--------------------------------------------------------------------------------

INTERNAL HISTORY
Quenya or High-elven is the most prominent language of the Amanya branch of the Elvish language family. In Aman there were two dialects of Quenya, Vanyarin and Noldorin. For historical reasons, only the latter was used in Middle-earth. The only other Eldarin language spoken in Aman, Telerin, could also be considered a dialect of Quenya, but it was usually held to be a separate language and is not discussed here (see separate article).

Compared to many other Elvish tongues, Quenya was archaic. It preserved the main features of the original Elvish language, invented by the Elves when they first awoke by the mere of Cuiviénen - a tongue with "many...beautiful words, and many cunning artifices of speech" (WJ:422). In fact, the Silmarillion Index refers to Quenya as "the ancient tongue, common to all Elves, in the form it took in Valinor" in Aman - as if Quenya was so similar to Primitive Elvish that it was merely as a later form of it, not a new language. Indeed Primitive Elvish and Quenya may have been mutually intelligible, but it must not be thought that they were almost identical. In Valinor, the ancient Elvish tongue underwent certain changes: "Its altering ...[came] in the making of new words (for things old and new) and in the softening and harmonizing of the sounds and patterns of the Quendian tongue to forms that seemed to the Noldor more beautiful" (WJ:20). The sounds b and d became v and l (or n) initially, final long vowels were shortened, unstressed medial vowels often disappeared, and many consonant clusters underwent metathesis or other changes, generally making them easier to pronounce. Quenya also adopted and adapted a few words from the language of the rulers of Aman - the Valar, the Angelic Powers guarding the world on behalf of its Creator. However, the Valar themselves encouraged the Elves to "make new words of their own style, or...translate the meaning of names into fair Eldarin forms" instead of retaining or adapting Valarin words (WJ:405). It is stated that the Noldor "were changeful in speech, for they had great love of words, and sought ever to find names more fit for all things that they knew or imagined" (Silm. ch. 5).

In Aman, Quenya was spoken not only by the Vanyar and the Noldor, but also by the Valar: "The Valar appear quickly to have adopted Quenya" after the arrival of the Elves, and their own tongue, Valarin, was not often heard by the Eldar: "Indeed it is said that often the Valar and Maiar might be heard speaking Quenya among themselves" (WJ:305). Pengolodh the sage of Gondolin notes: "In the histories the Valar are always presented as speaking Quenya in all circumstances. But this cannot proceed from translation by the Eldar, few of which knew Valarin. The translation must have been made by the Valar or Maiar themselves. Indeed those histories or legends that deal with times before the awakening of the Quendi, or with the uttermost past, or with things that the Eldar could not have known, must have been presented from the first in Quenya by the Valar or the Maiar when they instructed the Eldar." He mentions the Ainulindalë as an example: "It must...have been from the first presented to us not only in the words of Quenya, but also according to our modes of thought." Indeed even Melkor learnt Quenya, and learnt it well. "Alas," Pengolodh notes, "in Valinor Melkor used the Quenya with such mastery that all the Eldar were amazed, for his use could not be bettered, scarce equalled even, by the poets and the loremasters." (VT39:27)

When Rúmil invented letters, Quenya became the first language to be recorded in writing (Silm. ch. 6, LotR Appendix F). But outside the Blessed Realm of Aman, Quenya would never have been known if it had not been for the rebellion of the Noldor in the First Age. Most of this clan left Aman and went into exile in Middle-earth, bringing the High-elven tongue with them. In Middle-earth the Noldor were greatly outnumbered by the native Sindar or Grey-elves, who spoke a clearly related, yet quite different language. The Sindarin tongue had long since dropped the case inflections that were still preserved in Quenya, and the general sound of the two languages differed much - Quenya was much more vocalic than Sindarin and had a very limited distribution of the voiced stops b, d, g, that were frequent in Sindarin. As it turned out, "the Noldor...learned swiftly the speech of Beleriand [i.e. Sindarin], whereas the Sindar were slow to master the tongue of Valinor [i.e. Quenya]". Twenty years after the coming of the Noldor to Middle-earth, "the tongue of the Grey-elves was most spoken even by the Noldor" (Silm. ch. 13). When King Thingol of Doriath finally learnt that the Noldor had killed many of his kinsfolk among the Teleri and stolen their ships when they left Valinor, he banned the use of Quenya throughout his realm. Consequently, "the Exiles took the Sindarin tongue in all their daily uses, and the High Speech of the West was spoken only by the lords of the Noldor among themselves. Yet that speech lived ever as a language of lore, wherever any of that people dwelt" (Silm. ch. 15).

Thus Quenya did survive, even in the dark First Age. In fact, the vocabulary was expanded: The Noldor adopted and adapted some words from other tongues, such as Casar "dwarf" from Dwarvish Khazad and certa "rune" from Sindarin certh (WJ:388, 396). Some words already in use developed new or modified meanings in Exilic Quenya, such as urco, a word that in Valinorean Quenya was used of "anything that caused fear to the Elves, any dubious shape or shadow, or prowling creature" that was remembered in ancient tales of the March from Cuiviénen. In Exilic Quenya, however, urco was recognized as a cognate of Sindarin orch and was used to translate it; hence the meaning of urco was now simply "Orc" (WJ:390; the Sindarin-influenced form orco was also used). When the Edain arrived in Beleriand, they learned not only Sindarin, but "to a certain extent Quenya also" (WJ:410). Though Quenya "was never a spoken language among Men" (Plotz Letter), High-elven names like Elendil became popular among the Edain. Túrin gave himself the Quenya name Turambar or "Master of Doom", and his sister Nienor cried some High-elven words before she killed herself (Silm. ch. 21).

There are also numerous examples of Quenya being used or remembered by the Noldorin Exiles themselves: When Turgon built his hidden city, "he appointed its name to be Ondolindë in the speech of the Elves of Valinor", though the Sindarin-adapted form Gondolin became the usual name of the city. Even in Gondolin, Quenya "had become a language of books" for most people, "and as the other Noldor they used Sindarin in daily speech". Nonetheless, Tuor heard the Guard of Gondolin speak "in the High Speech of the Noldor, which he knew not". It is also stated that "Quenya was in daily use in Turgon's house, and was the childhood speech of Eärendil" (UT:44, 55). PM:348 confirms that "Turgon after his foundation of the secret city of Gondolin had re-established Quenya as the daily speech of his household". Aredhel left Gondolin and was captured by Eöl, to whom she bore a son, and "in her heart she gave him a name in the forbidden tongue of the Noldor, Lómion, that signifies Child of the Twilight" (Silm ch. 16). Eöl later called his son by the Sindarin name Maeglin, but Aredhel "taught Maeglin the Quenya tongue, though Eöl had forbidden it" (WJ:337).

However, Quenya as spoken by the Exiles early underwent some minor changes, probably before Thingol's edict against its use largely froze all processes of linguistic change. In a letter to Dick Plotz, Tolkien described the noun declension of an old form of Quenya, so-called "Book Quenya". Tolkien writes that "as far as was known to [mortal] men - to the Númenorean scholars, and such of these as survived in Gondor in [the Third Age] - these were the forms used in writing". But he further notes: "Quenya as a spoken language had changed to a certain extent among the Noldor before it ceased to be a birth tongue [i.e. early in their exile]... In this 'colloquial' form it continued to be spoken among Elves of Noldorin origin, but was preserved from further change since it was learned anew from writing by each generation." The implication seems to be that also this "colloquial" form of Quenya could be used in writing, and that this was the Quenya of the writings from which each generation learned the language anew. These would be writings written by the Noldor during their exile, after their language had diverged slightly from Amanian Quenya (in particular by the loss of the accusative case): "Exilic conditions...made necessary the writing down anew from memory of many of the pre-Exilic works of lore and song" (PM:332). The Númenorean scholars may have picked up a more archaic form of Quenya because they were in contact with the Eldar of Eressëa and Valinor, not only the Noldorin Exiles in Middle-earth. Today, most writers do not use Book Quenya, but the Exilic Noldorin form of High-Elven, the language of Galadriel's Lament (LotR1/II ch. 8).

The First Age ended in the War of Wrath. At the beginning of the Second Age, some of the Noldor returned to Aman, "but some lingered many an age in Middle-earth" (Silm. ch. 24). Thus, native speakers of Quenya were still present in the Hither Lands. Indeed even their greatest enemy made a Quenya name for himself when he appeared to the Elves in a fair form to deceive them: Annatar, the Lord of Gifts (Of the Rings of Power in Silm). His real name was also Quenya, but one may well understand that he did not like it: Sauron, the Abhorred (see Silm. Index). Later, the Smiths of Eregion gave Quenya names to their greatest works: Narya, Nenya, and Vilya, the greatest of the Rings of Power save the One Ring itself.

