Up to this point in the course, we have presented three major groups of Quenya verbs: the basic verbs, the u-stem verbs and the a-stem verbs. The discussion of the a-stem verbs have been intentionally simplified, and as previously mentioned they represent a larger set of “hidden verb classes”. The chapter discusses these specialized verb classes in more detail, and how their conjugation into verb tenses differs from the simplified rules we have presented so far in the course.
This entire section can be considered an “advanced topic”. As such, the information given here is more technical that other sections of this course, and includes both a description of recommended (Neo) Quenya grammar and of the source material from which those rules are derived. You can also ignore this entire section and stick to the simplified conjugations described earlier in the course, and you will still be correct most of the time.
Prior to 2015, we didn’t have a good picture of the verb classes in the Quenya language. Most (Neo) Quenya descriptions of verb tense formation divided verbs into groups based on the ending of the verb stem, with basic verbs ending in a consonant (mat-), u-stem verbs ending in the vowel u (liru-), and a-stem verbs ending in the vowel a (tulta-). While there was scattered information among Tolkien’s published writings indicating further complexities in the Quenya system of verbs, at that stage we didn’t have a complete picture of how they worked. This changed with the publication of Parma Eldalamberon #22 (PE22) in 2015. That journal included a number of previously unpublished documents on Quenya verbs written in the 1930s through 1960s.
Of the documents presented in PE22, the two most significant for our purposes are the Quenya Verbal System (QVS) and Late Notes on Verb Structure (LVS). QVS was written in 1948, and had a detailed breakdown of Quenya verb classes. The verb classes in that document are the basis for the classes described here. However, QVS has a number of problems when it comes to (Neo) Quenya use. In particular, it belongs to a brief period in the late 1940s where Tolkien used pronominal prefixes for verbs rather than suffixes (there was a similar period in the 1920s). Thus in QVS, Tolkien used formations like ni·tulë for “I come” rather that tulin as is the norm in his later writings. Thus some of the features described in QVS were abandoned later, but it isn’t always clear what was abandoned and what was retained.
The LVS documents were written around 1969. They are not a single document, but rather a bundle of documents having to do with verbs. The title Late Notes on Verb Structure for the collection was coined by Christopher Gilson rather than Tolkien himself. LVS touches on most of the same verb classes discussed in QVS, and thus gives us an idea of how Tolkien thought Quenya verb classes behaved towards the end of his life. In particular, there is evidence of all the QVS verb classes in Tolkien’s later writing, and little evidence of any other new major verb categories (with the exception of a few very specialized classes like inceptives, not discussed here).
QVS (PE22/99-127) and LVS (PE22/147-168) are the primary sources for the information in this section, though some information is drawn from other sources.
The a-verbs were a class of verbs described in QVS that were derived from a root with the addition of the vowel ā (QVS: PE22/113, 116). Tolkien’s preferred example for this verb class was fara- “to hunt”. In Ancient Elvish the a-verbs had an “inherently continuative sense”, and as such did not have a distinct present tense form: fara was originally used for both present and aorist. By the time of modern Quenya, though, they had developed a separate present tense.
Tolkien experimented with a variety of present tense forms for the a-verbs, but in rough notes within LVS he indicated that he would “make Q. ea as present tense invade other forms”; the exact reading is unclear (PE22/164). This seems to mean that the ea-presents from the causative verb class (see below) spread into other verb classes, and this note is followed by several examples of a-verbs both with and without ea-presents: ála vs. álëa (ala- “grow”), fára vs. fárëa (fara- “hunt”), áva vs. ávëa (ava- “do not”). One distinctive characteristic of these ea-presents for a-verbs is that the base vowel was lengthened (á), something impossible for the present tense of most other non-basic verbs.
Most of the attested a-verbs in QVS and LVS had weak pasts (faranë “hunted”) and simple perfects (afárië “has hunted”), though some strong pasts also appeared. The net result is that a-verbs mostly follow the simplified rules described in this course. Note that this course uses the term a-stem verb to refer to the larger group of verbs with suffixes ending in a (including -ta, -ya), but a-verb refers specifically to verbs ending in just a and nothing else.
The formative verbs are one of the pair of verb classes that Tolkien referred to as the “half-strong” verbs (PE22/113-115), the other being the talat-stem verbs discussed in the next section. The formative verbs were generally the result of an ancient suffix -t(ă) or -y(ă) being added to an otherwise defective verbal root. The ancient form of these verbs ended in ă in the aorist tense, but did not use ă in other tenses. These verbs were usually (though not always) intransitive, not taking a direct object.
