Also called: Danian, the Silvan tongue, Silvan Elvish
During the long hike from Cuiviénen to the Sea, some Telerin Elves refused to cross the terrifying Misty Mountains. They forsook the March to the sea, where Ulmo would bring the Elves to Valinor (Silm ch. 3). In Quenya, these Elves were later called the Nandor or "Those who go back", though it seems that none of them actually returned into the East; they just stayed behind the Hithaeglir (WJ:384). Lead by one Denethor some of the Nandor eventually did enter Beleriand, though they had lost the boat to Valinor by several millennia. They settled in Ossiriand, which region they renamed Lindon, and by the Sindar they came to be called Green-elves (Sindarin Laegil, Laegelrim). Concerning the relationship between the Green-elven and Grey-elven tongues, it is stated that "although the dialects of the Silvan Elves, when they again met their long separated kindred, had so far diverged from Sindarin as to be hardly intelligible, little study was needed to reveal their kinship as Eldarin tongues" (UT:257). WJ:385 confirms that the Sindar recognized the Green-elves "as kinsfolk of Lindarin origin (...), using a tongue that in spite of great differences was still perceived to be akin to their own".
However, all that is known of the Nandorin tongue is about thirty words, most of which are found in the Etymologies. Wrote Tolkien, "Though the comparison of the Silvan dialects with their own speech greatly interested the loremasters, especially those of Noldorin origin, little is now known of the Silvan Elvish. The Silvan Elves had invented no forms of writing, and those who learned this art from the Sindar wrote in Sindarin as well as they could" (UT:257). Elsewhere, he refers to "the now almost entirely lost tongue of the Nandor" (VT47:39).
Some of the Sindar that came to Thranduil's realm escaping the destruction of Doriath adopted the Nandorin tongue and took names of Silvan form and style, just like the Noldor had adapted their Quenya names to Sindarin centuries earlier. These Sindar "wished indeed to become Silvan folk and to return, as they said, to the simple life natural to the Elves before the invitation of the Valar had disturbed it" (UT:259). Yet Sindarin somehow sneaked into even the Silvan communities: "By the end of the Third Age, the Silvan tongues had probably ceased to be spoken in the two regions that had importance at the time of the War of the Ring: Lórien and the realm of Thranduil in northern Mirkwood. All that survived of them in the records was a few words and several names of persons and places" (UT:257). Nimrodel would only speak the Silvan tongue even after it had fallen into disuse in Lórien; see UT:241. UT:252-253 suggests that the name Lórien itself may be altered from Nandorin Lórinand, "Valley of Gold (golden light)", or even older Lindórinand "Vale of the Land of the Singers (= Lindar, Teleri)". According to a footnote in Appendix F of the LotR, not only Lórien but also the names Caras Galadhon, Amroth and Nimrodel "are probably of Silvan origin, adapted to Sindarin".
There is not much we can say about the structure of Nandorin. Very little grammar can be extracted from the few words we have. A Sindarin-style umlaut-plural is seen in urc "Orc" pl. yrc (Sindarin orch, yrch). This umlaut must have developed independently of the Sindarin umlaut on the other side of the Misty Mountains (there is no trace of umlaut in Quenya and the Telerin of Aman, languages that evolved from Common Eldarin after the separation of the Nandor from the other Eldar, just like Sindarin did). In Lindi, the name the Nandor had for themselves, a descendant of the old Primitive Quendian plural ending -î is still present. Does the ending -on of Caras Galadhon indicate genitive plural, cognate with and identical to the corresponding Quenya ending? This would give the name the plausible meaning *"fortress of trees". Galadh "tree" could be Sindarin, but this language has no genitive endings. (The name should probably be Galadon in pure Nandorin, since the word for "tree" is given as galad in MR:182; the form Galadhon would be "adapted to Sindarin" by changing d to dh.)
