Chapter 11 - Compound Tenses, Subjunctive, More Noun Classes

This chapter consists of advanced topics which are necessarily speculative in many particulars. As such, it has no exercises.

11.1 Compound Tenses

English uses auxiliary verbs when forming many of its verb tenses such as the present continuous (= imperfect), perfect and future tenses: “I am eating”, “I have eaten”, “I will eat”. English can combine verb tenses (past, present, future) with verbal aspects (imperfect, perfect) for various shades of meaning. For example, a future imperfect: “I will be eating” or a past perfect/pluperfect “I had eaten”. Quenya has some similar constructions, which in one place Tolkien called the Quenya compound tenses (PE22/122). These are formed by combining the various suffixes used in Quenya tense formation.

This section attempts to combine ideas on the compound tenses from the Quenya Verbal System (QVS) written in 1948 with the later Quenya verb system of 1950s and 60s. As such, the entire section is speculative and will involve a larger than normal amount of technical details.

11.1.1 Tense vs. Aspect

Many verb systems for various languages, including Quenya and English, use various factors when determining verb conjugations. Two common factors are tense (time) and aspect (completeness):

English conjugations have two basic tenses: past (“he/she ate”) and not-past (“he/she eats”). The other factors are expressed by various auxiliary verbs:

Quenya instead has five basic conjugations, none of which use auxiliary verbs:

Strictly speaking, máta and amátië are the present imperfect and the present perfect, as both describe actions as they affect the present. Quenya can combine these five basic tenses with other verbal suffixes for different shades of meaning.

11.1.2 Compound Pasts

Quenya has three compound tenses that put other tenses into the past by adding the past suffix -në:

These three past compounds are formed by taking the present/imperfect, perfect and future tenses and adding -në to shift the frame of the action into the past. In the case of perfect + -në and future + -në, the extra suffix triggers prosodic lengthening, as discussed in Chapter 6, Section §6.1.3, so that the vowel before the suffix is lengthened: amátiénë and matuvánë. As their use evolved over time, the augment in the past perfect became optional: mátiénë. Furthermore, the past-future suffix -uvanë came to be perceived as a single suffix, so that the prosodic lengthening shifted forward: matúvanë.

The a-stem verbs whose present imperfects were formed with -ëa likewise see prosodic lengthening in their past imperfects: fárëánë “was hunting”, lantëánë “was falling” (PE22/157). How do we know this? The past imperfect and the past future tenses appearing above are drawn from the Quenya Verbal System (QVS) written in 1948 (PE22/105, 109). The past imperfect was also mentioned in Late Notes on Verbs Structure (LVS) from 1969, where Tolkien retained the same mátanë form as in QVS after a brief flirtation with an alternate past imperfect form mátantë (PE22/157).

QVS also had a different past perfect/pluperfect form mantelyanë based on a 1948 “long perfect” form amantelyë, a variant of the ordinary perfect amátië. Since the long perfect was abandoned, I think it is best to use the ordinary perfect as the basis for the past perfect instead. The 1948 past perfect dropped the augment, so I think it is best to assume the augment was dropped in mátiénë forms as well.

Note that it is not entirely clear to what degree Tolkien retained compound tenses in Quenya as he imagined it in the 1950s and 60s; only the past imperfect form mátanë was mentioned (PE22/157). However, Tolkien did imagine a similar system of past and future active participles based on adding -(i)la to the perfect and future tense (PE22/155), an updated version of the system in QVS which used an earlier suffix -lya for active participles (PE22/107-108). This lends support to the notion that compound tenses remained possible in Tolkien’s later version of Quenya, even if we don’t know exactly how they were formed. As for the past/future active/passive participles, see the discussion in Chapter 9, §9.1.3.

11.1.3 Compound Futures

Quenya has two compound tenses that put other tenses into the future by adding the future suffix -(u)va:

Note that there is no “past in the future” tense, because the future perfect fulfills this function: yá anyuvalvë i osto, apátiévalvë ter auri neldë “when we arrive at the city, we will have walked for [through] three days”. There is, however a (rare) past future perfect, used in hypothetical past statements much like English “would”:

For example: quíta cestanes, túliévanen “if he/she asked, I would have come [afterwards]”. See section 11.2 Subjunctive for further discussion. How do we know this? As in the previous section, the future imperfect and the future perfect are based on the Quenya Verbal System (QVS) written in 1948. In QVS, the future imperfect was actually based on the 1948 active participle: matalyuva using the suffix -lya (PE22/109). Since that suffix was abandoned, I think it is better to use the normal present continuous/imperfect form as the basis instead: mátauva, analogous to past imperfect mátanë.

