Tolklang archive file articles/Appleyard.Quenya (version of 7 Mar 1995) From: A.Appleyard, C14, Materials Science, UMIST, Manchester 1, England (email: A.APPLEYARD@FS1.MT.UMIST.AC.UK) [QUENYA GRAMMAR REEXAMINED] [CONTENTS] SYMBOLS MAJOR REVISIONS PREFACE Constructivism Vinyakaarie THE LOTR-PERIOD CASE SYSTEM THE OLDER (1930'S) CASE SYSTEM NARQELION CASES THE SUFFIX `-ya' ADJECTIVES PRONOUNS VERBS Subjunctive, optative, subordination Being Causative `Becoming', inchoative Passive Translating the English infinitive VOCABULARY COMPOUND SENTENCES AND CONJUNCTIONS Vinyakaarie of conjunctions Place Time CASES OF CASES? PHONOLOGY OF DEVELOPMENT OF QUENYA FROM PROTO-ELDARIN Vowels Consonants Suffixes MISCELLANEOUS ANALYSES OF SOME TEXTS REFERRED TO FS (Fiiriel's Song) ONAM (original form of Namaarie) (TOI, pp284-5) NCPT (Notion Club Text) OM1 (The Last Ark, first 1930's form) OM2 (The Last Ark, second 1930's form) Nieninque Earendel OM-NEW (The Last Ark, post-LOTR form) Koivieneeni Sentence "Arctic" Quenya sentence TELERIN VANYARIN, NOLDORIN REFERENCE CODES [SYMBOLS] "Q" = Quenya, "S" = Sindarin, "PE" = Proto-Eldarin, "IE" = Indo-European, "BQ" = Book Quenya, "C" = any consonant, "V" = any vowel, "1p" "2p" "3p" = 1st/2nd/3rd persons. "incl." = "including you", "excl." = "excluding you", of 1st person duals and plurals. Curly-bracketed uppercase, eg {ANA}: see BOLT/QL under that PE root heading. Curly-bracketed lowercase, eg {ini}: see ETYM under that PE root heading. One capital letter in Proto-Eldarin indicates the stress. "MET" = "in Middle-Earth time", "RT" = "in real time (Tolkien's lifetime)". << = "changed by Tolkien from", >> = "changed by Tolkien to". In PE and Quenya, "ng'" (with apostrophe, as in Swahili) is the `ng' sound as in "sing", and "ng" is `ngg' as in "finger". A and B are nouns, and X and Y are clauses. Comments welcome re word use etc. Most words are from ETYM and BOLT. ETYM prints PE `gh' as `3'. `~' = "lengthens previous vowel if that vowel is short in an open syllable". `*' on PE words in ETYM with Tolkien means "treated as not recorded in Middle Earth text surviving to the present day" or similar. But I treat such Tolkien forms as "Middle Earth known fact", and only put `*' on words which are not in Tolkien. According to LOTR, circumflexed vowels in Sindarin words are superlong (like in modern Estonian) (here represented as triple vowel). The same presumably applies to circumflexed Ilkorin words in ETYM, and perhaps to the few known cirumflexed Quenya words such as (1930's) `Ooo' = "the ocean" and `Veee' = "Mandos", and the circumflexed PE root {dooo} = "night". [MAJOR REVISIONS] 22 Sept 1994: Treated 1930's case endings more fully after VT#36 published MPD in full: (1) many more 1930's case forms; (2) disproved my previous theory that the MPB pronouns `ensi entu enta' meant "this here, that near you, that yonder" like Latin `hic iste ille'; (3) MPD short ablative in `-l' coincides with a previous analogical vinyakaarie of mine; (4) I added analyses of the 1930's SV Quenya poems. Also I restored a chapter on suffixes that had got lost in editing a while before. 2-3 Dec 1994: new information from QAE in WOTJ: new 2p pronouns; patientive pl. case; a 3rd version of PE. Much information about Telerin and Valarin. [PREFACE] This article is a printed form (as at December 1993) of the TL archive file {articles/Appleyard.Quenya}, which is a revised extended reprint of some of my submissions to TL. It is likely that sometimes C.Tolkien, when transcribing ETYM for publication, sometimes due to the difficulties of the manuscript misread acute accent (PE stress) for macron (vowel length) and vice versa. Any such errors in matter taken from ETYM stand in this article as printed in ETYM. Some features of this article caused controversy when they appeared in TL:- [Constructivism] An opinion that Tolkien changed Quenya so much over his life that e.g. c.1930 Quenya (Q1) is effectively a markedly different dialect from LOTR-period Quenya (Q2), so that Q1 words and forms do not always have full authority when writing in Q2, and vice versa. There has been much vocabulary change: see Christopher Tolkien's remarks in LR on the amount of erasures and alterations that he found in ETYM. Also see his remarks in SD in the last two pages before the index, about his father's Aduunaic language matter, referring e.g. to "continued subcreation", i.e. a desire to keep tinkering! And compare OM2 with OM-NEW for the amount of vocabulary etc. changes that Tolkien made when translating `The Last Ark' from Q1 into Q2. [Vinyakaarie] i.e. "new making", = matter not in Tolkien and not derivable by analogy from Tolkien matter. It is irrelevant to people studying his ideas and purpose; but may in the end sometimes be a necessary evil for people writing in Quenya when practicality clashes with faithfulness to Tolkien: e.g. no Tolkien text so far published provides a Quenya (or Sindarin) for the common "if"-clause or for several common English verbs. One reason for the grammar vinyakaarier hereinunder is to show how much is still missing in our knowledge of Quenya. How risky vinyakaarie now is depends on what proportion of his total output containing new information is not yet published (e.g. a Quenya verb `faiqueta' = "curse" < {spay kwet} = "contempt say" needed and invented (not by me) was overtaken by MR p373 publishing Tolkien's Sindarin `Narn e.'Rach Morgoth' = "Tale of the Curse of Morgoth": if the apostrophe represents lenited-out `g', it could < a PE root *{garAk}, which would > Q. *`araka' (not `arka', which already = "narrow" ETYM {ak})). How much analogical extension and vinyakaarie can be tolerated in practice? Any writers in Tolkien languages should annotate any necessary vinyakaarier so that readers know that they do not have Tolkien's authority. Examples of types of extensions of varying acceptability or doubtfulness are:- (a) e.g. *`ostonna' = "to the city" < known `osto-' = "city" and `-nna' = allative case; (b) e.g. *`tyelpea' = "silver" (adj) by analogy of `laurea' = "golden"; (c) completing the personal pronoun and case ending tables as hereinunder; (d) deriving new words from known Proto-Eldarin roots; (e) arbitrary invention, e.g. my *`A-ki' = "if A is there" hereinunder, or inventing a new PE root and its family of Quenya and Sindarin descendants. [THE LOTR-PERIOD CASE SYSTEM] PLET gives the "Book Quenya" forms of these cases of the vowel stem nouns `cirya' and `lasse'. I add `elen' from ITE, and non-BQ forms from various Tolkien Quenya text. cirya lasse elen cirya & lasse [7] "ship" "leaf" "star" many- `t'-dual sing plural sing plural sing plural -plural Nominative[1] a ar e i[8] n ni li t Genitive[1] o aron eo ion no nion lion to Instrumental anen ainen enen iinen nnen niinen[3] liinen nten Allative anna annar enna ennar nna nnar[3] linna(r) nta Dative[4] an ain en in nen nin[3] lin nt[2] Locative asse assen esse essen sse ssen[3] lisse(n) tse Respective[4] as ais es is nes nis[3] lis tes[3] Ablative allo allon ello ellon llo llon[3] lillo(n) lto Dedative[5] al ail el il nel lil[3] lil tel[3] Patientive[9] ava aiva eva iiva[3] nwa[3] niiva[3] liiva twa BQ nominative a ar ii ii n nii lii t BQ accusative aa ai ee ii n nii lii t BQ genitive oo aron eo ion no niion liion to Pre-BQ acc?[6] a`e a`ii e`e e`ii n`e n`ii li`ii t`e [1] For the BQ forms of these cases, see bottom of table. Ordinary accusative is the same as nominative. [2] Earlier sources point to `-ten': see below re which vowel to interpolate. [3] These positions are blank in sources, e.g. PLET: I have filled them myself. [4] These two lines in the original have no case name and have each form bracketed, as if they are short forms of the next line above each. ITE, using only LOTR and NAM, gives some of these (in ordinary Quenya form), and also some consonant-stem forms. [5] This case is a 1930's Quenya form from MPD: see below [6] Before BQ times the accusative may have been as here, with a sound (probably glottal stop, perhaps `g' or `gh'), which I represent here by `, which in recorded Quenya has disappeared, often with vowel contraction. [7] `elen' probably forms many-plural *`elelli' with assimilation, and dative perhaps *`elenen'. [8] SD p302 says that `orne' pl. `orni' = "tree" < PE `ornee' pl. `ornei'. [9] QAE pp 368,369,407 showed that patientive pl. is `-iva' (not **`-van' as I for a long time had theorized) BQ is an old (MET, not RT) form of Quenya which had presumably dropped out of general spoken use (PLET). ETYM {thor(on)} mentions an Old Noldorin (ancestor of Sindarin) gen.sg `-en'. Plurals: `Vala-r' (LOTR) = "Vala"; `lasse' : `lassi' (NAM) = "leaf"; `Valataari' : `Valataari' (ETYM) = "Vala-queen"; `Uuvanimo-r' = "monster"; `Ainu-r' (SIL) = "Ainu"; but `tie' : `tier' (NAM) = "road" and `Valie-r' = "female Vala". But NCP `heruion' gen.pl. of `heru' = "lord". See below re adjectives in `-ea'. Nouns in `-ie' form gen.sg. `-io' (BQ *`-ioo'?), nom.pl. `-ier', gen.pl. probably `-ieron'. In NARQ adjectives in `-a' form their nom/acc plural in `-ai', and nouns in `-a' and `-e' form their nom/acc plurals in `-r' (`aldar', `lasser', `rotser'). QAE p366 mentions a PE allative suffix `-da' > pre-Q. `-d' > Q. `-r'. Nouns in `-o', e.g. `osto' = "fort", `uuvanimo' = "monster", would probably remain `-o' in ordinary gen.sg., and change to `-oo' in BQ gen.sg. Final vowels lengthen (= stay long as they were in PE) before `-li' in two examples in LOTR (`vanimaalion', `malinorneelion'), but not in PLET. Compare `Kuiv-iee-nen' = "Awakening-water" (SIL). Most stems ending in two consonants (the result of a final vowel dropping, see {Phonology}) drop the last consonant when it is word final, e.g. `pilin : pilind-' = "arrow", `tarkil : tarkild-' = "Nuumenoorean"; this often happens to a short noun when it is used as the last element in a compound. Before a consonant suffix that final vowel returns; it is usually an `-e', but not always, e.g. `osto' = "fort, city" : `Formenos' = "North Fort", loc. `Formenostosse'. Some ETYM PE words end in short `-i' and `-u', which in Quenya when final > `-e' and `-o' (or in polysyllables often disappear), but unless analogical levelling won would persist before suffixes and case endings, e.g. NAM `suurinen' not `suurenen' instr.sg. of `suure' = "wind", and in BD; c.f. also `atari-nya' = "my father", `hildi-nya-r' = "my heirs" (pronoun genitive suffixes are inserted before any case endings: see below re them). Distinguish this interpolated `i' from the plural `i' < PE `ii'. LR p61 short form `yonya' = "my son" < `yondo' = "son" is vocative as endearment; likely the full form `yondonya' was used for other purposes; `onya' = "my son" UT p175 is likely the same but influenced by {on} = "beget". `Mandos' varies between `Mandosto-' and `Mandosse-'. Stems in `-Cw': `ango' : `angw-' = "snake"; `orko' : `orqu-' = "orc"; `ranko : ranqu-' = "arm"; `telko' : `telqu-' = "leg" (analogical) ETYM. The `w' would > `u' before a consonant, e.g. instr.sg. `rankunen', unless analogy produced e.g. `rankonen'. Other irregular nouns in ETYM: `filit' : `filik-' = "small bird"; `haran : harn-' = "king, chieftain"; `hwan' : `hwand-' = "sponge, fungus"; `ilu', `iluuve' : `iluuv-' = "universe"; `kaimasan' : `kaimasamb-' = "bedroom"; `nelet' : `nelk-' "tooth"; `noa, noo' : `noow-' = "idea"; `peltas' : `peltaks-' = "pivot"; `raa' : `raav-' = "lion"; `seler' : `sell-' = "sister"; `soron' : `sorn-' = "eagle"; `talan' : `talam-' = "floor, ground"; `toron' : `torn-' = "brother"; `yat' : `yaht-' = "neck". `-no' : `-now-' = "-minded" is a likely form of compounds of `noo' : `noow-' "idea". Nominative singulars of more than one syllable shorten the last vowel before final `r', e.g. `Valatar' : `Valataar-' = "Vala-lord". (But often stays long in LOTR in `palantiir'.) These nouns in ETYM lengthen the nom.sg.:- `neen' : `nen-' = "water"; `neer' : `ner-' = "sentient male"; `taal' : `tal-' = "foot"; `yaar' : `yar-' = "blood"; but compare `Koivieneeni'. Stems ending in one consonant also should regain any lost final vowel before suffixes, but by analogical levelling or haplology often do not if the suffix starts with `CV-' or a doubled consonant. Known are `earello' < *`earen-llo' LOTR, `Elenna' < *`elen-nna', `nuumessier' = "are in the west" FS < *`nuumen-sse' + FS's suffixed verb `ye' = "be", cases of dual `ciryat' PLET (which other words in `-Vt' may decline as), and with pronoun genitives `anari-nya' = "my sun" FS (pl. "my suns" would be `anarinyar'), `atari-nya' (LR p70). UT p193 has `aranya' < *`aran-nya' = "my king" used as vocative: there seems to be a tendency to contract `A-nya' = "my A" forms when used familiarly as vocatives. `-VCnen' (inst.sg) and `-VCna(r)' (allative) would behave phonetically like verb `n'-pasts (see below). `-VClo(n)' (abl.) and `-VCse(n)' (loc.) would suffer assorted damage to stem and/or ending, as in {Phonology} below. In *`-VCva(n)' (pat.) `v' > `w' after consonant. Many words and cases for pronounceability or clarity would instead regain the vowel (or insert a vowel, for any genuine PE consonant stems), including all cases of any like `Valinoorenna' < `Valinor-nna' (c.f. `noore' = "land") whose nom.sg. lost a final vowel in compounding. When consonant stems add (or regain) short `i' before suffixes, would this `i' > `e' when unstressed in the short place cases, e.g. dat.sg *`anarin' > *`anaren'? (Note that persistent Quenya final `-in' < PE *`-iin' or *`-iim'.) Q. `taar' {taa} < PE `taaro' = "lord", but FS and NCPT `Iluuvataaren' (dat. used as gen.) (not `-ron') = "Universe-lord" = "God". Most transitive clauses are in the order S-V-O; but where as sometimes in poetry and around relative clauses and for emphasis this order is disarranged, the BQ accusative could perhaps be used for the object to distinguish it from subject to avoid ambiguity: in apparent consonant stem forms the PE final short vowel could be restored and lengthened here, e.g. `Silmarillee' (< pre-BQ *`Silmarille`e'?, see above). For Tolkien's 1930's solution to this, see below re BD. If the clause starts with its object, that object can be turned into a subject by putting the verb into the passive, which see under {Verbs}. LR p63 tries to disambiguate poetic O-S-V order (and to fill the meter) by inserting a surplus pronoun `en' = "it": `Ilu Iluuvatar en kaare' = "The universe: Iluuvatar it made". For the t-dual see above; LOTR `Alduu-ya' = "(day) of the Two Trees" and `Aldudeenie' = "Lament for the Two Trees" (CHJ) (with `d' between vowels, which is unusual for Quenya) indicate *`aldu' = "2 trees"; also found is `aldaru' = "2 trees" (KVNM), and `-aru' (crossed out in BD). Originally (MET) the t-dual (< PE `ata' {at} = "again, two") was for any two, and