Ancient Quenya Phonetics P35: [ln] became [ld]

Ancient Quenya Phonetics P35: [ln] became [ld]

AQ. [ln] became [ld]; [ln] > [ld]

In (Ancient) Quenya, the combination ln became ld, a change mentioned in both the Outline of Phonetic Development [OP1] from the 1930s (PE19/48) and the Outline of Phonology [OP2] from the 1950s (PE19/100). This change is most noticeable in adjective forms from roots ending with L, where the primitive adjective suffix -nā produced Quenya adjectives ending in -lda:

The exact timing of this sound change is hard to nail down, but as demonstrated by (one of) the etymologies of Elda, it must have occurred after the Quenya syncope:

  • elenā [> *elnā] > Q, Elda “[person] of the stars” (PE17/139, 152).

This change only occurred for the dental nasal n; the combination lm is a favored group in Quenya and remained unchanged (PE19/48, 100).

Conceptual Development: The combination ln is fairly common in the Early Qenya of the 1910s, for example: ᴱQ. pilna “arrow” (QL/74) and ᴱQ. †qalna “dead” (QL/76). This makes it unlikely that this sound change was part of Tolkien’s earliest conception of Qenya. Tolkien seems to have decided that ln mutated sometime in the 1930s, but the earliest entries in The Etymologies of the 1930s seem to show ln > ll:

The second example was rejected when Tolkien revised this root ᴹ√SKAL >> ᴹ√SKEL, at which point he change the adjective to ᴹQ. helda (with ld) as shown above. Furthermore, ᴹQ. nulla appeared beside an alternate form nulda, and may well represent a blending with nulla “dark, dusky, obscure” < ᴹ✶ndulla (Ety/NDUL). Thus, it seems Tolkien updated this phonetic rule to ln > ld fairly quickly.


Submitted by Lokyt Sun, 07/21/2019 - 11:01

I'm not convinced that *ndulna > nulla and *skalnā > halla necessarily show a divergent developement of -ln- to -ll- directly.

Considering that -nl- > -ll- (cf. *nenle > nelle) and that metathesis of "stop+liquid" / "liquid+stop" clusters is common, I'd say we're just witnessing another metathesis (-ln- > -nl-, changing *ndulna, but leaving an option of not changing for *skalnā) an then a perfectly regular developement (-nl- > -ll-).

Submitted by Paul Strack Sun, 07/21/2019 - 15:39

I thought about that possibility but didn’t think it was likely. I double checked and the metathesis usually goes in the opposite direction: dl, ml > ld, lm rather than the reverse. So I think ln > nl is unlikely.

Even it is true, though, it merely provides a mechanism for the theory rather than invalidating the theory itself.