The future perfect formation above is drawn directly from QVS (PE22/105), but in that document the future suffix was -(u)va, taking the form -uva only after basic verbs (matuva “will eat”) but having the form -va with derived verbs (ortáva “will raise”). In Tolkien’s later writing, the u was a more intrinsic part of the future suffix, and it was not lost when it was added to a-stem verbs: oryuva “will rise” (PE22/157), ortauva “will raise” (PE22/159). This makes the future perfect form (a)mátiéva somewhat dubious, but we don’t know how the perfect suffix -ië and the future suffix -uva would combine in Tolkien’s later versions of Quenya. Until we get further information, I recommend sticking to (a)mátiéva for now, along with the (rare) past future perfect (a)mátiévanë which is also directly based on QVS.

11.1.4 Vocabulary: Mind

Here are some vocabulary words having to do with the mind and thought:

11.1.5 Section Summary

The compound tenses combine basic verb tenses with -në to put them in the past or -(u)va to put them in the future:

11.2 Subjunctive

Most languages, including Quenya, have a mechanism to describe things that may or may not be real. The linguistic term for this kind of verbal mood is the subjunctive. English typically indicates this kind of “unreal” mood with auxiliary verbs like “might” or “would”, as in “he might help” or “he would help if you asked”. Quenya marks unreality with various adverbs and conjunctions.

These hypothetical expressions are another topic where Tolkien often changed his mind on exactly how Quenya worked. The ideas discussed here are pieced together from documents separated by decades, many of them mutually incompatible, making this discussion rather speculative in comparison to other sections in this course. This discussion draws heavily on the examples Tolkien gave, but some of these examples have been modified to make them more internally consistent.

11.2.1 Adverbs of Uncertainty

Quenya’s basic adverbs and conjunctions of uncertainty are:

These elements are used in various ways for conditional, hypothetical or counterfactual statements. Also of interest are a pair of adverbs/conjunctions meaning “then” but with slightly different connotations:

The three-fold division between if/may/might is a well-established idea in Quenya, dating back to the Early Quenya Grammar of the 1920s, which likewise contrasted the conjunction “if” against a weak uncertainty “may” and a strong uncertainty “might” (PE14/59). This idea reappeared in the Quenya Verbal System (QVS) written in 1948 (PE22/120-121) and Common Eldarin: Verb Structure (EVS2) from early 1950s (PE22/138-139). Each time Tolkien discussed it, however, the exact set of adverbs changed, along with the details on how they were used.

The paradigm above is based on a pair of roots √KE “may (be), have a chance to” and √QUI(S) “suppose” appearing in the late 1960s (PE22/158; VT49/20). I choose this paradigm because (a) it is fairly comprehensive and (b) appears in Tolkien’s later writings. However, it is not the only system Tolkien experimented with towards the end of his life. For example, in the Ambidexters Sentence discussed in Chapter 7, Section §7.5, Tolkien used for “if” rather than qui as suggested above.

11.2.2 Hypothetical Statements

Any statement about the future is necessarily hypothetical, because the future has not yet come:

A greater degree of uncertainty can be indicated by putting the uncertainty particle before the verb, or with quí if more uncertain:

The more uncertain adverb quí may also be used if the result is surprising or counterintuitive:

These particles of uncertainty can also be used in the present or past if the actual event is unknown () or unlikely (quí):

The above is extrapolated from two examples: lá caritas alasaila cé nauva “not doing this may prove unwise” (PE22/154) which uses la-negation, and carë mára quí tyarë naxa “doing good may cause evil” (PE22/154), appearing above with the more common word ulco for evil and the gerund form carië rather than (infinitive?) carë. Both these examples indicate that the uncertainty particle appears before the verb. This word order is confirmed in Common Eldarin: Verb Structure from the early 1950s:

There existed, however, already in Eldarin certain adverbial particles, of possibility, remoter possibility, supposition, and wish, that could be used for greater precision. These were placed in close connexion with [the] verb. Most frequently they appear to have immediately preceded it (PE22/138).

A similar system appeared in the Quenya Verbal System (QVS) from 1948 and Common Eldarin: Verb Structure (EVS2) from the early 1950s, except the uncertainty particles were ai and au(ve) instead of and quí (PE22/120, 139).

11.2.3 Conditional Statements

When one thing is a natural consequence of another, the usual syntax is to use “when”:

The conditional qui “if” may also be used and this has virtually the same meaning:

If, however, the result clause is in the future, the statement becomes hypothetical:

The conditional clause may itself be in the future, in which case it indicates the required event is also expected to occur at some future point:

The consequential nature of the result clause can be emphasized by adding en or san “then”, with slightly different meanings:

The result clause can be made more uncertain by using :

If the conditional clause is itself unlikely, the more emphatic quíta is used:

Both clauses can be marked as more uncertain at the same time:

Finally, a repeating conditional occurrence can be indicated by quiquië “whenever”:

As in the previous section, two of the examples above are from notes from the late 1960s, with some minor modifications. The rest of the above is based on notes from Quenya Verbal System (QVS) written in 1948 (PE22/120-122), but the examples in the QVS quotes below have been extensively rewritten because:

Regarding the use of qui “if” in the sense “when” [with qe from QVS replaced by later qui, and íqa “when” replaced by later ]:

Note in general statements where there is no real supposition (although a second event is conditioned by a previous one), i.e. when “if” is really equivalent to “when, whenever”, then in Quenya qui is seldom used. It is possible to say qui tulis, (san) inye tulë yú; qui túles, (san) inye túlë yú “if/when he comes, I come too; if/when he came, I came too”. But usually “when, whenever” are used: ... yá issë tulë, (san) inye tulë “when hé comes, Í come” [with accents in the English indicating emphatic pronouns] (PE22/121).