However, the history of the Second Age is dominated by the saga of Númenor, the great isle given to the Edain by the Valar. Originally all the Edain were Elf-friends, and most of them knew Sindarin (though the daily speech of the Númenóreans was Adûnaic, a Mannish tongue). We are told that "the lore-masters among them learned also the High Eldarin tongue of the Blessed Realm, in which much story and song was preserved from the beginning of the world... So it came to pass that, beside their own names, all the lords of the Númenóreans had also Eldarin [Quenya and/or Sindarin] names; and the like with the cities and fair places that they founded in Númenor and on the shores of the Hither Lands" (Akallabêth). Examples of Quenya names in Númenor include Meneltarma, Armenelos, Rómenna and the name Númenor itself. Yet the fact remained that "Quenya was not a spoken tongue in Númenor. It was known only to the learned and to the families of high descent, to whom it was taught in their early youth. It was used in official documents intended for preservation, such as the Laws, and the Scroll and the Annals of the Kings..., and often in more recondite works of lore. It was also largely used in nomenclature: the official names of all places, regions, and geographical features in the land were of Quenya form (though they usually had also local names, generally of the same meaning, in either Sindarin or Adúnaic [Númenorean]. The personal names, and especially the official and public names, of all members of the royal house, and of the Line of Elros in general, were given in Quenya form" (UT:216). The Kings took Quenya names because High-Elven was "the noblest tongue in the world" (UT:218). However, times would change.

The Númenoreans began to envy the immortality of the Elves, and the friendship with Aman gradually became cold. When the twentieth King of Númenor ascended the throne in the year 2899 of the Second Age, he broke with the ancient custom and took the sceptre with a title in Adûnaic instead of Quenya: Ar-Adûnakhôr, Lord of the West. In his reign "the Elven-tongues were no longer used, nor permitted to be taught, but were maintained in secret by the Faithful; and the ships from Eressëa came seldom and secretly to the west shores of Númenor thereafter" (UT:222). In 3102 Ar-Gimilzôr became the twenty-third King, and "he forbade utterly the use of the Eldarin tongues, and would not permit any of the Eldar to come to the land, and punished those that welcomed them" (UT:223). Indeed "the Elvish tongues were proscribed by the rebel Kings, and Adûnaic alone was permitted to be used, and many of the ancient books in Quenya or in Sindarin were destroyed" (PM:315).

However, Gimilzôr's son Inziladûn proved to be of a very different character when he became King in 3177 (or 3175 according to one source - see UT:227). He repented of the ways of the Kings before him and took a title in Quenya according to the ancient custom: Tar-Palantir, the Far-Sighted. Tar-Palantir "would fain have returned to the friendship of the Eldar and the Lords of the West", but it was too late (UT:223). His only child, a daughter, he named Míriel in Quenya. She should have been Ruling Queen after his death in 3255, but she was forced to marry Pharazôn, son of Tar-Palantir's brother Gimilkhâd. Pharazôn took her to wife against her will to usurp the sceptre of Númenor. Evidently he could not stand her Quenya name and changed it to Zimraphel in Adûnaic. Proud and arrogant, Ar-Pharazôn challenged Sauron in Middle-earth. The evil Maia cunningly pretended to surrender, whereupon Pharazôn "in the folly of his pride carried him back as a prisoner to Númenor. It was not long before he had bewitched the King and was master of his counsel; and soon he had turned the heart of all the Númenoreans, except the remnant of the Faithful, back towards the darkness" (LotR Appendix A). Sauron made the King believe that he would become immortal if he managed to wrest the rulership of Aman from the Valar, and eventually Pharazôn did attempt to invade the Blessed Realm. As Sauron well knew, the Númenoreans could never conquer the Powers, and as he had foreseen, Pharazôn's armada was utterly defeated. However, Sauron had not foreseen that the Valar would call upon the One Himself, and that He would use His power to change the entire shape of the world. The Blessed Realm was removed from the visible world into the realm of hidden things, and with it went all native speakers of Quenya save those of the Noldor who lingered in Middle-earth. Númenor itself disappeared in the sea, and we shall never know the number of the books written in Quenya that were lost in the ruin of the Isle of the Kings. The sunken isle was given new High-elven names: Mar-nu-Falmar, Land (lit. Home) under Waves, and Atalantë, the Downfallen.

The only survivors of the Downfall were Elendil, Isildur, Anárion and those who followed them on their ships. As their Quenya names give away, they were Elf-friends and had no part in the rebellion against the Valar. In Middle-earth they founded the Realms in Exile, Arnor and Gondor. Sauron soon attacked Gondor, but he was defeated in the Battle of Dagorlad, and after seven years of siege he had to leave the Barad-dûr and was slain by Gil-galad, Elendil, and Isildur; only the last of these survived. So ended the Second Age of the World, but the Realms in Exile survived into the Third Age, and among the scholars of Arnor and Gondor the knowledge of Quenya was preserved.

The Kings of Arnor and Gondor used Quenya names, as had the faithful Númenorean Kings of old. (861 years into the Third Age, however, Arnor was divided into the petty realms of Arthedain, Rhudaur, and Cardolan; the Kings of these realms used Sindarin names.) The Stewards of Gondor also used Quenya names until the time of Mardil, the first of the Ruling Stewards (so called because there was no King in Gondor in the period 2050-3019 of the Third Age, and the Stewards had to take on all responsibilities). However, Mardil's successors ceased to use High-elven names. The Stewards never took the title of King, and they may have thought it would be presumptuous to use Quenya names after the manner of the Kings. But when Aragorn was crowned King in 3019, he called himself Elessar Telcontar in Quenya, following the ancient custom. Then the Fourth Age began, and the last of the Noldor set sail from the Havens and left Middle-earth for ever, returning to Aman. The last native speakers of Quenya were gone from our world, but as Gandalf pointed out to Aragorn, it was his task "to preserve what may be preserved" (LotR3/VI ch. 5) - including the knowledge of the Eldarin languages. We know that Aragorn gave a High-elven name to his son Eldarion, who succeeded him on the throne of Gondor when he died in the year 120 of the Fourth Age. Though little is known about this Age, there can be little doubt that as long as the realm of Gondor endured, Quenya was remembered.

Designations of the Language
The word Quenya, in the Vanyarin dialect Quendya, is an adjective formed upon the same stem as Quendi "Elves"; the basic meaning is thus "Elvish, Quendian". But the word Quenya was also associated with the stem quet- "speak", and indeed the stems quet- and quen- may be related: Tolkien speculated that "the oldest form of this stem referring to vocal speech was *KWE, of which *KWENE and *KWETE were elaborations" (WJ:392). The Elvish loremasters held that Quendi meant "those who speak with voices", and according to Pengolodh, Quenya meant properly "language, speech" (WJ:393). However, this may simply reflect the fact that Quenya was the only known language when the adjective Quen(d)ya "Quendian" was first applied to Elvish speech (elliptically for Quenya lambë "Quendian tongue"). Later the word Quenya was used exclusively as a name of this language, not as a general adjective meaning "Elvish, Quendian". The Noldor, however, "did not forget its connection with the old word Quendi, and still regarded the name as implying 'Elvish', that is the chief Elvish tongue, the noblest, and the one most nearly preserving the ancient character of Elvish speech" (WJ:374).

Quenya is also called parmalambë "the book-tongue" and tarquesta "high-speech" (LR:172; cf. "the High Speech of the Noldor" in UT:44). As Quenya originated in Valinor, it could also be termed Valinorean (LotR3/V ch. 8) or "the speech of the Elves of Valinor" (Silm. ch. 15). After the end of the First Age, many Noldor dwelt on the isle of Tol Eressëa, near to the coast of Aman. Therefore, Quenya is also known as Eressëan, or Avallonian after the Eresseän city of Avallónë (LR:41, SD:241). To the Amanian Teleri, Quenya was Goldórin or Goldolambe, evidently meaning "Noldoic" and "Noldo-tongue", respectively (WJ:375). In Gnomish, Tolkien's first attempt to reconstruct the language that much later turned out to be Sindarin, the word for Quenya ("Qenya") was Cweneglin or Cwedhrin, but these words are certainly not valid in mature Sindarin (Parma Eldalamberon No. 11 p. 28).The Elf Glorfindel referred to Quenya as "the Ancient Tongue" (LotR1/I ch. 3), and being the most prestigious language in the world, it is also called "the High Speech of the West", "the high Eldarin tongue" (Silm. ch. 15, Akallabêth) or "High Ancient Elven" (WR:160). By the Númenóreans, Quenya was called Nimriyê or "Nimrian tongue", as the Dúnedain called the Elves Nimîr, the Beautiful. (SD:414, cf. WJ:386). Later, Frodo referred to Quenya as "the ancient tongue of the Elves beyond the Sea" and "the language...of Elven-song". (LotR1/II ch. 8) In English, Tolkien also used designations like "High-elven" (occasionally in Letters: "High-elvish") and "Elf-Latin, Elven-Latin" (Letters p. 176). In Middle-earth, Quenya eventually became a language of ceremony and lore, so Tolkien deemed it comparable to Latin in Europe.

