Something remarkable happened last week, something that may never happen again: A new Quenya text by J.R.R. Tolkien is now in print. Hundreds of pages of Tolkien’s linguistic writings have been published in Vinyar Tengwar and Parma Eldalamberon, and there’s reason to hope that we’ll see more. However, most of these writings talk about the meanings and origins of individual words, and longer texts are few and far between (and none are significantly longer than “Galadriels’s Lament” included in Lord of the Rings).

The book that came out last week is called “The Collected Poems of J.R.R. Tolkien“. Among many previously unpublished texts in English, it contains a new poem in Quenya, “Twenty years have flowed away down the long river” / Yénion Yukainen Nunn’ ar Anduine Lútie Loar! / Loä Yukainen Avar Anduinë Sí Valútier (#182 in the book). The poem was originally recited by Tolkien during his visit to Rotterdam in 1958, and there was a plan to publish it back in 2014, but it didn’t work out.

I’ve been playing around with Tolkien’s languages for quite some time as part as my work on Etymograph, so I was very excited to try my hand at analyzing the new text. Of course, there are people who’ve studied Quenya far more than I have, and it’s quite possible that I got something wrong or missed some important details. If you have something to add or correct, please leave a comment below; this will be much appreciated! (And thanks to Paul Strack for his work on Eldamo! I couldn’t have done my analysis without this extremely valuable resource.)

The text is printed in three versions. Manuscript A is an earlier draft, and the Quenya text in manuscripts B and C is almost identical, so I’ve provided the analysis of texts A and C.

Text A

Yénion: Genitive plural (-ion) from yén ‘long year’. Since Tolkien was talking about 20 years of his own life, the use of yén here is not quite appropriate, because a long year is 144 solar years. This is why Tolkien used the word löa ‘solar year’ subsequently and in the version C of the text.

yukainen ‘twenty’: A combination of the prefix yu- ‘both’ and the later rejected form cainen ‘ten’.

ar: In this context it seems to be the preposition ar(a) ‘beside, next (to); out’.

nunn’ ‘down’: Allative form (-nna) of nú ‘going down’.

anduine ‘long river’: Quenya form of the name of the river Anduin.

lútie ‘floated’: Perfect form of the verb *lut- ‘to float’, from the CE root √LUT with the same meaning. Previously this root was attested in Quenya only as the noun luntë ‘boat’ (PE18:99).

loar ‘years’: Plural (-r) from the noun löa ‘(solar) year’.

aluvallie ‘never again’: I cannot fully analyze this form. The beginning part is aluvar ‘will not’ seen in text C, which is the combination of the negative particle al(a)-, the future tense suffix -uva- and plural ending -r. The -ll- is a regular result of assimilation in the -rl- cluster. However, the only meaning of -lie I’m aware of is ‘people, folk‘, which doesn’t fit in the context. -li is the partitive plural suffix, but I can’t see how it would work on a verbal form. It’s possible that -lie is some other derivation of the CE root √LI ‘many’, applied here with an intensifying meaning.

koiveanyo ‘in my life’: Genitive (-o) from the 1st person singular possessive form (-nyo) from the noun *coivea ‘life’. The regular word for ‘life’ is coivie, and the version C of the text uses that form instead of coivea.

entule ‘return’: Aorist from the verb entul- ‘to come again’, attested in future tense in the Silmarillion. The verb is derived from tul- ‘to come’ with the prefix en- ‘again’.

?naina ‘upstream?’: If the word is transcribed correctly, its root would be √NAY, which has two attested meanings: ‘cause grief‘ and ‘have opportunity, chance or permission‘. Neither of these matches ‘upstream’ which seems to be the only word in translation to which this word could correspond. The attested roots for the meaning ‘up’ are √AM and √RŌ/ORO, neither of which looks similar to this form.

ai ‘alas‘

melle ‘beloved’: Plural from the adjective *mella. This specific form is not attested elsewhere, but it’s clearly derived from the root √MEL ‘love’ and related to melda ‘dear, beloved’ (PE17:56)

yassen ‘in which’: Locative plural (-ssen) from the relative pronoun ya ‘that’. This exact form is attested in Galadriel’s Lament.

ekkennen ‘I saw’: Past tense (-ne), 1st person singular (-n), from the verb *eccen-, derived from the root cen- ‘to see’ with the prefix et- ‘out’. The form of the prefix before c- is the same as in the word ecces- ‘to find out’ (PE17:156).

haia ‘in the distance’: From the CE root √KHAY ‘far, distant, remote’, with the adjectival suffix -a. PE17:65 gives a slightly different form with a similar meaning: haila ‘far beyond’.

palantírielya ‘looking afar’: This seems to be an error for palantírienye, corrected by Tolkien in version C of the text.
The word is a gerund (-ie) from the word palantír- ‘to look afar’, well known from LotR. -lya is normally the 2nd person polite ending, which is not appropriate in this context (in earlier Quenya -lya was the suffix of the imperfective participle). -nye is the first person singular ending. Note that Tolkien specifically says (PE22:119) that the gerund cannot take pronominal affixes, so either this form is unexpected or the analysis is incorrect.

yárier ‘ages’: Nominative plural (-r) of the noun *yárie, derived from the root √YA ‘there, back, ago’, with abstract noun suffix -ië. The -r- seems to be inserted for phonotactics reasons, and Tolkien changed it to -s- in version C of the text.

andavanwe ‘long-departed’: Plural form of adjective *andavanwa, compound of anda ‘long’ and vanwa ‘gone’ (attested in Galadriel’s Lament).

yallume ‘in times when’: Compound of relative pronoun ya ‘that’ and lúme ‘time’. This word is also attested in an earlier text “Firiel’s Song” (LR:72) where Tolkien translates it as ‘at last’.