The distinctive characteristic of verbs in this class is that, like basic verbs, they generally form their past tense with a nasal-infixed n: nahta- “to slay” from primitive *nak-t(ă) with past form nacantë “slew” (PE17/77; PE22/157, 164), orya- “to rise” from primitive *or-y(ă) with past form oronyë “rose” (PE17/64, 77; PE22/164). The nasal was inserted in between the formative suffix -t(ă) or -y(ă) and the root, and the base vowel was duplicated to make the past pronounceable: nac-a-n-t-e and or-o-n-y-e. As such they have past tenses ending in -Vntë or -Vnyë, where V is the base vowel. This course refers to these as half-strong past tenses.
Another feature of this verb class is that they have variant present tense forms using the suffix -ia instead of -ëa: nahtia “is slaying”, ória “is rising” (PE22/164). But in keeping with the notion that “ea as present tense invaded other forms”, Tolkien also gave past tense forms with -ëa as well: nahtëa, órëa (PE17/77; PE22/159, 164). In LVS the ya-formatives lost their y in the present tense, allowing the base vowel to lengthen since it was no longer before a cluster; these ya-verbs likewise lost their y in the perfect tense: (or)órië “has risen”.
One notable subset of the formative verb are those with roots ending in Y or W: caita- “to lie (down)” from *KAY+tă, auta- “to depart” from *AW+tă. The Y/W roots usually (but not always) require some sort of suffix to be used as verbs (PE22/103). In attested examples, this is typically -ta. Verbs from Y/W roots have some rather unusual half-strong pasts due to various phonetic developments in Quenya’s history, such as (archaic) †cëantë “laid” (PE22/157) or †öantë “departed” (WJ/366). In practice, they were frequently supplanted by a strong past tense from the verbal root, such as cainë or vánë (< WĀ+nē).
In other cases the half-strong pasts of formative verbs usually survive, though they are occasionally displaced by weak pasts instead, especially in cases where the verb can be used transitively (with direct objects). I am of the opinion that the ia-presents of these verbs are archaic as well, displaced by ea-presents. It is an open question whether or not the y is retained in present tense forms of ya-formatives. In LVS the y was always lost for ya-formative presents, but there is an example from 1964 where the y is retained: melya- → present melyëa, past melenyë (PE17/77).
The talat-stem verbs are another class of “half-strong” verbs, which were derived from (rare) triconsonantal verbal roots, the best known being talta- “to collapse, fall (down)” based on the root TALAT. In theory these verbs would have used the same i-suffix as basic verbs in their aorist forms, and thus would have ancient aorist forms like *tal’ti “collapse” (PE17/186). Likewise their ancient present tense forms would have vowel lengthening and suffixed -ā: *talātā or *tālatā “is collapsing”. Their past forms would have had nasal-infixion, producing half-strong pasts like the formatives: talantë “collapsed”.
However, it seems the talat-stem verbs adopted the suffix a in their aorist tense forms early on: talta “collapse”. This in turn led to ea-presents in this verb class: taltëa < taltaya (PE17/186; PE22/164). The ultimate result is that the talat-stems tended to merge with the formatives into a single class of half-strong verbs. Speculative: This was especially true after ea-presents displaced ia-presents in the formative class.
The main remnant of the talat-stems are some rather exotic-looking half-strong pasts, such as sulumpë “lapped up”, the half-strong past of sulpa- (PE22/115). These more exotic pasts tended to be displaced by weak pasts: sulpanë. As such, some talat-stem verbs changed into weak verbs rather than half-strong verbs.
The causative verbs are one of the verb classes I refer to as weak verbs (= those with weak pasts). Tolkien himself used the term “weak verb” to refer to any non-basic verb (PE22/113). The causative verbs had the sense “to make X”, such as tulta- “to send (for), summon” = TUL+tā “(lit.) make come”. As such these verbs were invariably transitive with an obligatory direct object (except for cases where they merged with a formative verb, see above). There are two main types of causatives: the ta-causatives (ancient suffix -tā) and the rarer ya-causatives (ancient suffix -yā), such as tulya- “to bring, lead”.