EXTERNAL HISTORY
Quenya, originally spelt "Qenya", goes back to at least 1915. It seems that it was this year 23-year-old Tolkien compiled the "Qenya Lexicon", one of the very first Elvish word-lists (see LT1:246). Countless revisions affecting both grammar and vocabularly separate the earliest "Qenya" from the more-or-less final form that is exemplified in The Lord of the Rings, but the general phonetic style was present from the beginning. Nearly mature Quenya gradually emerged in the thirties, but minor revisions were being done even while LotR was being written, such as changing the genitive ending from -n to -o. There are also a few changes in the revised second edition of LotR, like when Tolkien decided that the word vánier in Galadriel's Lament should rather be avánier.

Throughout his life, Tolkien continued to refine the High-Elven tongue, that according to his son Christopher was "language as he wanted it, the language of his heart" (from the TV program J.R.R. Tolkien - A Portrait by Landseer Productions). In one of his letters, Tolkien himself wrote: "The archaic language of lore is meant to be a kind of 'Elven-latin', and by transcribing it into a spelling closely resembling that of Latin...the similarity to Latin has been increased ocularly. Actually it might be said to be composed on a Latin basis with two other (main) ingredients that happen to give me 'phonaesthetic' pleasure: Finnish and Greek. It is however less consonantal than any of the three. This language is High-elven or in its own terms Quenya (Elvish)" (Letters:176). Quenya was the ultimate experiment in euphony and phonaesthetics, and according to the taste of many, it was a glorious success. The grammatical structure, involving a large number of cases and other inflections, is clearly inspired by Latin and Finnish.

The longest sample of Quenya in The Lord of the Rings is Galadriel's Lament, sc. the poem Namárië near the end of the chapter Farewell to Lórien (LotR1/II ch. 8, starting Ai! laurië lantar lassi súrinen...) Many of the examples referred to in the following discussion are drawn from this poem. Other important Quenya texts include the Markirya poem in MC:222-223 and Fíriel's Song in LR:72, though the grammar of the latter differs somewhat from LotR-style Quenya; it represents one of Tolkien's earlier "Qenya" variants. (Markirya is very late and fully reliable.)

THE STRUCTURE OF QUENYA: A BRIEF SURVEY
Elementary Phonology
Quenya has five vowels, a, e, i, o, u, short and long; the long vowels are marked with an accent: á, é, í, ó, ú. The vowel a is extremely frequent. The quality of the vowels resembles the system in Spanish or Italian rather than English. To clarify the pronunciation for readers used to English orthography, Tolkien sometimes adds a diaeresis over some vowels (e.g. Manwë rather than Manwe to indicate that the final e is not silent, or Eärendil to indicate that the vowels e and a are pronounced separately and not drawn together as in English ear - the dots are not necessary for the meaning and can safely be left out in e-mail). The diphthongs are ai, au, oi, ui, eu, iu. (A seventh diphthong ei seems to occur in one or two words, but its status is uncertain.) The consonants are for the most part the same as in English, with the sibilants as the main exception: Ch as in church does not occur, neither does j as in joy, and instead of sh, zh (the latter like s in pleasure), Quenya has a sound like the German ich-Laut, spelt hy by Tolkien (e.g. hyarmen "south"). The h of English huge, human is sometimes pronounced as a weak variant of the sound in question. Quenya also lacks th (unvoiced as in thing or voiced as in the); unvoiced th did occur at an earlier stage, but merged with s shortly before the rebellion of the Noldor (see PM:331-333). It should also be noted that the voiced plosives b, d, g only occur in the clusters mb, nd/ld/rd and ng (some varieties of Quenya also had lb instead of lv). There are no initial consonant clusters, except qu (= cw), ty, ny and nw if we count the semi-vowels y, w as consonants. Normally there are no final clusters either; words end either in one of the single consonants t, s, n, l, r or in a vowel, more often the latter. Medially between vowels, a limited number of consonant clusters may occur; those described by Tolkien as "frequent" or "favoured" are in italics: cc, ht, hty, lc, ld, ll, lm, lp, lqu, lt, lv, lw, ly, mb, mm, mn, mp, my, nc, nd, ng, ngw, nn, nqu, nt, nty, nw, ny, ps, pt, qu (for cw), rc, rd, rm, rn, rqu, rr, rt, rty, rs, rw, ry, sc, squ, ss, st, sty, sw, ts, tt, tw, ty, x (for ks). A few other combinations may occur in compounds. Quenya phonology is quite restrictive, giving the language a clearly defined style and flavour.

Note that in Quenya spelling, the letter c is always pronounced k (so cirya "ship" = kirya). Tolkien was inconsistent about this; in many sources the letter k is used, but in LotR he decided to spell Quenya as similar to Latin as possible. In some cases, k in the sources has been regularized to c in the following discussion.

The Noun
The Quenya noun is inflected for nine or ten cases. (There are also four numbers, but we will for the most part stick to the singular while listing the cases.) The learner should not be daunted by the large number of cases. Where English uses a preposition in front of a noun, Quenya often prefers to add an ending to the noun instead; there is very little more to it.

The nominative singular is the basic, uninflected form of the noun; it has no special ending. The typical function of a nominative noun is to be the subject of a verb, like lómë "night" or aurë "day" in the cries heard before and during the Nirnaeth Arnoediad: Auta i lómë! "The night is passing!" Aurë entuluva! "Day shall come again!" (Silmarillion chapter 20).

Quenya as spoken in Valinor had an accusative that was formed by lengthening the final vowel of the noun: cirya "ship" (nominative), ciryá "ship" (accusative). Nouns ending in a consonant presumably had no distinct accusative. In the plural, even nouns ending in a vowel had the ending i, e.g. ciryai "ships" (nominative ciryar). The function of the accusative was primarily to mark that the noun was the object of a verb; we have no examples but may construct one like haryan ciryá, "I have a ship" (haryan ciryai "I have [several] ships"). But in Middle-earth, the distinct accusative case disappeared from the speech of the Noldor (such things happen when you are busy fighting Orcs, Balrogs, and Dragons), and the nominative took over its former functions. So from now on it was OK to say haryan cirya, haryan ciryar. Modern writers seem never to use the distinct accusative.

The genitive has the ending -o, generally corresponding to the English ending 's, though a Quenya genitive is often better rendered by an of-construction in English. One example from Namárië is Vardo tellumar "Varda's domes" or "(the) domes of Varda". Notice that the ending -o displaces final -a, hence Vardo, not Vardao - but most other vowels (it seems) are not displaced: In MR:329 we find Eruo for "the One's, Eru's". (If the noun ends in -o already, the ending becomes "invisible"; normally the context will indicate that the noun is a genitive and not a nominative. An attested example is Indis i Ciryamo "the mariner's wife"; cf. ciryamo "mariner".) Infrequently the genitive carries the meaning "from", cf. Oiolossëo "from Mount Everwhite, from Oiolossë" in Namárië - but this is usually expressed by means of the ablative case instead (see below). The plural genitive ending is -on, that may be observed in the title Silmarillion, "of the Silmarils", the complete phrase being Quenta Silmarillion, "(the) Story of (the) Silmarils". An example from Namárië is rámar aldaron, "wings of trees", a poetic circumlocution for leaves. The ending -on is added, not to the simplest form of the noun, but to the nominative plural. So though "tree" is alda, "of trees" is not **aldon, but aldaron because the nominative plural "trees" is aldar. Cf. also Silmaril, plural Silmarilli, genitive Silmarillion. (The doubling of the final l of Silmaril before an ending is an example of stem variation; some stems change slightly when an ending is added, often reflecting an older form of the noun.)

Then there is the possessive, by some called the "associative" or "adjectival case"; Tolkien himself speaks of it as a "possessive-adjectival...genitive" in WJ:369. This case has the ending -va (-wa on nouns ending in a consonant). Its general function is like the English genitive, to express ownership: Mindon Eldaliéva "Tower of the Eldalië". The function of the possessive was long poorly understood. In Namárië it occurs in the phrase yuldar...miruvóreva, "draughts...of mead". This one example, that for more than twenty years was the only one we had, made many conclude that the function of this case was to show what something is composed of - indeed the case itself was called "compositive". Luckily, The War of the Jewels p. 368-369 finally gave us Tolkien's own explanation of the more normal functions of this case, and how it differs from the genitive. The possessive can, as already stated, denote possession or ownership. Tolkien gives the example róma Oroméva, "Oromë's horn", used of a horn that belonged/belongs to Oromë at the time that is being narrated (past or present). Genitive róma Oromëo would also translate as "Oromë's horn", but properly it would mean "a horn coming from Oromë", implying that the horn had left Oromë's possession at the time that is being narrated. However, the genitive intruded on the functions of the possessive in later ages. Cf. genitive Vardo tellumar, not possessive *tellumar Vardava, for "Varda's domes" in Namárië (if the genitive does not imply that the domes originated with Varda rather than that she owns them).