the u-dual (< PE `-uu') was only for natural pairs, but by normal Quenya times both were only used for natural pairs, and euphony decided which of `-t' and `-u' to use (LETT p427). Nouns not in `-a': PLET has `lasset' (not `lassat') < *`lasse-ata' = "2 leaves". Are the dual short place cases `-tan' etc or the analogical `-ten' etc? (See below re PLET `-nt' and BD.) Cases of the `u'-dual are easier to form: `aldunna' = "to the 2 trees", `alduo' = "of the 2 trees". If a verb's subject is in the dual, what number is the verb in? Ditto an adjective qualifying a dual? If singular, the dual is likely a collective singular = "a pair of A's". If so, can the dual be pluralized to mean "pairs of A's" (e.g. `enque taruku-t-i' = "8 yoke of oxen")? `Varda teluume-n' (ONAM) is dative used to mean "known for" ("Varda, known for her dome"), extended perhaps from `B na/ea A-n' = "B is to/for A" used also to mean "A has B" in Quenya as in Greek. `An-kalim-esse-lii-nen' = "[he/they] with [= having] many long-bright [= famous] names" (SD ch.V) is instrumental used as an epithet: it seems that `A-nen' can mean "who has A". The role of the case that I call `respective' may relate to Locative as Dative relates to Allative: when a noun A is Allative, something goes to(wards) A, and when A is Dative, some object or action or favour `goes to' A semi-literally or figuratively. Perhaps `-s' means "concerning", "with respect to", "paying attention to", and (most often) "about" in "think or talk or write about", whence my name for it: when A is Locative the action is at A but does not move towards A, and when A is Respective the action concerns but does not directly affect or `go towards' A. In NARQ and OM-1931 this `-s' > `-r' used as locative, and in BD and `Nieninque' as dative. I put the 1930's MPD `Dedative' case here also, as since there is a shortened allative used figuratively (dative), and a shortened locative probably used figuratively, there should also be a shortened ablative used figuratively, i.e. "from", "originated by", agent (not instrument) of passive verbs, and similar. I call the dative and respective and dedative the "short place cases". Of apparent consonant stem forms of them there are: PLET `ciryant' dative (< dual stem `kiryat-') with metathesis, ending unusually in two consonants; NCPT and FS `Iluuvataaren' (used as gen.). Many of the possible combinations of `-VCn', `-VCs', `-VCl' are unpronounceable, even allowing metathesis, so that often `-en', *`-es', *`-el' must be used. I suspect that the `-nt' was a slip-up (by analogy with allative `-nta' < `-tna') by Tolkien who when he wrote PLET was likelier wading through a pile of all sorts of reader etc correspondence than thinking only about his languages, and that this form should perhaps be `-ten' (or `-tan' < {ata}?). But MPD dual has some undeniable examples of Quenya final `-nt'. How far in Quenya should the bulldozer of `Analogical Levelling', destroyer of many real world languages' linguistic archaeology, be allowed to intrude and obliterate etymological forms to make easy roads for learners? Quenya as defined so far, like many real languages, distinguishes "in A" from "to/into A", but not "staying outside A" from "coming out of A". If it had a "position outside" case ("Abessive"?), its shortened form ("Irrespective"?) might be used for figurative ditto, e.g. "notwithstanding A" or "regardless of A". These cases could be made by putting `la' = "not" {la} before the locative or respective. (Not `uu', which would negate the whole clause.) `o' = "together" {wo} used separately as a preposition with the instrumental would mean "together with", to distinguish from "by means of" or "done by" where necessary. (Note: `A-nen' can also mean "A-water"! (e.g. `Kuivieenen').) The `-va' case has two uses: (a) LOTR `yuldar lissemiruvooreva' = "draughts of sweet mead" (NAM) (for long the only known use); SIL `Mar Vanwa Tyalieeva' = "Home of Lost Play", `Nurtale Valinooreva' = "Hiding of Valinor". It may be `objective genitive' = `patientive', i.e. `A-va B' means that A is (grammatically) a `patient', being affected by B: the `lost play' is housed; (the Valar) hid Valinor; (Galadriel) poured the mead out. (b) QAE pp368,369,407 says explicitly that `X-va Y' rather than `X-o Y' is the original correct genitive in non-partitive uses if X at the time described still has/had physical possession of Y, but `-o' intruded on `-va' later MET in Quenya and largely displaced it in Telerin. Mindon Eldalieeva' SIL = "tall tower of the Eldar", is now seen to be an example of this use, and perhaps also `hwesta sindarinwa' = "Sindarin's `hw'" LOTR (which would prove ITE's guess that `-va' > `-wa' after consonant). This agrees with the agentive / patientive use of `-o' / `-va' as genitives, as the action comes the agent. According to ETYM {gho}, this `-o' < PE `ghoo(m)', and was formerly MET specifically partitive. Its plural added no `-n' in Telerin (QAE p369). The "substance made of" meaning that ITE guesses for `-va' (and thus Quenya written by other people often uses) is mostly covered by partitive `-o' and the `-ea' (pl. `-ie') adjective: `laure' = "gold" : `laurea' = "golden". QAE p407 says that `-va' is a declinable adjective: pl. `arani Eldaiv' (archaic MET pl. `-vai')= "kings of the Eldar", like NARQ `kuluvai' = "golden" (pl) < QL `kulu' = "gold", but clashing with NAM `yuldar -ev'. In NAM in tengwar in RGEO, `-eeva' is written with the letter for `v', not `w'. As well as this MET spread of `-o', `-o' spread RT displacing `-(e)n' which was formerly used as genitive (often in ETYM; FS & NCPT `Iluuvataaren'; etc). There may be an unlisted case `-me': `leneeme' = "with leave" (NCPT), `turme' = "to get control" or "in control"(?) (ONAM), `sinome' = "here" (LOTR, Aragorn's coronation) < {si} = "here". If this is `mi' = "in" suffixed, it may be an experiment at a separate inessive case "in X": `leneeme' may < {den} = "hole, passage, gap" > "opened door" metaphorical for "permission". In `iire' = "when" and `taare' = "then" FS (two of each), the suffix `-~re' is `ree' = "day" ("on which day", "on that day"). `-vaa-' in `arkandavaale' ONAM may be a "via" or "through" case, but is marginally acceptable as Patientive ("the road (leads from) their gate"). In `uuye seere indo-ninya' FS = "no rest is in my heart" or "my heart is not at rest", `seere' has no case ending to mean "at", so perhaps `indo-ninya' should be `indon-inya', and `-n-' is a possessive dative used to mean "in". If so, these two examples are unusual in having the case ending before the possessive pronoun suffix. It is best not to put too much weight on stray misfit forms in matter which never received Tolkien's final pre-publication revision. In published text, the allative was `-nna' after 1937, and `-nta' before c.1931 (CH). `-nta' occurs in OM1 and OM2 and MPD. See below re BD for an allative `-nda'. NARQ has a preposition `ne' which seems to mean the same as the instrumental case and may be its ancestor. Compare S. `o' = "of, from", and ETYM {gho} as ancestor of the Q. `o'-genitive; distinguish from Q. `o' = "with" {wo}. ITE says that prepositions take the accusative; but `nu huinenna' in NCPT has allative: "under death-shadow-to" = "to a position under d.s." and not "being under d.s. throughout the event": compare Latin `sub' with accusative and with ablative. `Et earello' = "out of the great sea" LOTR has ablative. OM-1931 has `X-in' instead of `X-Vr' everywhere in noun nom/acc plurals. This `-n' may < PE `-m', see {gho}. This was a passing phase, perhaps to avoid confusion with `-r' as locative, after which he returned to `-r' for plural as previously in NARQ, but changed to `-i' the plural of nouns in `-e' (except `-ie'). This nom.pl. is also used in the intensitive past active participle `kakainen' (KVNS), and in `X-e.en' (with two short `e''s in hiatus, SV-1931) = "like an X", which may be an `-n'-plural (< *`Xee-yaa-iim') of *`-ya' derivative adjectives. Doriath Ilkorin Elvish had genitive in `-a' or `-an' (ETYM {nauk}, and `dagnir Glaurunga' = "slayer of Glaurung" NIHH). Finnish (his inspiration for Quenya) has these cases: nominative, accusative, genitive, partitive (part of, some), essive (when he/she/it was/is an X, at the time X), translative (converted to an X, for the period of time X), abessive (lacking X), comitative (together with), instructive (X-ishly, in X's manner), prolative (via), allative (onto, to near), adessive (on, in, near), ablative (from, from near), illative (into), inessive (in), elative (from inside). [THE OLDER (1930'S) CASE SYSTEM] This case system appears in:- [Marquette Pronoun Declension] (MPD and MPB) [Bodleian Declension] (BD) [Various poems published in SV]: Oilima Markirya (OM), Nieninque (NIEN), Earendel (EAR). (`SP' = these poems taken together.) MPD is a list of unnamed Quenya cases for four demonstrative and/or 3p personal pronouns `entu-', `enti-', `enta-', `en', written between 1924 and 1936 on a sheet of paper whose back was later reused for LOTR draft. `en' declines from two roots, `en-' and 'end-'. It has a full dual listing, with 'ent-' replaced by `ass-' in `entu-', `enti-', `enta-'. The plural is missing, perhaps on another sheet which is now lost. BD is a list of unnamed Quenya cases written in 1936, for `-a' and `-o' and `-e' stems, but not consonant stems. Some `-e' stems (if the `-e' < PE `-i'?) turn the `-e' to `-i' before case endings. `i' + `i' > `i' if in last syllable, else `ii'. I name the cases from uses in NARQ and OM-1931. BD also has `sarya-' with nom.2 sg. `sare', nom.2 pl. `sari', perhaps < PE `-ya' (with short `a') (c.f. PE `talrunya' > Q `tallune' (not *`tallunya') = "sole of foot" ETYM), but he deleted it, perhaps changing his mind to letting PE short final `-ya' > `-i' > Q `-e'. The BD and MPD case endings are thus. The stated case uses are as deduced in VT#28 and VT#36. $ = `MPD', % = `SP', # = `BD':- sing dual $ plural Nom/Acc 1 -n # -nt -(i)n %; -r, -i # Nom/Acc 2 - #$ -t, (-s) Genitive -o #($), -n $ -o, -u irreg. -ron # Dative -s #, -r $ -ur -is/-isi # Instrumental -in #$, -inen $ -imet -inen # Locative -r %, -sse #$, -s $ -sset -sser #, -ssen #% " -nye $ -nya(t) -isen % (OM1 once) Partitive? (CGM)-iko $ -hto `-ika case' -ika # Allative -nda #, -nta $% -nta(s) -ndar # Ablative -llo #$ -llo, -llut -llor/-llon # Agentive? (CGM) -l $ Patientive -va #, -(w/y)a $ -va Similative -ndon #$ -ndos Abessive? (CGM) -hta $ DUAL: MPD dual is formed with any of the dual suffixes `s' and `t' (after the case ending), and `u' (before the case ending) or with none. The root `ass-' may itself mean `those two' and thus would be innately dual and so not need a dual suffix. This dual suffix `s' (which is mentioned nowhere else) may be related to this root `-ass'. When formed from `en(d)-', there is always a dual suffix. This surprisingly irregular dual may have been only provisional (CG in TL 10.51); Tolkien likely later rejected and replaced it with the somewhat later BD dual, which is in `-to-' (listed for `kiryato' = "2 ships") declined as a regular singular `-o'-stem with the case endings after the `t'. PLURAL: In BD the many-plural `-li' declines as `-i' stems, but has instr. `-liinen'. See below re OM1 `ondolin'. NOM/ACC: Nom/acc plurals both `-ar' and `-ai' of `a'-stems occur in these sources. See also under [Adjectives]:- `-ar' `-ai' PLET nom pl acc pl BD (distinction not explained, no case names given) NARQ noun nom/acc pl adjective nom/acc pl LOTR-period Quenya has `-ar' with nouns and `-e' with adjectives. In SP: Nouns have `-i' on some `e'-stems (perhaps those where the `e' < `i', e.g. `manduloomi' OM1 : `loome, loomi-' QL {LOMO}), and in `a'-stems and other `e'-stems replace the final vowel by `in' < *`iim'. There are no `o'-stem examples due to competition from the `-li' plural. There are three `-e.en' plurals of adjectives in `-ea'. The 3p subject pronoun suffixes `-ro' and `-re' add `-n' in the plural. MPD's missing plural likeliest followed SP in its nom/acc, since MPD follows them in the allative `-nta' and dat.sg. `-r'. The noun rule would give *`entin', *`ensin', *`entin', *`en(d)in'. The pronoun suffix rule would give *`enton', *`enten', *`entan', *`enden'. All these as well as the `-r' form would coincide sometimes with the singular of some case. Thus CG in TL 10.51 thinks that MPD's plural may have derived from a different root again. The nom/acc in `-n' (rather than endingless) was likely only used as nom. (not acc.) and then only to distinguish subject from object where word order is disarranged from S-V-O, e.g. `falmari langon veakiryo kiirier' = "the throat of the sea-ship cut the waves" EAR which is metris causa in order O-S-V plus a throat being very unusually the of the verb "cut". OM1 `ondo-li-n' may be a use of this `-n' on a many-plural. The difference between nom. and acc. is clear in MPD dual; but it is unclear which of the first 3 lines in MPD is meant to be which of nom. and acc. and perhaps also root etc: they are `entu ento ento; ensi ente ente; enta enta etta; en en enda/en'. In BD he rejected that and has the optional nom. `-n' in sg, and (perhaps to disambiguate) `-o' everywhere in gen.sg, and reverted to the early (NARQ) `-r/-i' nom/acc pl; this ended his experiment with PE *`-(ii)m' in nom/acc pl. GENITIVE: `-a' + `o' > `-o'. In `-o' stems `-o'+ `o' > `-u', following Greek contraction rules. In `-e' stems `-e'+`o' > `-o' or remains as `-eo'. In MPD the `-o' occurs only in dual (where it contracts oddly) and `end-' (4th col). DATIVE: BD dat.pl. `-is(i)' was likely taken from Greek (e.g. Greek 1st decl. has dat.pl. Attic `-ais', Ionic `-aisi') Dat.sg. `s' -> `r' by Germanic-style final rhotacizing, but stays as `s' as a short locative. This likely prompted his SV nom/acc pl. in PE *`-(ii)m' > Q. `-(i)n', to disambiguate it from dat.sg; BD's dat.sg has reverted to `-s'. LOCATIVE: The `-ye' may < a root meaning "there" (GCM). PARTITIVE?: GCM says "`-iko' < *`ii-gho'"? (ETYM {gho} calls the `-o' case "the old partitive".) Or (AA) perhaps < *`-ighgho' with double `gh' (c.f. loc, abl) and thus `gh' did not drop between vowels here. Or (AA) perhaps < adjective suffix `-ka' + `gho' (see -IKA here). GCM says "`-hto' < *`-at-gho' < ETYM {gho} + dual `t'?". -IKA: A derived adjective (c.f. the `-ka' suffix in section [Suffixes] below) (BD)? In BD only; perhaps a revised form of the MPD `-iko' case. ALLATIVE: In MPD the `-nt-' > `-at-' after consonant Greek-fashion in singular and dual `end-' (4th column). PATIENTIVE: MPD calls it the `adjectival'. The ending variation in MPD is unexpected; BD reverts to regular `-va'. SIMILATIVE: e.g. `taure-lasse-li-ndon' = "like forest leaves" OM1. OM1 `wilwarindon' = "like a butterfly" may