The names Nimrodel "Lady of the White Grotto" and Amroth "Up-climber" are probably of Nandorin origin, but in a footnote in Appendix F they are said to be "adapted to Sindarin" and may be somewhat altered from their original form; hence, they are not included here. (See UT:457, 245 concerning their meaning.) The name Caras Galadhon is also said to be so adapted and is therefore excluded from this list, but caras, that is given independently in UT:257, is included. - In the etymological notes, primitive forms "reconstructed" by Tolkien himself are not asterisked.
alm "elm-tree", probably from *almâ, formed from the stem ÁLAM "elm-tree" (LR:348 - notice that Quenya alalmë and Sindarin lalf clearly descend from different, if related forms). Based on other Nandorin forms we might have expected *ealm or *elm instead.
beorn "man", stated to descend from besnô "husband" (stem BER "wed", LR:352), but "blended with ber(n)ô", sc. "valiant man, warrior", derived from the stem BER "valiant" (LR:352). The shift of e to eo is strange and has no direct parallels, but compare eo from i in meord "fine rain" (< primitive mizdê). Normally final -ô becomes -â in Nandorin (see golda), but here it is simply lost instead of producing *beorna. Cf. meord, the other word where we might have expected to see a final -a (in that case from -ê); it may be that final vowels are lost in words that would otherwise come to have more than two syllables. - The shift of primitive s to r in besnô > beorn may be ascribed primarily to the blending with ber(n)ô, but r from z is seen in meord < mizdê; perhaps the s of besnô first became z and then r. Such developments are common in Quenya.
caras "moated fortress" (UT:257), probably to be compared to Sindarin ("Noldorin") caras "a city (built above ground)", derived from the stem KAR "make, do" (LR:362); the basic meaning may be simply "something made, construction" (compare Quenya car "building, house"). Extensions involving a suffixed stem-vowel and a final -s are attested; cf. for instance SPAL and its extended form SPALAS (LR:387). Hence, KAR could easily have a longer form *KARAS. Sindarin caras evidently includes the Grey-elven derivative ending -as, though (-as is basically used to derive verbal nouns, like English -ing, but these may often take on a more concrete meaning; car-as can probably be compared to English build-ing); the Nandorin ending may be related to the Sindarin one. Yet another possibility would be equating this ending -as with the collective ending seen in Danas, q.v., and assume that car means something like "house" (as in Quenya); hence caras = "group of houses, village", later taking on the meaning "moated fortress" if the Nandor used to surround their villages with moats.
cogn "bow", primitive form given as ku3nâ, derived from KU3 "bow" (LR:365); it is probable that ku3nâ was originally an adjective "bow-shaped", since -nâ is predominantly an adjectival ending. As for 3 (spirant g) becoming a plosive g, compare garma.
cwenda "Elf" (a doubtful word according to Tolkien's later conception; in the branch of Eldarin that Nandorin belongs to, primitive KW became P far back in Elvish linguistic history [WJ:375 cf. 407 note 5]. This was not a problem in Tolkien's earlier conception, in which the Danians came from the host of the Noldor, not the Teleri [see PM:76; the idea of the Nandor being of Noldorin origin also occurs in VT47:29]. In his later version of Nandorin, the word cwenda is probably best ignored; simply emending it to *penda would produce a clash with primitive pendâ "sloping" [cf. WJ:375].) In the Etymologies, Tolkien derived cwenda from kwenedê "elf" (stem KWEN(ED) of similar meaning, LR:366; as for the shift of original final -ê to Nandorin -â, compare hrassa "precipice" from khrassê). But later the primitive word that yielded Quenya Quendë was reconstructed as kwende (WJ:360). No certain example shows how original short final -e comes out in Nandorin, so we cannot say whether kwende is also capable of yielding cwenda, ignoring the question of kw failing to become p.
