The Mithrimin dialect of Sindarin

Roman Rausch

Jul. 1st 2009

He [Beren] spoke Sindarin after a fashion (probably derived from North Sindarin); but his halting and dialectal use of it offended the ears of King Thingol.




In remarks scattered through several sources after The Lord of the Rings, Tolkien mentions a variety of Sindarin, the North Sindarin or Mithrimin dialect. The aim of this article is to gather and analyze its various phonological details. An earlier article on this matter has been written by Richard Derdziński [1] (2001) and the present work will necessarily coincide with it in part, but there is much new material to be considered that has been published since.

1  Internal and external history

The Noldorin language was intended to be rich in various dialects from the very outset. The account in PE14:61-62 mentions the dialects of Mithrim, Gondolin, the Feanorian speech of Nargothrond and several other varieties. The Lhammas mentions the division of Noldorin into the speech Mithrim (Fingolfin’s folk), Gondolin (Turgon’s folk), Nargothrond (Felagund’s folk) and the speech of Himring (the folk of Fëanor and his sons) (LR:177).
The Etymologies list two words from this last Feanorian dialect, namely maið ’pale, fallow, fawn’ (Noldorin meið) (VT45:30), hence also the name Maiðros for Meiðros; and nef ’bolster, cushion’ for N. nedhw (changed from ’scarred’).

But in the conception after The Lord of the Rings the Noldor do not speak Noldorin, but Quenya, while the previous Noldorin language appears as Sindarin, spoken by the Sindar or Grey-Elves in Middle-earth. Sindarin is divided into three main dialects: Doriathrin or South Sindarin (spoken in Doriath), West Sindarin (spoken at the coast) and North Sindarin (spoken in Mithrim, Hithlum and Dor-lómin). Further subdivisions include the speech of Ossiriand, North-West and North-East Sindarin (PE17:131-134).
North Sindarin is also called ’the language of Mithrim’ or simply ’Mithrim’ (WJ:400,411). The adjectival formation ’Mithrimin’ is mentioned in PE17:134. In the internal history it perished during the war against Melkor and left little or no records, but can be traced in personal and place names (Dor-lómin, Hithlum). Tolkien also mentions that some of its words and forms eventually became incorporated in Sindarin of Sirion.
The classical Sindarin of the following ages, as it is for instance encountered in The Lord of the Rings, is mainly derived from West Sindarin, since the survivors of the war against Melkor took refuge at the coasts while the other parts of Beleriand were destroyed.

Interestingly, North Sindarin in fact became the language of the house of Fëanor, but now because the Noldor made first contact with the Sindar of Mithrim and adapted their speech. So earlier Feanorian nef agrees with the sound changes outlined for North Sindarin, but the preference of ai for ei does not, both diphthongs appear in the ultimate syllable of Sindarin words.

See also the detailed account in [1].

2  Mithrimin phonology

unlenited m (PE17:108,128,131,132,134, WJ:400, VT41:10)

The lack of lenition of m > v is a frequently mentioned feature of North Sindarin and is also mentioned as a chief difference from West Sindarin (PE17:128).
In later Sindarin phonology final -m becomes -n (cf. talan, tâl ’flat space, platform’ < talam) where Noldorin had -m > -f (talaf ’ground, floor’ < TALAM-). In North Sindarin final -m is retained.


In this notation indicates a nasalized vowel, its nasality is later lost in Sindarin. Similarly, probably indicates a bilabial spirant with a nasalized preceding vowel.
Note that Hithlum is a pure Noldorin word in The Etymologies (LUM-, LR:370), with N. lhum ’shade’ < *lumbē (Q. lumbe).