Regarding the basic conditional expression [again with qe replaced by qui]:

The first simple pattern is expressed as in English: qui cestas nin, tuluvan [“if he asks me, I shall come”]. The aorist is used in the conditional clause since the aorist has no special time reference and qui alone marks the clause as hypothetical. The future can, however, in Quenya be used also after qui: qui cestuvas ni, tuluvan. The latter is more nearly equivalent to “when he asks me”; it is more particular: “if on some future occasion he asks me, then on that occasion I shall come” (PE22/120).

Regarding more unlikely hypotheticals [with aiqe from the original replaced by later quíta]:

The second, more “remote”, pattern is expressed in Quenya (as far as the equivalence goes: the correspondences are not exact) by quíta [in the original: ai + qe] with the future. “If he asked me/were to ask me, I should come” quíta cestuvas ni, tuluvan. The particle en “in that (future) case” can be also inserted: en tuluvan (PE22/120).

The two different adverbs for san and en “then” are also sourced from QVS, as seen in the (modified) examples above. Tolkien specifically denoted en as pointing to the future = “then (future)”, while the use of san to mean “then (consequently)” is speculation on my part. Both en and san reappeared in documents from the early 1950s as well (PE22/131; PE22/140)

11.2.4 Counterfactual Statements

Conditional statements can also be used in the past, but in such cases the conditional event is generally known to be false, describing what hypothetically could have occurred but in actuality did not. Therefore, such past conditionals are more accurately described as counterfactuals, describing things that did not occur. As such, in Quenya they require the stronger “less likely if” quíta, with the result clause in the perfect:

There is a specialized but rarely used past future perfect which can be used for past counterfactuals as well:

When used in this way, the past future perfect emphasizes that the (hypothetical) result clause would have happened after the counterfactual conditional clause.

The syntax above on counterfactuals is mostly drawn from Late Notes on Verb Structure (LVS) written in 1969, where Tolkien said:

quí or quíta is used when (a) supposition is known to be not in accord with fact, as in “if he had come (as he did not), I should have welcomed him”, or “if you had not returned (as you did), I should have been angry”. quí(ta) la tuldes, náne márie (nin) “[if he had not come] it was well to me (I was glad)” ... The unreality can be more explicitly expressed by insertion of , may be, into a negative protasis [= the first clause of the sentence] or céla, maybe not, into a positive. quíta céla tuldes (PE22/158).

There are a couple of challenges in using this syntax in conjunction with other grammar rules advocated by this course. In particular, this syntax was coined in the brief 1969 period where Tolkien used la-negation rather than u-negation. That makes forms like céla questionable, but I have no good alternative. The other issue is that this entire paragraph was marked through by Tolkien, but he did not write a replacement.

The notion that past conditionals which are functionally counterfactuals require a greater uncertainty marker is itself a somewhat well established idea, however. The same basic formula appears in QVS from 1948, but the uncertainty particles are different:

In the past tense the possibility of wish or supposition known to be "unreal" arises; and the patterns “if he asked me, I came; when/if he (had) asked me, I came” [consequential] are naturally differentiated from “if he had asked me, I should have come; I wish he had asked me; would that he had asked me!” [hypothetical] ... The “unreal” patterns are in Q., which never developed inflexional subjunctives or conditionals, expressed as follows. The particle ai is used in the conditional clause, usually joined to qe “if” in the latter en, in that case, is used with or without ai.

Tolkien then gave several examples, with subject prefixes “he” and ni· “I” rather than suffixes -s and -nye:

In this examples, qe is “if” rather than later qui, illume is “whenever” rather than later quiquië, aiqe is the “stronger if” rather than quí(ta) and ela is the element emphasizing unreality rather than céla. Thus the Tolkien was using the same basic syntax 20 years earlier, even if all the individual adverbial elements changed. I personally would “modernize” these examples as:

Nevertheless, the system I have pieced together here is drawn from some very scattered evidence, and it is therefore rather speculative in nature. Of the above, the use of céla “maybe not” is the most questionable, which is why I have omitted it from the main section and only discussed it in the speculative section.

