Old Sindarin - between Primitive Elvish and Grey-elven

Old Sindarin - between Primitive Elvish and Grey-elven

Also called: Old Noldorin (the only term Tolkien uses!)
INTERNAL HISTORY

Old Sindarin is the last stop before (Classical) Sindarin in the evolution from Primitive Quendian to mature Grey-elven (between PQ and OS we have Common Eldarin and Common Lindarin). It was developed and spoken in Beleriand, but it appears to have evolved into mature Sindarin by the time the Noldor returned. Old Sindarin preserves the general sound of archaic Elvish much better than Sindarin does. There are quite a few cases of Old Sindarin words being identical to Quenya words: Examples include ku, kua "dove", malina "yellow", míre "jewel", parma "book", randa "cycle, age", rauta "metal", rimba "frequent, numerous", rimbe "crowd, host", ringe "cold", runda "rough piece of wood", síre "river", yaiwe "mocking, scorn". Other words are identical to Telerin, the most closely related Amanya language, such as branda "lofty, noble, fine", Bana "Vána" (a Valië) and belda "strong".

EXTERNAL HISTORY

Old Sindarin explicitly identified as such is only known from the Etymologies, though a few early Sindarin forms mentioned in Tolkien's essay Quendi and Eldar have been included in the wordlist below. There are no OS texts. Tolkien would mention an OS word as the intermediate stage between Primitive Quendian and Sindarin simply to clarify the derivation of the Sindarin word. In the Etymologies the language is of course called Old Noldorin (abbreviation ON), since Tolkien still had not made the revision that turned the Welsh-sounding language in his mythos into the tongue of the Sindar. In his former conception, "Old Noldorin" was evidently the language that the Noldor spoke in Valinor, the language that turned into classical Noldorin in Middle-earth. Notice that in the Etymologies, "Noldorin" (> Sindarin) is sometimes called "EN" = Exilic Noldorin, as if implying that ON is "non-exilic Noldorin". Unlike other archaic forms, the "ON" words are not usually asterisked, as if they were attested in writing: This agrees with LR:173, where it is stated that the Noldor started to write their tongue in the days of Fëanor's pride. Tolkien later revised all of this. The Noldor came to be speaking Quenya, Noldorin became Sindarin, and we must assume that "Old Noldorin" likewise became Old Sindarin, though Tolkien never uses this term. Would Old Sindarin be attested in writing, or should we asterisk the entire corpus? Perhaps Old Sindarin would be written with Daeron's Runes? However, in WJ:370 Tolkien expresses uncertainty about whether a certain inflection had formerly occurred in Sindarin, as if the older stages of this language were not directly attested. Moreover, forms that seem to belong to about the same stage in the linguistic evolution as the "Old Noldorin" forms from the Etymologies are asterisked in later sources (see for instance ekla-mbar, ekla-rista in the wordlist below).

THE STRUCTURE OF OLD SINDARIN
Phonological development

We will not here attempt to give a full listing of the phonological changes that affected Old Sindarin; see the word-by-word discussions in the wordlist below. Some main points may be summarized all the same. As stated above, Old Sindarin preserves the general sound of primitive Elvish much better than Sindarin does. In particular, the final vowels that were later lost are still in place; the long final vowels so characteristic of Primitive Elvish had merely become short. For instance, primitive alkwâ "swan" yields OS alpha, bélekâ "mighty" becomes beleka. (In one case, though, a long final vowel seems to persist in Old Sindarin, primitive magnâ "skilled" yielding the unchanged form magnâ; this is probably an error, by Tolkien or the transcriber, for *magna.) When not final, the quality of three of the primitive long vowels was altered in Old Sindarin: â became ó (this change is explicitly mentioned in LR:392 s.v. THÔN), ê became í and ô became ú. For examples, see ndóko, khíril, rúma in the wordlist below. There are some words where such changes fail to take place (see gása, tára, róna), but these may be mere mistakes - by Tolkien or the transcriber. Primitive î, û were unchanged (normally spelt í, ú in Old Sindarin).

In very many cases, earlier t, p, k become th, ph, kh following another consonant, though this change fails to take place in some words (Tolkien forgetting his own rules?) See thintha in the wordlist below (where further references to other words are listed). However, there is no evidence that the medial clusters st, sp, sk ever became sth, sph, skh, though this change did occur initially: primitive stankâ "cleft", spangâ "beard" and skalnâ "veiled" yielding OS sthanka, sphanga, skhalla. (This change may not have occurred at the earliest Old Sindarin stage, since stabne "room" is listed as OS as well as the later form sthamne.) Before ph, s disappeared during the Old Sindarin stage, as indicated by the spellings (s)pharasse "hunting" and (s)pharóbe "hunt" (LR:387 s.v. SPAR). It may be that sth-, skh- likewise became th, kh in late OS (in later Sindarin we find th-, h-, which is also how original primitive th- and kh- appear in Sindarin, suggesting that the distinction between, say, primitive sk- and primitive kh- had disappeared altogether in late Old Sindarin). - Some words suggest that t, p, k became th, ph, kh also following another t, p, or k, producing aspirates (?) tth, pph, kkh. See rattha in the wordlist below.

Before a nasal, the unvoiced plosives t, p, k became voiced d, b, g by assimilation to the voiced nasal. Cf. for instance yadme "bridge" from primitive jatmâ, or tulugme "support, prop" from tulukmê. Other assimilations that took place at the Old Sindarin stage include bn > mn, medial sm > mm (unchanged initially), nm > mm, dn > nn, sr > rrh (= long unvoiced R?), ln > ll, ht > tt, hs > ss (for examples, see ammale, Boromíro/Borommíro, etlenna, gêrrha, khalla, matthô-be, watte, wasse in the wordlist below).

In polysyllabic words, final consonants were often lost, but they were preserved in the plural form of nouns (since they were not there final because of the plural ending -i). Hence we have pairs like nele pl. neleki, oro pl. oroti, pele pl. pelehi, skhapa pl. skhapati, thele pl. thelehi (see the next paragraph for the meaning of these nouns).

In Old Sindarin we seem the first hints of the lenitions so prominent in later Grey-elven. There is still no trace of the voicing of all post-vocalic unvoiced stops (OS beleka yielding Sindarin beleg) or the post-vocalic voiced plosives becoming spirants (OS ngolodo "Noldo" yielding Sindarin golodh), but the lenition of s to h following a vowel occurred during the Old Sindarin stage of the linguistic evolution. There are several examples in the Etymologies: barasa "hot, burning" later became baraha, khelesa "glass" became kheleha, pelesi the pl. of pele "fenced field" became pelehi, and thelesi the pl. of thele "sister" became thelehi. (For kheleha and pelehi, listed in the Etymologies in the entries KHYEL(ES) and PEL(ES), the misreadings "khelelia" and "peleki" occur in the text as printed in LR.)
The Noun

The Old Sindarin plural ending was -i, directly descended from Primitive Quendian *-î and cognate with Quenya -i: boron "faithful vassal" pl. boroni, toron "brother" pl. toroni. Nouns ending in a vowel normally drop this vowel before the plural ending -i is added: poto "animal's foot", pl. poti. However, in several cases nouns in -o have plurals in -ui instead: malo "pollen", pl. malui; orko "goblin, Orc", pl. orkui; pano "plank", pl. panui; ranko "arm" pl. rankui. This phenomenon occurs when the original, primitive form did not end in -o (or actually -ô), but in -u. Malo, orko, ranko derive, according to the Etymologies, from primitive smalu, órku, ranku, and while Primitive Quendian final short -u became -o already at the Common Eldarin stage, Old Sindarin preserves the original quality of the vowel before the plural ending (something Common Eldarin also must have done). However, pano "plank" is said to come from primitive panô, and yet its Old Sindarin plural is panui. This is because Old Sindarin shifted primitive oi to ui; compare OS muina with Quenya moina (primitive *moinâ, my reconstruction; stem MOY). The primitive pl. of panô must have been *panôi, later *panoi, becoming panui in OS. This, however, does not follow the pattern established by the noun poto "animal's foot" mentioned above (pl. poti instead of **potui). For more than one reason, it is tempting to dismiss the plural form panui as some kind of mistake and read simply pano pl. *pani.

As mentioned above, the plural forms of nouns in some cases preserve original final consonants lost in the singular: nele "tooth", pl. neleki (stem NÉL-EK), oro pl. oroti "mountain" (stem ÓROT), skhapa "shore" pl. skhapati (primitive skhyapat-), pele "fenced field", pl. pelesi (later pelehi) (stem PEL(ES)), thele "sister", pl. thelesi (later thelehi) (stem THELES), ró "lion" pl. rówi (primitive râu, stem RAW). Some of the lost final sounds were restored in Sindarin, evidently by analogy with the plurals. For instance, oro pl. oroti corresponds to Sindarin orod pl. ered (post-vocalic T being lenited to D in Sindarin).

In one case the vowel of the sg. and pl. form of a noun differs: mó "hand", pl. mai. This is because Old Sindarin shifted original long â to ó (like ndâkô "warrior" > ndóko), while the diphthong ai was unchanged (like gaia "dread", primitive gais-). So while primitive mâ3 "hand" became mó, the primitive pl. *ma3i, later *mai, remained mai.

Except for the plural inflection there is little direct evidence of inflections in Old Sindarin. There is the word thoronen, said to be the "gen.sg." of the word for "eagle"; the nominative is not given, but it may have been *thoron (as in Sindarin). This genitive ending -en also occurs in the Quenya (or "Qenya") of the Etymologies, but Tolkien later changed it to -o, descended from primitive -ho (WJ:368, cf. 3O in the Etymologies, LR:360). This probably throws considerable doubt upon the ending -en in mature Old Sindarin (if I may use such a term). In WJ:370, Tolkien argues that Sindarin had probably developed inflexional -ô in "the primitive period" (later presumably *-o, after the shortening of the final vowels). He notes that "the placing of the genitive noun second in normal Sindarin [like Aran Moria, "King of Moria"] is also probably derived from inflectional forms" in -ô, *-o. It seems that according to Tolkien's mature conception of the language, we should probably read *thorono for thoronen.

Indirect evidence from later Sindarin suggests that Old Sindarin may still have had a living locative in *-sse (a case ending well known from Quenya, High-elven being much more conservative than Grey-elven). Tolkien-linguists agree that the word ennas "there" in the King's Letter (SD:128-129) must be derived from earlier *entasse, sc. a word enta "that yonder" (known from Quenya, LR:356 s.v. EN) with a locative ending to express "in that yonder [place]" = "there". If Old Sindarin had a locative, perhaps the allative and the ablative were also still living cases?

The Verb

The infinitive: The infinitival ending -ie, known from Quenya (UT:317), also appears in Old Sindarin: bronie "last, endure, survive", etledie "go abroad, go into exile", ndakie "to slay", orie and ortie "[to] rise", tre-batie "traverse", trenarie *"to recount, tell to end", warie "betray, cheat". (In the Etymologies, trenarie is explicitly called an "inf." form in the entry NAR2, LR:374.) Another group of verbs show the infinitival ending -be: buióbe "to serve, follow", matthô-be "to handle", naróbe *"to tell a story", ortóbe "raise", phalsóbe "to foam", pharóbe (older spharóbe) "[to] hunt", phuióbe "feel disgust at, abhor", puióbe "[to] spit", rostóbe "to hollow out, excavate", wattóbe "to soil, stain". (Naróbe is actually glossed "he tells a story", but it can hardly be a 3rd person present tense form.) In light of these examples, we must conclude that parthóbi "arrange, compose" is probably a misreading for *parthóbe.

The past tense: The corpus contains only three examples of the past tense: lende "fared" (from the stem LED, other forms not attested but cf. the infinitive etledie "go abroad, go into exile", that is essentially the same verb with the prefix et- "out"), narne *"told a story" (infinitive naróbe), and ndanke, past tense of ndakie "to slay". The past tenses lende and ndanke are formed from their stems (LED, NDAK) by means of nasal infixion, a device also known from Quenya (indeed lende is also a Quenya word, with the same meaning). Narne shows the past tense ending -ne, also occurring in Quenya.

The present tense: One possible example of the present tense is the word persôs "it affects, concerns". The ending -s seems to be a pronominal suffix "it". Without it, this verb may appear as *persa "affects, concerns" (since the primitive verbal ending -â, like any other -â, would become -a instead of ô when final).

The aorist (?): There are a few verbs showing the ending -e, or -i- when an ending follows; this would appear to be the same form that has been identified as the aorist in Quenya, with the same endings: yurine "I run" (where -ne = "I") and trenare "he recounts". "He runs" and "I recount" would presumably be *yure, *trenarine.
Pronouns

Only two pronominal endings are attested: -s "it" in persôs "it affects, concerns" (which ending is also known from Quenya), and -ne "I" in yurine "I run". The latter ending may be a cognate of Quenya -nyë (since older *ny, *nj became n in Common Lindarin, the ancestor of both Telerin and Old Sindarin). Alternatively, this ending may represent primitive -ni, sc. the first person sg. stem NI "I" used as an ending (LR:378). The 3rd person singular may evidently be expressed by means of the verbal stem alone, as in trenare "he recounts, tells to end", where no pronominal element seems to be present.

OLD SINDARIN WORDLIST with etymological notes

Most of the words come from the Etymologies; a few words from other sources (UT, WJ, PM) have been included. The latter group of words is not explicitly identified as Old Sindarin, but it seems best to place them in this list. Usually, they are also included in the Primitive Elvish wordlist, with a note that they may be Old Sindarin.

Notice that the combinations ph, th, kh represent aspirated p, t, k (LR:322), not the spirant sounds heard in English philosophy and think and German ach. (It is possible, however, that the spirant sounds are intended in certain combinations; for instance, the word rattha may be meant to contain an aspirate tth, sc. t followed by th as in think). Both circumflexes (in the sources macrons) and accents are used. The circumflexes must mark long vowels; the accents seem in some cases to mark the stress (e.g. spíndele, ngalámbe, trenárna), but normally an accent indicates that a vowel is long (as in ró "lion"; of course, there could be no doubt about which syllable is accented in a monosyllabic word). Tolkien's spelling is not wholly consistent; for instance, we have both ô and ó for long o. There is no reason to assume that any subtle distinctions in vowel-length are intended here (as in later Sindarin, where Tolkien uses circumflexes to indicate super-long vowels). Cf. for instance such verbs as buióbe and matthô-be; surely these could just as well have been spelt buiô-be and matthóbe.

Primitive words "reconstructed" by Tolkien himself are not here asterisked, though he usually did asterisk them. Notice that in the primitive words, circumflexes (for macrons in the sources) always indicate long vowels, while accents indicate stress. In primitive words, we generally prefer the spelling j for the sound of y as in English you (Tolkien's spelling is inconsistent, and sometimes it was even altered by the transcriber).
abóro "Avar", "Elf who never left Middle-earth or began the march [from Cuiviénen]". In the Etymologies, this word was derived from a stem AB-, ABAR- "refuse, deny" (LR:347), and the primitive form was given as ábârô (or ábâro), including the personal (masculine/agental) ending -ô: An ábârô is thus a "refuser" (that is, one who refused the invitation of the Valar to come to Valinor). When not final, primitive long â becomes ó in Old Sindarin; for another example cf. ndóko (q.v.) from ndâkô. (In monosyllabic words, â may become ó even when final; cf. mó "hand" from mâ.) However, final primitive long -â, -ê, -î, -ô, -û merely become short; the quality of the vowels is unchanged: Hence the final -ô of ábârô becomes -o in abóro. - The reconstruction ábârô and the stem AB-, ABAR- are superseded by information provided by Tolkien in a source that is about a quarter of a century younger than the Etymologies, namely the essay Quendi and Eldar from about 1960 (WJ:360-423). In WJ:370, the stem from which the Quenya and Sindarin words for Avari are derived is said to be ABA. This stem "probably derived from a primitive negative element, or exclamation, such as *BA 'no'! (...) it expressed refusal to do what others might wish or urge". The oldest form is now said (in WJ:371) to be abaro, not as in Etym ábârô with long â and ô. While Tolkien originally held the r to be a part of the stem ABAR given in Etym (LR:347), abaro must be analyzed as the new stem ABA with the longer agental ending -ro. The new reconstruction abaro is said to have yielded Common Eldarin abar (WJ:371), which would also have been the Old Sindarin form. The plural form *abari would have yielded the Sindarin plural Evair, which is attested in WJ:380 (said to be a word known to the loremasters only). The earlier reconstruction ábârô, yielding Old Sindarin abáro, instead produced Sindarin ("Noldorin") Afor, pl. Efuir, later Efyr (LR:347 - f in these words is just a peculiar way of spelling v; read Avor pl. Evuir, Evyr). It would seem that all of these forms were obsoleted by Tolkien's more mature concepts as set out in Quendi and Eldar.
alpha "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, concerning which see katwe), 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. In the entire Lindarin branch of the Elvish language family, primitive kw very early became p (WJ:375 cf. WJ:407 note 5), so alk-wâ, alkwâ first became alpa (which form persisted in the Telerin of Aman), but in the evolution of Old Sindarin, a subsequent change turned p, t, k following a liquid l or r into ph, th, kh, respectively. Hence the form alpha (in turn yielding Sindarin alph pl. eilph - see UT:265, footnote). Compare salpha, and for other examples of p, t, k > ph, th, kh following l or r see bértha-, parkha, parthóbi, pelthaksa, sulkha. (For words where this change fails to take place, see awarta, ngurtu, orko/orkui, ortie, ortóbe. Since the later Sindarin forms are awartha, gurth, orch/yrch and ortho, it would seem that the change eventually occurred after all; the abnormal forms may be explained as early Old Sindarin, later becoming *awartha, *ngurthu, *orkho/orkhui, *orthie, *orthóbe.)
ammale, ammalinde "yellow bird, 'yellow hammer' ", derived from a stem SMAL "yellow" (LR:386). The primitive forms are given as asmalê, asmalindê; these examples indicate that in Old Sindarin, the cluster sm was assimilated to mm between vowels. Both words show a prefixed a that is probably just the stem-vowel reduplicated, and the suffix -ê is probably a feminine ending. Asmalindê shows a longer feminine ending -indê, not otherwise mentioned as a primitive form, but cf. Quenya -indë as in Serindë (Þerindë) "Needlewoman" (PM:333).
anda "long", only attested in the compound andatektha, q.v. Derived from the undefined stem ÁNAD, ANDA (LR:348); the primitive form is there given as andâ with the common adjectival ending -â, though no Old Sindarin form is listed there (but later Sindarin and, ann is mentioned).
andatektha "long-mark" (= Q andatehta, a symbol used in writing to indicate that a vowel is long; listed in LR:391 s.v. TEK). Compound of anda and tektha (q.v.), not otherwise attested in Old Sindarin. The "Noldorin" form descended from andatektha is given in Etym as andeith, corresponding to Sindarin andaith in LotR Appendix E: "In this mode [the Tengwar mode of Beleriand] length of vowel was usually indicated by the 'acute accent', called in that case andaith 'long mark'."
ango "grandchild, descendant". Derived from a stem ÑGYÔ, ÑGYON (LR:377) that is similarly glossed; this is probably to be understood as a strengthening of the stem YÔ, YON "son" (LR:400), since Tolkien made a cross-reference to that stem. The Quenya and Telerin cognates of ango are indyo and endo, respectively; these forms together point to a primitive form ñgjô, identical to the form ÑGYÔ listed by Tolkien (y = j). The details can be debated but we are probably to assume a development more or less like ñgjô > *ñdjô > *ñdô > *añdo > ango (compare the Quenya development *ñdjo > *iñdjo > indyo, and Telerin ñgjô > *ñdjô > *ñdô > *eñdo > endo). The reason why an a (in Quenya an i, in Telerin an e) developed before ñ is evidently that this initial sound had become syllablic; PM:360 gives evidence that before a syllabic ñ, a vowel a would develop in Sindarin (and an i in Quenya), as in the name Angolodh from ñgolodô when the initial consonant was syllabic (Quenya Ingoldo; the Sindarin form Angolodh was not actually used, but its Old Sindarin form would nonetheless have been *Angolodo). - The word ango "grandchild, descendant" did not yield any Sindarin word, probably because it became identical to ang "iron" (Old Sindarin *anga) after the loss of the final vowels.
anu "a male (of Men or Elves), male animal. Derived from a stem 3AN (LR:360), simply defined as "male". The primitive form would be *3anû, a fairly well attested masculine ending being added to the stem (cf. for instance kherû "master" [Letters.282], whence Quenya heru "lord"). For the loss of initial 3 (also reconstructed as h), cf. elwa from *3elwâ.
Araume < Oroume "Oromë", name of a Vala. In the Etymologies, this word was derived from ORÓM (LR:379), a reduplicated form of ROM "loud noise, horn-blast, etc." (LR:384; cf. romba, róma). The primitive form of the name was given as Orômê (LR.379). The ending -ê is surprising in the name of a male Vala, since this ending is normally feminine. In this case, -ê is perhaps best taken as an abstract ending; cf. WJ:400, where it is said that the Eldar interpreted the name as "horn-blowing". From Orômê Tolkien first derived Old Sindarin Oroume; we would rather expect **Orúme, since primitive ô yielded OS ú (see brúna for an example). Perhaps "ou" is just another way of spelling ú; see doume under dogme. This strange "ou" then turns into au in Araume (though ú should regularly be unchanged in Sindarin!), and at the same time, the initial o mysteriously becomes a (some kind of dissimilation?) The development seems to be pretty ad hoc; Tolkien is heading straight for the Classical Sindarin form Araw, perhaps already decided upon before he tried to sort out the phonological intricacies. The precise meaning of the name remains vague; it simply has something to do with horns or horn-blasts. However, this explanation of the name Oromë/Araw is obsoleted by information provided in the essay Quendi and Eldar, written about a quarter of a century later. Tolkien rejected the interpretation "horn-blowing" or "horn-blower" as a mere folk etymology on the part of the Elves; there was not really any connection with the stem ROM after all. Tolkien decided that Oromë's name was actually adapted from the form his name had in the language of the Valar. In Valarin, the name simply denoted Oromë and had no etymology beyond that (WJ:400-401). The original Valarin form was Arômêz (with a special open, A-like variety of ô); the oldest Elvish adaptation was Arâmê, later becoming Arômæ (with A-like ô), then Araum(a), then Araumh (mh = nasalized v) and Arauv, finally becoming Araw in classical Sindarin. The form Arômæ may be seen as Old Sindarin. In the phonological scheme used by Tolkien in the Etymologies, primitive Arâmê would have yielded Old Sindarin *Aróme, which is at least close (but Tolkien's ideas about the development of Sindarin were clearly modified over the years, so it is no surprise if there is not full agreement between the Etymologies and the essay Quendi and Eldar, written much later).
awarta "forsake, abandon". Derived from a stem WAR- "give way, yield, not endure, let down, betray" (LR.397); the primitive form would be awartâ-. The verbal suffix -tâ is very common; the prefixing of the stem-vowel is probably simply intensive. Contrast the simpler verb warie "betray", derived from the same stem, but with no prefixed vowel (and neither any ending). - The form awarta must be early Old Sindarin, for later, t following r became th (presumably producing awartha-, which form is attested in later Sindarin). Compare bértha- from *bertâ-.
Bala "Power, God, Vala". In the Etymologies derived from an undefined stem BAL (LR:350); a later entry BEL "strong" is however compared to BAL, so the basic meaning may have something to do with "power". The primitive form of bala is given as *bálâ, a formation that seems to parallel Bánâ from BAN; see Bana below. In WJ:403 Tolkien provides some information about Quenya Vala, the close cognate of Bala. Vala (and hence Bala) is properly a verb "has power", and the plural form Valar (Old Sindarin *Bali?) can be interpreted "they have power". Subsequently these verbs were also used as nouns: "a Power, the Powers". - The word Bala as such does not seem to have any independent descendant in Sindarin; the Sindarin word for Vala, Balan, descends from the longer forms Balane, Balano.
Balandor "Valinor", *"Vala-land" (LR:350 s.v. BAL). Compound of Bala "Vala" and ndor "land", q.v.
Balane fem. form of Bala, "Valië" (LR:350 s.v. BAL). Primitive *Balanê, sc. the more basic form Bala (see above) with the feminine ending -nê added to produce a gender-specific form. See Balano.