The dative has the ending -n. This ending generally translates as the preposition "for" or "to"; the dative pronoun nin "for me" (from ni "I") is found in Namárië: Sí man i yulma nin enquantuva? "Who now will refill the cup for me?" Often the dative corresponds to an indirect object in English: *I nís antanë i hínan anna, "the woman gave the child a gift" (gave a gift to the child).

The locative has the ending -ssë, that carries the meaning "on" or "in". In the Tengwar version of Namárië that is found in RGEO, the poem has the superscript Altariello Nainië Lóriendessë, "Galadriel's Lament in Lóriendë (Lórien)". In the plural, this ending has the form -ssen, seen in the word mahalmassen "on thrones" in UT:305 cf. 317 (mahalma "throne"). This ending also occurs on the relative pronoun ya in Namárië: yassen "wherein, in which" (Vardo tellumar...yassen tintilar i eleni, *"Varda's domes...in which the stars tremble"). Refering back to a singular word, "in which" would presumably be yassë. The use of case endings rather than prepositions to express "in, from, to, with" (cf. the next paragraphs) is a characteristic feature of Quenya grammar.

The ablative has the ending -llo, that carries the meaning "from" or "out of". An example from Namárië is sindanóriello, "out of a grey country" (sinda-nórie-llo: "grey-country-from"). There is also the word Rómello, *"from (the) East", contraction of *Rómenello (Rómen "[the] East"). Cf. also the word Ondolindello "from Ondolindë (Gondolin)" in J. R. R. Tolkien - Artist and Illustrator p. 193.

The allative has the ending -nna, meaning "to", "into" or "upon". Both the ablative and the allative are exemplified in the words spoken by Elendil when he came to Middle-earth after the Downfall of Númenor, repeated by Aragorn at his coronation (LotR3/VI ch. 5): Et Eärello Endorenna utúlien. "Out of [lit. out from] the Great Sea to Middle-earth I am come" (Endor(e)-nna "Middle-earth-to"). The allative may also carry the meaning "upon"; cf. i falmalinnar "upon the foaming waves" in Namárië (-linnar being the ending for partitive plural allative; see below).

The instrumental case has the ending -nen and marks the instrument with which something is done, or simply the reason why something happens. Examples from Namárië are laurië lantar lassi súrinen, "like gold fall [the] leaves in [or by] the wind", i eleni [tintilar] airetári-lírinen, "the stars tremble in her song, holy and queenly", literally *"the stars tremble by holyqueen-song". An example of a more typical "instrumental" instrumental is provided by the sentence i carir quettar ómainen, "those who form words with voices" (WJ:391), ómainen being the plural instrumental of óma "voice".

Respective (?): This is what some have called a case that is listed in a letter Tolkien sent to Dick Plotz in the second half of the sixties (the so-called Plotz Letter is indeed our main source of information about the Quenya cases). The ending is -s (plural -is), but Tolkien did not identify this case by any name, nor have we ever seen it used in a text. Its function is therefore wholly unknown; it has indeed been called the Mystery Case. Some writers have used it simply as an alternative locative ending. They have had no nightly visits by Tolkien afterwards, so perhaps this is acceptable to him.

If case endings are added to a noun ending in a consonant, an e is often inserted between the noun and the ending to prevent a difficult cluster from arising: Elendil with the allative ending -nna "to" becomes Elendilenna "to Elendil" (PM:401), not **Elendilnna. However, if the noun is plural, an i is inserted between the noun and the ending: elenillor "from (the) stars" (elen "star") (MC:222).

The Quenya numbers: The numbers are singular, plural, partitive plural and dual. The singular and plural need no explanation. The function of the partitive plural (so called by Tolkien in WJ:388) as opposed to the normal plural is not fully understood, but it seems that it denotes some out of a larger group. Combined with the definite article i, it may simply denote "many": The element li in the phrase i falmalinnar "upon the foaming waves" in Namárië was translated "many" by Tolkien in his interlinear translation in RGEO:66-67. Since -li is the ending for partitive plural, it was long called "multiple plural"; indeed it was thought that it simply meant "many" of the thing in question, while the normal plural only meant "several". This may be correct in some cases, but it cannot be the whole story. The dual is used with reference to a natural pair, like two hands belonging to one person (cf. the word máryat "her hands" in Namárië, -t being a dual ending, literally "her pair of hands").

The nominative plural is formed with one of two endings. The ending -r is used if the noun ends in any vowel except -ë; well-known examples are Vala pl. Valar, Elda pl. Eldar, Ainu pl. Ainur. If the noun ends in a consonant or in -ë, the plural ending is -i, and it displaces the final -ë: Atan pl. Atani, Quendë pl. Quendi. (But if the noun ends in -ië, it forms its plural in -r to avoid one i following another: tië "path", tier "paths" - not **tii.) In the other cases, the plural ending is either -r or -n; for instance, the allative ending -nna has a plural form -nnar, the locative ending -ssë becomes -ssen, and ablative -llo can form its plural both in -llon and -llor. In the dative, instrumental and "respective", the plural is indicated by the element i, inserted between the stem of the noun and the same case ending as in the singular. (See the full list of endings below.)

The partitive plural has the ending -li, presumably *-eli on a noun ending in a consonant, but a contraction or an assimilated form may also be used (for instance, the partitive plural of casar "dwarf" is casalli, for *casarli). The endings for other cases are simply added following the ending -li, e.g. ciryali "some ships" > allative ciryalinna (or ciryalinnar) "to some ships". Note, however, that the vowel of -li is lengthened before the endings -va and -nen for possessive and instrumental, respectively: -líva, -línen.

Like the nominative plural, the nominative dual is formed with one of two endings. Most nouns take the ending -t, as in the word máryat "her hands" (two hands, a pair of hands) in Namárië. "Two ships, a couple of ships" is likewise ciryat (cirya "ship"). But if the last consonant of the stem is t or d, the ending -u is preferred: Alda "tree", Aldu "the Two Trees". In the other cases, a t is somehow inserted into or added to the various endings; for instance, the endings -ssë, -nna and -llo for locative, allative and ablative, respectively, turn into -tsë, -nta and -lto (ciryatsë, ciryanta, ciryalto = "on/to/from a couple of ships"). The instrumental ending -nen becomes -nten, while the dative ending -n becomes -nt (ciryant "for a couple of ships" - this is, by the way, the sole known case of a consonant cluster being allowed at the end of a word in Quenya).

These, then, are the Quenya case endings:

Nominative: Sg. no ending, pl. -r or -i, part. pl. -li (Book Quenya -lí), dual -t or -u.
Accusative (in Book Quenya only): Sg. lengthening of the final vowel (if any), pl. -i, part. pl. lí, dual: probably lengthening of the final u to ú (no distinct accusative in the case of t-duals?)
Dative: Sg. -n, pl. -in, part.pl. -lin, dual -nt (but possibly -en following a dual in -u)
Genitive: Sg. -o, pl. -on (added to the nom.pl.), part.pl. -lion, dual -to.
Possessive: Sg. -va, pl. -iva, part.pl. -líva, dual -twa.
Locative: Sg. -ssë, pl. -ssen, part.pl. -lisse(n), dual -tsë.
Allative: Sg. -nna, pl. -nnar, part.pl. -linna(r), dual -nta.
Ablative: Sg. -llo, pl. -llon or -llor, part.pl. -lillo(n), dual -lto.
Instrumental: Sg. -nen, pl. -inen, part.pl. -línen, dual -nten.
Respective: Sg. -s, pl. -is, part.pl. -lis, dual -tes.

(See the Appendix for examples of Quenya nouns declined in all cases.)

The Article
Quenya has a definite article i = "the", e.g. i eleni "the stars" in Namárië. There is no indefinite article like English "a, an"; the absence of the article i usually indicates that the noun is indefinite: Elen "star" must be translated "a star" when English grammar requires an article, as in the famous greeting Elen síla lúmenn' omentielvo "a star shines at the hour of our meeting" (LotR1/I ch. 3). But sometimes Tolkien's translations introduce a "the" where there is no i in the original, cf. the first line of Namárië: Ai! laurië lantar lassi... "Ah! like gold fall the leaves..." rather than just "(some) leaves".

The Verb
Most Quenya verbs can be divided into two categories. The smallest group may be termed basic (or primary) verbs. The stem of such a verb represents a basic root with no additions. For instance, the verb mat- "eat" comes directly from the Eldarin root-word MAT- of similar meaning (LR:371); the verb tul- "come" is simply a manifestation of the root-word TUL "come, approach" (LR:395).

