Probable Errors in the Etymologies

The Etymologies in The Lost Road pp. 347-400 (pagination as in the first edition) is our main source of Elvish vocabulary. However, the ultimate expert of Elvish had been dead for years when Etym was prepared for publication, and some errors seem to have crept into the transcript of his original manuscript. We are here talking about real errors, namely misreadings and misprints, as opposed to later revisions and changes done by Tolkien himself (including the total revision of the history of the Welsh-sounding language: all of a sudden "Noldorin" turned into Sindarin).
          By pointing out some probable errors in the Etymologies as published in LR we are by no means implying that Christopher Tolkien did a sloppy job when he prepared Etym for publication. From his notes it is clear that Etym is a very difficult document, in faded ink and with layers upon layers of revision and addition, and often quite confused from the start. We should also remember that while JRRT sometimes produced wonderful calligraphy, his rapid handwriting was often anything but calligraphic; some published examples can only be called a scrawl. All lovers of Tolkien's languages should be eternally grateful to Christopher Tolkien for undertaking what must have been a long, difficult and tiresome work of transcription: Where would we have been without the Etymologies? It is exactly because Christopher Tolkien provided us with so much material that we are able to make out the general rules for sound-changes etc. and sort out the few abnormal forms. What follows may seem like a long list, but the probable errors are actually very few compared to the total number of words.

[UPDATE, autumn 2004: Since I originally wrote this article some years ago, Carl F. Hostetter and Patrick H. Wynne have examined photocopies of Tolkien's Etymologies manuscript. Their findings were published in Vinyar Tengwar #45 and #46, in the two-part article Addenda and Corrigenda to the Etymologies. In their article, most of the likely misreadings to which I and others had drawn attention were confirmed as erroneous. Below I have added, in brackets, some of the results of the Hostetter/Wynne study.]

1: Problems with n

The letter n is involved in very many of the probable misreadings, either being introduced by the misreading where Tolkien actually intended another letter, or itself being misinterpreted (as r, ri or u), or - in one or two cases - being omitted altogether. It seems that Tolkien's handwriting was especially difficult and ambiguous on this point.

Quenya nyano "rat" from primitive nyadrô (stem NYAD) is certainly a misreading for *nyaro; compare the "Noldorin"/Sindarin cognate nâr. It is true that initial primitive D can sometimes produce Quenya n by assimilation to an n occuring later in the word (see WJ:414), but the d of nyadrô is not initial, there is no n later in the word, and even if the initial n caused the d to become n, an intermediate form *nyanrô would probably have produced Quenya *nyarro or *nyarno rather than nyano.

[Hostetter & Wynne think the intended reading in the manuscript is nyarro with a double rr (VT46:7). This may also be suggested by the primitive form nyadrô "gnawer", post-vocalic d regularly becoming r in Quenya.]

Another case of confusion n/r is ontani "parents" as the common pl. of Quenya ontaro, ontare "begetter", masculine and feminine (ONO). As suggested by the singular forms and as confirmed by SD:73, this should read ontari. But SD and the published LotR also suggest that Tolkien scrapped this word and decided that the Quenya word for "parents" should be nostari instead (sg. *nostar; gender-specific forms *nostaro m. and *nostarë f.)

[According to Hostetter & Wynne, the word Christopher Tolkien read as ontani is actually ontaru (VT46:7). This would be a dual form in -u, denoting two parents considered as a natural pair.]

Under the stem KHOR a Sindarin verb hoeno- or heno- "begin suddenly" is listed. This should certainly be *hoerio, *herio; ri has been misread as n. The final R of the stem could not possibly yield Sindarin n, and many Sindarin verbs end in -io.

[As the infinitive of A-stem verbs in -ia, that is. Hostetter & Wynne agree that the correct reading is hoerio, herio: VT45:22.]

Another case of n for ri is the Sindarin verb glin- "to sing"; the form of the stem GLIR makes it virtually certain that this is a misreading for *gliri. In Etym, many Sindarin verbs have infinitives in -i; compare giri "to shudder" from GIR.

[Hostetter & Wynne confirm that the correct reading is gliri, VT45:15.]

