√GAL¹ “light; shine, be bright”
A Sindarin-only root for “light; shine”, a variant of √KAL of the same meaning. Its most notable derivative is S. galad “radiance, light”, an element in the names Galadriel and Gil-galad. This root did not explicitly appear in the Qenya Lexicon of the 1910s, where the base for Qenya light words was given as ᴱ√KALA “shine golden” (QL/44), but nearly all the Gnomish derivations of this root begin with g-, as in G. gal- “to shine” and G. glarw(ed) “bright, light” (GL/39). Given that initial [g] > [k] in early Quenya, this makes it very likely the actual early root was *ᴱ√GALA, or at the very least a blending of ᴱ√KALA and ᴱ√GALA.
The first explicit appearance of the root ᴹ√GAL¹ “shine” was in The Etymologies of the 1930s (Ety/GAL), where it was given along with an extended root ᴹ√GALAN “bright” (EtyAC/GAL¹). Tolkien considered but rejected having some Quenya derivatives of these roots: ᴹQ. ala “day” and ᴹQ. alan “daytime” (EtyAC/GAL¹), but he seems have to decided that ᴹ√GAL was a Noldorin-only root, as described in the ᴹ√KAL entry from the same set of documents (Ety/KAL).
The root √GAL reappeared in various etymological notes from the 1950s and 60s, where Tolkien again iterated it was not used in Quenya, and was a root specific to Sindarin (PE17/59) and possibly also Nandorin (PE17/50). The root was potentially problematic in the name Gil-galad, however, in that it was not mutated to Gil-’alad. It seems likely that when Tolkien first coined this name, the second element was from N. calad “light” from the root ᴹ√KAL. This root and word survived into Sindarin (UT/65), and Tolkien sometimes still considered it the basis for Gil-galad (PE17/50).
However, at one point he decided the elements in the names Galadriel and Gil-galad were the same; to explain the lack of mutation in Gil-galad, he posited that the root was actually strengthened to √ÑGAL (PE17/59). Galadriel’s name in the Shibboleth of Fëanor from 1968 was based on the root √ÑAL “shine by reflection” (PM/347), so it seems this was the path Tolkien eventually followed. This make the eventual status of √GAL “shine” rather unclear, especially since some of its other derivatives like S. uial twilight (PE17/153) were sometimes derived from √ÑAL (PE17/169).
√GAL² “grow (like plants), flourish, be healthy, be vigorous, bloom, [ᴹ√] thrive”
A root meaning “grow, flourish” used in both Quenya and Sindarin, though in Quenya it was influenced by other roots such as √AL(A) “good, blessed”. Its precursors in the earliest versions of Tolkien’s languages seem to be ᴱ√ALA² “spread” (QL/29) and a Gnomish-only root ᴱ√cala [kala], unglossed but with derivatives like G. calw “green shoot, sapling, sprout” and G. caltha- “wax, grow, flourish” (GL/25).
The root ᴹ√GALA first appeared in The Etymologies of the 1930s with various glosses, the last being “thrive” (Ety/GALA). The relevant entries went through considerable revision. For example, the meaning of this root was first given as “grow”, but this meaning was rejected and Tolkien said the verb (or root?) for “grow” was ol- (EtyAC/GAL(AS)). The entry where this appeared was deleted, but the verb ᴹQ. ola- “grow” appeared in the Quenya Verbal System as a replacement for ᴹQ. ala- (PE22/113). Since ᴹ√GALA was retained, most likely the meaning of this root was changed from “grow” to “thrive” in the 1930s and 40s.
By the 1950s and 60s, however, the sense “grow” was restored (Let/426; PE17/153; PE22/160), whereas √OL seems to have shifted in sense to “become”, for example in the primitive verb ✶olā- “become, come into being, turn into (another state)” in notes from early 1950s (PE22/134). In this restored meaning, √GAL served as the basis for the Elvish verbs for “grow” such as S. gala- and Q. ala-¹ (PE17/131-132), though this verb may have been more limited in sense to the growth or growing of plants (PE17/100, 153).