Balano masc. form of Bala, "(male) Vala" (LR:350 s.v. BAL). Primitive *Balanô, sc. the more basic form Bala (see above) with the masculine ending -nô added to produce a gender-specific form. In later Sindarin, after the loss of the final vowels, masc. Balano and fem. Balane converged to produce a gender-neutral word for Vala, Balan.
Balthil a name of the White Tree of Valinor (LR:350 s.v. BAL). The umlauted Sindarin form descended from it, Belthil, is translated "Divine Radiance" in the Silmarillion index. The word seems to be simply a compound of two stems, BAL and THIL (LR:350, 392). BAL yields words for "Vala" and hence "divine" (see Bala above), while THIL is stated (in LR:392) to be a variant of the stem SIL, meaning "shine silver" (LR:385, compare Letters:425, third footnote). The formation Balthil seems to be quite similar to Narsil, said (in Letters:425) to be "composed of 2 basic stems without variation or adjuncts".
Bana "Vana", name of a Valië (the Quenya form is spelt Vána in the published Silmarillion). Primitive form given as Bánâ. The ending -â is usually adjectival and never explicitly feminine; it may simply be the stem-vowel suffixed and lengthened, but bánâ may originally have been an adjective that was later used only as an epithet of this goddess (compare the Quenya name of another goddess, Varda, originally an adjective "lofty, sublime"; see barada). Bánâ is derived from a stem BAN (LR:351) that is not defined as such, but it seems to have to do with beauty; it is the source of Quenya vanya "beautiful". If bánâ is properly an adjective, it may have a similar meaning. In Old Sindarin, the compound Bana-wende (*"Vána-maiden" or even *"Beautiful-maiden") was also used; see wende. - In WJ:383, reproducing a late source (ca. 1960), it is said that the Quenya name of this Valië, Vána, is derived from a stem WAN instead; this stem has to do with pale colour, also associated with beauty (much like English fair). Here, the primitive form is probably meant to be *Wânâ, that would yield *Wóna in Old Sindarin.
barada "lofty, sublime" (or "steep" if = Sindarin baradh, the word it yielded). Derived from a stem BARÁD (LR:351), itself undefined but suggested to be an expanded form of BAR, tentatively said to mean "raise" (LR:351); BARÁD could mean something like *raised, uplifted" and hence "lofty, sublime". The primitive form is given as barádâ, showing the normal adjectival ending -â (in Quenya, the name of the goddess Varda, The Sublime, is derived from this primitive adjective).
baraha, earlier barasa, "hot, burning". Derived from an undefined stem BARÁS (LR:351) said to be found only in "Noldorin" (read: Sindarin). The primitive form is given as barasâ (accented on the final syllable), showing the common adjectival ending -â. The oldest OS form was barasa, later becoming baraha with the lenition of intervocalic s to h. (For other examples of this lenition, cf. pelehi, thelehi under pele, thele; see also kheleha.)
barane "brown, swart, dark brown". The primitive form would be *baráni, sc. the stem BARÁN (LR:351) with the ending -i, very common in primitive colour-adjectives (compare in particular karani "red" from KARÁN, LR.362). The Quenya compound form varni- preserves the original quality of the final vowel (uncompounded varnë, with the same change of final short -i to -e as in Old Sindarin; this change seems to have occurred at the Common Eldarin stage, and is therefore common to Old Sindarin, Quenya and Telerin).
barasa "hot, burning"; see baraha (the later form).
Barathi, also Barathil "Varda, Elbereth" (LR:351 s.v. BARATH). The form Barathil seems to be secondary, with a longer feminine ending -il also seen in khíril "lady" (and in Bradil, another name of Varda, later struck out). The second element in the normal Sindarin name of the goddess Varda, Elbereth, descends from Barathi (with an element el- "star" prefixed); see elen-barathi. According to RGEO:74, Sindarin bereth (from barathi) means "spouse", used of the spouse of a king, hence coming to mean "queen": Varda is both the Queen of the Valar and the spouse of Manwë. In Letters:282, Elbereth is translated "Star-lady", providing yet another gloss for bereth and hence barathi: "lady". In the Etymologies, the stem BARATH from which Barathi is derived is not clearly defined, but merely suggested to be "related to BAR and BARÁD". Concerning the latter, see barada. The most basic stem here is clearly BAR, tentatively defined as "raise" and hence capable of yielding words for something high or lofty; BARÁD and BARATH are clearly simply extended forms of BAR, and BARÁD (yielding words for "lofty, sublime, steep") is said to have been "blended" with BARATH. Tolkien derived Barathi from primitive Barathî, showing the feminine ending -î. Since Barathî was defined as "spouse of Manwe, Queen of Stars", it seems implied that this was the first name the primitive Elves gave to Varda. Barathî should have yielded *Varsi in Quenya; however, Tolkien imagined that this primitive name had been blended with the adjective barádâ "lofty, sublime", which was the true origin of the Quenya name Varda. However, the ancient name persisted in Sindarin, to provide the final element of the later name Elbereth. - It seems somewhat doubtful that Tolkien intended barathî to mean simply "spouse" when he wrote the Etymologies, though this was how he defined Sindarin bereth in RGEO:74 many years later. Barathî was probably first meant to signify *"she that is sublime/lofty". Perhaps Tolkien later, after adopting the "spouse" interpretation, imagined that bereth included the Sindarin feminine ending -eth, primitive -ittâ (PM:345); if so the primitive form would rather be *berittâ or *barittâ (and the Old Sindarin form *berittha or *barittha).
bata "beaten track, pathway". Derived from a stem BAT (LR.351); the primitive form is given as bátâ (or báta, but this would have yielded Old Sindarin **bat instead). The ending -â is sometimes used to form nouns (more often adjectives), but in this case it may be simply the stem-vowel reduplicated and suffixed. It is also possible that -â has a local meaning here; compare rattha, yura.
batthô "trample" (indicated to be accented on the final syllable). Derived from a stem BAT "tread" (LR:351); the primitive form is given as battâ- (like batthô indicated to be accented on the final syllable). It is said that in battâ, the "medial consonant [is] lengthened in frequentative formation": The T of the stem BAT "tread" is doubled to symbolize the repeated action: *"tread repeatedly" = "trample". In Old Sindarin, earlier tt becomes tth (compare rattha, q.v., from rattâ; it is probable that primitive kk, pp similarly became kkh, pph: Though we have no explicit examples, later Sindarin words suggest such a development in OS). Normally, final long -â in polysyllabic words becomes -a in Old Sindarin words (see abóro). Therefore, we might have expected battâ- to yield *battha- instead of batthô. Non-final long â may regularly becomes ó (here spelt ô), and so does final â in monosyllabic words (see mó). What makes the difference is evidently that in battâ-, the final -â was stressed (just like the final vowel of batthô still is). The stress would have moved to the first syllable in batho, the later Sindarin form. When Tolkien wrote this, he must already have decided that the original final vowels were lost in Sindarin. Here he seems to be toying with a concept that final stressed vowels were not lost, that batthô (stressed on -ô) first became bathó in Sindarin, the final vowel surviving because it was accented, the accent moving to the first syllable only later, as the classical accent pattern evolved. A Sindarin verb in an earlier entry in the Etymologies points in the same direction: berio "to protect", derived from a primitive form barjâ (baryâ); once again, Tolkien carefully indicated that the final -â was accented (LR:351 s.v. BAR). He probably assumed that barjâ became Old Sindarin *baryô (baryó), still stressed on the final vowel, which in turn yielded Sindarin *berió, later berio with the stress on another syllable. However, the evidence is that Tolkien shortly afterwards decided upon a quite different way of deriving Sindarin verbs in -o from primitive forms in -â, rejecting the "stressed final vowel"-idea he experimented with in the early entries of the Etymologies. (In later entries, even stressed final -â becomes -a rather than -ô in Old Sindarin; see khalla.) Instead of assuming that -â was originally stressed, he introduced an infinitive ending -be; in the Etymologies, it first turns up in the entry BEW (where buio "to serve, follow" is derived from Old Sindarin buióbe, q.v. concerning the phonological development). Perhaps Tolkien would have derived Sindarin batho from Old Sindarin *batthóbe, not batthô, if he had found the time to revise all his notes to bring them into accord with his new insights.
belda "strong", derived from a stem BEL of similar meaning (LR:352); the primitive form would be *beldâ with medial fortification l > ld and adjectival -â. (Compare kuldâ "golden-red" from KUL, LR:365.)
beleka "mighty, huge, great". Derived from the same stem BEL "strong" (LR:352) as belda above. Tolkien first mentions a form bélek, which is probably to be understood as an extended variant of BEL, with the stem-vowel reduplicated and suffixed (so-called ómataina, vocalic extension) and a consonant -k added. Then follows the primitive form bélekâ, which is clearly this extended stem with the common adjectival ending -â.
belka "excessive" - attested as a Telerin word, but "possibly from ON", Old Noldorin (LR:352). Of course, no Old Sindarin word could get across the ocean to enter the speech of the Teleri of Aman. Apart from the difficulties caused by Tolkien's revision of the history of this language, belka would be simply a variant form of beleka, derived from the same stem BEL "strong". In the case of belka, the primitive form would be *belkâ with an adjectival ending -kâ added directly to the stem (rather than the shorter ending -â added to the extended stem bélek- as above). Concerning the adjectival ending -kâ, cf. for instance poikâ "clean, pure" from POY, LR:382.
belle "(physical) strength". Derived from same stem BEL (LR:352) as belka and belda above; the primitive form of belle would most likely be *bellê, since the ending -ê is often abstract. The doubling of the l could be a "medial fortification", though l > ld is more common. (Notice that Ilkorin bel is said to derive from belê with no doubling of the l.) Conceivably bellê could also be the stem BEL with the universal and abstract ending -lê (VT39:16).
berina "bold, brave". Derived from a stem BER "valiant" (LR:352); this is an adjectival formation paralleling malina (q.v.) from smalinâ: primitive *berinâ.
bértha- "to be bold" (the accent probably indicates that it is the first syllable that is accented, not that the e is long). Derived from a stem BER "valiant". The primitive form would be *bertâ-, the common verbal suffix -tâ being used in a rather uncommon way: to form a stative verb, describing something one is rather than something one does. Normally, -tâ forms transitive or even causative verbs (e.g. tultâ- "make come" from TUL- "come", LR:395). - Following r and l, primitive t becomes th in Old Sindarin; see alpha. Since ngurtu (not *ngurthu) is said to be an Old Sindarin form, it may be that this change occurred during the Old Sindarin stage, so that an earlier form of bértha- was *bérta-
besse "wife" (or "woman"?) Derived from a stem BES "wed" (LR:352); the primitive form is given as bessê. The word could include the (feminine?) ending -sê seen in NDIS-SÊ (LR:375), whence Quenya nissë "woman". Alternatively, the doubling of the S of BES is some kind of medial fortification, and the ending is simply the feminine suffix -ê. - We cannot be absolutely sure whether besse means "wife" or simply "woman"; primitive bessê is defined as "wife" (reflecting the root meaning "wed"), while Sindarin bess had a generalized meaning "woman". The intermediate Old Sindarin form besse is mentioned, but not defined. It probably still meant "wife"; the wording in the entry BES seems to indicate that it was Sindarin bess and not the earlier form besse that was used to replace earlier words for "woman". See also LR:378.
bioro, also biuro, "follower, vassal". Derived from a stem BEW "follow, serve"; the primitive form is given as beurô with the masculine (often, as here, agental) ending -rô (WJ:371). It seems that primitive eu first became iu and then io; biuro is probably to be understood as an older form, later (but still during the Old Sindarin stage) becoming io. Compare sniuma "snare" from *sneumâ; the later "Noldorin" form hniof (Sindarin *nýw) suggests that iu later became io (*snioma). - In the Etymologies, the name Bëor is derived from bioro. The published Silmarillion (ch. 17) agrees that Bëor means "Vassal", but in this book, the name is said to be derived from the language of Bëor's own people, not from Elvish: Tolkien's concept seems to have undergone some revisions. (In Sindarin, older bioro, biuro would yield býr rather than beor; perhaps Tolkien, wishing to keep the long-established name Bëor even after he had revised Sindarin phonology, transferred it from Grey-elven to Mannish for this very reason.)

Boromíro masc. name, *"Steadfast Jewel", Boromir (LR:353 s.v. BOR; the variant Borommíro is found in LR:373, under MIR). The entry BOR indicates that Boromíro represents Boronmíro; the form Borommíro in the entry MIR would reflect nm in the form of a double mm, n being assimilated. The first element, unassimilated boron, is apparently more or less identical to the independent word boron, see below. The element míro is evidently a variant of míre "jewel", q.v., the masculine ending -o substituting -e when the word is used as the final part of a masc. name. - In a footnote in LotR Appendix F, it is said that the name Boromir is a "mixed" form, the context indicating that this means that it contains both Quenya and Sindarin elements; this is probably Tolkien's way of explaining why the m of Boromir is not lenited to v. In the Etymologies, where the forms Boromíro, Borommíro, Boronmíro are quoted as the historical origin of this name, there is no trace of this later notion that the name is somehow "mixed". In the Etymologies, Boromir is described as "an old N[oldorin] [later: Sindarin] name of ancient origin". So while the name existed before LotR was written, Tolkien's ideas about its precise history would seem to have undergone revisions.
boron "steadfast, trusty man, faithful vassal", pl. boroni (LR:353 s.v. BOR). Derived from a stem BOR- "endure" (LR:353); a primitive form is given as bóron- (the hyphen suggesting that some final vowel, perhaps masculine -o, was originally present). The form bóron- shows ómataina (reduplication of the stem-vowel) and the ending -n; indeed the stem BORÓN, also listed in LR:353, is said to be an "extension" of BOR and a "verbal form" of borón- (notice the difference in stress). The plural ending seen in boroni descends directly from primitive -î. - For a possible parallel to boron pl. boroni from BOR, see toron pl. toroni from TOR (cf. also *thoron).
Boronmíro > Boromíro (masc. name) (LR:353 s.v. BOR). See Boromíro.
[Bradil] (struck out) "Varda". This rejected name of the Star-goddess was derived from a stem BARÁD (LR:351), concerning which see barada. The form brad- has lost the first, unaccented vowel of BARÁD (cf. branda below); the ending -il is a feminine suffix also found in Barathil and khíril (q.v.)
branda "lofty, noble, fine". The same word occurred in the Telerin of Aman. Derived from a stem BARÁD, not itself defined; see barada for some comments on this stem. The primitive form of branda is given as b'randâ, showing loss of the first A, nasal infixion (or medial fortification d > nd) and the common adjectival ending -â. Perhaps we are to understand that the very oldest form was *barandâ with the first vowel intact; later it was lost (because it was unaccented?) Compare a primitive form like b'rônâ from the stem BORÓN; see brûna.
brasse "white heat". Derived from an undefined stem BARÁS (LR:351); it yields words for "hot, burning, fiery". A primitive form b'rás-sê is given (simply defined as "heat"); an even older form may have been *barás-sê with the first A intact (the loss of unaccented vowels in such primitive forms is not uncommon; compare *b'ron- from BORÓN; see bronie below). The precise meaning of the ending -sê is uncertain (though the final vowel -ê could denote abstracts). In some words, -sê apparently denotes something that is made by the action denoted by the stem: khotsê "assembly" from KHOTH "gather" (LR:364), sjadsê (later sjatsê) "cleft, gash" from SYAD "shear through, cleave" (LR:389), wahsê "stain" from WA3 "[to] stain, soil" (LR:397). If the undefined stem BARÁS means something like "burn" or "heat up", brás-sê might fit into this pattern.
bronie "last, endure, survive". Derived from a stem BORÓN (LR:353), said to be an extended form of BOR "endure", supposedly originally a verbal form. As is usual in such extended forms, the stem-vowel has been reduplicated and suffixed (ómataina, vocalic extension), before a consonant has been added (here -n). Bronie shows the infinitive ending -ie also known from Quenya (see Tolkien's comments on the word en-yalië in UT:317; for other examples of Old Sindarin infinitives in -ie, see etledie, ndakie, orie, ortie, tre-batie, trenarie, warie). Removing this ending, we are left with bron-, that must descend from *b'ron-, a form of the stem BORÓN that has lost the first, unaccented vowel; see brûna.
Bronwega *"Enduring Man"; masc. name (LR:353 s.v. BORÓN, LR:398 s.v. WEG). The element bron- is the same as the stem of the verb bronie, see above. Concerning the ending -wega "-man", see separate entry.
brûna (simply another way of spelling brúna) "that has long endured, old" (of things only; implies that they are old, but not changed or worn out). Derived from the same stem BORÓN (LR:353) as bronie above. The primitive form is given as b'rônâ, with loss of the first O of BORÓN; a similar reduced stem must underlie the verb bronie (see above). B'rônâ shows lengthening of the stem-vowel and the common adjectival ending -â. As we see, long non-final ô yields Old Sindarin ú (compare rúma from *rômâ, or wanúre from *wonôrê).
buióbe "to serve, follow". Derived from a stem BEW of similar meaning (LR.352); the primitive form is given as beujâ- (beuyâ-) with the frequent verbal ending -jâ, that here adds nothing to the meaning of the stem itself. There seem to be some inconsistencies regarding the development of primitive eu in Old Sindarin. In beurô > bioro, biuro, this primitive diphthong becomes io or iu. It similarly becomes iu in *sneumâ > sniuma. However, eu becomes ui in *pheujâ > phuiobe, and so also in the word before us: beujâ > buióbe. It may be noted that Tolkien first derived núma, not sniuma, from *sneumâ (see núma). He changed it later, but he seems to have considered the possibility that eu became ú (û) instead of iu or io. If so, we may assume a development beujâ- > *bûjâ- > buió-. If we ignore the rejected example núma, we may simply assume that while eu became iu, euj (probably via -eui- > -iui-) yielded ui instead of becoming a "triphthong" **iui (compare puióbe, tuio, q.v.) - The ending -be occurring in the word buióbe is interesting. It is not really descended from any element in the primitive form beujâ-, but is a later addition. It seems to be an infinitive ending, and it has been suggested that it is somehow related to the Quenya preposition ve "as, like", Sindarin be (attested in the form ben, "according to the", in the King's Letter), if Tolkien imagined that primitive *be (bi?) was some ancient adverbial element. The adverbial ending -vë seen in Quenya, as in andavë "long" (as adverb, adjective anda) also points in this direction. - Primitive *beujâ-bê (or *beujâ-bi?) becomes buióbe because non-final â yields Old Sindarin ó (see abóro). In later Sindarin, buióbe became *buiauv > *buiau > buio. The infinitive ending -be is thus responsible for the fact that in later Sindarin (or at in least the "Noldorin" of the Etymologies), verbal stems in -a have infinitives in -o. For other examples of the infinitive ending -be, see matthô-be, naróbe, ortóbe, phalsóbe, parthóbi (read *parthóbe), spharóbe (pharóbe), phuióbe, puióbe, rostóbe, wattóbe.
daer "bridegroom" (LR:354 s.v. DER). One wonders whether this may be a misreading for *dair; compare ndair (also said to mean "bridegroom") from NDER. But an even more probable error is that the language is misidentified. The context in the entry DER is as follows: Tolkien, explaining why the initial D of the stem DER "man" irregularly becomes N in the Quenya word nér (also meaning "man"), states that this unexpected development is partly due to the "strengthened stem ndere bridegroom, ON daer". The "strengthened stem ndere" forms an independent entry in the Etymologies: NDER, expressly stated to be a strengthened form of der "man" (LR.375). But in this entry, the "ON" (Old "Noldorin"/Sindarin) word for "bridegroom" is said to be, not daer, but ndair. On the other hand, the entry NDER lists an "N" ("Noldorin"/Sindarin) word doer, that is listed as the descendant of ndair. While writing the Etymologies, Tolkien kept changing his mind about whether older ai became ae or oe in later Sindarin, so doer clearly corresponds to daer in the entry DER. It seems, then, that the word daer in this entry is actually Sindarin, not Old Sindarin. Perhaps Tolkien accidentally wrote "ON" in front of this word when he intended to write "EN", Exilic Noldorin, or perhaps the transcriber misread "EN" as "ON". Daer would be a most atypical Old Sindarin ("Old Noldorin") word; it would be the sole word in the language to have the diphthong ae. The true Old Sindarin form, from which Sindarin daer descends, is ndair (q.v.)
dalma "palm of hand". The dal- part derives from a stem DAL- "flat". It would be tempting to identify the element -ma with a common Eldarin nominal ending - compare parma "book" from the stem PAR "compose, put together" (LR:380), hence parma = *"composed thing". A dalma would then simply be a "flat thing, something flat". However, this was not what Tolkien had in mind: In LR:353 s.v. DAL we are informed that dalma is actually a compound, the second element being a form of mâ "hand" (cf. LR:371 s.v. MA3-). Thus, the literal meaning of dalma is *"flat-hand". The independent Old Sindarin form of this word is mó (q.v.) showing the normal change â > ó when â is not the final sound in polysyllabic words, but a final â in polysyllabic words is simply shortened to -a, so also in dalma. This Old Sindarin word might still have been vaguely felt to be a compound by speakers of the language, but the descendant Sindarin word dalf would definitely be perceived as a unitary word.
dî "woman", said to be rare and poetical: "bride, lady" (LR:378; this word is also mentioned in LR:354 s.v. DER). This word was earlier nî, essentially identical to the stem NÎ1 "woman". The shift n > d did not occur for phonological reasons; according to LR:378 it was due to the influence of the word dîr "man" (q.v.) - not that the words for man and woman were actually confused, but since the word for "man" was dîr, it was apparently felt to be appropriate and symmetrical that its feminine counterpart should also be a word in d-.
dîr "adult male, man" (elf, mortal, or of other speaking race). Derived from a stem DER of similar meaning (LR:354); this Old Sindarin word is also mentioned in LR:378 (under NÎ1, where Tolkien tells how it influenced nî "woman" to become dî). The primitive form from which dîr must be derived is mentioned in the beginning of the same entry (LR:377): dêr with a long vowel. This example demonstrates that long ê became î (í) in Old Sindarin; compare khíril.
Dirghel a masc. name (LR:354 s.v. DER), apparently meaning *"man of joy". This form is mentioned simply as an older form of Sindarin Diriel and is not said to be Old Sindarin as such; we include it here, though it is probably intended to be younger than the rest of the OS forms (the "pure" OS form may be *Dirgelle). The first element is clearly identical to dîr "man" (see above); the long vowel is apparently shortened before a consonant cluster. The later name Diriel is also mentioned in the entry GYEL (LR:359) in a context suggesting that the element -ghel (later -iel) is to be equated with Sindarin gell "joy" (primitive *gjellê, perhaps showing the abstract ending -lê; the OS form would be *gelle). In Dirghel, the lenition of g to gh following a liquid (here r) has already occurred; this is a later development that does not belong to the Old Sindarin stage. On the other hand, neither has it reached the Classical Sindarin stage, since gh no longer occurred in Sindarin (in this particular environment, gh apparently became i: Dirghel > Diriel). - Narrative texts outside the Etymologies employ the form Díriel with a long í, preserving the long vowel of dîr (dír-) "man".

