda-r-on ‘trees-of’ cf. Finnish talo-in ‘of the houses’, vuor-ten ‘of the mountains’ 
It has suffixes for possessive pronouns, as in atar-inya ‘my father’ cf. Finnish isä-ni ‘my father’. When there are both inflections and possessive pronouns, the pronouns precede the inflections: 
óma-ry-o ‘voice-her-of’
tie-lya-nna ‘path-your-on’ 
The Finnish arrangement is the opposite: 
 tie-llä-si ‘way-on-thy’ 
There isn’t time to go into the intricacies of Quenya verbal inflection, but I will just mention the remarkable system of first person plural morphology, which affects both verbs and nouns. In the first person plural (‘we’) there is a distinction between ‘inclusive’ use, whereby the speaker includes the person addressed, and ‘exclusive’ use, whereby only the speaker and anyone he speaks for are included, for example laituva-lme-t ‘we (exclusive) shall bless them’, omentielvo = omentie -lva -o ‘of our (inclusive) meeting’. This is an especially original idea: no European language this side of the Caucasus makes this distinction grammatically. 

In the first edition of The Lord of the Rings, the reading was omentielmo, because at the time –lma was the first person plural inclusive. In later editions it was changed to omentielvo, because the ending –lma was transferred to the first person plural exclusive and –lva became the new inclusive. (See notes to Lett. 205.) That Tolkien devised such a grammatical category shows that he was very well informed about grammatical systems worldwide, and although his invented languages are very European in some ways, he was open to influences from elsewhere. 

But what is primary: the ingenious grammatical feature, or the intriguing sound patterns it produces? Surely what is important is the ebb and flow of musical endings: 
 Et Eärello Endorenna utúlien
Sindanoriello caita mornie i falmalinnar imbe met
Andave laituvalmet
Nai tiruvantes i hárar mahalmassen mi Númen 
 The grammatical devices are there simply to produce this harmonious end result. 


4.3. Historical development 

English philology as popularly understood in the Oxford of Tolkien’s day was historical linguistics, regarded as having a rather narrow slant towards Anglo-Saxon sound changes. I don’t think Tolkien discussed this study much, but of course he practised it constantly. When he was an OED lexicographer he wrote etymologies connecting words to their possible roots; through an academic lifetime he continued to investigate the origin and development of word forms; and as a language inventor, of course, he constructed a whole glossary of Elvish words arranged under their roots, showing how the various forms developed in a range of half a dozen languages. 
 In inventing a language he mentions two contrasting processes of word form generation: 
 You may, for instance, construct a pseudo-historical background and deduce the form you have actually decided on from an antecedent and different form (conceived in outline); or you can posit certain tendencies of development and see what sort of form this will produce. [‘A Secret Vice’, in The Monsters and the Critics, p. 211–12]
Tolkien must have used both of them. He clearly at times, and certainly in the early stages of his languages, started from the ‘pseudo-historical background’, i.e. the Legendarium. He ‘decided on’ a form and he deduced it from an antecedent by running the ‘tendencies’—the sound laws and so forth—backwards. Supremely, as he shows (Lett. 297) in the case of the Old English name Éarendel, of uncertain meaning, which, in 1914, he adopted into the incipient Legendarium and adapted as the Qenya name Eärendil. Subsequently he reanalysed it as Eäre + ndil and generated the two roots *AYAR ‘sea’ and *(N)DIL- ‘to love’. 

It is very likely that this occurred in other cases, e.g. the name of Eärendil’s boat Wingilot, taken from Middle English and re-etymologized as from wingë ‘foam’ and lótë ‘flower’. He said ‘To me a name comes first and the story follows’ (Lett. 165) but since almost every name had a meaning, an etymology must also follow, whether from new roots invented for the purpose or conveniently fitting existing roots. Another example is the place name Rohan, which he admits to liking and borrowing as the Sindarin name of the Mark. It needed an etymology: roch ‘horse’ + -and ‘region’. This etymology necessitated two sound laws: medial –ch- is weakened to /h/, and final –nd is simplified to /nn/ and then /n/. 