11.2.5 Vocabulary: Supernatural

Here are some vocabulary words having to do with the supernatural:

11.2.6 Section Summary

Quenya’s basic adverbs and conjunctions of uncertainty are:

11.3 More Noun Classes

In this section we will revisit noun declensions. Up until this point, we have grouped nouns into three basic declensions: vocalic nouns, e-nouns and consonantal nouns. However, the system of declensions used by Tolkien was a bit more complex. In a document labeled the Declension of Nouns (DN) from the early 1930s (PE21/1-41), Tolkien grouped nouns by (a) their final vowel or consonant and (b) whether or not they triggered prosodic lengthening when suffixes were added. Although the noun cases from this document underwent extensive revisions in Tolkien’s later writings, the basic classification seems to have remained valid.

As with the discussion of advanced verb classes in Chapter 8, Section 8.1, the information below can be considered an advanced topic. These advanced declensions draw heavily on Raccoon’s research in his Tenguesta Goldorinwa, most notably his examination of Noun Declensions, though I do not agree with him on all the particulars. All of the following can be considered optional. If you stick with the simpler system of noun cases described earlier in the course, you will still be correct most of the time.

Note that in the tables below, the “long” nouns are those that trigger prosodic lengthening (see Chapter 6, Section §6.1.3), namely those with three or more syllables whose second-to-last syllable is light: having a short vowel followed by zero or one consonant. For example, for purposes of prosodic lengthening lícuma “candle” counts as a long a-noun, but mahalma “throne” and tincotéma “t-series” do not. In addition, entries whose form significantly differs from normal declensions are marked in bold with a footnote explaining the variation.

11.3.1 a-nouns and o-nouns

The a-nouns and o-nouns are those ending in the vowels a and o. They are the best representatives of the ordinary declensions for vocalic nouns, differing only in the genitive case. In the ancient form of these nouns, the vowels were originally long and in DN Tolkien referred to these classes as the ā and ō declensions. a-nouns: The declensions for a-nouns are represented by short cirya “ship” and long lícuma “candle”, as well as coa “house” which represents the small class of oa-nouns which (speculative) have a specialized genitive form: cuo.

a-nounSg.Du.Part. Pl.Pl.
Genitiveciryo ¹ciryatociryalionciryaron

oa-nounSg.Du.Part. Pl.Pl.
Genitivecuo ¹coatocoalioncoaron

long a-nounSg.Du.Part. Pl.Pl.
Nominativelícumalícumatlícumáli ²lícumar
Dativelícumanlícumantlícumálin ²lícumain
Genitivelícumo ¹lícumáto ²lícumálion ²lícumáron ²
Possessivelícumáva ²lícumatwalícumalívalícumaiva
Instrumentallícumánen ²lícumantenlícumalínenlícumainen

The replacement of the final a with o in a-noun genitives dates as far back as the Declension of Nouns from the early 1930s (PE21/4), and reappeared in the Plotz Letter of the late 1960s (VT6/14). It is the result of the ancient sound change whereby final ā + ō > ō (PE19/63). Genitive forms like cuo for coa “house” are more speculative, however. Such words were based on ancient forms like kawā, which became coa over Quenya’s history of sound changes (PE17/108; PE19/63). However, if the ancient genitive form was kawā+ō > kawō, this would develop into cuo (PE19/63). o-nouns: The declensions for o-nouns are represented by short rocco “horse” and long ciryamo “sailor”.

o-nounSg.Du.Part. Pl.Pl.
Genitiverocco ¹roccotoroccolionroccoron

long o-nounSg.Du.Part. Pl.Pl.
Nominativeciryamociryamotciryamóli ²ciryamor
Dativeciryamonciryamontciryamólin ²ciryamoin
Genitiveciryamo ¹ciryamóto ²ciryamólion ²ciryamóron ²
Possessiveciryamóva ²ciryamotwaciryamolívaciryamoiva
Instrumentalciryamónen ²ciryamontenciryamolínenciryamoinen

The form of o-noun genitives is confirmed by phrases like Indis i Ciryamo “Wife of the Mariner” (UT/8). The example Tolkien used in the Declension of Nouns from the early 1930s was ondo “*of stone”, although in this 1930s document it had a variant genitive form onduo that was not seen later (PE21/5). This course uses rocco rather than ondo for its declension of o-nouns because ondo would have less-representative dual forms like ondu as opposed to the more typical roccot; see Chapter 5, Section §5.2.1 for a discussion of duals.

11.3.2 e-nouns

The e-nouns differ from other vocalic nouns primarily in their plural forms, which use i-plurals rather than r-plurals, with the exception of e-nouns that end in potentially awkward combinations -ië and -lë. Because the ancient form of these nouns had long vowels, in DN Tolkien referred to this class as the ē declensions. The declensions for e-nouns are represented by short lassë “leaf” and long telumë “roof”, as well as tyalië “play, mirth” to represent ie-nouns.