dissa "young woman". In the Etymologies, this word is mentioned in the entry BES (LR:352), but the stem is obviously another. Dissa seems to be basically the same word as ndissa, q.v. for a discussion of the endings involved. A form with initial d- instead of nd- is rather difficult to explain. The stem of ndissa is NDIS, said to be a strengthened form of NIS "woman" (LR:378). The form dissa would require a stem *DIS, that is nowhere attested. Did Tolkien consider the idea that NDIS is a strengthening of *DIS instead of NIS? In LR:375 s.v. NDIS- it is said that this stem-form was made as a strengthening of NIS to parallel the relationship between NDER and DER, two stems meaning "bridegroom" and "man", respectively (LR:375, LR:354). Perhaps this parallel was carried one step further, so that the Elves extrapolated a new stem *DIS from NDIS to fully mimic the relationship between NDER and DER, and that a noun dissa was formed from it, co-existing with the more correct form ndissa? Or was ndissa, like dî (for nî) influenced by dîr "man" to produce a form with simple initial d-? In later Sindarin, both ndissa and dissa would become dess (but the behavior of this word in mutation position depends on which derivation is preferred).
dogme, dougme, doume "Night [as phenomenon], night-time, shades of night". Derived from a stem DO3/DÔ (LR:354), clearly indicating a basic stem DO3 with an alternative form DÔ that arose when the back spirant 3 was lost and the vowel was lengthened in compensation. The stem DO3/DÔ is not defined, but yields words for night and gloom. A primitive form do3mê is mentioned in LR:355; the ending -mê is abstract. It often forms a kind of verbal nouns; see ragme (cf. also tulugme), but here -mê > -me seems to denote something that is simply intangible: Night. In the form dogme, 3 has become g; this is our sole explicit example of this shift. However, later Sindarin forms suggest that 3 regularly became g before nasals; for instance, taen "height", derived from a stem TA3 (LR:389), must represent Old Sindarin *tagna (in turn derived from primitive *ta3nâ; a form ta3na is mentioned). - The alternative forms dougme, doume are strange. If the "ou" of doume represents long ú, this word could be derived from *dômê, sc. an alternative form of do3mê where 3 was lost instead of becoming g before a nasal, the vowel o being lengthened to ô in compensation. Primitive long ô yielded long ú in Old Sindarin; perhaps "ou" is just another way of spelling ú; cf. Oroume (see Araume). If doume is the same as *dúme, it could also be derived from dômi- "twilight", a form mentioned in LR:354 s.v. DOMO-: It is said that Quenya lómë "night, twilight" derives from both do3mê "night" and dômi "twilight", since these forms fell together in Quenya. Perhaps they were confused in Old Sindarin as well, dogme representing do3mê and doume representing dômi-, with dougme as an historically unjustified compromise between the two.
Eide "Rest, Repose", used as the name of a Valië (in Quenya called Estë), the wife of Lórien. An earlier Old Sindarin form was Ezde, a later form was Ide. In the Etymologies, Eide (Ezde, Ide) is listed under an entry-head EZDÊ (LR:357), but with a cross-reference to SED, which is the basic stem here: EZDÊ is rather a reconstruction of a primitive word. (Under SED [LR:385] a form Ezda is mentioned; this is probably simply a misreading for Ezde.) In WJ:403, Tolkien derives this name from a stem SED, stating that esdê became ezdê in Common Eldarin (s being voiced to z by contact with the voiced consonant d). Esdê would seem to be the most primitive form, with a rearranged form of the stem SED (vowel-consonant-consonant instead of normal consonant-vowel-consonant); the ending -ê can be both abstract and feminine. Originally esdê was evidently a common noun "repose"; in WJ:404, Pengolodh observes that while the Quenya and Telerin forms (Estë, Êde) had come to be used only as a name of the goddess, the Sindarin word îdh still had a general meaning "rest". This suggests that here, the ending -ê was originally abstract only, though the fact that there was a feminine ending of the same form would have made the word easily applicable as a fem. name. The Common Eldarin form Ezdê connects with the Old Sindarin form Ezde listed in the Etymologies. Before a consonant, z became i (compare mazga "soft" > maiga), producing Eide. Later, ei was monophthongized, yielding Ide (which in turn yields Idh as the Sindarin/"Noldorin" name of Estë; compare the noun îdh from WJ:404 mentioned above).
ekla-mbar "Eglamar", a name of Beleriand: "Home of the Forsaken", with reference to the Elves that were left there. This form, mentioned in WJ.365, is there asterisked as unattested. This form, like ekla-rista below, is not explicitly said to be Old Sindarin, but seems to belong to the same stage of the linguistic evolution as the Old Sindarin ("Old Noldorin") forms from the Etymologies, regardless of the revised history of the language (transplanted from the Blessed Realm to Middle-earth). The element ekla must be derived from heklâ (WJ:361), said to be an adjectival form (with the common adjectival ending -â) of hekla, a Primitive Quendian word defined as "any thing (or person) put aside from, or left out from, its normal company" (later used of Elves that did not go to Aman, but stayed in Beleriand). The stem is HEKE, said to be probably an adverbial element "aside, apart, separate" (WJ:361; this stem is not found in the Etymologies). The ending -la (in hekla) is simply a noun-former here. In the Etymologies, -la is found in the names of a number of implements: makla "sword" from MAK "sword, fight with sword" (LR:371), tekla "pen" from TEK "write" (LR:391, hence *"thing for writing"), and, with a nasal-infixed stem, tankla "pin, brooch" from TAK "fix, make fast" (LR:389). But in the word magla "stain" from the stem SMAG- "[?to] soil, stain" (LR:386) the ending simply acts as a noun-former, as in hekla. Heklâ, the adjectival form of hekla, becomes ekla in Old Sindarin because "P[rimitive] Q[uendian] h- only survived in the dialects of Aman. It disappeared without trace in Sindarin" (WJ.365). - The second element of ekla-mbar comes from the stem MBAR "dwell, inhabit" (LR:372), here used as a noun "home". In the Etymologies, it is suggested that MBAR is a strengthened form of BAR, probably meaning "raise", but the how the meaning "dwell, inhabit" could develop from "raise" is not explained.
ekla-rista "Eglarest", a place-name apparently including rista "cut; ghyll, ravine". Like ekla-mbar above, this form is asterisked as unattested in WJ:365. Concerning the element ekla, see ekla-mbar. The element rista must be referred to the stem RIS "slash, rip" or "cut, cleave" (LR:384, where two different entries are devoted to this stem). An Old Sindarin verb rista- "rend, rip" (q.v.) is here mentioned. The rista of ekla-rista would seem to be a noun derived from this verb, referring to a cut in the landscape, hence a ravine. The ending -ris in Sindarin Imladris "Rivendell" would be related. See rista.
elen-barathi a name of Varda, ancestral form of Sindarin Elbereth (intermediate form between Old Sindarin and Classical Sindarin: Elmbereth) (MR:387). Concerning barathi, see separate entry. Elen "star" is "according to Elvish legend" derived from the primitive exclamation ELE "lo!", "behold!" ("made by the Elves when they first saw the stars"); elen represents an extended form of ELE, with -n suffixed to the reduplicated stem-vowel (ómataina). This was Tolkien's later concept, as described in WJ:360 (also alluded to in the Silmarillion Appendix, entry êl, elen). In the Etymologies, the corresponding stem (EL, LR:355) is simply defined as "star" or "starry sky", and there is nothing to suggest that the later connection with "behold!" had yet arisen in Tolkien's mind.
elle "sky". Derived from a stem 3EL (LR:360) of similar meaning; Tolkien imagined that after the loss of 3, the Elves confused this stem with the originally distinct stem EL "star, starry sky" (cf. LR:355). This word, together with the Quenya cognate hellë, point to a primitive form *3elli or more probably *3ellê. The doubling of the l could be some kind of medial fortification; the ending -ê may simply be the stem-vowel suffixed and lengthened. Conceivably a longer ending -lê could be present, but this ending is normally abstract or universal (see belle), and "sky" is a relatively concrete phenomenon. - In the Etymologies, the primitive sound spelt 3 is said to be a "back spirant" (LR.360, in an editorial note); this would be the spirant equivalent of g, spelt gh in Orkish (as in ghâsh "fire"). It is very much possible that Tolkien later decided that this primitive sound was rather a normal h; notice that while the Quenya word ho "from" and the plural genitive (partitive) ending -on are derived from a stem 3Ô in the Etymologies (LR:360), the prefix hó- of similar meaning, as well as the genitive ending, are derived from a stem HO in the essay Quendi and Eldar written decades later (WJ:368-369). 3 in Tolkien's early ideas about Primitive Quendian would therefore seem to correspond to H in his later conception. Notice that this 3, like later H, is lost in the Sindarin branch, but yields h in Quenya (elle = Quenya hellë; compare Ekla-mbar = Q Heceldamar above). It seems, then, that the primitive form of elle could be reconstructed as hellê just as well as 3ellê (and the stem 3EL should perhaps be updated to *HEL).
elwa "(pale) blue". Derived from the same stem 3EL "sky" (LR:360) as elle above; the primitive form would be *3elwâ with the adjectival ending -wâ (concerning which see katwe): literally *"sky-like" with reference to colour, hence "(pale) blue".
elyadme "rainbow", lit. "sky-bridge" (LR:360 s.v. 3EL). The initial element el- represents the stem 3EL "sky" itself. (Conceivably, elyadme could also be interpreted *"star-bridge" after the loss of 3; compare elle above - but rainbows obviously do not appear at night when the stars are visible.) Concerning yadme "bridge" (only attested in this compound), see separate entry.
et- prefix "forth, out", in Old Sindarin only attested in the word etledie (see below). This prefix has the same origin as the identical Quenya prefix mentioned in the entry for the stem ET (LR:356), though no OS form is there listed (but the later Sindarin form ed- is).
etledie "go abroad, go into exile". An infinitive consisting of three elements: the prefix et- "forth, out" (see above), the verbal stem led- "go, fare, travel" (LR:368 s.v. LED - this is where the OS form etledie is listed) and the infinitive ending -ie (concerning which see bronie). The past tense of etledie would be *etlende; see lende.
etledro "exile" (LR:368 s.v. LED). This word may refer to an exiled person rather than "exile" as an abstract: The literal meaning may be *"out-goer", used with reference to the Noldor who left Valinor to become Exiles in Middle-earth. To the verb etled- "go out" (see etledie above) is added the (masculine/) agental ending -ro (primitive -rô, WJ:371; compare bioro). However, -ro does seem to function as an abstract ending in the Old Sindarin word ndakro "slaughter, battle" (verb ndak- "slay"). Therefore, etledro may be intended to mean "exile" as an abstract after all.
etlenna "exiled" (LR:368 s.v. LED). The must be considered a past participle of the verb etled- "go out" (q.v.), hence literally "outgone". The primitive form must be *etlednâ with the ending -nâ, often used to form adjectives and past participles. Notice that dn is assimilated to nn in Old Sindarin.
Ezde > Eide "Rest", name of a Valië, the wife of Lórien (LR:356 s.v. EZDÊ; the form Ezda given under SED [LR:385] may be a misreading). See Eide.
Findekâno masc. name, "Fingon". (LR:381 s.v. PHIN). This language this form is supposed to belong to is not identified; it is mentioned as the ancestral form of the Sindarin name Fingon, so we include it here. It seems to be a hybrid that could never actually have occurred at any stage in the linguistic evolution; Tolkien mixes things up. The element finde did not have this form in Old Sindarin; it was phinde (q.v.), for ph became f only in later Sindarin. On the other hand, the element kâno must be earlier than even Old Sindarin, for in this language non-final â became ó. The actual Old Sindarin form must have been *Phindekóno. In the scenario of the Etymologies, phinde means "skill", while the element *kóno is to be derived from the stem KAN "dare" (LR:362). In this entry the later Sindarin form caun, in compounds -gon, is mentioned; these forms would be derived from Old Sindarin *cóno, primitive *kânô (later kâno) with the masculine ending -ô; the meaning would then be *"daring one", *"brave one". *Phindekóno or "Findekâno" thus means *"Skillful Brave One". However, this is not how Tolkien later explained the name Fingon. In PM:345, Fingon is said to be a Sindarized form of Quenya Findecáno (Findekáno). The first element is said to be findë "hair" (a tress or plait of hair); it is mainly meant to echo the fin- of the name of Findecáno's grandfather, Finwë. Findë "hair" is derived from primitive phindê, which would yield Old Sindarin *phinde, clashing with phinde "skill" from the Etymologies (in the Etymologies, Quenya findë "tress, braid of hair" corresponded to Old Sindarin sphinde instead, since in Tolkien's earlier concept the stem for "hair" was SPIN [LR:387] instead of PHIN). Cáno in Tolkien's revised concept means "commander", derived from a stem KAN "cry, call aloud" (PM:361-362) that is obviously not the same as KAN "dare" in the Etymologies. The primitive form of cáno is given as kânô (said to be the older and simplest agental form, with lengthening of the stem-vowel and the masculine ending -ô), which would again yield Old Sindarin *kóno. However, since the name Fingon is now said to be Sindarized from Quenya Findecáno, there is not now any need to involve any Old Sindarin forms.

gaia "dread". Derived from a stem GÁYAS "fear" (LR:358). This stem seems to be an expanded version of a simpler stem GAY; see gêrrha below. Gaia must be derived from such a simpler stem: primitive *gâjâ (*gâyâ), *gaiâ. If the stem has a verbal meaning "fear", this would be some kind of verbal noun: "fearing, fear" = "dread". For a similar formation, compare kânâ "outcry" from the stem KAN "cry, call aloud" (PM:361-362; not the same stem as KAN "dare" in the Etymologies, LR:362).
gása "the Void". Derived from a stem GAS "yawn, gape" (LR:357; the form gása is listed on the next page; it is not glossed but simply equated with Quenya cúma, in turn glossed said to refer to the Void, sc. the emptiness beyond the World: LR:365 s.v. KUM). A primitive form gâsa is given (LR:358), but the most primitive form must have been *gâsâ, before the shortening of the long final vowels (an original short final -a would not appear in Old Sindarin, but be lost already at the Common Eldarin stage). The word shows lengthening of the stem-vowel and an ending -â that may here be used as a noun-former; it may be simply the stem-vowel suffixed. It is surprising that the first â of gâsa, *gâsâ comes out as á in Old Sindarin, since non-final â normally becomes ó instead: *gósa. This may be a misreading; compare tára. - A later OS form would be *góha, after the change of s to h in such positions (cf. kheleha from khelesa).
gêrrha (earlier gæ^sra) "dreadful". Derived from a stem GÁYAS "fear" (LR:358). This looks like an expanded form of a simpler stem GAY, but while such a stem is indeed listed above GÁYAS, it does not seem to have a suitable meaning (GAY itself is undefined but yields words for "red, copper-coloured, ruddy"). However, a stem GAYA "awe, dread" is actually mentioned in later writings (PM:363; Silmarillion Appendix s.v. gaer). Whatever the case, the primitive form of gêrrha is given in LR:358 as gaisrâ, the gais- part representing the stem GÁYAS and -râ being a primitive adjectival ending (concerning which see tára). The first Old Sindarin form of gaisrâ was gæ^sra, the original diphthong ai becoming a long æ (like a in English cat, but longer). It seems that primitive ai turned into long æ only before a consonant cluster; notice that ai is unchanged in a word like yaiwe, q.v. (primitive *yaiwê). In the later form gêrrha, the long æ has turned into a long e (ê). The earlier medial cluster sr now appears as rrh; this is probably to be understood as long unvoiced R (since unvoiced r is often spelt rh in Tolkien's works). It seems that sr was assimilated to rr (compare medial sm becoming mm, see ammale), but this double R was unvoiced like the s that had become assimilated. This word gerrha in turn produced Sindarin gaer. With the introduction of a new stem GAYA in Tolkien's later writings, possibly replacing GÁYAS of the Etymologies, it may be just as well to derive gaer from a simpler adjectival form gairâ (WJ:400) Old Sindarin *gaira. This would make gaer the cognate of Quenya aira "holy" (PM:363).
Gondambar "Stone of the World", a name of Gondolin. It is not said what language this form (mentioned in LR:359 s.v. GOND) belongs to; it is just mentioned as an "old" form underlying Sindarin Gondobar, so we include it here. The element gond- "stone" probably represents a full Old Sindarin word *gondo, cognate with Quenya ondo, clearly derived from primitive *gondô (cf. Letters:410, PM:374). Ambar "world" (as an Old Sindarin word only attested in this compound, unless Phind-ambar is also taken as OS) is derived from the stem MBAR "dwell, inhabit" (LR.372). In the form ambar (the same word is used in Quenya), the stem-vowel has been prefixed to produce an "intensive" form: cf. the spelling a-mbar in LR:372 s.v. MBAR. In the Etymologies, Tolkien defined a-mbar as "oikumenê", a Greek word for the world as the home or dwelling-place of the human race. Compare the meaning of the stem MBAR. - The compound Gond-ambar is literally "Stone-world"; the second element is to be understood as a genitive, hence "Stone of World" rather than "World (made) of Stone" or similar. Perhaps an explicit genitive marker was present at an earlier stage; compare WJ:370, where it is suggested that Sindarin may well have developed inflectional -ô in the "primitive period". A long final -ô would belong to the Common Lindarin stage or only shortly after it; Old Sindarin would already have -o. Perhaps Gondambar is a form intermediate between later Gondobar and Old Sindarin *Gondo-ambaro.
*gósa possible correction of gása, q.v.
gæ^sra > gêrrha "dreadful" (LR:358 s.v. GÁYAS). See gêrrha.
hwesta "puff, breath, breeze". Derived from a stem SWES "noise of blowing or breathing" (LR:388). A primitive form swesta- mentioned by Tolkien is evidently a verb "to puff"; read probably *swestâ- with a long final â, for final short a would have disappeared at the Common Eldarin stage. The ending -tâ is a frequent verbal ending, sometimes causative (see bértha-), but here it simply forms a verb from a non-verbal stem. The noun hwesta is in turn derived from this verb; Quenya had the same word. Notice the precise wording in LR:388 s.v. SWES: "Q hwesta- to puff; hwesta breath, breeze, puff of air; ON [Old "Noldorin"/Sindarin] hwesta". Tolkien's intention may be that Old Sindarin hwesta is both noun and verb, corresponding to both meanings of the identical Quenya word. - This word is our sole explicit example of how primitive sw behaves in Old Sindarin; it becomes hw. This digraph is surely meant to represent the same sound as in Quenya, unvoiced w (English wh, in dialects where which is audibly distinct form witch). - Initially, primitive st becomes sth in Old Sindarin, but this does not happen here (**hwestha). This example, as well as rista, may suggest that this change did not occur medially.
hyúle "incitement" (or "cry of encouragement in battle", if = Sindarin hûl, the word it yielded). Derived from a stem SIW "exite, egg on, urge" (LR:386). The Quenya cognate siulë points to a primitive form *siulê with the abstract ending -lê (VT39:16), the W of the stem SIW becoming a full vowel before a consonant, producing a diphthong iu. It seems that at the Common Lindarin stage, this diphthong became yu (ju) following a dental consonant, the i turning into a semi-vowel y before u. (A footnote in LotR Appendix E states that by the Third Age, Quenya iu had likewise come to be pronounced like yu in English yule; in the Lindarin branch, a similar change had evidently happened ages earlier, if only in certain environments.) The Common Lindarin form of *siulê was evidently *syûlê (*sjûlê); notice how u becomes long ú to maintain the prosodic length of the lost diphthong iu. The combination sy (sj) later becomes hy in Old Sindarin; the digraph hy undoubtedly represents German ich-Laut, as in Tolkien's normal spelling of Quenya (e.g. hyarmen "south"). Indeed Quenya hy also comes from older sy- in many cases, though this change must be wholly independent from the similar change taking place in Middle-earth. (Hy did not survive into later Sindarin, but became h, hyúle yielding hûl.)
Ide "Rest", name of the wife of Lórien, a Valië (Quenya Estë) (LR:357 s.v. EZDÊ). See Eide.
ien-rinde "year" (LR:400 s.v. YEN). Literally "year-circle", apparently referring to the year as a cyclus (compare the Quenya term coranar, "sun-round"). The language this word belongs to is not identified; it is mentioned as the ancestral form of Sindarin idhrin, so we include it here. Ien represents the stem YEN "year" (so in the Etymologies; in LotR, Tolkien used the Quenya derivative yén to denote a "long year", an Elvish century of 144 solar years - but this is apparently not the meaning here). Normally, primitive initial y remains y in Old Sindarin (compare yaiwe, yura and possibly even yen in yen-panta, though it is not explicitly said that the latter is Old Sindarin). The development ye > ie seems to have happened after the stage normally called Old Sindarin, so perhaps ien-rinde would be *yen-rinde in "pure" OS. The second element rinde is clearly the same as an identical Quenya word for "circle" mentioned in the Etymologies in the entry RIN (LR:383); the later Sindarin word descended from ien-rinde (idhrind > idhrin) is also mentioned here. The stem RIN itself is not glossed, but all the words derived from it have to do with circles or something circular. Rinde should probably be derived from *rindê or possibly *rindi, with medial fortification N > ND.
impanta see yen-panta
in-fant see yen-panta

