Like most languages, Quenya distinguishes between singular (one) and plural (multiple) nouns. Quenya has two special plural forms, the dual (for pairs) and the partitive plural (for portions of groups = “some”). This entry discusses the general plural, which is used for all the other forms of noun plurality.
Quenya has two general plural suffixes: -i used after consonantal nouns and -r used after (most) vocalic nouns. Of the two plural noun suffixes, the i-plural is the most ancient:
Plural. The most used sign was ī appearing especially between the stem and other affixes, but also finally in the uninflected or “acc.” pl. form (PE21/56, Primitive Quendian Final Consonants, 1936).
Plurality … In nouns the most used element was [ī]. This was added to the stem direct, and since it preceded the addition of any other affixes, as those for “case”, it was probably the oldest element (PE21/72, Common Eldarin Noun Structure, early 1950s).
The plural element in nouns is [i] as a suffix [ī]. Plurals formed with this are in Quenya “general” … (PE17/62, draft of a 1955 letter to David Masson).
In Common Eldarin, the plural suffix -ī was used with all nouns, forming a (long) diphthong when added to vocalic nouns: ✶tekmāi “letters” (PE17/43), ✶lassḗi “leaves” (PE19/106). At some point, though, Quenya adopted the r-plural suffix from verb inflections to use with (most) vocalic nouns:
r (more rarely l) was originally, it seems, employed chiefly in verbs to mark the plural subject, especially when this was unspecified or indeterminate. The invasion of noun-inflexion by r as a (nominative) plural sign is peculiar to Q. and an event that occurs within the period of the oldest records (PE21/73, Common Eldarin Noun Structure, early 1950s).
Thus it was that when the name Banyai of old was changed to Vanyar this was done only because the sound b was changed to v throughout the language (save in certain sequences) — and this change, it is recorded, began among the Vanyar; whereas for the showing of many the new device of r was brought in and used in all words of a certain shape — and this, it is said, was begun among the Noldor (PM/402, footnote in Dangweth Pengolodh, mid 1950s).
In addition to -ī (nouns) and -r (originally just verbs), one final ancient plural marker of note was ✶-m used in a variety of circumstances in Common Eldarin (VT/42/26; PE21/72). This suffix did not remain a general plural marker in Quenya, but survived as a secondary plural marker in a few noun cases, notably genitive plural -ron/-ion, ablative plural -llon and locative plural -ssen (see below). This was after the sound change whereby final -m became -n.
Exactly when the switch to the r-plural for nouns took place isn’t clear, but it was probably established by early Classical Quenya (Parmaquesta). Though not explicitly stated above, it is clear from the examples that -r largely took over the pluralizing function in vocalic nouns: Vanyar vs. ancient ✶Banyai, Eldar vs. ancient ✶Eldai. This replacement had one major exception, however: the very large class of Quenya nouns ending in -e, which I call e-nouns.
The sg. quende (not much used) was made in Quenya from Quendi, on the model of other nouns in -e, the majority of which formed their plurals in -i (WJ/361).
Here the ancient diphthong ei became ī as it usually did: “So lassḗi (pl. of lassē “leaf”) > lássei, PQ lassī” (PE19/106). By the time of the Exile (Tarquesta), the long final -ī shortened to -ĭ. There are numerous examples of e-nouns with i-plurals:
- essi “names”, sg. essë (MR/216).
- fairi “phantoms”, sg. fairë (MC/221).
- lómelindi “nightingales”, sg. lómelindë (MR/172).
- maiwi “gulls”, sg. maiwë (MC/222).
- tinwi “sparks”, sg. tinwë (MR/388).
However, this exception itself has some exceptions, which I call er-plurals. In particular, nouns ending -ie and -le form plurals with -r rather than -i. In case of -ie it is probably because -ii is phonologically impossible. In the case of -le it is probably because the resulting i-plural would look too much like the partitive plural suffix -li:
- tier “paths”, sg. tië (LotR/0377).
- Valier “Queens of the Valar”, sg. Valië (S/25).