Eldalie: ‘the Eldar’, literally ‘Elven-folk‘

enwa ‘still’: Seems to be a variation of en(a) ‘still’, attested in PE17:167, from the CE root √EN ‘again’.

marnie ‘dwelt’: Perfect tense (-ie) from the verb mar- ‘to abide’. The form seems to be formed from the past tense marn-, rather than from the root mar-. This is unusual because similar forms existed in the 1940s (PE22:102) but were abandoned in later writings, and even then roots ending in -r seemed to have perfects without the -n- infix (PE22:103).

endoresse ‘in Middle-Earth’: Locative (-sse) of Endórë ‘Middle-Earth’.

luminke ‘a little while’: From the noun lúme ‘time’, derived with diminutive suffix -incë.

ar ‘and‘

laiqavinye ‘young green’: Plural form of substantivized adjective *laiquavinya, from laiqua ‘green’ and vinya ‘young’.

ólar ‘grew’: Aorist plural (-r) of the verb ola ‘grow, become’. The long vowel is unexpected, because other derived words from the root √OL are attested with a short root vowel.

i ‘to’: The word i is normally a relative pronoun, not a preposition, so maybe a more literal translation would be ‘grew so that trees’.

aldar ‘trees’: Nominative plural (-r) from the noun alda ‘tree’.

epeni ‘before’: This is clearly related to the preposition epë, which, interestingly enough, specifically means ‘after’ when speaking about time and ‘before’ in other contexts (VT49:12).

ilya ‘all‘. Attested in Galadriel’s Lament.

fentane ‘began’: Past tense (-ne) from an unattested verb *fenta. Based on the crossed-out form ‘epphenne’ in the manuscript, it seems clear that it’s related to the root √PHEN, glossed as ‘door’ in PE17:181. If a broader meaning of this root is taken, such as ‘entrance’, it can be seen how the meaning ‘begin’ seen in this text is related.

hwirya ‘wither’: This seems to be derived from an unattested root *√SWIR with the formative suffix -ya. Early Qenya had words for ‘wither’ with two other roots, hesta- and narqa-, but both were abandoned in later writings.

an ‘but‘. Note that from this point on the drafting of the Quenya text doesn’t seem to match the English text (“in the cold breath of the wizards”). The Quenya text says something more like “But then I saw before me…”

sinna ‘thither?’: Allative (-nna) from the pronoun si ‘this’.

‘ekkene: This seems to be another past tense form of eccen- ‘see’. I’m not sure why this word is written with an elision marker: I don’t see what vowel could have been elided.

opá: This seems to be a variant of opo ‘before’, or at least a word derived from the same root.

ni ‘I/me‘

Text C

avar ‘down?’: This seems to be a preposition derived from the root √AWA ‘away (from)’.

sí ‘now‘. Attested in many texts including Galadriel’s Lament.

valútier ‘have flowed away’: Same as lútier in text A, with an additional prefix va- ‘away from’.

i ‘that‘

aluvar ‘will never’: This can be seen as a future tense of the verb ala- ‘to deny’, but more likely it has a purely grammatical meaning. See aluvallie in text A.

coivienyo ‘in my life’: See koiveanyo in text A.

eärello ‘from the Sea’: Ablative (-llo) from the noun eär ‘sea’. This form is also attested in Elendil’s Oath.

nantule ‘return’: Aorist of nantul- ‘to return’, from the verb tul- ‘to come’ with the prefix nan- ‘again’.

palantírielya ‘looking far away’: See palantírienye in text A.

yásier: See yárier in text A.

tenn’ ‘still‘: This form is also attested in Elendil’s Oath, with a different meaning ‘until’.

lente ‘free’: Plural form of previously unattested adjective lenta ‘free’, from the root √LED ‘go, proceed’. Related to lenda- ‘go free’ (PE22:157).

landanoresse ‘in a wide country’: Locative (-sse) of landanore, from landa ‘boundary’ and nóre ‘country’. The Quenya meaning doesn’t quite match the translation because ‘wide’ rather implies ‘lacking boundaries’, however landanore doesn’t seem to contain any element expressing negation.

ólaner ‘grew’: Past tense (-ne) plural (-r) of the word ola ‘to become’. See ólar in text A.

tó ‘for’: Tolkien gives the meaning ‘thence‘ for tó (VT49:11), which doesn’t seem to fit the context. A variant to with a short vowel with a prepositional meaning ‘on’ is mentioned in PE22:168, but it’s unclear whether it refers to the same idea.

ilyama ‘all’: Compound of ilya ‘every’ and ma ‘something’.

menta ‘begins’: Aorist of menta-, attested in PE17:93 with the meaning “have as object, (in)tend, proceed, make for, go towards”. See fentane in text A. Looks like Tolkien rejected the idea of deriving a word for ‘begin’ from the root √PHEN ‘door’ and used a much more common root √MEN ‘go, move, etc.’ instead.

hondoringe ‘cold-hearted’: Plural from the adjective hondoringa, compound of hondo ‘heart, seat of deepest feelings’ and ringa ‘cold’.

fúmenen ‘in the breath’: Instrumental (-nen) of *fúme ‘breath’, derived with the abstract noun suffix -me from the root *-fú. The form of the root is somewhat unexpected because normally CE roots starting with PHU- produce hu- in Quenya, however, forms with fu- are also attested. The root is most likely connected to √PHUY ‘breathe out’, however, the details of the connection are unclear. The root √PHŪ is glossed as ‘interjection of displeasure’ in PE17:181, but this doesn’t feet the context.

istarion ‘of wizards’: Genitive plural (-ion) of istar ‘wizard’.