The causatives can be described as “weak” because their tense-forms almost always involve some kind of suffix being added to the stem. In particular, these verbs are among those most likely to have weak pasts: tultanë “sent for”, tulyanë “brought”. This is also the verb class in which ea-presents originated: tultaya > tultëa “is sending for” (PE22/164), with the usual Quenya sound change of aya into ëa (the same change leading to cëantë < cayantē above). The ea-presents also appear for ya-causatives: tulyëa “is bringing” (PE22/164).
One unusual variation for the class of causative verbs is their future forms. In LVS, the futures of causative all have -auva instead of -uva: tultauva “will send for”, tulyauva “will bring” (PE22/164). They are the only verb class with this kind of future form. The only place these auva-futures appear are the 1969 notes in LVS, and as such they may be very late invention on Tolkien’s part. However, the causatives did have specialized future forms in QVS as well (PE22/117): tultáva, though these áva-futures were abandoned.
Another variation are the perfect forms of some causative verbs. There are many examples of causatives verbs whose perfect is formed from their (weak) past tense: tultanë → utultanië (PE22/157); I call these forms the “weak perfect”. Other times they have the simple perfect formation based directly on the verb stem, as described elsewhere in this course tulta- → utultië (PE22/164). It seems these were two competing tense formations in Tolkien’s mind and it is not clear when one should be used over the other. However, these weak perfect forms can be particularly useful with the (relatively rare) ya-causatives to make them more distinct: tulya- “to bring” → *utulyanië “has brought”.
Derivative verbs are those verbs that are not based on a verbal root. The derivative verbs were not originally verbs at all, but were based on nouns or adjectives that were transformed into verbs, either directly or with the addition of a suffix. Examples of the first include remba- “to net, trap”, which is the noun rembë “net” turned into a verb (VT42/12), and cúna- “to bend”, which is the adjective cúna “bent” turned into a verb (MC/223). Examples of suffixed derivative verbs include henta- “to read, scan” = KHEN “eye” + -ta (PE17/77) and móta- “to work” = MŌ “labour” + -ta (LR/373).
Because the derivative verbs are not based on verbal roots, they needed to “borrow” their conjugations from other verb classes. In Tolkien’s later writing, the derivative verb for which we have the most information is henta- (PE17/77), and its conjugations mostly agree with the simplified rules presented above for a-stem verbs.
One possible exception is the perfect tense. Some attested perfects of derivative verbs all have a “weak perfect” based on their (weak) past tense, such as hentanë → ehentanië (PE17/77). Others have a simple perfect like tengwanë → etengwië (VT49/48). This variation between simple and weak perfect may reflect a similar variation within the other major class of weak verbs: the causatives (see above).
Here is a set of tables illustrating the conjugations of the full set of Quenya verb classes. For completeness, the tables include some basic verbs and u-verbs. Verb forms marked with a * are not attested for that specific verb, but may be attested for other verbs from the same class.
|Present||nahtëa ¹||órëa ¹|
W/Y root (half-strong):
Causative Verbs (weak):
|Future||tultauva ¹||tulyauva ¹|
|Perfect||utultië ²||*utulyanië ²|
Derivative Verbs (weak):
|Gloss||“read, scan”||“read (writing)” ¹|
188.8.131.52 Conjugating A-stem Verb Classes: As a reminder, up until this point the course has treated all a-stem verbs (verbs with stems ending in a) as a single class which is conjugated as follows:
The more specific verb classes above diverge from these basic rules in a few circumstances:
Tolkien himself experimented with various systems of verb conjugation in QVS and LVS, as well as other documents written between those two. Given the variety of forms, other interpretations are possible. Therefore other systems of Neo-Quenya verb tenses might not agree with the above, especially if they try to incorporate the more specialized verb classes. The above is my attempt to put the attested examples into a single system.
184.108.40.206 Identifying Verb Classes: If you want to use the more advanced conjugations of the specialized verb classes, you need to be able to identify which class a verb belongs to. Some dictionaries (like Eldamo) may have this information, but not necessarily for every verb.
A-verbs: These are the easiest to identify. For the most part, if a verb ends in a preceded by a single consonant, it is an a-verb. If, however, the verb stem has a long vowel, it is usually some kind of derivative verb: cúna- “to bend” (from adjective cúna “bent”), móta- “to work” (from MŌ + ta).
Half-strong verbs: If you know a verb’s past tense, it is easy to identify whether it is a half-strong verb by whether it has a half-strong past: one that ends in -ntë or -nyë. If you don’t know the past, then an intransitive ta- or ya-verb is probably half-strong. Intransitive verbs favor ya-suffixes and transitive verbs favor ta-suffixes (PE22/115), so when in doubt assume ya-verbs are half-strong and ta-verbs are weak.