- fintaler “tricks”, sg. fintalë (PE17/119).
- tyeller “grades”, sg. tyellë (LotR/1118).
These are not the only e-nouns that break the rules. A few others show an er-plural for no obvious reason:
- lámatyáver “sound tastes”, sg. lámatyávë (MR/216).
- lotser “(small) flowers”, sg. lotsë (PE17/160).
- nasser “natures”, sg. nassë (PE17/175).
There are even some e-nouns that appear with both plural forms: Ingwi (PM/332) vs. Ingwer (PM/340). It is hard to say whether these examples are conceptual vacillations on Tolkien’s part, irregular nouns or represent some dialectical variant. For purposes of Neo-Quenya, I’d treat them as irregular nouns, and assume that the noun plural of e-nouns is usually -i with the exception of nouns ending in -ie and -le.
To summarize:
- Consonantal nouns form their plural by adding -i.
- Most vocalic nouns form their plural by adding -r.
- The e-nouns form their plural by changing the final -e to -i.
- Nouns ending in -ie and -le form their plural by adding -r.
The noun cases frequently preserve more ancient plural formations. Based on the Plotz Letter, the plural inflections for most vocalic nouns are:
- ciryain [plural dative] = cirya-i-n.
- ciryaron [plural genitive] = cirya-r-o-n
- ciryannar [plural allative] = cirya-nna-r.
- ciryallon [plural ablative] = cirya-llo-n.
- ciryassen [plural locative] = cirya-sse-n.
- ciryainen [plural instrumental] = cirya-i-nen.
- ciryaiva [plural possessive-adjectival] = cirya-i-va.
The plural possessive-adjectival does not appear in Plotz, but can be seen elsewhere in examples like Eldaiva “of the Eldar” (WJ/369). Note how the dative, instrumental and possessive-adjectival plurals retain the older i-plural suffix, thereby forming diphthongs with the preceding vowel.
The allative, ablative and locative attach the case ending first, and then the plural marker second, and in the case of ablative and locative use the ancient plural marker -n (originally -m), one of the few situations where this marker survives in Quenya. This -n marker also appears in the genitive plural. For the ablative, a variant -llor is sometimes seen as an alternative to -llon: raxellor “*from dangers” (VT44/9).
The e-noun plural inflections are very similar:
- lassin [plural dative] = lass-i-n.
- lassion [plural genitive] = lass-i-o-n
- lassennar [plural allative] = lasse-nna-r.
- lassellon [plural ablative] = lasse-llo-n.
- lassessen [plural locative] = lasse-sse-n.
- lassínen [plural instrumental] = lass-í-nen.
- lassíva [plural possessive-adjectival] = lass-í-va.
The possessive-adjectival plural for e-nouns does not appear in Tolkien’s later writing, but it does appear in charts from the early 1930s (PE21/47). The form lassíva remains plausible in Tolkien’s later conception of Quenya phonology: *lasseiwā becoming either lassīvā, or becoming lassēvā which then leveled to lassīvā via analogy. Compare this with the attested developments of lasséinen > lassēnen > lassī́nen (PE19/106). The above declensions for e-nouns are nearly the same as the vocalic declensions, except that the case suffix for the dative and genitive is added to i-plural form, and likewise for the instrumental and the possessive-adjectival cases but with a lengthened í as well.
The consonantal nouns do not appear in the Plotz tables, but they are probably very similar to the e-nouns, since they also use i-plurals. The main differences are the allative, ablative and locative cases, where the case suffix was added directly to the i-plural form:
- atanin [plural dative] = atan-i-n.
- atanion [plural genitive] = atan-i-o-n
- ataninnar [plural allative] = atan-i-nna-r.
- atanillon [plural ablative] = atan-i-llo-n.
- atanissen [plural locative] = atan-i-sse-n.
- atanínen [plural instrumental] = atan-í-nen.
- ataníva [plural possessive-adjectival] = lass-í-va.