Indlour, a masc. name (LR:361 s.v. ID). The language this word belongs to is not identified; it seems to be an older form of the name Inglor, so we include it here. Two primitive forms are suggested: Indo-klâr or Indo-glaurê. The element indo means "heart", derived from the stem ID, not itself defined but cf. îdî "heart, desire, wish". The most primitive form would be indô with nasal-infixion and an ending -ô that may be simply a noun-former, perhaps also an agental ending (if the heart is considered a "desirer"). Klâr occurs nowhere else, but it must be derived from the stem KAL "shine" (LR:362); a form k'lâ is said to underlie Quenya cala "light", so perhaps we are to assume that there was a primitive agental form *k'laro "shiner" (the agental ending -ro is mentioned in WJ:371), later becoming klâr in Common Eldarin. The alternative suggestion glaurê (in Indo-glaurê) must be taken as a g-prefixed variant of laurê, "gold, golden light", derived from the stem LÁWAR - which stem in LR.368 is said to have the alternative "Noldorin" (later: Sindarin) form GLÁWAR. Thus, Indo-glaurê must mean something like "heart of gold", while Indo-klâr may be interpreted "Shining Heart". The later form Indlour has lost the final o of indo and the initial consonant of glaurê or klâr (g could be lost through normal Sindarin lenition, though this process belongs to a later stage than Old Sindarin; the k of klâr was perhaps assimilated to g by the preceding d, this g later disappearing through lenition again). The earlier diphthong au, alternatively the long vowel â, now appears as ou - a strange combination that sometimes turns up in Tolkien's spelling. In Oroume, later Araume, ou becomes au - but here earlier au becomes ou instead! Alternatively, long â (as in klâr) should normally become Old Sindarin ó (see abóro). In the entry dogme, dougme, doume we suggest that ou may there be a way of spelling long ú; perhaps ou in Indlour is a way of spelling long ó? (Compare "Féanour" from Phay-anâro in LR:381 s.v. PHAY; normally, primitive â produces Old Sindarin ó and Classical Sindarin au, aw.) In short, Indlour is a rather curious form. In LR:381, "Féanour" is listed as a "N[oldorin]" form, corresponding to "ON"/OS Phayanôr, so perhaps the name Indlour should similarly be taken as later "Noldorin" for "ON" *Indoklôr. If so, the name Indlour should not be included here at all.
kamba ?"(hollow) of hand". In the Etymologies, this word is mentioned in the entry MA3, LR:371. However, kamba itself is derived from the stem KAB "hollow" (LR:361). No Old Sindarin form is there mentioned, but Quenya kambe (better spelt cambë) is there translated "hollow (of hand)" (which gloss we here tentatively assign to kamba as well). This Quenya form may suggest that kamba in the entry MA3 is a misreading for *kambe. If we accept kamba, the primitive form would most likely be *kambâ with nasal infixion and the (here nominal) ending -â. Since this ending is more commonly adjectival, we could speculate that *kambâ was originally an adjective "hollow" (same meaning as the stem), later also being used as a noun "a hollow", the meaning eventually being specialized: the hollow of a hand. The later Sindarin word cam came to mean simply "hand".
katwe "shaped, formed". The primitive form is given as katwâ, derived from a stem KAT "shape". The ending -wâ is adjectival; compare for instance some primitive colour-adjectives "reconstructed" by Tolkien: laik-wâ "green", smalwâ "fallow, pale", narwâ "red" (LR:368 s.v. LÁYAK, LR:386 s.v. SMAL, LR:374 s.v. NAR1 - narwâ is not asterisked, but because of the long final vowel it cannot be Quenya). Since final -â normally yields Old Sindarin -a, we might have expected **katwa instead of katwe. However, a special phonological rule may be in effect here. Compare some forms given in LR:400 s.v. YAT, where jatmâ (yatmâ) "bridge" yields Quenya yanwë instead of **yanwa (cf. also Old Sindarin yadme in elyadme). It seems that the endings -mâ, -wâ were altered to -mê, -wê following a t (or possibly following any dental consonant, but if so, pathwa rather than **pathwe from *pathmâ is a curious exception). Hence, katwe may actually descend from an altered form *katwê.
kelepe "silver". Derived from a stem KYELEP of similar meaning (LR:366; the form TELEP also given there is the later Telerin form of this stem). The primitive form is not explicitly given in the Etymologies, but Letters:426 has kjelepê (there spelt kyelepê). This is our only explicit example of primitive kj (ky) being simplified to k in Old Sindarin. (This change occurred already at an earlier stage, in Common Lindarin, when all the primitive palatalized sounds were depalatalized: kj > k, nj > n etc. This change is therefore reflected in the descendant languages: Sindarin, old and classical, and the Telerin of Aman.) The ending -ê in kjelepê could be simply the stem-vowel suffixed and lengthened, but -ê is also found in a number of other primitive nouns denoting substances, e.g. mazgê "dough" (LR:371 s.v. MASAG) or srawê "flesh" (MR:350).
khalla "noble, exalted" (compare orkhalla), derived from the stem KHAL2 "uplift" (LR:363). The primitive form is given as khalnâ (indicated to be accented on the final syllable) with an ending -nâ that sometimes simply forms adjectives (compare magnâ), but often it also functions as a past participle ending. In this case, khalnâ is literally *"uplifted", the past participle of the verbal stem "uplift". For other examples of the assimilation *ln > ll, see skhalla (< skalnâ) and skhella (< skelnâ).
kheleha "glass", from earlier khelesa. (Kheleha was misread as "khelelia" by the transcriber of the Etymologies: LR:365 s.v. KHYEL(ES). Primitive khjelesê could not possibly yield OS khelelia, but for the change of intervocalic s to h, cf. for instance baraha [q.v.] from barasa. Khelesa > kheleha is therefore wholly plausible. Another, similar misreading turned pelehi into "peleki"; see pele.) The stem, already mentioned, is given as KHYEL(ES) (simply defined as "glass"), apparently indicating a simple stem KHYEL with an expanded form KHYELES showing reduplication and suffixing of the stem-vowel (so-called ómataina, vocalic extension) and a suffixed -S. Tolkien's "reconstructed" primitive form khjelesê (spelt khyelesê in LR:365) shows the ending -ê, one function of which is to derive nouns denoting substances (see kelepe above for some examples). It is surprising that this original final -ê becomes -a in Old Sindarin khelesa/kheleha, for normally, -ê yields -e (cf. for instance kelepe from kjelepê, see above). It may be that this is a misreading and that the OS forms should actually be *khelese, *khelehe; it would not be the first case of an editor confusing e and a in Tolkien's difficult handwriting. - This is the only explicit example showing how primitive khj- comes out in Old Sindarin; it becomes kh-, merging with original kh- (unchanged in OS). The depalatalization of khj- to kh- simply reflects the general loss of palatalization in Common Lindarin; compare kj > k (kjelepê yielding kelepe). - Kheleha produced "Noldorin" hele, but in Tolkien's later vision of Sindarin, the word for "glass" is heledh, and this is now said to be a loan from Khuzdul (Dwarvish) kheled and not originally an Elvish word at all: See the Silmarillon Appendix, entry khelek-.
khelelia - misreading for kheleha, q.v.
khelesa (later kheleha) "glass" (LR:365 s.v. KHYEL(ES))
khéro "master". Derived from a stem KHER "rule, govern, possess" (so glossed in LR:364; simply "possess" in Letters:178). The Quenya word heru "master" is also mentioned in this entry, and the primitive form of this noun is given in Letters 282: kherû. In this word, the masculine/agental ending -û appears. However, kherû should have yielded Old Sindarin *kheru, not khéro. The latter may rather represent another example of the "older and simplest agental form" (PM:362) derived by means of the masculine/agental ending -ô combined with lengthening of the stem-vowel: *khêrô "ruler, governor, possessor". This is still not wholly unproblematic, for since long non-final ê yielded Old Sindarin í (see dîr), we would have expected *khíro instead. In the Sindarin word descended from khéro, hîr, former ê, é has indeed become î. It may be that khéro is an error, by Tolkien or the transcriber, for *khíro. Compare the word khíril (not **khéril) "lady", that is listed immediately after khéro "master". - Letters:282 also lists a simpler primitive form khêr. The context suggests that Tolkien had now decided to derive Sindarin hîr from this form, in which case the Old Sindarin form should be altered to *khír.
khíril "lady". Derived from the same stem KHER "rule, govern, possess" (LR:364) as its masculine counterpart khéro above. Instead of masculine -o, this word shows the feminine ending -il. Here the expected change from primitive long ê to Old Sindarin í does take place: khêr- > khír-, supporting the assumption that khéro should similarly read *khíro.
ku, kua "dove". The primitive form Tolkien derived from the undefined stem KÛ is kukûwâ. Kukû- may be onomatopoetic, while the ending -wâ is probably adjectival (see katwe concerning this ending). Perhaps kukûwâ is an adjective referring to the sound made by a dove, later used as a noun referring to the animal itself (for another primitive bird-name in -wâ, compare alpha "swan" from alk-wâ). Old Sindarin kua must come from a shorter form *kûwâ, perhaps formed from kukûwâ simply by haplology. The semi-vowel w evidently merged into the u before it, producing *kûâ, *kuâ > kua. The alternative form ku has lost the final vowel as well, though final -â regularly becomes -a in Old Sindarin.
kúma "empty, void". Derived from the stem KUM "void" (LR:365); the primitive form would be *kûmâ with lengthening of the stem-vowel and the adjectival ending -â.
lende "fared". A past tense formed from the verbal stem LED "go, fare, travel" (LR:368) by nasal infixion and the ending -e; this is also a common way of forming the past tense in Quenya, which language indeed has the same word with the same meaning. The infinitive of this verb is ledie, attested with a prefix (see etledie). Compare ndakie "to slay", nasal-infixed past tense ndanke.
linde "singer / singing", used as a name or part of the name "of many rivers of quick course that make a rippling sound". Mentioned in WJ:309 as the origin of the second element of the Sindarin river-name Taeglind (the published Silmarillion has the form Teiglin). In WJ:309, this linde
is not explicitly said to be Old Sindarin, but it seems to belong to the same stage of development as the "Old Noldorin" forms of the Etymologies. The most primitive form would probably be *lindê. The stem is obviously LIN, having to do with "melodious and pleasing sound" (WJ:382; compare the stem LIN2 "sing" listed in the Etymologies, LR:369). The form *lindê shows medial fortification N > ND and the ending -ê, that may be both abstract (hence the interpretation "singing") and feminine/agental (hence the interpretation "singer"; cf. the final element of Quenya l&oacutemelindë "nightingale, *night-singer").
líre "row, range". Derived from a stem LIR1, itself not defined. The primitive form would most likely be *lîrê, with lengthening of the stem-vowel and final -ê, in this case just a noun-former (not abstract or feminine as this ending often is).
litse , later litthe "sand". The stem LIT (LR:369) is not defined. Litse would derive from *litsi or more probably *litsê. Exactly what the ending -sê would mean here is far from clear (see brasse for some thoughts about this ending); the final vowel -ê is not uncommon in words denoting substances. The change ts > tth evidently occurred during the Old Sindarin stage; this seems to be our sole explicit example of this change, though earlier tt also becomes tth in Old Sindarin (see batthô, rattha).
loga "fenland". This form, mentioned in UT:263 as the origin of Sindarin lô, is not explicitly said to be Old Sindarin (only identified as an "earlier" form). However, loga is said to be derived from a stem log- "wet, soaked, swampy" (not in the Etymologies) and could represent an intermediate form between primitive *logâ and Sindarin ló. A form *logâ may simply be an adjective (with the common adjectival ending -â), having much the same meaning as the stem; later this adjective may have been used with reference to a concrete "swampy" place, hence developing into a noun "fenland". (OS loga may still have been an adjective; the gloss "fenland" in UT.263 primarily applies to Sindarin lô.)
magnâ "skilled". Derived from a stem MAG "use, handle", said to be related to MA3 "hand" (LR:371); "handle" would thus seem to be the most literal definition of MAG. The form magnâ cannot be correct Old Sindarin, but is an obvious mistake, by Tolkien or the transcriber, for *magna. The primitive form, also mentioned in LR:371, was indeed magnâ; perhaps the two were confused: In Old Sindarin, earlier long final â was shortened to -a in polysyllabic words; all the long final vowels were so shortened (see abóro). - The meaning of this word is somewhat surprising. Since the adjectives formed with the ending -nâ are often seen to be a kind of past participles, we might have expected magnâ, derived from a stem meaning "use, handle", to mean "used, handled" (compare skalnâ "hidden" from a stem SKAL1 "hide"; see skhalla). Instead, -nâ is here used as a very generalized adjective-former; the meaning "use, handle" is developed in the direction of "being good with regards to using or handling things" and hence "skilled". Compare maite "handy, skilled", derived from the related stem MA3 "hand".
mai pl. of mó, q.v. (LR:371 s.v. MA3)
maiga "pliant, soft", earlier mazga. Derived from a stem MASAG "knead, make soft by rubbing, kneading, etc." (LR.371). A primitive form mazgâ is listed; the very oldest form may have been *masgâ, before s was voiced to z in contact with the following voiced consonant (something that seems to have happened in Common Eldarin; see Eide). Mazgâ shows the common adjectival ending -â. In (later) Old Sindarin, z became i before a consonant; compare Eide, from earlier Ezde. Hence mazga > maiga.
maite "handy, skilled". Derived from the stem MA3 "hand" (LR:371); the primitive form is given as ma3iti with an adjectival ending -iti. An shorter ending -ti is seen in such adjectives as neiti- "moist, dewy" (LR:376 s.v. NEI), phoroti "right" or "north" (LR:382 s.v. PHOR). There seem to be no direct attestations of the longer ending -iti in words "reconstructed" by Tolkien himself, but the Quenya ending -itë in adjectives like uruitë "fiery" (LR:396 s.v. UR) is clearly descended from -iti. Primitive ma3iti became maite already at the Common Eldarin stage, after the loss of medial 3 (H) and the change of final short -i to -e. - In "Noldorin", maite yielded moed (LR:371); this would correspond to maed in later Sindarin. (A similar, though semantically distinct word maed "shapely" is mentioned in PM:366; this is said to be derived from magit-, presumably indicating a complete form *magiti.)

malda "gold" (as metal). In LR:386 derived from a stem SMAL "yellow"; the primitive form is given as smaldâ, evidently showing medial fortification l > ld and a (here nominal) ending -â, possibly simply the stem-vowel reduplicated and suffixed. This example, as well as malina and malo below, demonstrate that the primitive initial cluster sm- is simplified to m- in Old Sindarin. Tolkien also considered letting Old Sindarin ("Old Noldorin") retain sm-, in turn yielding Sindarin/"Noldorin" hm- (unvoiced m); see LR:387. However, this idea was evidently dropped (Anthony Appleyard has pointed that if this idea were valid at the LotR stage, mallorn would have been hmallorn instead). In David Salo's reconstruction of Old Sindarin, primitive initial sm yields OS hm- (unvoiced m), which in turn becomes voiced m- in Sindarin. - It seems doubtful that the explanation of the Eldarin words for "gold" that is offered in the Etymologies was still valid in Tolkien's later scenario. For one thing, according to the entry SMAL in Etym, the Quenya word for "gold" was also malda. In LotR Appendix E, the Quenya word for "gold" is said to be malta instead (mentioned as the name of tengwa #18). In accordance with this, a late source states that the Eldarin stem for "gold" was MALAT (PM:366) - not SMAL. Tolkien had evidently decided to derive Quenya malta from *malatâ instead. The latter would produce Old Sindarin *malata, in turn yielding Sindarin malad (as in the name Rathmalad, WJ:191).
malina "yellow". Derived from a stem SMAL (LR:386) of similar meaning; the primitive form is given as smalinâ. The ending -inâ seems to be basically an extended variant of the common adjectival or participial ending -nâ. Other Old Sindarin words exemplifying this ending are berina "brave" and pikina "tiny", that must be derived from *berinâ, *pikinâ. Via Old Sindarin -ina, primitive -inâ gave rise to Classical Sindarin -en, one of the most common adjectival endings in Grey-elven.
malo pl. malui "pollen, yellow powder". Derived from a stem SMAL "yellow" (LR:386); the primitive form is given as smalu with a very unusual ending; where -u occurs, it tends to denote body-parts (e.g. ranku "arm", LR:382 s.v. RAK) or localities (e.g. jagu- "gulf", LR:400 s.v. YAG). Smalu is the sole word in -u that denotes a substance; it seems that it is here simply a noun-former used to derive a word for "something yellow". - In Common Eldarin, smalu presumably became *smalo, producing malo when initial sm was simplified to m in Old Sindarin. As suggested in the main article above, the fact that the plural form still shows -u- as a part of -ui could be explained by assuming that the final syllable of the primitive plural *smaluî, *smalui did not change in Common Eldarin (while final -u became -o).
map- "seize, take away by force". Simply the stem MAP "lay hold of with hand, seize" (LR:371) with no additional elements. Tolkien probably intended MAP to be related to MA3 "hand", found on the same page in the Etymologies.
matthô-be "to handle" (this is evidently simply another way of spelling matthóbe, matthó-be; compare for instance buióbe). Derived from a stem MA3 "hand" (LR:371). The primitive form is given as ma3-tâ (the hyphen emphasizing that -tâ is a derivational ending added to a stem ma3-). The suffix -tâ is a very common verbal ending, and it can be seen that the gloss "to handle" is very literal, since ma3- means "hand". Tolkien notes that ma3-tâ in turn yielded (Common) Eldarin mahtâ. Here, the letter h probably represents [x], sc. German ach-Laut: The back-spirant 3 was eventually devoiced in contact with the unvoiced sound t, becoming [x]. Compare wahtê "stain, soil" from the stem WA3, evidently Common Eldarin for Primitive Quendian *wa3tê. Wahtê became Old Sindarin watte (q.v.), ht being assimilated to tt, and mahtâ- must similarly have become *mattó- in early OS, later turning into matthó- when double tt became tth (see batthô - we must likewise assume that watte later became *watthe). Matthô-be shows the infinitive ending -be (see buióbe), and since the primitive verbal ending -tâ thus was not final, the â becomes ó (here spelt ô) instead of -a.
mazga "pliant, soft" (LR:371 s.v. MASAG). See maiga (the later form).
míre "jewel". Derived from an undefined stem MIR (LR:373); the primitive form would be either *mîri or (more likely) *mîrê. The ending -ê can denote substances, so perhaps *mîrê originally meant "jewel" as a substance rather than as a concrete gem. However, Quenya mírë can be used of a (single, concrete) jewel, and this is probably also true of the identical Old Sindarin word (we know it is true of the descendant Sindarin word mîr; compare The Lays of Beleriand p. 354, where the moon is apparently called menel-vîr or *"jewel of heaven" [-vîr being the lenited form of mîr]).
mó pl. mai "hand". Derived from a stem MA3 (LR:371), similarly defined as "hand". In this entry in the Etymologies, the Primitive Quendian word for "hand" is said to have been mâ3 with a long vowel (but short ma3- before an ending). After the loss of the guttural 3, only mâ would be left, and this form is attested in the entry DAL, LR:353 (where it is said that Old Sindarin dalma "palm of hand" probably derives from a compound of dal "flat" and mâ "hand"). Normally, final long -â becomes Old Sindarin -a (as the word dalma demonstrates), but in a monosyllable like mâ this vowel turns into ó, as â normally does in polysyllabic words when not final (see abóro). The plural form mai preserves the original quality of the stem-vowel: In the linguistic scenario of the Etymologies, the plural form of primitive mâ3 would be *ma3î, becoming *mâî, *maî, mai after the loss of 3: If the a was ever long (it may have been lengthened in compensation when 3 was lost), it quickly became short again when it merged into the plural ending -î to produce a diphthong aî/ai that remained ai in Old Sindarin. - This scenario may have to be modified a little in light of Tolkien's later vision of the development of the Eldarin word for "hand". In VT39:11, in a document dating from about 1960 (hence well over two decades after the Etymologies was written), it is said that the primitive word for "hand" was maha rather than mâ3. After the early loss of medial h in Common Eldarin (WJ:368), maha would become *maa = mâ with a long vowel, which would again yield old Sindarin mó as above. The plural form of primitive maha would presumably be *mahaî, *mahai, which could easily evolve into mai after the loss of h (probably via *mâi).
muina "familiar, dear". Derived from an undefined stem MOY (LR:374); the primitive form would be *moinâ (compare Quenya moina) with the adjectival (or participial) ending -nâ. If MOY could be assigned the meaning "love", *moinâ could be considered a past participle "loved" and hence "dear". Notice the Old Sindarin change oi > ui (cf. Uigolosse corresponding to Quenya Oiolossë).
naróbe is glossed "he tells a story" (pa.t. narne), but this gloss clearly cannot be taken literally: naróbe is transparently an infinitive *"to tell a story", showing the infinitive ending -be seen in many other words glossed as infinitives, while the past tense narne simply means *"told a story", with no pronominal elements. Derived from the stem NAR2 "tell, relate" (LR:374). This corresponds to Quenya nyar- (cf. words like nyarna "tale, saga"). The wording in the entry NAR2 in the Etymologies suggests that NAR is the basic form of this stem, with nyar- as a Quenya variant. Actually it would have made no difference if one assumed that the original stem was *NYAR, for since ny (nj) like all other palatalized consonants was de-palatalized at the Common Lindarin stage (see kelepe), the Old Sindarin form would still have been nar-. It seems that this was not how Tolkien imagined it, though. - The form naróbe shows the Old Sindarin infinitive ending -be (see buiobe). Naró- would seem to represent a primitive verbal stem *narâ-, sc. the stem NAR2 with a suffixed -â that is apparently a verbal ending here; such formations are relatively rare (see spharóbe for another example). The stem NAR2 would much rather be expected to function as a "basic" verbal stem, the verb simply being nar- (aorist nare), with the infinitive narie instead of naróbe. We do not asterisk nare, narie, for these forms actually do occur in Tolkien's material, only with a prefix: trenare, inf. trenarie "recount, tell to end". Naróbe seems to be an alternative formation, but its past tense narne is formed directly from the stem nar- (not **naróne or something). Surely the past tense of trenarie is similarly *trenarne. The past tense ending -ne is well known from Quenya, but in our small Old Sindarin corpus, it is attested here only. Most derived verbs (with infinitives in -be) probably form their past tenses in -ne, e.g. buiobe "to serve", pa.t. *buione. However, the pa.t. of spharóbe "to hunt" may be *spharne (instead of *spharóne), following the pattern established by narne.
ndagno "slain [as noun], corpse". Derived from a stem NDAK "slay" (LR:375), concerning which see ndakie. If there is a primitive form, it would be *ndaknô with the masculine ending -nô. However, this ending is generally agental when added to a stem with a verbal meaning. Cf. such primitive words as *besnô "husband" from BES "wed" (LR:352; hence literally "one who weds, one who has wedded somebody") and tirnô "watcher" from the stem TIR "watch, guard" (LR:394, mentioned as part of the compound khalatirnô "fish-watcher"). Hence, a form *ndaknô would mean "slayer" rather than "slain one". It may be better, then, to assume that ndagno is actually a personalized form of *ndagna, a past participle "slain" derived from *ndaknâ with the frequent adjectival/participial ending -nâ (see khalla). - Before a nasal consonant, unvoiced plosives become voiced in Old Sindarin, hence kn > gn. Compare ragme, tulugme, yadme.
ndair "bridegroom". Derived from a stem NDER (LR:375), said to be a strengthened form of DER "adult male, man" (LR:354); the "strengthening" manifests as nasalization of the initial consonant. The primitive form of ndair is given (in LR:375) as ndêro with the masculine ending -o; this form yields "Eld." (= Common Eldarin) ndæ^r. The long æ then turns into Old Sindarin ai. This development seems rather strange; it is difficult to find any parallel to it. Old Sindarin ai must in turn become Sindarin ae (or "Noldorin" oe; in the entry NDER, ndair yields "Noldorin" doer, corresponding to mature Sindarin daer - which is attested, but the language is misidentified; see daer in this list). However, other examples of primitive non-final ê produce Sindarin î, not ae: Examples include nêthê > S nîth and thêrê > S thîr (LR:376-377, 392 s.v. NETH, THÊ); cf. also rênê > S rîn in a late source (PM:372). The example dîr (q.v.) from dêr indicates that this change had already occurred in Old Sindarin; we can assume that nîth, thîr, rîn represent Old Sindarin *níthe, *thíre, *ríne. If primitive non-final ê regularly produced Old Sindarin ai (hence *naithe, *thaire, *raine), the Sindarin forms should have been **naeth, **thaer, **raen, which are not found. If ndair is indeed to be derived from ndêro, we must assume that the development is quite irregular. It may be that the long æ of the Eldarin form should be equated with ae, representing an A-infixed form of the stem NDER, hence ndaer, the diphthong ae becoming Old Sindarin ai (that later reverted to ae in Sindarin). Compare waide, q.v. Such a development would not be wholly in accord with the scenario sketched by Tolkien in VT39:10, where he says ae became long â in the Telerin (Lindarin) branch of the Elvish language family. If so, the Old Sindarin form of *ndaer (later *ndâr) should perhaps be *ndór instead of nd, since earlier â became OS ó (in turn becoming au in later Sindarin: **daur - but the actual Sindarin descendant of ndæ^r was daer). But since VT39:10 reproduces a document about 25 years younger than the Etymologies, some changes in Tolkien's linguistic scenario are to be expected. Etymological intricacies aside, the Sindarin word for "bridegroom" should be daer.
ndakie "to slay", pa.t. ndanke. Derived from a stem NDAK "slay" (LR:375), itself probably a strengthened form of NAK "bite" (LR:374). The form ndakie is formed directly from the stem with the infinitive ending -ie, concerning which see bronie. The past tense ndanke is derived by nasal infixion and the ending -e; this is also a common way of forming the past tense in Quenya. See lende.
ndakro "slaughter, battle". Derived from the same stem NDAK "slay" as the verb ndakie above (LR:375). The primitive form of ndakro would be *ndakrô, which is a somewhat surprising form given the abstract meaning; -rô is a (masculine/) agental ending (WJ:371), so we would expect ndakro to mean "slayer" instead of "slaughter". (Compare bioro.) Perhaps a battle is somehow personified as a "slayer" of the people involved? For another possible case of an Old Sindarin abstract in -ro, see etledro.
ndangwetha "answer" (noun). This form, mentioned in PM:395 as an older form of Sindarin dangweth, may be taken as Old Sindarin (primitive *ndangwethâ with a long final -â?) In the source, the final element is said to be derived from a stem gweth "report, give account of, inform of things unknown or wished to be known". This stem is attested here only. The first element, ndan-, is not explained in the source, but it is obviously the same as the stem NDAN in the Etymologies (LR:375), meaning "back". A ndangwetha or "answer" is thus literally a "back-report". - Clearly the prefix ndan- here simply means "back, in return", which does not agree very well with Tolkien's explanation of it in WJ:412, where he refers to "the element *dan, *ndan- indicating the reversal of an action, so as to undo or nullify its effect". Then ndangwetha would mean, not "back-report" = "answer", but rather "withdrawing of information previously divulged" (to the extent the word would make any sense at all). Of course, the basic idea of ndan- = "back" may be assumed to have developed in various directions in various Eldarin languages.
ndîs "bride" (LR:375 s.v. NDIS-SÊ/SÂ). The actual stem is NDIS, said to be a strengthened form of NIS "woman"; the entry NIS itself (LR:378) describes NIS as an "elaboration" of simpler elements INI, NÎ, which in turn are glossed "woman" (LR:377) and "female" (LR:361), respectively. Notice how the derivation of a stem for "bride" by strengthening a stem basically meaning "woman" parallels the derivation of "bridegroom" from a simpler element meaning simply "man"; see ndair. The primitive form of ndîs is given (in LR:375) as ndîsê, showing lengthening of the stem-vowel and the feminine ending -ê. The Old Sindarin form is surprising, though: we would rather expect *ndíse with the original -ê preserved as a short -e. (Compare kelepe from kjelepê.) However, the "final" Sindarin form would be dîs in both cases.

