Light and Tree

A Survey Through the External History of Sindarin

by Helios De Rosario Martínez


There are certain lexical areas in which the Elvish languages created by J.R.R. Tolkien are especially rich. Light and trees, two motifs with a great importance in Tolkien’s sub-created world, from either the mythologic or the pictorial point of view, are indeed among these elements with permanent and extensive presence in the Elvish vocabulary.

Moreover, there seems to be some poetical connexion between light and trees. John R. R. Christie notes it in various poems and descriptions of characters and places throughout the entire legendarium, where light and trees (and other elements, such as eyes and hair) often interact, like sunlight filtering through leaves and branches, or are metaphorically connected (Christie and Shuttleworth, 2005). See for instance the first version of Kortirion among the Trees (I:25–26), where the twinkle of early stars is tangled in the bars of the trees and the seven lampads of the Silver Bear swing slowly in their shrouded hair (lines 28–31), or naked elms entwine … the Pleiades, and long-armed armed poplars bar the light of golden-rounded moons (lines 98–100). And of course, the mythical conception of the Moon and the Sun as the fruits of the Trees of Valinor may also be considered.

In Elvish linguistics there is also an apparent connexion, at least in some families of words related to light and trees, although in Tolkien’s published texts we cannot find explicit etymological links between them. In relation to this, the Sindarin word galad, that only in The Lord of the Rings occurs as derivations of originally different words with different meanings, in names such as Galadriel, Gil-galad, Caras Galadon or Galadrim, was the subject of a thread of discussion in the Lambenor list (in Spanish), resulting in the article El problema de Galad, published by this author in July 2001 (and printed in the journal of the Spanish Tolkien Society, Estel 38:31–36).

That article dealt chiefly with the Sindarin term mentioned in its title, but one could glimpse a long history behind it, that goes back to the stage of literary and linguistic creation in which the Sindarin tongue did not yet exist, when Noldorin and even earlier Gnomish occupied its place. Encouraged by Carl F. Hostetter, editor of Tengwestië, that study has been extended to cover these stages. In Spanish the article La historia de Galad, specifically concerning the term galad, was published in April 2004, as a continuation of the previous one. The article published here, however, is not limited to the term galad. This is a thorough analysis of the etymological families that include some words with the meanings of light and tree, in the evolution of the Elvish language inspired by Welsh, from the early Gnomish as conceived before 1917, to the latest stages of Sindarin in 1972.

This study reflects, through the history of a small set of words, the intricate scheme of the Elvish languages, the complex intertwining of their etymological roots, and how Tolkien’s conceptions were permanently flowing around some central immutable items (as the verbal root gil‑ shine white or silver in Gnomish and successive tongues, or the Q(u)enya noun alda tree), while most other words and etymologies were in a perpetual state of change, usually subtle, and in some instances striking, like the evolution of the legendarium in which these languages were set.

The analysis will be structured in sections, corresponding to three conceptual stages: Gnomish and early Noldorin, from the very first texts to approximately mid-twenties (the lingustic material of that stage having been thoroughly published in Parma Eldalamberon); Later Noldorin and early Sindarin covering the later stage prior to the publication of The Lord of the Rings or immediately after (with the Etymologies as the main published source of linguistic information of that stage); and Changes in later Sindarin, in which some details are noted, on the basis of published information dealing with that stage. After each section there is a chart with selected representative roots related to light and trees, and derived stems and words in Gnomish, Noldorin or Sindarin, as well as in Q(u)enya and the common ancestor to these languages, here labeled as Eldarin (Eld.).

Gnomish and early Noldorin stage

Words for light

In the earliest conception of the Elvish languages, there were several words related to the light in general, and to certain types of light, but the principal were probably those coming from the roots kal‑ and gal‑, obviously interrelated. In The Book of Lost Tales we may easily find names such as Kalaventë Ship of Light, Kalormë Hill-crest over which Sun rises or Galmir Goldgleamer (I:254, 256). Actually the Qenya Lexicon (henceforth QL, published in PE12) composed by Tolkien contemporaneously to those tales explicitly included the root KALA shine golden, with many words related to the daytime, the morning, light, etc., among which there is kala daytime, (sunlight), 12 hours (QL:44). On the other hand, the complementary Gnomish Lexicon (GL, published in PE11 together with the Gnomish Grammar, GG) had the verbal root gal‑ shine (golden as the Sun), from which are likewise derived a number of words related to the daylight, sunrise, daytime, etc. (GL:37). Some of these words are explicitly connected with the Qenya ones, as happens with gala light, daylight = Q kala, and glâ day, daytime (time Sun is above horizon) = Q kala < kalā & kălaȝ. (According to this etymological note, glâ could be interpreted as a syncopated form of earlier *galā).1

However, the roots kal‑ and gal‑ were apparently exclusive variants for Qenya and Gnomish, respectively: in this last tongue no word beginning with cal‑ was directly related to the light, day or Sun. It must be noted that there are many other cases of Gnomish words beginning with voiced stops, with cognates in Qenya that begin with its unvoiced counterpart. See, for instance:

  • Labial stops: Gn. Baul body, trunk cp. Q pūle, pulka in GL:22, and pulko body, trunk, bole of tree in QL:75.
  • Dental stops: Gn. high (adj. and av.) in GL:29 cp. Q. (1) adj. † high (2) av. high above, high up in QL:87.
  • Velar stops: Gn. gobos haven cp. Q kôpa, kôpas in GL:40, and kōpa haven, bay cited in QL:47.