Although he says relatively little about it, I think that the processes of historical sound change were a further source of Tolkien’s linguistic pleasure. If you catch the bug (or perhaps it is a gene), there is something inexplicably fascinating, and at times rapturous, about the way in which forms in one stage of a language morph into others or diversify into several in daughter languages. 

A highly original set of Elvish sound changes is the one he devised to explain the fact that in some words (series 1) Sindarin initial consonants b-, d-, g- = Quenya v-, l-, zero, while in others (series 2) Sindarin b-, d-, g- = Quenya m-, n-, ng-; e.g. 
(1) Sindarin Balan ‘Power’, Quenya Vala; Sindarin ‘night’, Quenya lómë; Sindarin galadh ‘tree’, Quenya alda.
 (2) Sindarin bar ‘dwelling’ (in Bar-en-Danwedh), Quenya mar (in Eldamar); Sindarin dûn ‘west’, Quenya númen; Sindarin Golodh, Quenya Ngoldo
Tolkien arranges that in Primitive Elvish, there are two initial consonantal series, (1) B, D, G, and (2) MB, ND, NG; they produce the same outcomes in Sindarin, but quite different ones in Quenya. Series 1 is not really surprising, but series 2 has no parallel in any well-known European language, and rather resembles a feature in some African languages (compare mbira, mbongo, Ndebele, ngege). The divergent development in each language is equally ingenious. Sindarin has ‘strong’ voiced initial consonants in both cases. Quenya ‘softens’ the first series into fricatives and the second into nasals. 

There is a huge array of sound-changes underlying the Elvish languages, some paralleling known language families (e.g. Sindarin vis a vis Welsh) and some entirely original. The ‘Etymologies’ list strings of cognate words exhibiting these sound changes. Tolkien surely enjoyed the devising of these sets of cognates, so absolutely like those in real world languages. Here, I think, was the heart of Tolkien’s aesthetic enjoyment of language. 


6. The making, moulding, and marring of languages. 

 The idea that a language can be devised by an agent and that its development can be influenced or directed by its speakers is recurrent within the Legendarium. 

 Most notably, in the case of the Elves. Although for the most part the development of the Elvish languages is described as if directed by impersonal linguistic changes, in ‘Dangweth Pengoloð’ (HME 12, pp. 395–402), Tolkien puts forward an original theory of language change in Elvish. The question it sets out to answer is why, if the Elves do not die and if their memories reach back to past ages, their languages do not remain unchanged. Part of the answer lies in ‘the changefulness of Eä (the created world)’. However, Pengoloð says, the Eldar also for aesthetic reasons alter the sounds of their speech from time to time, in a regular and consistent way. Because they know and are aware of the whole of their language at every moment, they will introduce a sound change throughout the language ‘as a weaver might change a thread from red to blue, either throughout his web, or in such parts thereof as were suitable to the new pattern, but not randomly here and there nor only in one corner’ (p. 399). 

After the shape of the Elvish languages had been fixed by the publication of LR, Tolkien spent much time trying to explain and reconcile words and forms that didn’t quite fit into the overall schema. For example, by making Quenya the ancient ‘Elven Latin’ and Sindarin the developed language of Middle-earth, he had a problem explaining why Primitive Elvish /þ/ remained in the latter but became /s/ in the former; it was the subject of an essay called ‘The Shibboleth of Fëanor’ (HME XII. 332–366). Here he writes: 
The change þ > s must therefore have been a conscious and deliberate change agreed to and accepted by a majority of the Ñoldor, however initiated, after the separation of their dwellings from the Vanyar… The change was a general one, based primarily on phonetic ‘taste’ and theory, but it had not yet become universal. It was attacked by the loremasters, who pointed out that the damage this merging would do in confusing stems and their derivatives that had been distinct in sound and sense had not yet been sufficiently considered. (HME XII. 332–3) 
It’s a long story, but essentially a rift occurs between the Noldor loyal to Fëanor, who maintain /th/ in memory of his mother Míriel Therindë in whose nickname the sound occurs, and those loyal to his father Finwë and his second wife Indis. Hence the use of /þ/ becomes a shibboleth, a test of loyalty, for Fëanor. 