The second, larger group of Quenya verbs may be termed A-stems, or derived verbs. They all show the final vowel -a, but it is not part of the basic root; their stems have added some ending to this root. The endings -ya and -ta are by far the commonest. For instance, the root TUL "come" yields not only the basic verb tul- "come", but also the longer A-stems tulta- "to summon" and tulya- "to bring". Here the endings are seen to modify the root meaning; in this case both -ta and -ya are causative, since "to summon" and "to bring" are variations of the idea "to make come". But often, the endings seem to make no difference for the meaning (the root SIR "flow" produces the Quenya basic verb sir- of the same sense, but in a related language the derived verb sirya- is used for the same meaning: LR:384). Some A-stem verbs show less frequent endings like -na (e.g. harna- "to wound", apparently derived from the adjective or participle harna "wounded"); there are also a few A-stems that end in the simple ending -a, e.g. ora- "to urge".

Five Quenya tenses are known: Aorist, present, past, perfect, and future. (In all likelihood, Tolkien also imagined yet other tenses, like the pluperfect - but such forms are not exemplified in our material.)

The aorist is the simplest form both by its meaning and shape. The basic meaning of the verb is not modified or limited in any particular way. The aorist may express general, timeless truths, as when Elves are described as i carir quettar "those who make words" (WJ:391). However, it can equally well describe a simple, ongoing action, as in the battle-cry heard before the Nirnaeth Arnoediad: Auta i lómë! "The night is passing!" In this context the translation "is passing" is the most natural English rendering, but the aorist auta as such simply means "passes" and does not explicitly mark the action as on-going (as does the Quenya present tense, see below). Generally speaking, the Quenya aorist apparently corresponds to the simple English present tense (as opposed to the "is ...-ing" construction). So Tolkien often translated it, e.g. in the first line of Namárië: Ai! laurië lantar lassi súrinen, "ah! golden fall the leaves in the wind".

The aorist of a basic verb originally showed the ending -i. In Quenya, the final short -i of earlier stages of Elvish had changed into -ë, so now the aorist of a primary verb like car- "make, do" appeared as carë instead (this form may be translated "makes" or "does"). However, since the vowel had only changed when it occurred at the end of a word, we still see -i- whenever any kind of further ending is added. When a finite Quenya verb occurs with a plural subject, the verb receives the plural ending -r, so the aorist carë "makes" correspond to carir "make" in the sentence "those who make words" cited above. We also see -i- before all pronominal endings; indeed Tolkien very often cites Quenya primary verbs as aorist forms with the ending -n "I" attached (e.g. carin "I make", LR:362, tulin "I come", LR:395). A-stem verbs show no variation, but end in -a whether or not any further ending follows (e.g. lanta "falls", lantar "fall" with a plural subject, lantan "I fall", etc.)

What is sometimes called the Quenya present tense is also referred to as the continuative form. It refers to an action that is explicitly identified as on-going, and it is often best translated by means of the English "is ...-ing" construction. The present tense of a basic verb is formed by adding the ending -a and lengthening the vowel of the verbal stem itself (the long vowel being marked by an accent). Thus the verb sil- "shine" has the present-tense form síla "is shining"; the verb mat- "eat" has the present-tense form máta "is eating" (or with plural subjects sílar "are shining", mátar "are eating"). Sometimes Tolkien translates Quenya present-tense forms by means of the English simple present tense, as in the famous greeting elen síla lúmenn' omentielvo = "a star shines on the hour of our meeting". It seems that the aorist and the present tense are to some extent interchangeable; in one draft version of this greeting, Tolkien indeed used the aorist silë "shines" instead of the present-tense form síla "is shining" (see RS:324).

As for the present tense of A-stem verbs, the traditional interpretation of Quenya grammar had it that the ending -a simply merged with the final -a already present at the end of the verbal stem, so that (say) lanta- "fall" would have the present tense lanta "is falling". This form appeared to be attested in the first line of Namárië (with the ending -r to go with the plural subject lassi "leaves"): Laurië lantar lassi, "golden fall leaves", or "golden leaves are falling". However, it seems that the form lantar is actually an aorist (cf. above). In July 2000, new examples appeared in Vinyar Tengwar #41: The A-stem verb ora- "urge" is shown to have the present tense órëa "is urging". This would seem to indicate that A-stem verbs actually have present-tense forms in -ëa: Apparently the final -a of the stem undergoes dissimilation to -ë- to avoid two a's in sequence when the ending -a associated with the present tense is added. Where there is no consonant cluster following the stem-vowel, it is lengthened just as in the case of basic verbs: thus órëa from ora-. However, most A-stems do have a consonant cluster following the stem-vowel, and in front of a cluster, Quenya cannot have a long vowel. We have no examples, but it may be assumed that an A-stem like lanta- "fall" has the present tense lantëa "is falling" (pl. lantëar "are falling"): Since the stem-vowel cannot be lengthened in this position, only the ending -ëa tells us that this is a present-tense (or continuative) form.

The Quenya past tense always shows the final vowel -ë (though secondary endings may of course be added; for instance, we see -er where the verb has a plural subject). This vowel -ë is very often part of the ending -në, which seems to be the most general past tense marker in Quenya. A-stem verbs typically add this ending. For instance, a verb orta- "to rise/raise" is listed in the Etymologies (entry ORO, LR:379), and the song Namárië in LotR demonstrates that its past tense is ortanë. Other attested examples:

ulya- "to pour", past tense ulyanë (LR:396 s.v. ULU)
hehta- "to forsake", past tense hehtanë (WJ:365)
ora- "to urge", past tense oranë (VT41:18)

As for primary verbs, the picture is somewhat more diverse. When such verbs end in -r or -m, they can take the ending -në just like A-stem verbs, since the resulting clusters rn and mn are permitted by Quenya phonology. Examples from the Etymologies include tirnë as the past tense of tir- "to watch" and tamnë as the past tense of tam- "to tap" (see entries TIR, TAM). Since double nn is also a permissible Quenya cluster, it may be assumed that primary verbs in -n also receive the ending -në; for instance, it seems a fair guess that the verb cen- "to see" has the past tense cennë "saw" (not attested in published material).

However, primary verbs in -p, -t, and -c cannot receive the ending -në, since the clusters pn, tn, cn that would arise are not permitted in Quenya. This problem is solved by replacing the nasal element of the ending -në with nasal infixion intruding before the final consonant of the verbal stem. The nasal infix appears as m before p; before t it is simply n, and before c (k) it is pronounced like the ng of English king, but it is still spelt n: Compare the pronunciation of English orthographic "nk" as in think. Examples from the Etymologies (see entries TOP, SKAT, TAK):

top- "to cover", past tense tompë
hat- "to break", past tense hantë
tac- "to fasten", past tense tancë

Primary verbs that have -l as their final consonant seem to surrender the nasal element completely; the past tense of vil- "to fly" is given as villë (LR:398 s.v. WIL). Perhaps vil- (older wil-) originally formed its past tense by means of nasal infixion much like the verbs above, but *winle eventually turned into villë by assimilation. (For this development of the older group nl in Quenya, compare the noun nellë "brook", which Tolkien derived from older nenle: See the entry NEN in the Etymologies, LR:376.)

The perfect tense expresses the idea of an action that has been completed in the past, but that is still "relevant" for the present moment, usually because its effects are still felt. English has no unitary perfect tense, but employs a circumlocution involving the verb "to have", as in "he has come". Quenya does have a unitary perfect tense. Its formation is somewhat complex. All perfect forms receive the ending -ië (or with a plural subject, -ier). The vowel of the verbal stem is, if possible, lengthened. Thus from the stem tul- "to come" we have the perfect tense utúlië "has come". As we see, a prefix u- also turns up here. This prefix, called the augment, is actually variable in form, since it is always identical to the vowel of the verbal stem itself. Hence:

mat- "to eat" vs. amátië "has eaten"
tec- "to write" vs. etécië "has written"
ric- "to twist" vs. irícië "has twisted" (attested in VT39:9)
not- "to count" vs. onótië "has counted"
tuv- "to find" vs. utúvië "has found" (upon finding the sapling of the White Three, Aragorn exclaims utúvienyes = "I have found it")

In some of Tolkien's earlier forms of Quenya, we find perfect-tense forms without augments, for instance lendië (rather than elendië) for "has gone" or "has traveled" (the form lendien in SD:56 - incorporating the suffix -n "I" - may be interpreted "I have traveled"). Occasionally, the augment is omitted in post-LotR sources as well, e.g. fírië rather than ifírië for "has breathed forth" in MR:250 (verb fir- "to expire, fade, die"). The example lendië also lacks the long stem-vowel normally associated with the perfect tense (not *léndië). This is because Quenya cannot have a long vowel immediately in front of a consonant cluster, and most A-stem verbs do have a consonant cluster following the stem-vowel that would normally be lengthened in the perfect tense. Good Tolkienian examples are lacking, but it may be assumed that except for this lengthening, the perfect tense of derived verbs is formed as in the case of primary verbs: by prefixing the stem-vowel as an augment and adding the ending -ië (this ending replacing the final -a of such derived verbs):

panta- "to open" vs. apantië "has opened"
menta- "to send" vs. ementië "has sent"
tinta- "to kindle" vs. itintië "has kindled"
costa- "to quarrel" vs. ocostië "has quarreled"
tulta- "to summon" vs. utultië "has summoned"

