As a general rule, the subject precedes the verb in Quenya: i nér cenne “the man saw”. The biggest exception is when the subject is a pronoun, in which case the pronoun is attached to the verb as a suffix: cennenye “I saw” = cen-ne-nye “see-(past)-I” or “see-(past)-1sg”. This is not strictly speaking subject-verb agreement, a feature of many European languages. The Quenya pronominal subject suffix doesn’t “agree” with the subject, it is the subject.
The subject suffixes resemble but do not match the forms of independent pronouns. Quenya Pronominal Elements (QPE) written in 1968 gives a complete set of pronominal suffixes (VT49/51), and most Neo-Quenya writers use this document as the basis for the Quenya pronoun paradigm. Many of these forms, especially in the singular, exist in long (-nye) and short (-n) variants, often represented as -n(ye). The short forms are used more frequently than the long forms, except when there is an object suffix in which case the long form is required. The full set of Quenya pronominal subject suffixes is as follows:
|1st Person||-n(ye) “I”||-lme “we (but not you)”||-mme “us two (but not you)”|
|-lve “we (including you)”||-nque “you and I”|
|2nd Person||-t(ye) “you (familiar)” ¹||-lde “you”||-ste “you two”|
|-l(ye) “you (polite)”|
|3rd Person||-s(se) “he/she/it” ²||-lte “they” ³||-tte “they two”|
¹ In QPE Tolkien said that -tye did not have a short form, but such a short form did appear in the Quendi and Eldar essay written around 1960, in the short imperative hecat “be gone [you]” (WJ/364). The short 2nd singular familiar form -t also appeared in an inflection chart from the mid-60s (VT49/48; PE17/75). Many Neo-Quenya writers therefore this as the short form of -tye, and it is worth learning so you can recognize it. Personally, though, I recommend always using the longer form -tye. The suffix -t serves several other grammatical functions, such as the dual suffix or the 3rd person plural object suffix.
² In QPE Tolkien said that the long 3rd singular suffix was (rare) -se. In documents elsewhere he gave -sse (VT49/28). In yet another place he gave the long form as -s(te) for animates, and -s(sa) for inanimates (PE17/57). All three of these long variants appear in inflected verbs in the 1950s and 60s, but -sse is slightly more common and that is what I recommend using. In practice, the short third singular -s is by far the most common option, however.
³ Tolkien gave -lte in QPE, but the suffix -nte appears frequently as the 3rd plural suffix in Tolkien’s later writing as well (for example on UT/317). I personally prefer -nte over -lte, but the two appear so often it is hard to say one is better than the other. In one place Tolkien gave -nta as the inanimate 3rd plural suffix (PE17/57), but personally I would use -s and -nte/-lte for both animate and inanimate in verbal inflections.
Other pronominal verbal inflections: In several places (PE17/75, VT49/51) Tolkien listed the “impersonal” verbal suffixes -r (plural) and -t (dual); the “singular impersonal” was unmarked. The main use for these impersonal forms is for subject-verb agreement when there is an independent subject. This agreement is in number only, since any 1st or 2nd person subject is attached directly to the verb. See the entry on impersonal verbs for further discussion.
Tolkien mentioned a variety of reflexive constructions, as discussed in the entry on reflexive pronouns. Finally, Tolkien mentioned “honorific” 3rd person pronouns in a couple places (PE43/29; PE17/57), all from the 1950s. On PE43/29 these honorifics are combined with the preposition ó: sg. ótar, pl. ótari. On PE17/57, they are suffixed to the verb car-: sg. caritar, pl. carintar(i), du. caristar, followed by 2nd person variants sg. cariltar, pl. cariltar(i), du. cariltar(o). Christopher Gilson explained these forms as follows:
The forms caritar, etc., are explained in unpublished notes on the pronominal endings contemporary with the conjugation given here. The 2nd person familiar category “was abandoned in the colloquial” — these are the forms marked here with a “†” — and with the earlier polite forms coming to be used in ordinary contexts “a new ‘honorific’ form was made by adding (vocatively) tar, pl. tári or tar” to these forms. “This was only done to persons of high rank or seniority, like sir, madam.” (PE17/58).
Thus these 1950s honorifics are part of a paradigm where the familiar form -tye fell out of use, -lye became the general 2nd sg. suffix, and -tar was introduced to restore the now missing “polite” forms. In the 1960s there is no sign that -tye was archaic, so I would not use these honorifics in Neo-Quenya, even though it would be nice and super-geeky to have a way to say “Elf-san” in Quenya.