The Sindarin word egledhriur "exiles" under , TA3 should probably be *egledhruin instead.

[Hostetter & Wynne prefer the reading egledhrim, VT46:16. But *egledhruin, later *egledhryn, would also be a possible plural form of egledhron "exile", the form given in the entry LED. Egledhrim would on the other hand be a so-called "class plural" in -rim.]

Under YA, we read: " 'old' (decripit, worn) of things is gem [GENG-WÂ]. See GYER." As we see, Tolkien himself included a reference to GYER, while Christopher Tolkien improperly (though with good intentions!) added a reference to GENG-WÂ in brackets. Sure, there really is a word gem listed under GENG-WÂ, but it does not mean "old" with reference to things - it means "sickly" and can obviously apply to living beings only. As we have seen, Tolkien himself added a reference to GYER, not GENG-WÂ - and under GYER we find the word gern "worn, old (of things)". So to cut short a story that is starting to get too long: The word gem in the entry YA is a misreading for gern. Having misread this word, Christopher Tolkien naturally thought it was connected to the actual word gem that is found under GENG-WÂ, but the words gem and gern do not mean quite the same.

[Hostetter & Wynne concur that in the entry YA, the word for "old" is gern rather than gem: VT46:22.]

The Sindarin word damna- "to hammer" under NDAM should perhaps be *damma; cf. the past tense dammint.

[Hostetter & Wynne agree that we should "perhaps read" damma- here, and they also suggest that the past tense dammint should be dammant, which form would fit the patterns elsewhere observed much better. - VT45:37.]

Similarly, Sindarin tamno- "to knock" from TAM should evidently read *tammo-; the primitive form is given as *tambâ-, and b cannot become n in Sindarin. (The primitive from of *damma- must similarly have been *ndamba-, as the Quenya cognate is namba-.)

[Hostetter & Wynne agree that tammo is correct, VT46:17.]

Sindarin Dadhrin "Nandor" under NDAN should probably read *Dadhrim, given that -rim is an ending used in the names of peoples; see RIM.

[Hostetter & Wynne have no comment.]

Under the stem AM1 the plural of Ilkorin aman "mother" is given as emuin. This should undoubtedly read *emnin, given the singular form. Compare Ilkorin boron "man", pl. burnin - not **buruin (BOR). Tolkien's n's were often all too similar to his u's; compare Letters:279, where the word nin of the invocation a tiro nin (found in LotR) is misspelt niu.

[Hostetter & Wynne agree that emnin is correct, VT45:5.]

A similar case is Sindarin moru "black" under MOR. It should have been morn as in Letters:382 and 427; cf. also mornedhel "Dark-elf" (morn + edhel) in WJ:377.

[Hostetter & Wynne concur: VT45:35.]

Yet another case is Quenya leuka- "loose, slacken" from LEK "loose, let loose, release"; read *lenka-. Quenya verbs often show nasal infixion, compare panta- "to unfurl" from the stem PAT, but the u-infixion required to derive leuka- from LEK is unprecedented. Luckily, an alternative verb lehta- of the same meaning is listed under LEK, so writers can use this word and ignore leuka- altogether. (According to a footnote in LotR Appendix E, Quenya leuca means "snake".)

[Hostetter & Wynne agree that lenka is the correct reading: VT45:27.]

Sindarin hamnia "clothe" under KHAP should evidently be *hamma-, an m being misread as ni. Compare hammad "clothing", that is basically the corresponding verbal noun.

[Hostetter & Wynne concur: VT45:21.]

Under BOR we find a Quenya word vorogandele "harping on one tune", but Quenya does not permit g in this position; this consonant occurs solely in the combination ng, or ñg according to the spelling used in Etym. The word should evidently read *voroñgandele; compare the stem ÑGAN and the words derived from it.

[Hostetter & Wynne concur: VT45:7.]

Another possible case of a missing nasal may be Quenya makar "tradesman" (MBAKH); given the verb manka- "trade" and the noun mankale "commerce" one wonders whether makar should not read *mankar. In LotR, macar (makar) means "swordsman" (as in Menelmacar *"Swordsman of the Sky", Orion - see the first footnote in Appendix E).