However, the simplest derived verbs, with only the short ending -a, may show the normal lengthening of the stem-vowel as well (since there is no consonant cluster following the vowel). A verb like mapa- "grasp, seize" may therefore form its perfect tense just as if it were a primary verb **map-: perfect tense presumably amápië. The numerous A-stem verbs in -ya may behave in a similar way. When the perfect ending -ië is added to a verb in -a, regularly displacing the final -a, we would expect the final combination **-yië, but Quenya cannot have y + i. Therefore, -yi- is simplified to -i-, so that nothing is left of the original ending -ya. In short, the perfect tense of a verb in -ya can be formed as if this ending were not present at all, leaving only the basic root of the verb, which necessarily behaves as a primary verb in the perfect tense. We may have one attested example: In the Etymologies, Tolkien listed a verb vanya- "go, depart, disappear" (LR:397 s.v. WAN). What could be its perfect tense, avánië, appears in Namárië (there with the plural ending -r to go with a plural subject; Galadriel sings yéni avánier, translated "the long years are gone", more literally "...have passed away"). The form avánië displays all the features a perfect-tense verb can have: prefixing of the stem-vowel as an augment a-, lengthening of the stem-vowel to á in its normal position, and suffixing of the ending -ië (completely displacing the -ya of vanya-). In one post-LotR source, Tolkien discussed this perfect-tense form occurring in Namárië and (re-?)interpreted it as the perfect tense of the very irregular verb auta- "go away": See WJ:366. However, it seems entirely possible that when Tolkien actually wrote Namárië more than a decade earlier, he thought of avánië as the perfect tense of the verb vanya-. If so, this example gives away how the numerous verbs in -ya behave in this tense. Perhaps the verb vanya- was later abandoned because it clashed with the adjective vanya "beautiful".

The future tense has the ending -uva (or with a plural subject, -uvar). For instance, the future tense of the basic verb mar- "dwell, abide" occurs in Elendil's Oath: maruva "will abide" (there with the pronominal ending -n "I" attached: sinomë maruvan, "in this place will I abide"). The final vowel of A-stem verbs seems to drop out before the ending -uva; attested examples include linduva as the future tense of linda- "sing" and apparently oruva as the future tense of ora- "urge". The form antáva as the future tense of anta- "give" (LR:63) is a divergent example; here the simpler ending -va is employed, combined with lengthening of the final vowel of the verb-stem. However, the form antáva comes from a pre-LotR source; Tolkien may have revised the language later. The other examples we have suggest that in LotR-style Quenya, the future tense of anta- should be antuva (not attested).

The future tense is also used in a kind of "wishing formula" introduced by the word nai "be it that" or "wish that". Thus a future-tense form like tiruvantes "they will keep it" (tiruva "will watch/keep" + -nte "they" + -s "it") may be used to expressed a wish by prefixing nai: Cirion's Oath has nai tiruvantes for "may they keep it" or "wish that they will keep it" (UT:305, 317).

Other forms: In addition to the five (known) tenses, the Quenya verb also may also appear in such forms as infinitive, gerund and imperative. The infinitive of a basic verb (e.g. quet- "to speak") shows the ending -ë, as in the example polin quetë "I can speak" (VT41:6). A-stem verbs apparently show no special endings in the infinitive; the stem and the infinitive are simply identical (for instance, the verb lelya- "go" would have the infinitive lelya "to go"). Thus, the infinitive is identical in form to an (endingless) aorist. Quenya also has an extended infinitive which adds the ending -ta; when added to the infinitive of a basic verb, its ending -ë instead appears as -i-: While the simplest infinitive of the verb car- "to do, to make" is carë, its extended infinitive is thus carita. The extended infinitive may receive pronominal endings denoting the object of the infinitive, e.g. -s "it" in caritas "to do it".

The extended infinitive may also function as a gerund, that is, a verbal noun the English equivalent of which is formed with the ending -ing. (This English ending is also used to form active participles, but they are quite distinct in Quenya, ending in -la instead.) An extended infinitive is seen to function as an gerund in the sentence lá carita i hamil mára alasaila ná, "not to do [or, not doing] what you judge good would be unwise" (VT42:33). Another infinitival/gerundial ending is -ië, as in the verbal noun tyalië "play(ing)" vs. the basic verb tyal- "to play". The final -a of an A-stem verb would apparently be omitted before the ending -ië is added (and in the case of a verb ending in -ya, this entire ending would have to go, since **-yië would be an impossible combination). Combined with the dative ending -n "to, for", such gerunds in -ië may express the meaning of English infinitives signifying "[in order] to do": The verb enyal- "recall" is attested in the form enyalien "[in order] to recall/commemorate", which Tolkien explained as a gerund inflected for dative (UT:317).

The imperative may be formed by placing the independent imperative particle á in front of a form similar to the simplest infinitive (or endingless aorist). Thus from carë "to do" may be formed the imperative phrase á carë "do!" or "make!" The imperative particle may also appear in the shorter form a, as when the Ring-bearers were hailed with the cry a laita te! "bless them!" on the Fields of Cormallen. (Possibly the short form a is preferred when there is a long vowel or diphthong in the following verbal stem, like the diphthong ai of the verb laita- "bless, praise".) A negative imperative may be introduced by áva "don't!", e.g. áva carë "don't do [it]!" (WJ:371).

There are also a few attested imperative forms that do not include the independent imperative particle, but employ the related ending -a. Thus we have exclamations like ela! "look!" or heca! "be gone!" (WJ:364). Cf. also ëa! as the imperative of the verb "exist", used by Eru Ilúvatar when he grants independent existence to the Music of the Ainur: "Eä! Let these things Be!" It is not known how productive Tolkien intended this formation to be. It may be that basic verbs may have alternative imperative forms with the ending -a (distinct from the present tense since the stem-vowel is not lengthened). For instance, perhaps the imperative of car- "do" may be *cara! as well as á carë!

Special verbs: Not all Quenya verbs fit readily into the system sketched above. "Irregularities" are often historically justified in terms of the underlying phonological evolution Tolkien had in mind, and in this perspective the verbs are not irregular at all. So let us speak of "special" rather than "irregular" verbs.

One example of an historically justified "irregularity" is provided to the verb rer- "to sow". We might expect its past tense to be **rernë; cf. such examples as tir- "watch", past tense tirnë. But the actual past tense "sowed" is rendë. The discrepancy is easily explained: Whereas the verb tir- "to watch" directly reflects a primitive root TIR, so that the r had been r all along, the final -r of the verb rer- "to sow" actually comes from an earlier d: The original root is RED (LR:383). The past tense rendë is formed by simple nasal-infixion of this root, so this past tense is actually wholly analogous to (say) quentë as the regular past tense of quet- "to say, speak". However, as Quenya evolved from primitive Elvish, an original d following a vowel normally became z and then r. So the root RED produced a primary verb rez- > rer-, but in the past tense form rendë, the nasal infix "shielded" the original d from the preceding vowel. Thus it remained d. - Other verbs that may belong to this category include hyar- "to cleave", ser- "to rest" and nir- "to press, thrust", since they are derived from roots SYAD, SED, NID (see VT41:17 about the latter). However, the past tense forms hyandë, sendë, nindë are not explicitly mentioned in published material.

One attested form that would belong in this category is lendë "went" as the past tense of the verb "to go, travel". The form lendë arises by nasal-infixion of a base LED (listed in the Etymologies; according to the WJ:363 it is re-formed from even older DEL). The basic form of the verb "to go" is lelya- (from older ledyâ), so the past tense lendë also illustrates another phenomenon: some verbs in -ya drop this ending in the past tense, which is then formed just as if we were dealing with a primary verb. Especially interesting in this regard is the verb ulya- "to pour", which Tolkien noted has a double past tense: "Poured" in the transitive tense, as in "the man poured water in a cup", is ulyanë with the normal past tense ending -në added to -ya: a perfectly "regular" form according to the rules set out above. However, the past tense "poured" in the intransitive sense, as in "the river poured into the gorge", is ullë - as is this were a primary verb *ul- rather than ulya-. (Compare villë as the past tense of vil- "to fly".) This suggests that intransitive verbs in -ya surrender this ending in the past tense. Some other intransitive verbs may seem to support this theory, though Tolkien only listed the deviant past tense forms without explaining them: We have already mentioned (ledyâ >) lelya- "to go", past tense lendë; other examples include farya- "to suffice", pa.t. farnë, and vanya- "go, depart, disappear", pa.t. vannë (instead of the "regular" forms **lelyanë, **faryanë, **vanyanë).