Subject suffixes in other contexts: Most of the time, subject suffixes are appended to a verb, but they sometimes appear in other contexts. More notablely, they can also be added to particles like imperative á or negative lá/ua-:
- alye anta men siar ilyarea mastamma “give us this day are daily bread” (VT43/11).
- uan carne “I did not make” (PE17/144).
- melin sé apa lanye hé “I love him but not him (the other)” (VT49/15).
The subject pronouns are expressed in all of the above English translations, but a-lye “(imperative)-you-(polite)” is probably just a more polite imperative, and la-nye = “not-I” probably modifies the (unexpressed) verb mele “love” in the second clause. The case of negatives lá and ua- is more ambiguous, because Tolkien sometimes treated them as negative verbs with full inflections of their own for both tense and person.
There are also examples of subject suffixes used with prepositions; see that entry for discussion.
Origin of subject suffixes: The fact that Quenya pronominal subject suffixes follow the verb even though ordinary subjects precede is a bit strange. In one place, Tolkien said that this was because ancient Quenya used VSO (Verb, Subject, Object) word order:
The normal order in Quenya had been verb first, subject, direct object, indirect object. This survives in cases of “persons” inflexionally expressed, but the classical and normal order was expressed subject, verb, object (PE17/72).
This shift from Common Eldarin VSO to Quenya SVO is discussed in more detail in red-ink revisions to Quendian & Common Eldarin Verbal Structure. This document dates to the mid-to-late 1940s, and was originally based on a paradigm of pronominal subject prefixes. The red-ink revisions switched to a paradigm when Quenya had subject suffixes, as described by Christopher Gilson in his introduction to the document (PE22/82). Thus, the red-ink revisions probably at a minimum date to Tolkien’s reintroduction of subject suffixes, around 1950, as suggested by Lokyt in an discussion on the Aglardh form, January 2020:
The normal order in Eldarin was with pronouns as with nouns (see Declension): subject: object: indirect object. “give a book the man” : “give it him.” It seems fairly clear that in a normal simple sentence the verbal form was placed as near the beginning as possible, though particles, prefixes or simple adverbial forms that qualified the verbal notion could precede it. The subject normally followed, but if it was displaced and put earlier for special emphasis the appropriate pronominal element was often (not always) placed next after the verb, so that the normal order of a simple sentence was: Gives man it me or Man gives he it me (PE22/93 note #4).
The unemphatic subjective pronominal object thus originally as a rule followed immediately after the verbal stem. From this already in Common Eldarin grew up a system of inflexions for the subjective pronoun (as in Indo-European) (PE22/93 note #5).
This order was largely normalized in Eldarin languages, though these may differ in detail. In most it became more usual for a full noun subject (espec. personal name) to precede verb; and since it would as pronominal affix in that case come next. Emphatic subjective pronouns normally preceded for their purpose was full emphasis. In most Eldarin languages they have no pronominal affix (in a difficult-to-read margin note, PE22/93 note #5).
As described here, the normal Common Eldarin word order was VSO, but the subject (noun or pronoun) could appear before the verb when emphatic. In such cases, the non-emphatic pronoun also appeared after the verb, as: ✶nēr anta se parma ni “man give he-(nonemphatic) book [to] me”. The following pronoun became a clitic, then an inflection: anta-se > antas. Meanwhile, the “emphatic” positioning of ordinary nouns before the verb became the normal word order for subjects other than non-emphatic pronouns. The “double-marked” subject eventually vanished, leaving only the noun or emphatic pronoun before the verb: Q. nér anta parma nin, esse anta parma nin vs. antas param nin. There are examples of such (archaic?) double-marked subjects elsewhere in Tolkien’s writing, however: esse ūpa nas “he is dumb” (PE17/126), more literally “he-(emphatic) dumb is-he-(non-emphatic)”.
As for the origins of the individual suffixes, the Common Eldarin pronominal elements were as follows (PE49/50):
|1st Person||✶ni||✶me (exclusive) or ✶we (inclusive)|
|2nd Person||✶ki (familiar) or ✶le||✶de|
When used as subject suffixes, ✶ni and ✶ki had an added e, so that -nı̯e, -kı̯e > -nye, -tye with [ky] > [ty] as usual. The 2nd singular polite form became -lye by analogy, and this form replaced the independent pronoun as well (PE19/80).