For longer consonantal nouns, the í would normally lengthen in the instrumental and the possessive-adjectival cases due to prosodic lengthening. In theory this lengthening might not happen for monosyllables or nouns whose stem ends in a consonant clusters, but it seems likely the long í would intrude in those cases via analogy with longer consonantal nouns and e-nouns (or perhaps the primitive long ī would simply be preserved in those circumstances). Thus maybe from singular nat “thing”: natíva, natínen “things’, by means of things”.
Only a few of the above plural forms for consonantal nouns are actually attested in Tolkien’s later writings:
- aranion “of the kings” genitive plural of aran (LotR/864; PE17/49).
- mindoninnar “upon towers” allative plural of mindon (MC/222).
- elenillor “[from] the stars” ablative plural of elen (MC/222).
The last example uses the alternate ablative plural ending -llor vs. -llon as given in Plotz.
Conceptual Development: The use of the suffix -i for the plurals of consonantal nouns dates all the way back to Tolkien’s earliest conception of Qenya from the 1910s, for example tavari or tavarni plural of ᴱQ. tavar “dale sprites” (LT1/66; QL/90). In the 1910s, however, there were two competing plural suffixes for vocalic nouns: -r (PE11/10) and ᴱQ. -li (QL/53). The plural forms Noldoli (LT1/47) and Noldor (LT1/162) both appeared in the Lost Tales, though Noldoli was more common.
In the Qenya Lexicon plurals of vocalic nouns mostly used -r: Eldar (QL/43), Torqeler “Tropics” (QL/94), Valar vs. variant Vali (QL/99). There was also evidence of special i-plurals for e-nouns: arauki plural of arauke “demon” (QL/32).
In the Early Qenya Grammar (EQG) of the 1920s, Tolkien seems to have settled on the paradigm of -i for vocalic nouns and -li for consonantal nouns:
Vocalic.
- Sg. N. kalma, A. kalmat, G. kalman, D. kalmar
- pl. kalmali, kalmalin, kalmqlion, kalmalir.
In a word of this type there is no shift of accent or quantity except in the G. pl. where according to uniform trissyllabic law of Qenya the accentuation is kàlmắlion …
- Sg. tantare, … ondo …
- pl. tantarēli, etc.; ondoli, etc.
Consonantal.
- Sg. peltas, peltaksa, peltakso, peltakse
- pl. peltaksi, peltaksin, peltaksion, peltaksir
Similarly, pilin from stem pilind-; kar from stem kas- (PE14/43-44).
Note how there is no special i-plural for the e-noun tantare. It uses the regular vocalic -li suffix, with prosodic lengthening: tantaréli. The above is from the manuscript version of EQG. The system described in the typescript version of EQG is the same, and Tolkien gave the following set of plural endings (PE14/73):
| N. | A. | G. | D. | |
|---|---|---|---|---|
| Vocalic | -li | -lin | -lion | -lir |
| Consonantal | -i | -in | -ion | -ir |
In both the manuscript and typescript versions of EQG, Tolkien also described a special set of plural forms for the allative, ablative and locative cases: plural -ntar, -llon (or -llor), -ssen vs. singular -nta, -llo, -sse (PE14/47, 79). In the manuscript EQG, Tolkien said these plural forms would use the joining vowel -i- with consonantal nouns (PE14/47), though he didn’t use those exact words. It is interesting to see that the -llon/-llor variants date all the way back to the 1920s.
Tolkien further explored Qenya noun declensions in a series of tables posthumously labeled Qenya Declensions with various versions composed from the 1920s through 1940s (PE16/111-115, PE21/42-54). There is a longer document labeled the Declension of Nouns written in the early 1930s marking the mid-point of this series of conceptual developments (PE21/1-41).
Version 1 (PE16/111): The first version had the same basic declensions as the Early Quenya Grammar, but the only examples it gave were vocalic nouns:
- Nominative -lin; Accusative -li; Genitive -lion; Dative -lir.