be haplology for *`wilwarindendon'. I prefer the name `similative' to CG's `comparative', as `comparative' also means the "more X" or "X-er" form of adjectives. ABESSIVE?: < *`-tka', c.f. `-ika' (MPB)? Or < *`-gho-ta' (ETYM {gho} + `-ta' as in MPD allative, being an abessive case (`X-hta' = "lacking X") (CGM)? In TL 10.51 CG says "... some of the points he [= AA in TL 10.44 & 10.46, also stated here] touches on involve chains of reasoning that depend on an assumption that the chart makes sense as a whole, even though with no explicit guidance from Tolkien everything anyone says about it remains theoretical.". Perhaps thus MPD was only a trial working out of forms, and can be trusted for the list of case endings as above not as a final decision as to which of the many differences from a regular `entu- enti- enta- en(d)- + case endings as above' format are to be used as authorititive when writing in 1930's Quenya, or as to what the LOTR-period independent 3p personal pronouns would be. MPD has many declension irregularities: original `ti' > `si' as in Attic Greek; apparent dissimilations; unpredictability of when `u' and `i' open to `o' and `e'; variations in the dual suffix used, etc, `ei' > `ii'. The third nom/acc row form for `enta' is unexpectedly but undoubtedly `etta'. MPD `end-' and `en-', which are the only 1930's examples of consonant stems, behave unlike each other; `end-' interpolates usually `i', a few times `u' or `a', at `dry joints' (i.e. consonant meeting consonant at stem+ending junctions) when the case ending also starts with a consonant; `en-' has instr.sg. `iinen', partitive sg. `enko', and forms most of its available cases as from `e-'. In the dual abl. `-llo + t' > `-llut' as if `-llo' < PE *`-llu'. [NARQELION CASES] NARQ, written about 1915, should give a very early form of the Quenya case system, long before the 1930's `-(ii)m' nom/acc plural experiment and its consequences. In it, nom/acc noun plurals add `r' after vowel, e.g. `lasser' "leaves", `sangar' "throngs". The `-i' plural `sinqi' = "jewels" probably < singular *`sinco', stem *`sinqu-'. All adjective nom/acc plurals are in `-ai' from singular `-a'. The only other likely case forms are: locative `tarasse' = "aloft", and (hyphens in original) two forms with case prefixes: `n-alalmino' = "(begotten) of the elm" (twice), and `ne-suume' = "by means of a gust of wind". For `kuluvai' = "golden" (pl.) see above. [THE SUFFIX `-ya'] `-ya' < PE `-yaa' makes adjectives from nouns:- Proto-Eldarin Quenya -aa -aayaa -a -ea 2 uses in LOTR and many in ETYM -ee -eeyaa -e/(nil) -ea -uu -uuyaa -u -uuya Alduuya' LOTR = "(day) of the Two Trees)" < *`aldu' = "2 trees" < PE *`alduu' -ii -iiyaa -i -iiya so likely *`Ainuuya' = "of an Ainu" & *`taariiya' = "queenly" < `taari' = "queen" khaya khayayaa hai- haaya/haiya' "far" {khaya} ETYM oi oiyaa oi oia "everlasting" < `oi' = "ever" {oy} ETYM ulu uluyaa ul(u)- uulea (analogical) "pouring, flowing" {ulu} ETYM There are so far no known uses of `-ya' after other vowels. Would *`-oya' and *`-uya' > `-oa', or even analogical `-ea'? Examples of `-ya' on apparent non-compound consonant stems are six of the LOTR weekday names: `Elenya', `Anarya', `Isilya', `Alduuya' / `Aldea', `Menelya', `Earenya', `Valanya'. Either final vowel loss (if there was a final vowel) early spread by analogy through the declension as in `Elenna' < *`elen-nna', or PE unstressed short `a' (and other short vowels?) dropped before `-yV' (c.f. `-aye' (with short `a') > `-ie' in VT#27 p36). [ADJECTIVES] When qualifying a nom/acc plural noun: adjectives in `-a' form nom/acc plural in `-e' < *`-ai' < *`-aaii' in LOTR period: `ilye tier' NAM = "all roads", `yeeni uunootime' NAM = "years uncounted", etc. Adjectives in `-ea' < *`aayaa' have LOTR nom.pl `-ie' < *`-aayaaii' (`laurie' NAM), and SV-1931 (3 uses) nom.pl `-e.en' (two e's in hiatus) (CHJ) < *`-aayaaiim' (see below re `n'-plurals). LOTR usage agrees with NARQ's very early use of `-ar' and `-ai'; thus likely the same was done when he wrote BD. But PLET (which was written after LOTR) disagrees, and uses `-ai' as a separate BQ noun accusative. In other cases (and also in nom./acc.pl. in OM-NEW `raamar sisiilala' = "wings shining"), the adjective remains uninflected: allative pl. `atalantea mindonnar' = "onto fallen towers" OM-NEW. In poetry an adjective can be inflected instead of a noun that it qualifies, but the noun agrees in number: `axor ilka-la-nnar' = "on gleaming bones" OM-NEW; there is no evidence what would happen in the gen.pl., which in nouns is formed by nom.pl.+ PE `-ghoom' > Q `-on' (`ilya' : *`ilyaaiighoom' > *`ilyeon', or *`ilyaron'? `laurea' : *`glauraayaaiighoom' > *`lauri(e)on', or *`laurieron'?). They would decline like nouns in plural cases where also nouns added `-i-' before the ending, the `-ai' persisting before a consonant (`ilya' : `ilyain', `laurea' : `laureain' dat.pl., and thus the other short place cases and the instrumental). If an adjective is used as a noun, e.g. `laurea' as "golden object", would it decline as a noun (nom/acc pl. `laurear', etc)? In SP the rule is often different, with unusual case endings on the adjectives:- Nien. wingild-in wilwarinde.en nom.sg foam-fays butterfly-like Nien. taal-in paptalasselinde.en nom?sg feet falling-leaf-like Ear. tyulm-in aiqa-li-n nom.pl mast tall OM1 kirya ninqe pinilya acc.sg ship white small OM1 fuin oilima-ite (twice) acc.sg evening last (after preposition) OM2 rusta kirya acc.sg broken ship OM2 hui oilima acc.sg evening last OM2 hui oilima-ite acc.sg evening last Ear. o silqelosse.en acc.sg with (hair) silver-blossom-like Ear. laiqa-li linq-i falmar-i acc.pl green wet waves OM1 ailini-sse alkara-in loc.sg shore shining OM1 aili-ssen oilima-isen loc.pl beaches last OM1 alkari-ssen oilima-in loc.pl light-rays last OM2 laiqa ondo-li-ssen loc.pl green rocks and many examples of adjective + noun expressed by compound words (e.g. Ear. `ninqe-ruvi-sse' = "white horse on"). [PRONOUNS] ITE, with NAM and LOTR as its only available sources, tried to reconstruct the personal pronouns; but forms published since have changed the picture, particularly in 3p. They represent a long series of his ideas down his years; it is not possible to make a consistent complete personal pronoun table using all known old and new forms without clashes. I know of these forms:- "I":- `ni' ({ni} PE), `ni' (ARCT, before verb), `ni-' (SD), `inye' (BOLT; LR p61), `-~ne' (LR), `-nye-' (LOTR before another suffix), `-n' (LOTR) "for me":- `nin' (NAM, dative) "to me":- `nin' (LR p63) "my":- `-nya' (LOTR `hildi-nya-r' = "my successors", etc) "us two" (exclusive):- `met' (NAM) "our" (inclusive):- `-lva' (`omentie-lv(a)-o' = "of our meeting" (LOTR) "our" (inclusive):- `-mma' (`omentie-mma-n' = "of our meeting" (LOTR draft) "our" (exclusive):- `-lma' (HOME, message from Tolkien to his printer re `omentielvo') "we" (exclusive):- `-mme' (WOTJ p371) "for us" (probably):- `ng'win' (1954, VT21 pp 6,7,9 CH) "thou":- `elye, -lye' (in `elye hiruva', `hiruvalye') (NAM) "thee":- `le' (LOTR, Sindarin; RGEO says it was taken from Quenya) "thee":- `tye-' (LR (hyphen in original)), `tye' (ARCT after verb) "thine":- `lyaa' (NARQ) "your":- `-lda' (WOTJ p.369; LOTR (Aragorn in Loorien, WOTJ p369 gives meaning.) "you" (nom. pl.):- `-lle' (ONAM, VT#24) (meaning proposed by CH) 2p imperative: `-t' sg, `-l' pl (WOTJ p364); and with no ending. "it" (acc.):- `-s' (LOTR,COTE), `-~ta' (FS), `en' (SD before verb, FS p63), `-t' (SD), `-n-'? (EAR) "he", "it" (nom.):- `-r / -~ro' (OM1, EAR) (usage varies between these two forms) "he" (nom.) + "it"(acc.):- `-roota' (LR p63) "he":- `-~ro' (FS), `e-'? (KVNS); `su', `suu', `so', `soo', `-so' ({s} (PE)) "she":- `-~re' (OM1, EAR (for a ship), KNVM); `si', `sii', `se', `see', `-se' ({s} (PE)) "her" (genitive):- `-rya' (NAM, NARQ) "he, she, it" and their dual: see MPD above. "they":- `-~ron' (OM1, EAR), `-lto' (FS, 3 uses), `-nte' (COTE `tiruvantes', SD `sakkante' haplology < *`sakkantente' `n'-past < *`sakkata' emphatic < {skat}) "they" (feminine):- `-~ren' (OM1, EAR) "them":- `sen-' (KVNS), `-ren-' (KVNS after another prefix) "them" (dual? pl?):- `te' (LOTR), `-t' (LOTR, see below) impersonal subject:- `-o' (OM1 `kildo' = "one saw") ETYM at {s} gives these Sindarin 3rd person pronouns:- MASC FEM NEUT singular ho, hon, hono he, hen, hene ha, hana plural huin hiiin hein Demonstratives often become or are used as 3p pronouns. In Quenya I know of:- ass- "those two" CGM en "this", "that", 3p pronoun SD (= "it"), MPD, etc en- "yonder", "future" ETYM {en} end- "that" (?) CGM ensi "she", "that female" CGM enta "it", "that neuter" CGM enta "that yonder" ETYM {en} entu "he", "that male" CGM i "the", "who", clause-introducer LOTR, COTE, FS, etc s-,-r- 3p pronoun SV, LOTR, etc {s} demonstrative ETYM {si} "this", "here", "now" ETYM sina "this" COTE {ta} demonstrative ETYM tana "the abovementioned" (= anaphoric) ETYM {ta} tanya "that" OM-1931 toi "they" LR p72 ya "who", "which", "what" NAM, ARCT, NARQ {ya} "there", "over there", "ago" ETYM yana "that", "the former" ETYM {ya} The `-nt-' in the MPD independent 3p personal pronouns is likely intended to recall the `-nt-' of the 3p verb suffix. Declined LOTR-style they would be `ento (entu-), ente (enti-), enta (enta-), en(d)-'. Reanalysis of the evidence leads to something like this:- Pronouns at LOTR period. Uppercased forms are attested:- NOM ACC GEN DAT 1 sg INYE,enye,-NYE-,N ni enya,-NYA NIN 2 sg familiar etye,-tye,(-t?) ti etya,-tya tin 2 sg respectful ELYE,-LYE,(-l?) li elya,-lya,-LDA lin 3 M sg eero??,-~RO so,-S erwa,-rwa son 3 F sg eere??,-~RE se,-s erya,-RYA sen 1 excl pl elme,-lme,-MME ME elma,-LMA men? elmen? 1 incl pl elve,-lve ve elva,-LVA ven? elven? 2 pl familiar elte,-lte te elta,-lta ten? elten? 2 pl respectful elle,-LLE LE ella,-lla len? ellen? 3 pl ente,-NTE (TE) enta??,-nta enten? 1 excl dual elmet?,-lmet?? MET elmeto? meten? elmeten? 1 incl dual elvet?,-lvet?? vet elveto? veten? elveten? 2 dual familiar eltet?,-ltet?? tet elteto? teten? elteten? 2 dual resp ellet?,-llet?? let elleto? leten? elleten? 3 dual entet?,-ntet?? (tet?) enteto? enteten? impersonal -O (known as a 1930's form) imperative sg Nil or `-T' imperative pl Nil or `-L' 3 sg en en EN eno enen 3 pl en eni eni enion eniin 3 dual en enet? enet? eneto? eneten? The separate nominative forms are used for emphasis. Words ending in consonants add/regain a final vowel before genitive uffixes. The demonstrative `en' is used as "it", "they", perhaps as neuter only. Do not confuse cases of the demonstrative `enta' with the `-nt-' 3p pronoun. After dropping `s/r' as 3 pl, he tried 3 pl nom. `-LTO' (FS) before settling on `-nt'. `ng'win' in a NAM variant fragment (`sii man i.yulmar ng'win enquatuva') is plural, if pl. `yulmar' implies one cup each; it may < unrecorded PE *{ngwi} as a revival of BOLT PE `nguo' > Gnomish `go' (>> ETYM {gho} > S. `o', Ilkorin `go') = "together" (CH in VT21 pp 6,7,9). Gen. of *`ng'wi-' would likely be *`-ng'wya' > *`-nya' coinciding with 1p sg, or perhaps > *`-ngwa'. More evidence is needed before deciding the status of `ng'wi-' as a pronoun. WOTJ p.407 has Telerin `-nguo' = "our (excl.)". The WOTJ & (said in WOTJ to be) LOTR 2p resp. `-lda' is a very unexpected form, even if I treat it as < *`-lna' or *`-dla'. (See re `Meletyalda' below.) WOTJ p364 may imply that `-t' sg., `-l' pl., are imperative only; it says that the endingless imperative is also valid. 1920's and 1930's forms, probably dropped while revising LOTR drafts:- 1sg 2 sg 2 sg 3 sg 3 sg 3 pl 3pl 3 dual 3 dual fam resp masc fem masc fem masc fem NOM NI-,etc TYE- lye -~RO,- -~RE -~RON -~REN -~rot? -~ret? ACC ni tye lye so se son,-REN sen sot? set? GEN nya tya LYAA hwa? hya? sonon?? senon?? soto? seto? or for 3rd person use a demonstrative or the MPD forms. No other cases are known for 1p and 2p, but they were probably formed from the acc. In 3rd person `s-' and `h-' > `-r-' if used as a suffix. 1p pl and 2p pl not known. `e-' in `eren-' KVNS may be an emphatic prefix as in LOTR-period `e-lye' NAM. Dropping the `s-'/`-r-' 3 pl forms before FS and LOTR may be part of removing all remaining nom and acc uses of the `-n' (< PE `-m') noun etc plural which was common in SV. With a separate noun subject (or in LOTR times with a separate emphatic pronoun subject), the verb has no suffix with a singular subject, and with a plural subject adds `-R'; no dual examples known. But verbs with noun subjects have subject pronoun suffixes in:- (1) OM1 (not NIEN or EAR; all OM2's verbs are `man X-uva?' "who will X?"); (2) `sakkante' NCP; (3) `nai i haarar' (COTE) = "I-wish-that they who sit ... will guard it" where `i' used as a relative pronoun is effectively subject of both verbs (unless `-nte' = "they" is treated as the antecedent of the `i'). Distinguish from e.g. `ree' = "day", `-ve' = "-ly", words < {tegh} = "line". `-l' 2 sg nom is not in Tolkien but occurs in Quenya written by other people. Other cases were likely formed from the stems used here for the dative, e.g. `nillo' = "from me" (ablative), `liva' (or `liiva'?) = "of [= done to] thee". `lyaa' written c.50 years before NAM may or may not invalidate my and ITE's guess that LOTR-period pronoun genitives prefix `e-' when free-standing like NAM nom. `elye' / `-lye'; but e.g. LOTR `-lva' = "our" would not be a phonetically valid Quenya form as free-standing *`lvaa'. `-ne' = "I" LR >> `-n' and `-nye-' LOTR likely to avoid resemblance to `n'-past in `-CCa'-type verbs, e.g. `antaane' = "I give" and `antane' = "gave". Both uses of `tye' are between friends, but LOTR `elye' and `-lye' and `le' are where respect or reverence is likelier (CH). `le' = "thee" fits easier if it was plural used respectfully as singular as in French etc or influenced by native Sindarin. `nimaruva' >> `maruvan' = "I will dwell" and `nilendie' >> `lendien' >> `utuulien' = "I have come" (SD >> LOTR), and LOTR "praise them!" as `laita te!' (not *`senlaite' or the like), are likely part of a total rejection of pronoun prefixes as he was preparing LOTR, in which case any LOTR-period familiar 2p pronoun would be the same as respectful but with `t' for `l' everywhere - which rule extended into the plural gives `te' and its cases (`tenna' etc) three meanings: "you" pl fam and "they" (not nom or gen), and time prepositions < PE {tegh} = "line" (`tenna' = "until" LOTR and COTE). Context resolves this ambiguity in `laita te' but not necessarily in all uses. The change `-lt-' >> -`nt-' in 3 pl may have been partly to free `-lt-' for another use, which is why I suspect that -`lt-' 2 pl familiar existed in LOTR times. I suspect that Tolkien never finally cleared of uncertainties all less-used parts of the pronoun table. I could disambiguate