In a private exchange I was made aware of a possibility that this dialectal unlenited m could be a possible explanation for medial -m- in the name Boromir. PE17:134 states that the Beor-folk of the Edain were influenced by the Finrodians[?], especially in personal names, during their residence in the highlands. Both Boromir and Belemir are names from the house of Beor (WJ:230-231) and thus could be originally North Sindarin names, adapted in the south, and eventually by the Edain.
In The Etymologies Boromir is a pure Noldorin name, from Old Noldorin Boronmíro (BOR-). In a footnote in the Appendix F to LotR, Tolkien explains it as a mixed Sindarin-Quenya form.

unlenited initial s-, preserved medial -h- (PE17:134)

There is actually a whole series of formulations, one replacing another:

Initial s remained and medial h
Thus probably sad ’[…] place, spot’ (VT42:20) > WS *i had, NS *i sad. The sentence looks unfinished, but maybe intends to say that medial h also remains. This seems to refer to the contraction when h is found between vowels, e.g. the past tense of sav- ’have’: ahawv > aw *’did not have’ (SAM-, PE17:173) where in NS one would hence expect *ahǭm, *ahum.

>> S unlenited initially and medially
This apparently implies NS *i sad, *asǭm, *asum.

>> S unlenited initially; kh > h initially and medially
Now, -s- seems to be lenited again medially (NS *i sad, *ahǭm, *ahum), but kh > h would imply a lack of mutation of h, e.g. CE khēr > WS i·chír (VT41:9), NS *i·hír.

>> S unlenited initially; h (preserved) medially
Now apparently medial -kh- > -ch- [χ] (mentioned further on the same page), hence i·chír; but probably medial lenition -s- > -h- without further change, i.e. *ahǭm, *ahum.

The Beorian name Barahir contains unlenited medial -h- which could be explained by an adaptation from Northern Sindarin (the conception kh > h initially and medially), similar to Boromir, Belemir (see above).

pp, tt, kk > p, t, k

(but ph, th, kh and ps, ts, ks > f, th [þ], ch [χ], as in West Sindarin (PE17:134))

This is a remarkable development, as unvoiced intervocalic consonants are not found anywhere else in Tolkien’s Celtic branch of invented languages (Goldogrin, Noldorin, Late Ilkorin, Sindarin). But there is a similar development in Welsh where bb, dd, gg become pp, tt, cc, e.g. Cad-deyrn > Catéyrn for Catteyrn, wyneb-bryd > wynepryd ’countenance’ (J. Morris-Jones, An Elementary Welsh Grammar §127).

We can construct some examples:

unchanged combinations ’nasal + voiceless stop’ mp, nt, nc (VT42:27, PE17:132,134), unchanged lt (PE17:132,133,134)

In a note to the essay The Rivers and Beacon-hills of Gondor (1969) Tolkien states that the forms canthui [4th], enchui [6th], tolthui [8th] are those of the southern Sindarin dialect adopted by the Noldor. In the Northern dialect (which perished in the course of the war against Morgoth) nt, nc, mp had remained unchanged (VT42:27). In another note cantui, canhui ’fourth’ encui, enchui ’sixth’ (VT42:10) are cited where cantui, encui have apparently to be taken as NS cognates and examples of unchanged -nt-, -nc-. Another example is cantu >> centu with an unclear meaning (PE17:133).
The combination lt also remains unchanged, although lp, lc, rt, rp, rc become spirantal lph, lch, rth, rph, rch (PE17:132). Hence we expect NS *toltui ’8th’, *maltorn < malt ’gold, orn ’tree’ (VT42:27).

The conception at that time may be summarized as following:

Old SindarinNorth SindarinWest Sindarin
mpmpearly -mf- [bilabial f], late -mh-, final -mp
ntntearly -nþ-, late -nh-, final -nt
ncncearly -ŋχ-, late -ŋh-, final -nc
ltltearly -lþ-, late -lh-, final -lt

(Tolkien himself uses nch, ñx for ŋχ and ñh for ŋh.)

In this notation -mh-, -nh-, -ŋh- represent long or geminated voiceless nasals [m̥m̥], [n̥n̥], [ŋ̥ŋ̥] and -lh- a long or geminated voiceless l [ɬɬ].

There is also a different earlier conception: A note from 13 December 1962 describes the following development:

Old SindarinNorth SindarinWest, South, Ossiriandic Sindarin
mpmedial -mf-, final mpmedial -mh-, -m-, final -mp
ntmedial -nth-, final -nt (late)medial -nh-, -nn-, final -nt
ncmedial -ŋch- > ̄ch- (late), final -ncmedial -ŋh-, -ŋg-, final -nc
ltmedial -lth-, final -ltmedial -lh-, -ll-, final -lt

(Tolkien uses ngh for ŋh.)