The basis for Elvish “tree” words, this root first appeared in The Etymologies of the 1930s as an extension of ᴹ√GALA “thrive” (Ety/GALAD). This replaced the earliest derivation of “tree” from the Qenya Lexicon of the 1910s, where the Qenya word for “tree” ᴱQ. alda was derived from ᴱ√ALA² “spread” (QL/29). In The Etymologies, the Quenya form of this word remained the same, but the 1910s Gnomish words G. âl “wood” and †alwen (GL/19) became the 1930s Noldorin word N. galadh “tree” (Ety/GALA). Quenya and Sindarin retained these words for “tree” thereafter, and while Tolkien did not mention the root √GALAD again, his continued use of primitive ✶galadā “tree” (Let/426; PE17/153; PE21/74; UT/266) made it clear this root remained valid.
√GAP “bend (tr.)”
A root from etymological notes from the late 1960s glossed “bend (tr.)” with the derivatives Q. ampa and S. gamp “hook” (VT47/20). This root first appeared (unglossed) as ᴹ√GAP in The Etymologies of the 1930s with Quenya and Noldorin words having the same form (Ety/GAP), albiet slightly different meanings: the Noldorin word was glossed “hook, claw” whereas the Sindarin word from the late 1960s was glossed “hook, crook”.
ᴹ√GAR “keep, hold, possess; maintain, defend”
A root in The Etymologies of the 1930s, apparently a variant of ᴹ√ƷAR, and having extensions ᴹ√GARAD and ᴹ√GARAT (Ety/ƷAR). It was given various glosses in separate entries such as “hold, possess” and “keep, maintain, defend”, but these glossed entries were all deleted (Ety/GAR; EtyAC/GAR). Two notable derivatives of ᴹ√ƷAR/GAR were ᴹQ. arda “realm, region” and N. ardh “realm” (Ety/ƷAR). Both these words reappeared in the Quendi and Eldar essay from 1959-60 as Q. arda and S. gardh “region” from primitive ✶gardā (WJ/402), hinting at the continued validity of the root √GAR, though its later meaning is unclear.
ᴹ√GAS “yawn, gape”
A root in The Etymologies of the 1930s glossed “yawn, gape” with several derivatives such as ᴹQ. assa/N. gas “hole” and N. gaw “void” (Ety/GAS). The names based on this root like N. Belegast and ᴹQ. Ilmen-assa do not appear in later versions of the Legendarium, so the eventual fate of this root is uncertain, especially given that there are other roots of similar meaning like ᴹ√YAG “yawn, gape” (Ety/YAG) that do reappear later (Let/383; PE17/35, 42). For purposes of Neo-Eldarin, however, derivatives of this root are among the best attested words for “hole”, so it’s probably best to retain ᴹ√GAS and its derivatives in Neo-Eldarin writing.
An unglossed root in The Etymologies of the 1930s with derivatives like N. gath “cavern” and N. gador “prison, dungeon” (GAT(H)), so probably meaning something like “*cave”. Its most notable derivative was N. Doriath “Land of the Cave”. But later on in this same entry, Tolkien added Doriathrin words Dor. gad “fence” and Dor. argad “outside the fence”, indicating a conceptual shift in the meaning of this root. Indeed, in later versions of The Silmarillion, S. Doriath was translated “Land of the Fence” containing the element S. iath “fence” (WJ/370, 378), so most likely ᴹ√GAT(H) “*cave” was abandoned.
ᴹ√GAY “red, copper-coloured, ruddy”
A root in The Etymologies of the 1930s glossed “red, copper-coloured, ruddy”with the derived adjectives ᴹQ. aira and N. goer of similar meaning (Ety/GAWA). It replaced a (deleted) root ᴹ√GAIRĀ (EtyAC/GAIRĀ). Its most notable derivative is the name N. Taragaer “Ruddyhorn” (Ety/TARÁK), which was the precursor in Lord of the Rings drafts of 1940s to the name of S. Caradhras “Redhorn” (RS/419). Given the eventual replacement of this name in the final version of The Lord of the Rings Tolkien likely abandoned this root, especially given its conflict with the later root √GAY(AR) “awe, dread” that was the basis for words for “sea”.
√GAY(AR) “awe, dread; astound, make aghast; sea”
In Tolkien’s later writings, the root for Elvish “sea” words seems to be √GAY(AR) or √AY(AR). The first hints of this root may be words from the Gnomish Lexicon of the 1910s: G. ail/ᴱQ. ailo “lake, pool” and G. ailion/ᴱQ. ailin “lake” (GL/17), perhaps indicating a (hypothetical) early root ᴱ√AYA(LA) “lake”. The last of these Early Qenya words also appeared in the contemporaneous Qenya Lexicon, but Tolkien gave its root only as “?” (QL/29). This early root might also be the basis for 1920s ᴱQ. ailin “shore” which appeared in the Oilima Markirya poem (MC/213).