Version 2 (PE16/112): The second version added a plural instrumental case, which unlike EQG was distinct from the singular instrumental form. It provided examples for both vocalic and consonantal nouns. This is the first set of declensions where Tolkien explicitly defined a special set of rules for e-nouns. The e-nouns could either be declined like other vocalic nouns, or could be declined with the same endings as the consonantal nouns, with the -i replacing the final e. The vocalic-style declensions for e-nouns were in parenthesis, indicating these variants were probably obsolete, and the consonantal-style declensions were preferred.
| Noun Class | N. | A. | D. | G. | Inst. |
|---|---|---|---|---|---|
| Vocalic¹ | -lin | -li | -lir | -lion | -línen |
| Consonantal | -in | -i | -ir | -ion | -ínen |
¹ e-nouns typically inflected like consonantal nouns, replacing the final e with the suffix.
Version 3 (PE16/113-115): The third version added the allative, ablative, locative, comitative and adverbial cases. The four basic declensions remained the same as EQG, and the newer allative, ablative, locative follow the pattern established in version 3 of adding the case suffixes to the plural forms. The adverbial case had previously appeared in EQG but without a distinct plural form. The comitative case was entirely new.
| Noun Class | N. | A. | D. | G. | Inst. | All. | Abl. | Loc. | Com. | Adv. |
|---|---|---|---|---|---|---|---|---|---|---|
| Vocalic¹ | -lin | -li | -lir | -lion | -línen | -linta(n) | -lillon | -lissen | -linko | -lindon |
| Consonantal | -in | -i | -ir | -ion | -ínen | -inta(n) | -illon | -issen | -inko | -indon |
¹ e-nouns typically inflected like consonantal nouns, replacing the final e with the suffix.
Version 4 (PE21/1-41): The fourth version added possessive and partitive cases, but removed the comitative case. This version also added a large number of noun classes, but plural forms were still mostly based on the same three categories: vocalic, e-noun and consonantal. Of these, only the vocalic instrumental declension changed from version 3 in normal declensions. The version also described an archaic set of “short plurals” for vocalic nouns, based on more ancient i-plurals. This is the earliest clear evidence that the i-plural was the older the two plural forms.
| Noun Class | N. | A. | D. | G. | Inst. | All. | Abl. | Loc. | Adv. | Pos. | Part. |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Vocalic¹ | -lin | -li | -lir | -lion | -inen | -lintan | -lillon | -lissen | -lindon | -linwa | -ika |
| (short) | -in | -i | -er/-ir³ | -ion | -inen | -ntan | -llon | -ssen | — | — | -ika |
| Consonantal | -in | -i | -ir | -ion | -ínen | -intan | -illon | -issen | -indon | -inwa | -íka |
¹ e-nouns typically inflected like consonantal nouns, replacing the final e with the suffix.
³ With -e replacing the final vowel for a-nouns.
Version 5a (PE21/42-45): Version 5a restored the comitative. It only gave declensions for a-nouns and e-nouns, but presumably consonantal nouns still used the same suffixes as the e-nouns. The allative changed from -(l)intan to -(l)inden and the possessive from -(l)inwa to -(l)íva but otherwise the long form declensions mostly remained the same as version 4. In the (still archaic) short vocalic forms Tolkien reintroduced the r-plural for the first time since the Qenya Lexicon:
| Noun Class | N. | A. | D. | G. | Inst. | All. | Abl. | Loc. | Com. | Adv. | Pos. | Part. |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Vocalic | -lin | -li | -lir | -lion | -inen | -linden | -lillon | -lissen | -líle | -lindon | -líva | -líka |
| (short) | -r | -r | -(i)re(n) | -ron | -inen | -nden | -llon | -ssen | -ile | -ndon | -iva | -ika |
| e-noun² | -in | -i | -ir | -ion | -ínen | -inden | -illon | -issen | -íle | -indon | -íva | -ka |
² With -i replacing the final vowel for e-nouns.
