It has long been known that, under some conditions, the final vowel in Quenya words might lengthen when a suffix is added. One notable example is the phrase a vanimar, vanimálion nostari “O beautiful ones, parents of beautiful children” (Let/308, 448; LotR/981): why is the a short at the end of vanimar but long in vanimálion? I’ve seen this topic discussed in several different Quenya courses (including those of Helge Fauskanger, Thorsten Renk and Tamas Ferencz), but I don’t think those descriptions adequately capture the true nature of the phenomenon. In particular, while these descriptions address how the word changes, they don’t discuss why. I think this phenomenon can be best understood as at least partially a morphological one (based on how grammatical elements are assembled) rather than an exclusively phonological one (based only on how words are pronounced).

I called this phenomenon “prosodic lengthening”, a term that I believe (but have not confirmed) was first used by Alexander Zapragajev. The basic rule is simple: in long compounds (of four or more syllables), the stress “wants” to fall on the last syllable of the first element of the combination, to better mark the boundary between the two elements (or morphemes). This rule applies to grammatical suffixes as well, provided they consist of at least one syllable.

Nancy Martsch (author of Basic Quenya) proposed a remarkably similar system in a letter to Vinyar Tengwar written in 1990 (VT13/2). She connected the lengthening to patterns of stress, noted that it occurred before suffixes, and that the phenomenon was not limited to nouns. It seems her suggestions were largely forgotten, and I myself was unaware of them until Lokyt pointed them out to me. Properly speaking, Martsch was the first one to propose this system of prosodic lengthening, though she did not use that term.

A clear example of this rule is the compound Cuiviénen (S/48) = cuivie “awakening” + nén + “water”. When combined, the final syllable of cuivie is lengthened so that it attracts the main stress: cùiviḗnen (here and in all examples of stress: ¯ marks length, ´ marks primary stress and ` marks secondary stress). Without this lengthening, in theory the stress would retract back to the third-to-last syllable: cùivíenen, which is problematic for a number of reasons: stress falling on the first of a pair of adjacent short vowels, a light stressed syllable following a heavy stressed syllable, and so forth.

Another example is the plural Atanatári “Fathers of Men” (S/103), from singular Atanatar. Again, the vowel at the end of Atanatar is lengthened to attract stress when the plural suffix is added, àtanatā́ri, otherwise the stress pattern would be àtanátari. This is easily pronounceable but does not accurately mark the boundaries between the morphemes (grammatical elements). This example does a good job illustrating that prosodic lengthening is probably a morphological phenomenon, because from a purely phonetic perspective there is no reason it couldn’t be àtanátari or àtanā́tari.

This phenomenon of lengthening the last vowel in the inflected forms of longer (in particular compound) words ending in a single consonant was mentioned in the Declension of Nouns (DN) from the early 1930s:

With regard to lengthening of stem vowel of 2nd element [of a compound] (when it ends in a single consonant) it should be noted that Q. (cf. suffixes ending in singles cons., see above) lengthened a short vowel of a stem before a single cons. + inflection if the preceding syllable was short. That is, Q. lost ` secondary accent on short syllable normally (¯˘˘̀˘ to ¯˘̀˘˘: kíryằlĭŏn) — but where ` fell on {second >>} initial of a recognized noun, it kept its place; then a vowel bearing ` between short syllables was lengthened. Cf. nouns ending -mat, names of meals: ahtumat “supper”, [accusative] ahtumāta (PE21/32).

Prosodic lengthening is not triggered every time a suffix is added, however. In many circumstances, it is unnecessary. In Aratarya “her Sublimity” (WJ/369) = arata + -rya, the suffix begins with a consonant cluster and the syllable before it already attracts the main stress without any lengthening: àratárya. The phenomenon is not triggered by the dative suffix -n or the plural suffix -r, since these do not add a syllable to the word. The prosodic lengthening is only triggered if (a) the suffix adds a syllable and (b) a single consonant separates the suffix from the base word. The second condition can be fulfilled if either the word stem ends in a vowel and the suffix begins with a single consonant (vanimá-lion), or the word stem ends in a single consonant and the suffix begins with a vowel (Atanatár-i or Atanatár-ion).