 The key thing to notice that we have a sound change such as has occurred in real world languages    (/þ/ > /s/ has happened in both Latin American Spanish and Yiddish), but with the following differences: 
  1.  it is based on taste—essentially the same mechanism underlying one’s attraction to a language, according to Tolkien 
  2. it is consciously directed to all instances of the phoneme, and 
  3. the loremasters object to it on the ground that words will get confused. 
In the real world, as far as we know, sound changes arise randomly like evolutionary mutations and are not based on anyone’s predilections; they spread by unconscious mechanisms through the phonology of a language, though sometimes they do not reach every corner; and however much the learned, if they notice them at all, object, their establishment is completely irresistible. 

A degree of conscious control is also attributed to the Dwarves in their linguistic history: 
The father-tongue of the Dwarves Aulë himself devised for them… The Dwarves do not gladly teach their tongue to those of alien race; and in use they have made it harsh and intricate… Yet in secret they use their own speech only, and that (it is said) is slow to change; so that even their realms and houses that have been long and far sundered may to this day well understand one another. (HME XI. 205) 
In commenting on the brutality and ugliness of Orkish speech Tolkien points towards similar processes: 
The Orcs had a language of their own, devised for them by the Dark Lord of old, but it was so full of harsh and hideous sounds and vile words that other mouths found it difficult to compass, and few indeed were willing to make the attempt. And these creatures, being filled with all malice and hatred, so that they did not love even their own kind, had soon diversified their barbarous and unwritten speech into as many jargons as there were groups or settlements of Orcs. [‘The Appendix on Languages’ HME XII. 35]  
It is said that they had no language of their own, but took what they could of other tongues and perverted it to their own liking; yet they made only brutal jargons, scarcely sufficient even for their own needs, unless it was for curses and abuse. [LR Appendix F]  
Many indeed of the older tribes, such as those that still lingered in the North and the Misty Mountains, had long used the Westron as their native language, though in such a fashion as to make it hardly less unlovely than Orkish. [LR Appendix F]  
It is said that the Black Speech was devised by Sauron in the Dark Years, and that he had desired to make it the language of all those that served him, but he failed in that purpose. From the Black Speech, however, were derived many of the words that were in the Third Age wide-spread among the Orcs, such as ghâsh ‘fire’, but after the first overthrow of Sauron this language in its ancient form was forgotten by all but the Nazgûl. [LR Appendix F] 
The concepts here are worth examining. First there’s the idea that a being like Sauron can ‘devise’ a language. (Originally, he the Orcs’ language, the later scenario is that Sauron devised the Black Speech for all his servants, but failed to get it adopted by any but the Nazgûl). But equally, it seems that lesser beings can diversify the speech given to them, or take other languages and ‘pervert’ them. 

Then there is the idea that, at least when used by a race as debased and malicious as the orcs, a language, as regards its sound, could be ‘full of harsh and hideous sounds’, unpronounceable by other races, and, as regards its meaning, could be merely ‘barbarous’, a ‘brutal jargon’, ‘perverted’; a jargon that is ‘scarcely sufficient for their own needs’ – i.e. critically lacking in expressiveness, presumably in terms of both lexis and syntax. Useful only ‘for curses and abuse’. 