e-nounSg.Du.Part. Pl.Pl.
Nominativelassëlassetlasselilassi ¹
Dativelassenlassentlasselinlassin ¹
Genitivelassëolassetolasselionlassion ¹
Possessivelassevalassetwalasselívalassíva ¹
Instrumentallassenenlassentenlasselínenlassínen ¹

long e-nounSg.Du.Part. Pl.Pl.
Nominativetelumëtelumetteluméli ²telumi
Dativetelumentelumenttelumélin ²telumin
Genitivetelumëo ¹teluméto ²telumélion ²telumion
Possessiveteluméva ²telumetwatelumelívatelumíva
Instrumentalteluménen ²telumententelumelínentelumínen

ie-nounSg.Du.Part. Pl.Pl.
Nominativetyaliëtyaliettyaliéli ²tyalier
Dativetyalientyalienttyaliélin ²tyaliain ³
Genitivetyaliëo, tyaliéno ¹tyaliéto ²tyaliélion ²tyaliéron ²
Possessivetyaliéva ²tyalietwatyalielívatyaliaiva ³
Instrumentaltyaliénen ²tyaliententyalielínentyaliainen ³

Most of the declensions for lassë “leaf” appear in the Plotz Letter of the late 1960s (VT6/14). The prosodic lengthening for long e-nouns is easy to deduce, but the stress pattern for the genitive singular is unclear. There is an example Óromèo (accents marking stress) in the Declension of Nouns from the early 1930s which indicates that the e would be stressed (PE21/41), though in that document it seems to receive the secondary rather than primary stress.

The dative, possessive and instrumental plurals of ie-nouns are also problematic. In declensions from the early 1930s, these nouns all show iai (PE22/7, 47, 50) as given above: tyaliain, tyaliaiva, tyaliainen. These forms are the result of a 1930s sound change whereby iei > iai (PE22/7), or more exactly ei > ai after ı̯/y (PE21/6). DN gave similar examples for the declensions of short ie-nouns: niainen the instrumental plural of nië “tear” (PE21/6). However, there is strong evidence that Tolkien abandoned the iei > iai phonetic development, because in notes from the late 1960s he said instead that ei > é after y, as in *skeyeiti > xiétë “passing, impermanent” (PE22/155).

I think that the 1930s iai-forms might be retained by assuming there was an exaggerated sound change only for plural noun cases of ie-nouns to better differentiate them from their singular forms. If, however, you are not comfortable with this and prefer to use the known phonetic rules from the late 1960s, then the dative, possessive and instrumental plurals for ie-nouns would be identical to their singular forms: tyalien, tyaliéva, tyaliénen.

11.3.3 i-nouns and u-nouns

The i-nouns and u-nouns differ from other vocalic nouns in that they frequently have long final vowels in their inflected forms, as if they underwent prosodic lengthening even though they are short nouns rather than long nouns. Because the ancient form of these nouns had long vowels, in DN Tolkien referred to these classes as the ī and ū declensions.

This class is relatively small, mostly consisting of feminine nouns ending in -i and masculine nouns ending in -u. The plural declensions for ru-nouns differ slightly from u-nouns with other consonant; the same is true to a lesser degree for ri-nouns vs. other i-nouns. These declensions are represented by tári “queen”, ainu “angel”, and heru “lord” to represent ru-nouns.

i-nounSg.Du.Part. Pl.Pl.
Nominativetáritárittáríli ¹tárir
Dativetárintárinttárílin ¹tárin ²
Genitivetáriotáríto ¹tárílion ¹tárion ³
Possessivetáríva ¹táritwatárilívatáríva ¹
Instrumentaltárínen ¹tárintentárilínentárínen ¹

u-nounSg.Du.Part. Pl.Pl.
Nominativeainuainutainúli ¹ainur
Dativeainunainuntainúlin ¹ainuin
Genitiveainuoainúto ¹ainúlion ¹ainúron ¹
Possessiveainúva ¹ainutwaainulívaainuiva
Instrumentalainúnen ¹ainuntenainulínenainuinen

ru-nounSg.Du.Part. Pl.Pl.
Nominativeheruherutherúli ¹heruvi ²
Dativeherunheruntherúlin ¹heruin
Genitiveheruoherúto ¹herúlion ¹heruion ²
Possessiveherúva ¹herutwaherulívaheruiva
Instrumentalherúnen ¹heruntenherulínenheruinen

The uses of long final vowels in short i-nouns and u-nouns is a feature of their declensions from the early 1930s (PE21/14-15, 51). There is some evidence for this in Tolkien’s later writings as well, notably singular possessive táríva “of a queen” rather than **táriva (PE17/76).

The avoidance of repeated r’s (r-r) is a feature of Quenya (PE19/99). The best evidence for this in noun cases are the examples heruvi and heruion, the plural and genitive plural of heru “lord” (SD/246, 290), rather than expected **herur and **herúron. The genitive plural form tárion also appears in DN (PE21/14). However, I think the plural of ri-nouns would still use r as in tárir “queens”, because (a) otherwise they would not have a distinct plural and (b) the sequence rir does appear in verb inflections, for example: carir “do [pl.]” (PE17/132; WJ/391).