Danas "Green-elves, Nandor". In Etym derived from the stem DAN (LR:353), simply defined as an "element found in names of the Green-elves", and tentatively compared to NDAN "back" (since the Nandor "turned back" and did not complete the march to the Sea). Tolkien's later view on the derivation of the name of the Green-elves, as set down in WJ:412, is that the stem dan- and its strengthened form ndan- do indeed have a similar meaning: these forms have to do with "the reversal of an action, so as to undo or nullify its effect", and a primitive form ndandô, "one who goes back on his word or decision", is suggested. However, it seems unlikely that the Nandor would have called themselves by such a name, and indeed Tolkien in WJ:385 states that "this people still called themselves by the old clan-name Lindai [= Quenya Lindar], which had at that time taken the form Lindi in their tongue". It may be, then, that Tolkien had rejected the idea that the Nandor called themselves Danas. - As for the ending -as, it is probably to be compared to the Sindarin class plural ending -ath; indeed a Sindarin ("Noldorin") form Danath evidently closely corresponding to Danas is given in LR:353.
dóri- "land", isolated from Lindórinan. The independent form of the word may differ; it is unclear where the i of the compound Lindórinan comes from. In the Etymologies, the Eldarin words for "land" are derived from a stem NDOR "dwell, stay, rest, abide" (LR:376). No Nandorin word is there listed, but Sindarin dor is derived from primitive ndorê. Notice, however, that Tolkien many years later derived the Eldarin words for "land" from a stem DORO "dried up, hard, unyielding" (WJ:413). However, this later source does confirm that the Primitive Quendian form was ndorê, now thought to be formed by initial enrichment d > nd. This is defined as "the hard, dry land as opposed to water or bog", later developing the meaning "land in general as opposed to sea", and finally also "a land" as a particular region, "with more or less defined bounds". Whether dóri- actually comes from ndorê is highly doubtful (this would rather yield *dora in Nandorin), but it must be derived from the same set of stems.
dunna "black"; this might seem to be derived from *dunnâ, sc. the stem DUN "dark (of colour)" (LR:355) either with the adjectival ending -nâ or with medial fortification n > nn and the simpler adjectival ending -â. However, other Nandorin words seem to have lost their final -â's, e.g. ealc "swan" from alk-wâ, and (to quote a wholly parallel example) cogn "bow" from ku3nâ. The descendant form is not *cogna with the final vowel intact as the case would seem to be in dunna. However, primitive -ô does come out as -a in Nandorin, cf. golda "Noldo" from ñgolodô, so a form *dunnô might be capable of yielding dunna, but this primitive form would rather be a noun *"dark person/thing", since primitive -ô, -nô are nominal rather than adjectival endings. Of course, Nandorin may have turned an original noun into an adjective, or developed an adjectival ending -a afresh. But all things considered *dunnâ still appears to be the best reconstruction of the primitive form. The words dunna and scella (see below) raise the question of whether original final -â is actually preserved as -a following double consonants (as opposed to clusters of different consonants) in Nandorin.
ealc "swan", primitive form given as alk-wâ, derived from a stem ÁLAK "rushing" (LR:348); alk-wâ would seem to be an adjectival formation (ending -wâ), so the primitive word probably had the same meaning as the stem: "rushing", later used as a noun "rushing (one)" and applied to an animal. According to Tolkien's later conception, kw should probably have come out as p rather than c in Nandorin; see cwenda. Primitive a becoming ea is a strange shift with no direct parallels even where it might have been expected, but compare eo from i in meord (and from e in beorn), as well as ie from a in sciella. Perhaps we are to understand that the liquids l, r trigger such changes in a preceeding vowel, but then we might expect for instance *ealm instead of alm as the word for "elm-tree".
edel "Elda, High-elf". Stated in the Etymologies to be derived from a stem ÉLED (LR:356), defined as "Star-folk"; Tolkien points out that Doriathrin and Danian used a "transposed" form, clearly referring to the sounds L and D changing place. In Etym the development is apparently meant to be eledâ (this primitive form is explicitly given in Letters:281) > edela > edel. Later Tolkien reconstructed the primitive form of Quenya Elda as eldâ (WJ:360); whether this could regularly yield Nandorin edel is doubtful, unless final -ld metathesized to -dl and a vowel developed to break up this final cluster. - In Etym, Tolkien first gave the Nandorin form as elda, then changed it. *Eledâ could not yield elda, since final -â is regularly lost in Nandorin. Since in this word we do not see loss of the second of two identical vowels (compare golda), we must conclude that
final -a was lost before this could happen.