The word ᴹQ. ailin “pool, lake” reappeared in The Etymologies of the 1930s derived from the root ᴹ√AY, with its Noldorin form N. oel reflecting Noldorin rather than Gnomish phonology (Ety/AY). This root had an extended form ᴹ√AYAR “sea” from which Tolkien derived ᴹQ. ear and N. oer of the same meaning; among other things this extended root was the basis for a new etymology for the name ᴹQ. Earendil from the 1940s forward as “Friend of the Sea” (SD/241, 305); in early writings ᴱQ. Earendel was connected to ᴱQ. earen “(young) eagle” (QL/34).
The root √AY(AR) “sea” continued to appear in Tolkien later writings, for example in the second version of Tengwesta Qenderinwa from around 1950 (TQ2: PE18/97). However, Tolkien also considered alternate derivations of “sea” words from new root √GAY(AR), for example changing √AY(AR) >> √GAY(AR) in etymological notes written between the 1st and 2nd editions of The Lord of the Rings (PE17/27). As opposed to √AY(AR) which meant simply “sea”, Tolkien also glossed √GAY(AR) as “astound, make aghast” (WJ/400) or “awe, dread” (PM/363) and it was the basis for other words such as S. gaer “awful, fearful” (WJ/400) as well as S. goe “terror, great fear”, Q. aica “fell, terrible, dire” and Q. aira “holy, sanctified” (PM/363).
Tolkien seems to have been unable to make up his mind between these two derivations, as reflected in an ongoing vacillation between S. gaear (PE17/27; PM/363; WJ/400) and aear (Let/386; RGEO/65) as the Sindarin word for “sea”. This word appeared in The Lord of the Rings in the phrase nef aear, sí nef aearon “here beyond the Sea, beyond the wide and sundering Sea”, but that does not resolve the question as the word in this phrase seems to be lenited, and hence would lose its initial g (if any). This vacillation continued late into Tolkien’s life: in the published corpus Tolkien used √AYAR “sea” in a letter from 1967 (Let/386) and √GAYA “awe, dread” in the Shibboleth of Fëanor written in 1968 (PE/363).
Neo-Eldarin: For purposes of Neo-Eldarin writing, I think it best to use √GAY(AR) as the form for this root, though admittedly this does create problems for the etymologies of Q. ailin and S. ael “lake”.
A root in The Etymologies of the 1930s glossed “sick” with derivatives ᴹQ. engwa/N. gem “sickly” (Ety/GENG-WĀ). It was given as ᴹ√GENG-WĀ, and thus perhaps represented some extension of an otherwise unattested root ᴹ√GENG. In the entry for ᴹ√YEN from the The Etymologies it appeared as ᴹ√GEM in the discussion of N. ingem “old, (lit.) year-sick” (EtyAC/YEN), but I believe this represents the Noldorin phonetic developments of the true primitive form rather than a conceptual variation. The continued appearance of Q. Engwar in The Silmarillion narratives of the 1950s and 60s hints that this root may have remained valid as well.
ᴹ√GLAM “inarticulate voiced sounds”
A (Noldorin-only) root in The Etymologies of the 1930s glossed “inarticulate voiced sounds”, a strengthened form of the root ᴹ√LAM which was the basis of general sound words (Ety/GLAM). The derivatives of ᴹ√GLAM include N. glamm “shouting, confused noise” and N. glavra- “babble” but also N. glamor “echo” and N. glamren “echoing”, the last of these seen in N. Dorlamren, the pure Noldorin form of N. Dor-lómen “Land of Echoes”, which incorporates the Ilkorin word lómen “echoing” (LR/249; Ety/LAM).
The word S. glam “din, uproar, the confused yelling and bellowing of beasts” reappeared in the Quendi and Eldar essay written in 1959-60, where it was an element (as it was earlier) in the collective name for orcs: S. Glamhoth “Din-horde” (WJ/391). This strongly indicates that the “confused sounds” aspect of this root survived into Tolkien’s later conception of the language. However, the “echo” portion seemed to have been transferred to the unstrengthed form of this root √LAM as with (West) Sindarin loven “echoing” vs. North Sindarin lómin, which remained an element in Dor-lómin “Land of Echoes”, albiet reconceived of as a North Sindarin name (PE17/133).