In this notation ̄ch- means that the preceding vowel receives compensatory lengthening when the velar nasal ŋ is lost (however cf. NS dancen, dachen (PE17:133) without lengthening). The medial voiceless spirants now become voiced -m-, -nn-, -ŋg-, -ll- unless followed still by a stressed syllable. Interestingly the same pattern is found in Welsh, e.g. *brénhin > brénin ’king’, pl. brenhíni̯aeth (J. Morris-Jones, An Elementary Welsh Grammar §88, 90,126).
In this conception mp, nt, nc, lt are not preserved in North Sindarin, but universally become spirantal at some stage of Old Sindarin. But North Sindarin among all dialects retains the spirants and thus still remains more archaic. The combinations lf < lt, rth < rt, rf < rp; rch < rc remain spirantal in both dialects.

long ā > ǭ (WJ:414, PE17:133,134)

The long vowel ā universally becomes ǭ in Sindarin. The hook indicates an open pronunciation of the vowel, hence IPA [ɔ:] (as in English law, but short). North Sindarin retains this sound, but in West Sindarin it breaks into the diphthong au, in some positions it just shortens to o (WJ:414). In unstressed position ǭ seems to become u in North Sindarin. Another explanation would be that this is due to the following labial m, hence ǭm > um, with a different development if followed by other consonants.

As examples we have:

The shift lǭmin > lomin might indicate the adaptation of this word into classical Sindarin in the place name Dorlomin or Dor-lómin (Silm), genuine North Sindarin would still be lǭmin.

[Note that WJ:400 mentions that in North Sindarin diphthongization of ō […] did not occur, but this is probably a misreading for ǭ – only that would make sense in the particular context of the passage (namely the development of arǭmæ) and in the context of Sindarin phonology in general, where ō becomes ū instead.]

probably au > ô (PE17:133)

A couple of variants of the name of ’The Mound of Slain’, appearing as Haudh-en-Ndengin in The Silmarillion are found in PE17:133 (cf. also WJ:79, LR:363). Among these forms, one has apparently to identify Hauð i nenghin as West Sindarin exhibiting the long voiceless velar nasal ngh < nk. The form ð indencin then has apparently to be identified as North Sindarin since it shows preserved -nc-. The variant đ i ndechina with inflected genitive -a might be Doriathrin (interestingly it is a genitive singular suffix, maybe because plurality is expressed by vowel mutation, cf. -(i)on, PE17:97). At least by the time of The Etymologies the inflected genitive is a Doriathrin feature, e.g. Nauglamîr, N. mîr na Nauglin, Nauglvir > Nauglavir (LR:375). Alternatively, đ i ndechina might be a parallel North Sindarin form – note that NS dancen, dachen ’slain’ is mentioned in the line above.
If Tolkien did not changed the derivation of hauð ’mound, grave, tomb’, it should still be derived from khagda (LR:363, KHAG-) with vocalized g; or from khabdā (PE18:91, √KHAB-) with vocalized b, in any case via the diphthong au. So it seems that North Sindarin changes au (from a + consonant or from original aw) to ô (in all probability identical to ǭ), but might still develop au, for example from ou.

absence of a-mutation (PE17:133,134)

A-mutation in Sindarin is caused by final -a and influences i > e, u > o in the preceding syllable. Tolkien states that u and o and i/e remained distinct – no a-muation of i in North Sindarin (PE17:134).
We have only an example of i, but can construct examples for u:

However, this rule seems to be taken back again with the following development:

where one might have expected NS *dancin or *dencin.

combinations with final -w (VT42:8-9)

Old SindarinNSWS
-tw > -dw-b-du
-dw > -ðw-v-ðu
-thw > -þw-f [f]-þu

Tolkien says that Q. Finwe would have been adapted as S. Fim, but was actually sindarized as -fin, Curufin (VT41:10). PM:344 confirms that Fim would have been North Sindarin, and otherwise Finu.
In The Etymologies the sound change -ðw > -v was already attributed to the Feanorian dialect, e.g. nidwō ’bolster, cushion’ > ON nidwa > N.nedhw, Feanorian nef (NID-, LR:378, VT46:4); also deleted ON nidwa ’scarred’ > N. neðw, Feanorian nef. Here, -f is probably just a spelling of [v]. To avoid confusion one might use -ph for [f] and -v for [v].
There are not many examples with the above combinations to be found in Noldorin and Sindarin:

Note that for instance pathmā becomes Old Noldorin pathwa > pathw ’level, space, sward’ (PATH-, LR:380), but if this is to be taken into later Sindarin, the change m > *v > w would not be universal and the Mithrimin form would be *pathm (later maybe also twosyllabic *?pathum). The same applies to khithme > N. hithw ’fog’ (NS *hithm, *?hithum).

retraction of accent (?), retention of -nd- after accent (?) (PE17:133)

In what appears to be somewhat rough notes, Tolkien mentions an accent retraction in N[orth] (PE17:133). The following examples are Góndolind, S. Gonnólen, Gonðolen, Gondolen. This seems to be an attempt to explain the name Gondolin as a Northern Sindarin word: Apparently North Sindarin keeps the accent on the first syllable (gónd(o) línd > Góndolind) and nd which follows this accented syllable is not changed. But the development outside Mithrim would be *gónd dólen > *Gon(d)-ðólen > Gonnólen with medial -nn-. Somewhat unclear about this note is the different composition – Góndolind is obviously *’stone of music’ (as explained on the same page below), but Gonnólen must be *’hidden rock’ with dolen ’hidden’ (cf. DUL-, LR:355). One would have expected a comparison of different developments of the same primitive form.

3  Ideas taken from earlier languages

Edward Kloczko has long ago made the observation that the North Sindarin dialect seems to incorporate some aspects of the Ilkorin of The Etymologies.for which there was no place in the legendarium anymore after Tolkien had decided that the Grey-Elves should speak Sindarin instead [2]. For instance, Ilkorin changes ā > ō instead of a diphthongization to au as in Noldorin and leaves m unmutated. The same is, however, also the case in Doriathrin [3]:

At this stage Dorlómen, Ered Lómin is in fact a noldorized Doriathrin name Lómendor, Lóminorthin, while later Dor-lómin is an adapted North Sindarin name.
In Noldorin tinmē ’sprakle, glint’ becomes tinw, but tim in Ilkorin (TIN-, LR:393). In later Sindarin TIN- yields tim ’star, spark’ (PE17:22). Kloczko suggested that this could be a loan from North Sindarin, but in fact it might be that Tolkien simply changed the phonological conception where now tinmē > S. tim regularly (essentially by nasal mutation, cf. Annúminas ’Tower of the West’ (UT:419), evidently < annûn + minas). In any case, the correspondence of WS -nu to NS -m is separately attested in VT42:8-9.

But it seems that Tolkien also introduced many features from even earlier stages into North Sindarin, specifically from Goldogrin. Goldogrin leaves both m and s unmutated (GL:7), although they rarely appear intervocalically. Primitive ā also becomes ō in Goldogrin; and final -nw > -m. For example:

There is also a hesitation in the Gnomish Lexicon between combinations like mf, nth, ngh; lf, lth, lch; rf, rth, rch and mp, nt, nc; lp, lt, lc; rp, rt, rc. It seems that Tolkien sometimes decided for one or the other pattern; and sometimes for both. For example:

See [4] for more details.

The change -nw > -m also appears in Early Noldorin: hen(n) ’eye’, pl. hinn, but dual him < *henwi (PE13:122). Early Noldorin also seems to have a preference for u before m, it always appears as the connecting vowel in the 1st person pl. present excl., e.g. glathrum *’we polish’, madum *’we eat’, dagum *’we slay’ (PE13:126-130) – compare NS Arum.


Richard Derdzinski. Northern dialect of Sindarin
Helge Fauskanger. Ilkorin – a ”lost tongue”?
Helge Fauskanger. Doriathrin - the mothertongue of Lúthien
Roman Rausch. Historical Phonology of Goldogrin

update: Jun 29 2010 — discussion of Boromir, Belemir, Barahir added


This document was translated from LATEX by HEVEA.