enel "in the middle, between". The sole known Nandorin preposition, this form was supposedly recorded by the Loremasters (VT47:39). It is derived from a variant of the root ÉNED- "centre" (LR:356), since "d and l interchanged frequently in Common Eldarin" (VT47:39).
galad "tree" (MR:182). Derived from galadâ "great growth", "tree", applied to stout and spreading trees such as oaks and beeches (UT:266, Letters:426; in the latter source, the root GAL is defined as "grow", intransitive). It is interesting to notice that this word, given in a source much later than the Etymologies that
provides most of the Nandorin material, nonetheless agrees well with the older words cited by Tolkien: again we see the loss of original final -â, whereas original post-vocalic d is unchanged as in the word edel.
[garma "wolf" -3ARAM (LR:360, struck out)] Since Quenya and "Noldorin" = Sindarin cognates were given as harma and araf (rejected along with garma), the primitive form would be *3aramâ. The word golda "Noldo" confirms that Nandorin would lose the second of two identical vowels in adjacent syllables; however, other examples indicate that final -â would simply disappear instead of yielding -a. See for instance ealc.
golda "Noldo". The primitive form of Quenya Noldo (and hence also Nandorin golda) is given in WJ:364, 380 as ñgolodô. This example demonstrates that in Nandorin, like in Quenya, the second of two identical vowels in adjacent syllables is lost in words that had another syllable following the lost vowel. This word alone provides a clear example of the change of primitive final -ô to -a. The form golda also suggests that in Nandorin as in Sindarin, the original initial nasalized stops ñg, nd, mb were simplified to g, *d, *b, though examples for *d and *b are lacking in our very small corpus. The stems involved are found in LR:377: ÑGOL "wise" and the extended form ÑGOLOD "one of the wise folk". Ñgolodô is thus either formed from ÑGOL by ómataina (suffixed base-vowel), suffixed D and the nominal (often masculine or agental) ending -ô, alternatively simply the longer ending -dô (of similar meaning) suffixed to the ómataina-form of the stem ÑGOL (sc. ñgolo-).
hrassa "precipice". Primitive form given as khrassê, derived from a stem KHARÁS (LR:363) that is not defined, but compared to a stem KARAK "sharp fang, spike, tooth" (LR:362). The form khrassê displays the loss of an unaccented stem-vowel often seen in primitive words (cf. for instance d'râk- "wolf" from DARÁK-); the ending -ê is found on a number of words denoting inanimates (though it is also a feminine ending). For the doubling of the final s, compare lassê "leaf" from LAS1 (LR:367). This hr- is our only example of how primitive khr- comes out in Nandorin; hr is surely meant to denote an unvoiced r, as in Tolkien's later spelling of Quenya words (e.g. hroa "body"). For primitive -ê becoming Nandorin -a, compare cwenda (q.v.) from kwenedê.
Lindi, what the Nandor called themselves, a cognate of Quenya Lindar (Teleri) (WJ:385). The sg. is probably *lind, perhaps attested in the name Lindórinan. This form is stated to descend from the older clan-name Lindai (WJ:385), or at the oldest stage Lindâi (WJ:378). Lindâ was originally the name of a member of the Third Clan of the Elves, among the Eldar also called the Teleri; the Nandor came from this branch of the Eldarin peoples. In WJ:382, Lindâ is stated to be derived from a stem LIN, the primary reference of which is to "melodious or pleasing sound"; Lindâ, derived by medial fortification and adjectival -â, would seem to be in its origin an adjective, but later applied to the third clan of the Elves and eventually used as a noun. The reference was to their love of song (notice that Tolkien translated the name Lindórinan as "Vale of the Land of the Singers"; UT:253). The Nandorin word Lindi alone in our small Green-elven corpus shows a direct descendant of the Primitive Quendian ending -î, while the sole other attested Nandorin plural is formed by umlaut: urc "Orc" pl. yrc. Perhaps the ending -i persisted in the case of words that had the stem-vowel i, since this vowel could not be changed by umlaut (being already identical to the vowel causing the umlaut so that no assimilation was possible); therefore, singular and plural would become identical if the plural ending -i had been dropped as in yrc. (It may not be necessary to invoke the simple "real-world" explanation that Tolkien's ideas about Nandorin had changed during the thirty years that separate the source that has yrc from the source that provides the word Lindi.)