√(G)LAN “rim, edge, border, boundary”
A root appearing in Tolkien’s notes on The Rivers and Beacon-hills of Gondor from the late 1960s, glossed “rim, edge, border, boundary” and with numerous derivatives of similar meaning (VT42/8). Tolkien coined this root to explain the river name S. Glanduin, which he translated as “Border-river” in this context. In these same notes (or perhaps other notes that were close contemporaries) Tolkien also said the Sindarin name of the Mering Stream in Rohan was S. Glanhír “Boundary Stream” (UT/318).
The Sindarin and Telerin derivatives of this root have an initial gl-, with the sole exception of S. lanc/T. lanca “sharp edge, sudden end”. According to Tolkien: “It is debated whether gl- was an initial group in Common Eldarin or was a Telerin-Sindarin innovation (much extended in Sindarin)”. For the purpose of Neo-Eldarin, I think it is probably best to assume the gl- is from Common Eldarin, to keep this root distinct from √LAN which in earlier writing was the basis for various words having to do with threads and weaving (PE17/60; Ety/LAN).
√(G)LAWAR “good (physically), blessed, fortunate, prosperous, health(y)”
Tolkien’s Elvish words for “gold” were fairly stable in his conception of the languages, resembling Q. laurë and S. glaur for most of his life. The main exception was the Early Noldorin and Gnomish forms of the 1910s and 20s, which were ᴱN./G. glôr (PE13/144; GL/40). These early forms survived to some degree in Tolkien’s later conceptions, because in Sindarin compounds au often reduced to o, and thus the name Glorfindel “Golden-hair” retained the same form and meaning for Tolkien’s entire life despite the revision of G. glôr >> S. glaur.
Tolkien’s first primitive root for these words was ᴱ√LOU̯RI in the Qenya Lexicon of the 1910s (QL/51). This reflected the Early Qenya sound change whereby [ei], [ou] became [ai], [au]; later on these ancient diphthongs usually because [ī], [ū]. In this earliest conception, the root and its derivatives referred to the physical substance of “gold”, though not quite its mundane nature: Tolkien said that ᴱQ. laure was the “mystical” or “magic” name of gold as opposed to its more mundane name ᴱQ. kulu (QL/42; LT1/100).
Tolkien introduced a new root ᴹ√GLAWAR or ᴹ√LAWAR for these words in The Etymologies of the 1930s, along with the modified form for its Noldorin derivative N. glaur (Ety/GLAW(-R), LÁWAR). Tolkien said that ᴹ✶laurē referred to the “light of the golden tree Laurelin” (Ety/LÁWAR), indicating a conceptual shift in the meaning of the root to “golden light or color” rather than “mystical gold”. Indeed, in etymological notes Tolkien wrote sometime around 1960, Tolkien said of √LAWAR: “The application to gold of this stem was poetic and referred to colour primarily ... not to material (malta)” (PE17/159).
Tolkien’s representation of this root as both √GLAWAR and √LAWAR indicates some uncertainty on whether the initial gl- was from Common Eldarin or was only a later Ancient Telerin innovation. However, in later writings Tolkien typically represented the primitive word as ✶glawar(e) (PE17/17; PE21/80; VT41/10), and in the Outline of Phonology from the early 1950s Tolkien gave the Ancient Quenya word AQ. alaurē “sheen of gold” whose vowel augment arose from an abnormal vocalization of the ancient spirantalized initial g-: *glawarē > glaurē > ʒlaurē > alaurē (PE19/79). This strongly indicates the primitive root was √GLAWAR.
√(G)LOS “snow, whiteness”
In Tolkien’s writing the root √(G)LOS is mostly used for “snow” but also for “white”. The earliest indications of this root are words from the Gnomish Lexicon of the 1910s such as G. gloss “white, clear white”, G. glost “whiteness” and (possibly) G. glui “snow” (GL/40), indicating an (unattested) early root *ᴱ√LOSO of similar meaning. The (plural) adjective ᴱQ. losse “white” appears in the ᴱQ. Oilima Markirya poem of the late 1920s (MC/213), indicating this root spread to other branches of the Elvish languages.
The first clear mention of this root is in the The Etymologies of the 1930s where ᴹ√GOLOS is given as the basis for ᴹQ. olos(se) and N. gloss “snow, fallen snow” (Ety/GOLÓS). In later writings the root √LOS appears several times as the basis for snow words (PE17/26, 69, 160-161; RGEO/61). The last mention of this root is in notes on The Rivers and Beacon-hills of Gondor from the late 1960s, where Tolkien gave the root as √(G)LOS with the underlying meaning “white”, with two distinct Sindarin derivatives S. loss “snow” [noun] and S. gloss “snow white” [adj.] (VT42/18). Given this l/gl variation in Sindarin, likely the strengthened form √GLOS was a sporadic, Sindarin-only innovation.