Weak verbs: A verb with a consonant cluster and a weak past is probably a weak verb. If it is a ya-verb or especially a ta-verb, it is likely a causative, and it is definitely a causative if its meaning is “make X”, as in tulta- “to summon = make come”. Verbs with a consonant cluster ending in consonants other than t and y are generally derivative verbs, such as remba-, tengwa-. The same is true of ta- and ya-verbs with weak pasts but without meaning “to make X”.
You need to distinguish true causatives like tulta- from derivative verbs because causatives use auva-futures like tultauva.
220.127.116.11 Advanced Topic: Adjectival Causatives: One final special case are the “adjectival causative” verbs, such as ninquita- “to whiten, make white” or tancata- “to fasten, make fast”. In QVS from 1948 Tolkien said these had weak pasts like other causatives (PE22/117), but examples in Tolkien’s later writing and a brief note in LVS indicate that in Tolkien’s later conception they were treated like half-strong verbs: “tā causative & tă when not made from ver[bal] stem” (PE22/164). These verbs therefore have weak pasts in QVS (ninquitánë, tancatánë) but half-strong pasts later on (ninquintë, tancantë).
The shift from weak to half-strong (assuming this is what happened) appears to have been motivated by the need to explain Sindarin past tenses of a-stem verbs, which consistently show -ant in Sindarin. The clearest explanation for these Sindarin pasts is that they were generalized from past tenses like tancant(e) > S. tangant after final vowels were lost in Sindarin (PE17/44).
Here are some vocabulary words for various actions:
Based on the tenses provided, fill in the remaining verb tense conjugations (past, perfect, present, future) and identify the verb class:
As vocabulary practice, translate the following to Quenya:
Answers are in Answer Key 8.1 at the end of this chapter.
This section discusses how to ask questions in Quenya.
The basic question marker in Quenya is the element ma. This is the basis for several interrogative pronouns and question words in Quenya:
We don’t know the words that Quenya uses for “when” and “why” in questions, but the neologisms malumë and manan are popular choices for (Neo) Quenya writing. All the above words are used to form questions:
The question word replaces the subject or object of the phrase as appropriate, and is placed first. Many of the question words are simply the interrogative particle ma with some noun case ending: ma + -ssë “where = at what”, ma + -llo “whither = from what”, etc. Speculative: We have no direct evidence, but presumably mana can also be used as an interrogative adjective:
Note that English also uses “who, what, which” as relative pronouns to create subordinate clauses, but Quenya has a separate set of pronouns for this purpose: i, ya, ye. These were discussed briefly in Chapter 1, Section §1.2.6, and are discussed in more detail in Chapter 10, Section §10.1. Thus if you want to say “I want to see the man who came yesterday”, you don’t use man but rather say merin cenë i atan i túlë nöa.
Complex questions in Quenya can be formed using the question words given above. Simple yes/no question can be expressed by intonation alone, but you may also begin the phrase with the interrogative particles ma to indicate a question. The particle ma is almost always used in writing where there is no tone to indicate a question. For example ma túlalyë? “are you coming?” is simply the regular declarative phrase túlalyë “you are coming” with ma added to indicate it is a question. As another example, the statement i elda halla “the elf [is] tall” can be turned into the question ma i elda halla? “is the elf tall?” by adding ma to the beginning.
The usual word order is ma + subject + verb (or verb with pronominal suffix), but sometimes other words (such as adverbs) may be placed in various locations in the phrase for emphasis, but they can never be placed between ma and the subject. Thus the following are all possibilities:
But the following is not possible:
Finally, like other particles á, vá, and ui, the interrogative ma might receive a pronominal subject suffix if the verb is omitted, such as: Ma túlalyë? Mantë? “Will you come? Will they?” This formation might also be used when responding to a statement with a brief question:
Finally, here is some additional question-related vocabulary:
The interrogative element ma is mentioned in several places (PM/357; PE17/162), but the longest discussion of questions appears in Late Notes on Verb Structure (LVS) written in 1969 (PE22/160-161). Much of the above discussion is based on that source. As for question words other than ma, they are drawn from a variety of documents (LotR/377; PE22/124; PM/395), whereas !malumë “when” (ma + lúmë “what time”) and !manan “why” (= dative mana + -n “for what”) are popular neologisms for otherwise unattested question words.