[Hostetter & Wynne comment on the entry MBAKH but present no correction of makar: VT45:33.]

2. Miscellaneous

Sindarin rhinn "circular" under RIN should probably have been *rhenn if it is a cognate of Quenya rinda (primitive *rindâ, not given; in Sindarin, the â would cause the stem vowel i to become e by umlaut before the final vowel was lost). Perhaps the editor confused *rhenn with rhinn "circle", given in the same entry?

[Hostetter & Wynne concur: VT45:11.]

The form Duveledh *"Dark Elf" (and not plural "Dark Elves", as a too literal interpretation of the wording in the entry MOR would indicate) should undoubtedly read *Dureledh: One of Tolkien's r's was too similar to a v for its own good. (Compare duredhel under DO3/; cf. also Barad-Dûr, not *Barad-Dûv, for "Dark Tower".) But in WJ:377, in a document thirty years younger than the Etymologies, the Sindarin word for "Dark Elf" is given as Mornedhel, so Duveledh/*Dureledh/duredhel can be ignored altogether.

[Hostetter & Wynne have nothing to say about the reading Duveledh though they comment on the entry MOR, VT45:35.]

The Old Sindarin ("ON") form peleki "fenced field" should certainly be *pelehi. The stem is PEL(ES), and while S is in some cases lenited to h in Sindarin, there is no (other) known case of s becoming k. Neither is such a development phonologically probable.

A similar case, but with a different misreading, is khelelia as a later form of Old Sindarin khelesa "glass" (KHYEL(ES)). Khelelia should read *kheleha.

[Hostetter & Wynne concur in both cases: VT46:8, VT45:23.]

Sindarin rhaes "horn" (RAS) must be a misreading for *rhass. Compare the form -ras in names like Caradhras and the Quenya cognate rassë.

[Hostetter & Wynne think we should "probably read" rhas rather than rhaes: VT45:11. Rhas and rhass would be the same word, since Tolkien is not consistent in his orthographic treatment of final -s(s); see separate article.]

Quenya raime "hunt, hunting" should evidently read *roime, the stem being ROY and the other Quenya derivate being roita- "pursue" - not **raita.

[Hostetter & Wynne concur: VT45:12.]

Under GIL we get the impression that geil is the plural of a Sindarin noun gîl "star"; according to everything we think we know about Sindarin, gîl must be plural "stars" and geil the singular form "star" (though rather gail in LotR-style Sindarin; this had also been the form of the word in Tolkien's early "Gnomish" language).

[Hostetter & Wynne concur that geil is singular and gîl plural, not vice versa: VT45:15.]

Quenya helk "ice-cold" (KHEL, KHELEK) is actually an impossible Quenya word; High-Elven hardly permits any final consonant clusters at all, and certainly not lk. Read helka, as in LT1:254 and the Silmarillion Appendix.

[Hostetter & Wynne concur: VT45:21.]

The Sindarin verb gwedi "bind", derived from the stem WED, should evidently be *gwedhi. Following a vowel, primitive *D becomes dh (ð) in Sindarin: Compare the past tense gweðant. A ð (dh) has been misread as a d, unless Tolkien himself misspelt the word.

[In his article The Past-Tense Verb in the Noldorin of the Etymologies, Hostetter states that both the infinitive gwedi and the pa.t. gweðant "are very clearly written as such in the manuscript". If so, this is likely a slip by Tolkien himself; it is very difficult to understand why he would want to introduce such a phonological irregularity in the infinitive but not in the past tense.]

Under KHAL, a Sindarin word orchel "superior, lofty, eminent" is listed; in a note, Christopher Tolkien admits that the e is uncertain. The stem KHAL suggests and WJ:305 confirms that it should read orchal.

[Hostetter & Wynne agree that it "may be possible" (VT45:20) to read orchal instead of orchel, and general principles strongly favour this reading.]

Another case of e for a may be sogennen as the past participle of Sindarin sogo- "drink" (SUK); other examples of Sindarin participles point rather to sogannen. (Cf. mae govannen, not *mae govennen, for "well met" in LotR.)

[Hostetter & Wynne have no comment.]