There is in fact some indirect remainder of kal‑ in Gnomish, though it is probably a Qenya influence: Gn. alc, alchor (shrine, fane, temple) are cognates of Q alkar, alkarin, themselves seemingly derived from kal‑ after syncope and consonant transposition, cp. alka < ak’lā ray (GL:18, QL:30 s.v. ḶKḶḶK or KAL). But in contrast, there are Gn. glarosta‑, glartha‑ to dazzle, compared to Q alkar but apparently developed from syncopated gal‑ directly (GL:39); there are also Gn. agla flash, aglar glory, as probable cognates of Q alka, alkar (with the prefixed stem-vowel too).

It seems certain that gal‑ was a variant developed from kal‑. At least in one instance related with these roots, this is actually the type of evolution noted in Tolkien’s texts: gâl < kalda, or gaul a light. kāle (see the Gnomish Lexicon Slips — henceforth GLS, PE13:114). Such a change might have been influenced by the verbal root gil‑ gleam, shine pale and silver, as of the moon — the relation between both roots is noted in the entry gilm moonlight, silver light, which is compared with găla, daylight; and Gn. glôr gold and derivatives also might have had some influence. But it could be argued whether gal‑ was a later Gnomish variant, or the separation an Eldarin development. The first option seems most probable: the voicing k‑ > g‑ is a development parallel to that of the Gnomish soft mutation (cf. Derdzinski, 2004). But it could be that words from both gal‑ and kal‑ coexisted in Eldarin, and each subsequent language adopted only the phonologically preferred words. Patrick Wynne and Christopher Gilson suggest that at least the Eldarin word *galā could have existed, as it apparently occurs in Q otsola week < ot·g’lāta, cognate of Gn. ochlad << oglad, an idea supported by the existence of the Q root ALA, a variant of KALA in QL, perhaps originally conceived as cognate with an Eldarin root *GALA (PE14:16–17). This variant would have been developed like Q ‘il cp. Gn. gil (GL:38 s.v. Gilweth, QL:42 s.v. Ingil).

In writings later than The Book of Lost Tales, the language of the Gnomes or Noldoli (>> Noldor) was usually called Noldorin, instead of Gnomish (or Goldogrin or Noldorissa). This Noldorin language showed significant grammatical and phonological differences from the earlier Gnomish, but its vocabulary was very close to that observed in the Gnomish Lexicon (cf. PE13:119ff.). Hence the words related to light were still similar, though it is interesting to note that in this stage syncopated forms prevailed: see agladhren radiant, gladwen sunshine, etc. in the Noldorin Word-lists (PE13:136, 144), or AGLANN ray of light, glavaith (< *k’lamektā) a blaze, burning, blazing light, etc. in the Noldorin Dictionary (PE13:158, 162). In fact the only word in this stage that shows the full form of the root is galad dawn (PE13:144), which as far as we know occurs here for the first time,2 and which is very remarkable by itself: in this stage it was only another derivative of gal‑, but we will see that later it played an interesting role, in conjunction with other words.

Words for tree

In some texts contemporary to The Book of the Lost Tales we can find other Gnomish words that begin with gal‑, but have nothing to do with those related to the light. The Official Name List (ONL) connected to the Lost Tales includes Galdor lord of Thlim galdon (PE13:104). And the name Galdor was also included in the subsequent Name-list to the old tale of The Fall of Gondolin (NFG), where the name of his folk was originally written Nos Galdon, and it was added that (Galdon) is a tree, and thereto Galdor’s name akin. In emendations of NFG, however, Nos Galdon was changed to Nos nan Alwen, with a similar note explaining that alwen is a tree, though no etymological connection with Galdor’s name was then noted (PE15:24).

Let us first examine the versions prior to the change Galdon >> Alwen. The statement of NFG, that identifies Galdon with a tree and relates it with Galdor’s name, may also be applied to ONL. There the Eldarissa (i.e. Qenya) version of Galdor is given as Aldaron or Aldar, that contains Q alda tree (one of the Elvish words that Tolkien never changed), and Thlim galdon most likely meant *House of the Tree, the name of Galdor’s folk in Gondolin (cf. thlim race, folk as a suffix in GL:73). Galdon was probably a proper name derived from Gn. *gald‑ = Q alda, with the suffix ‑on that apparently also occurs in Gn. Baithon = Q Vaitya the Outer-Airs (GL:21), Gn. Ilon = Q Ilu Iluvatar, God (GL:50), etc.3 In this case, the word for tree in both languages should have come from a common Eldarin term, with an initial g‑ lost in Qenya as usual.

The change introduced in NFG, however, is a sign of a shift in the etymology of these words. The emended form is consistent with the Gnomish Lexicon, where galdon does not occur, but alwen tree does. And it is also a shift towards the Gnomish Grammar: according to it, Nos nan Alwen accords with the description there of the use of the Gn. article in genitive prevocalic form, nan· (GG:9); while its predecessor Nos Galdon could not be regularly formed from the word Galdon mentioned in NFG. It would be regular if it came from *Galda, which inflected into genitive would actually give Galdon (GG:13); but the wording of NFG seems to indicate that Galdon is the nominative form, and according to GG such a noun would be compounded by postposition, either with the genitive prefix, the article or nothing, but always with grammatical mutation, thus yielding *Nos a·Ngaldon, *Nos na·Ngaldon or *Nos ’Aldon, respectively, but not Nos Galdon (GG:8, 12).4

On the other hand, Q alda is cited in the Gnomish Lexicon, but in this case not related to any word with g‑. The cross-reference is actually from âl wood (material) — though its semantic cognate was †alwen tree, (GL:19). All related words begin likewise with a‑ (or o‑ < ā‑), and the only ones that begin with g‑ in GL and GLS are goloth (GL:41) and gawlas (GLS:114), both glossed forest, but in these cases the g‑ belonged to the collective/intensive prefix go‑, ŋwa‑ (GG:8).

