The allative suffix -nna indicates motion towards the declined noun, and is variously glossed “to, towards, onto, upon”. Thus Endorenna “to Middle Earth” (LotR/967), tielyanna “upon your path” (Let/448), Amanna “towards Aman” (VT49/26). It is not restricted to horizontal movement:
- nai laure lantuva parmastanna lúmissen tengwiesto “may (a) golden light fall on your book at the times of your reading” (VT49/47).
In this sentence the light is falling downwards towards parmasta “your [dual] book”, so here -nna indicates motion “on[to]” the book. The allative suffix can also be used temporally, of motion through time towards a specific
(future) moment. As such, it is often used with the future tense:
- nai elen atta siluvat aurenna veryanwesto “*may two stars shine upon the day of your wedding” (VT49/45).
But it can also be used with the present tense, most famously in:
- Elen síla lúmenn’ omentielvo “a star shines on the hour of our meeting” (LotR/81).
Perhaps the use of the allative for lúmenna “upon the hour” (WJ/367) emphasizes the arrival at the current moment from some time in the past, though it could also have
to do with the shining of the starlight “towards” the hour. Presumably the allative could also be used with the past tense
when indicating arrival at a moment from some time further in the past.
Forming the allative: With vocalic nouns, the allative suffix -nna is simply added to the noun. With consonantal nouns ending in -n, the suffix is almost always assimilated to the end of the noun as a “short allative”, so that from Aman you get Amanna “towards Aman” (VT49/26). But with other consonants a “joining vowel” -e- is usually inserted between the noun stem and the suffix: Elendilenna “to Elendil” (PM/401). See the joining-vowel discussion in the entry on the adverbial cases for further details.
The allative plural adds the suffix -nnar to vocalic noun stems: ciryannar, lassennar (Plotz). Consonantal nouns use a joining vowel of -i- rather than -e-: mindoninnar “to the towers” vs. assimilated plural allative mindonnar (MC/222). The suffix -nna(r) is added to partitive plural forms: ciryalinna(r) (Plotz). Since the partitive plural suffix already marks the plural, the final plural marker -r is redundant and therefore optional: either ciryalinna or ciryalinnar is acceptable.
In the case of t-dual nouns, the ancient inflection was -t + na which (after metathesis) became -nta: ciryanta “to the pair of ships” (Plotz). We don’t have any u-dual allative examples, but presumably the -nna suffix is added directly to the dual form: aldunna “towards the two trees”.
To summarize:
- For vocalic nouns (including e-nouns), the suffix -nna is used in the singular and -nnar in the plural.
- For consonant nouns, the suffix -enna is used in the singular and -innar in the plural.
- If the noun ends in n, the suffixes can reduce to “short allative” -na and -nar.
- For partitive plurals, the suffix -nna(r) is added; the (r) is optional.
- The t-dual allative suffix is -nta, for u-duals (probably) the suffix -nna is added.
| Allatives | Sg. | Du. | Part. Pl. | Pl. |
|---|---|---|---|---|
| vocalic: cirya | ciryanna | ciryanta | ciryalinna(r) | ciryannar |
| e-noun: lasse | lassenna | lassenta | lasselinna(r) | lassennar |
| consonantal: nortil | nortilenna | *nortilunna | *nortillinna(r) | nortilinnar |
| consonantal (final n): atan | atanna | *atanunna | *atanelinna(r) | atannar |
Forms marked with a * are unattested and hypothetical.
Origins of the allative: As discussed in the entry on adverbial cases, the allative was original just a suffix used to form adverbs from noun, roughly equivalent to the English suffix “-wards”.
Amanna originally mean “Aman-wards”, and this old function can be seen in some ancient names like Elenna “Starwards” (S/261).