ndissa "young woman" (cf. dissa). The entry-head in the Etymologies, NDIS-SÊ/SÂ (LR:375), would seem to be a reconstruction in itself, pointing to primitive forms *ndissê, *ndissâ. Concerning the stem NDIS, see ndîs above; here it means "woman" rather than "bride". Ndissa must come from *ndissâ, and Tolkien's "NDIS-SÊ/SÂ" suggests that a separate ending -sâ is involved, but of this ending little can be said. In WJ:416, in a source much younger than the Etymologies, Tolkien mentions a primitive word neresâ. This is said to be a "feminine adjectival formation" from NER "man", meaning "she that has manlike valour or strength". One could speculate that *ndissâ is another such "feminine adjectival formation" in -sâ, so that it was originally an adjective *"she that is womanly" (!), later evolving into a noun "young woman" in Old Sindarin. - The alternative form *ndissê would seem to include the common feminine ending -ê; this would yield ndisse in Old Sindarin. However, the ending -sê in b'ras-sê "heat" (see brasse) may be wholly unrelated.
ndóko "warrior, soldier". Derived from the same stem NDAK "slay" as the verb ndakie above (LR:375). The primitive form is given as ndâkô, with lenghtening of the stem-vowel and the masculine ending -ô, often agental; hence ndâkô is literally *"slayer". (In PM:362, Tolkien refers to a similar formation as an example of "the older and simplest agental form".) As usual, long non-final â yields Old Sindarin ó; cf. abóro.
ndolo "head". Derived from an undefined stem NDOL (LR:376). The primitive form would be *ndolô with an ending that could be masculine, but here it simply forms a noun, or it could be seen simply as the stem-vowel suffixed. Quenya nóla cannot descend from *ndolô; this Quenya word must represent a distinct primitive formation *ndôlâ.
ndor "land" (only attested in the compound Balandor, "Vala-land", Valinor). In the Etymologies, the Eldarin words for "land" are derived from a stem NDOR "dwell, stay, rest, abide" (LR:376). No Old Sindarin word is there listed, but Sindarin dor is stated to come from primitive ndorê. The intermediate Old Sindarin form would be *ndore, of which -ndor in Balandor may be a shorter compound form (from *ndore). - Notice, however, that Tolkien many years later derived the Eldarin words for "land" from a stem DORO "dried up, hard, unyielding" (WJ:413). Even so, this later source does confirm that the Primitive Quendian form was ndorê (that Tolkien now thought to be formed by initial enrichment d > nd). Ndorê 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". (The bounds of Balandor, Valinor, were of course well defined by the Sea.)
nele pl. neleki "tooth". Derived from NÉL-EK (LR:376), apparently a simpler stem *NEL with an extended form showing -K added to ómataina (the reduplicated and suffixed stem-vowel). A stem NEL is indeed listed in LR:376, but its meaning ("three") seems to preclude connecting it with NÉL-EK "tooth". Nele pl. neleki must represent primitive *nelek pl. *nelekî, compare Quenya nelet pl. nelci (final primitive -k became -t in Quenya). Notice that final -k following a vowel is lost in Old Sindarin, but survives in the plural (since it was "shielded" by the plural ending and hence was not final: *nelekî > neleki). Compare oro pl. oroti and skhapa pl. skhapati, q.v.
nestak- "insert, stick in" (LR:388 s.v. STAK). Sindarin nestegi must derive from a fuller form of this word: *nestakie with the infinitive ending -ie (see bronie). The final element in nestak- is plainly simply the stem STAK "split, insert" itself, but where does the prefix ne- come from? It is probably an element meaning "in". A Sindarin (or "Noldorin") sentence mentioned in J. R. R. Tolkien - Artist and Illustrator has neledhi for "walk in" (Old Sindarin *neledie, compare the attested form etledie with another prefix). An independent Sindarin word ned occurs in one variant of the King's Letter, in the phrase nelchaenen ned Echuir "the thirty-first day of the Stirring" (SD:129). If this "of" is actually "in", it could be connected to the first element in nestak-, that would then mean quite literally "stick in". A Sindarin word ned would normally be assumed to descend from Old Sindarin *net. Others want to connect ned to the stems NÉD, ÉNED (LR:376, 356) meaning "middle", "centre", assuming that ned "in" originally meant "in the middle of". (But if the stem is NÉD, we would expect the Sindarin word to be *nedh instead of ned.)
ngalámbe "barbarous speech". This word is not explicitly said to be Old Sindarin, but seems to belong at approximately that stage of the linguistic evolution. (In the Etymologies, this word is asterisked as unattested, though - the Old Sindarin or "Old Noldorin" words are usually not asterisked. Moreover, we might have expected the first, unstressed a to have disappeared already in OS; it is gone in the later Sindarin form glamm.) Ngalámbe (the accent simply indicates which syllable is stressed, not that the second a is long) is derived from a stem ÑGAL (extended form ÑGALAM) "talk loud or incoherently" (LR:377; Tolkien later changed the stem to ÑGYAL, but since palatalized consonants were depalatalized at the Common Lindarin stage [see kelepe], ngalámbe as an Old Sindarin form would not be affected by this revision). While ngalámbe could probably have been explained as the stem ÑGALAM (or *ÑGYALAM) with a medial fortification m > mb and the abstract ending -ê, -e, Tolkien actually imagined that it was influenced by the word lambe "tongue" (not otherwise attested in Old Sindarin and not given a separate entry here, but there can be no doubt that this was also the form of this word - well known from Quenya - in OS). Lambe is obviously derived from the stem LAB "lick" (LR:367), but in the Etymologies, this word is not listed in that entry (though the word lamba referring to the physical tongue is mentioned; it is listed as a Quenya word, but surely had the same form in Old Sindarin). In an essay written much later than Etym, Quendi and Eldar from about 1960, Tolkien explained that lambe is literally "tongue-movement, (way of) using the tongue" (WJ:394), being derived "probably" (WJ:416) from lab-mê, sc. a stem LABA "move the tongue, lick" (cf. LAB in Etym) with the abstract or verbal noun ending -mê (concerning which see ragme). The cluster bm became mb in Common Eldarin or even earlier, producing the primitive form lambê (WJ:394), whence Quenya and Old Sindarin lambe. The form ngalámbe is mentioned in the Etymologies as the origin of Sindarin glamm "clamour" (Glamhoth being a name of the Orcs), but this word is also derived from a stem GLAM (LR:358), in turn said to be the "Noldorin" (Sindarin) form of LAM (such G-prefixion being very common in Sindarin; see WJ:411, note 13). Tolkien's idea was that GLAM had been "influenced" by ÑGAL(AM). The stem LAM, that GLAM is a variant of, is listed in LR:367. LAM itself is not defined; all the words it yields have to do with sound. In a much later source, Tolkien stated that LAM, LAMA "refers to sounds, especially to vocal sounds, but was applied only to those that were confused or inarticulate. It was generally used to describe the various cries of beasts" (WJ:416). In the same source, glam (glamb, glamm) is explicitly said to be derived from LAM (called an "elaboration" of this stem). This scenario is certainly more or less the same as the one set out in the entry GLAM in the Etymologies (LR:358), since GLAM is there said to be a variant of LAM. But in this later source, nothing is said about another stem ÑGAL(AM) or ÑGYAL(AM), or about any influence from lambe "tongue".
ngolfine "magic skill" (LR:381 s.v. PHIN). This word is not explicitly said to be Old Sindarin (or "ON"). It is mentioned as the origin of the second element of the Sindarin name Fingolfin, so we include it here. The proper Old Sindarin form would rather be *ngolphine, since the change ph > Sindarin f occurred later. As suggested by the gloss "magic skill", this is a compound of two elements. Ngol- can be safely identified with the stem ÑGOL "wise, wisdom, be wise" (LR:377), also associated with magic (compare Quenya ingolë "deep lore, magic" and Sindarin gollor "magician"). The second element fine, or better *phine, comes from the stem the word ngolfine is listed under: PHIN "nimbleness, skill" (LR:380). The primitive form could be *phini or *phinê (cf. the longer form phinde, q.v.; in this entry we discuss the reasons for assuming that this stem PHIN may no longer be valid in Tolkien's later scenario). Tolkien says ngolfine contains either "phinya or -phini"; the first must be more or less identical to the adjective phinya "skillful" (q.v.), while -phini must be a suggested primitive form of the final element of ngolfine; this word should evidently be derived from *ñgolphini. - It should be noted that Tolkien later explained the second element of the name Fingolfin quite differently. In PM:344, Fingolfin is said to be the double Quenya name Finwë Ñolofinwë given a Sindarin style and fused into one name. Ñolofinwë is the name Finwë with a prefixed element ñolo- meaning "wise" (obviously derived from the same stem ÑGOL as ngol- in ngolfine; see PM:344). Finwë gave his own name to his all his sons, but added prefixes to distinguish them: Curufinwë, Ñolofinwë, Arafinwë (later, in Sindarin, known as Fëanor, Fingolfin and Finarfin, respectively, though Fëanor is a Quenya-Sindarin hybrid). Concerning Tolkien's various explanations of the name Finwë itself, see Phinwe.
ngolodo "one of the wise folk, Gnome" = Quenya Noldo (LR:377 s.v. ÑGOL, ÑGOLOD). The primitive form is given in PM:360 and WJ:383 as ñgolodô (MR:350: ngolodô), derived from a stem ÑGOL "wise, wisdom, be wise" (LR:377) or "knowledge, wisdom, lore" (WJ:383). The form ñgolodô shows reduplication of the base-vowel (ómataina) and the masculine/animate ending -dô (that also seems to appear in the nasalized form -ndô; see sthabro, sthabrondo). The clan-name Noldor (Old Sindarin *Ngolodi) meant "Lore-masters" (MR:350) or "the Wise" (WJ:383) ("but wise in the sense of possessing knowledge, not in the sense of possessing sagacity, sound judgement" - Silmarillion Index, entry "Noldor").
nguru, ngurtu "Death". Derived from a stem ÑGUR (LR:377), not itself defined but only yielding words for death. The primitive forms must be *ñgurû, *ñgurtû; it is not at all clear where the t in the latter form comes from (this would be the sole attestation of a derivational ending -tû). The ending -û could be interpreted as a masculine ending, if this is Death personified as the capital D suggests (Quenya Nuru, as opposed to nuru, is explicitly said to be Death personified, a name of Mandos). - The form ngurtu must be early Old Sindarin, since primitive rt became rth (see alpha and cf. bértha-; for another word "lacking" the expected change rt > rth, see ortóbe).
nî "woman". This word is identical to the stem NÎ1 of similar meaning (LR:378-9). Later, nî was replaced by dî; see dî concerning the reasons for this change.
nidwa "bolster, cushion". Derived from a stem NID "lean against" (LR:378); the primitive form is given as nidwô. The ending -wô is most unusual, attested in this one word only. It could be a nominal counterpart of the adjectival ending -wâ (concerning which see katwe). A nidwô is apparently *"a thing one can lean against". Normally, primitive final -ô becomes -o in Old Sindarin, not -a as in nidwa. In another entry in the Etymologies, WÔ (LR:399), it is said that when wo (with a short o, explicitly marked as such with a diacritic) was stressed, it became wa in "Eldarin" (Common Eldarin). However, the final vowel of nidwô is long, and whether it is stressed we do not know. Since this is our sole example of the ending -wô, we must simply accept that it comes out as -wa in Common Eldarin - perhaps ô was somehow dissimilated following w.

nui "lament", probably noun. The stem NAY is also defined as "lament" (LR:375), but this gloss is perhaps meant as the verb "to lament" (like the Quenya verb naina- derived from this stem). Tolkien mentions a primitive form naje (naye); the -e may be an abstract ending (more often long -ê). However, naje seems quite unable to yield Old Sindarin nui; we would rather expect *nai. Two Quenya forms, nai and noi, are listed; the latter is evidently related to nui, since oi became ui in Old Sindarin (cf. OS muina, q.v., corresponding to Quenya moina). It would seem that an irregular side-form noje arose in Common Eldarin at the latest, and that this is the origin of Quenya noi and Old Sindarin nui.
[núma] "snare" (and "noose", like Sindarin nû?) (LR:379 s.v. SNEW). Tolkien rejected this form, changing it to sniuma, snýma, but all of these words are clearly to be derived from the same primitive form (*sneumâ). See sniuma.
orie "rise". Evidently an infinitive "to raise"; see bronie concerning the infinitive ending -ie. The stem or- comes directly from the stem ORO "rise" (LR:379), in turn apparently a stemvowel-prefixed variant of RÔ (LR:384) of a similar meaning. Compare ortie.
orkhalla "superior" (LR:363 s.v. KHAL2). This is khalla "noble, exalted" (q.v.) with the prefix or- "above", derived from the stem ORO "up; rise; high; etc." (LR:379). The compound thus expresses *"super-exalted". - Notice that the Sindarin descendant of orkhalla is orchal; the form orchel occurring twice in the Etymologies (entries KHAL2, ORO) is a misreading. Under KHAL2, Christopher Tolkien admits that the e of his reading is uncertain; the correct form orchal is found in WJ:305 (there translated "tall").
orko pl. orkui "goblin" (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", if 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. The plural form orkui preserves the original quality of the primitive final -u, while final short -u otherwise became -o at the Common Eldarin stage, hence sg. orko (compare malo pl. malui and ranko pl. rankui). The forms orko, orkui must be early Old Sindarin, since k became kh following r during this stage (see alpha): We must assume that the "proper" forms *orkho, *orkhui eventually developed. - However, this information from the Etymologies may be invalidated by the scenario set out in the essay Quendi and Eldar, written decades later. Here, Tolkien derived the Elvish words for "Orc" from a stem RUKU having to do with fear (WJ:389, not in the Etymologies) and listed tentative primitive forms: urku, uruku, urkô. The first and the last would both become *urkho in Old Sindarin; uruku would appear as *uruk, in turn yielding Sindarin urug (WJ:390). But the commonest Sindarin word for "Orc" was orch, which according to the same source must be derived either from urkô (with the masculine ending -ô; OS *urkho pl. *urkhi) or from urkâ (OS *urkha), originally an adjective "Orkish" (with the common adjectival ending -â).
oro "mountain", pl. oroti. There is also a longer sg. oroto (its pl. would probably be oroti as well). Derived from a stem ÓROT "height, mountain" (LR:379). ÓROT is itself an extension of the verbal stem ORO "rise" (LR:379), in turn apparently a stemvowel-prefixed variant of RÔ (LR:384) of a similar meaning. The forms oro pl. oroti evidently represent *orot pl. *orotî, the t being lost when final, but preserved in the plural form because it was there "shielded" by the plural ending. Compare the parallel examples skhapa "shore", pl. skhapati, and nele "tooth", pl. neleki. - The longer form oroto descends from the a primitive form with some extra element suffixed to the stem ÓROT, either *orotô or *orotu. In the first case, the final -ô is perhaps simply the stem-vowel suffixed and lengthened. If the original ending was -u, *orotu could belong to a larger group of primitive words in -u denoting localities: jagu "gulf", tumbu "deep valley" and tundu "hill, mound" (LR:400 s.v. YAG, LR:394 s.v. TUB, LR:395 s.v. TUN). However, if the primitive form was *orotu, we would expect the plural form of oro to be **orotui rather than oroti, so it seems that we must opt for *orotô.
Oroume > Araume "Oromë"; see Araume, the later form.
ortie "rise". Like orie of similar meaning, this form is apparently the infinitive of a verb; cf. bronie concerning the infinitive ending -ie. Ortie is derived from the same stem ORO "rise" (LR:379) as orie, but while orie seems to be derived directly from the stem, ortie must be the infinitive of *orta-, from primitive *ortâ- showing the verbal ending -tâ. Since -tâ is sometimes used to form causative verbs (see bértha-), we might have expected ortie to mean "to raise" rather than "to rise". (Normally, verbs in -tâ are at least transitive.) Another infinitive form derived from the same stem, ortóbe, does indeed mean "to raise"; see below. (In Quenya, the verb orta- covers both "rise" and "raise".) It is most unusual that a derived verb like *ortâ- has an Old Sindarin infinitive in -ie; normally, such verb have infinitives in -be instead (and as already mentioned, a form ortóbe actually does occur in the Etymologies). Indeed ortie looks like a confused compromise of orie (the infinitive of the primary verb or-) and ortóbe (the infinitive of the derived verb orta-); perhaps Tolkien imagined that the two had been blended. - The cluster rt becomes rth in Old Sindarin (see alpha), so ortie must be an early OS form, later becoming *orthie.
ortóbe "raise". Derived from the same stem ORO "rise" (LR:379) as ortie above; again, we might have expected *orthóbe instead. Otherwise, as suggested above, the verb ortóbe in all respects behaves more like we would expect. It is transitive, like a tâ-verb usually is (*ortâ-, see ortie), and in Old Sindarin it shows the infinitive ending -be, like derived verbs normally do.
pano "plank, fixed board, especially in a floor", pl. panui. Derived from a stem PAN "place, set, fix in place (especially of wood)" (LR:380). As a verbal stem, PAN evidently has to do with fitting something into its place in a construction. The primitive form of pano is given as panô. The ending -ô is here simply a noun-former, denoting 'something that is fixed in its place', hence a "fixed board". Panô almost looks like an agental formation "placer, one who places", but normally the stem-vowel is also lengthened in such formations (**pânô); cf. kânô from KAN (see Findekâno). - The plural form panui reflects the primitive plural *panôi > *panoi, since the diphthong oi became ui in Old Sindarin (cf. muina, Uigolosse). However, this plural form is quite surprising. We would expect the final diphthong to be simplified to -i already before the change oi > ui. Compare poto "animal's foot", pl. poti (primitive potô pl. *potôi). The plural form panui would be expected to yield *peny is later "Noldorin", but the plural form actually listed in the Etymologies is pein (read pain in later Sindarin). Pein, pain must be derived from *pani, not panui. It is tempting to dismiss the plural panui as some sort of mistake; reading *pani removes all problems.
panta "full" (compare yen-panta). Derived from a stem KWAT (LR.366); the primitive form would be *kwantâ, an adjective derived by nasal infixion and the adjectival ending -â. (Except for the ending, this is not the commonest way of deriving adjectives; compare randa and runda for nouns derived in a similar way.) As pointed out in the entry alpha, primitive kw very early became p in the Lindarin branch of the Eldarin language family (WJ:375 cf. WJ:407 note 5); cf. also póre below. - The stem KWAT as such is not defined in the Etymologies. A much later source states that KWATA is a verbal stem expanded from a simpler stem KWA, that "evidently" referred to completion (WJ:392). - A few examples suggest that at one stage in the period designated as Old Sindarin, panta may have appeared as *pantha; see thintha.
parkha "dry". The stem PÁRAK (LR:380) is not defined, but seems to mean the same. This stem could have been an extended form of a simpler stem PAR, but while such a stem is indeed listed in the Etymologies, it can hardly yield a longer stem for "dry"; PAR means "compose, put together". Old Sindarin parkha and Quenya parca would descend from *parkâ, that shows the common adjectival ending -â. Following the liquid r, primitive k becomes kh in Old Sindarin; see alpha.
parma "book". The primitive form is given as parmâ, derived from a stem PAR "compose, put together" (LR:380). The ending -mâ is frequent in the names of implements (see WJ:416, note 33; compare LR:389-390, where takmâ derived from TAK "fix, make fast" is defined as "thing for fixing", yielding words like Quenya tangwa "hasp, clasp" and Sindarin taew "holder, socket, hasp, clasp, stable"). However, a parmâ is a "composed thing" rather than a "thing for composing". The ending -mâ here seems to function simply as a noun-former, which is its normal meaning when it is added to a stem with an adjectival meaning (see pathwa below).
parthóbi [read *parthóbe?] "arrange, compose". Since ten other Old Sindarin words have infinitives in -be (see buióbe for a list), we must conclude that the ending -bi (found here only) is most likely a misreading for -be. This word is derived from a stem PAR "compose, put together" (LR:380); the primitive form would be *partâ- with the common verbal ending -tâ, that in this case adds little to the meaning of the stem. The cluster rt regularly becomes Old Sindarin rth; see alpha. By itself, *partâ would appear as *partha, but when the infinitive ending -be is present, -â is not final and regularly becomes ó instead.
pathwa "level space, sward". Derived from a stem PATH (LR:380), itself undefined, but the words derived from it suggest a basic meaning "smooth, level". The primitive form is given as pathmâ. The ending -mâ is the same as in parmâ (see parma above). Here it is simply a noun-former, as in *kormâ "round thing" > Quenya corma "ring" (compare the stem KOR "round" in the Etymologies, LR:365, though these derived words are not mentioned there). Pathmâ is etymologically simply *"something smooth", being used of a level space or sward. Notice the shift thm > thw in Old Sindarin; this is our only explicit attestation of it.
pattha "smooth". Derived from the same stem PATH *"smooth, level" (LR:380) as pathwa above; the primitive form is given as pathnâ with the frequent adjectival ending -nâ. This is our sole explicit example of the shift thn > Old Sindarin tth.
pele pl. pelesi, later pelehi "fenced field", Old English tún. (Pelehi was misread as "peleki" by the transcriber of the Etymologies: LR:380 s.v. PEL(ES). Earlier pelesi could not possibly turn into "peleki", but for the change of intervocalic s to h, cf. for instance baraha [q.v.] from barasa. Pelesi > pelehi is therefore wholly plausible. Another, similar misreading turned khelesa into "khelelia"; see khele.) The stem, given as PEL(ES), would seem to indicate a basic, shorter stem PEL with an extended form PELES (with ómataina, or reduplicated stem-vowel, and the ending -S). PEL(ES) itself is not defined, but the shorter stem PEL, listed above it, could well be the basic form here. PEL means "revolve on fixed point"; Quenya pel- means "go round". It could be that the extended form PELES develops the basic idea of "go round" towards "encircle" and hence "fence in". There must have been a primitive form *peles (possibly with a short final vowel that was later lost). *Peles becomes Old Sindarin pele, the final consonant being lost. In the plural, shielded by the plural ending so that it was not final, s persisted: *Pelesî > pelesi. Parallel cases are nele pl. neleki, oro pl. oroti, skhapa pl. skapati. Later, the plural form pelesi became pelehi, since intervocalic s became h (cf. baraha, representing older barasa). For a wholly parallel case, see thele pl. thelesi, later thelehi.
pelthaksa "pivot" (LR:380 s.v. PEL). The primitive form is given as pel-takse. The first element is simply the stem under which this word is listed, PEL "revolve on fixed point". Concerning takse, Tolkien added a cross-reference to TAK "fix, make fast" (LR:389). It may be that the independent primitive form is actually *taksê with a long final -ê, since the long final vowels of primitive Elvish are commonly shortened when they occur as part of the second element of a compound (cf. WJ:403, where it is explained that primitive khînâ "child" appears as -khîna when it is the second element of a compound). An ending -sê is found in a number of primitive words; in some words, it apparently denotes something that is made by the action denoted by the stem: See brasse. A *taksê may then be *"something that is fixed or made fast". A pel-takse is *"a fixed thing that yet has something to do with revolving", *"the fixed thing that something else revolves around", hence a pivot. In Old Sindarin pelthaksa, earlier lt has regularly become lth (see alpha), but the ending -a is surprising. The short final -e of pel-thakse should have been lost already in Common Eldarin. Quenya peltas pl. peltaxi may indeed represent CE *peltaks pl. *peltaksî. Perhaps final -ks was not allowed at the Common Lindarin stage, so that a new ending -â (later OS -a) was added to the word?
persôs "it affects, concerns". Derived from a stem PERES "affect, disturb, alter" (LR:380). Persô- must represent a verbal stem *persâ- (compare phalsâ- from the stem PHÁLAS; see phalsóbe below). The ending -s, "it", is apparently derived from the demonstrative stem S- (LR:385). Here, primitive pronouns su/sû or so/sô "he" and si/sî or se/sê are mentioned. The later Sindarin pronoun ha "it" may suggest that there was also a primitive pronoun *sa, *sâ "it". Compare the Quenya pronominal ending -s "it" (as in tiruvantes "they will keep it", UT:305). Persôs may be derived from *persâ-sa.
phaire "radiance". Derived from a stem PHAY "radiate, send out rays of light" (LR:381). The primitive form is most likely intended to be *phairê; the ending -rê here has about the same meaning as in the word thêrê "look, face, expression" vs. the verbal stem THÊ "look" (LR:392).