Here are some vocabulary words having to do with happy emotions:
Vocabulary words having to do with questions and answers include:
Translate the following to English:
Translate the following to Quenya:
Answers are in Answer Key 8.2 at the end of this chapter.
The previous section talked about how to formulate questions. This section discusses how to answer questions, in both words and gestures.
As mentioned in the previous section, a yes/no question can be formulated by adding a ma at the beginning of the phrase: ma tuluvalyë? “will you come?” In speech, a question can also be indicated by tone alone: tuluvalyë? Tolkien didn’t describe the question intonation for Elvish, but presumably it involves a rising tone at the end of the phrase, like questions in English (and many other languages).
The answer to a yes/no question depends on whether it is a question of a fact or a request for action. A question of fact is generally answered with either ná “yes = it is” or ui “no = it is not”. Thus if you were asked ma i elda halla? “is the elf tall?” your answer would likely be ná for “yes” or ui for “no”, or possibly nás “he/she is” or uis “he/she is not”.
An answer to a request or a command uses either sá/asa [þ] “will” or vá “will not”. Thus in answer to ma tuluvalyë enar? “will you come tomorrow?” you might answer sá “yes = I will” or vá “no = I will not”. The same responses may be given to a command: á tulë enar “come tomorrow” could be answered either sá for “I will” or vá for “I won‘t”.
Of sá and asa, sá is for simple agreement = “I will”, while asa (especially if used alone) is more emphatic, meaning something like “of course I will”. The element vá can receive pronominal suffixes as appropriate: ványë “I won’t”, válmë “we (excluding you) won’t”. It is also possible to say válvë “we (including you) won’t” in answer to a debate among friends or allies: ma mahtuvalvë i orcor? “will we fight the orcs?”, with the response válvë “we won’t”. Vá is a rather forceful response; a softer response would be uin asa “I do not agree”.
The short form sá “[I] agree” cannot receive pronominal suffixes, but longer asa can: asanyë “I will”, asalmë “we (but not you) will”, asalvë “we (and you) will”. Both vá and asa can be used as semi-verbs before another verb to indicate disagreement or agreement: válmë tulë “we won’t come” vs. asalmë tulë “we agree to come”. The verb can also be in present or future but not in the past, since one cannot refuse or agree to an action that has already occurred. Thus ványë túla “I am not coming (now)” or ványë tuluva “I will not come (in the future)”.
Finally, as discussed in Chapter 5, Section §5.5.3, when vá is used of another party rather than yourself, it indicates prohibition, not refusal. Hence vásë tulë means “he/she is not to come”, and vályë mates “you are not to eat it”. This can be further strengthened to a negative command: áva mates! “don’t eat it!”, or isolated áva “don’t” or strongest of all avá! “don’t!”, the last used only in anger or urgently.
18.104.22.168 How do we know this? As with questions in general, the main source for the information above is the Late Notes on Verb Structure (LVS) written in 1969, which had lengthy discussions of vá and sá, though the latter appeared only in its more archaic form þá (PE22/161-3, 165-6). These documents also described ná “yes, it is so” (PE22/166), but the negation of fact in LVS was lá (PE22/160). In keeping with this course’s preference for u-negation, negative lá “no, it is not” has been replaced by ui in the above discussion based on other documents from the late 1960s (VT49/28); see the discussion in Chapter 5, Section §5.5.4 for a more information on the choices associated with the two forms of negation.
Likewise, LVS examples are based on the paradigm of negative adverbs rather than the approach preferred in this course of negative semi-verbs. Where the above has ványë tuluva “I won’t come”, LVS had vá tuluvanyë with the pronominal suffix on the main verb and not the negative element (PE22/162). If you prefer the negative adverb approach, you should place the pronouns on the main verb, with the exception of phrases like asalvë tulë “we agree to come”, because asa- (aþa-) remained a semi-verb (or even a full verb) within the LVS paradigm.
Elves often accompanied responses to questions with some kind of gesture, but the gestures used by Elves may differ from those used by Men. The two different ways of saying “no” had distinct gestures. A simple ui “no” to questions of fact was accompanied by a shaking of the head with pursed lips, but a vá “no” in response to a request was accompanied by a jerking back of head. These descriptions of head movements are from notes on negation written around 1959 (PE17/145).