by forming those cases of 3p pl from a demonstrative (`en' or `enta' etc) instead, and time prepositions from other time roots such as `luu' or `ree' or `yeen' as suitable for each use; `te' = "them" (in `laita te' LOTR) is undeniable but is a late and perhaps hurried change to the text from SD ch.V `laitalle' = "you (pl) praise", perhaps to avoid using a statement as an imperative. The competition in 3p between the roots {s} and {ta} and {en} is shown by `en' = "it" (acc) (SD ch.VI) >> `-t' in `intuuviet' (SD ch.VI) >> `-s' in `Yee! Utuuvienyes!' (LOTR). The `-r-' suffixes in OM1 and EAR are not a middle voice (intransitivising) ending, as in `langon veakiryo kiirier falmari' (words rearranged here to S-V-O order) = "the throat of the sea-ship cut the waves" EAR it occurs on a verb with an object. `langon' is the subject, not agentive dative, as if `falmari' pl. was the subject, `kiirier' would be `kiirieeron' pl. agreeing with it (c.f. `kautaaron' EAR line 9). Duals: `met' NAM = "me and her" is the plural plus the dual suffix `-t'; for forming dual from plural compare `aldaru' = "two trees" KVNM. `laituvalmet' LOTR = "we(excl) will praise them", with "them" probably dual, is a late, perhaps hurried, change << `laituvalme' = "we(excl) will praise", and could be haplology for *`laituvalme tet': compare `laita te' (not *`laitat') = "praise them". In `maa-rya-t' = "hand-her-two" NAM the `-t' dualizes not the pronoun but the thing owned. I venture a complete tentative reconstruction of the dual. We have no Tolkien reflexive pronouns ("self", "each other") yet, and they are needed in at least the 3rd person. We cannot form them Greek-fashion by suffixing "same" to ordinary pronouns, as there is still no known Quenya for "same". If they ever turn up and are brief, some may want to use them as an intransitivizer to produce a middle voice, as in some real languages: see below. They should have nominatives, to distinguish e.g. "A said that he (= A) was there" from "A said that `he' (not A) was there". Compare OM-NEW where he had to supply a prefix `na-' or `a-' to distinguish active from middle. If this distinction is kept to, as in Greek etc, "the man burns (active)." would clearly mean "the man burns something (no need to say what)", whereas English "the man burns." without object is liable to misunderstanding. Being finally driven to vinyakaarie here when translating `The Annals of Valinor' into Quenya, I invented `atya' = "same", `atso' = "himself", "itself", `atse' = "herself", < {at} = "again" + {s}, like Greek `autos' = "same", "self" < Greek `au' = "again". Or try a form with `p', which is unused in Quenya pronouns so far. OM1 `kildo' = "one saw" may < *`kilneo' (c.f. `kiluva' (OM2,CHJ)) and contain an `-o' impersonal subject pronoun suffix (see below re the `n'-past); if this one-off c.1930 form is valid in LOTR-style Quenya (where it would be *`kildeo', as `eo' > `o' optionally in BD but remains as `eo' in PLET), it provides a short passive with no agent possible (but not a reflexive): `etekeo parma' = "one has written a book" = "a book has been written". Quenya, like English, uses feminine pronouns only for biological females and does not have random gender of inanimate nouns (except feminine for ships in SV). The distinct forms for "we" as to whether the listener is included is a useful feature to avoid ambiguity misunderstandings followed by delaying wordy explanations of what was meant. Another is distinct singular and plural 2p pronouns ("thou"/"ye"). English and French polite disuse of "thou"/`tu', making "you" ambiguous as to whether others as well as the listener are included, led to the USA colloquial plural "youse", and to French pronunciations like `vuzet Alzherien?' (sg., no liaison) / `vuzet zAlzherien?' (pl., liaison) for `vous etes Algerienne?' = "are you Algerian?". Modern Greek developed a 2p plural like "youse" when the classical plural coincided with "we" due to phonetic change; Dutch got a familiar 2p plural `jullie' to restore singular / plural distinction after the old singular `du' died out leaving the old plural `jij' as both singular and plural (and now `jij' is itself familiar singular, and a new polite form `U' has arisen). QAE p369 gives new uses of pronouns: (1) Instead of proper name genitives: e.g. `Olwe's house' could be `koarya Olwe' = "his house, Olwe" (c.f. Turkish `Mehmet cami' = "Mehmet( hi)s mosque", and a former English habit of spelling e.g. "St.James's Park" as "St.James his park".) This form (as `cavaria Olue'), plus simple apposition without case ending (`Olue cava' or `cava Olue'), was a usual way of expressing ownership by one person in Telerin. (2) An adjective with a proper name or title (usually vocative) had or could have a pronoun suffix referring to the name, e.g. `(Aran) Meletya-lda' = "Your Majesty", and explains LOTR `Arwen vanim(a|e)lda, namaarie!' as "Arwen beautiful-your, farewell". This pronoun suffix form `-lda' is very unexpected and un-analogical: several known singular pronouns listed above all point to 2p resp. sg. *`-lya' or similar like the Narqelion form. Also, I would have expected e.g. *`vanim-ie-l(y|d)a' = "your beautifulness": `vanim(a|e)-l(y|d)a' should mean "the beautiful one belonging to you", as if addressed to a parent or husband than to herself. It could be that `vanimalda' (1st edition), `vamimelda' (2nd edition) was first composed with `vanima-Elda' = "beautiful Elf" in mind, and that the idea that `-lda' is a pronoun, and also the idea of putting pronouns on adjectives thus, are a later retro-fit to `vanim(e|a)lda' as published and thus not easily altered. (3) WOTJ p363 may say (the wording is unclear) that the 2p pronominal `l' < `d' and that that `d' persists in 2p pronouns in other Elvish languages. [VERBS] LOTR-period Quenya verbs form: present `-a'; future `-uva'; past `-e' (and `-ie'); present participle active `-ala' (OM-NEW); perfect passive participle `-ina' or `-aina' (OM-NEW, FS), or `-na' (LOTR `nuquerna' = "inverted", NAM `yulda' < *`yulna' = "draught of drink" (formed like the `n'-past), and `turuun'' (elided) = "overmastered" NIHH < {tur} which may be an experiment at `-uuna' by analogy with `-uva'. The infinitive `-ie' is a fully declined noun. Clause subordination tense sequence rules are known for past purpose clauses: see below. For wishes and 3p imperative, put `nai' = "I command or wish that" at the start of the sentence. ITE guesses `-uve' as "future in the past", e.g. *`tekuve' = "he was going to write", Latin `scripturus erat'. The imperative is the present stem, e.g. `laita' = "praise!"; QAE p364 quotes optional imperative suffixes `-t' sg., `-l' pl. The `n'-past is formed thus: Remove the final vowel and then: If the stem ends in two consonants, append `-ane', else append `-ne'. If an illegal consonant pair results, assimilate and/or metathesize. Irregular `n'-pasts: `fary-' : `farne' = "suffice", `ist-' : `sinte' = "know", `rer-' : `rende' = "sow", `taap-' : `tampe' = "stop, block", `top-' : `tompe' = "cover", `u-, um-' : `uume' = "do not, be not", `uly-' : `ulyane' (trans.), `ulle' (intrans.) = "pour", `vany-' : `vanne' "go, depart, disappear", `vil-' : `ville' = "fly", `kil-' : `kildo' = "see" (OM1: see above re the `-o'). `-ane' occurs as an active past participle in OM2, but >> `-ala' in its later version OM-NEW. Perhaps infinitives (replace the final vowel by `-~ie', e.g. `na-maar-ie' = "towards home (go)ing" = "farewell" : `mar' = "home") and active participles (append `-la') were formed from any tense, e.g. *`tekuuvie' = "to be going to write", *`utuuliela' = "having found". There are these past and related tense forms in `-ie' in LOTR and its drafts:- PRESENT (ETYM) PAST PERFECT `tulin' {tul} "I come" - `utuulie-n' (LOTR) "come" `wanya' {wan} `vaanier' (NAM in LOTR) `avaanier' (NAM in RGEO) "go" - `tuuvien' (WOTR) `utuuvie-nye-s' (LOTR) "find" The `-ie' pasts here may be from verbs in `-ya' (`tulya', `wanya', `tuvya'), as if `-Cye' > `-~Cie', (also in infinitives, but not in pronoun nominatives): `tulin' could come from `tula-' or `tulya-'. If so, the perfect may be likewise, e.g. "A has made" would be `A akare' rather than `A akaarie'. But in QAE p366 he changed the etymology of `vaan-' and made it irregular: {wa} (instead of ETYM {wan}) > pres. `auta-' = "go away", archaic MET past `anwe', past `vaane' & perf. `avaanie' = "vanished, lost, etc"; past `oante', perf. `oantie' = "gone to another place" (`oa-' < `awa-'); perf. `avaanie' < `awaani.iee' (with two short i's) (influence from `n'-past) < `awaawi.ie'. If so, "A has made" would be "A akaarie"; but perfect would clash with infinitive and optative in unreduplicatable forms such as `oant-'; QAE p66 says that `vaanie' (as well as being an optative, see below) is not uncommon in poetry as unaugmented perfect. What dropped to leave the vowel cluster `i.ie'? Does the perfect derive from an ancient compound tense, e.g. *`akare-aya-' with the root of `ea', = "he made, and in the resulting state."? If all words starting with a vowel originally started with glottal stop as in Arabic and German (here spelt `#'), perhaps *`#akaree #ayaa' > *`#akare#yaa' < *`#akari#yee' (analogical change of ending to a past form), and the (second) `#' survived long enough to support the `y' and allow `-ye' > `-ie' as if after a consonant. QAE p367 gives some Telerin verb past and perfect forms: they are similar to Quenya's. QAE p366 also says there was a period in Quenya history when the stress receded to the beginning but not onto verb augments and reduplications. I know of no Tolkien Quenya durative tenses so far; an imperfect if needed could perhaps be formed English-fashion e.g. `ne tekala' = "was writing". QAE p415 note 30 has one of the few bits of information about Sindarin verbs. [Subjunctive, optative, subordination] ARCT `X nea' = "I hope that X will be" is a hope-optative. KVNS `i' shows that NAM `nai C' < `na i C' = "let be the state or event that C". Between commanding and hoping are many intergrades; RGEO says `nai' is "a wish rather then a hope". Likely `nai C' = "get C done", i.e. imperatives and requests, and hoping with no ability to help is `-ea' (< PE *`-aya'?). But extending analogy too far from one form is risky, e.g. KVNS invalidates a subjunctive e.g. `tirai A' = "let A see" guessed by analogy with `nai' (TL 1.93). Perhaps `nai' should be `eai' when referring to conditions permanent by nature rather than temporary. KVNS `i -koitanie' = "in order that ... should awaken", and NCPT `ullier' = "they might pour" (conjunction missing, as NCPT is written as fragments), are a subordination-to-past optative formed from the `n'-past (not the `X-e' past, as `X-ie' is also infinitive). We now know that past purpose clauses are headed by `i' and have their verbs in the `-(a)nie' tense. I do not know yet how present purpose clauses were made: they could not use the present indicative, as that would not distinguish "in order that" from e.g. "with the (perhaps unintended) result that"; perhaps for present purpose clauses they used the `-ea' tense. `nilendie' >> `lendien' = "I have come" < {led} is not optative but perfect tense, but it >> `utuulien' before publication. We await more new text. Verbs in `-ya' may have changed both `-Cyie' and `-Cye' to `-~Cie'`', or to `-~Cye', or may have kept them distinct as `-~Ciye' (or analogical `-Cyie') and `~Cie'. WOTJ p365 and p404 mentions a PE imperative particle `aa' (which became the Sindarin imperative in `-o'): a Quenya example is the wish-optative `Aa vala Manwe!' = "May Manwe order it!". [Being] COTE uses two verbs "be", `na' and `ea'. From the usage there I suspect that they differed like the two verbs "be" in Irish and Spanish:- Quenya Irish Spanish be temporarily or as an event na ta estar be permanently or longterm or by nature ea is ser They may form their tenses thus:- Present Pres.pl. e-Past n-Past Perfect Fut Pr.Part Infin na nar ne nane ane nuuva nala nie ea ear e.e? ie? aine aaye ayuva eala aiye Distinguish present plurals from `naar' = "fire" and `ear' = "sea". What is the intensitive of `na'? `Nan' and `nen' both have other meanings. If `ea' < PE `eya' or `aya' (ETYM does not list it), `Ainu' and `aina' < {ayan} may be related to it: "one who (always has) existed". NAM has `naa', but ETYM has `na'; `na' may be long only when emphasized. In LR (FS etc), and perhaps in KVMN, for "be" he used `ye' (future `yeeva'), or suffixed to predicate adjectives etc as `-ie' (future `-ieeva'); but he later rejected this form and went back to `na' and `ea'. [Causative] In some English verbs, e.g. "burn", the transitive use is causative of the intransitive use. Some languages have separate active and middle voices to represent these uses; but not Quenya as so far known except "pour" = `ulya' present, `ulyane' past tr., `ulle' past intr. (ETYM), and 3 uses of a causative `-ta': "burn" `usta' tr., `urya' intr. (BOLT); "wake" `koiva' (>> `kuiva') intr., `koita' tr. (KVNS); `tulya' = "come", `tulta' = "send for" (ETYM). This `-ta' would give e.g. `karneta' = "redden" (transitive), `tirata' = "cause to see" = "show". Using `kara' = "make" can be ambiguous: `kara so karne' = "make it red" = "cause it to be red", or "make it and make it red". EAR has `alka-n-tam-ee-ren' = "shine it? make past they(fem)" = "(foam maidens) made it shine". ['Becoming', inchoative] This is lacking in available Quenya; but ETYM {es} = "start, first" >> *{yes} (LOTR `yestare' = "first day of the year") (LMBR) (disambiguating from {es} "name" but clashing with ETYM {yes} = "desire" (or obsoleting it, as {id} and {mer} also = "desire") (but he did not mark this change in ETYM)). If so, we could have: `yera' = "become", `yesta' = "begin", `X-ieera' = "start to X", "become an X", "start to be X". [Passive] FS `hosta-in-iee-va' = "number-ed will be", which with the usual word for "be" becomes `hostaina nuva', i.e. a passive participle with "be". Thus `ea' and `na' = "be" need full sets of tenses; and nouns need an agentive case, which my `dedative' case (e.g. `kiryal', see above) could be used for (distinguish from the instrumental case). Passives could use `yera' = "become" (c.f. German `werden'): present `tenka yera' > `tenkieera' = "it is being written", perfect `tenka na' = "it is (already) written" = "it has been written", so the passive participle is always perfect. The English passive participle often slides into present tense meaning. Or the two FS forms `-aina' and `-ina' could be used one each as present and perfect passive participles. [Translating the English infinitive] In Quenya, as in Vedic Sanskrit, the verb infinitive is a declinable noun which can also be used to translate "X-ing" (gerund, distinguish from