However, we do find in GL some words beginning with c‑ such as the noun caltha‑ a plant, herb or calw green shoot, sapling, sprout (GL:25), that may be compared with the noun altha shoot, sprig, scion, sapling and the other a‑ words for tree previously mentioned. They are not the only words with c‑ (or k‑ in Qenya) with variants where this consonant is lost. There are also Gn. caithl, source, fount, origin and aithl a spring (cp. Q kektele and other words under KEKE(1) or KELE, KELU‑ and Ektele fountain, in QL). And more interesting, Q KALA is equated with ALA light in QL:29. This last case virtually coincides with that of the Gn. cal‑ vs. al‑ words related to plants and trees; therefore we find that, though Tolkien experimented with diverse families of Gn. words cognate with Q alda, all of them (gal‑, al‑ and cal‑) could be confounded with one or another Gnomish or Qenya root for light (gal‑, ALA or KALA, respectively).

Was this coincidence intentional? There was no apparent reason for an etymological or semantic connection. According to QL, the basic gloss of the root ALA(2) that yielded the words related to trees had nothing to do with light, but meant spread, and was related to broad things (cp. LAHA), such as the broad of the back (see Q aldo, aldamo, QL:29, or Gn. alm, GL:19, GLS:109). It was metaphorically related to trees, as growing things; cp. in GL the noun altha shoot, sprig, scion, sapling, and in GLS the verb altha‑ to shoot up, grow (high) or the adjective alw lofty, of living things: trees, men. And there was also a name for a specific tree, the elm (a broad tree, in fact), derived from it; see Q alalme and Gn. lalm, etc. in QL:29, GL:52. A similar metaphor was probably behind the other Gn. word for tree, orn (GL:62, GLS:115), cp. or‑ on, onto, up, in addition to, etc. and following entries in GL, and ORO(1) in QL:70. But nothing of this seems to be related to the light.

If there was some connection between tree and light, it perhaps was of a mythical type: note that the words cited related to light did not refer to any kind of light, but specially to the golden light of the Sun, which was (as the Elves knew) the last light of the Golden Tree of Valinor. However, nothing in the published texts points to this direction or to any other one.

In the later Noldorin Word-lists and Noldorin Dictionary the presence of these Gnomish words denoting trees and growing things decreases. In fact, the only entry glossed tree is orn (PE13:151, 164), and the words similar to alwen, etc. are reduced to alt branch, aulos forest (both deleted), and the base ALT bough, branch, that in fact is equated with Q alda (PE13:136, 137, 159). Ólin elm (PE13:151) is probably also related.

This could be an indication of a progressive change in the conception of the Noldorin etymology of these words. There is also a remarkable case that points to this direction, though it is only indirectly related to the words about trees. In the Noldorin Word-lists there was the entry all wide, broad < *aldá, equated to Q alda (PE13:136). It was doubtless related to that sense of ALA explained in QL. But it was then deleted and replaced by gall, which had the same gloss, while its etymology also changed to *ȝaldá, and the Q cognate to halda. We will see in the next section how this type of change was subsequently extended, and the g‑ finally reintroduced.

Selected roots, stems and words for light in this stage
Roots Stems and Words
Gn. gal‑ shine (golden as the Sun) Gn. gala light, daylight; glâ day, daytime; N. gladwen sunshine; galad dawn.
Q KALA shine golden Q kala daytime (sunlight); kalma (day)light; kalta‑ kindle, set light to; kalaina serene.
Eld. *KAL‑ Probable origin of terms from Q KALA and Gn. gal‑
Eld *GALA Eld. *g’lāta in ot·g’lāta, Gn. ochlad, oglad, Q otsola week.
Gn. gil‑ gleam, shine pale and silver, as of the moon Gn. gilm moonlight.
Selected roots, stems and words for tree in this stage
Roots Stems and Words
Gn. Galdon a tree; Galdor
Gn. *AL‑, cp. Q ALA Gn. alwen tree; âl wood (material); perh. altha‑ to shoot up, grow (high), alw lofty, of living things: trees, men.
Q ALA spread Q alda tree; alalme elm (tree).
Gn. caltha‑ a plant, herb; calw green shot, sapling, sprout.
Gn. orn tree

Later Noldorin and early Sindarin. The matter of galad

Words for light

When Tolkien wrote the Etymologies at the end of the thirties, some significant changes had been made. In Etym. we find two entries to GAL‑ glossed as variants of KAL‑, one deleted, glossed shine, and another (probably its substitute) accompanied by the extended form GALÁN‑ and glossed bright (VT45:13). The Noldorin derivatives are very close in form and meaning to the Gnomish words in gal‑, gla‑ referring to the light, and specifically to the sunlight, daytime, etc. The great difference with the earlier conception is that in Etym. KAL‑ and GAL‑ are not limited to one tongue; both Qenya and Noldorin (and even other languages as Doriathrin) have derivatives from both roots. There was, for instance, Q ala day from GAL1;5 and conversely under KAL‑ we find N calad light (likely < *kalatāˊ) or calen bright-coloured = green (< kalina). Both would become well-known Sindarin terms as well: calad occurs for instance in the Narn i Chîn Húrin, as a part of the cry of the Edain Lacho calad! Drego morn! (UT:65); and calen is part of names such as Calenardhon, Calenhad — and in lenited form Parth Galen, etc. — occurring in LR, among other places. The difference between GAL‑ and KAL‑ might be semantic: the terms under KAL‑ are not so related to the daylight as some under GAL1, and in fact the basic gloss of KAL‑ is shine (general word). The fact that this more specific meaning of GAL‑ was developed (or retained), was perhaps influenced by the root GLÁWAR‑, the N variation of LAWAR‑, light of the golden Tree Laurelin; cp. earlier LAURE (‘LOU̯RI‑, Gn. glôr) in QL:51; though there it was rather related to the metal, not to the light of the Tree. However, Christopher Tolkien noted that the idea that GAL‑ was a Noldorin variant of KAL‑ was rejected in the text under this last base, and as he points out, it is not clear at this stage how these roots were related.

But regardless of the precise relation between the roots, it is especially interesting to compare the aforementioned calad with galad, which occurs also glossed as light under the deleted base GAL‑. The lenited form of calad would be virtually indistinguishable from galad, as happens in two compounds found in Etym.: gilgalad starlight (s.v. the base GIL‑ shine (white or pale), virtually the same as the Gnomish root mentioned above) and Cilgalad (N cognate of Q Kalakilya Pass of Light, cf. KIL‑).6 The consonant mutation pattern in Noldorin as observed in Etym. is a complex matter that concerns various types of phonological changes, many possible circumstances that may condition their development, and many irregularities, either real or apparent. Therefore, no easy rules may be defined in order to ascertain whether gilgalad and Cilgalad are formed by unmodified galad or lenited calad. However, the second option is for various reasons the most probable in both cases.

In the case of gilgalad, there is another compound morphologically alike, and even semantically very similar, that is therefore a reliable instance for making comparisons. It is gildin silver spark; and its occurrence under the base TIN‑, just after N tinw spark, small star, tint spark positively indicates that the second part of the compound underwent the lenition t‑ > d‑. Hence, it is most likely that gilgalad here also has lenited calad.

Regarding Cilgalad, comparisons with other cases are also favourable to the hypothesis of lenited calad. First, its Qenya cognate Kalakilya shows kala‑, though parallel names in different languages may have different etymologies. But the comparison with the case of gilgalad also gives phonological support to this idea: both compounds share the same suffixed element (‑galad), and the prefixed elements differ only in the first consonant, probably even in their primitive form — most likely *kilya‑ and *gilya‑ (cp. primitive *gilya for N gíl star s.v. GIL‑, and Q kilya kilya cleft, pass between hills, gorge before N cîl); so the phonological development of the suffixed element should be the same in both compounds. And from a syntactical view, it may be noted that similar compounds, with the modifier element positioned after the modified noun (a normal pattern in Noldorin, though less frequent than the opposite, at least in Etym.), also show soft mutation of the second element, in the vast majority of cases (cp. Dagor Vregedúr Battle of Sudden Fire s.v. BERÉK‑, Amon Uilos Mount Everlasting-snow s.v. GEY‑ or Eredwethion < ered + gwethion *Mountains of Shade s.v. WATH‑, inter alia).

After The Lord of the Rings was completed, Noldorin was replaced by Sindarin, but though this was a great change in the conception of the Elvish languages, it does not seem to have substantially altered galad, calad and other related words. This can be observed in the linguistic explanations that Tolkien gave in the late fifties of the names of two important Elves: Galadriel and Gil-galad.

Regarding the first one, he wrote in a note to the later version of the Quenta Silmarillion that it meant Lady with garland of sunlight, and in its primitive form had been galata-rīg-elle (X:182). From this we may corroborate that the noun galad still existed in Sindarin, and that it was derived from *galata. Note also that the meaning sunlight is more specific than the bare gloss light in Etym., though it is consistent with other words derived from the same root (see above), and such an idea goes back to the terms derived from gal‑ in the Gnomish Lexicon.

Regarding Gil-galad, Tolkien wrote to Rhona Beare in 1958 that it meant star-light, and that it was a case of grammatical change or k, c > g in Grey-elven, seen in the initials of words in composition or after closely connected particles (L:279), which basically coincides with the meaning and etymology suggested above for gilgalad in Etym. In that letter he also wrote about the stem kal used for words referring to light, and about the more or less synonymous stem gal (corresponding to gil which only applied to white or silver light), all of them concepts identical to those found in Etym.

Words for tree

When comparing Etym. with the earlier lexicons, a major change of a number of interconnected roots and derived terms is observed, in the middle of which are Q alda tree and alalme elm, virtually untouched since they first emerged in the Qenya Lexicon (see above), while everything around them was rearranged into a new set of roots.

No direct successor of the root ALA spread from QL is found in Etym., but there is the pair of interconnected bases AL‑ and GAL‑ (cf. VT45:5). The first one is unglossed but includes words related to riches, good fortune and bliss. The second one is glossed grow, spread, increase, thrive, prosper, be healthy, be glad, etc., depending on the version.7 Spread is in fact among these glosses, but the other meanings of the new bases do not resemble so much those of the old ALA, but rather those of the Gn. root gwal‑ fortune, happiness (a blessed fortune, related to the Valar, whose Gnomish name Gwala came in that stage from such a root; cf. GL:44). It is not certain whether the underlying root GAL(A)‑ was a Noldorin variant of AL(A)‑ with a false g‑ (as was usual in that language), or conversely AL(A)‑ was a Qenya variant of GAL(A)‑. A deleted note under GALÁS‑ explicitly states the former, but the apostrophe in Q ’al after GALA‑, apparently the Qenya cognate stem, suggests the latter. Another evidence for assessing this question would be the base LAR‑, LAS‑, originally related to bliss and fortune, though later only to richness and fatness (VT45:26). However, this is also a contradictory clue: it is cross-referenced to GALAS‑ (marked with a question mark), but a variation characterised by the loss of the entire first syllable is rare, and it could be also expected to come from *ALAS‑ (cp. ALA‑, LA‑ no, not; ÁLAK‑, LAK2 rushing and swift; etc.).

Other roots and words that in the earlier stage were related to ALA were still present in Etym., though such a relationship was now less evident, if it existed at all. Q alalme elm-tree, for instance, is found in QL, but in Etym. its root was LÁLAM, and apparently bore no direct relationship to the previous ones.8 LAHA and other roots related to open spaces, moors, to spread, extend, etc. (QL:50) are reflected in the variants DAL‑, LAD‑, LAT‑ in Etym., but no relation with ALA was noted at this later stage. It is also interesting to see the evolution of another word that years before had started its own course: N all >> gall wide, broad and its primitive form *ȝaldá (see above) are probably the predecessors of the root KHAL2 in Etym., here glossed uplift, whose derivatives include the N adjective hall exalted, high < *khalnāˊ. No relationship to the other roots in Etym. is noted for this one, either.

But among these exceedingly mutable roots, there was GÁLAD‑ tree, the one to which we have been leading. The relation with the abovementioned bases is clear from words such as galw growth, galas growth, plant or galo‑ to grow, in both deleted and retained versions of GAL‑, GALA‑, and some explicit cross-references. But in comparison with the others, this root was quite stable and simple. It was written down in two very similar versions, though neither was deleted or altered, and together only comprised one word for tree in each of Qenya, Noldorin and Doriathrin: alda, galadh and gald, respectively; and some nouns or names that contained it. One of these names is Galdor (which in the contemporary Quenta Silmarillion was a Man’s name), which shows how this root beginning with g‑ was a shift back to the old conception, when the Gnomish word for tree was galdon.

And the fact is that at least Q(u)enya alda and Noldorin >> Sindarin galadh tree, and their newly devised etymology, would be kept for many years; in fact there is no sign of further change in the published materials. In The Notion Club Papers (c. 1945) Arundel Lowdham reproduces Alboin Errol’s explanation about these Elvish nouns (V:41), providing moreover their primitive form, galadā, which would fit perfectly in Etym., though it is not recorded there. Q orne and N orn (pl. yrn), the other word for tree (cp. Gn. and early N orn above), is also explained in The Notion Club Papers, as coming from ornē (pl. ornei), and specially applied to smaller and more slender [trees] like a birch or rowan (IX:302); in Etym. the forms coincide, but it was rather applied to a high, isolated tree (s.v. ÓR-NI‑). This linguistic explanation of galadh and orn (and their Quenya cognates) would be repeated in later texts, with identical primitive forms, and only slight variations concerning the type of tree that each term described: In UT:266 Christopher Tolkien mentions that in the latest writings his father noted that ornē was originally applied to straighter and more slender trees such as birches, and galadā great growth to stouter, more spreading trees such as oaks and beeches; but that the former fell out of use in Sindarin, and hence all trees came to be called galadh. Similarly, in L:426 (as late as 1972) Tolkien explained that galadā came from GAL grow, and ornē from OR/RO rise up, go high.

Selected roots, stems and words for light in this stage
Roots Stems and Words
Eld. KAL‑ shine (general word) N calad light, N gilgalad, S Gil-galad starlight; calen bright-coloured, green; Q kala light; kalma a light; kalta‑ shine; kalina light (adj.).
Eld. GAL‑ shine, variant of KAL‑ N galad light; glan clear and daylight; Eld. gala, Q ala day; Q alan daytime; Eld. galata sunlight in S Galadriel.
Eld. GIL‑ shine (pale or white) Eld. gilya, N gíl star, cf. gilgalad starlight. Q Ilma starlight.
Selected roots, stems and words for tree in this stage
Roots Stems and Words
Eld. GÁLAD‑ tree, perh. related to GAL(AS)‑, GALÁS‑ grow, thrive, prosper, be glad, etc. Eld. galadā tree, N, S galadh, Q alda tree; S Galad(h)rim *tree-folk.
Eld. ÁLAM‑, LÁLAM‑ elm-tree N lalf, lalven, lhalwen, Q alalme, lalme.
Eld. ÓR-NI‑ high tree Eld. ornē tree (small and slender), N, S orn, Q orne high isolated tree.

Galad vs. galadh: an orthographical vacillation

Galadh would be the cause of further false occurrences of galad. There should not be reason for such a confusion, since d and dh represent different phonemes in the Elvish languages; but when the first edition of LR was published, Tolkien leveled their spelling to d, since he thought that readers would find dh or ð uncouth, and for instance he wrote Caras Galadon and Galadrim, though their regular spelling would have been Caras Galadhon City of the Trees and Galadhrim Tree-folk. This levelling may also be observed in the drafts of the Appendices to LR, in names that were not finally published: In XII:260 it is explained that the queen Vidumavi (an adaptation from Gothic Widumawi Wood-maid, cf. UT:311) was called Galadwen in Gondor, although we would have expected it to be *Galadhwen. This could also be the reason for the name of Galador, the son of Imrazôr and Mithrellas, and first Lord of Dol Amroth (UT:248, XII:220–221); there is no actual evidence that this name is related to galadh, but it is interesting to compare it with Galadhor in Etym. s.v. GALA‑ and GÁLAD‑ (although here it is not meant to be related in any way with the Lord of Dol Amroth, a place whose existence was not yet conceived).

But at least for Caras Galadon and Galadrim, an inner explanation was also provided by Tolkien. According to the Appendices, Caras Galadhon was not a genuine Sindarin name, but the Sindarinization of a Silvan name (LR:1101). And in X:182 we learn that galad was the Silvan cognate of S galað. From this we may infer that Caras Galadon would not be an irregular spelling, but the actual Silvan name. Had Galadwen and Galador been published in the Appendixes, a similar explanation could also have justified their apparently heterodox spelling, for they could be Gondorian variants, even influenced by the Silvan tongue in these cases of a woman from Rhovanion and the son of a Silvan maid.

In the note concerning Silvan galad tree, however, Tolkien’s intention was not to explain why galad occurred in names that should have galadh, but the opposite. The name to which this happened was Galadriel, which, as stated above, actually contained Sindarin galad light according to that note. Elsewhere Tolkien noted that many associated her name with that of the Galadhrim’, so that among those whose memories of the ancient days and Galadriel’s history had grown dim her name was often altered to Galadhriel (UT:266, cp. S:436). But in fact what Tolkien was describing was his own experience! In the first drafts in which the Lady appeared, her name was changed from Finduilas to Rhien >> Galdrien >> Galaðrien < Galað-rhien, meaning (tree)-lady (VII:249). And though in the manuscripts of LR it was subsequently changed to Galadrien and finally to Galadriel, in later texts such as The Annals of Aman the form Galaðriel is found instead. Moreover, J.R.R. Tolkien. Life and Legend shows a draft of the annotated Namárië later published in The Road Goes Ever On (p. 84), in which the subtitle reads Aldariello nainië Lóriendesse, translated as Galaðriel’s lament in Lórien, in contrast with the published subtitle Altariello nainie Lóriendesse, Galadriel’s lament in Lórien. Therefore, the note where Tolkien mentions the Silvan variant galad seems to have been written as an attempt to resolve this doubt about the name of Galadriel.

Changes in later Sindarin

Though future publication of later texts may reveal subsequent changes here unnoticed, we have seen that nothing substantial was apparently altered about the words for tree, after the great change observed in Etym.9 And the words for light were not much changed in form either (the names of Galadriel and Gil-galad were in fact set in stone, as it were, by their publication in LR); but their etymology did undergo some revision.

In The Shibboleth of Fëanor, a text of the late sixties or early seventies that dealt with many names of the Noldor, we see that the root of the first part of Galadriel was no longer GAL, but ÑAL (ñ being used for the sound [ŋ]), and from this we can reconstruct the Primitive form *ñalatā-rīg-elle. With this variation Tolkien perhaps intended to make the roots for light and tree more different than before. On the other hand, the meaning of ÑAL was specifically shine by reflection (XII:347); so, in this modern version, the meaning of Galadriel is also different, Maiden crowned by a garland of bright radiance, not by sunlight as before. Likewise, the etymology and meaning of the root GIL and the word gil star, which had previously been stated to be related to GAL, was also reset: according to another late text, Sindarin gil came from the root ngil silver glint, which would correspond to a base *ÑIL (X:388). This shift GIL >> *ÑIL, identical to that of GAL >> ÑAL, seems to reveal that gil kept its former relation to other terms related to light, although the form of the stems was modified; its meaning was then sparkle with silver light, very similar (although some subtle difference may be found) to previous shine white or pale.

Note also that such a shift would not affect the Sindarin forms substantially, since both g‑ and ñ‑ at the beginning of a Primitive word give g‑ in Sindarin (except in compounds and other constructions); but it is different in Quenya, since in this tongue ñ‑ gave n‑. So related words in Quenya should start with n‑, as seen in the true Quenya name of Galadriel, stated to be Ñaltariel (in post-exilic Quenya it would be Naltariel, as Ñoldor became Noldor, etc.), although she was given the name Altariel by influence of Alatāriel(lë), its form in Telerin, the language in which her name was first devised.

Finally, we must also comment on a change suffered by the name Gil-galad: In the same text in which it is said that Galadriel came from the stem ÑAL, it is also noted that this stem is likewise present in Gil-galad. So this name’s structure was essentially changed, because now it was not Starlight (being gil star, the modifier, and calad, the main element), but Star of radiance (gil star as main element, modified by galad radiance). On the other hand, in this modern version of the name no lenition occurs to the initial consonant, as it did before. This was also explicitly commented on by Tolkien in the letter of 1972 mentioned above: there he said that there was no mutation, among other cases, when a name is virtually an adjective, as in Gil-galad (L:426). This does not mean, of course, that calad and other words related to the root KAL were not retained in the Elvish vocabulary, but simply that they do not occur in these specific names.

Selected roots, stems and words for light in this stage
Roots Stems and Words
Probably Eld. KAL‑, as previously See previous stages.
Eld. ÑAL shine by reflection Eld. *ñalatā in Q Ñaltariel, S Galadriel; S Gil-galad.
ngil, *ÑIL silver glint S gil star, cf. Gil-galad.
Selected roots, stems and words for tree in this stage
Roots Stems and Words
Eld. GAL grow Eld. galadā, S galadh, Q alda lit. great growth, tree, stouter and more spreading.
Eld. OR/RO rise up, go high Eld. ornē, S orn, Q orne tree, straighter and more slender.

Complications in the Appendix of The Silmarillion

Probably, one of the first approaches of any Tolkien reader to Elvish linguistics is, after the Appendices of The Lord of the Rings, the Appendix on Elements in Quenya and Sindarin names of The Silmarillion. When comparing that book with the raw material in The History of Middle-earth from which it was edited, it becomes quite clear that Christopher Tolkien tried to represent the latest conceptions, both of historical and linguistic motifs when possible, in order to produce the most coherent and internally self-consistent narrative (S:vi). But this was quite difficult, especially when dealing with elements so thoroughly interconnected and complicated as the names we have studied here and their etymologies.

We may mention first the case of the entry alda, related to the Sindarin form galadh, where the names Caras Galadhon and Galadhrim were spelt in the first editions Galadon, Galadrim, as his father wrote them in the first edition of LR. However, this mistake Christopher Tolkien noted in UT and emended in later editions.

But more subtle is the case of the entry kal‑ (gal‑). It is not really inconsistent with the latest conception described in the previous section, but is nearer to the conception of the early stages of Sindarin, for the primitive form ñal‑ (or ngal‑ with the orthography used in that Appendix) was not given for the Sindarin variant gal‑; and the explanation about Galadriel was adapted by Christopher Tolkien from the note of the Quenta Silmarillion, in which it was told that her High-elven name was Altariellë (referred to primitive galata-rīg-elle). In fact, studies prior to the publication of The History of Middle-earth deduced from this entry that the Common Eldarin form of al(a)ta was most likely *galatá (PE6:9), which matches the conception in the early stages of Sindarin, but not the latest.

However, the explanation presented in Quenta Silmarillion was slightly changed in the Appendix of The Silmarillion, obviously pointing to the latest conception: the terms and glosses of the elements (especially alata radiance) are taken from the later discussion of this name in The Shibboleth of Fëanor, and Altariellë was replaced in the Appendix by Al(a)táriel, reflecting together Telerin Alatāriel(lë) and its form as adapted into Quenya, Altariel, as given in the Shibboleth (though the Quenya form in the Shibboleth, Ñaltariel, was not mentioned). It may be noted that the adaptation of the earlier explanation of the name is especially canny in keeping the term High-elven for the language of the alternative form of the name (Altariellë in the original, Al(a)táriel in the adaptation). In the original version High-elven most certainly meant Quenya, as it means throughout the whole narration of The Silmarillion (see S:418, s.v. Quenya), but in the adapted explanation of the Appendix it could be interpreted as either Quenya or Telerin.

One may wonder why Christopher Tolkien chose to adapt an earlier text about Galadriel’s name to represent the latest conception, instead of directly citing the latest text. There are various possible reasons: maybe it was just a matter of simplicity, since with the note from the Quenta Silmarillion he was able to elude the entire matter of the root ÑAL and its distinct evolution in the Elvish languages, not needed for a coherent and self-consistent account of the published forms of the name (Sindarin Galadriel, and Quenya Altariel in R:66 — actually its genitive form Altariello). But it could also have been for contextual reasons, since The Silmarillion to which the Appendix was attached was to a great extent based on the Quenta Silmarillion, from which the note was adapted, while The Shibboleth of Fëanor was a purely philological essay, a source of interesting details that were added to the narrative, but not a primary source.

1. The entry glâ was rejected in the pencil layer of the manuscript. For further discussion about this see PE14:16, and Lambengolmor list messages 702704.

2. galad could conceivably have existed prior to the Noldorin Word-lists. In GL:62 we can find the abovementioned term ochlad week, formerly written as oglad, for which Tolkien also provided its older form ot·glāta. Wynne and Gilson explained in PE14:16 that the precise reading of the etymology is in fact ot·g’lāta, and that it probably means *(a period of) seven days, while deducing that the second part of the word, g’lāta, came from earlier *galāta, or maybe even older *kalāta. It may be seen that, had a variant with different accentuation or lengths of the vowels existed, the root-vowel could have been preserved, and the resultant word would have been actually galad.

3. The Gnomish Grammar gives distinct grammatical functions to the suffix ‑on: a genitive morpheme for nouns ending in a vowel (GG:13), and a plural ending of adjectives ending in ‑a or ‑u (GG:15). However, in the Gnomish Lexicon it is revealed as a far more productive element. Among other functions, here we see that it characterised some proper names, perhaps of ancient or poetic motifs: Sirion is glossed in GL:67 as †river — properly name of the famous magic river…, cp. Sîr simply glossed as river.

4. This indicates that the original version of NFG may have been written before GG and GL, though there cannot be much difference in time between their composition, if they were not contemporary: Tolkien was working on GL already in 1917 (PE11:4), while the manuscript of ONL, the predecesor of NFG, physically follows the Poetic and Mythologic Words of Eldarissa (PE13:100), which must have been composed after 1916–17 (PE12:xxi). There are other changes in NFG that suggest that the first layer of NFG preceded the Gnomish Grammar. See for instance Los Glōriol >> Los ’lōriol (II:216, PE15:25): both variants of the name transparently mean *Golden Flower, the name of the House of Glorfindel; cp. lôs a flower and glôriol golden, like gold in the Gnomish Lexicon. However, GG:12, 15 indicate that adjectives postpositioned to nouns are usually mutated, which in NFG occurs only in the emended forms. Amon Gwareth >> Amon ’wareth hill of vigilance (PE15:20), or Falas a·Gwilb >> Falas ’Wilb beach of peace (PE15:24) are changes of the same type.

5. And in the title page of Etym. there is *gala day: Q ala a day. (VT45:13). Cp. this primitive form with the Eldarin *galā > *g’lā > Q ‑la *a day in otsola week, as suggested by Wynne and Gilson and mentioned in the previous section.

6. In V:362 s.v. KAL‑ we find a cross-reference to the name Gilgalad, next to calad; this would lead one to think that this name shows a lenited form of calad. However, in A&C Hostetter and Wynne note that the cross-reference in the MS is in fact to Glingal (VT45:19).

7. It is difficult to follow the trail of this root, repeatedly revised in the manuscript. Apart from the cross-reference in AL‑, we can find two rejected versions of GAL‑: one marked with a superscript 2 (apparently for distinguishing it from GAL1 mentioned in the previous section), glossed grow, spread, increase, and a longer one that was changed to GAL(AS)‑, and originally glossed grow, thrive, prosper, be healthy, be glad, etc. The contents of the latter were nearly as many as those of AL‑. All these mentioned roots were rejected, and some contents of GAL(AS)‑ were marked for incorporation into the replacement root GALA‑, the only one included by Christopher Tolkien in The Lost Road. Obviously related, there was also GALÁS‑ with some derivatives glossed joy, that was included by Christopher Tolkien too (VT45:13–14).

8. It is interesting to see that Gn. lalm, cognate of Q alalme, occurs also in Etym., though it is not a Noldorin word, but Doriathrin. On the other hand, LÁLAM‑ does have an indirect relation to AL(A)‑, etc.: about its variant ÁLAM‑ it is said that some hold it related to ALA‑ since the elm was held blessed and beloved by the Eldar.

9. Only in the secondary Elvish tongues do we find new forms, due to the changes in the history of the languages. For instance, Telerin became more important for its great relation to Sindarin; and of this tongue Tolkien wrote that the cognate of Q alda and S galadh was galla (VT39:7).

Christie, John R.R. and Rebekah Shuttleworth. Galadhremmin Ennorath. Lecture at Tolkien 2005 — The Ring Goes Ever On. August 14th, 2005. Birmingham.

Derdzinski, Ryszard. Consonants mutations in conceptual evolution of Noldorin/Sindarin phonology. Published in Gwaith-i-Phethdain, April 14, 2004.

De Rosario, Helios. El problema de Galad. Published in Lambenor, July 2001, last revised in February 2004.

——— La historia de Galad. Published in Lambenor, April 2004, last revised in July 2004.

——— What did replace glâ in GL? Post to the Lambengolmor list, message no. 702, June 28, 2004. Reply by Christopher Gilson in message no. 704.

——— calad or galad? Post to the Lambengolmor list, message no. 706, July 4, 2004. Various replies by David Kiltz in message no. 709 and following.

Hostetter, Carl F. and Patrick H. Wynne. Addenda and Corrigenda to the Etymologies. Published in Vinyar Tengwar 45 and 46, Nov. 2003 and July 2004.

Priestman, Judith, ed. J.R.R. Tolkien. Life and Legend. An Exhibition to Commemorate the Centenary of the Birth of J.R.R. Tolkien (1892–1973). Bodleian Library. Oxford. 1992.

Tolkien, J.R.R. Early Noldorin Fragments. Edited by Christopher Gilson, Bill Welden, Carl F. Hostetter and Patrick Wynne. Published in Parma Eldalamberon 13. 2001. 92–165.

——— Name-list to The Fall of Gondolin. Edited by Christopher Gilson and Patrick H. Wynne. Published in Parma Eldalamberon 15. 2004. 19–29.

Welden, Bill. The Role of Telerin in the Development of Sindarin. Published in Parma Eldalamberon 6. 1983. 6–14, 28.

See also the general Tengwestië Bibliography.

  • 2008-11-08 14:27:38: Formatting changes only: converted text to Gentium/Basic, deprecated all Gentium tags, converted combining diacritics to modifiers where possible

Copyright ©2005 Helios De Rosario Martínez

Quotations from the works of J.R.R. or Christopher Tolkien are the copyright of their publishers and/or the Tolkien Estate, and are used here with their kind permission. The word TOLKIEN is a registered trademark of The J.R.R. Tolkien Estate Limited. The characters and scripts of Tolkien’s invented languages and works in those languages are the copyright of the Tolkien Estate.


First published on November 22nd, 2005

Last modified on Friday, November 5th, 2010 9:18 am PDT

This page has been visited 11588 times since November 22nd, 2005

Page design by and copyright ©2005 Carl F. Hostetter

Site owned and operated by Carl F. Hostetter