11.3.4 Semivocalic Nouns

In addition to feminine i-nouns and masculine u-nouns ending in ancient long ī and ū, Quenya had a set of more general nouns ending in ancient short ĭ and ŭ. Tolkien called these semivocalic nouns or the ĭ and ŭ declensions (PE21/3, 10-13). In the simplex, the ancient ĭ became e and ancient ŭ became o, as in sírë (síri-) “river” derived from ancient *sīrĭ and malo (malu-) “pollen” derived from ancient *malŭ. Inflected forms generally restore the ancient vowel, for example: singular instrumental sírinen “by means of a river” and singular possessive maluva “pollen’s”. However the short vowels ĭ, ŭ could sometimes become semivowels y, w provided that the preceding consonant was suitable. y-stem and e/i-noun: For purposes of discussion, this course refers to nouns whose consonants are suitable for combination with y as the y-stem nouns. These nouns show y in the dative and dual, as well as the genitive singular and (speculative) dual. Where the stem’s consonant or consonant cluster is not suitable for y, however, then i is seen throughout the noun declensions; this course refers to these as e/i-nouns. These declensions are represented by sírë (síri-) “river” and lindë (lindi-) “singing, musical sound”.

y-stemSg.Du.Part. Pl.Pl.
Nominativesírësiryat ¹sírilisíri ²
Dativesiryen ¹sírintsírilinsírin
Genitivesiryo ¹siryato ¹sírilionsírion

e/i-nounSg.Du.Part. Pl.Pl.
Nominativelindëlinditlindililindi ²
Dativelindien ¹lindintlindilinlindin

The tables above are mostly based on the Declension of Nouns (DN) from the early 1930s (PE21/10) with forms modernized to fit the case system of the 1950s and 60s. In DN, the stem form of “river” was siri- with a short i in the first syllable, and lindi- was glossed “pool” rather than having its later meaning “musical sound”. The descriptions in DN are the best information on the semivocalic declensions, but the dual form siryat appeared in notes from the late 1960s (VT47/11) and is the best evidence that something like these semivocalic declensions remained valid. w-stem and o/u-noun: For purposes of discussion, this course refers to nouns whose consonants are suitable for combination with w as the w-stem nouns, and these nouns show w in the dative and genitive singular, as well as several of plural forms. There is a special case for w-stem nouns with a preceding c consonant: this course refers to them as qu-nouns, where the resulting cluster cw is written qu which also appears in the plural. Nouns which cannot form clusters with w show u instead in all declined forms. These declensions are represented by malo (malu-) “pollen”, ranco (rancu-) “arm” and luppo (luppu-) “lump”.

w-stemSg.Du.Part. Pl.Pl.
Dativemalwen ¹maluntmalulinmalwin
Genitivemalwo ¹malutomalulionmalwion

qu-stemSg.Du.Part. Pl.Pl.
Nominativerancorancutranculiranqui ²
Dativeranquen ¹rancuntranculinranquin ²
Genitiveranquo ¹rancutoranculionranquion ²
Possessiverancuvarancutwaranculívaranquíva ²
Instrumentalrancunenrancuntenranculínenranquínen ²

o/u-nounSg.Du.Part. Pl.Pl.
Dativeluppun ¹luppuntluppulinluppuin

The tables above are mostly based on the Declension of Nouns (DN) from the early 1930s (PE21/11-12) with forms modernized to fit the case system of the 1950s and 60s. In DN, the examples include malo (malu-) “rust” rather than with its later meaning “pollen”, as well as telco (telcu-) “leg” which was a true qu-noun in DN, rather than reformed to align with nouns like ranco (rancu-) “arm” as it was in Tolkien’s later writings (LR/391).

The DN example for o/u-noun lesto (lestu-) “journey” was long abandoned, and in the table above it has been replaced by later luppo (luppu-) “lump” from the 1950s (PE19/92). The -uen dative singular appears as the normal singular form in DN: lestuen (PE21/12), but whether forms like luppuen survived in Tolkien’s later conception of the languages is unclear. I assume that such forms (if they exist) are archaic.

11.3.5 Consonantal Nouns

Consonantal nouns have different declensions than vocalic nouns, mainly in that they have i-plurals, u-duals and their forms frequently have a joining vowel with case endings that begin with a consonant: e generally but a for the instrumental. This section examines nouns whose stem ends in a single consonant, where the instrumental and partitive plurals often show some sort of consonant assimilation. As with vocalic nouns, long consonantal nouns whose second-to-last syllable is light (having a short vowel followed by zero or one consonant) show prosodic lengthening, as discussed in Chapter 6, Section §6.1.3. Nouns ending in l, n, r, t: Nouns ending in l, n, r, t have stem forms identical to their uninflected forms, but generally show consonant assimilations in the instrumental and partitive-plural forms. The examples here are tecil “pen”, atar “father”, tirion “watch tower”, and imbilat “deep valley”; these last two represent long consonantal noun as well.