Lindon region in eastern Beleriand where the Green-elves settled, formerly called Ossiriand (WJ:385). The idea that Lindon is a Nandorin word is not found in the Etymologies; here the word is said to be Ilkorin instead, derived from Lindân-d (LR:369 s.v. LIN2) and defined as "musical land" ("because of water and birds"). However, the name Lindon in Tolkien's later conception represents primitive Lindânâ (WJ:385), which is clearly Lindâ "Linda, Elf of the Third Clan" + the well-attested adjectival ending -nâ. Lindânâ therefore means simply "(Land) of the Lindar", "Lindarin (Land)". Interestingly, this Nandorin word from a source much later than the Etymologies confirms the loss of original final -â seen in many words listed in Etym. Lindon from Lindânâ is also our sole example of how medial â comes out in Nandorin; it seems to become o. (Cf. Doriathrin, in which language primitive medial â becomes ó.)
Lindórinand "Vale of the Land of the Singers (= Lindar, Teleri)", "Lórien" (UT:253). The elements must be lind- "singer, Linda" (cf. pl. Lindi above), dóri- "land" (the independent form may differ; see dóri-) and nand "valley" (q.v.).
Lórinand "Valley of Gold (golden light)", "Lórien". Altered from Lindórinand, q.v. (UT:252-253). This word would seem to point to lóri- (the independent form may differ somewhat) as the Nandorin word for "gold, golden light", transparently a derivative of the stem LÁWAR (LR:368) that covers precisely this meaning; a primitive form laurê is given in the Etymologies. This word alone testifies a Nandorin shift au > ó. (However, the final vowel of laurê would be expected to come out as -a in Nandorin; cf. hrassa from khrassê; it may be that lóri- rather represents a colour-adjective *lauri; if so the final -i may be preserved before endings and in compounds only, the independent form being *lór.)
lygn "pale". Primitive form given as lugni "blue", sc. the stem LUG1 (LR:370, not defined) with an ending -ni not otherwise attested, though -i is an ending found on many primitive colour-adjectives. The ending -i causes umlaut u > y; compare yrc as the plural of urc "Orc". That a short original final -i is capable of causing such an umlaut at the Common Eldarin stage is somewhat surprising, since Primitive Quendian lugni should have become *lugne at this stage, and final e would hardly cause umlaut. Perhaps we are to understand that the change of final i to Common Eldarin e happened relatively late, after the Eldar had crossed the Hihtaeglir and parted with the Nandor?
meord "fine rain". Primitive form given as mizdê, derived from a stem MIZD (LR:373) that is not defined, but Christopher Tolkien is undoubtedly right in observing that the stems MISK (yielding words for "wet") and MITH (yielding words for "wet mist" and "grey") are probably meant to be related to MIZD. The ending -ê seen in mizdê seems in this case to denote a substance. While final -ê sometimes becomes -a in Green-elven, it has here been lost; see beorn for some thoughts about this. This word alone shows eo from i, but cf. eo from e in beorn.
nand "valley", isolated from Lindórinand, Lórinand (q.v. for reference). While this word is not given in the Etymologies, it is clearly derived from the stem NAD (LR:374) and hence a close cognate of the similar Doriathrin word nand "field, valley". The Quenya cognate nanda (meaning "water-mead, watered plain") indicates a primitive form *nandâ; as in most cases, the final -â is lost in Nandorin.