5. Language as music 

Reviewing these areas of language: the sound system alone and the sounds in relation to the vocabulary, the grammar, and the historical development, the thread that I discern running through it all is the primacy in Tolkien’s heart of the heard sound of a language; normally, as an attractive, pleasurable sound; but where malice intervenes, as a harsh, ugly sound: in fact, language as music. Take these two passages from LR
The singing now drew nearer. One clear voice rose now above the others. It was singing in the fair elven-tongue, of which Frodo knew only a little, and the others knew nothing. Yet the sound blending with the melody seemed to shape itself in their thought into words which they only partly understood. [LR I. iii. 79] 
At first the beauty of the melodies and of the interwoven words in elven-tongues, even though he understood them little, held him in a spell, as soon as he began to attend to them. Almost it seemed that the words took shape… [LR II. i. 233] 
In both cases the hearers have little or no literal comprehension of the speech they hear, but the sound of it works on their minds to produce shapes: words or visions. Again, in A Secret Vice’, after he has quoted several passages of verse in Elvish, he writes: 
 I may say that such fragments, nor even a constructed whole, do not satisfy all the instincts that go to make poetry. It is no part of this paper to plead that such inventions do so; but that they abstract certain of the pleasures of poetic composition (as far as I understand it), and sharpen them by making them more conscious. It is an attenuated emotion, but may be very piercing—this construction of sounds to give pleasure. The human phonetic system is a small-ranged instrument (compared with music as it has now become); yet it is an instrument, and a delicate one. [‘A Secret Vice’ in The Monsters and the Critics, p. 217–8] 
Very strikingly, Tolkien writes: 
The purely communicative aspect of language—the one that seems usually supposed to be the real germ and original impulse of language. But I doubt this exceedingly; as much as one doubts a poet’s sole object, even primary one, being to talk in a special way to other people. 
He granted that 
the communication factor has been very powerful in directing the development of language; but the more individual and personal factor—pleasure in articulate sound, and in the symbolic use of it, independent of communication though constantly in fact entangled with it—must not be forgotten for a moment. [‘A Secret Vice’ in The Monsters and the Critics, p. 208] 
The idea that the communicative aspect of language might not be the primary aspect one is quite extraordinary and fascinating. 


7. Conclusions… 

In modern linguistic theory, no language is inherently better than any other; the relationship between signifier and signified is arbitrary and the same idea can be expressed in innumerable phonetic forms, none apter than another; the development of language is not under the conscious control of speakers of the language, but is controlled by psychomotor and sociolinguistic factors of which they’re largely unaware; grammatical and syntactic usage is also largely unconscious and cannot be altered by so-called ‘rules’; and, of course, ‘giving pleasure’ is not a relevant category to the discussion of a language itself (as opposed to its use in discourse); you can’t say that ‘beauty’ is inherent in the forms or structures of a language. 

 In Tolkien’s writings, we read of the language learner being drawn by the beautiful sound of a language to his or her true ‘native language’. The sounds of a language elegantly and pleasingly fit on to the meanings of the words. A mind devises a language for its creatures to use. These language users vary their language according to their taste, in some cases to make it produce more beautiful sounds, in others to produce ugly, violent, and aggressive utterances. These are the scenarios we seem to visit in Tolkien’s writing about language and language making. Of course, he seems to be saying, language communicates; but this is a subordinate purpose. Rather, like the primeval gods in ‘The Music of the Ainur’, the ultimate purpose of the speaking peoples in using language is to make beautiful music.

Note that:

  1. This topic is investigated in Dr Dimitra Fimi’s book, Tolkien, Race and Cultural History (Palgrave Macmillan, 2008) (see  especially chapter 3: ‘“Linguistic Aesthetic”: Sounds, Meaning and the Pursuit of Beauty’), which I had not read before giving this talk.
  2. An extended edition of ‘A Secret Vice’, co-edited by Dr Dimitra Fimi and Dr Andrew Higgins, is due to be published in a few days (7th April). It includes a hitherto unpublished essay by Tolkien on ‘Phonetic Symbolism’.