Common Eldarin originally had a different allative suffix, however: ✶-da (WJ/366; PE21/76) which was lost in Quenya as an inflectional element. As Tolkien described it in Common Eldarin: Noun Structure written in the early 1950s:
It is also notable in Quenya that the “allative” has the forms -n, nna in declensions, and not da, or nda. This was probably one of the causes of the loss of the -n genitive in Q. It might be assumed that na, nna is a primitive variant of da, nda; but this is not found in other Eldarin languages, not even in the closely associated V[alarin] Telerin, and it is more probable
that nna was developed in Q. from nda as a parallel to llo, sse. The influence of the adverbial/prepositional stem ANA/NA has also been assumed, but the sense and functions of ANA are not
really very close. It is true that in Q. na appears with sense “to”, but this at nearest means “towards, to a position near, alongside”; whereas the allative/dative
n/nna means “up to, to, at (arriving at the point)”. The original sense of Eldarin ana was plainly “at side of, alongside, besides,” hence also “moreover, in addition, plus” (seen in use of an- as an intensive prefix), and so an or na in some languages has the sense “along with, with, accompanied by, provided with, associated with” and the like. Cf. Bel.
[Sindarin] na which forms virtually adjectival expressions: as Taur na Foen “The Forest of Foen” (i.e. which included the mountain called the foen) (PE21/79).
Thus Q. -nna was derived from or influenced by the primitive root √NA/ANA of similar meaning. Tolkien indicated that the Quenya dative was derived from the same root, so that the dative and allative were etymologically related:
Tolkien wrote in a late isolated note, “The ‘dative’ -n was of course in origin a reduction of -na ‘to’.” This note also states that adverbial -nna (i.e., the allative case ending) was of the same origin (notes by Patrick H. Wynne, VT49/14).
Etymological notes from the late 1950s also confirm that these forms were related to the Sindarin dative an and genitive na (PE17/146-7).
Conceptual Development: The first mention of the allative suffix was in a list of “adverbial suffixes” in the Early Qenya Grammar from the 1920s but the suffixes were -nta or -tta, as mentioned in both the manuscript (PE14/46) and typescript (PE14/78) versions. In the manuscript version the consonantal
forms use the joining vowel -i-: -inta, -itta (PE14/47), but in the typescript version the joining vowel was -a-: -anta, -atta (PE14/78). In both versions “true consonantal” nouns (those that primitively ended in a single consonant) often used short
allative -ta, presumably with various assimilations. Originally the suffixes were not true inflections and could not be applied to plural
nouns, but eventually a plural form -ntar/-ttar developed (PE14/47, 79).
This system persisted in declension charts later in the 1920s, except the variant -tta vanished (PE16/113). Tolkien also introduced a dual form -ntat, and the plural form became -nta(n), with a final n rather than r. The (n) was marked as optional and redundant, since the inflection was added to the plural form:
- kirya → (plural) kiryali → (allative plural) kiryalinta(n).
- pilin(d-) → (plural) pilindi → (allative plural) pilindinta(n).
Tolkien retained this basic system in the lengthy Declension of Nouns from the early 1930s except the (n) was no longer optional in the plural: kiryalintan (PE21/4). For consonantal nouns, Tolkien also described the occasional use of an “abbreviated” inflection derived from -tā (PE21/18). He gave quite a few examples of consonantal inflections, some with “short allative” -ta and others with the full allative and a joining vowel, which was now -u-. Allative plurals for consonantal nouns continued to use the joining vowel -i-, though archaic allative sometimes added just the plural marker -n (PE21/20-37):
- nēr → (allative) nerta → (archaic allative plural) †nertan beside more regular nerintan.
- tāl → (allative) talta → (allative plural) †taltan vs. talintan.
- kas → (allative) kasta → (allative plural) †kastan vs. karintan.
- nēn → (allative) nēnunta.
- hōn (hom-) → honta.
- hun → hunta.
- sūt → sutta.
- yāt (yak-) → yahta.
- māl → malta or malunta.
- tet → tetta or tetunta.
- sat (sap-) → †sapsa (archaic) or sapunta.
- qen (qend-) → †qendata or qendunta.
- let (leps-) → †lepsata or lepsunta.
- fas (fass-) → †fassata or fassunta.
- nin (ning-) → †ningata or ningunta.
- nil (nild-) → †nildata or nildunta.
- mar (mard-) → †mardata or mardunta.
- hat (haht-) → †hahtata or hahtunta.
- laman (lamn-) → lamnunta or lamunta.
- aran (arn-) → arunta.
- olar → †olallo or olarunta.
- ambor (ambos-) → ambosta or †ambusta.
- qilir (qiles-) → qilesta or qilirunta.
- andul → andulta or andulunta.
