Sindarin Phonetic Development (Part 27)

Sindarin Phonetic Development (Part 27)


OS. [ei], [ou] became [ī], [ū]; [ei|ou] > [ī|ū]

In (Old) Sindarin the ancient primitive diphthongs [ei], [ou] developed into simple long vowels [ī], [ū]. The clearest representation of this change is in a chart describing the phonetic development of Primitive Elvish diphthongs in all three languages (Quenya, Telerin and Sindarin), appearing in notes associated with Tolkien’s discussion of Hands, Fingers and Numerals from the late 1960s (VT48/7). Here Tolkien made it clear that [ei], [ou] > [ē], [ō] first, and then later [ē], [ō] became [ī], [ū]. Tolkien described these changes elsewhere in his notes on Sindarin:

  • In Sindarin ... the diphthongs so formed also followed the development of original diphthongs (aı̯ > ae; iı̯ > ī; eı̯ > ī; oı̯, uı̯ > ui) [emphasis added] (PE19/91).
  • Q Orome gives S. Araw, from Oromē ... ow new > aw, but old ow- > ū [emphasis added] (PE17/99).

As the second example makes clear, these early phonetic developments of primitive diphthongs were different from the developments of newer diphthongs [ei], [ou] produced by other phenomenon such as vocalization of spirants. The later [ou] became [au], whereas the later [ei] became [ai] in final syllables. These differing early vs. late diphthongal developments can help identify the timing of other phonetic changes. For example, it helps establish that the first in a pair of voiced stops vocalized very early, since the resulting diphthongs produced the same result as primitive diphthongs as pointed out by Tolkien himself (PE19/91).

Clear examples of this sound changes in Sindarin are not numerous, but there are at least a few:

  • negdē [> neidē] > S. nîdh “juice” (PE19/91).
  • thowō > thow [tʰou] > S. Thû “Sauron” (PE17/99).

Conceptual Development: The phonetic developments for the primitive diphthongs [ei] and [ou] evolved quite a bit in the various conceptual stages of the language, and each time period deserves a detailed discussion.

Gnomish (1910s): In the early chart of Gnome Vowels from the 1910s, the diphthongs ei, ou produced ê, ô when stressed, but various short vowels (e/i or o/u) when unstressed (PE15/13). As pointed out by Roman Rausch, the diphthongal development in the Gnomish Lexicon mostly follow these rules (HPG/§1.2):

Early Noldorin Developments (1920s): There is similar chart of diphthongal developments in Early Noldorin from the 1920s, but these show [ei] > [ai] and [ou] > [au] (PE15/64). As pointed out by the editors (PE15/64, note #21) there is a nearby example with lopse > laus (presumably [lopse] > [louse] > [laus]), so this probably represents the later diphthongal developments after vocalization of spirants rather than the development of primitive diphthongs. In fact, in the contemporaneous Noldorin Wordlists and Noldorin Dictionary circa 1923, it seems generally [ei] > [ui] and [ou] > [ū]:

These sound changes were described by Roman Rausch in his Historical Phonologies of Ilkorin, Telerin and Noldorin around 1923, along with their parallels to Welsh development (HPITN/§4.2.4). In Welsh, primitive ei > wy (pronounced [ui̯] or [u̯i]) before a consonant (WGHP/§75iii), and generally [ou] > [u] (WGHP/§76i), making Early Noldorin’s development very “Welsh-like”. There are, however, a few aberrant Early Noldorin developments in which [ei] becomes [ai]:

  • ᴱ✶peia > ᴱN. hai “scorn” (PE13/146).
  • ᴱ✶leyé > ᴱN. lhai “folk” (PE13/148).

Roman Rausch attributes this alternate development to variations in stress, but I think, like Welsh, the variation is due to the presence or absence of a following consonant. Consider the following example:

  • ᴱ✶ʒolweı̯á or golwé(i)ı̯a > ᴱN. golwai “stinking” vs. goluidh of the same meaning (PE13/162).

The primitive form of goluidh is not given, but it is derived from some variant of ʒolweı̯á + d (?ʒolweı̯d), whereas the primitive form without a consonant produced -ai. Likewise in Old Welsh, eı̯a > iı̯a > -ai, Modern Welsh -ae (WGHP/§75iii-2, §75v-4), so this seems to be another “Welsh-like” development for Early Noldorin.