phalsóbe "to foam". Derived from a stem PHÁLAS, an extended form of a simpler stem PHAL, itself defined as "foam" (LR:381; it is not clear whether this is "foam" as a noun or as a verb "to foam"; perhaps it covers both, since both verbs and nouns are derived from this stem). Phalsó- must represent an old verbal form *phalsâ- that by itself would appear as *phalsa, but when it has the infinitive ending -be (see buióbe), the original old ending -â becomes ó instead, since it is not then final. - Phalsóbe yielded Sindarin faltho; this word thus demonstrates that ls became lth only at a later stage.
pharasse (also spharasse) "hunt, hunting" (LR:387 s.v. SPAR). See spharasse.
pharóbe (also spharóbe) "hunt" (verb) (LR:387 s.v. SPAR). See spharóbe.
phasta "shaggy hair". Derived from the stem PHAS (LR:381), not itself defined. The Quenya verb fasta- "tangle" may reflect the most basic form here, a verb *phastâ- with the frequent verbal ending -tâ. This would become phasta in Old Sindarin, but now used as a noun *"something that is tangled" = "shaggy hair" rather than a verb "tangle". For other possible cases of original verbs being used as nouns, see hwesta, rista.
phauka "thirsty". Derived from a stem PHAU "gape" (LR:381). The primitive form would be *phaukâ with the adjectival ending -kâ (as in poikâ, see ruska - compare also túka). Quenya fauca is defined as "open-mouthed" as well as "thirsty, parched"; since the stem means "gape", "open-mouthed" or *"gaping" is probably the most literal meaning of *phaukâ.
Phayanôr (indicated to be accented on the final syllable) masc. name., "Fëanor" (LR:381 s.v. PHAY). The name Fëanor was ever present in Tolkien's mythos, but he repeatedly reinterpreted its meaning. In his early Gnomish Lexicon (1917), it was interpreted "Goblet-smith" (LT1:253). In the Etymologies, the name is said to mean "radiant sun": The initial element phay- is identical to the stem PHAY "radiate, send out rays of light", under which the name is listed. Anôr is just another way of spelling anór, "sun" (in later Sindarin Anaur, in turn becoming Anor: cf. SD:306). Anôr is derived from the stem ANÁR (LR:348, where the primitive form anâr- is listed; the hyphen suggesting that some final vowel is not included - the full form may here be intended to have been anâro, since Phayanôr is derived from Phay-anâro in LR:381 s.v. PHAY; on the other hand, SD:306 quotes Anâr with no final vowel as the "oldest form" of the word for Sun). The stem ANÁR itself is identified as a "derivative of NAR1", the Elvish stem for "flame, fire" (LR:374). The stem vowel is prefixed to form an intensive variant of this stem, the Sun being the ultimate "flame". However, Tolkien abandoned this etymology in later sources. His final interpretation of the name Fëanor is that it means "Spirit of Fire" (MR:206, footnote). He came to think of Fëanor as a hybrid of pure Quenya Fëanáro and pure Sindarin Faenor. In Quenya Fëanáro, the elements are fëa "spirit" and nár "fire", plus a (probably masculine) ending -o. Fëa "spirit" comes from a stem phaya of similar meaning (PM:352; in early Quenya fëa was evidently faya, since a form Fayanáro is mentioned in PM:343). This does not obsolete the stem PHAY "radiate, send out rays of light" in the Etymologies (LR:381), for Tolkien also stated that "the ancient significance [of fëa and fairë, later used for 'soul'] seems to be rather 'radiance'" (MR:250; the word fairë "radiance" is also listed in the Etymologies, though there is not there any hint that it can also mean "soul"). The second element of Fëanáro is nár "fire"; the Etymologies lists nár and náre, derived from the stem NAR1 "flame, fire" (LR:374). As observed above, the final -o is probably to be understood as a masculine ending, though conceivably it could also be the Quenya genitive ending -o (fëa náro = literally "spirit of fire"). However, it seems that Quenya rather uses the possessive-adjectival case in -va to describe what something consists of, so this is a more doubtful interpretation. - The Old Sindarin form of Fëanáro, had one existed, would presumably have been *Phayanóro - not so different from the form Phayanôr found in the Etymologies, for though Tolkien later revised the interpretation of the name, the primitive stems that are involved remained the same. But of course, the name Phayanôr was not intended to be Old Sindarin when it was written; it was "Old Noldorin", the language the Noldor spoke in Valinor. When Tolkien revised the history of the Welsh-sounding language of his mythos, making it the tongue of the Sindar of Middle-earth instead of the tongue of the Noldor of Aman, he incidentally eliminated the possibility that a form of the name Fëanáro (known only from the Blessed Realm) could have occurred in the older form of that language. The Noldor, adapting the name Fëanáro to Sindarin, did not do their homework; they "should" have reconstructed the Common Eldarin form *Phayanârô, Old Sindarin *Phayanóro, and derived the "true" Sindarin form Faenor from these earlier forms. Instead, as mentioned above, the Quenya-Sindarin hybrid Fëanor arose. (This story is not found in the Etymologies; here Feanoúr, Féanor is genuine "Noldorin". In Tolkien's later vision of Sindarin phonology, it would probably be impossible to derive such a form, but - perhaps to keep the long-established name Fëanor - Tolkien adopted the explanation that it was not pure Sindarin after all.)
phelga "cave". Derived from a stem PHÉLEG (LR:381) that is similarly defined. Phelga and its cognates (Quenya felya and Telerin felga) point to a primitive form *phelgâ. The ending -â, often adjectival, seems to indicate simply an inanimate noun here. - Later developments in Tolkien's vision of the languages of Arda throw considerable doubt upon the entire entry PHÉLEG and the Elvish words derived from it. This entry was meant to explain the first element in the name Felagund, which in the Etymologies is implied to mean *"Cave-prince" (compare LR:366 s.v. KUNDÛ); Felagund was held to be an Elvish name at this stage. Later, Tolkien explained Felagund as a Sindarin adaptation of Khuzdul (Dwarvish) Felakgundu, Felaggundu "cave-heaver" instead (PM:352; see also the Silmarillion Index, entry Felagund - here it is pointed out that the translation "Lord of Caves" occurs in the Silmarillion text itself, but this seems to reflect Tolkien's earlier interpretation of Felagund). In Dwarvish Felakgundu, it is gundu, not felak-, that means "cave" or "underground hall" (PM:352).
phenda "threshold". The stem PHEN (LR:381) is not defined; the primitive form must be *phendâ with medial fortification n > nd and the ending -â, here indicating an inanimate noun.
pheren "beech". Derived directly from a stem PHÉREN (LR:381), the extended form of a simpler stem PHER, formed by adding -n to ómataina (reduplicated stem-vowel). The stem PHER (PHÉREN) is simply defined as "beech". Compare LR:352 s.v. BERÉTH, where it is stated that "the beech-tree was probably originally called *phéren"; the Old Sindarin word pheren is mentioned here as well.
pherna "mast" (beechnuts). Perhaps originally an adjective *phernâ, derived from the stem PHER "beech" (LR:381) with the frequent adjectival ending -nâ. Later this adjective was perhaps used as a noun "something having to do with (something coming from) beeches" and applied to beechnuts.
Phind-ambar masc. name. This name is not explicitly said to be Old Sindarin, but it is mentioned as the ancestral form of the later Sindarin name Findabar (LR:381 s.v. PHIN), so we include it here. The meaning of this name is rather obscure; it seems to be a compound of phinde "skill" (q.v.) and ambar "world"; see Gondambar for a discussion of this element. Since Gondambar is defined as "Stone of the World", perhaps Phind-ambar similarly means *"Skill of the World", sc. the most skillful one in the world?
phinde "skill". Derived from a stem PHIN "nimbleness, skill" (LR:381); the primitive form is probably meant to be *phindê with medial fortification n > nd and the abstract ending -ê. This produces a clash with Tolkien's later scenario, as set out in PM:362: Primitive phindê is now redefined to mean "tress", while in the Etymologies, the primitive word for "tress, braid of hair" had been spindê (LR:387 s.v. SPIN; Old Sindarin sphinde). See Phinwe.
*Phindekóno possible correction of Findekâno, q.v.
Phinwe masc. name, "name of chief Gnome [= Noldo]", Quenya Finwë. In the Etymologies, this word was derived from a stem PHIN "nimbleness, skill" (LR:381). According to LR:398 s.v. WEG, the ending -we seen in this name descends from the abstract suffix -wê (concerning which see yaiwe). It would seem that when the Etymologies was written, Tolkien intended the name Phinwe to mean "Skill". However, in a much later source he stated that the name Finwë did not necessarily have any meaning at all, "that is any intentional reference to or connexion with other stems already existing in primitive Eldarin" (PM:340). In the same source, the ending -we is also reinterpreted: No longer being explained as an abstract suffix, it is said to be an old word for "person", derived from a stem EWE (not in the Etymologies). In PM:344, however, it is suggested that the name Finwë was at least associated with words for "hair" (like Quenya findë), though Tolkien made it clear that this should not be seen as "conclusive proof" that the name Finwë was actually related to such words. This refers to Tolkien's revised ideas about the meaning of the stem PHIN, now having to do with hair instead of skill - see phinde above.
phinya "skilled". Derived from the same stem PHIN "nimbleness" (LR:380) as phinde "skill". The primitive form would be *phinjâ with the well-attested adjectival ending -jâ (the origin of the -ya of the name Quenya = "Elvish, Quendian", primitive kwendjâ, WJ:360, 393). - Notice that primitive *nj persists in the Old Sindarin word (spelt ny). This demonstrates that there is a distinction between nj = palatalized n and nj = n followed by j (as in this case, since phinya = phin- + jâ). Palatalized sounds were depalatalized already at the Common Lindarin stage (cf. kelepe from kjelepê), so if the nj of *phinjâ had been a palatalized n instead of a cluster n + j, the Old Sindarin form would have been **phina.
phuine "night". Derived from a stem PHUY (LR:382), not itself defined, but the Quenya cognate fuinë, huinë means "deep shadow" rather than "night" (while Quenya words for "night" are derived from the stem PHUY with no additions: Fui, Hui). Phuine (and Q fuinë, huinë) must come from a primitive form *phuinê or conceivably *phuini. However, no primitive ending -ni is mentioned in Tolkien's papers, while an ending -nê is does appear in a few of his "reconstructed" primitive forms. In neinê "tear" (from a stem NEI of similar meaning, LR:376) it seems to be simply a noun-former. Another noun in -nê is slignê "cobweb" (LR:386 s.v. SLIG). The word ornê "(slender) tree" may provide the best clue to the (or a) basic meaning of -nê: Since ornê is stated to be related to the adjective ornâ "uprising, tall" (UT:266), we can speculate that -nê is a nominal ending corresponding to the common adjectival ending -nâ, an ornê being literally a "tall thing", used with reference to slender trees. Similarly, there could be a primitive adjective *phuinâ "dark" or "shadowy", with a corresponding noun *phuinê producing Old Sindarin phuine "night" and Quenya fuinë, huinë "deep shadow".
phuióbe "feel disgust at, abhor". A formation wholly parallel to buióbe: The stem PHEW of similar meaning (LR:381) would have yielded a primitive verb *pheujâ- with the common verbal ending -jâ, that here adds nothing to the meaning of the stem itself. As in buióbe, primitive euj here becomes ui (though eu by itself becomes io or iu; see bioro, biuro). *Pheujâ- by itself would normally appear as *phuia-, but when the infinitive ending -be is present, â is not final and regularly becomes ó, hence phuióbe.
pikina "tiny". Derived from a stem PIK that is not defined in the Etymologies (LR:382), but it seems to have to do with smallness: Quenya pitya "little" (as in Pityafinwë "Little Finwë", PM:353) probably represents primitive *pikjâ, sc. PIK with an adjectival ending (cf. also Pitya-naucor or Picinaucor for "petty-dwarves", WJ:389; the prefix form pici- confirms that the stem is PIK). However, pikina cannot represent *pikjâ; this Old Sindarin word must come from *pikinâ with another adjectival ending, concerning which see malina.
póre "fist". The stem is KWAR, that in the Etymologies is defined as "clutching hand, fist" (LR:366). Póre, as well as Quenya quárë, points to a primitive word kwârê or less probably kwâri. (In the Etymologies, Quenya quárë was changed from quár, but quár reappears in a late source in PM:318.) The form póre demonstrates both the Common Lindarin change kw > p (see alpha, panta) and the Old Sindarin change of non-final â to ó (see abóro). However, in PM:318, the stem KWAR is defined "press together, squeeze, wring" and the primitive form is given as kwâra (the ending -a is probably simply the stem-vowel suffixed). This would have yielded Old Sindarin *pór instead (cf. Telerin pár), since final short -a was lost at the Common Eldarin stage. Both póre and *pór would become paur (-bor) in Sindarin, though.

poto "animal's foot", pl. poti (LR:382 s.v. POTÔ). The entry-head in the Etymologies, POTÔ, is clearly a complete reconstructed word in itself, representing a shorter stem *POT. There is not much to say about a primitive form potô, which Tolkien did not further explain in any way; the ending -ô is normally masculine or agentive, but here it may be simply the stem-vowel suffixed and lengthened, or just a noun-former (as in panô; see pano). Potô regularly yields Old Sindarin poto, but the plural poti is interesting. The primitive plural *potôi would probably appear as potoi at the stage immediately preceding Old Sindarin, but it seems that the final diphthongs (except -ui) were simplified at this stage, so that nouns ending in a vowel formed their plurals by dropping this vowel before the plural ending -i was added. Hence poto pl. poti instead of *potui for potoi (oi became ui, cf. muina, Uigolosse). However, one Old Sindarin noun, pano "plank", surprisingly has the pl. panui (instead of *pani) and hence does not follow this pattern. See pano.
puióbe "spit" (verb). Derived from a stem PIW, also defined "spit" (LR:382). The primitive verb was evidently *piujâ- (compare tuio, q.v., from the stem TIW, where the Quenya cognate tiuya- clearly points to a primitive form *tiujâ-). The ending -jâ is a common verbal ending (but Quenya piuta, also meaning "spit", must represent *piutâ- with another ending). *Piuj- was probably *piui- at some stage, but the "triphthong" iui was simplified to ui in Old Sindarin (cf. euj- also becoming ui; it is indeed probable that euj- became iuj, iui at some stage, then being simplified to ui; see buióbe for a possible example). *Piujâ- by itself would appear in Old Sindarin as *puia-, but when the infinitive ending -be (concerning which see buióbe) is present, earlier â was not final and regularly became ó, hence producing the form puióbe.
ragme "fathom", noun. Derived from a stem RAK "stretch out, reach" (LR:382); the primitive form is given as rakmê with an ending that forms abstract nouns. Compare some forms mentioned in WJ:416: julmê "drinking, carousal", from the stem JULU "drink" (WJ:416) or labmê "the action of *LABA", sc. a stem having to do with licking or moving the tongue (WJ:416). So rakmê is literally a verbal noun "reaching", later being used in the sense "fathom". - K becomes voiced g by assimilation to the following m: indeed all unvoiced plosives become voiced before a nasal. Compare ndagno, tulugme, yadme.
ragna "crooked". Primitive form given as ragnâ, sc. the undefined stem RAG (LR:382) with an ending -nâ that is either adjectival or forms past participles (see khalla). This may suggest that the stem RAG has a verbal meaning *"crook, bend".
randa "cycle, age" (100 Valian Years). Derived from a stem RAD "back, return" (LR:382). The primitive form randâ shows nasal infixion and an ending -â that is here used to form a noun; for another example of a similar derivation, compare rundâ from RUD (see runda). Given the meaning of the stem, it seems that randa literally refers to something that returns or comes back, hence a time-cycle (and therefore also used to mean age).
ranko pl. rankui "arm". Derived from the same stem RAK "stretch out, reach" as ragme above; an "arm" is thus conceived as "that which stretches/reaches out". The primitive form is given as ranku. The ending is interesting; several primitive words denoting body-parts show the ending -u: besides ranku we have mbundu "snout, nose" (LR:372 s.v. MBUD) and tûgu "muscle, sinew" (LR:394 s.v. TUG; see túgo concerning the latter; the Old Sindarin form of mbundu is not given, but would have been *mbundo). In the cases of tûgu and mbundu, the ending -u could be simply the stem-vowel suffixed, but this cannot be the case in ranku, since the stem-vowel is here a. This word is therefore valuable in establishing that there really is an independent ending -u sometimes occurring in the names of body-parts. Beside this ending, ranku also shows nasal infixion, which seems to make this form wholly parallel to mbundu "snout, nose" (from a stem MBUD "project", hence "that which projects", just like ranku is "that which stretches/reaches out" < stem RAK "streach out, reach"). Ranku becomes Old Sindarin ranko because final short -u became -o at the Common Eldarin stage; the original quality of the vowel is however preserved before the plural ending -i, creating a diphthong ui that did not change. Compare malo pl. malui and orko pl. orkui.
rattha "course, river-bed". Derived from a stem RAT "walk" (LR:383). The primitive form is given as rattâ (or ratta; there is a diacritic indicating that the final vowel can be either long or short - but rattha must derive from rattâ, for final short -a was lost at the Common Eldarin stage). The doubling of the t may be seen simply as a medial fortification of some kind; the final -â may have a local meaning here, denoting a place where something (namely water) "walks" or runs. Compare the synonym yura (*jurâ), derived from a stem meaning "run" by means of the same ending. - Older tt becomes tth in Old Sindarin; cf. batthô from battâ.
rauda "hollow, cavernous". Derived from a stem ROD "cave" (LR:384). This word must be derived from a Common Eldarin form *raodâ with the adjectival ending -â added to an A-infixed variant of the stem ROD. This seems to be our sole explicit example of Common Eldarin *ao yielding Old Sindarin au. In VT39:10, Tolkien suggests that ao became long â in the Telerin (Lindarin) branch of the Elvish language family. If so, the Old Sindarin form of *raodâ should perhaps be *róda instead of rauda, since earlier â became OS ó, but in Classical Sindarin, both rauda and *róda would become raudh anyway. (Compare VT39:10, where Tolkien derives Sindarin caul "great burden" from kâlô, itself representing an older, A-infixed form of the stem KOL, evidently meant to be *kaolâ. In this scheme, the Old Sindarin form intermediate between kâlô and caul would be *kólo; in Tolkien's earlier concept as exemplified in the Etymologies, the form would be *kaulo. But as already mentioned, the discrepancy has no effect on the "final" Sindarin form of the word.)
rauta "metal" (meaning changed by Tolkien from "copper"). The "stem" given in the Etymologies, RAUTÂ (LR:383), seems to be a complete reconstructed word. However, it cannot be further analyzed with any certainty. It could, among several possibilities, represent an A-infixed variant of a simpler stem *RUT.
ríge "crown". Derived from a stem RIG that is not defined in the Etymologies (LR:382); however, PM:347 informs us that RIG means "twine, wreathe". In the Etymologies, the primitive form of ríge is given as rîgê, which may be understood as an original abstract formation "twining, wreathing" later being applied to a concrete wreath and hence a "crown". (See síre for a possible example of a parallel development.) However, PM:347 gives the primitive form as rîgâ (defined as "wreath, garland"). This -â would be simply a nominal ending. Both rîgê and rîgâ would yield Sindarin rî (= rhî in the "Noldorin" of the Etymologies), but the Old Sindarin form of rîgâ would be *ríga.
rimba "frequent, numerous". Derived from an undefined stem RIM (LR:383); the primitive form is given as rimbâ, with medial fortification m > mb and adjectival -â.
rimbe "crowd, host". Derived from the same undefined stem RIM as rimba above; the primitive form is given as rimbê, with medial fortification m > mb and the nominal ending -ê, that can have many meanings. It occurs in a number of words denoting substances (see kelepe for some examples) and since a "crowd" or "host" is people considered as a mass rather than as individuals, this meaning may be relevant.
ringe "cold". The "stem" or entry-head given in the Etymologies, RINGI (LR:383, simply defined as "cold") seems to be a complete "reconstructed" word in itself. It shows the adjectival ending -i (frequent in colour-adjectives; see barane). Final short -i turned into -e already in Common Eldarin; therefore, Quenya similarly has ringe "cold" (but also ringa with another adjectival ending, though this Q form is not listed in the Etymologies - its Old Sindarin counterpart, if one existed, would also be *ringa).
rista- "rend, rip". Derived from a stem RIS "slash, rip" or "cut, cleave" (LR:384). In the second entry RIS-, a primitive form rista- is given; the very oldest form must have been ristâ- with a long final vowel (a short -a would have been lost already in Common Eldarin, not appearing in Old Sindarin). The ending -tâ is a common verbal suffix; it may be causative but here adds nothing to the meaning of the basic stem. A noun rista is seen in the compound ekla-rista (q.v.); this noun would be derived from the verb. It refers to a cut in the landscape, hence a ghyll or ravine. For another case of a verbal stem in -ta (also) being used as a noun, compare hwesta; cf. also phasta.
ró "lion", pl. rówi. Derived from a stem RAW (LR:383), itself undefined; a primitive form is given as râu, with lengthening of the stem-vowel and a somewhat surprising shift W > U. Since long â became ó in Old Sindarin (see abóro), a form *róu must have arisen, but u was evidently assimilated into the preceding vowel: *róu > *róo > *ró. This does not happen in the plural, where we evidently have *râwî (*râuî?) > rówi.
romba "horn, trumpet" (this word also occurs in Quenya). Derived from a stem ROM "loud noise, horn-blast, etc." (LR:384). The primitive form would be *rombâ, which in relation to the stem ROM shows medial fortification m > mb and the (in this case nominal) ending -â. The meaning of a word derived in this way is not predictable; compare rimba, where an adjective is derived from the stem RIM using exactly the same devices.
róna "east", derived from the stem RÔ "rise" (LR:384), itself a variant of ORO "up; rise; high; etc." (LR:379). This identifies the East as the direction of the rising sun. It is, however, possible that róna was originally "east" as an adjective rather than as a noun; the primitive form must be *rônâ with the adjectival ending -nâ. Since primitive ô normally yields Old Sindarin ú, û (cf. brûna from b'rônâ), we might have expected *rúna instead. Róna yielded Sindarin rhûn, where the shift O > U has indeed occurred, so perhaps we are to assume that róna is early Old Sindarin, later becoming *rúna. Of course, it is also possible that róna is simply a misreading for *rúna in Tolkien's original manuscript. - It may be that Sindarin rhûn cannot be derived from róna after all. In the "Noldorin" of the Etymologies, earlier initial r- becomes unvoiced rh-, so in "Noldorin", róna could easily yield rhûn. However, Tolkien revised the phonology when "Noldorin" became Sindarin; now earlier r- was unchanged in Classical Grey-elven. An Old Sindarin word róna, *rúna should therefore have become **rûn, not rhûn, in Sindarin. Rhûn would now require an earlier form *sróna, *srúna. It may be that Tolkien, when he revised "Noldorin"/Sindarin phonology, for some reason wanted to keep rhûn as the Sindarin word for "east" (instead of emending it to rûn), and devised some kind of historical explanation for this.
rostóbe "to hollow out, excavate". Derived from a stem ROD, itself defined as "cave" (LR:384). The primitive form would be *rodtâ- with the common, sometimes causative verbal ending -tâ (concerning which see bértha-); hence *rodtâ- has to do with "causing" a cave or hole to be, or simply "excavate". This word alone explicitly demonstrates that primitive dt became st in Old Sindarin (but cf. also Sindarin hast from primitive sjad-ta [LR:389 s.v. SYAD]; the Old Sindarin form is not given, but it would have been *hyasta). *Rodtâ- by itself would appear in Old Sindarin as *rosta, but when the infinitive ending -be (concerning which see buióbe) is present, earlier â was not final and regularly became ó, hence producing the form rostóbe.
rúma "loud sound, trumpet-sound". Derived from a stem ROM "loud noise, horn-blast, etc." (LR:384). This, as well as the Quenya cognate róma, points to a primitive form *rômâ with lengthening of the stem-vowel and the ending -â, here a noun-former (more commonly used to derive adjectives). For the change ô > û, compare brûna from b'rônâ.
runda "club". Derived from a stem RUD (LR:384), itself undefined; the primitive form is given as rundâ, with nasal infixion and the ending -â, here used as a noun-former (cf. rúma above). Runda yielded Sindarin grond; Grond is best known as a name of Morgoth's mace.