- Earendel → Earendilta, Earendelta or Earendelunta.
- pilen → pilinta or pilenta.
- aman → amanta or amanunta.
- soron → sorunta or soronunta.
- oron (orum-) → orunta or oronunta.
- helen (helem-) → helinta, helenta or helemunta.
- qelet → qelta or qeletunta.
- kelut → kelutta or kelutunta.
- talat → talatta or talatunta.
- filet (filek-) → †filihta, filehta or filekunta.
- arat (arak-) → arahta or arukunta.
- Tinúviel → Tinúvielta or Tinúvielunta.
- falmarin → falmarinta or (rare) falmarinunta.
- pilin (pilind-) → †pilinta, †pilindata or pilindunta.
- miqilis (miqilist-) → †miqilistata or miqilistunta.
- veaner → veanerta or veanerunta.
- Valinor → †Valinórata or Valinorunta.
- Koiviénen → Koivienenta, †Koivienénata or Koivienenunta.
- ahtumat (ahtumát- or ahtumatt-) → ahtumatta or ahtumatunta.
- telumet (telumett-) → telumetta or telumet(t)unta.
- Astulat (Astulaht-) → Astulahtta or Astulahtunta.
- peltas (peltaks-) → peltaksata or peltaksunta.
- kaimasan (kaimasamb-) → kaimasanta, †kaimasambata or kaimasambunta.
The general trend was to add -ta where the combination with the stem was phonologically suitable, and -unta (older †-ata) when it was not. There were various irregularities and minor vowel shifts in what was originally the last syllable of the
noun (e → i, o → u) along with occasional assimilations like kt > ht and pt > ps. Plurals tended to added -intan to the stem even in cases where the singular added -ta.
In declensions later in the 1930s the allative became -nde with plural -nden (PE21/42, 46, 50). For consonantal nouns, the short allative became -de and the joining vowel switched back to -a- (PE21/52):
- nēr (ner-) → nerde or nerande.
- pilin (pilind-) → †pilinde or pilindande.
- hōn (hom-) → †honde or homande.
There was then a brief restoration of -nta (PE21/54), but starting in the mid-1930s the -nna allative (along with its plural -nnar) finally appeared in The Etymologies and the contemporaneous poem Firiel’s Song:
- ᴹQ. telmello talmanna “crown to foot” (Ety/TEL; EtyAC/TEL).
- ᴹQ. Tar-Kalion ohtakáre Valannar “[Tar-Kalion] war-made on-Powers” (LR/47).
Tolkien stuck with -nna thereafter. There aren’t many examples of allative declensions for consonantal nouns in the 1940s, 50s or 60s, so details
there (when short allatives could be used, how joining vowels work) are hard to deduce.
The full set of conceptual developments is given in the table below, using the version numbers for the declension charts from
PE16 and PE21, with version 0 for the Early Quenya Grammar and LQ for Late Quenya forms (in Plotz and elsewhere). Sh. Pl. = “Short Plural” are for shorter plural forms where they exist, adding the allative plural suffix directly to a vocalic stem
without the plural suffix; this became the norm as of version 6. This table omits archaic or less typical forms.
?
| V | Allatives | Sg. | Du. | Pl. | Sh. Pl. |
|---|---|---|---|---|---|
| 0 | vocalic | -nta/-tta | -linta(r)/-litta(r) | -ntar/-ttar | |
| 0 | consonantal | -anta/-atta | -intar/-ittar | ||
| 3 | vocalic | -nta | -ntat | -linta(n) | |
| 3 | consonantal | -(a)ta | -(a)tat | -inta(n) | |
| 4 | vocalic | -nta | -ntau/-nto | -lintan | -ntan |
| 4 | consonantal | -(un)ta | -(un)tau/-(un)to | -intan | |
| 5a | vocalic | -nde | -ndet | -linden | -nden |
| 5b | vocalic | -nde | -ndet | -linde(n) | -nden |
| 5c | vocalic | -nde | -ndet | -linde(n) | -nden |
| 5c | consonantal | -(an)de | -(an)det | -inden | |
| 6 | vocalic | -nta | -ntas | -ntar | |
| LQ | vocalic | -nna | -nta | -nnar | |
| LQ | consonantal | -(en)na | -innar |