Noldorin Developments (1930s): In the Comparative Tables of phonetic developments from the 1930s, Tolkien had primitive [ei] > [ui] and [ou] > [ū], much like the Early Noldorin phonetic rules of the 1920s. There are some examples of the first of these sound changes in the early (deleted) layer of entries in the Etymologies:

  • ᴹ√GEY > N. guir “eternity” (Ety/GEY)
  • ᴹ√EY > N. uir “eternity” (Ety/EY)
  • ᴹ✶Eigolosse > N. Uiloss “ever-snow” (Ety/EY)

Tolkien ultimately rejected the roots ᴹ√GEY and ᴹ√EY, inventing a new root ᴹ√OY¹ from which these forms were derived. In later, undeleted entries in the Etymologies, the normal development was [ei] > [ī], the same as it is in Sindarin:

The last example demonstrates this was an early sound change, in Old Noldorin. Similarly, the attested examples in the undeleted entries in the Etymologies indicate [ou] > [au]:

  • ᴹ√NOWO > N. naw “idea” (Ety/NOWO)
  • ᴹ√NOWO > N. nauth “thought” (Ety/NOWO)

This ultimate result is identical to the phonetic development of the later Noldorin diphthong [ou] which developed from vocalized spirants:

  • ᴹ✶oktā [> oktʰa > ouθa] > N. auth “war” (Ety/KOT)
  • ᴹ√LOKH [> loksa > lousa] > N. lhaws “hair” (Ety/LOKH)

Thus, it seems likely that the primitive diphthong [ou] remained unchanged in Old Noldorin, and in fact attracted several other primitive vowel-sounds: all of [au], [ā], [ǭ] became [ou] early on. The change of [ou] > [au] did not occur until the Noldorin stage of phonetic changes.

This means that the phonetic development of primitive [ei] in the Etymologies is already consistent with Sindarin, but the phonetic development of primitive [ou] was not. By the 1950s, Tolkien switched the development of this primitive diphthong to [ou] > [ū], which seems to be a partial reversion to Early Noldorin sound changes. To summarize the most common sound changes in each conceptual period:

Gnomish (1910s): [ei] > [ē] [ou] > [ō]
Early Noldorin (1920s): [ei] > [ui] [ou] > [ū]
Noldorin (1930s): [ei] > [ī] [ou] > [ou] (> [au])
Sindarin (1950s): [ei] (> [ē]) > [ī] [ou] (> [ō]) > [ū]
Submitted by Lokyt Wed, 12/19/2018 - 08:35

Wouldn't it be better to postulate a change (o)w > (o)u̯ (analogical to j) first, rather than to treat ow as ou̯ silently? Or to ask more comprehensively :) - don't your data, presented above, imply that ow > ou̯ must have taken place before the changes treated here were possible?

Anyway, nice work with the early materials.

Submitted by Paul Strack Thu, 12/20/2018 - 02:06

In reply to by Lokyt

I haven’t discussed it yet, but I do assume that in Primitive Elvish [VwC] > [Vu̯C] just as [VjC] > [Vi̯C]. The intervocalic development of [j] and [w] seem to be different, though. In particular, intervocalic w is possible in both Noldorin and Sindarin:

  • S., N. tawar “forest”.
  • S. leweg “worm”.
  • N. tawaren “wooden”.
  • N. glawar “sunlight”.

Now, its possible this is an orthographic convention, and these w represent [u̯] rather than [w]. There are definitely places where Tolkien uses w for [u̯], such as at the end of words (S. raw “wing”) but sometimes medially as well (S. hawn “brother” and N. lhaws “hair”). However, I believe that ia in leithia- is two syllables, while wa in hadhwa- is only one.

There a probably also subtle developmental differences having to do with stress patterns. We know that áwa and awá had different phonetic developments in Quenya (the later to öa) and it seems likely there are similar variations in Sindarin/Noldorin. I’m deferring full analysis of these questions for now, because there isn’t really enough evidence for me to draw conclusions one way or another. I’m hoping the picture becomes clearer with future publications.

Submitted by Lokyt Thu, 12/20/2018 - 11:54

In reply to by Paul Strack

See, this is why I say that the semivowel-diphtong area is tricky and hard to puzzle out :)

So just to exchange wiews, my belief is that whatever there was in PQ. or CE., Noldorin/Sindarin had no actual [w] any longer (just []) and all the au vs. aw etc. vacillation is just Tolkien playing with aesthetics of Welsh orthographical conventions.

Submitted by Paul Strack Fri, 12/21/2018 - 01:31

In reply to by Lokyt

I agree that you’re probably right. For the time being, though, I am hedging my bets and using phonetic rules that are close to Tolkien’s orthographic conventions, in the hope that things become clearer later, especially once the as-yet-unpublished discussion of Noldorin phonetic history becomes available. It’s supposed to be from the 1940s, so it may give a nice bridge between Noldorin and Sindarin and make certain things clearer.