Prosodic lengthening is also inhibited if the base word is “too short”. If the base word has only two syllables, then lengthening its final syllable would put the new stress immediately after the original stress. Thus atari “men” or lasseli “some leaves” don’t show lengthening; the stress remains where it was in the original word stem: átar-i, lásse-li. Prosodic lengthening is also inhibited in longer words where the stress falls on the second-to-last syllable for similar reasons: miruvóreva = miruvórë + -va (LotR/377), Ingoldova = Ingoldo + -va (VT39/16). Prosodic lengthening only applies if the stress in the word stem falls on the third-to-last syllable. Another way to put it is that the word must end in two light syllables (each with a single short vowel followed zero or one consonant).

Thus, prosodic lengthening is triggered when three conditions are met:

1. A suffix is added to a word with three or more syllables ending in two light syllables (stressed on the third-to-last syllable).
2. The suffix adds a (light) syllable to the word.
3. A single consonant separates the word stem from the suffix.

In this circumstance, the vowel in the last syllable of the original word stem is lengthened, to attract the stress and better mark the boundary between the word stem and the suffix.

Certain noun suffixes are much more likely to induce prosodic lengthening than others. In particular, the partitive-plural suffix -li, the instrumental suffix -nen and the possessive suffix -va all begin with a single consonant, and thus frequently trigger prosodic lengthening when added to a trisyllabic word: lintiénen “with swiftness” (PE17/59), Oroméva “Orome’s” (WJ/368). The possessive suffix is a bit of a special case: there are examples of this suffix triggering prosodic lengthening even in cases of disyllables: Ulmóva (PE19/60; PE21/80), taríva (PE17/76; PE21/14), huinéva (Ety/PHUY). This is not a pattern Tolkien followed consistently, however: forms like Ingoldova and lasseva also appear. There seems to be something special about the diphthong ui, given other examples examples like Uinéniel = Uinen + -iel (UT/182) and tuilérë = tuile + ré (LotR/1109); hat tip to Vyacheslav Stepanov for the 2nd example.

The partitive-plural suffix also has some distinct behavior: when added to a disyllabic word, the result is a trisyllable that can take further suffixes. This new trisyllable is also subject to prosodic lengthening: ciryali “some ships” becomes (instrumental) ciryalínen and (possessive) ciryalíva (Plotz). Another interesting set of examples come from the genitive partitive-plural suffix -lion, which is a disyllabic suffix that also triggers prosodic lengthening: vanimálion “of beautiful ones” (LotR/981), malinornélion “of yellow trees” (LotR/467; PE17/80). Similar behavior is seen in the genitive plural suffix -ion: Atanatárion “of the Fathers of Men” (MR/373).

These examples demonstrate that it is not really the “last syllable of the first element” that gets the stress, but rather the “last syllable of the second-to-last morpheme”: the morpheme break down of ciryalíva is cirya + -li + -va. As further suffixes are added to the word, the stress continues to shift toward the end. Thus we would expect:

• caraitië “activity” [karáitie].
• caraitiéli “some activities” [kàraitiḗli].
• caraitielínen “by means of some activities” [kàraitielī́nen].

As shown by the examples Atanatárion (atan-atar-i-on) and malinornélion (malin-orne-li-on), the second-to-last morpheme cannot always accept the stress, if this would lengthen the first vowel of a pair. In those cases, the stress moves back to the preceding morpheme: [àtanatā́rion] and [màlinornḗlion]. Such “fusional” suffixes might be treated as a unit, a single suffix in and of itself, rather than two distinct morphemes.

Note that the word malinornélion breaks the rule that the lengthening cannot occur immediately after the main stress of the original word stem. Perhaps “long suffixes” like this are an exception because the originally stressed syllable is now far enough back that it can no longer hold the main stress. If so, a more accurate description of the conditions where prosodic lengthening occurs would be “after suffixion the word ends in three light syllables” (but this is the only example where this alternate rule would be required). Also of interest is lintieryanen = lintie-rya-nen “with his speed” where lengthening does not occur, vs. lintiénen where it does (PE17/58-59). This can be explained by the fact that the stem lintierya “his speed” is stressed on the second-to-last syllable before the instrumental suffix -nen is added.