As we have already touched on, Tolkien may have replaced the verb vanya- "to disappear" with auta- "go away, leave", which has a double set of past and perfect tenses: past oantë with perfect oantië if the verb is used with reference to physically leaving one place and going to another, but past tense vánë with perfect avánië when the verb is used of disappearing or dying off. The word vanwa "gone, lost, vanished, departed" is said to be the "past participle" of this verb, though it seems so irregular that it could just as well be treated as an independent adjective. See WJ:366.

While the past tense of most verbs involves a nasal, either in the form of nasal infixion (as in quentë "said" from quet- "to say") or as part of the long past tense ending -në, there are some verbs that show no nasal element in the past tense. Instead, the past tense is formed by lengthening the stem-vowel and adding the ending -ë. The past tense of lav- "to lick" is seen to be lávë (compounded in Namárië: undulávë = "down-licked", i.e. "covered completely"). The negative verb um- "not be" or "not do" likewise has the past tense úmë. However, most words formed from verbs in this manner are not past-tense forms, but abstract nouns. For instance, the word sérë connects with the verb ser- "to rest", but sérë is not the past tense "rested"; it means "rest" as a noun. The past-tense formation represented by words like lávë and úmë is thus rather ambiguous by its form, and maybe this is why it is not much used.

A few verbs in -ta may drop this ending in the past tense, and what remains of the verbal stem forms its past tense form after the úmë pattern. For instance, the verb onta- "to beget, create" may have the past tense form ónë (as an alternative to the regular form ontanë). Indirect evidence from LotR suggests that the verb anta- "to give" may behave in the same manner: past tense ánë rather than (or as well as) the regular form antanë, itself unattested (whereas ánë appears in early material). The Sindarin form onen "I gave", mentioned in a LotR appendix, would correspond to Quenya ánen (the ending -n signifying "I").

The verb "to be": The only attested forms of this verb are ná "is", nar "are" and nauva "will be". The past tense "was" may be né. The infinitive and perfect forms are unattested and uncertain.

The Adjective
Many Quenya adjectives end in the vowel a:

laiqua "green"
alassëa "happy" (from alassë "happiness")
númenya "western" (from númen "west")
vanya "fair"
morna "black"
melda "dear, beloved" (originally *melnâ; the endings -na and -da may sometimes have the same origin, n being dissimilated to d following l)

There are also a number of adjectives ending in ë, like carnë "red", varnë "swart" or inimeitë "female". It may be noted that in mature Quenya, there seem to be no adjectives in -o or -u. Relatively few adjectives end in a consonant - typically n, as in firin, qualin "dead" (by natural cause and by accident, respectively).

Adjectives agree in number with the noun they describe. Adjectives in -a have plural forms in -ë, adjectives in -ë or in a consonant have plural forms in -i, and adjectives in -ëa have plural forms in -ië:

vanya vendë "a beautiful maiden" > vanyë vendi "beautiful maidens"
carnë parma "a red book" > carni parmar "red books"
laurëa lassë "a golden leaf" > laurië lassi "golden leaves"
firin casar "a dead dwarf" > firini casari "dead dwarves"

Hence in the first line of Namárië we find laurië lantar lassi, "like gold (lit. golden) fall the leaves", while "golden falls a leaf" would be laurëa lanta lassë (both the verb and the adjective agreeing with lassë, lassi "leaf, leaves" in number).

The present writer once thought that the name of the journal Vinyar Tengwar contained an error; if the intended meaning was "New Letters", it would have to be Vinyë Tengwar (vinya "new", tengwa "letter"). But as Carl F. Hostetter subsequently explained, the intended meaning is "News Letters", so vinya is inflected like a noun. This writer was still skeptical about the whole construction and thought it should have been Tengwar Vinyaron "Letters of News" or something similar, but material that has since been published shows that "loose compounds" of this kind really are possible. (Last line of querulous defence: Tengwa "letter" is only attested with the meaning "character", not "letter" = "mail, post"!) It may be noted that in some earlier variants of Quenya (or "Qenya"), adjectives actually did have plural forms in -r; cf. LR:47, where raikar is used as the plural form of raika "bent". Tolkien revised the grammar later.

An intensive or superlative form of the adjective is derived by prefixing an-: Calima "bright", Ancalima "most bright" (Letters:279). We don't know how to construct the comparative ("brighter, more bright"), though an element yanta- is compared to Gnomish gantha- "more" in a very early wordlist compiled by Tolkien (see Parma Eldalamberon No. 11 p. 37, where the misreading "yonta" occurs; this error was pointed out in Parma Eldalamberon No. 12 p. 106 - however, the latter source also states that yanta- is actually a verb "enlarge, increase", so this may not be a word for "more" after all).

The Participles
The present (or active) participle describes the condition you are in when you do something: If you go, you are going; if you think, you are thinking. In English, present participles are derived from the corresponding verbal stems by adding the ending -ing. The corresponding Quenya ending is -la. There are many examples of it in the Markirya poem (MC:221-222 cf. 223). For instance, the participle falastala "foaming" is derived from a verbal stem falasta- "to foam". If the stem-vowel is not followed by a consonant cluster (or another vowel), it is lengthened: the participle of hlapu- "fly" (in the wind, of spray etc.) is hlápula. Basic verbal stems like sil- may be turned into "continuative stems" (with long vowel and final a: síla-) before the participal ending is added, so "shining" may be sílala (attested in the Markirya poem with a "frequentative" stem sisílala, with reduplication of the first syllable). But the connecting vowel may also be i, with no lengthening of the stem vowel; cf. itila "twinkling, glinting" in PM:363 (stem it-, though a verbal stem ita- is also given).

The past (or passive) participle describes the condition you enter if you are exposed to the action of the corresponding verb (if someone sees you, you are seen; if someone kills you, you will thereafter be killed), or, in the case of some verbs, the condition you are in after having completed the action described by the verb (if you go, you will thereafter be gone). In Quenya, most past participles are derived from the corresponding verb with the ending -na or -ina. The past participle of car- "make" is carna "made"; the stem rac- means "break", while rácina is "broken" (if there is not a consonant cluster following the stem vowel, this vowel seems to be lengthened when the participal ending is added, like a > á in this case). If the stem ends in l, the ending -na is dissimilated to -da: mel- "love", melda "beloved" (Tolkien's gloss of the latter, "beloved, dear" rather than "loved", indicates that the distinction between adjectives and participles sometimes becomes blurred).

The past participle probably agrees in number with the noun it describes (by final -a becoming -ë in the plural, just as with normal adjectives), but the present participle does not change -la to -lë as one might expect; it seems to be indeclinable (MC:222: rámar sisílala "wings shining", not **rámar sisílalë). Perhaps this is to avoid confusion with the verbal noun ending -lë "-ing" (as in Ainulindalë "the Music of the Ainur", literally *"Ainu-singing").

Pronouns
The pronouns have always been a problem. There are many uncertain points, and the subject is further muddied by the fact that Tolkien seems to have revised the pronominal system repeatedly. The system sketched here is pieced together from many sources and involves both extrapolations, reconstructions and some undeniably arbitrary choices. Not for a second do I think it is 100 % correct according to Tolkien's final intentions.

One thing, at least, is perfectly clear: Quenya pronouns usually appear as endings directly suffixed to a verb or noun, not so often as independent words, as in English. Examples from Namárië are the words máryat and hiruvalyë. Máryat means "her hands", "her" being expressed by the pronominal ending -rya (here followed by the dual ending -t to denote a natural pair of hands). Hiruvalyë is "thou shalt find", "thou" being expressed by the pronominal ending -lyë added to the verb hiruva "shall (shalt) find". Cf. also the ending -n "I" in Elendil's words Endorenna utúlien, "to Middle-earth I am come" (utúlië-n "have-come I").

This is an attempt, and nothing more, to compile a table of the pronominal endings used on verbs:

1. person sg: -n or -nyë "I"
2. person sg and pl, courteous: -l or -lyë "thou, you"
2. person sg and pl, familiar: *-ccë "you" (based on a Sindarin ending -ch, very hypothetical!)
3. person sg: -s "he, she, it" (there may also be gender-specific forms -ro "he", -rë "she")
1. person pl. -mmë: "we" (exclusive), -lmë "we" (inclusive)
1. person dual *-lvë: "we" (inclusive, "thou and I" - it is possible that this should rather be *-lwë)
3. person plural -ntë "they"

Note that there is a distinction between exclusive and inclusive "we", despending on whether the person addressed is included in "we" or not. Also note that -lme is the inclusive, not the exclusive "we" - exclusive "we" is -mmë! The ending *-lvë (or *-lwë?) is the inclusive dual "we", sc. "we" meaning "thou and I", not the general (plural) inclusive "we". This part of the Quenya pronominal system has long been misinterpreted (in Jim Allan's An Introduction to Elvish, Nancy Martsch' Basic Quenya etc. etc.)