participle), "X-tion", "X-ance" and similar abstract nouns, and care must be taken what case to use. The English infinitive "to X" has various uses. Examples of `X-ie' in various cases are:- Nom X-ie ea maara er Melkooren X-ing is useful only to Melkor Acc meran X-ie I want to X Acc nyaren so X-ie I told him to X Gen X-io of X-ing Dat X-ien with the purpose or likely result of X-ing Abl X-iello from X-ing (or other cases, see {NOUN CASES} and {COMPOUND SENTENCES AND CONJUNCTIONS}) Vedic renders "go or make a journey for the purpose of X-ing" by accusative (Q. allative) of X instead of dative, e.g. "I went to town bread". With `kakainen', compare `sisiilala' = "shining" and `fifiirala' = "fading away" (OM-NEW), and `lalantila' = "falling one by one" (NARQ), and (KVNS) (CHJ) `se-alaalan' = "they grow (intrans)" < {gala} (ETYM) = "thrive". How much is it allowed to form two tenses at once to produce other tenses? E.g. `wanya' = "go" : `wanne' = "went" :: `awaanie' = "have gone" : `awanne' = "had gone" (pluperfect); `awanyuva' = "will have gone" (future perfect); `X awanyea' = "I hope that X has gone" (perfect hope-optative). Other verbs in `-ea' < *`-eya' or *`-aya' may conjugate like `ea'; PE `ei' > Q. `ai' (e.g. *`kea' = "lie down" (intens. part `kakainen' KVNS, OM1 `kaire' "she (= a ship) lay"). He formerly (FS, ETYM) conjugated Quenya verbs somewhat differently for a while. The present ended in various vowels; the future added `-~va'. To get the LOTR-period present of these verbs, replace the end vowel by `-a'. If the available form is 1p sg present, e.g. `karin' = "I make", the final `-in' could represent stem `-a' or `-ya', as `-Cyin' probably > `-Cin'. FS has one `-uva' future and two `-aava' futures. A prefix or suffix is needed to make verb forms durative ("be in the course of X-ing"), e.g. "he was writing" v. "he wrote": see below re {Time}. [VOCABULARY] When translating randomly chosen English into Quenya, not long passes before meeting a word with no good translation, e.g. "buy", "sell", "obtain", "carry", "take", and many other small everyday matters that never happened to occur in anything that Tolkien needed in Quenya. E.g. ETYM's {mbak} = "trade" S. `bakh-' does not distinguish buying from selling, and in Q. it coincides with {mak} = "slay, sword": `makar' = "trader" in ETYM but "swordsman" in LOTR (in `Menelmakar' = "Orion" (constellation)). `kolindo' = "bearer" LOTR is the only evidence for *`kola' = "bear", "carry" (< PE *{kol}?), and then only because ETYM `melindo' < {mel} confirms that `-indo' is an agent suffix. There will be need to record and coordinate the resulting new roots, and new words (including many verbs) from existing roots, which people will have to invent. [COMPOUND SENTENCES AND CONJUNCTIONS] Known Quenya is very short of conjunctions, except "and" and the relative pronoun. This lack forced Bjorn Fromen in his poem `Valinoorenna' in ITE to write two simple sentences often where he likely wanted to (and the English does) use a compound sentence. The Quenya clause conjunctions that I know of are:- `yassen' = "in which", "where" in Namaarie. This needs a noun antecedent. `iire' / `taare' = in FS "when / then", but the suffix `-~re' = "day" seems to limit its use to specifying date, not time of day. `ten' = "for them" (dative of `elte') used as the conjunction "for" (in "for now is the ending") in FS. This (like the English which it literally translates) is probably short for "for those (reasons)"; but it is ambiguous. `i' with optative = "in order that" in the past: KVNS `i erenekkoitanie'. [Vinyakaarie of conjunctions] Many languages develop conjunctions from relatives and demonstratives; but these then are liable to be ambiguous between referring to a whole clause and to one word. KVNS shows that `i' = "the" and sometimes "which" is also used before a clause to mean "the following information or event or state". If it could take case endings (as it perhaps does in `iire' in FS, see below), we have:- `Queten Mardil na sinome'. = "I said Mardil is here". `Queten Mardil na sinome'. = "I spoke Mardil's being here". (PLET's `short locative' used as `respective', see above) `Si apsa na wanyalme ostonna'. = "This food is we go to town". `Terhatelye weere atarilya quete sos'. = "You broke the agreement (letting) your father talk about it". (`sos' is respective of `so' = "him, it", see above) `Kola si apsa omentuvalme'. = "Carry this food we will meet". `Ilya si uu nelye sinome'. = "All this you were not here". (`ne' = past of `na'; `i' treated as place/time = "place/time of when--".) `Uu melan ero ar erwar oi mankar ar uu karar vinyar'. = "I do not like him and his (people)'s forever trading and not [= instead of] making [their own] new (items)". (In Quenya, case of `i' + clause with finite verbs here.) Coordination between sentences can be conveniently expressed by these new case endings which I invented and are primarily for use on `i' but can be used on infinitives and ordinary nouns etc:- A-ki if A exists/is there X iki Y X if Y; Y implies X A-ke caused by A X ike Y X because Y A-kei intentionally caused by A X ikei Y in order that X, Y A-ka if and only if A exists e X ika Y X if and only if Y; X eqv Y if and only if A is there A-ko causing A X iko Y X therefore Y A-koi intentionally causing A X ikoi Y X in order that Y A-ku implying that A exists X iku Y if X then Y; X implies Y implying that A is there Plural: add `-n', e.g. `kiryakan' = "if and only if (there are) ships". With consonant stems: `Elendilke' = "because of Elendil", `elenke', `earke' present no problems; `lanatke' would metathesize to `latakte'. Two-consonant stems add a vowel as before other case endings: `Silmarilleken'. Duals decline like `lanat', e.g. `Aldakte' = "because of the Two Trees". Unreal conditions (e.g. "X would have happened if Y had happened"): insert a particle (say `ken' (from Greek)) into the X clause, and/or use endings `-kt-' instead of `-k-': `Elendilekti sii' = "if Elendil (was alive) now"; `Uuvea uumea uu ainer ken ikti Yavanna varne ristier Aldato neunasse tana-re-n' = "Much evil would not have happened if Yavanna had saved cuttings of the Two Trees in a second (place) before that day". Prefixing `ilya' = "all" would produce "-ever", like Latin `-cumque':- `X ilyaiki Y' = "X on every occasion that Y", `Ilyainna tana raa tulya, (tasse) makaro tarukuli ar karaaro uuve uumea.' = "Wherever that lion goes, it kills many cattle and does much harm". If the relative word urgently wants to be in one case in its role in the main clause and in another in its role in the relative clause, the `ilya' could be separated to carry the other case ending: `Ilyasse inna tana raa ...'. Thus interrogatives:- manka? if and only if what? manke? why? (= what caused it?) mankei? who/what intended that ...? manki? if what? manko? if ------, what results from it? mankoi? why? (= intending what?) manku? if ------, what? Some say this `i' neither takes nor needs case endings; but often in real languages multi-use clause-headers have prefixes (e.g. `mennai' KVNS) and suffixes (e.g. `iire' FS) and accompanying words to distinguish uses. Declining this `i' is tidier than inserting a word merely to carry the case ending. KVNS `X i Y-(a)nie' means the same as my `X ikoi Y', and may imply that `X i Y' (with Y's verb in an indicative tense) as "with result that" means the same as my `X iko Y'; but this still leaves the known Tolkien Quenya clause relation system very gappy. Context as an ambiguity resolver is unreliable: `i' undeclinable in various conjunction roles as well as "the" and as relative pronoun is to my mind overloaded and prone to unclearness, and it is easy e.g. to compose a sentence `queten i X' which = "I said that X" and also "I spoke with the result that X" and "I spoke by means of the fact that X"; in my usage these last two meanings would be expressed as `queten iko X' and `queten inen X'. With my `X ikoi Y' with its specific purpose suffix `-koi', compare English "A that B", with B as a purpose clause, being replaced in general usage by "A in order that B" (of learned origin) or "A so (that) B". QAE p372 has {kwen} > `queen' pl. `queni' = "person", `quen' unstressed = "someone", `ilquen' = "everybody", `aiquen' = "if anybody, whoever". So we have a sentient indefinite pronoun at last (but not an inanimate indefinite pronoun), and perhaps also a word for "if". [Place] Suffix `-men' = "place of" (cf. `il-men' = "Star-place" = "Space"). ETYM `tar' = "to there"; but `tar' also = "king" and "beyond", and lacks the expected allative ending. "Where" etc with noun antecedent: use case of `ya', e.g. `tellumar yassen' = "halls in which" (NAM), `ostollo yasse X' = "from the town where X". No antecedent: provide one by using `men' = "place", if the place wants to be a different case in the main clause from in the relative clause: `Wanyuvan Elendil tuulie, nu-nen-ea suuy-il-inen.' = I will go "where" Elendil came "from", with underwater breathing apparatus. (not `suuyilnen' > `suuyillen', with the instrumental `-nen' damaged.) `Wanyuvan Mardil wanya.' = I will go where Mardil goes. If the `men' and the `ya' are in the same case, shortening is tempting:- `Wanyuvan Mardil wanya'; but `ya-men' = "a/the place of which"; `Wanyuvan Mardil wanya.' is better. `men' suffix on infinitive:- (`-messe' < `-mensse') `Tulkanen alda omentielvamesse minya' = "I planted a tree where we met first". Or use place cases of `i', as above:- `Tulkanen alda omentanelve minyave' = "I planted a tree where we met first". [Time] Tolkien tried three approaches, in this order:- [0] ARCT has `X an Y' = "X until Y", with only the preposition "to(wards)" and nothing else between the clauses: see below. [1] A case `-re' (< `ree' = "day"): FS `iire' - `taare' = "which day - that day" (twice). Here `i' has a case ending. Some would not call `-re' a case ending; but the `A-o' and `A-nen' cases also arose from separate words becoming suffixes: see above. Also, `iire' is not merely the sum of its parts: "on which day" as two words would probably be `i resse', not `i ree'.) This is for when time is counted in days; other usable suffixes would be `-lu' < `luu' = "occasion" {lu}, `-yen' < `yeen' = "year":- `Tuula tenalye nyelle'. = "Come when you hear the bell". `Tuula yaanar lantar'. = "Come when (= on the day) the fruit falls". `tanaaren' (see above) = "on that day". `Teka tenalye nyelle'. = "Write until you hear the bell". `Nostaneo ni Sauron lante'. = "I was born when (= the year) Sauron fell". `ilunna' is a case-of-case. See [2] for time meanings of the place cases. A previous vowel lengthens in open syllables in these cases of monosyllables. For "when and if", try some combination such as `-luki'. [2] {men} > KVNS `mennai' [place-to-that] = "until". Like LOTR `tenna' [line-to] > {tegh/teng'} (LOTR) = `kenna' [earth-to] > {kem} (SD) = "until" (preposition) it treats time as a space dimension; but perhaps time words could be used here also, e.g. `ree' = "day", `yeen' = "year", `luu' = "occasion", `luume' = "time". (A place conjunction might be expressed as a place case of `ya'.) If we allow the other place cases, and allow the short place cases for "for part of the duration", we have:- {teng'} {re} {teng'} {re} tenna renna until tenen ren before tesse resse same time as tenes res within the time of/that tello rello forever after tenel rel after and thus from the other roots. Append `-i' when these words are used as conjunctions. The short cases of the stems ending in nasals (`men', `kem', `yen', `teng') may in use shorten to one syllable by analogy with the vowel stems (`re', `lu'). Beware confusion with `menel' = "after" and "heaven", and with `te-' as here and as pronoun stem = "them". (See QT43 re my `dedative' case; compare `ala' = "after" (prep.) (OM1) for its `l'.) (I was once criticized for using `ree' in matter set before the Sun was made; but `ree' may also have been used to refer to the light variation cycle of the Two Trees.) If I supply a prefix `mu-' (altered form of `mi' = "in") to make verb tenses durative (e.g. `muteke' = "he was writing", `teke' = "he wrote"), it could also be used in a case `-mu', e.g. `imu' = "during" (conjunction), `omientienyamu' = "during my meeting". [CASES OF CASES?] In some languages the possessive genitive is often replaced by a declinable adjectivizing suffix, e.g. Homeric Greek `-eios' instead of `-oio', Russian `-ov' instead of `-a'; and early Latin `cujus cuja cujum' = "whose", so the noun can be genitive and also in the case of the noun that governs it. Greek uses genitive instead of the usual case with "to" and "in" to make forms with ellipsis like `en Heeraas' = "in Hera's (temple)", `eis Priamoio' = "into Priam's (house)"; c.f. English "to John's". Are these possible in Quenya? If `omientie-lva-sse' = "at our meeting" is used much, then (since gen. `elva' = "(something of) ours") `elvasse' = "at that object/event of ours" is a likely abbreviation, being in effect allative of genitive of `elve'. Likewise `Vardo tellumassen' = "in Varda's halls", if used much, may get shortened in usage to `Vardo-ssen' = "in Varda's", being locative pl. of genitive sg. of `Varda'. Other cases than genitive may be treated this way: `Nuumenoore-llo (atan)-in' = "for the (men) [who came] from Nuumenor" > `Nuumenoore-llo-in' (dat.pl of abl.sg); `osto-sse (atan)-ion' = "of the (men) in the town" > `osto-ss-ion' (gen.pl. of loc.sg.); `ta na er selda-in anna-ron' > `ta na er selda-in-ion' = "that's one of the (presents) for the children" (gen.pl. of dat.pl.). [WHICH TO ACCEPT WHERE ETYM CONTRADICTS LOTR?] (This section has also appeared as a `letter from a reader' in VT#36.) The study of Quenya and Sindarin, and the changes that Tolkien made to them in his lifetime, causes queries and shows contradictions that should be resolved for people who want to write in these languages, even though deciding which forms to reject is irrelevant for those who merely want to study what he wrote. The first rule of thumb is to use the latest version; but, after many years of using The Lord of the Rings, particularly its tengwar and month-name lists, as an important reliable source for vocabulary, new publications such as the Etymologies suddenly contradict some points. Tolkien seems to have altered Etym. little after starting LotR, despite having to invent various new Elvish words and meanings while writing LotR; some LotR forms (which people have treated as authoritative for many years) ignore or reverse changes that he made in Etym. Some of these changes seem to have been made to eliminate homophones. It is true that natural languages have homophones; but it is also true that sometimes a natural language puts up with a pair of homophones for