One final special noun case is that of the genitive suffix -o, which when added to a noun ending in -ie results in three sequential short vowels: sindieo “of darkness” = sindie + o (PE17/72). Tolkien seems to have recognized this as a problem, and considered inserting an -n- in such situations, in which case prosodic lengthening applies: máriéno = márie [+ n] + o (PE17/59); see the entry on the genitive for further discussion. If such a construction isn’t used, I suspect the stress pattern would still be sìndiéo even if it is not reflected in a lengthened vowel.

The phenomenon of prosodic lengthening is most frequently seen in noun declensions, but there are hints that it may be part of the verbal system as well. In notes on the active participle from the late 1960s, Tolkien described a “perfect active participle” formed from an augmentless perfect karie + active participle suffix -la to produce kariéla. There is also an example where the short 3rd person subject suffix added to a verb form likewise results in lengthening: tuluváse “he [shall] come” = tul- + -uva + -se (PE22/162). This is not a common pattern for subject suffixes because most of them begin with two consonants.

In the Quenya Verbal System document written in 1948, Tolkien described various compound verb tenses assembled from multiple verb suffixes which may or may not demonstrate prosodic lengthening: for the verb [ᴹQ.] kar-, there is future karuva, perfect akárie, future-perfect akáriéva (PE22/109). Another example adds the past tense suffix -ne to the future karuva to produce the past-future †karuváne (archaic), later reformed to karúvane (PE22/105). This document has a larger variety of object suffixes than what appears in Tolkien’s later writing. Many of these begin with a single consonant and seem to trigger lengthening of the preceding vowel: [ᴹQ.] efarále “[he] hunts you” = e + fara- + le (PE22/110) and ni·karnéte “I made it” = ni + kar- + -ne + te (PE22/118).

It is hard to tell whether these verb formations represent genuine prosodic lengthening or whether they represent the preservation of an ancient long vowel in the phonetic development of the inflection. In the example of the future-past, Tolkien explicitly said it was originally derived from ancient káruvā̀-nē (PE22/105). Indeed, it is plausible that such preserved long vowels are the origin of prosodic lengthening in the first place. The words undergoing such lengthening very often primitively ended in a long vowel. They may have been immune to the ancient Quenya stress shift in their inflected forms, thereby preserving the vowel length in such cases. However, there are definitely cases of prosodic lengthening where no long vowel existed primitively, such as Atanatári (S/103) and Ilúvatáro (PE21/83), so it seems the lengthening rule was generalized.

There are some cases where prosodic lengthening is expected but does not occur. Some of the examples that show lengthening also appear without it: malinornelion, vanimalion and Ilúvataro are all attested (PE17/80; PE17/111; S/322). There are also some compound forms that should lengthen but don’t, such as Tasarinan “Willow-vale” = tasari + nan (LotR/469). It is unclear whether these represent some real variation or are just cases where Tolkien forget to apply the rule.

Conceptual Development: There are examples of “special” vowel lengthening rules at the end of suffixed long words dating back as far as the Early Qenya Grammar of the 1920s:

Indefinite article. “a”, in pl. “some, certain”, is suffixed -ma. Trissyllabic nouns of which the penultimate syllable is short, lengthen the final vowel, as: tantare “dance”; tantaré·ma “a dance” (PE14/42).

Where the simple forms occur in long words the preceding vowel is usually long in inflected forms: as falmarin “sea-fay” (masc.), pl. falmaríni (PE14/43).

Sg. tantare, ondo ... Pl. tantarēli, ondoli (PE14/44).

It seems likely Tolkien introduced the idea of prosodic lengthening very early, and kept it as part of Quenya throughout his life. It is a reasonable compromise between the rigid stress system of Quenya and basic pronounceablity.