For plural pronominal subject suffixes, the plural element ✶l intruded between the verb and the pronoun. -lwe > -lve as usual (PE49/51), but the ancient pronominal element survived in -lme, -lde, -lte. In the case of 2nd plural -lde, the suffixal form was distinct from the independent pronoun, where ✶de > le because initial [d] became [l] in Quenya. The origin of the variant 3rd plural -nte is probably similar, except the “intruding plural element” was ancient ✶m instead, with m > n via assimilation to following t.
As for the dual forms, Tolkien described their origin this way in QPE:
The dual inflexions are peculiar to Quenya. They would be explicable as produced by an inserted number sign -s- parallel to -l- for plural. But there is no trace of any such dual element elsewhere in CE. The probable explanation is that in dual inflexions duality was shown by duplication of the consonant: 1a -mme b ? wwe, 2 d-de, 3 t-te. But since -wwe did not yield a satisfactory form it was replaced by nki 1a + 2a since dual was mainly familiar and maintained as such (PE49/51).
For 2nd person dual exclusive, ✶-d+de > -zde > -ste following usual Quenya phonetic developments. Primitive 3rd person dual ✶-t+te > -ste as well, but it seems it was reformed back to -tte, perhaps to distinguish it from 2nd dual exclusive. Likely 2nd dual inclusive/familiar developed as -nkĭ > -nke > -nkwe (-nque) via melding with inclusive plural -we.
Conceptual Development: Tolkien changed his mind rather frequently on how pronouns were used in verbal inflections. In his earliest writings, there are plenty of examples of pronominal suffixes, such as ᴱQ. tulielto “they have come” (LT1/114). No pronominal paradigm from the 1910s has survived, unfortunately. By the Early Qenya Grammar (EQG) of the early 1920s, Tolkien had switched to pronominal prefixes:
A single proclitic before the ordinary verbal forms is taken as a NOMINATIVE, as: HUMATE; usually (to show the accent) written HU·MATE “he eats”; HIMANTE “she ate” (PE14/85).
For the most part, these were simply the independent pronouns prefixed to the verb. Such forms worked fine for verbs beginning with consonants, but where the verb began with a vowel some adjustments were needed (PE14/86):
- SG. 1 NI before vowel NY-, N before I as nyanta, nista “I give, know”
- 2. KE before vowel TY-, T before I as ty’anta, t’ista “thou givest” &c.
- 3m. HU [before vowel] HW-, before Ū̆, H- as hwanta, hule = “he pours”
- f. HI [before vowel] HY-, before Ī̆, H- as hyanta, hista “she knows”
- n. HA [before vowel] H- as hanta, hule, hista.
The EQG plural pronominal prefixes were quite variable, depending on the initial vowel of the verb.
It seems Tolkien did not retain this system for long. By the late 1920s pronominal suffixes were once again the norm, for example in the various drafts of the Oilima Markirya poem (MC/213-5; PE16/53-87). There are several published verb conjugations that are near contemporaries of these poems, from the late 1920s or early 1930s, all with pronominal suffixes (PE16/124-128). These charts exist in three versions labeled by the editors as V1 to V3. The pronominal paradigm in these charts was as follows:
|2nd Person||-l/-ste ||-le(n)/-lle|
|3rd Person (m.)||-r(o)/-ndo ||-ron/-ldo(n) |
|3rd Person (f.)||-r(e)/-nde ||-ren/-lde(n) |
|3rd Person (n.)||-t/-ta||-tan/-lta(n)|
Some of the suffixes in V1 were different:
-  2nd sg. short form -t.
-  3rd sg. masc. short form -nt.
-  3rd sg. fem. long form -sse.
-  3rd pl. masc. short form -ton.
-  3rd pl. fem. short form -sin.
All of the pronominal suffixes had both “short” and “long” forms. Rather surprisingly, many of the verb tenses were only distinguished by which of the two pronominal forms was used. For example, the aorist, past and conditional tenses all used the short forms (tulin, túlien, tuluvan) and the present, perfect and future tenses were identical except that they used the long forms (tulinye, túlienye, tuluvanye).
There is not (yet) a published pronominal paradigm from the mid-to-late 1930s. The next complete published paradigm appears in Quenya Verbal System (QVS) from the late 1940s, but in it Tolkien once again reverted to pronominal prefixes:
Q ni·kave antas = I can give it … The normal order in Eldarin was with pronouns as with nouns (see Declension): subject: indirect object: object. “give the man a book”; “give him it.” The subject usually preceded the verb, so that the normal order of a simple sentence was: The man / he, (to) him, gives it (PE22/92-93).