Quenya tolle as the plural form of tol "island" (TOL2) should probably read tolli, as in LT1:269. Quenya nouns do not have plurals in -e.

[Hostetter & Wynne concur: VT46:19.]

Under ÑOL, a Quenya word holme "odour" is given. Should it read *ñolme? We have no other examples of initial Primitive Quendian Ñ (as opposed to ÑG), so we cannot be sure how it developed in later Elvish tongues. Emending holme to *ñolme would produce a clash with a later word for "science" given in PM:360, but this is not conclusive (at best it suggests that it may be practical to accept holme as it stands). I would accept holme until there is any evidence that it is wrong, but it would be nice to have the matter settled.

[According to Hostetter & Wynne, "the initial h of Q...holme was subsequently struck out", VT46:6. Tolkien himself was apparently not certain what became of initial Ñ in Quenya: did it become h- or did it disappear altogether? The reading olme would seem to be the form we should adopt.]

3: What language is it?

Under NAR1 the Sindarin ("N") word for "sun" is given as Anar, but this is actually Quenya; the Sindarin form is Anor as under ANÁR. Compare names like Minas Anor in LotR.

[Hostetter & Wynne confirm the "N" form Anar here, VT45:36, so the error - at least relative to the implications of the LotR - is Tolkien's own.]

Under KAL the names Kalamor and perhaps also Kalamando are listed as Sindarin ("N"), but this is obviously Quenya. (In Sindarin, m following a vowel would lenit to v.)

[According to Hostetter & Wynne, there is not really any "N" before these words in Tolkien's manuscript; it would seem to be an improper editorial addition. -VT45:18.]

The verb helta- "strip" would seem to be Sindarin according to the wording in the entry SKEL, but in Sindarin t became th following l. Either helta- is Quenya or the form must be assumed to be a misreading for Sindarin *heltha-, probably the first. Tolkien simply failed to place a "Q" before this verb; the editor is probably not to be blamed in this case.

[Or maybe he was after all; according to Hostetter & Wynne, the manuscript reading is indeed heltha, the expected Sindarin/"Noldorin" form. -VT46:14.]

Under YUL, the words iolf "brand" and iûl "embers" are said to be "ON", Old Noldorin (read: Old Sindarin). This should probably have been "N"; the words look like normal "Noldorin".

[Hostetter & Wynne confirm that the manuscript reads "ON", but concur that "ON" is here "apparently a slip for N", VT46:23.]

4: What does it mean?

Under PHI a Quenya word fion pl. fiondi or fioni is given; its gloss was "unfortunately not certainly legible; the likeliest interpretation would be 'haste', but 'hawk' is a possibility". As "haste" would have no plural form, we must definitely go for "hawk". Besides, a quite different word for "haste" is listed under GOR.

[Hostetter & Wynne concur that "the reading 'hawk' appears more likely", VT46:9.]

Under KHOR, Quenya hórea is glossed "impulsion". According to what we think we know about Quenya word-formation, the meaning ought to be "impulsive" instead; the word seems to be an adjective derived from the noun hóre "impulse".

[Hostetter & Wynne concur, VT45:22.]

The meaning of the stem KHUG (see KHUGAN) is given as "bark, bay". Some editions of The Lost Road read "bar" for "bark", obviously a misprint.

Under the stem YUK "employ, use" a Sindarin verb iuitho is given; its gloss was almost illegible, but Christopher Tolkien suggested "enjoy". Given the stem meaning, "employ" seems a more likely interpretation. On the other hand, it seems strange that this seemingly "obvious" reading should not have occurred to the editor himself, so perhaps the scrawl in question really does look more like "enjoy" than "employ". We cannot know. A facsimile edition of the Etymologies would definitely be of interest to serious students of Tolkien's languages.

[Hostetter & Wynne agree that "employ" is perhaps the better reading, VT46:23. They present facsimiles of 3 manuscript pages in their two-part article; otherwise we must still depend on their and Christopher Tolkien's readings. Their Addenda and Corrigenda to the Etymologies article suggests many more corrections than the ones covered here. As far as Quenya is concerned, many of these corrections and additions can be found in the Quenya wordlists offered on this site; search for the references "VT45" and "VT46".]

Ardalambion Index