Version 5b (PE21/46-49): Version 5b still had no consonantal examples. The long vocalic plurals are more or less unchanged from the previous version, and the short form changed only in the accusative.
| Noun Class | N. | A. | D. | G. | Inst. | All. | Abl. | Loc. | Com. | Adv. | Pos. | Part. |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Vocalic | -lin | -li | -lir | -lion | -línen | -linde(n) | -lillo(n) | -lisse(n) | — | -lindon | -líva | -líka |
| (short) | -r/-in | -e/-i³ | -ire(n) | -ron | -inen | -nden | -llon | -ssen | -l/-ile | -ndon | -iva | -ika |
| e-noun² | -in | -i | -íre(n) | -ion | -ínen | -inden | -illon | -issen | -íle | -indon | -íva | -íka |
² With -i replacing the final vowel for e-nouns.
³ With -e replacing the final vowel for a-nouns.
Version 5c (PE21/50-52): Version 5c was just a list of suffixes, allowing it to cover more noun classes in a smaller space. It also had more variations as a result, not all of which are reflected below. The long and short vocalic declensions remained mostly the same, but some of the short vocalic inflections appeared in e-nouns. As a result, this is the first (only?) version where e-noun suffixes were distinct from both the vocalic nouns and the consonantal nouns. In this version, the e-nouns had OQ [Old Quenya] variants of the allative, ablative and locative plurals where -nden, -llon, -ssen were added directly the stem (lassessen) rather than the plural (lassissen). This OQ pattern became the normal pattern in later declensions.
| Noun Class | N. | A. | D. | G. | Inst. | All. | Abl. | Loc. | Com. | Adv. | Pos. | Part. |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Vocalic | -lin | -li | -lir | -lion | -línen | -linden | -lillon | -lissen | -líle | -lindon | -líva | -líka |
| (short) | -r | -e/-i³ | -(i)re(n) | -ron | -inen | -nden | -llon | -ssen | -l/-ile | — | -iva | -ika |
| e-noun² | -i/-r | -i/-r | -íre(n) | -ion | -ínen | -inden | -illon | -issen | -íle | -indon | -íva | -íka |
| Consonantal | -in | -i | -ir/íre | -ion | -ínen | -(i)nden | -(i)llon | -(i)ssen | — | — | -íva | — |
² With -i replacing the final vowel for e-nouns.
³ With -e replacing the final vowel for a-nouns.
Version 6 (PE21/53-54): The sixth version lost the partitive, possessive and (plural) adverbial cases, and labeled the accusative simply as the “base” form to which it was identical (in this and earlier versions). This version only had vocalic examples. The long vocalic variants with -li vanished, and the r-plurals became the normal pattern for vocalic nouns. The e-nouns retained a distinct set of declensions, and for the first time since their split from vocalic nouns shared the same declensions for the allative, ablative and locative. Many of these plural declensions are identical to those found in Plotz, excluding only the dative and allative cases, as well as the change in the partitive from a noun case to the partitive plural form.
| Noun Class | N. | A. | D. | G. | Inst. | All. | Abl. | Loc. | Com. |
|---|---|---|---|---|---|---|---|---|---|
| Vocalic | -r | -i | -ino | -ron | -inen | -ntar | -llon | -ssen | -iko |
| e-noun² | -i | -í | -íno | -ion | -ínen | -ntar | -llon | -ssen | -íko |
² With -i replacing the final vowel for e-nouns.