As with EQG, the QVS pronominal system simply prefixed the independent pronoun to the verb as the subject. Possibly Tolkien consulted EQG as he was composing QVS, and becoming enamored with his earlier ideas decided to restore them. These pronominal prefixes do not appear in Lord of the Rings drafts except briefly in drafts for Elendil’s Oath (SD/56). Thus it seems the restoration of prefixes was short lived:
- et Ëarello Endorenna nilendie (draft); ni-lendie = “I have-come”.
- et Eärello Endorenna utúlien (published); utúlie-n = “have-come I”.
- Sinome nimaruva, yo hildinyar tenn’ Ambar-metta (draft); ni-maruva = “I will-abide”.
- sinomë maruvan ar Hildinyar tenn’ Ambar-metta (published); maruva-n = will-abide I”.
We currently have four complete pronominal paradigms for Quenya verbs from the 1950s and 60s, three of which were published in VT49 and one more in PE17; one of the VT49 paradigms was republished in PE17 as well. These paradigms are (1) from the late 50s or early 60s (PE17/57), (2) from 1964 (PE17/75; PE49/48), (3) Jan. 1968 (VT49/16) and (4) later (maybe?) in 1968 (VT49/51). Using 1a/1b for 1st person exclusive/inclusive and 2a/2b for 2nd person familiar/polite, these paradigms in their probable chronological order are:
The first two paradigms (1) and (2) were composed before the second edition of The Lord of the Rings was published, the last two (3) and (4) were composed after. One notable changed between the two editions is Frodo’s greeting to Gildor, “a star shines on the hour of our [inclusive] meeting”:
This shift is reflected in the changes in 1st person plural inclusive/exclusive between (1)/(2) paradigms and the (3)/(4) paradigms. In notes preparing for the second edition, Tolkien explained the origins of the second person plural forms like this:
The first person plurals a and b were merged in one form karilme … in EQ pronunciation the forms of 1a and b -lme; -lbe had become insufficiently distinct. The distinction remained active in the dual, where the distinction between exclusive and inclusive “we” is most clear, and Dual 1b. “thou and I” — the so-called oholima (or “confidential”) form — was in frequent colloquial use. Nonetheless the use of -lme for Pl. 1b was classed as “incorrect”. Still more the consequential [was] confusion of the independent pronouns mē̆ (exclusive) and wē̆ > vē̆ (inclusive), me being used for both. But in spite of the “teachers” the loss of the distinction in the plural became established. This was largely assisted by the fact that the change of w > v had caused the pronoun we to merge in sound with vē̆- “similar, like, as” (PE17/129-130).
Thus -mme/-lme is part of a paradigm where the independent inclusive pronoun ve fell out of use, me was repurposed for both exclusive and inclusive, and -lme came to represent inclusive. Perhaps a new exclusive form -mme was introduced by analogy with duals; there are also 2nd person exclusive suffixes -mme in Quenya prayers from the 1950s such as Átaremma (VT43/8-12). Tolkien ultimately changed his mind though, saying:
All this to-do is really to accomodate Frodo’s omentielmo. But either he was wrong (and this could be noted). Or alter to elvo. Or simply have no distinctive 1a, b in plural in Q. but in dual met is exclusive because of special CE in-ke pronoun (PE17/130).
Tolkien then proceeded to spill a lot of ink trying to figure out how to resolve this question before ultimately deciding to just change the form to omentielvo in the second edition.
Other than this shift of -mme/-lme >> -lme/-lve and some vacillation over -lle vs. -lde and -lte vs. -nte, the basic pronominal suffixes seem comparatively stable in the 1950s and 60s. Dual forms, however, continued to shift, and I won’t attempt an analysis of the conceptual development until more information is published. There are definitely unpublished paradigms, such as one for the version of Frodo’s statement in the drafts of LotR, where he said omentiemman (RS/324).
Neo-Quenya: My suggestions for Neo-Quenya usage are discussed above. To summarize, they are:
- Mostly use the paradigm from Quenya Pronominal Elements.
- Use only long 2nd person familiar -tye, but be aware of short -t since others might use it.
- Mostly use -s for 3rd person singular, but where a long form is required use -sse.
- For 3rd plural use either -lte or -nte as dialectical variants, and be able to recognize both.
- Don’t use distinct forms for 3rd person inanimate; use -s (sg.) and -nte/-lte (pl.) for both animate and inanimate.