Hand gestures in greeting were also different among the Elves. According to Tolkien, Elvish hand gestures include:
The above are excerpted from Tolkien’s notes in Eldarin hands and fingers written around 1968 (VT47/9, 13). Tolkien further added:
Extension of the fingers modified the significance [of a gesture]. The gesture of a receiver or asker, if the fingers and thumb were opened, indicated distress and urgency of need or poverty. The gesture of prohibition in the same way was made more hostile and threatening, indicating that if the command was not immediately obeyed force or weapons would be used (VT47/9).
Here are some vocabulary words having to do with unhappy emotions:
Translate the following to English:
Translate the following to Quenya:
Answers are in Answer Key 8.3 at the end of this chapter.
The best known Quenya text is Galadriel’s Namárië poem, which we’ve already partially examined in Chapter 3, Section §3.5. There is another version of that text in non-poetic form which Tolkien said was of “a clearer and more normal style”. It was thus basically a prose version of the poem more like ordinary Quenya speech. This prose version was published in the 1967 song book The Road Goes Ever On (RGEO/58-59), and we will examine its first few lines here.
First, some vocabulary:
The first four phrases of the prose text are:
Ai! Lassi lantar laurië súrinen,
yéni únótimë ve aldaron rámar.
Yéni avánier ve lintë yuldar
lissë miruvóreva mí oromardi ...
Broken down into its basic elements:
Ai! Lass-i lanta-r laur-ië súri-nen
Ah! Leaf-(pl.) fall-(pl.) golden-(pl.) wind-by
yén-i únótim-ë ve alda-r-on ráma-r
Year-(pl.) uncountable-(pl.) as tree-(pl.)-of wing-(pl.)
Yén-i avánie-r ve lint-ë yulda-r
Year-(pl.) (has departed)-(pl.) as swift-(pl.) drink-(pl.)
lissë miruvóre-va mí oromard-i
sweat mead-of in-the (lofty hall)-(pl.)
First phrase: Ai! Lassi lantar laurië súrinen. The text begins with the interjection Ai! “Alas!”, a cry of lamentation. The subject of the first phrase is lassi “leaves” with the verb lantar “fall” in the plural to agree with it. This is followed by the adjective laurië, the plural of laurëa “golden” again in agreement with plurality of the subject. Since the adjective follows the verb, it is being used adverbially: “golden[ly]”, the colour of the leaves perhaps indicating they are falling in autumn. This is followed by instrumental súrinen “by (means of) wind”. The i in this last word is because the stem form of súrë “wind” is súri- and it is not a plural form; the plural would be súrínen “by (means of) winds”. Thus “Ah! Leaves fall golden[ly] by (means of) wind”.
Second phrase: Yéni únótimë ve aldaron rámar. The second phrase is descriptive of the falling leaves and the circumstances of their fall. In particular they have fallen over yéni únotimë “long years uncountable”, where únotima “uncountable” is plural to agree with its noun. A yén is an Elvish long year, equal to 144 solar years, a common time unit among the immortal Elves. The remainder of the phrase also describes the leaves: ve aldaron rámar “like trees’ wings”. Thus “long years uncountable as trees’ wings”. In context, the falling leaves are like the wings of trees and are a metaphor for the numberless years.
Third phrase: Yéni avánier ve lintë yuldar. The subject of the third phrase is yéni “long years” with the verb avánier “have departed”, plural to agree with the subject. Their passing is ve lintë yuldar “like swift drinks”, or put more poetically “like swift draughts”. Here linta “swift” is made plural to agree with yuldar “draughts”. Thus “long years have departed like swift draughts”, or more appropriately for years: “long years have passed like swift draughts”.
Fourth phrase: Lissë miruvóreva mí oromardi. The fourth phrase is descriptive of the draughts. In particular, they consist of lissë miruvóreva “of sweet mead”, and were located mí oromardi “in the lofty halls”. Recall that preposition mí is an elided form meaning “in the” = mi “in” + i “the”; see Chapter 6, Section §22.214.171.124. Thus the phrase means “of sweet mead in the lofty halls”.
Putting it all together:
Ah! Leaves fall golden[ly] by (means of) wind,
long years uncountable as trees’ wings.
Long years have passed like swift draughts
of sweet mead in the lofty halls ...
Tolkien’s actual (and somewhat loose) translation was:
Ah! Like gold fall the leaves in the wind,
long years numberless as the wings of trees.
The years have passed like swift draughts
of the sweet mead in lofty halls ...
Questions and Answers:
Conjugations of Specialized Verb Classes: