New Theme! What do you think?

Study, speak, and hang out with fellow Elvish students!

Elvish Roots – Ideas from Early Works

So far in this series I’ve focused on the Primitive Elvish of Tolkien’s later works, from the 1930s to the 1960s. This is because in the early to mid-1930s, Tolkien went through a major overhaul of Primitive Elvish, and the conceptual framework for his languages’ development was very different prior to that point. This Early Period of the 1910s to 1920s needs to be discussed separately.

Before I get into it, though, it’s worth answering the question “Why look at Tolkien’s early works at all?” Most Neo-Elvish writers (myself included) are primarily interested in Tolkien’s later ideas on his languages, especially those from around the time The Lord of the Rings was published, and afterwards. Many of Tolkien’s earliest ideas are incompatible with his later conception of the languages. Despite this, I think the earliest works are worth examining.

The Early Period of the 1910s and 20s seems to be the only period of Tolkien’s life where he attempted to make his languages function as actual languages. The Early Qenya and Gnomish Lexicons of the 1910s have vocabulary spanning the whole breadth of human (and Elvish) experience, and so were much more practical for ordinary day-to-day communication. In Tolkien’s later writings, the Elvish languages became more intertwined with his stories. Much of the later vocabulary he invented seems to be focused on explaining the Elvish names of characters and places in his stories, as well as supporting the few examples of Elvish speech and poetry that appear in his writing. There is nothing with the same breadth as the early lexicons in Tolkien’s later writings (the Etymologies of the 1930s is the closest, but even that is more limited in scope).

That means that there are many words and concepts for which we have no words later than Tolkien’s oldest ideas from the 1910s and 1920s. For example, this is true for all words having to do with “cruder” bodily functions like sexual intercourse and defecation. There are times when digging through the older material is the only available option for certain concepts. In some cases the earlier material simply can’t be made consistent with Tolkien’s later linguistic aesthetics, but there are other case where early words can be salvaged (though which words fall into which case can be a matter of dispute among Neo-Elvish writers).

Having addressed why the the earliest material is worth looking at, let me outline Tolkien’s earliest conception of Primitive Elvish, and how it differed from his later ideas. Unfortunately, the remainder of this article gets pretty technical.

As mentioned in previous articles, Tolkien’s later conception of Primitive Elvish included three “series” of sounds: the p-series (labials), t-series (dentals) and k-series (velars). As variants of the t-series and k-series, Tolkien also had palatalized consonants (with added y-sound) and labialized consonants (with added w-sound). In Tolkien’s earliest writings, however, he treated these palatalized and labialized consonants sounds as their own series, resulting in five series total: p-series, t-series, c-series, k-series and q-series. The clearest description appears in the Qenyaqetsa (published in Parma Eldalamberon #12), which includes a description of Elvish phonology from the 1910s.

The q-series are simply labializeds velars, much as they appeared in Tolkien’s later writings. Tolkien’s notation for the c-series is a bit ambiguous, but I am of the opinion that they were probably pure palatal sounds. Such sounds are rare but possible in human languages, although they don’t exist in English. In IPA notation the series are:

p-series t-series c-series k-series q-series
1. [p] [t] [c] [k] [kʷ]
2. [b] [d] [ɟ] [g] [gʷ]
3. [ɸ] [θ] [ç] [x] [xʷ]
4. [β] [ð] [ʝ] [ɣ] [ɣʷ]
5. [m] [n] [ɲ] [ŋ] [ŋʷ]
6. [mb] [nd] [ɲɟ] [ŋg] [ŋgʷ]

In Tolkien’s own notation (or as close as I can get to it using only unicode characters):

p-series t-series c-series k-series q-series
1. p t c or ty k q or kw
2. b d j or dy g gw
3. or f þ χ̑ or þy x xw
4. ƀ or v đ ɣ̑ or zy ɣ ɣw
5. m n ñ ŋ ŋw
6. mb nd ñj ŋg ŋgw

Note that in this Early Period, Tolkien used ñ to represent a palatal nasal (IPA [ɲ]), as opposed to the 1930s and later when he used it to represent the velar nasal [ŋ]. This was probably because ñ was easier to produce on 20th-century typewriters than either [ɲ] or [ŋ].

Tolkien also included the sounds r, l; s, z; w, y (IPA [j]) in the Primitive Elvish of the 1910s, but they were outside this system of series and grades. Three of these Early Period grades also appear in Tolkien’s later writings: #1 (voiceless stops), #2 (voice stops) and #5 (nasals). Grade #6 (nasalized stops) is analogous to the strengthened nasals and voiced stops of Tolkien’s later writing. The remaining two grades, #3 (voiceless spirants) and #4 (voiced spirants) were unique to the Early Period. Grade #3 corresponds to the voiceless aspirates of Tolkien’s later conception of the languages (pʰ, tʰ, kʰ). Grade #4, the voiced spirants, didn’t exist in any form in Tolkien’s later conception of primitive Elvish, though in terms of their phonetic evolution they are closest to the voice stops in Tolkien’s later works (see below).

Puzzling out the phonetic development in the child languages of the Early Period (Gnomish and Early Qenya) is challenging, since Tolkien organized the languages into two separate lexicons. Furthermore root forms appeared only in the Qenya Lexicon, not the Gnomish lexicon. The notations Tolkien used for primitive roots in the Qenya Lexicon are not precisely same as given above from the Qenyaqetsa. This means there are some additional phonetic symbols you need understand from the Early Period:

  • Instead of đ/Đ Tolkien often wrote ř/Ř since the voice dental spirant usually developed into r in Qenya.
  • For palatal voiced spirant [ʝ] Tolkien often wrote ẏ/Ẏ since these usually developed into y in Qenya.
  • Likewise, for labialized velar voiced spirant [ɣʷ] Tolkien wrote ẇ/Ẇ since this usually became w.
  • For simple velar voiced spirant [ɣ] Tolkien often wrote ʒ/Ʒ as he did in his later writings.
  • For the voiceless labialized velar spirant [xʷ], Tolkien often wrote ƕ/Ƕ.
  • For the voiceless palatal spirant [ç] Tolkien often wrote hy, the Qenya sound it ultimately developed into.
  • Similarly, for the other palatals, Tolkien generally wrote ty, dy, ny.
  • For roots beginning with primitive voiced stops, Tolkien often wrote the voiceless stops they ultimately developed into in Qenya.

Thus, the symbols as Tolkien generally used for primitive roots in the Qenya Lexicon are as follows. These are mostly (though not entirely) the Kor-Eldarin form of the roots, as described on Parma Eldalamberon #12, p. 16. The modified root forms in the Qenya Lexicon partially reflect the Qenya sound developments rather than being a pure representation of the original primitive roots:

p-series t-series c-series k-series q-series
1. p t ty k q
2. p [b] t [d] dy k [g] q [gw]
3. f þ hy x xw
4. v đ or ř ʒ
5. m n ny ŋ ŋw
6. mb nd ndy ŋg ŋgw

As partially represented in the above chart, the basic Early Qenya developments are as follows:

  • Voiceless Stops: Voiceless stops generally survived unchanged.
  • Voiced Stops: Medial voiced stops merged with voiced spirants as they did in later iterations Quenya, with the same developments as given below. Initially, however, they became voiceless stops. Tolkien rarely showed any voiced stops in roots within the Qenya Lexicon, often representing them as the voiceless stops they ultimately became in Early Qenya. The main way this to detect this change is by comparing the frequent b/p, d/t, g/k, gw/q variations in Gnomish vs. Early Qenya, where the Gnomish forms reflect the original voiced stop.
  • Voiceless Spirants: These generally weakened in Early Qenya, with [ç] > voiceless y (hy), [x] > [h], [θ] > [s], but the voiceless labial spirant [ɸ] ( or f) survived as the voiceless spirant [f]. The voiceless labialized velar spirant [xʷ] (ƕ) also usually became f, but it became h before u. There are mostly the same as the aspirate developments in later versions Quenya.
  • Voiced Spirants: These also weakened in Early Qenya, with [ʝ] () > [j] (y), [ɣʷ] () > [w], and [ɣ] (ʒ) vanishing. However, the voice labial spirant [β] (ƀ or v) survived as the voiced spirant [v]. The voiced dental spirant [ð] (đ or ř) > [z] > [r] medially, but became [s] (also from [z]?) initially.
  • Nasals: These generally survived unchanged except that [ŋ] became [n] initially and [ŋg] medially.
  • Nasalized stops: These survived medially but became simple nasals initially, as with later version of Quenya.
  • Palatals: Some palatal consonants became palatalized dentals in Early Qenya: [c] > [tʲ] (ty) and [ɲ] > [nʲ] (ny), but the voiced palatal stop and spirant [ɟ] (dy) and [ʝ] () both became [j] (y), and the voiceless spirant [ç] became voiceless y (hy).
  • Others: The other sounds r, l; s, z; w, y mostly survived, except that generally [w] > [v] and [z] > [r] as was the case in later versions of Quenya as well. Final [s] became [r], but not intervocalic [s], which was the reverse of later Quenya development.

The basic Gnomish developments are as follows:

  • Voiceless Stops: Voiceless stops survived unchanged initially but generally became voiced stops medially in Gnomish, as also happened in Noldorin/Sindarin.
  • Voiced Stops: Voiced stops survived unchanged initially but generally became voiced spirants medially in Gnomish, as also happened in Noldorin/Sindarin.
  • Voiceless Spirants: These survived mostly unchanged in Gnomish, just as aspirates became voiceless spirants in Noldorin/Sindarin. Initial [x-] then became [h-], as was also the case in Noldorin/Sindarin.
  • Voiced Spirants: It seems voiced spirants most survived medially, but became voiced stops initially in Gnomish. The main way to distinguish these is the initial b/v, d/s, g/-, gw/v pairings in Gnomish vs. Qenya.
  • Nasals: These mostly survived unchanged in Gnomish except that [ŋ] became [g] initially and [ŋg] medially. Unlike Noldorin/Sindarin, medial m did not become v in Gnomish.
  • Nasalized stops: These survived medially but became simple stops initially, as with Noldorin/Sindarin.
  • Palatals: Initial palatal consonants mostly became velars in Gnomish, having c- vs. ᴱQ. ty-, g- vs. ᴱQ. y- from dy- or ẏ-, but n- (from [ŋ-]?) vs. ᴱQ. ny- and h- (probably from voiceless velar spirant [x-]) vs. ᴱQ. hy-. This is mostly the same as the development of later primitive palatalized velars in Noldorin/Sindarin.
  • Labialized Velars: Unlike Noldorin/Sindarin, labialized velars like cw- and gw- often survived unchanged in Gnomish. The one exception is medial [kʷ], which became [p].
  • Others: The other sounds r, l; s, z; w, y mostly survived in Gnomish, except initial w- became gw- (as it did in Noldorin/Sindarin), and y- usually became g- (unlike Noldorin/Sindarin where it usually survived although was written as i-). The unvoicing of initial r, l did not occur in Gnomish; Tolkien introduced that idea in Noldorin only to abandon it again in Sindarin. I have no idea how z developed in Gnomish, since there are so few examples.

Although the above is very complicated, it does give a rough heuristic on how Early Period roots might be updated for better consistency with Tolkien’s later languages:

  • Voiceless stops can be left mostly as-is.
  • Nasals and nasalized stops can be left mostly as-is.
  • Voiceless spirants can revised to aspirates.
  • Voiced spirants can revised to voices stops.
  • Palatals (c-series) can be revised to palatalized velars.

Voices stops present a problem, since the b/p, d/t, g/k, gw/q pairings between Gnomish and Early Qenya simply can’t be reproduced in the phonology of later languages. Labialized velars (q-series) present a similar problem, since these frequently survived in Gnomish but became simple labials in Noldorin/Sindarin. Tolkien himself frequently updated Gnomish words into Noldorin/Sindarin by replacing labialized velars with labials, so it is best to leave early labialized velars alone when updating Early Period roots but update the Gnomish words to fit later Noldorin/Sindarin development.

Full vocalic developments are too complex to cover here, but one interesting item of note is that Primitive Elvish of the 1910s and 1920s also used syllabic nasals () and liquids (ṛ, ḷ) as vowels, and occassionally syllabic as well. These syllabic consonant generally developed into vowel/consonant combinations like > ar or > il in the child languages. Such early roots with syllabic consonants could become KALAT or KALTA-stem roots in Tolkien’s later writings, for example KṚNṚ “red” become KARAN.

Note that you can’t just blindly update Early Period roots and expect to produce something compatible with Tolkien’s later languages. The above only outlines the most basic phonetic developments; there are many minor variations with their own incompatibilities. Tolkien sometimes gave old roots new meanings in his later writings, for example: the early root ᴱ√MIRI “smile” was replaced by later root √MIR “precious”. But hopefully this gives you some idea of what is possible, so that you aren’t completely lost when I talk about early words.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.