l-nounSg.Du.Part. Pl.Pl.
Nominativeteciltecilutecilli ¹tecili
Ablativetecillo ²tecilullotecillillo(n)tecilillon
Instrumentaltecilmen ³tecilúnentecillínentecilínen

long n-nounSg.Du.Part. Pl.Pl.
Nominativetiriontiriónutiriolli ¹tirióni
Allativetirionna ²tirionunnatiriollinna(r)tirioninnar
Instrumentaltirionwen ³tirionúnentiriollínentirionínen

r-nounSg.Du.Part. Pl.Pl.
Nominativeatarataruatalli ¹atari
Instrumentalatarmen ³atarúnenatallínenatarínen

long t-nounSg.Du.Part. Pl.Pl.
Nominativeimbilatimbilátuimbilateli ¹imbiláti
Instrumentalimbilatwen ³imbilatúnenimbilatelínenimbilatínen Nouns ending in s, þ, m, c: Nouns with ancient stems ending in s, m, c have uninflected forms that differ from their noun stem. The m-nouns and c-nouns retain their ancient consonant for inflected forms but have -n, -t in the uninflected singular. Nouns whose uninflected singulars have final s (s-nouns) instead show r in their inflected forms. The exception to this are nouns were the s was derived from archaic þ (þ-nouns), in which case the s is preserved in inflected forms as well.

The examples here are cas (car-) “head”, lós [þ] “inflorescence”, hón (hom-) “heart (physical)”, and filit (filic-) “sparrow”.

s-nounSg.Du.Part. Pl.Pl.
Nominativecascarucalli ¹cari
Locativecaressë ²carussëcallisse(n)carissen
Instrumentalcarmen ³carúnencallínencarínen

þ-nounSg.Du.Part. Pl.Pl.
Nominativelóslósulóseli ¹lósi
Instrumentallóswen ³lósúnenlóselínenlósínen

m-nounSg.Du.Part. Pl.Pl.
Nominativehónhomuhomeli ¹homi
Instrumentalhomnen ³homúnenhomelínenhomínen

c-nounSg.Du.Part. Pl.Pl.
Nominativefilitfilicufiliceli ¹filici
Instrumentalfiliquen ³filicúnenfilicelínenfilicínen

When we get into declensions of consonantal nouns, things become increasing speculative since there are no detailed discussions and relatively few examples of inflected forms for such nouns in Tolkien’s later writings. The most detailed discussion on this class of nouns appears in the Declension of Nouns (DN) from the early 1930s (PE21/16-37). This document predates many changes in Quenya’s system of noun cases and phonological developments. Updating the DN declensions to be compatible with Tolkien’s later vision of his languages involves a lot of guesswork.

11.3.6 Noun Stem Clusters

Some Quenya nouns, especially longer ones, tend to have noun stems ending in a consonant cluster. This is due to the tendency towards vowel loss in longer words from early in Quenya’s history (PE19/59). Such consonant clusters are incompatible with the kinds of assimilations seen in noun stems ending in a single consonant.

The examples here are pilin (pilind-) “arrow” and caimasan (caimasamb-) “bedroom”, a compound of caima “bed” and sambë “room”. In the second example caimasamb-, the final e was lost and the cluster mb reduced to n in the uninflected singular, but was retained in inflected forms. Words like pilind- have similar origin, but since they are not compounds and the vowel was lost so long ago, in many cases it is not possible to reconstruct the lost vowel.

clusteredSg.Du.Part. Pl.Pl.
Nominativepilinpilindupilindeli ¹pilindi
Possessivepilindua ²pilindúvapilindelívapilindíva
Instrumentalpilindanen ³pilindúnenpilindelínenpilindínen

clusteredSg.Du.Part. Pl.Pl.
Nominativecaimasancaimasambucaimasambeli ¹caimasambi
Possessivecaimasambua ²caimasambúvacaimasambelívacaimasambíva
Instrumentalcaimasambanen ³caimasambúnencaimasambelínencaimasambínen Syncopated Consonantal Nouns: One of the basic sound changes in Quenya’s history is the Quenya syncope: the loss of a repeated identical vowel when the result would be a suitable consonantal cluster. For example, neldë “three” is derived from more ancient neledē (VT47/10). For consonantal nouns, the syncope is not possible in the uninflected singular, because the result would be a final noun cluster. With inflected forms, however, the syncope becomes possible again. The net result is that their noun stems end in consonant clusters, and they behave much like longer nouns whose stems end in clusters.

The examples here are laman (lamn-) “animal” and nelet (nelc-) “tooth”.

syncopatedSg.Du.Part. Pl.Pl.
Nominativelamanlamnulamneli ¹lamni
Possessivelamnua ²lamnúvalamnelívalamníva
Instrumentallamnamen ³lamnúnenlamnelínenlamnínen

syncopatedSg.Du.Part. Pl.Pl.
Nominativeneletnelcunelceli ¹nelci
Possessivenelqua ²nelcúvanelcelívanelcíva
Instrumentalnelcanen ³nelcúnennelcelínennelcínen

Nouns with stems ending in clusters require the same kind of guesswork as nouns whose stems end in a single consonant. The examples lamn- (laman), pilind-, and caimasamb- are drawn from the Declension of Nouns (PE21/28, 35, 37) with modifications to bring them in line with Quenya’s noun cases as Tolkien envisioned them in the 1950s and 60s. The singular/plural nelet/nelci “tooth/teeth” appeared in The Etymologies written around 1937 (LR/376), an example I used mainly so I can demonstrate rare assimilated possessive forms like nelqua “tooth’s”.