ruska "brown". The entry-head RUSKÂ given in the Etymologies (LR:385) seems to be a reconstructed primitive word rather than just a "stem". It shows the frequent adjectival ending -â, or possibly even the longer adjectival ending -kâ (cf. for instance poikâ "clean, pure" from POY, LR:382). However, if the ending -kâ is present, we must postulate a new basic stem *RUS (clearly not the same as RUS "flash, glitter of metal" in LR:384). - Initially, sk becomes skh in Old Sindarin (see skhalia), but it would not seem that this change occurred medially (not **ruskha).
russe "polished metal". Derived from a stem RUS "flash, glitter of metal". The primitive form, most probably *russê, may show a simple ending -ê sometimes occurring in the name of substances (see kelepe) combined with doubling of the medial S as some kind of fortification (cf. such a primitive word as rossê "dew" [Letters:282] derived from a stem ROS [LR:384, where primitive rossê is not mentioned, but its Quenya descendant rossë is). If a longer ending -sê is present, its meaning is obscure. An ending -sê does occur in a few words; under brasse we argue that it may be used to derive words for something that is made by the action denoted by the stem, but in this case, the stem does not have a verbal meaning at all.
salape "herb, green food plant". The "stem" or entry-head given in the Etymologies, SALÁK-(WÊ) (LR.384), seems to be a complete reconstruction of a primitive word *salakwê. Originally, this word apparently meant "grass", which is still the meaning of Quenya salquë and Ilkorin salch. The ending -wê evidently forms a noun here (perhaps it is a nominal counterpart of the adjectival ending -wâ; see comments on katwâ under katwe). As usual, kw becomes p (see alpha). - No information is given about the basic stem-element SALAK; it presumably simply means "grass" (already Tolkien's earliest "Qenya" wordlist had a word salki "grass"; see the Qenya Lexicon p. 84).
salpha "liquid food, soup, broth". Derived from an undefined stem SÁLAP (LR:385); Quenya salpa- "lick up, sup" may provide a clue to its basic meaning. Both this Quenya verb and the Old Sindarin noun probably derives from a form *salpâ. As usual (cf. alpha), p following l becomes ph in Old Sindarin.
síre "river", derived from a stem SIR "flow". Síre (with lengthening of the stem-vowel and the ending -e) is actually an abstract or verbal noun formation, literally *"flowing", but the word was applied to something concrete (a river). Similar formations are frequent in Quenya (which language indeed has sírë with the same meaning). The primitive form would be either *sîrê or *sîri. Compare ríge from rîgê; see also slíwe, representing slîwê.
sirya- "flow", derived from the stem SIR (LR:385) of similar meaning: primitive form surely *sirjâ- with the frequent verbal ending -jâ, that here adds no meaning (in Quenya, the simpler verb sir-, derived from the stem with no additions, was used). In light of the verb skhalia- derived from skaljâ, we might have expected siria- instead; see skhalia-.
skhalia- "veil, conceal, screen from light", derived from the stem SKAL1 "screen, hide (from light)" (LR:386); the primitive form would be *skaljâ- with the frequent verbal ending *-jâ. (The adjective skhalla, see below, comes from an ancient past participle derived directly from the stem.) For other examples of primitive initial sk- becoming Old Sindarin skh-, see skhalla, skhella. The ending *-jâ here comes out as -ia, the semi-vowel *j (= English y) turning into a full vowel i; in the word sirya- (< *sirjâ) above, the semi-vowel persists. We might have expected *siria. Since this verb became sirio "to flow" in Sindarin, the y did eventually turn into a full vowel also in this case. Perhaps *sirjâ-, *skaljâ- became sirya-, *skhalya- in early Old Sindarin, later (but still during the OS period) becoming *siria-, skhalia-.
skhalla "veiled, hidden, shadowed, shady", derived from the stem SKAL1 "screen, hide (from light)" (LR:386); the primitive form is given as skalnâ, with an ending -nâ that sometimes simply forms adjectives (compare magnâ), but often it also functions as a past participle ending. In this case, skalnâ literally means "hidden", the past participle of the verbal stem "hide" (but probably not recognizable as a past participle at the Old Sindarin stage). For other examples of the assimilation *ln > ll, see khalla (< khalnâ) and skhella (< skelnâ).
skhapa pl. skhapati "shore". Derived from a stem SKYAP (LR:386), itself undefined. The primitive form is given as skjapat- (skyapat-), the final hyphen evidently indicating that there was originally some final vowel that cannot be reconstructed (the full word may have been *skjapata, or possibly *skjapate or *skjapato; it could not be **skjapatu or **skjapati, since this would have yielded Old Sindarin **skhapato or **skhapate, respectively). The final vowel (*-a, *-e or *-o) was lost already in Common Eldarin, producing a form skjapat. It has been suggested that t is here some kind of dual marker (compare the Quenya dual ending -t), skjapat supposedly referring to the two shores of a river. However, we may just as well assume a primitive expanded stem *SKYAPAT (with a T suffixed to *SKYAPA, sc. the stem SKYAP with ómataina, reduplication of the stem-vowel). Compare such variants as ESE > ESET (LR:356), or ORO > ÓROT (LR:379). - The original initial cluster skj- was simplified to sk- at the Common Lindarin stage, when the original palatalized consonants were depalatalized (see kelepe). In Old Sindarin, sk- regularly becomes skh-; compare skhalia, skhalla above and skhella below. From the Old Sindarin form alone it is thus impossible to tell whether the original primitive word had sk- or skj-. - In OS, the final consonant of skjapat is lost in the singular (skhapa), but t persists in the plural form, since it was there "shielded" by the plural ending -i (primitive -î) and hence was not final. Compare nele pl. neleki, oro pl. oroti, pele pl. pelehi, thele pl. thelehi.
skhella "naked", derived from the undefined stem SKEL (LR:386); the primitive form is given as skelnâ, with an ending -nâ that sometimes simply forms adjectives (compare magnâ), but often it also functions as a past participle ending. If the undefined stem SKEL is assigned a verbal meaning *"strip naked, denude", skelnâ can be understood as a past participle (but on the Old Sindarin stage it would probably be felt to be an adjective). For other examples of the assimilation *ln > ll, see khalla (< khalnâ) and skhalla (< skalnâ).
slaiwa, later thlaiwa, "sickly, sick, ill", derived from a stem SLIW "sickly" (LR:386). The primitive form is given as slaiwâ. The adjectival ending -â is very common. The A-infiction turning the stem SLIW into slaiw- is less usual, but there are a number of parallel formations, like thausâ "foul, evil-smelling, putrid" from the stem THUS (LR:393) and taurâ "masterful, mighty" from TUR (LR:395); outside the Etymologies we also have maikâ "sharp" from a stem MIK, see WJ:337, and naukâ *"stunted" from NUKU, WJ:413. - As we see, slaiwa has an alternative (apparently later) form thlaiwa; compare slíwe/thlíwe below. In the Etymologies, Tolkien's scenario was that initial sl first became thl-, in later "Noldorin" also fl-: Slaiwa, via thlaiwa, became "Noldorin" thlaew (or thloew - Tolkien kept changing his mind back and forth concerning whether ai yielded ae or oe), and in later "Noldorin", thlaew in turn became flaew. Slíwe "sickness" likewise became thliw (via thlíwe), and thliw eventually became fliw. In Sindarin as it later emerged in Tolkien's notes, there were certain revisions: Earlier sl- now yields Sindarin lh- (unvoiced l), so the Sindarin descendants of slaiwa and slíwe would be *lhaew and *lhiw (or better *lhîw) rather than flaew and fliw as in "Noldorin". If we want to keep as much as possible of the Etymologies material, we may choose to assume that Old Sindarin sl first became thl, and that thl in turn yielded Sindarin lh.
slíwe, later thlíwe, "sickness". Derived from the same stem SLIW "sickly" (LR:386) as slaiwa above, but without A-infixion. The primitive form is given as slîwê, with lengthening of the stem-vowel and the abstract ending -ê. If we dare to assign a verbal meaning to the stem SLIW, namely *"be sick", slîwê could be seen as a kind of verbal noun with several parallels: See síre. - Concerning the variation sl/thl, see slaiwa above.
sniuma, also snýma, "snare", or "noose" if the word means the same as the Sindarin word it yielded. Tolkien first wrote núma, then changed it to sniuma, snýma (the wording in Christopher Tolkien's bracketed note in LR:387 s.v. SNEW is not entirely clear, but sniuma, snýma are clearly meant to be "ON" forms, not "N"). The stem SNEW means "entangle"; sniuma is clearly to be referred to a form primitive form *sneumâ (compare Quenya neuma). The ending -mâ is often used to derive words for implements; see parma. A sneumâ is thus a "thing for entanglement". - The diphthong eu here becomes iu; compare biuro (or bioro) from beurô. It would seem that sniuma later became snýma, indicating a further development iu > ý; in later Sindarin, we would indeed expect sniuma to end up as *nýw (and not hniof, as in the "Noldorin" of the Etymologies). The rejected Old Sindarin form núma would seem to indicate that eu became ú instead, but there are no clear parallels to such a development elsewhere; perhaps this was simply a mistake, soon corrected. The form núma would also indicate that initial sn was simplified to n already at this stage, but in sniuma earlier initial sn persists, and the latter was evidently Tolkien's final decision. (Later Sindarin only has n anyway; one revision Tolkien did in the "Noldorin" of the Etymologies - that earlier sn, sm became unvoiced hn, hm - was undone later. See Christopher Tolkien's note in LR:387, indicating that even in Etym itself, this revision was not carried through consistently.)
sóba "juice", derived from the undefined stem SAB (LR:385). The Quenya cognate sáva points to a primitive form *sâbâ. As usual, long non-final â becomes ó in Old Sindarin (see abôro). The ending -â is often adjectival, but is also quite common in nouns denoting inanimate objects (not quite so common in nouns denoting substances, like "juice").
sphanga "beard". Derived from a stem SPÁNAG (LR:387), not itself defined: primitive spangâ. The ending -â could be simply the stem-vowel suffixed and lengthened, or it may be an independent ending -â used to form an inanimate noun. Just like sk becomes skh- (see skhalia), sp becomes sph. Other examples are listed below: spharasse, spharóbe, sphinde, sphíndele. The alternative forms of spharasse and spharóbe, namely pharasse and pharóbe, seem to indicate that s before ph was lost during the Old Sindarin stage, so perhaps sphanga "beard" later appeared as *phanga. The Sindarin form was in any case fang.
spharasse (also pharasse) "hunt, hunting" (concerning the variation sph- vs. ph-, see sphanga above). Derived from a stem SPAR "hunt, pursue" (LR:387). The primitive form would be *sparassê, sc. the stem SPAR + an ending *-assê having various functions. The Old Sindarin form -asse is explicitly attested in this word only, but the later Sindarin ending -as (spharasse > faras) is found in collective nouns, verbal nouns and nouns derived from adjectives. This is obviously a verbal noun. Compare for instance Sindarin galas "growth" (Old Sindarin *galasse) vs. the verb galo "to grow" (OS *galóbe); see LR:357 s.v. GALA. - It is possible that the ending *-assê is somehow connected to the Quenya locative ending -ssë, though the precise relationship between them is obscure.
spharóbe (also pharóbe) "hunt" (verb). Concerning the variation sph- vs. ph-, see sphanga. Derived from a stem SPAR "hunt, pursue" (LR:387). The form naróbe shows the Old Sindarin infinitive ending -be (see buiobe). Spharó- would seem to represent a primitive verbal stem *sparâ-, sc. the stem SPAR with a suffixed -â that is apparently a verbal ending here; such formations are quite rare (see naróbe for another example). The stem SPAR would much rather be expected to function as a "basic" verbal stem, the verb simply being **sphar- (aorist **sphare), with the infinitive **spharie instead of spharóbe.
sphinde "lock of hair". Derived from a stem SPIN (LR:387), itself undefined, but the primitive form spindê is defined as "tress, braid of hair". Spindê is formed from SPIN with medial fortification n > nd and an ending -ê (that here seems to be neither abstract not feminine, its usual functions; it appears to be simply a noun-former). However, Tolkien in a source much younger than the Etymologies derived Eldarin words for "tress" from phindê instead of spindê (PM:362). Phindê would become *phinde, not sphinde, in Old Sindarin - but this in turn clashes with a word for "skill" listed in the Etymologies. See phinde.
sphíndele "(braided) hair" (LR:387 s.v. SPIN). The accent over the i probably indicates simply that this is the accented syllable, not that the vowel is long. This word is sphinde "lock of hair" or *"tress" (see above) with the ending -le, itself undoubtedly the Old Sindarin (and Quenya) form of the ancient abstract ending -lê (VT39:16). Now "(braided) hair" is of course not an abstract as such, but we may think of sphíndele as meaning literally something like *"tressing", used with reference to hair arranged in braids.
stabne, sthamne "room, chamber". Derived from the stem STAB (LR:387), itself undefined. The words derived from it suggest that the basic meaning may have something do to with wood. Two primitive forms are suggested: stabnê and stambê, one showing the independent ending -nê, the other nasal infixion and the shorter ending -ê. The endings -nê, -ê seem to function simply as noun-formers here; while they could be feminine endings, this is clearly not their function here. It would seem that both stabne and sthamne descend from stabnê, not stambê: In Old Sindarin, primitive mb is unchanged, cf. for instance rimba from rimbâ (LR:383 s.v. RIM), so stambê would have yielded OS stambe > sthambe. It seems, then, that stabnê first yielded OS stabne (the only change being the normal shortening of the final vowel), but later the b was assimilated to the following n, becoming m. Another OS development turned initial st into sth. (Compare the words listed below: sthabro/sthabrondo, sthalga, sthanka, stharna, sthina, all derived from stems in ST-; the example stabne suggests that the oldest OS forms preserved the original cluster st- unchanged: *stabro etc.) The change st > sth does not seem to occur medially, though; cf. words like hwesta, nestak-, rista (not **hwestha, **nesthak-, **ristha).

sthabro, sthabrondo "carpenter, wright, builder". Derived from the same stem STAB (LR:387) as stabne, sthamne above. If STAB basically has to do with wood, as we suggested above, a sthabro is basically one who makes wooden constructions. The primitive form would be *stabrô with the agental/masculine ending -rô (WJ:371; compare bioro); the most literal meaning may be something like *"woodman", hence a carpenter. In the longer form sthabrondo, the ending -ro has been expanded by adding yet another masculine ending after it: -ndo, primitive -ndô: Compare a primitive word like la(n)sro-ndo (read perhaps *-ndô) "listener" (LR:368 s.v. LAS2), where Tolkien's hyphen indicates that two distinct endings are present. The ending -ndô may be a nasalized version of the ending -dô seen in the primitive form of the word Noldo, ñgolodô (Old Sindarin ngolodo, q.v.) Because the early Lindar so often expanded the agentive ending -rô, -ro to -ro-ndô, primitive words in -rô often end in -ron in later Sindarin. Sthabrondo itself yielded Sindarin thavron (while the shorter form sthabro would have yielded *thavr, later *thavor, but these forms do not appear).
sthalga "stalwart, steady, firm". Derived from the stem STÁLAG (LR:388), itself undefined. The primitive form is given as stalga, but the very oldest form must have been *stalgâ with the normal adjectival ending -â. (No Primitive Quendian short final -a made it into Old Sindarin or any other Eldarin tongue; such vowels disappeared already in Common Eldarin.)
sthamne (< stabne) "room, chamber"; see stabne.
sthanka "cleft, split". Derived from a stem STAK "split, insert" (LR:388). Two primitive forms are here suggested: stankâ and staknâ. The latter would have yielded Old Sindarin *sthagna instead, so we must go for stankâ; perhaps Tolkien wanted to indicate that a primitive form staknâ was very early metathesized to stankâ (not a regular development). It is probable that the glosses "cleft" and "split" are to be understood in an adjectival/participial sense, not as nouns "a cleft, a split". In that case, staknâ is clearly to be understood as a past participle derived from STAK with the frequent ending -nâ (see khalla). If stankâ is not simply a metathesized variant of staknâ, it could be seen as an adjective (meaning much the same as the past participle) derived by means of nasal infixion and the common adjectival ending -â. (Compare sthinta.)
stharna "sapless, stiff, rigid, withered". Derived from the stem STAR "stiff" (LR.388); the primitive form would be *starnâ with the frequent adjectival ending -nâ. Here, the ending does not modify the meaning of the stem, that is already adjectival.
sthinta "short". The entry-head STINTÂ in the Etymologies (LR:388), itself defined as "short", seems to be a complete word in itself. It could be a nasal-infixed variant of a simpler stem *STIT- with the adjectival ending -â. - A few examples suggest that at one stage in the period designated as Old Sindarin, sthinta may have appeared as *sthintha; see thintha.
sulkha "root" (especially as edible). The stem SÚLUK not defined. The primitive form would be sulkâ (cf. Quenya sulca); here -â functions as a noun-former again. Following l, earlier k turns into kh; see alpha.
taika "boundary, limit, boundary line". This form, mentioned in WJ:309 as an older form of the Sindarin element taeg in Taeglin (= Teiglin), is not explicitly identified as Old Sindarin. However, the most primitive form would probably have a long final -â, so we include taika here. Unlike most of the words from the Etymologies, this form is asterisked as undefined (but then Tolkien thought of this as the ancestor of a Sindarin word from the beginning, while the "Old Noldorin" words from the Etymologies were originally imagined to have been spoken in the Blessed Realm, where writing was invented at an early point). The stem is said to be taya "mark, boundary line", with an alternative form tayak that is probably to be understood as an extension. The primitive form from which taika derives would probably be *taikâ (for *taykâ), with -â used as a noun-former (or conceivably it is just the stem-vowel suffixed and lengthened).
tára "lofty". The primitive form is given as târâ, derived from a stem TÂ/TA3 "high, lofty, noble" (LR:389) with the adjectival ending -râ (for the latter, compare a primitive word like wa3râ "soiled, dirty" from the stem WA3- "soil, stain", LR:397; see wóra). The oldest form of târâ may have been *ta3râ; later the back spirant 3 was lost and the preceding vowel was lenghtened to compensate for this loss. - It is surprising that târâ did not yield OS *tóra, since non-final â normally becomes ó (see abóro). It may be that tára is a misreading for *tóra in Tolkien's original manuscript: The wording "Q tára, ON tára" found in LR:389 seems rather redundant; Tolkien could simply have written "Q, ON tára" instead. It seems likely, then, that he actually wrote or at least intended to write "Q tára, ON tóra". Compare the derivation of wóra (q.v.) from wa3râ, that would be entirely parallel to *tóra from *ta3râ.
targa "tough, stiff". Derived from the undefined stem TÁRAG (LR:390); the primitive form is given as targâ, with the common adjectival ending -â.
tarsa "trouble", noun. Derived from the undefined stem TARAS (LR:391); the primitive form would be *tarsâ. The ending -â is here used to form a noun (it is more commonly an adjective-former). Abstract nouns like "trouble" rarely show this ending; a more common abstract ending is -ê. Perhaps -â is here simply the stem-vowel suffixed and lengthened. The Sindarin descendant of tarsa (tars, tass) has indeed taken on a slightly less abstract meaning: "task" (but also "labour").
tektha "mark" (isolated from andatektha, q.v.). Derived from the stem TEK- "make a mark, write or draw" (LR:391); the primitive form would be *tektâ. The suffix -tâ is most often a verbal ending, but here it is used to form a noun. This is the sole explicit example demonstrating that primitive *kt became kth in Old Sindarin. The Quenya cognate of tektha, tehta, is also used with reference to diacritical marks (especially the vowel-marks of Fëanorian writing).
thele "sister"; pl. thelesi , later thelehi. Derived from THEL, THELES, also defined as "sister" (LR:392; compare wathel). The basic stem THEL is seen to have an expanded form THELES, formed by adding the ending -S to ómataina (reduplicated stem-vowel, here E). Tolkien compared this to the derivation of toron "brother" from a shorter stem TOR. There must have been a primitive form *theles (possibly with a short final vowel that was later lost). *Theles becomes Old Sindarin thele, the final consonant being lost. In the plural, shielded by the plural ending so that it was not final, s persists: *Thelesî > thelesi. Parallel cases are nele pl. neleki, oro pl. oroti, skhapa pl. skapati. Later, the plural form thelesi became thelehi, since intervocalic s became h (cf. baraha, representing older barasa). For a wholly parallel case, see pele pl. pelesi, later pelehi.
thintha- "fade". Derived from a stem THIN that is not itself defined, but the first primitive word listed in this entry, thindi "pallid, grey, wan", is probably indicative of the basic meaning (LR:392). The Old Sindarin verb thintha- must come from *thintâ- with the common verbal ending -tâ; the literal meaning of this verb is perhaps "become grey". Interestingly, earlier nt here becomes nth (cf. also wintha- from earlier winta-, though wintha was struck out). In the words panta "full", sthinta "short" and wanta "depart, die", no change nt > nth can be observed. (Wanta- is not explicitly identified as Old Sindarin, but the other two are.) Perhaps these words were *pantha, *sthintha, *wantha- at some stage during the Old Sindarin period. To summarize our observations of similar changes throughout this wordlist: The plosives t, p, k are seen to become Old Sindarin th, ph, kh following another t, p, or k (see batthô), following a k (see tektha), following l and r (see alpha), following initial s- (see skhalia), and now also following n, though in the case of most of these rules, there are also words where such changes apparently fail to occur. It is tempting to generalize a rule "t, p, k following any consonant > th, ph, kh in Old Sindarin" - with some extra rules listing the environments where th, ph, kh swiftly reverted to t, p, k (e.g. "medial sth, sph, skh only survive initially, otherwise reverting to st, sp, sk"). The word sthinta "short" interestingly does show the change st > sth, but not the change nt > nth. Perhaps primitive stintâ briefly appeared as Old Sindarin *sthintha, as suggested by thintha, but then became sthinta- because nth reverted to nt medially? If so, the word before us, thintha- "fade", may have become *thinta- later.
thlaiwa < slaiwa "sickly, sick, ill" (SLIW). Possibly not a valid form in Tolkien's later conception of Old Sindarin (as opposed to "Old Noldorin"); see slaiwa.
thlíwe < slíwe "sickness" (SLIW) Possibly not a valid form in Tolkien's later conception of Old Sindarin (as opposed to "Old Noldorin"); see slíwe.
*thoron "eagle"; only the gen.sg. thoronen is attested in Old "Noldorin"/Sindarin (LR:392 s.v. THOR, THORON). However, the genitive ending -(e)n, that also occurs in the Quenya of the Etymologies (e.g. Ar Manwen "Manwë's Day", LR:368 s.v. LEP), is not valid in Tolkien's later vision of Elvish. In Quenya, he changed the genitive ending from -(e)n to -o while writing Namárië for LotR, and in WJ:370, he argues that Sindarin did have the corresponding ending *-ô in "the primitive period" (that must here mean at the Common Lindarin stage: in Old Sindarin, after the shortening of the final vowels, we would have *-o). Therefore, we should probably read *thorono for thoronen. - The word thoron itself represents THORON, an extended form of the stem THOR, formed with ómataina (reduplicated and suffixed stem-vowel, here O) and the ending -N. (Compare boron "trusty man" from BOR, toron "brother" from TOR.) According to LR:393, the stem THOR itself means "come swooping down"; Christopher Tolkien is undoubtedly right when he takes "this to be an indication of the root-sense of THOR". It seems, then, that thoron etymologically speaking means "one who comes sweeping down", used with reference to eagles.
tó ?"that, it". The wording in LR:389 s.v. TA would most naturally be taken to indicate that tó has the same meaning as Quenya tar, namely "thither". However, tó can hardly be a cognate of tar. Tar is said to represent primitive tad; this must be Common Eldarin for Primitive Quendian *tada, sc. the demonstrative stem TA "that" with the primitive allative ending -da "to" (WJ:366), hence *"to that" = "thither". But PQ *tada, CE tad would remain tad in Old Sindarin (in later Sindarin *tadh, not attested). The form tó would seem to require a primitive form *tâ (compare mó from mâ). Could tó actually be the cognate of Quenya ta "that, it"?
*tóra possible correction of tára, q.v.
toron pl. toroni "brother". Derived from a stem TOR, that is similarly glossed (LR:394). Toron clearly represents an extended stem *TORON, with ómataina (reduplicated and suffixed stem-vowel, here O) and the ending -N. Compare thele "sister" from THEL (LR:392); in the case of this word, Tolkien explicitly listed the expanded stem THELES (with ómataina, here E, and the ending -S; perhaps this ending is perceived to be feminine, while -N is masculine). In the entry THEL "sister", Tolkien urged the reader to "cf. tor, toron- brother", indicating that this is also an example of an expanded stem. For similar formations, see boron and *thoron.
tre- (unstressed prefix) *"through" (LR:392 s.v. TER/TERES). Concerning the etymology of this prefix, see trí. The prefix occurs in the words tre-batie "traverse" (LR:352, in the entry BAT), evidently also in trenare "he tells to end" and in the corresponding noun trenárna "account, tale" (LR:374 s.v. NAR2). In tre-batie "traverse", this prefix appears with its basic meaning "through" (compare the independent preposition trí); this is literally "through-walk". In trenare "he tells to end" the prefix indicates completeness: *"he through-tells", recounts the whole story. The noun trenárna likewise refers to a complete story, an "account".
trenare "he recounts, tells to end", infinitive trenarie. Derived from the stem NAR2 "tell, relate" (LR:374); for more information about this stem and its relationship to Quenya nyar-, see naróbe. Concerning the meaning the prefix tre- has here, see the separate entry above; concerning its etymology, see trí. Removing this prefix, we are left with a verb nare, an aorist that must descend from *nari, final -i becoming -e at the Common Eldarin stage, but remaining i where not final: Compare yurine "I run", which would appear as *yure "(he) runs" if the pronominal suffix -ne were removed; conversely, "I recount, tell to end" is undoubtedly *trenarine. - While trenare is translated "he recounts, tells to end", no pronominal element "he" is actually present. It may be that the form with no explicit pronominal ending was sometimes used by itself as a third person sg. (or specifically third person masculine sg.) form, but it may also be that Tolkien simply included the pronoun "he" in the translation to indicate that this is a 3rd person sg. form. - The infinitive trenarie has a well-attested infinitive ending -ie that is also found in Quenya; see bronie. - While nar- in trenare/trenarie seems to behave as a "basic" verbal stem, the root NAR2 also yields a longer, "derived" verb *narâ-; see naróbe.
trenárna "account, tale" (LR:374 s.v. NAR2). The accent mark in trenárna probably indicates simply that the middle syllable receives the accent, not that the vowel is long. Trenar- is obviously the stem of the verb trenare "recounts, tells to end" discussed above (removing the aorist ending -e). The suffix -na added to the stem is seen to form an abstract or verbal noun here. Undoubtedly this is properly a past participle *"accounted, told" that is subsequently used as a noun: something that is accounted or told, hence an "account" or a "tale". The ending -na is undoubtedly a descendant of the frequent adjectival or participial ending -nâ (see khalla, representing khalnâ). While trenárna had come to be used as a noun, it seems that words in -na normally retained their original adjectival meaning in Old Sindarin (see for instance muina).