scella, sciella "shade, screen" (prob. noun). Primitive form given as skalnâ, derived from the stem SKAL1 "screen, hide (from light)" (LR:386). Since -nâ is an adjectival ending, often taking on the meaning of a kind of past participle, skalnâ must mean "screened, hidden (from light)"; this has become a noun "shade, screen" in Nandorin. The word scella, sciella alone tells us that ln is assimilated to ll in Nandorin, and as in dunna, spenna a primitive final -â, usually lost, seems to persist as -a following a double consonant. The shift of a to e in skalnâ > scella is parallelled by the similar shift in *spannâ > spenna, q.v. However, such a shift does not occur in what might seem to be similar environments (before a double consonant?); cf. hrassa, not *hressa, from khrassê. It would seem that e might further break up into ie, scella having the alternative form sciella.
snæ^s "spear-head, point, gore, triangle". Original form not entirely clear; the stem is SNAS/SNAT (LR:387), not defined but evidently to be understood as a strengthened form of NAS "point, sharp end" (LR:374). A primitive plural form natsai is mentioned under SNAS/SNAT; snæ^s may derive from something like *snatsâ via *snats, *snas. The shift of original a to long æ^ (presumably the same vowel as in English cat, but longer) is found in this word only, but there are several examples of e from a, see spenna, scella. Perhaps a became æ^ in stressed monosyllables where there was no following consonant cluster (as in nand).
spenna "cloud". Derived from a stem SPAN "white" (LR:387), but hardly a direct cognate of Quenya fanya and Telerin spania (both probably from *spanjâ), nor a direct cognate of Sindarin faun, stated to be derived from spâna. Rather spenna must derive from *spannâ, sc. the stem SPAN with the adjectival ending -nâ (or possibly the simpler adjectival ending -â combined with a medial strengthening n > nn). As for the change of a to e, cf. scella from skalnâ.
swarn "perverse, obstructive, hard to deal with". Derived from a stem SKWAR "crooked" (LR:386); the primitive form was undoubtedly *skwarnâ with the adjectival ending -nâ. In this case the final vowel of this ending is lost, while it seems to persist in dunnâ < *dunnâ, scella < skalnâ and spenna < *spannâ; may the vowel be preserved following double consonants only?
urc (pl. yrc) "Orc". In the Etymologies, the primitive form of this word is given as órku (defined as "goblin"), derived from an undefined stem ÓROK (LR:379). This stem may be understood as a vowel-prefixed variant of the stem ROK "horse", assuming that this originally referred to the steed of the monstrous "dark Rider upon his wild horse" that haunted the Elves by Cuiviénen, assuming that the stem ROK was originally associated with Melkor's creatures. However, Tolkien later derived the Elvish words for "Orc" from a stem RUKU having to do with fear (WJ:389) and listed tentative primitive forms: urku, uruku, urkô. Since primitive final -u is lost in Nandorin (cf. Utum from Utubnu), the forms urku and uruku would evidently be capable of yielding Green-elven urc (while urkô would rather come out as *urca; cf. golda "Noldo" from ñgolodô). The plural form yrc clearly shows umlaut caused by the lost Primitive Quendian plural ending -î; cf. the umlaut caused by the primitive adjectival ending -i, primitive lugni "blue" yielding lygn.
Utum "Utumno", Melkor's first stronghold. The primitive form is given as Utubnu, derived from a stem TUB (LR:394), not defined as such but yielding a series of words that suggest a basic meaning "deep, lowlying". The prefixing of the stem-vowel is a common feature in strengthened primitive forms; the ending -nu seems to be used nowhere else, but Utubnu is clearly to be understood as *"very low [place]". The original cluster bn comes out as m in Nandorin; cf. Quenya Utumno. The development is evidently meant to be Utubnu > *Utumnu > *Utumn > Utum.