11.3.7 Vocalic Monosyllables

There is one final small class of nouns Tolkien mentioned the Declension of Nouns (PE21/38-40) and in some later writings as well (VT47/12, 34-36): monosyllabic nouns ending in a vowel. These nouns are the result of either a lost weak consonant in ancient Elvish such as *maha > “hand”, or derivation from ancient nouns that were themselves monosyllabic such as * > “(closed) mouth”. Such nouns almost invariably have a long vowel. For the most part, they are declined like vocalic nouns, but with shortening of their long vowel (1) before suffixes with clusters or (2) with suffixes beginning with i which then form a diphthong. Their duals and plurals tend to be irregular, sometimes showing -r, -t but sometimes preserving more ancient i-plurals and u-duals instead.

The examples here are “hand”, “(closed) mouth”, “small insect, fly”, “owl”, “sound of wind”.

a-monosyllableSg.Du.Part. Pl.Pl.
Nominativemát ¹málimár ¹
Dativemánmant ²málinmain ³
Possessivemávamátwa ²málívamaiva ³
Allativemanna ²manta ²málinna(r)mannar ²
Ablativemallo ²malto ²málillo(n)mallon ²
Locativemassë ²matsë ²málisse(n)massen ²
Instrumentalmánenmanten ²málínenmainen ³

e-monosyllableSg.Du.Part. Pl.Pl.
Nominativepeu ¹pélipér ¹
Allativepenna ²peunnapélinna(r)pennar ²
Ablativepello ²peullopélillo(n)pellon ²
Locativepessë ²peussëpélisse(n)pessen ²

i-monosyllableSg.Du.Part. Pl.Pl.
Nominativepít ¹pílipír ¹
Dativepínpint ²pílinpín
Possessivepívapítwa ²pílívapíva
Allativepinna ²pinta ²pílinna(r)pinnar ²
Ablativepillo ²pilto ²pílillo(n)pillon ²
Locativepissë ²pitsë ²pílisse(n)pissen ²
Instrumentalpínenpinten ²pílínenpínen

o-monosyllableSg.Du.Part. Pl.Pl.
Nominativehót ¹hólihór ¹
Dativehónhont ²hólinhoin ³
Possessivehóvahótwa ²hólívahoiva ³
Allativehonna ²honta ²hólinna(r)honnar ²
Ablativehollo ²holto ²hólillo(n)hollon ²
Locativehossë ²hotsë ²hólisse(n)hossen ²
Instrumentalhónenhonten ²hólínenhoinen ³

u-monosyllableSg.Du.Part. Pl.Pl.
Nominativesút ¹súlisúvi ¹
Dativesúnsunt ²súlinsuin ³
Genitivesúosútosúlionsuion ¹
Possessivesúvasútwa ²súlívasuiva ³
Allativesunna ²sunta ²súlinna(r)sunnar ²
Ablativesullo ²sulto ²súlillo(n)sullon ²
Locativesussë ²sutsë ²súlisse(n)sussen ²
Instrumentalsúnensunten ²súlínensuinen ³

Tolkien discussed the declensions of this noun class in the Declension of Nouns (DN) from the early 1930s (DN: PE21/38-40), but there have been so many grammatical and phonological changes in Quenya that is hard to know which ideas survived in Tolkien’s later conception of Quenya. The only one of these nouns for which we have a significant number of declensions from the 1950s and 60s is “hand” (PE17/160), and it is from this example that I derive the notion that vocalic monosyllables preserve their long vowel before genitive singular -o. The genitive of o-monosyllables do in fact have uo-genitives as an alternate forms in DN (PE21/38, 40), but in that document the u is short and the notion that they have long úo (húo “of two owls”) is speculative on my part.

A lot of the remaining of the declensions are guesswork on my part, especially for the e-, i-, o- and u-monosyllables.

11.3.8 Vocabulary: Law and Society

Here are some vocabulary words having to do with law and society:

Here are miscellaneous nouns introduced in this section:

11.4 Chapter Summary

11.4.1 Chapter Vocabulary



Law and Society:


11.4.2 Grammar Summary Compound Tenses: The compound tenses combine basic verb tenses with -në to put them in the past or -(u)va to put them in the future. Subjunctive: Quenya’s basic adverbs and conjunctions of uncertainty are as follows. More Noun Classes: The information in the section on noun classes is too complex and specialized to summarize cleanly. Please consult the main discussion in section §11.3.