Example: lendë "went", lenden or lendenyë "I went", lendel or lendelyë "you [polite] went", lendeccë "you [familiar] went", lendes "he/she/it went", lendemmë "we [exclusive] went", *lendelmë "we [inclusive] went", lendelvë (or rather lendelwë?) "we [= thou and I] went", lendentë "they went". The object can also be expressed as a pronominal ending added directly to the verb, following the ending denoting the subject. Cf. Aragorn's exclamation when he found the sapling of the White Tree: Utúvienyes!, "I have found it!" (utúvie-nye-s "have found-I-it"; LotR3/VI ch. 5), or a word from Cirion's Oath: tiruvantes "they will keep it" (tiruva-nte-s "will keep-they-it, UT:317).

As indicated by the word máryat "her hands" discussed above, even possessive pronouns like "her, his, my" are expressed by endings in Quenya, added directly to the noun (in this case má "hand"). The possessive endings used on nouns for the most part correspond to the pronominal endings used on verbs, but have the ending -a:

1. person sg: -nya "my"
2. person sg and pl, courteous: -lya "thy, *your"
2. person sg and pl, familiar: *-cca "your" (based on a Sindarin ending, very hypothetical!)
3. person sg: -rya "his, her" (and possibly "its")
1. person pl: *-mma: "our" (exclusive), *-lma "our" (inclusive)
1. person dual: *-lva: "our" (inclusive, "thy and my" - perhaps it should rather be *-lwa)
3. person pl.: *-nta "their"

Example: parma "book", parmanya "my book", parmalya "your (polite) book", parmacca "your (familiar) book", parmarya "his/her/?its book", parmamma "our (exclusive - not your!) book", parmalma "our (including your) book", parmalva (or parmalwa?) "our (thy and my) book", parmanta "their book" (the last of which must not be confused with the dual allative "to a couple of books"). In the case of nouns ending in a consonant, an e may be inserted between the noun and the possessive ending, e.g. macil "sword", *macilerya "his sword". In the plural, the plural ending -i may serve to separate noun and ending, e.g. macili "swords", maciliryar "his swords" - but as we see, an additional plural ending r appears after the suffix; cf. the next paragraph. There are some indications that the ending -nya "my" always prefers i as its connecting vowel, even in the singular, as in Anarinya "my Sun" in LR:72 (Anar "Sun"). Hence *macilinya "my sword".

The forms with possessive endings are inflected like normal nouns. Constructed examples: Nominative parmanya "my book" (pl. parmanyar "my books"), genitive parmanyo "of my book" (pl. parmanyaron), possessive parmanyava "my book's, of my book" (pl. parmanyaiva), dative parmanyan "for my book" (pl. parmanyain), locative parmanyassë "in my book" (pl. parmanyassen), allative parmanyanna "to my book" (pl. parmanyannar), ablative parmanyallo "from my book" (pl. parmanyallon, parmanyallor), instrumental parmanyanen "by my book" (pl. parmanyainen) - and respective parmanyas pl. parmanyais, whatever that means. Attested examples are tielyanna "upon your path" in UT:22 cf. 51 (tie-lya-nna "path-your-upon") and omentielvo "of our meeting" in the famous greeting Elen síla lúmenn' omentielvo "a star shines upon the hour of our meeting" (omentie-lva-o "meeting-our-of", the genitive ending -o displacing the final -a of the pronominal ending; cf. Vardo for **Vardao).

However, Quenya does possess independent pronouns in addition to the numerous endings discussed above. Some of them are emphatic. The final lines of Namárië provide us with a good example of this. In the sentence nai hiruvalyë Valimar "maybe thou shalt find Valimar", "thou" is expressed with the ending -lyë attached to the verb hiruva "shall find", as explained above. But in the following sentence, nai elyë hiruva "maybe even thou shalt find [it]", the corresponding independent pronoun elyë is used for emphasis: hence the translation "even thou". Another attested independent pronoun is inyë "(even) I". It is assumed that most of the independent pronouns are formed by prefixing e- to the corresponding pronominal ending, like *elmë "(even) we", but most of these forms are not attested in our small corpus. The emphatic words for "he, she, it" are uncertain.

Other independent pronouns, apparently not emphatic, include ni "I" (dative nin "for me" in Namárië), nye "me", tye "thee, you (as object)", ta "it", te "them" (and *"they"?), me "we" (dual met "we two" in Namárië). "He, she" may be so, se (cf. LR:385).

APPENDIX: EXAMPLES OF QUENYA NOUNS FULLY INFLECTED
The following examples are for the most part the ones listed by Tolkien in the so-called Plotz Letter, sent to Dick Plotz in the mid-sixties; reproduced by Nancy Martsch in Basic Quenya Appendix A:

1. CIRYA "ship" (an R-plural)
Singular: nominative cirya "a ship", (accusative ciryá in archaic Book Quenya only,) dative ciryan "for a ship", genitive ciryo "a ship's, of/from a ship", possessive ciryava "of a ship", locative ciryassë "on/in a ship", allative ciryanna "to a ship", ablative ciryallo "from a ship", instrumental ciryanen "with/by a ship", respective ciryas (meaning unknown).

Plural: nominative ciryar "ships", (accusative ciryai in Book Quenya, later ciryar,) dative ciryain, genitive ciryaron, possessive *ciryaiva (not in Plotz), locative ciryassen, allative ciryannar, ablative ciryallon (or *ciryallor, not in Plotz), instrumental ciryainen, respective ciryais.

Partitive plural: nominative ciryali *"some ships" (in archaic "Book Quenya" ciryalí both in the nominative and the accusative), dative ciryalin, genitive ciryalion, possessive ciryalíva, locative ciryalissë or ciryalissen, allative ciryalinna or ciryalinnar, ablative ciryalillo or ciryalillon, instrumental ciryalínen, respective ciryalis.

Dual: Nominative ciryat "two ships, a couple of ships" (no distinct accusative even in archaic Quenya?), dative ciryant, genitive ciryato, possessive ciryatwa, locative ciryatsë, allative ciryanta, ablative ciryalto, instrumental ciryanten, respective ciryates. In the case of a u-dual, however, duality is already sufficiently expressed by the suffix -u, so the normal case endings with no t are (presumably) used: Nominative aldu "two trees", (accusative *aldú,) genitive *alduo, possessive *alduva, dative *alduen, allative *aldunna, ablative *aldullo, locative *aldussë, instrumental *aldunen, respective *aldus.

2. LASSË "leaf" (an I-plural)
Singular: Nominative. lassë "leaf", (acc. lassé,) dative lassen "for a leaf", genitive lassëo "a leaf's", possessive lasséva "of a leaf", locative lassessë "in/on a leaf", allative lassenna "to a leaf", ablative lassello "from a leaf", instrumental lassenen "with a leaf", respective lasses (meaning unknown).

Plural: Nom. lassi "leaves", (acc. lassí,) dat. lassin, gen. lassion, poss. *lassiva (not in Plotz), loc. lassessen, all. lassennar, abl. lassellon or lassellor, inst. lassinen, resp. lassis.

Partitive plural: Nom. lasseli (in "Book Quenya" lasselí both in nom. and acc.), gen. lasselion, poss. lasselíva, dat. lasselin, loc. lasselisse/lasselissen, all. lasselinna/lasselinnar, abl. lasselillo/lasselillon, instr. lasselínen, resp. lasselis.

Dual: Nom/acc lasset "a couple of leaves", dat. lassent, gen. lasseto, poss. lassetwa, loc. lassetsë, all. lassenta, abl. lasselto, inst. lassenten, resp. lassetes.

The Plotz letter gives no example involving a noun ending in a consonant, but it must go something like this:

3. NAT "thing"
Singular: Nominative nat "thing", dative *naten "for a thing", genitive *nato "a thing's", possessive *natwa "of a thing", locative *natessë "in/on a thing", allative *natenna "to a thing", ablative *natello "from a thing", instrumental *natenen "by/with a thing", respective *nates (meaning unknown).

Plural: Nom. *nati "things", (acc. *natí,) dat. *natin, gen. *nation, poss. *nativa, loc. *natissen, all. *natinnar, abl. *natillon or *natillor, inst. *natinen, resp. *natis.

Partitive plural: Nom. *nateli (in "Book Quenya" *natelí both in nom. and acc.), dat. *natelin, gen. *natelion, poss. *natelíva, loc. *natelisse/natelissen, all. *natelinna/natelinnar, abl. *natelillo/natelillon, instr. *natelínen, resp. *natelis.

Dual: Nom/acc *natu "a couple of things" (the ending -u being preferred since the stem ends in a t): dat. *natuen, gen. *natuo, poss. *natuva, loc. *natussë, all. *natunna, abl. *natullo, instr. *natunen, resp. natus. But a consonant stem not ending in -t or -d, like elen "star", would presumably go like this: Nom/acc. *elenet "a couple of stars", dat. *elenent, gen. *eleneto (*elento?), poss. *elenetwa, loc. *elenetsë, all. *elenenta (perhaps contracted to *elenta), abl. *elenelto, inst. *elenenten (perhaps contracted to *elenten), resp. *elenetes (*elentes?).