a while, then rejects one for the sake of clarity - e.g. English "sewer" (big drain) vs. obsolete (e.g. Shakespearean) "sewer" (servant who tells guests where to sit) - and it is likely that Tolkien sometimes did the same with his languages. An example is seen in the two bases {ulu} 'pour, flow' (LR:396) and {yul} 'smoulder' (LR:400), compared with *{yul} 'pour' seen in `yuldar' "draughts" (of drink) (? < passive participle *`yulna') and `yulma' "cup" (NAM); and note that, given the habitual loss of the distinctive `g', some Quenya and Noldorin derivatives of {ulug} "evil" (e.g. Q `ulundo', N `ulund' "monster") may appear to be derived from {ulu} "pour, flow". It may be that Tolkien changed {ul} to {yul} when it means "pour" when the long-tolerated homophony of `ul'- = "pour" and "evil" proved inappropriate in a poem by Galadriel; but in the process he created a homophony involving {yul}. In `henulka' "evil-eyed" (SD:68) he tried another way to separate the `ul'- homophones, by changing {ulug} "evil" to *{uluk}, whose distinctive `k' persists in Quenya and Sindarin. If this second solution is adopted, {ul} could stay as "pour", and {yul} as "smoulder", yielding *`ulma' "cup" and *`ulda' "draught", distinct from `yulma' "brand" < {yul} "smoulder". QAE p416 derives the two NAM words from {yulu} = "drink". Also, Etym. has {es} "indicate, name" and {ese}, {eset} "precede" (LR:356), with identical derivatives from both: Q `esse' "name" and "beginning", `este' "name" (verb) and "first". This homophony was long tolerated, but was eliminated in LotR with `yestare' *"first day" as if < *{yes} "first", which would yield *`yesse' "beginning" and *`yeste' "first". However, this form clashes with {yes} "desire" (LR:400); dislikers of homophones could ignore {yes} "desire", as {id} *"heart, desire, wish" (LR:361) and {mer} "wish, desire, want" (LR:373) provide ample means to express the meaning "desire". The base {ung} (LR:396) yields Q `ungwe' "gloom" and `ungo' "cloud, dark shadow"; Ilkorin `ungol' "darkness" and `ungor' "black, dark, gloomy"; and N `Ungoliant' (the Great Spider, a name taken from Quenya). But later apparent derivatives from {ung} suggest that the base means "spider": Q `ungwe' "spider's web" (III:401) and S `Cirith Ungol' "Pass of the Spider" (III:430). It may be that Tolkien had not needed generic Elvish nouns for "spider" and "spider's web" before the composition of LotR; before that, he had only needed proper names and titles for the Great Spider, and use of roots meaning "gloom" in such words (`Ungoliant', `Wirilome' [LT1:254 s.v. `Gwerlum']) is to be expected, given the role of spiders in his mythos. Rooted dislikers of homophones could perhaps let the base {ung} = "spider or its web" only and let {domo} "faint, dim" (LR:354) and {lum} (LR:370) provide words for "gloom". The base {ur} "be hot" (LR:396), which had stood since QL times (cf. `Uur', LT1:271), and all its Quenya and Noldorin derivatives, were for some reason crossed out in ETYM and replaced by (roughly written) {ur} "wide, large, great", whence Q `uura' "large"; but later, in LOTR Tolkien gives Q `uure' "heat" (III:401) and `Uurime' *"hot" = August (III:388). There is no reason not to keep {ur} "be hot": this replacement of it is an example of a temporary change which he later thought better of. Another change to ETYM that may be safely ignored is the change in the development of original initial *`sm'- in Old and Exilic Noldorin. In ETYM as first written, *`sm'- > ON `m'- > N `m'-; but Tolkien subsequently altered this scheme so that *`sm'- > ON `sm'- > N `(h)m'- (see LR:387 s.v. {smal}). If this change is followed through it would produce forms such as **"hmallorn" (< {smal}) instead of LotR's "mallorn". To this change is related the vacillation as to whether original initial `l'- yields Noldorin/Sindarin `l'- or `lh'-; here I would stick to `l'-, as Tolkien seems to in LotR. A minor contradiction without apparent easy solution is that between the base {yat} "join", whence Q `yanta' = N `iant' "yoke", and Q `yanwe' = N `ianw' "bridge" (LR:400); and Q `yanta' (III:401) = S `iant' = "bridge" (cf. `Iant Iaur' "Old Bridge"; S:336, 360). Presumably `yanta' < passive participle *`yatna' "(something) joined". The Silmarillion is a post-LotR rewriting of matter written in the 1930's and earlier: how much in the rewriting did Tolkien update the Elvish in its text? There are changes that result in contradictions, for example the differing etymologies of `Vala'. QL (LT1:272 s.v. `Valar') has {VALA} *"blessed, happy" (a usage likely modelled on Greek `makaros' "blessed, happy", which is often used of the Greek gods), but Etym. has {bal} "power" (LR:350). For a while Tolkien also used {bal} "power" in a bad sense as in `Balrog' and Balchoth', but later he re-etymologized such words; but in WOTJ he again derived `Balrog' from `vala'. These changes mean e.g. that the early (Narqelion) word `karne-valin-ar' *"red-happy", i.e. 'rejoicing in redness [of autumn vegetation]', is invalidated by ETYM. There have also been revisions to pronouns and noun cases; for example, the 3p plural pronoun suffix -`lto', (e.g. in `meldielto' "they are beloved" FS, as opposed to the later -`nte' (e.g. in `tiruvantes' "they will guard it" in COTE and (possibly) in `sakkante' "[they] rent" (SD:246). Tolkien may have made this change to release the -`lt'- pronoun suffix family for another use (perhaps as the 2p plural intimate?). Another subject worthy of comment is the history of words meaning "big". QL has Q `velike' "great" (LT1:254 s.v. `Haloisi Velike'), obviously taken from Russian `velikiy', and given a corresponding Gnomish `beleg' "mighty". ETYM gives N `beleg' "great" < ON `beleka' "mighty, huge, great" < PE `beleka' (LR:352 s.v. {bel} "strong"). Later, Tolkien may have rejected Q `velike' (perhaps for being too obviously Russian) and re-etymologized N `beleg' as < *{mbelek} (not in ETYM), cognate with Q `melk'- in `Melko(r)'. (The derivation in ETYM of Q `Melko' < PE *`Mailikoo' < {mil-ik} *"greed" is likely a re-etymologization, and is to be queried in its handling of the vowel between `m' and `l'.) SIL says that Q `Melkor' = S `Beleguuur' = "He who arises in Might" (S:340). QAE p369 has `meletya' = "mighty" (< *`melekya'?) In ETYM the Quenya word for "big" seems to be `alta', `alat'-, < {alat} "large, great in size" (LR:348), the real-world origin of which may be involved with LOTR's S `galad' "light" < *{galat} (not in ETYM), which would yield Q *`al(a)t'-; and with attempts to decide on a Quenya form and an etymology for the Sindarin name `Galadriel'. The only other words in this semantic group that I know of are: (a) Q `ura' "large" and N `uur' "wide", derivatives of the base {ur} "wide, large, great" that Tolkien wrote in ETYM as a replacement of the earlier base {ur} "be hot" (LR:396); but as discussed above, these seem to be have been invalidated by his apparent return to {ur} "be hot" in LOTR. (b) The family of words from {taa}, {tagh} = "high, lofty, noble", which sometimes shifts into "large", e.g. S. WOTJ `daer'. (These roots may << Chinese `ta' = "big", Central Asian Turkic `tagh' = "mountain(s)" (the latter probably got from atlas maps). Tolkien certainly studied languages well!) [PHONOLOGY OF DEVELOPMENT OF QUENYA FROM PROTO-ELDARIN] This follows ETYM and ignores the very different early part of the phonology development shown in BOLT. As shown by KVNS, at that time RT the Elves made PE by trying to learn Valarin from Orome at Koivieneeni and afterwards; Tolkien likely made ETYM PE / Valarin uneuphonic with roots such as {snas} and {skyap} and {kugh} to make it sound stern and commanding. (Did Valarin have a history of change MET before that? E.g. PE {magh} = "hand", {mag} = "use, handle". From the description in Ainulindale (SIL), Ea before Arda was made was likely big enough and lasted long enough for even Ainur's languages to evolve different ways in different areas.) (This is changed completely RT by WOTJ, where (p5) after the Elves woke 35 Valian Years passed before Orome found them; in that time they developed PE by themselves. As in that form PE (called there PQ and later MET CE) is not tied to Valarin, QAE's PQ roots are more euphonic and much like the old QL roots, e.g. {kwene}, {aba}, {dele}, {ho} instead of ETYM {kwen}, {ab(ar)}, {led}, {gho}, and include a new root {hek} = "begone!". QAE pp397-404 describe a distinct Valarin language, chosen to sound stern and commanding but with a different phonology from ETYM PE roots; a limited number of its words got into Quenya later MET.) [Vowels] Indo-European-type nom.sg.-only lengthening (`vrddhi') occurs in e.g. {nen} > Q `neen' (stem `nen-') water (but note `Koivieneeni'). The only stem syllable lengthens if it is short and open:- {el} > Q `eel' = "star" (poetic) {khop} > Q `hoopa' = "haven" {kwam} > Q `quaame' = "sickness" {kwar} = "clutching hand, fist" > Q `quaare' {top} > Q `toopa' = "roof" N & V, but `topa' = "cover" V PE stress on 1st of 2 stem syllables:- {gAlad} > PE `galad-' > Q `alda' = "tree" {tArak} > Q `tarka' = "horn (of animal)" {dOron} > PE `dOron-' > Q `norno' = "oak" {barAn} > Q `varne' = "brown" (stem `varni-') {barAd} = "lofty" > PE `barAdaa' > Q `Varda' but {nArak} = "tear V" > PE `naraaka' > Q `naraka' ("violent", etc) {ab(ar)} > PE `Abaaro, Abaaroo' > Q `Avar, Avaro,' pl. `Avari' {ter(es)} = "pierce" > PE `tereewaa' > Q `tereva' = "fine, acute" PE stress on 2nd of 2 stem syllables:- {erEd} > PE `eredee' > Q `erde' = "seed, germ" {turUm} > PE `turumaa' > Q `turma' = "shield" {barAd} > PE `barAdaa' > Q `Varda' but {darAk} > PE `draak' > Q `raaka' = "wolf" Final long vowels shorten:- {bal} > PE `bAlaa' > Q `Vala' = "Vala" {bal} > PE `balii' > Q `Vali' = "Valar" "formed direct from stem, cf. `Valinor'" {ad} > PE `adnoo' > Q `ando' = "gate" {bes} > `besuu' > Q `veru' = "married pair" This stopped before `gh' etc dropped between vowels before Book Quenya times, but resumed afterwards to shorten results of contractions. Nom/acc plural `-ii' > `-i' is still long in Book Quenya. `ao' > `oo':- PE *`kiryaaghoo' > BQ `kiryoo' > Q `kiryo' = "ship" (gen.sg.) Original final short vowels:- `ya' > `e':- {tal} {run} (ETYM) > PE `talrunya' > Q `tallune' = "sole of foot" `i' > `e':- {lug} > PE `lugni' = "blue" > Q `luune' {mor} > PE `mori' > Q `more' = "black" {nik-w} > PE `ninkwi' > Q `ninque' = "white" `o' > nil:- {ab(ar)} > PE `Abaaro'/`Abaaroo' > Q `Avar(o),' pl. `Avari' `u' > `o':- {Orok} > PE `Orku' > Q `orko' = "goblin" {tug} > PE `tuugu' > Q `tuo' = "vigour, physical strength" and Tolkien said in RGEO (I think) that most PE final short vowels dropped in Quenya. `-aye' > `-ie':- `liinaye' (with short `a')> `liinie', but `-aaye' > `-e.e':- `liinaaye' > `liine.e' (KVNM in VT#27 p36) `-ayV' > -`eV':- PE `*-aya' > Q. `-ea' e.g. `laurea' {day} > PE `daio' > Q `leo' = "shadow" (so PE unstressed short `a' drops before any `-yV'??) `-aayV' > -`aay'/`-aiy' (this may < PE `-ay' + `ya'):- {way} = "enfold" > PE `waayaa' > Q `Vaiya' Q `haaya' (spelt `haiya' in SD) {khaya} = "far, distant" `eyV' > `Ve':- Q `silqe-losse-en' (pl) < `*-eyaaiim' (SV `Earendel') `e.e' > `ie':- {tegh} > PE `teghee' > *`te.e' > Q `tie' = "road, line, direction" {weg} > PE `weghee' > *`we.e' > Q `vie' = "manhood" `ei' > `ai':- {ey} > Q `aira' = "eternal", `aire' = "eternity" (deleted entry) {wey} > PE `weiree' > Q `Vaire' = "weaver; a Vala" but:- {gey} > Q `iira' = "eternal", `iire' = "eternity" (deleted entry) {gey} > PE `geiaa' > Q `ia' = "ever" (deleted entry) Long vowel shortens in hiatus:- {tug} > PE `tuugu' > Q `tuo' = "muscle, etc" Any final vowel of polysyllable noun nom.sg. often drops, e.g. `noore' / `Valinor'; `hilde' / `tarkil' (with simplification of any resulting illegal final consonant cluster), but returns before case endings etc. [Consonants] 3 consonants:- {kwet} > PE `kwentro' = "narrator" > Q. `quentaro' {tak} > PE `tankla' = "pin, brooch" > Q. `tankil' `b' > `v':- {bal} > PE `bAlaa' > Q. `Vala' = "Vala" but stays in `mb' between vowels. `Cb' > `Cw':- {gOlob} > PE `gOlob-' > PE `golbaa' > Q `olwa' = "branch" `d' > `l' init:- {dab} > Q `lav-' = "yield, allow, grant" `d' > `r':- {syad} = "shear through, cleave" > Q `hyar-' = "cleave" {red} > Q `rer-' = "sow" (verb), past `rende' but stays in `ld', `rd', `nd' between vowels. `dt' > `st':- {wed} > PE `wedtaa' > Q `vesta' = "swear (promise)" (deleted) `dy' > `y' init:- {dyel} > Q `yel-' `g' > nil:- {gAlad} > Q `alda' = "tree" {Ulug} > PE `ulgundoo' = "monster" > Q `ulundo' but stays in `ng' between vowels. `g' > `~':- {yagh} > PE `yagwee' > Q `yaawe' = "ravine, cleft, gulf" {reg} = "edge, border, margin" > Q `reena' = "margin" `gd' > `ht':- {khag} > PE `khagda' = "pile, mound" > Q `hahta' `gh' > nil:- {wegh} > PE `weghee' > Q `vie' = "manhood" {magh} > PE `magh'/`maagh' > Q `maa' = "hand" {dogh}, {dooo} > PE `doghmee' > Q `loome', ON `dogme'/`dougme' = "night" {kugh} > Q `kuu' = "bow" `gh' > `h' init:- {ghan} > PE `ghan' > Q `hanu' = "male sentient" (but initial `gh' drops in ON & S, > `g-' in Ilkorin) `gn' > `ng'/`~n':- {tug} > PE *`tugnaa' > Q `tunga' = "taut, tight" c.f. Q `tuo' < PE `tuugu' = "muscle", etc -- Here the `n'-infix relationship is obscured in the Quenya result. {stag} = "press, compress" > PE `stangaa' > Q `sanga' = "crowd" {slig} > PE `slignee'/`slingee' > Q `liine' = "cobweb" {lug} PE `lugni' = "blue" > Q `luune' `k' > `t' final:- {philik} > Q `filit' (pl. `filiki') = "small bird" {nEl-ek} = "tooth" > Q `nelet' (pl. `nelki') = "tooth" `kh' > `h' init:- {kham} > Q `ham-' = "sit" `kh' > `h'/`k':- {khaya} > Q `avahaira' = "remote" {khil} > PE `taarakhil' > Q `tarkil' = "Nuumenoorean" `khs' > `ks':- {lokh} > Q `lokse' = "hair" `khy' > `hy' init:- {khyar} = "left hand" > Q `hyarmen' = "south" `kl' > `kil':- {tek} > PE `tekla' > Q `tekil' = "pen" `km' > `ngw':- {rak} = "stretch out, reach" > PE `rakmee' > Q `rangwe' = "fathom" {tek} > PE `tekmee' = "letter, symbol" > Q `tengwe' = "writing" `kt' > `ht':- {ek(te)} > PE `ekte-' > Q `ehte' = "spearman" `lf' > `lp':- but Q {il} {phir} > `ilfirin' not *`ilpirin' = "immortal" `lg' > `ly':- {phEleg} > Q `felya', Telerin `felga' = "cave" `ln' > `ld'/`ll':- {skal} > PE `skalnaa' = "veiled, hidden, shadowed" > Q `halda' {skel} > PE `skelnaa' > Q `helda' = "naked" {dul} > PE `ndulna' = "secret" > Q `nulla' or `nulda' {khal} > PE `khalnaa' = "noble, exalted" > Q `halla' = "tall" LOTR `lr' > `ll':- {tal} {run} (ETYM) > PE `talrunya' (short final `a') > Q `tallune' = "sole of foot" {kal} > PE `kalroo' > Q `kallo' = "noble man, hero" `ls' > `ll':- {khyel(es)} > PE `khyelesee' > Q `hyelle' = "glass" {teles} > PE `tElesaa' > Q `tella' = "hindmost, last" {theles} = "sister" > Q `seler' (pl. `selli') `m' > `n' final:- {kem} = "earth" > Q `keen,' `kemen' = "earth, soil" `mb' > `m' init:- {mbal} > Q `malle' = "street" `mr' > `mbr' > `mbar':- {tam} > PE `tamroo' > Q `tambaro' = "woodpecker" (= "knocker") `n' + `g' > `ng':- {spAnag} > PE `spangaa' > Q `fanga' = "beard" `nd' > `n' init:- {ndew} > PE `ndeuna' > Q `neuna' = "second" {nil,ndil} > Q `nilda' = "friendly" {ndol} > Q `noola' = "round head, knoll" `ndy' > `ndy'/`ny':- {Ened} > PE `Ened-' > Q `en(d)ya' = "middle" `ng' > `ng'' init:- {ngal(am)} (talk loud or incoherently) > Q `ng'alme' = "clamour" {ngan(ad)} > Q `ng'ande' = "harp" `ng'w' > `ngw'??:- PE `ling'wi' = "fish" > Q `lingwe' (or `ling'we'?) (how to tell `ngw' from `ng'w' medial in usual Quenya Roman spelling?) `ngm' > `~m':- {teng'} > Q `teema' = "row, line, series" `ngy' > `ndy':- {ngyoo,ngyon} > Q `indyo' = "grandchild, descendant" `nm' > `nw'?:- {lan} = "weave" > Q `lanwa' = "loom" `owa' > `oa':- {tow} > Q `toa' = "woollen" `ph' > `f':- {phor} > Q `forya' = "right-hand" `ps' stays:- {lap} > Q `lapse' = "babe" `rg' > `ry':- {tArag} > PE `targaa' = "tough, stiff" > Q `tarya' `s' > `z' > `r' in `VsV':- {thus} > PE `thausaa' > Q `saura' = "foul, evil-smelling"; but QAE has many Vanyarin forms with persisting `z'. QAE pp 388,413 says this `z' reverted to `s' when near an `r': `Khazaad' > Q, `Kasar' = "dwarf". {bes} > PE `besuu' > Q `veru' = "married pair" `s' > `r' final:- {sol} > PE `solos' > Q `solor' = "surf" {teles} > Q `Teler' = "Telerin Elf" and in the short locative case around 1931 (SV) but not later. `sk' > `h' init:- {skar} > PE `skarwee' > Q `harwe' = "wound" (noun) `sl' > `l' init:- {sliw} > PE `sliiwee' > Q `liive' = "sickness" `sn' > `n' init:- {snar} > Q `narda' = "knot" else > `rn':- {bes} > PE `besnoo' > Q `verno' = "husband" `sp' > `f' init:- {spin} > PE `spindee' = "tress" > Q `finde' `st' > `s' init:- {stak} = "split, insert" > PE `staknaa' > Q `sanka' = "cleft, split" else stays:- {bes} > PE `bestaa' > Q `vesta' = "matrimony" `sw' > `hw' init:- {swad} > PE `swanda' > Q `hwan' pl. `hwandi' = "sponge, fungus" `sy' > `hy' init:- {syal} > PE `syalma' > Q `hyalma' = "seashell" `sy' > `ry' medial:- {sus} = "hiss" > Q `surya' = "spirant consonant" `th' > `s':- {khis,khith} > PE `khiithi' > Q `hiise' = "mist, fog" `th' > `s', not > `r':- {nowo} > `nauthe' > Q `nause' = "imagination" (but is written `th' in tengwar) `thm' > `sw':- {khith} > PE `khithme' > Q `hiswe' = "fog" `thm' > `tum':- {koth} > Q `kotumo' = "enemy" (ETYM) (if via PE *`kothmoo') `thr' > `ss':- {neth} > PE `nethraa' > Q `nessa' = "young" (via `sr'?) `thy' > `ty':- {koth} > `kotya' = "hostile" (ETYM) `tm' > `nw':- {mbarat} > PE *`mbaratmee' > Q `maranwe' = "destiny" `thw' > `sw':- {khith} > PE `khithwa' > Q `hiswa' = "grey" `w' > `v' sometimes:- {weg} > PE `weghee' > Q `vie' = "manhood" `w' > nil final:- {raw} > PE `raaw' > Q `raa' (pl. `raavi') = "lion" {tow} > Q `too' = "wool" `wo' > `o':- {woo} > PE `woo' > Q `oo-' (stressed), `o-' (unstressed) = "together" `zd' > `st':- {mizd} > PE `mizdee' > Q `miste' = "fine rain" Metathesis:- `bm' > `mb':- {kab} = "hollow" > Q `kambe' = "hollow of hand" `bn' > `mb':- PE `lebnaa' > Q `lemba' = "left behind" {leb} `dn' > `nd':- PE `adnoo' > Q `ando' = "gate" {ad} `kn' > `nk':- {stak} = "split, insert" > PE `staknaa' > Q `sanka' = "cleft, split" `kn' > `nk':- {tak} = "fix, make fast" > Q `tanke' (past of `take' = "he fastens") `kr' > `rk':- {ak} > PE `akraa' > Q `arka' = "narrow" `tk' > `ht':- {et} {kel} > PE `etkelee' > Q `ehtele' = "spring, issue of water" `tn' > `nt':- {kat} = "shape" > Q `kanta' = "shaped" `tr' > `rt':- {nEter} > Q `nerte' = "nine" `ts' > `st':- {khoth} > PE `khotsee' = "assembly" > Q `hosta' = "large number" `pn' > `mp':- {top} > Q `tompe' (past of `topa' = "cover" V) `pn' > `mp':- {rep} = "bend, hook" > Q `rempa' = "bent, hooked" (passive participle) `sg' > `zg' > `ks':- {masag} = "knead" > PE `mazgaa' > Q `maksa' = "pliant, soft" and often in `n'-pasts and `n'-participles of verbs. Dissimilation:- PE `kukuuwaa' = "dove" > `kuu.ua' > Q & ON `ku,' `kua', Sind. `cugu,' listed at {kuu}. Natural phonological development of onomatopoeic words sometimes but not always is held back by re-imitating the original animal noise: thus Anglo-Saxon `cucu' > English "cuckoo" unaltered instead of its vowels becoming as in "furrow"; but English "bleat" followed other words with "ea" in as "ea" changed from like the sheep's noise to its modern pronunciation: and now "bleat" is being replaced by a new onomatopoeia "baa". PE `daala' or `laada' > Q `laara' = "flat", under {dal}: Tolkien was undecided here. Haplologies:- `wilwarindon' < `wilwarindendon' = "like a butterfly" SV `tuilind-' < `tuilelind-' = "spring (season) singer" = "swallow" (bird) NARQ. [SUFFIXES] This list gives some, not necessarily all, uses of each suffix, from ETYM. -at: Originally the dual?? {skyap} > PE `skyapat' > Q `hyapat' "shore" (as rivers have two banks. The Elves marching from Koivieneeni would have seen and named and crossed many rivers before seeing the sea.) {lan} "weave" > Q `lanat' "weft" (possibly dual, ref. warp & weft of cloth) But not dual in:- {na} "be" > `nat' "thing" -e: Female: see also `-me'. ana > `ante' "female giver" "-ness", often as part of a longer suffix. -ea: Adjective former, < PE `-eya' and `-aya' `laurea' "golden" NAM -ie: Verb infinitive; "-ness". Perhaps < `-e.e', as `tie' "road" < `teghe'. `mornie' "darkness" NAM LOTR-period nom/acc plural of `-ea' adjective former. -il: `X-il' "instrument for X-ing", < PE `-la' (short a) {tek} > PE `tekla' > Q `tekil' "pen" {tak} > PE `tankla' "pin, brooch" > Q. `tankil' -ima: Adjective-former: `X-ima' "characterized by X" `fiirima' "mortal" < {phir} "dead of natural causes" `nessima' "youthful" < `nesse' "youth" < {neth} -ina: Adjective-former. {smal} "gold metal" > PE `smalinaa' > Q `malina' "yellow" -indo: Agent {mel} "be friend of" > Q `melindo' "lover" (masc.) *{kol} "carry" V > Q `Cormacolindor' "Ring-bearers" -inqua: Adjective-former. Q. `alkarinqua' (S. `aglareb' < *`-ikwaa' > Common Telerin `-ipaa') (QAE p412) -ka: Adjective-former. {spay} "despise, have contempt for" > `faika' "contemptible, mean" {poy} > PE `poikaa' > Q `poika' "clean, pure" {phau} "gape" > `fauka' "open mouthed, thirsty" perhaps the `-ika' and `-hta' (< *`-tka'?) BD & MPD cases; but not in `raika' "crooked, bent, wrong" < {rAyak}. -le: Condition or result. Probably < PE `-li' and so > `-li-' before other suffixes. {thuu} "blow" V > Q `suule' "breath" > Q `Suulimo' "Manwe" (*PE `suuli'?) -ma: Usually "instrument for". Sometimes appears as `-wa'. {dyel} > PE `dyel-' > Q `yelma' "loathing" {par} "compose, put together" > PE `parmaa' > Q `parma' "book" {snew} > Q `neuma' "snare" {skel} > PE `skelmaa' "skin, fell" > Q `helma' {syal} > PE `syalma' > Q `hyalma' "seashell", and in many other words (but not in Old Noldorin `dalma' "palm of hand" < PE `dal-magh' {dal} {magh}) -me: Condition or result. Sometimes appears as `-we'. {dogh, dooo} > PE `doghmee' > Q `loome' ON `dogme'/`dougme' "night" {rak} "stretch out, reach" > PE `rakmee' > Q `rangwe' "fathom" Here it is `-mo' feminized with `-e', "female agent":- {ser} > `serme' "female friend" -mo: Agent. (Probably PE `-kmo' and the like > `-ngo' rather than `-ngwo'.) {thuu} "blow" V > Q `suule' "breath" > Q `Suulimo' "Manwe" {ser} > `sermo' "friend" -na: Usually passive participle. Note PE stress. {bar} > PE `barnAA' > Q `varna' "safe" (passive participle) {khal} > PE `khalnAA' "noble, exalted" > Q `halla' LOTR {skel} > PE `skelnaa' > Q `helda' "naked" (passive participle) {kat} "shape" > Q `kanta' "shaped" {phay} > Q `faina' "send out rays of light" (verb) -ne: {phuy} > Q `fuine' "deep shadow" {lan} "weave" > Q `lanne' "tissue, cloth" (not < PE `lanme': c.f. Q `lanwe') Also verb `n'-past. -no: Agent sometimes. {wo,woo} {tor} > Q `otorno' "sworn brother, associate" {stab} "wooden post" > PE *`stabnoo' > Q `samno' "carpenter, wright, builder" {bes} "marry" > PE *`besnoo' > Q `verno' "husband" -o: Male: see also `-mo' {ban} > `vanima' "beautiful" > `Vanimo' "Fair One" {ana} > `anto' "male giver" `mundo' "bull (animal)" (in Tolkien's Letters) -ra: Adjective former. {ak} > PE `akraa' > Q `arka' "narrow" {neth} "young" > PE `nethraa' > Q `nessa' "young" (via *`nesra'?) -re: Condition. {phay} > Q `faire' "radiance" Female agent. {wey} > PE `weiree' > Q `Vaire' "weaver" and name of a Vala -ro: Agent. {kwet} > PE `kwentro' "narrator" > Q `quentaro' {tam} > PE `tamroo' > *`tambro' > Q `tambaro' "woodpecker" (lit. "knocker") -se: Usually forms inanimate concrete nouns. {phas} > `fasse' "tangled hair, shaggy lock" {ras} "stick up" (itr.) > Q `rasse' "horn" {kwes} > PE `kwessee' > Q `quesse' "feather" {es} > PE `es-' "name" > Q `esse' {galAs} > PE `galAs-' "joy, be glad" > Q `alasse' "joy, merriment" {lokh} > Q `lokse' "hair" {khoth} > PE `khotsee' "assembly", c.f. Q `hosta' "large number" -sse: {khop} > Q `hopasse' "harbourage", `hoopa' "haven" {er} "one" > `eresse' = "solitude" Locative case. Is this how the locative case arose? Or is `hopasse' a locative used as nominative of a new noun? -ta: Causative verb former. {dyel} > PE `dyel-' > Q `yelta' "loathe" V {tul} > PE `tultaa' "make come" > Q `tulta' "send for" -ve: Adverb former. Q `anda' "long" (adj) > Q `andave' "long" (adv) LOTR -wa: Distinguish from `-wa' < `-ma'. {dyel} > PE `dyel-' > Q `yelwa' "loathsome" {wil} > PE `wilwaa' "lower air" > Q `wilwa' > `vilwa' (>> `wilma') {khith} > PE `khithwa' > Q `hiswa' "grey" -we: Condition or result. Distinguish from `-we' < `-me'. {ngol} > Q `ng'olwe' "wisdom, secret lore" {skar} > PE `skarwee' > Q `harwe' "wound" (noun) {yag} > PE `yagwee' > Q `yaawe' "ravine, cleft, gulf" -ya: Verb former. {phew} > Q `feuya' "feel disgust at" {bar} > PE `baryAA' > Q `varya' "protect" V Adjective former, see under [NOUN CASES]. Miscellaneous:- {rom} > Q `romba' "trumpet" {smal} > PE `smaldaa' > Q `malda' "gold metal" (< PE *`smalnaa'?) {phin} "nimbleness, skill" > PE `phinde' "skill" {spin} > PE `spindee' "tress" > Q `finde' {pal} > Q `palla' "wide, extensive" {gyel} > PE `gyel-' > Q. `yello' "call, shout of triumph" {nyel} "ring, sing" > Q `nyelle' "bell", `nyello' "singer" {rip} "rush, fly, fling" > Q `rimpe' "rushing, flying" {bat} "tread" > Q `vanta' "walk, to walk" [MISCELLANEOUS] In words stressed on the last syllable but two, a final short vowel may be lengthened to as in PE metris causa or in exclamation, e.g. `namaariee' (RGEO p61). COTE uses `i' as "the", and twice as "who". But `yassen' (NAM) ("at which", pl.) points to `ya' for "who", "which". It seems that Quenya, like English, had a habit of using "that" as a relative pronoun as well as a demonstrative/article. `andave' (LOTR (Cormallen)) shows that `-ve' forms adverb from adjective. SD on p129 has a valuable big new length of connected Sindarin text. `man' means "who?" and also "which?" (NAM, FS). I am surprised that there is no distinction between animate and inanimate interrogative pronouns. Does `man' decline (e.g. `mano' = "whose?", `man-sse' > `masse' = "where?" (and `manwe' "in which manner"!, c.f. `andave' = "long" (adv) (LOTR))? There are in English and Latin two families of indefinite pronouns, "someone, somewhere, etc" and "anyone, anywhere, etc" (note that "some" and "any" are not always synonymous) but so far no known way in Quenya to say them (except the `-o' pronoun subject as its nom.: e.g. `tirneo so' = "someone saw him"). A short Quenya word for "sentient" would be useful to distinguish "who" from "what", "someone" from "something" when necessary. There are `nat' = "thing" < {naa}; `kelva' = "animal or sentient" SIL < {kel}; `olva' = "plant" SIL < {golob}; but what about "sentient"? {nowo} = "think", so we may have e.g. `noowa' "sentient", `Mannat tirnelye?' = "What (not people) did you see?", `Manno tirnelye?' = "Who (not objects or animals) did you see?" (stem `mannow-'). FS `yallume' = "at last" < `ya-n luume' = "to which time" = "which time ended it" (with a perfect tense), and `ullume' = "forever" < `uu-n luume' = "to no time" = "no time ended it". Note `uu' = "not" with a case ending. The stress pattern seems to have changed twice: ETYM PE disyllabic roots show a random stress, but RGEO says that Quenya stress was initial before it moved to its classical position, "and there is still some tendency to stress the first syllable". [ANALYSES OF SOME TEXTS REFERRED TO ABOVE] (My hyphens. `= 'represents original hyphen) [FS (Fiiriel's Song)] (with Tolkien's changed lines 5 & 6) (my analysis and some punctuation changes, using LR's translation): (1) `Ilu' [universe] `Iluuvatar' [Il.] `en' [there] `kaare' [made] `elda-in' [elves for] `a' [and] `fiirimo-in' [mortal-s for] (2) `ar' [and] `anta-roo-ta' [gave he it] `ma-nnar' [hand-s to] `Val-ion' [Valar of]: `nuume-ss-ie-r' [west in are pl.]. (3) `To-i' [they are] `aina' [holy], `maana' [blessed], `meld-ie-lto' [beloved are they] = `enga' [except] `morion' [black one (= Morgoth)]: (4) `talant-ie' [fallen is]. `Marde-llo' [Earth from] `Melko' [Morgoth] `lende' [went]: `maar-ie' [good is]. (5) `en' [there] `kaar-ie-lto' [make past they] `elda-in' [elves for] `Isil' [moon], `hild-in' [men for] `Uur-anor' [Fire sun]. (6) `To-i' [they are] `iirima-r' [beautiful pl.]. `Ilya-in' [All (pl) to] `anta-lto' [give they] `anna-r' [gifts] `lesta-nen' [measure instr.] (7) `Iluuvataar-en' [Il. of]. `Ilu' [universe] `vanya' [fair], `fanya' [sky], `ear-i' [seas], (8) `i=mar' [the earth], `ar' [and] `ilqa' [all] `ii-men' [that place (= in them)]. `Iirima' [lovely] `ye' [is] `Nuumenor' [N.]