trí "through". Derived from a stem TER "pierce" with an extended variant TERES showing reduplicated stemvowel (ómataina) and a suffixed -s (LR:392). As we see, TER is itself a verbal stem, but it yields a preposition - the idea of "through" evidently evolving from "piercing". Trí may represent *terês, accented on the final syllable (for the long ê, compare the primitive adjectives terêwâ "piercing, keen" and terên(ê) "slender" listed in the same entry in the Etymologies). About the Common Lindarin stage, the first, unaccented vowel of *terês may have disappeared, reducing the word to *t'rês (compare brasse from b'rás-sê, stem BARÁS). In Old Sindarin, earlier long ê became long í (compare dîr, khíril), while final -s was lost (compare such words as pele, thele, derived from the stems PELES, THELES). Hence, *t'rês became trí. It is also possible that the primitive form was simply *terê (with no final s to be lost later); compare Quenya ter, terë (final s was not lost without trace in Quenya). A form *terê would represent the simplest form of the stem, TER. - As an unstressed prefix, trí appeared as tre-. This must represent *tere(s)-, the long ê of *terê(s) being shortened when unstressed, and short e did not change to i in Old Sindarin. (In later Sindarin we find both tre- and tri- as prefixes; the latter is evidently reformed on analogy with trî, the descendant of trí.)
túgo "muscle, sinew, vigour, physical strength". Derived from a stem TUG, itself undefined (LR:394). The primitive form is given as tûgu, with lengthening of the stem-vowel and an ending -u. While this could be simply the stem-vowel suffixed, it should be noted that several primitive words denoting body-parts end in -u; see ranko (itself from ranku) for examples. The most basic meaning would then be "muscle, sinew", with "vigour" and "physical strength" as secondary, more abstract meanings. (These glosses do not necessarily apply to túgo or tûgu, but rather to later words: Quenya tuo, Sindarin tû, Ilkorin tûgh or tû.)
tuio- "swell, grow fat". Derived from a stem TIW "fat, thick" (LR:394). The Quenya cognate tiuya- clearly points to a primitive form *tiujâ- (for *tiwjâ-), with the common verbal ending -jâ. It seems that *tiuj- turned into *tiui- at some stage, but the "triphthong" iui was simplified to ui in Old Sindarin (cf. euj- also becoming ui; it is indeed probable that euj- became iuj, iui at some stage, then being simplified to ui; see buióbe for a possible example). However, tiujâ- should have yielded Old Sindarin *tuia-, not tuio-. It may be that Tolkien or the transcriber confused *tuia- with tuio, the later Sindarin form. But notice the hyphen following tuio-, suggesting that it is not a complete word in itself; it may be short for *tuióbe, the same verb with the infinitive ending -be. When this ending is present, the -â of *tiujâ is not final and therefore regularly becomes -ó- instead of -a. Later Sindarin tuio must be derived from Old Sindarin *tuióbe, not simply tuio.
túka "thick, fat". Derived from a stem TIW "fat, thick"; the primitive form is given as tiukâ (for *tiwkâ) with the well-attested adjectival ending -kâ (compare for instance phauka). It seems that tiuk- at some early stage became *tjûk- (but later than Common Eldarin, since Quenya has tiuca instead of **tyúca). Following a t, the i of the diphthong iu became a semi-vowel j (= y as in you), but to maintain the prosodic length of the original diphthong iu, the new monophthong u became long û. (Compare *siulê > *sjûlê; see hyúle.) The new initial combination tj evidently merged into a single sound, palatalized t, which was regularly de-palatalized to become normal t in Common Lindarin (cf. kelepe); after this change, the long ú was the sole trace of the original iu, maintaining the length of this lost diphthong: Hence túka in Old Sindarin.
tulugme "support, prop" (noun). Derived from a stem TULUK (LR.395), not itself defined, but apparently meaning "firm, steadfast", or as verb "make firm". The primitive form is given as tulukmê, with an ending -mê that normally forms abstracts or verbal nouns (see ragme). It seems, then, that tulukmê was originally "support" as an abstract, later being used of a concrete prop (in the concrete meaning, we would rather have expected *tulukmâ > *tulugma; see parma concerning the ending -mâ often used to form words for implements). - Before a nasal consonant, unvoiced plosives become voiced in Old Sindarin, hence km > gm. Compare ndagno, ragme.
Túna, name of an Elf-city in Valinor, or the hill upon which it was built (a word that for historical reasons could occur in "Old Noldorin", but hardly in Old Sindarin). In the published Silmarillion, the city itself is called Tirion, while Túna was the hill it stood upon; to some extent the names would be interchangeable. The stem TUN (LR:395) is not defined; the words derived from it suggest that it basically has to do with hills, mounds, or simply with being tall. The primitive form of Túna is given as Tûnâ; actually the final vowel -â has a diacritic indicating that it can be both short or long, Tûna or Tûnâ. Túna must be derived from the latter, since short final -a was lost already at the Common Eldarin stage (compare Quenya Tún, an alternative, shorter form of Túna).
uia "envelope", especially of the Outer Sea or Air enfolding the world within the Ilurambar or world-walls. Derived from a stem WAY "enfold" (LR:397); the primitive form is given as wâjâ (wâyâ). The ending -â may be a noun-former or simply the stem-vowel suffixed. Since the stem-vowel is also lengthened to â in its normal place, it first became ô (= ó) in Old Sindarin, like any other non-final â (see abóro). The oldest OS form was therefore wôya, which Tolkien asterisked as if it were "unattested". It may be that wôya later became *woia, and when earlier oi became ui (compare muina < *moinâ), *wuia was simplified to uia, the semi-vowel w being lost before the corresponding vowel u.
Uigolosse "Everlasting snow" = Taniquetil (Oiolossë); later Sindarin Uilos (Taniquetil being called Amon Uilos, Mount Everwhite). Uigolosse is listed in LR:379 in the entry OY, which stem means "ever, eternal". Letters:278 similarly mentions oio as a "Primitive Elvish" element meaning "ever", with a later Sindarin form ui. As demonstrated by the word Uigolosse, the ancient diphthong oi became ui already at this stage (compare muina; see also nui). The second element is golosse "snow", as an Old Sindarin word attested in this compound only. The stem is obviously GOLÓS "snow" (LR:359). The primitive form would be *golossê , the doubling of the s perhaps being simply a medial fortification. The ending -ê occurs in several primitive words denoting substances; see kelepe. OS golosse would thus be a cognate of Quenya olossë. - In the entry GOLÓS, Tolkien also mentioned a Sindarin adjective gloss "snow-white" (RGEO:70 has glos(s) "dazzling-white"). This is evidently derived from a form of the stem that had lost the first, unaccented vowel: *g'lossê (concerning the final vowel, compare Quenya lossë "snow-white" [RGEO:69]). The Old Sindarin form would be *glosse. Since Tolkien in RGEO:69 also stated that the stem was los ("applied to fallen snow") rather than GOLÓS as in the Etymologies, we might reconstruct the primitive form simply as *lossê, assuming that a g- was prefixed later (initial l often being elaborated to gl in Sindarin; see WJ:411, note 13; indeed Tolkien in RGEO:70 stated that this was the case in the word gloss). The Old Sindarin form could then be simply *losse. The point is that since Tolkien in later sources translated Sindarin Uilos and Quenya Oiolossë as "Ever-white" or "Ever-snow-white" (RGEO:69) rather than "Everlasting snow", we should now perhaps assume that the Old Sindarin form was *Uilosse (or less likely *Uiglosse) rather than Uigolosse.
Uinenda, name of a Maia, the wife of Ossë; Quenya Uinen. In the Etymologies, this name is listed in the entry UY (LR:396). The stem UY is not defined as such, but Quenya uilë, Sindarin uil means "seaweed". (Compare the description of Uinen in Valaquenta in the Silmarillion, that her "hair lies spread through all waters under sky" and that she loves "all weeds that grow there".) According to the Etymologies, the final element of the name is to be referred to the Elvish stem for water, NEN (LR:376). In this entry, the Sindarin form Ui-nend is connected to the adjective nend "watery", in turn the cognate of Quenya nenda, pointing to a primitive form *nendâ with medial fortification n > nd and the common adjectival ending -â. A primitive form *Uinendâ would produce Old Sindarin Uinenda (while Quenya Uinen may represent *Uinenda with a short final -a). Precisely how the name as a whole would be translated or interpreted is not clear: *"watery seaweed"? In the post-LotR period, Tolkien himself gave up trying to interpret this name: In the essay Quendi and Eldar from about 1960, he had Pengolodh observing that Uinen was one of the names that "are not Elvish, as far as can now be seen...[they] may...represent titles in the Valarin tongue, or such part of them as the Eldar could adapt" (WJ:404).
wa- (prefix) "together". As an Old Sindarin element, this prefix is mentioned in the Etymologies in the entry TOR (LR:394, Tolkien explaining the first element in the word wator), but TOR is obviously not the stem wa- is derived from. The stem is WÔ "together" (LR:399); Tolkien notes that if wo (with a short o) received the accent, it became wa already in "Eldarin", evidently meaning Common Eldarin. The prefix wa- occurs in the words wanúre, wanúro, wator, wathel, q.v.
waide "bond, troth, compact, oath". Derived from a stem WED "bind" (LR:397). A primitive form wæ^dê is given. The ending -ê clearly denotes an abstract here. It may be that æ does not here represent the vowel [æ] (= the a of English cat), but rather a diphthong ae, produced by A-infixion in the stem WED (*waed-). For another example of earlier "æ" yielding Old Sindarin ai, see ndairo. Such a development would not be wholly in accord with the scenario sketched by Tolkien in VT39:10, where he says ae became long â in the Telerin (Lindarin) branch of the Elvish language family. If so, the Old Sindarin form of *waedê (later *wâdê) should perhaps be *wóde instead of waide, since earlier â became OS ó (in turn becoming au in later Sindarin: **gwaudh - but the actual Sindarin descendant of wæ^dê was gwaedh). But since VT39:10 reproduces a document about 25 years younger than the Etymologies, some changes in Tolkien's linguistic scenario are to be expected.
wanta- "depart, die". This word is not explicitly said to be Old Sindarin, but it is mentioned as the ancestral form of Sindarin gwanno and by its form it seems to belong to the OS stage of the linguistic evolution. The primitive form is obviously *wantâ-, sc. the stem WAN "depart, go away, disappear, vanish" (it is said that in Sindarin, or "Noldorin", this stem had replaced KWAL as the stem referring to death and dying; compare LR:366 s.v. KWAL). Wanta- is the stem WAN with the common verbal ending -tâ, that in this case adds nothing to the meaning of the stem itself. The later Sindarin infinitive gwanno cannot descend directly from wanta-, but represents *wantóbe with the infinitive ending -be (see buióbe). - It may be noted that one word derived from this stem WAN, Quenya vanwa "gone, departed", is assigned a quite different etymology in a later source; here the stem is simply wâ, or AWA (WJ:365-366). We may still chose to accept these words from the Etymologies. - A few examples suggest that at one stage in the period designated as Old Sindarin, wanta may have appeared as *wantha; see thintha.
wanúre "kinswoman". Mentioned in the Etymologies in the entry THEL/THELES (LR:392), but this is not the actual stem of this word. Wanúre must descend from *wonôrê, the feminine equivalent of masc. wonôrê; see wanúro below.
wanúro "brother" (LR:378 s.v. NÔ) or "kinsman" (LR:394 s.v. TOR). The stem is not TOR, but NÔ "beget". Wanúro clearly represents *wonôrô, literally *"together-begotten one", one begotten "together" with another, hence a brother, or a kinsman (one being born within the same family). Concerning the prefix wo- "together", see wa-. *Nôrô is the stem NÔ with the masculine ending -rô. This ending is most often used to derive agental formations (WJ:371), but *nôrô means "begotten one", not "begetter". The short variant -ro functions simply as a masculine ending in the primitive word târo "king" (from TÂ, TA3 "high, lofty, noble", LR:389; hence literally "high one, noble one"), and it seems that -rô has a similar non-agental meaning in nôrô. Wanúro "kinsman" has a feminine counterpart wanúre "kinswoman", itself representing *wonôrê. The feminine ending -rê occurs in only one primitive word listed by Tolkien himself, weirê "[female] weaver", from WEY "weave" (LR:398). Here it is agental, just like -rô usually is, but in *wonôrê it would be simply a feminine ending. *Wonôrô and *wonôrê were still distinct words in Old Sindarin, wanúro and wanúre (ô regularly becoming ú, see brûna), but in Sindarin, after the loss of the final vowels, these words fell together as gwanur, a gender-neutral word for relative ("kinsman or kinswoman", LR:392 s.v. THEL-, THELES-).
warie "betray, cheat": This is simply the stem WAR "give way, yield, not endure, let down, betray" (LR:397) with the infinitive ending -ie (concerning which see bronie). Contrast the more complex formation awarta derived from the same stem.
wasse "stain", noun (cf. the synonym watte). Derived from a stem WA3 "stain, soil" (it is not clear whether these glosses are intended as nouns or verbs; perhaps they cover both). A primitive form is given as wahsê, in which the letter h probably stands for [x], sc. German ach-Laut: The 3 (the back spirant gh) of the stem WA3 evidently become unvoiced by contact with the following unvoiced s (perhaps the very oldest form was *wa3sê), the cluster hs [xs] being assimilated to ss in Old Sindarin (compare ht similarly producing tt; see watte). The ending -sê may denote something that is produced by the verbal action described by the verbal stem (see brasse), hence wahsê "something that is made by staining or soiling" = "a stain".
watha "shade", noun. Derived from a stem WATH that is also glossed "shade". The primitive form would be *wathâ; here, -â functions as a noun-former (perhaps with some kind of local meaning in this case; compare rattha, yura).

wathel "sister, associate". It appears that this is simply the stem THEL "sister" (LR:392) with the element wa- "together" (q.v.) prefixed. The corresponding masculine form wator "brother" was "especially used of those not brothers by blood, but sworn brothers or associates" (LR:394 s.v. TOR). Since wathel is glossed "associate" as well as "sister", we can assume that this word, too, primarily refers to relationships other than blood-kin. See wator.
wator "brother" ("especially used of those not brothers by blood, but sworn brothers or associates"). Literally *"together-brother"; tor is essentially identical to the stem TOR "brother" (LR:394), and the element wa- "together" (q.v.) has been prefixed. This is thus wholly parallel to wathel "sister, associate" above. In both cases, the prefix wa- "together" seems to suggest that the word refers to a "brother" or "sister" that has been joined "together" with another person by some close relationship other than blood-kin.
watte "a stain" (cf. the synonym wasse). Derived from a stem WA3 "stain, soil" (LR:397). A primitive form wahtê is given; this is probably Common Eldarin for Primitive Quendian *wa3tê (compare wattóbe below concerning this development). The ending -tê is very rare, but wahtê "a stain" seems to be a noun based on the verb wahtâ- "to soil, stain" (see wattóbe below). Compare a primitive word (actually said to be a "verbal derivative") mentioned in WJ:396: kirtê "cutting" (the origin of Sindarin certh "rune"). The ending -tê here seems to denote something that is made by the action denoted by the stem (here obviously KIR "cut, cleave"; concerning the meaning of this stem, listed but not defined in the Etymologies, see kir- in the Silmarillion Appendix). If we take Tolkien's glosses to WA3 - "stain, soil" - as verbs rather than nouns, wahtê has much the same relationship to its root WA3 as kirtê has to KIR. - The medial cluster ht has been assimilated to tt in Old Sindarin. We would have expected that tt, in turn, became tth (compare matth- from maht- [see matthô-be], probably via matt-). We must assume that watte represents an early stage of Old Sindarin, later becoming *watthe (compare later Sindarin gwath).
wattóbe "to soil, stain". Derived from the same stem WA3 "stain, soil" (LR:397) as watte above. A primitive form wahtâ- is given. This is evidently Common Eldarin for Primitive Quendian *wa3tâ, before 3 was devoiced to h (probably = German ach-Laut here) in contact with unvoiced sounds like t. Compare LR:371 s.v. MA3, where a primitive verb ma3-tâ yields Common Eldarin mahtâ- (Old Sindarin matthô-be, q.v.) *Wa3tâ-, wahtâ- shows the common verbal ending -tâ. Notice that as in watte above, earlier ht is assimilated to tt in Old Sindarin. Again we miss the subsequent change tt > tth (as in mahtâ > *mattô-be > matthô-be), and again we must assume that this is early Old Sindarin, later becoming *watthóbe (compare later Sindarin gwatho). - *Wahtâ- by itself would normally appear as *watta- (*wattha-), but when the infinitive ending -be is present, â is not final and regularly becomes ó, hence wattóbe.
weda "bond". Derived from a stem WED (LR:397), defined as "bind". The primitive form is given as wedâ, which is a good example of -â functioning as a noun-former. However, the meaning of such a derivative is not entirely predictable in relation to the stem meaning: a wedâ is simply a noun somehow having to do with binding. In this case, -â indicates an impersonal agent; a bond is a thing that binds. But contrast for instance yura (*jurâ) from the stem YUR "run"; this noun does not denote an impersonal "runner", but the place where running (of water) is taking place: yura refers to the "course" of a river.
-wega (compound form) "-man" as an element in masculine names, like Bronwega (q.v.) Derived from a stem WEG "(manly) vigour" (LR:398). The primitive form wegô shows the masculine ending -ô and apparently meant simply "man" (that, at least, is the meaning of the Quenya word vëo, descended from wegô). Tolkien also mentions a "compound form" -wego with shortening of the final -ô to -o. There are parallels to this both in the Etymologies and in later sources; the shortening of final long vowels in nouns when they function as the second element of a compound seems to be a general rule. In LR:395 s.v. TUR, tûrô "lord" is said to have the form turo (or just tur) in compounds (an example is Spanturo "Cloud-lord", LR:387 s.v. SPAN). See also WJ:403 concerning khînâ "child", appearing as -khîna in compounds. The Old Sindarin compound form -wega must represent -wego (though short -o was normally lost at the Common Eldarin stage; this is our sole example of it becoming -a instead).
wende "maiden", only attested in the compound Bana-wende (see Bana), but clearly derived from the stem WEN, also defined as "maiden" (LR:398). The primitive form would be *wendê with the feminine ending -ê; the d could come from an extended stem WENED (and such a form is actually listed in LR:398) or be due to a medial fortification n > nd.
[wintha] "it fades, advesperascit" (Latin: 'evening approaches'). This word, listed in LR:399 s.v. WIN, WIND, was struck out along with the rest of this entry in the Etymologies. The stem WIN (also with medial fortification: WIND) was not itself defined, but the gloss of the first primitive word that was listed in this entry (windi), namely "blue-grey, pale blue or grey", is probably more or less the same as the basic meaning of the stem. Other primitive derivatives, such as winjâ (winyâ) "evening", seem to develop the basic idea of "grey" (in winjâ used of twilight). Wintha was derived from a primitive form given as winta-; at the most primitive stage this must have been *wintâ-, for final short -a would have disappeared already at the Common Eldarin stage. *Wintâ- would be the stem WIN *"grey" with the common verbal ending -tâ; the literal meaning would perhaps be "become grey" (and hence "fade"). The form wintha shows the otherwise poorly attested, but interesting change nt > nth; see thintha for some thoughts about this. The verb thintha "fade" (LR:392 s.v. THIN) is indeed similar to wintha in both form and meaning; when he struck out the entry WIN, WIND, Tolkien made a cross-reference to THIN, as if to suggest that WIN was to be suppressed in favour of THIN. - The gloss to wintha, namely "it fades" rather than "(to) fade", is interesting: it indicates that wintha- is properly a present-tense form rather than an infinitive (the infinitive would be *winthóbe). Surely the same goes for other Old Sindarin verbs in -a, though they are glossed as infinitives: awarta "forsake, abandon", rista- "rend, rip", sirya- "flow", skhalia- "veil, conceal", thintha "fade". The "it" of the gloss "it fades" does not indicate that any pronominal element "it" is present (it would be -s, as in persôs "it affects", q.v.) "It fades" here means *"it is getting dark" (compare the Latin gloss advesperascit, evening approaches), so "it" is just a concession to English idioms and has no real meaning here.
wôia "envelope", especially of the Outer Sea or Air enfolding the world within the Ilurambar or world-walls. In the source (LR:397 s.v. WAY), this form is asterisked as unattested. See uia (the later form).
wóra "soiled, dirty". Derived from the same stem WA3 "stain, soil" (LR:397) as watte; the primitive form is given as wa3râ with a well-attested adjectival ending (see tára). The spirant 3 was early lost, but the previous vowel was clearly lengthened in compensation, producing an intermediate form *wârâ (whence Quenya vára); as usual, long non-final â then yields Old Sindarin ó (see abóro).
yadme "bridge", only attested in the compound elyadme "sky-brigde" = rainbow (LR:360 s.v. 3EL). Derived from a stem YAT "join" (LR:400). The primitive form is given as jatmâ (yatmâ), with an ending -mâ that is often used to form words for implements (see parma; compare sniuma). Before a nasal consonant, unvoiced plosives become voiced in Old Sindarin, hence tm > dm (for other examples of this phenomenon, see ndagno, ragme, tulugme). Since long final -â normally becomes -a in Old Sindarin, we might have expected *yadma instead. However, it seems that the endings -mâ, -wâ were altered to -mê, -wê following a t; compare katwe (q.v.) from katwâ. This change jatmâ > *jatmê must have occurred already in Common Eldarin, since it is also reflected in Quenya yanwë.
yaiwe "mocking, scorn". Derived from a stem YAY "mock" (LR:400); the primitive form would be *yaiwê, where the ending -wê is seen to form what could be considered a verbal noun. Other primitive words exemplifying this ending, "reconstructed" by Tolkien himself, include et-kuiwê "awakening" from KUY "awake" (LR.366) or wanwê "death" from WAN "depart" (LR:396). In LR:398 s.v. WEG, Tolkien explicitly states that -wê is an abstract suffix.
yen-panta "aged, long-lived", literally *"year-full" (LR:400 s.v. YEN). This word is not explicitly said to be Old Sindarin, but it is listed as the ancestral form of Sindarin ifant (better spelt iphant if we are to use the spelling outlined in LotR Appendix E), so we include it here. Moreover, the word panta "full" (q.v. for etymology) is attested by itself as an Old Sindarin word. The element yen here means "year", representing the stem YEN itself (according to the Etymologies, this stem also conveys "year", but in LotR, Tolkien used the Quenya derivative yén to denote a "long year", an Elvish century of 144 solar years - but this is apparently not the intended meaning here). Tolkien outlines a development yen-panta > impanta > in-fant (in turn evidently becoming iphant). This cannot be taken literally; in particular, we can hardly get from impanta to in-fant. Perhaps Tolkien simply meant to indicate that Sindarin ifant (iphant) represents in (shortened form of în, "year") and fant (phant), a nasal-mutated form of pant "full" (LR:366 s.v. KWAT). The actual development of Sindarin iphant must rather be imagined something like this: To Quenya yén corresponds Old Sindarin *yín (since long é became í in OS). A compound *yín-panta becomes impanta when i is shortened before a consonant cluster, y is lost before i, and n is assimilated to the following p, becoming m. Later, mp is further assimilated to pp, which produces Sindarin ph = f. Hence ifant, iphant.
yura "course". Formed from a stem YUR "run" (LR:400), here referring to the running of water, though this stem is not especially associated with water - contrast KEL, defined as "go, run (especially of water)" (LR:363). The primitive form must be *yurâ with -â as a noun-former; here it denotes the place where the activity denoted by the stem occurs - sc. where something, namely water, runs. Compare the synonym rattha (q.v.) from rattâ, showing the same ending (like yura, rattha is derived from a stem having to do with movement: RAT "walk").
yurine "I run". Formed from the same stem YUR "run" (LR:400) as yura above. The yuri- part is simply the primitive I-stem *yuri unchanged. This is a simple aorist "runs" (as opposed to the continuative form *yûrâ "is running"). By itself, this word would appear as *yure in Old Sindarin (compare the ending -e in trenare), since final short -i became -e in Common Eldarin. Where not final - before an ending, like the pronominal suffix -ne "I" - the final vowel is unchanged. The precise development of the suffix itself is uncertain. The Quenya ending -nye could point to an earlier form *-njê; this would appear as -ne in Old Sindarin: The primitive palatalized sounds were depalatalized in Common Lindarin (see kelepe), so nj would become normal n, and the final long -ê would be shortened to -e. However, it is also possible that the suffix -ne originally had the same form as the stem NI2 "I" (LR:378). *Yurini would become yurine in Old Sindarin, final short -i again turning into -e in Common Eldarin. (If so, Quenya -nye might be explained as a later elaboration of a simpler ending -ne, perhaps to avoid confusion with the past tense ending -ne.)