The Elvish words for “stone” were established very early as Q. ondo and S. gond. In the Qenya Lexicon of the 1910s Tolkien gave the root of these words as ᴱ√ONO “hard” with derivatives like ᴱQ. ondo “stone, rock” and ᴱQ. onin “anvil” (QL/70). But its Gnomish derivatives like G. gonn “stone” and G. gontha “pillar” (GL/41) indicate the actual root was *ᴱ√ƷONO, since initial ʒ > g in Gnomish.
In The Etymologies of the 1930s Tolkien gave the root as ᴹ√GONOD or √GONDO “stone” with essentially the same Elvish forms: ᴹQ. ondo and N. gonn (Ety/GOND). The root itself did not appear in later writings, but Tolkien continued to state, with great frequency, that the primitive form of the word was ✶gondō (Let/410; PE17/28; PE18/106; PE21/81; PM/374; RC/347).
ᴹ√GOR “violence, impetus, haste”
A root in The Etymologies of the 1930s glossed “violence, impetus, haste”, with derivatives like ᴹQ. orme “haste, violence, wrath” and N. gorn (Ety/GOR). The names derived from this root include N. Huor “Heart-vigour”, Tuor “Strength-vigor”, Celegorn “*Swift-impetuous” (Ety/KHŌ-N, TUG, KYELEK). In later version of the Legendarium, Huor and Tuor were reconceived as being from the languages of Men (PM/348, 364 note #49) and (North) S. Celegorm was retranslated as “Hasty Riser”, with its final element being “riser” rather than “impetious” (PM/353). The eventual fate of this root is unclear, but since it remains the best source of words having to do with “haste”, I think it is best to retain this root for purposes of Neo-Eldarin.
A root in The Etymologies of the 1930s glossed “dread” with two variations: ᴹ√GOS and ᴹ√GOTH (Ety/GOS). It was an element in quite a few names in Silmarillion drafts from the 1930s, such as ᴹQ. Mandos “Dread Imprisoner” (Ety/MBAD), ᴹQ. Osse (Ety/GOS), N. Gothmog (Ety/MBAW), N. Tauros “Forest-Dread” (Ety/TÁWAR), and N. Dor-Daideloth “Land of the Shadow of Dread” (LR/405), the last of these containing N. deloth “abhorrence, detestation, loathing” = ᴹ√DYEL + ᴹ√GOTH (Ety/DYEL).
In Tolkien’s later writings, many of these names were given new forms or etymologies: Q. Mandos “Castle of Custody” = mando + osto (MR/350); Q. Ossë as an adaptation of his Valarin name Oš(o)šai (WJ/400); S. Tauron “Forester” (PM/358). This calls into question whether ᴹ√GOS or ᴹ√GOTH survived as a root. It does have a few useful derivatives for the purposes of Neo-Eldarin, however, such as N. gosta- “fear exceedingly”.
√GUR “hard, stiff, difficult, cumbrous, slow”
A root appearing in a set of documents referred to as Definitive Linguistic Notes (DLN) written around 1959, with the gloss “hard, difficult” (PE17/154). An apparent draft of this note had a longer gloss “hard, stiff, difficult, cumbrous, slow” for √GUR (PE17/172). In the more extensive notes, Tolkien gave a variety of Quenya (ur-) and Sindarin (gor-) derivatives of this root, but noted that “Sindarin owing to approach of √GUR- to other stems (as ÑGUR- “death”, ÑGOR- “terror, dread”) tends to use gor- in a very strong sense of things very painful and horrible to do; and uses dír- (tough) for lesser efforts” (PE17/154). As an example, Tolkien contrasted S. gorbedui “only to be said with horror or grief, lamentable to tell” versus S. dirbedui “hard to utter, difficult to pronounce”.
Thus it seems as a prefix, this root took on a distinctly negative tone in Sindarin, though some of its other derivatives were more neutral, as in S. gordh “difficult, laborious” and S. gorn “hard, stiff, thrawn”. The last of these was given as a word used for the Dwarves (PE17/46), although there were several other competing forms such as S. dern and S. dorn of similar meaning, all coined around 1959-60.