Putting all of the above in one big chart with “Version 0” for the Early Qenya Grammar and LQ for the Late Quenya paradigm from the Plotz letter:
| V | Noun Class | N. | A. | D. | G. | Inst. | All. | Abl. | Loc. | Com. | Adv. | Pos. | Part. |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | Vocalic | -lin | -li | -lir | -lion | -inen | -ntar | -llor | -ssen | — | — | — | — |
| 0 | Consonantal | -in | -i | -ir | -ion | — | — | — | — | — | — | — | — |
| 1 | Vocalic | -lin | -li | -lir | -lion | — | — | — | — | — | — | — | — |
| 2 | Vocalic¹ | -lin | -li | -lir | -lion | -línen | — | — | — | — | — | — | — |
| 2 | Consonantal | -in | -i | -ir | -ion | -ínen | — | — | — | — | — | — | — |
| 3 | Vocalic¹ | -lin | -li | -lir | -lion | -línen | -linta(n) | -lillon | -lissen | -linko | -lindon | — | — |
| 3 | Consonantal | -in | -i | -ir | -ion | -ínen | -inta(n) | -illon | -issen | -inko | -indon | — | — |
| 4 | Vocalic¹ | -lin | -li | -lir | -lion | -inen | -lintan | -lillon | -lissen | — | -lindon | -linwa | -ika |
| 4 | (short) | -in | -i | -er/-ir³ | -ion | -inen | -ntan | -llon | -ssen | — | — | — | -ika |
| 4 | Consonantal | -in | -i | -ir | -ion | -ínen | -intan | -illon | -issen | — | -indon | -inwa | -íka |
| 5a | Vocalic | -lin | -li | -lir | -lion | -inen | -linden | -lillon | -lissen | -líle | -lindon | -líva | -líka |
| 5a | (short) | -r | -r | -(i)re(n) | -ron | -inen | -nden | -llon | -ssen | -ile | -ndon | -iva | -ika |
| 5a | e-noun² | -in | -i | -ir | -ion | -ínen | -inden | -illon | -issen | -íle | -indon | -íva | -ka |
| 5b | Vocalic | -lin | -li | -lir | -lion | -línen | -linde(n) | -lillo(n) | -lisse(n) | — | -lindon | -líva | -líka |
| 5b | (short) | -r/-in | -e/-i³ | -ire(n) | -ron | -inen | -nden | -llon | -ssen | -l/-ile | -ndon | -iva | -ika |
| 5b | e-noun² | -in | -i | -íre(n) | -ion | -ínen | -inden | -illon | -issen | -íle | -indon | -íva | -íka |
| 5c | Vocalic | -lin | -li | -lir | -lion | -línen | -linden | -lillon | -lissen | -líle | -lindon | -líva | -líka |
| 5c | (short) | -r | -e/-i³ | -(i)re(n) | -ron | -inen | -nden | -llon | -ssen | -l/-ile | — | -iva | -ika |
| 5c | e-noun² | -i/-r | -i/-r | -íre(n) | -ion | -ínen | -(i)nden | -(i)llon | -(i)ssen | -íle | -indon | -íva | -íka |
| 5c | Consonantal | -in | -i | -ir/íre | -ion | -ínen | -inden | -illon | -issen | — | — | -íva | — |
| 6 | Vocalic | -r | -i | -ino | -ron | -inen | -ntar | -llon | -ssen | -iko | — | — | — |
| 6 | e-noun² | -i | -í | -íno | -ion | -ínen | -ntar | -llon | -ssen | -íko | — | — | — |
| LQ | Vocalic | -r | -i | -in | -ron | -inen | -nnar | -llon | -ssen | — | — | — | — |
| LQ | e-noun² | -i | -í | -in | -ion | -ínen | -nnar | -llon | -ssen | — | — | — | — |
¹ e-nouns typically inflected like consonantal nouns, replacing the final e with the suffix.
² With -i replacing the final vowel for e-nouns.
³ With -e replacing the final vowel for a-nouns.
Thus in version 5a the r-plurals were introduced as variant (archaic) short plurals of vocalic nouns, and as of version 6 they were firmly established as the normal plural for vocalic nouns (excluding e-nouns). Such r-plurals were the norm in The Etymologies from the 1930s. In the contemporaneous Silmarillion drafts, forms like Noldoli were phased out in favor of Noldor. Tolkien seems to have stuck to these basic patterns thereafter.
Neo-Quenya: There is broad consensus on plural forms among Neo-Quenya authors, even the unattested consonantal plural forms. The only area of ambiguity is whether or not the possessive plural should be -íva or -iva for e-nouns and consonantal nouns. I favor -íva based on historical phonology and attested forms from the late 1920s and early 1930s. Of the major Quenya courses, only Thorsten Renk advocated for -iva (QLE/33).