Works by Tolkien: 
The Lord of the Rings. References are to ‘LR’ followed by the book and chapter number (i.e. books I to VI of the complete work, not the three named volumes). 
The History of Middle Earth (12 volumes). References are to ‘HME’ followed by volume and page. 
Letters of J. R. R. Tolkien. References in the text are to Lett., generally followed by the letter number rather than the page number. 
‘English and Welsh’, in The Monsters and the Critics 
‘A Secret Vice’, in The Monsters and the Critics 
‘Valedictory Address to the University of Oxford’, in The Monsters and the Critics 
The Monsters and the Critics and other Essays. Edited by Christopher Tolkien. London: George Allen & Unwin. 1983
Other works: 
K. Bosley 1989. The Kalevala. Oxford: World’s Classics. 
Risto Pottonen, ed. 2008. Kalevala suomeksi Helsinki: Books on Demand GmbH. 
Peter Rickard, 1992. The French Language in the Seventeenth Century. Cambridge: D. S. Brewer. 
E. S. C. Weiner and Jeremy Marshall ‘Tolkien’s Invented Languages’ in Michael Adams, 2011. From Elvish to Klingon. Oxford: OUP.


  1. Thanks for sharing this. We cover much the same ground - see my book, Tolkien, Race and Cultural History (Palgrave Macmillan, 2008), especially chapter 3: 'Linguistic Aesthetic’: Sounds, Meaning and the Pursuit of Beauty (section titles include: Theorizing language invention: A Secret Vice; The theory of ‘inherent linguistic predilections’: English and Welsh; Language attitudes; Sound symbolism and sound experiments). See here for the book’s TOC:

    I also thought you may like to know that an extended edition of A Secret Vice, co-edited by me and Dr Andrew Higgins, is due to be published in a few days (7th April). It includes a hitherto unpublished essay by Tolkien on "Phonetic Symbolism" (I reference it in my book, but I didn't have permission back then to quote extensively from it). See here for details:

    All the best,

    Dr Dimitra Fimi

    1. Thank you for your generous response. I did not know about 'Tolkien, Race and Cultural History' and will try to read it (and also mention it in the blog). I did know about the new edition of 'A Secret Vice' and it was the stimulus to my posting the article.

  2. Thank you for posting this -- it's a wonderful essay, and I've greatly enjoyed reading it. I thought this comment was particularly on the mark:

    'He loved the individual form which resulted from the general law. In devising his languages, he would be concerned to introduce really good morphological, inflectional, and syntactic devices because of the pleasurable forms they would produce, not for their own sake.'

    This dovetails with things I've been thinking about a lot lately, like the idea of 'congruence' Tolkien dabbled with to explain the similarity of Quenya Eärendil to OE Ēarendel, but not to earlier Germanic forms:

    One very minor point I'd quibble with is the suggestion that the phoneme /g/ doesn't occur at all in Quenya. It's actually precisely comparable to /b/, appearing only after a homorganic nasal in medial position -- rare, but definitely there in words like _anga_ 'iron', [aŋga]. This follows from Appendix E, where we're told that:

    'NG represents the ng in finger, except finally where it was sounded as in English sing. The latter sound [i.e. /ŋ/] also occurred initially in Quenya...' (TLotR, p. 1114)

    Tolkien is also pretty explicit in his private phonologies:

    'After nasals. In this position the stops remained as such, and mb, nd, ñg were among the most favoured consonantal groups of Quenya.' (Outline of Phonology, PE 19, 92; probably written within a few years of the Appendices)

    None of this really matters for your (excellent) main points, of course, but I thought it was at least worth mentioning, given your focus on phonological details and patterns here.

    1. Quite right, the remark about /g/ was careless. Thank you for the appreciative comments.

    2. Quite right, the remark about /g/ was careless. Thank you for the appreciative comments.

  3. This comment has been removed by the author.

  4. Mr. Weiner, Hello!

    I'm Tianluo_Qi, one of the members of a fan organization devoted to Tolkien Fandom in China. Via the shortlist of Tolkien Society Award given in their website, we got to know this article written by you. We found this one very inspiring, and want to translate it into Chinese, but not sure if we have the honor to do so.
    Because we are just a non-profit organization, and the platform we are trying to post this translation is a free one just like reddit, so we are not able to pay you any sort of royalty. Of course, we will stress that we are just translators and the original writer is you, and stick this blog under that article.
    If you are willing to, you can contact me via, looking forward to it.