. (9) `Nan' [but] `uu-ye' [not is] `seere' [rest] `indo=nin-ya' [heart in? my] `ullume' [forever]; (10) `te-n' [for] `sii [now]' `ye' [is] `tyel-ma' [end-ing], `yee-va' [be will] `tyel' [end] `ar' [and] `i=narqelion' [the fading]. (11) `Iire' [when] `ilqa' [all] `yee-va' [be will] `noot-ina [count-ed],' `hosta-in-iee-va' [number-ed will be] `yallume' [at last], (12) `ananta [but' yet] `uu-va' [not will (be)] `taare' [then] `faarea' [enough], `u-faarea' [not enough]! (13) `man' [what] `taare' [then] `antaa-va' [give will] `ni-n' [me to] `Iluuvatar' [Il.], `Iluuvatar' [Il.], (14) `en-yaa-re' [in that day] `tar' [beyond] `i' [the] `tyel' [end], `iire' [when] `Anar-i-nya' [sun (svarabhakti) my] `qel-uva' [fade will]? Notes:- (a) `te-n' = "for them" (dat.sg.), used as conjunction "for". (b) `tar' = "beyond": misetymologized, should be `sar', as PE & Sind `thar'. (c) Tolkien seems to have changed the Quenya for "all" from `ilqa' to `ilya' while writing FS; he changed `ilqain' to `ilyain' in line 6. (d) "be" is `ye', `-ie' , not `na', `ea'; "they" (nom) is `to', `-lto', not `-nte'. Verbs often use the preclassical conjugation described above. FS was written about 1940; between then and when NAM was finished and COTE was written, he seems to have changed some features of Quenya grammar. He likely rejected `-ie' as "is" because `-ie' has too many other uses, risking ambiguities. It will likely never be possible to bring into one scheme all features of all Tolkien's discoverable Quenya text; we will have to choose. (e) `u' used as verb = "not to be" [ONAM (original form of Namaarie) (TOI, pp284-5)] (My translation and analysis as far as I can. Comments and help welcomed.) (1) `Ai' [alas]! `laurie' [golden pl.] `lanta-r' [fall pl.] `lassi' [leaves] `suuri-nen' [wind instr.] (2) `inya-lemiine' [year numberless?] `raamar' [wings] `alda-ron' [of trees] (3) `inya-li' [year many] `et-tul-ie-lle' [out came you(pl)] `tur-me' [control to-get?] `maar-ion' [of lands] (4) `andunie-sse' [in west] `la' [not] `miiruvoor-ion' [of meads] (5) `Varda' [V.] `teluume-n' [to/for dome] `falmar' [waves] `kiir-ie-n' [for dividing] (6) `laurea-lass-ion' [of golden leaves] `oomar' [voices] `mailinon' [(< `ma' = "hand"?)] (7) `Elen-taari' [star queen] `Varda-n' [to/for Varda] `Oiolosse-a-n' [to/for dweller on O.] (8) `Tintalle-n' [to/for Kindler] `maali' [yellow?] `or-teluume-nen' [high dome instr.] (9) `ark-anda-vaa=le' [narrow gate via? of-you(pl)] `qanta-malle' [full road] `tuul-ie-r' [came pl.] (10) `e' [from?] `falma-li-llon' [from many waves] `morne' [dark?] `sinda-norie' [grey country] (11) `no' [?] `miir-inoite' [jewel ?] `kalla-silya' [bright? silvershine] `Valimar' [V.]. Notes:- (a) `inya' = "female" in ETYM {ini}, but "year" in CAED. (CH) (b) for `turme', see `-me' under [Noun Cases] above. (c) Any help is welcome to translate `mailinon' and `no miirinoite'. (d) The text is disjointed and unpunctuated (and judging by the meter line 2 lost a word). It could be merely bits that came to Tolkien's mind as he thought, and he intended to interpolate other matter later to complete the sense; but in the end he scrapped it all except line 1. It may mean this: "Alas! The golden leaves fall in the wind. Years numberless [as] the wings [= leaves] of trees, many years. You came out to [get] control of lands in the West, not to [partake] of mead, [and] Varda of the dome [came out] to divide the waves. The voices of the golden leaves `mailinon'. To the Star-queen Varda, to she of Oiolosse, to the Star-kindler [I call]. Golden [is the] high dome. A full road[ful of you Elves came] through your narrow gate [= the harbour entrance gate at Alqualonde?]. From many waves dark, a grey country. Valimar `no miirinoite' [= shines with jewels??] bright silver-shine.". [NCPT (Notion Club Text)] (SD's translation mostly) (omitting Aduunaic parts) `O' [and] `Sauron' [S.] `tuule' [came] `nu-kum-na' [humble-d] ... `lantane-r' [fell pl.] `tur-kildi' [Nuumenooreans] `nu' [under] `huine-nna' [shadow to] ... `Tar-kal-ion' [Ar Pharazoon] `ohta' [war] `kaare' [made] `Vala-nnar' [Valar to] ... `nuume-heruvi' [west lords] `Arda' [Earth] `sakkante' [rent] `leneeme' [with leave] `Iluuvataar-en' [Il. of] ... `eari' [seas] `ull-ier' [should pour, intrans. pl.] `i-kilya-nna' [the chasm to] ... `Nuumenoore' ["] `ataltane' [fell down] ... `haiya' [far] `va-haiya' [away far] `Andoore' [Giftland]. `Malle' [road] `teena' [straight] `lende' [went] `nuumenna' [west to], `ilya' [all] `sii' [now] `maller' [roads] `raikar' [bent pl.] ... `tur-kildi' [Nuumenooreans] `roomenna' [east to] ... `nuuru-huine' [death shadow] `lumna' [heavy] = `va-haiya' [away far] `siin' [now? now is?] `Atalante' [(the) Downfallen]. Notes:- (1) `ull-ier' = "should pour" is the subordinating optative middle of `ulya' = "pour": `eari ullier i kilyanna' is a subordinate clause, but its conjunction is missing, since the text was written as fragments. (2) Note plural `maller' not `malli'. This type of plural of nouns in `-e' only occurs elsewhere (as far as I know) in the very early poem `Narqelion'. [OM1] (`Oilima Markirya' = "The Last Ark") `Kildo' [one saw] `kirya' [ship] `ninqe' [white] / `pinilya' [small] `wilwarindon' [butterfly-like] / `veasse' [sea on] `luunelinqe' [blue stream] / `talainen' [wings with] `tinwelindon' [stars like]. `Vean' [sea] `falastaneero' [was loud with surf] / `lootefalmariinen' [flower-waves-with (= waves `flowering' with surf)], / `kirya' [ship] `kallieere' [shone] / `kulukalmaliinen' [gold-lights-with]. `Suuro' [wind] `laustaneero' [made wind noise] / `taurelasselindon' [forest leaves like]; / `ondolin' [rocks] `ninqaneeron' [were white] / `Silmeraano' [silver moon in] `tindon' [glint (case & number unclear)]. `Kaivo' [(like a) corpse] `i' [which] `sapsanta' [grave into] / `Raana' [moon] `nuumetar' [sank in west], / `manduloomi' [hell shadows] `anta' [raised (lit. gives?)] / `moori' [black] `Ambalar' [East (subj)]; / `teluumen' [heaven vault] `tollanta' [hills on] / `naiko' [(not in the English: c.f. ETYM {nayak}: `naike' = "sharp pain")] `lunganar' [sagged]. `Kaire' [lay] `laiqa'ondoisen' [green rocks on] / `kirya' [ship]; `karnevaite' [red-sky-(some suffix)] / `uuri' [sun] `kilde' [wet-eyed?] `hiisen' [mist of] / `niie' [wept] `nienaite' tears], / `ailissen' [beaches on] `oilimaisen' [last] / `ala' [after] `fuin' [dark, night] `oilimaite' [last], / `alkarissen' [light-rays in] `oilimain' [last]; / `ala' [after] `fuin' [night] `oilimaite' [last] / `ailinisse' [shore on] `alkarain' [shining]. [OM2] (`Oilima Markirya' = "The Last Ark") `Man' [who] `kiluva' [see will] `kirya' [ship] `ninqe' [white] / `oilima' [last] `ailinello' [shore from] `luute' [sail (verb)], / `niive' [pale] `qimari' [phantoms] `ringa' [cold] `ambar' [bosom (in)] / `ve' [like] `maiwin' [gulls] `qaine' [wailing]? `Man' [who] `tiruva' [heed will] `kirya' [ship] `ninqe' [white] / `valkane' [vague] `wilwarindon' [butterfly like] / `luunelinqe' [blue flowing/wet] `vear' [sea in] / `tinwelindon' [stars like] `talaliinen' [wings with], / `vea' [sea] `pustane' [surging], / `raamali' [wings] `tiine' [shining] / `kalma' [light] `histane' [fading]? `Man' [who] `tenuva' [hear will] `suuru' [wind] `laustane' [making wind noise] / `taurelasselindon' [forest leaves like], / `ondoli' [rocks] `losse' [white] `karkane' [snarling] / `silda-raanar' [gleam moon in], / `minga-raanar' [wane moon in], / `lanta-raanar' [fall moon in], / `ve' [like] `kaivo-kalma' [corpse candle (= will o' the wisp?)]; / `huuro' [storm] `ulmula' [muttering], / `mandu' [abyss, Hell] `tuuma' [moving]? `Man' [who] `kiluva' [see will] `loomi' [clouds] `sangane' [gather], / `teluume' [heavens (sg)] `lungane' [bending] / `tollalinta' [hills on] `ruste' [crumbling], / `vea' [sea] `qalume' [heaving], / `mandu' [abyss] `yaame' [yawning], / `aira' [old [1]] `moore' [darkness] `ala' [beyond] `tinwi' [stars] / `lante' [falling] `no' [on] `lanta-mindon' [fallen towers]? `Man' [who] `tiruva' [heed will] `rusta' [broken] `kirya' [ship] / `laiqa' [green] `ondolissen' [rocks on] / `nu' [under] `karne' [red] `vaiya' [sky], / `uuri' [sun] `nienaite' [bleared] `hiise' [(with) mist] / `piike' [fading)] `assari' [bones (on)] `silde' [gleaming] / `ooresse' [morning in] `oilima' [last]? `Hui' [evening] `oilima' [last] `man' [who] `kiluva' [see will?], / `hui' [evening] `oilimaite' [last]? [1] = "primaeval"? (c.f. `aire' = "holy" NAM; `ea' = "be by nature" < *`aya' :: {ayan} = "holy" ETYM) [Nieninque] `Norolinde' [tripping lightly] `pirokendea' [whirling lightly] / `elle' [to there] `tande' [came] `Nielikkilis' [little Nielle] / `tanya' [that] `wende' [maiden] `nieninqea' [snowdrop like] / `yar' [to who] `i' [the] `vilya' [air] `anta' [gives] `miqilis' [kisses]. / `I' [the] `oromandin' [wood spirits] `eller' [to there] `tande' [came] / `ar' [and] `wingildin' [foam fays] `wilwarinde.en' [butterflies like], `losselie' [white people] `telerinwa' [of the shores (of Elfland)], / `taalin' [with feet] `paptalasselinde.en' [falling leaves like]. [Earendel] `San' [then] `ninqeruvisse' [white horse on] `luutier' [sailed] / `kiryasse' [in ship] `Earendil' [Earendel] `or' [on] `vea' [sea], / `ar' [and] `laiqali' [green] `linqi' [wet] `falmari' [waves (obj)] / `langon' [throat (subj)] `veakiryo' [sea ship of] `kiirier' [clove]; / `wingildin' [foam maidens] `o' [with] `silqelosse.en' [silver blossom like (hair)] / `alkantameeren' [made it shine] `uurio' [sun's] / `kalmainen' [lights in (instr)]; `i' [the] `lunte' [boat] `linganer' [hummed like harp], `tyulmin' [masts] `talaliinen' [sails with] `aiqalin' [tall] / `kautaaron' [bent], `i' [the] `suuru' [wind] `laustaner' [made wind noise]. [OM-NEW (The Last Ark, post-LOTR version)] This was written in 1963 to 1973, well after LOTR (MCOE p221). Tolkien provided a linguistic commentary of some words. The question marks are in the original. I quote here the latest of three slightly different versions. `Men' (thus the MS; error by Tolkien for `Man'?) [who] `ken-uva' [see will] `faane' [white] `kirya' [ship] / `meetima' [final] `hresta-llo' [beach from] `kiira' [leave], / `i' [the] `fairi' [visible ghosts] `neeke' [vaguely seen] / `ringa' [cold] `suuma-rya-sse' [hollow-cavity her in] / `ve' [like] `maiwi' [seagulls] `yaimie' [wailing pl.]? `Man' [who] `tir-uva' [heed will] `faana' [white] `kirya' [ship], / `wilwarin' [butterfly] `wilwa' [fluttering about], / `ear=kelume-ssen' [sea-flood tides-in] / `raama-inen' [wings with] `elvie' [starlike pl.], / `ear' [sea] `falasta-la' [foaming], / `winga' [foam, spray] `hlaapu-la' [flying in wind] / `raama-r' [wings] `sisiila-la' [freq.-shine-ing], / `kaale' [light (noun)] `fifiiru-la' [freq.-fading away]? `Man' [who] `hlar-uva' [hear will] `raavea' [roaring] `suure' [wind] / `ve' [like] `tauri' [forests] `lil-lassie' [many-leaved pl.], / `ninqui' [white] `karkar' [rocks] `yarra' [snarl, growl] / `isilme' [moonlight] `ilka-la-sse' [gleam-ing-in], / `isilme' [moonlight] `piika-la-sse' [dwindl-ing-in], / `isilme' [moonlight] `lanta-la-sse' [fall-ing-in], / `ve' [like] `loiko-liikuma' [corpse candle]; / `raumo' [(noise of) storm] `nurrua' [grumble], / `undume' [abyss] `ruuma' [move heavy object]? `Man' [who] `ken-uva' [see will] `lumbor' [clouds] `a-hosta' [assemble] / `Menel' [heaven] `a-kuuna' [curved, bent] / `ruxa-l'' [crumbling] `ambo-nnar' [hills on], / `ear' [sea] `amorta-la' [heaving], / `undume' [abyss] `haaka-la' [yawning], / `enwina' [old] `luume' [darkness] / `eleni-ll-or' [star many genitive] `pella' [beyond] / `talta=talta-la' [slip-slipp-ing] / `atalantea' [downfallen] `mindo-nnar' [towers on]? `Man' [who] `tir-uva' [heed will] `raak-ina' [broken] `kirya' [ship] / `ondo-li-sse' [rock many on] `morne' [black] / `nu' [under] `fanyare' [sky (where clouds are)] `ruukina' [disordered], / `anar' [sun] `puurea' [discoloured] `tihta' [blink, peer] / `axo-r' [bone-s] `ilka-la-nnar' [gleam-ing-on] / `meetim'' [final] `aure-sse' [dawn in]? `Man' [who] `ken-uva' [see will] `meetim'' [final] `anduune' [evening]? Notes:- (1) Present participle in `-la'. (2) The `a-' in `ahosta' and `akuuna' (altered from `na=') shows that the bare stem used as infinitive after "hear", "see", etc, is intransitive here. (3) Note two frequentatives formed by repeating first syllable of verb. (4) "corpse candle" = "will o' the wisp". Q. `liiko' = "wax". [Koivieneeni Sentence] (condensed and modified from Chris Gilson and Patrick Wynne's analysis in VT#14) MS Quenya: `Eldar ando kakainen loralyar Koivienenissen mennai Orome tanna lende [i erenekkoitan(n)ie / na senekkoita].' MS English: "The Elves were long lying asleep at Koivieneeni until Orome came thither [that he might awake them / to awake them]." My `[-/-]' represents alternate endings in the MS, which is on the back of a bit of LOTR draft. `Elda-r' [Elves] `and-o' [for a long time (gen. as adverb)] `ka-kai-ne-n' [durative lie past pl] `loralya-r' [asleep (pl.) {los}] `Koiv-ie-nen-issen' [Wake (intrans.) -ing (not participle) water loc.pl] `men-na-i' [place towards that = until] `Orome' [O.] `ta-nna' [that to {ta}] `lende' [came {led}] `i' [that (clause-header)] `e-ren-ek-koit-anie' [he? them out wake(trans.) might (`ek' assimilated < `et'; `ren' < `sen' {s})] / `na' [to, towards] `sen-ek-koita' [them out wake (trans.)] Notes:- (1) The spelling `Koiv'-: the spelling `Kuivieenen' dates to when, after writing LOTR, he thoroughly revised his 1930's First Age stories to produce SIL. (2) The last `n' in `-an(n)ie' is an intruding part of irrelevant matter. (3) Infinitive suffix `-ie' in *`-koitie' omitted. Compare `an ... veela' in ARCT, and his notes to OM-NEW about bare stem sometimes used as infinitive. ["Arctic" Quenya sentence in `Father Christmas Letters' by Tolkien] `Maara' [good] `mesta' [journey? (`maara mesta' = "goodbye")] `an' [towards (= "until" here)] `ni' [I] `veela' [see] `tye' [you (acc.)] `ento' [next (adverb)], `ya' [which] `rato' [soon] `nea' [I hope that ... will be]. Notes:- (1) `an' preposition used as conjunction with no `i' clause-header. (2) `ya' with clause as antecedent (unless `veela' is a bare stem used as infinitive (see his notes with OM-NEW), or a participle `vee-la' = "see-ing" in apposition with `ni'. (3) `nea' hope-optative, see above. (4) I suspect that as ARCT is not set in Middle-Earth and was written by Tolkien for his son Christopher, who was young then, its grammar may be distorted from Quenya normal towards English so that its intended reader could more easily work out what word meant what. [TELERIN] Quenya varies RT as Tolkien developed it; but one variant of Quenya that existed within Middle Earth was Telerin, the language of the Teleri on Eressea and around Alqualonde. It preserves more PE consonants than standard Quenya, and to a Vanya or Noldo would likely be a somewhat rough-sounding sailors' and sea-fisherman's dialect. These Telerin words are known from ETYM. It likely also had many sailing and sea-fishing terms which standard Quenya would have little need for. `alpa' = "swan" {Alak}, `Bala' = "Vala" {bAlaa}, `Bana' = "Vana PN" {ban}, `Barada', `Baradis' = "Varda" {barAd}, `belda' = "strong" {bel}, `belka' = "excessive" {bel}, `belle' = "physical strength" {bel}, `branda' = "lofty, noble, fine" {barAd}, `bredele' = "beech-tree", `buuro' = "vassal" {bew}, `endo' = "grandchild, descendant" {ngyoo}, `Fallinel' = "Telerin Elf" (pl. `-elli') {phal} {nyel}, `felga' = "cave" {pheleg}, `ferne' = "beechen" {pher(en)}, `Findo' = "Thingol" {thin}, `golodo' = "Noldo" {ngolod}, `linda' = "singer, Telerin Elf", `Soloneldi' (pl.) = "Teleri" {sol} {nyel}, `spalasta-' = "foam, froth V" {spal(as)}, `spanga' = "beard" {spAnag}, `spania' = "cloud" {span}, `stalga' "steady, firm A" {stAlag}, `Spanturo' = "Mandos or Loorien (Vala) ('Cloud-lord')" {span}, `telpe' = "silver" (noun) {kyelep}, `trumbe' = "shield" {turUm}, `tuua' = "serve" V {bew}, `ulga' = "hideous, horrible" {Ulug}, `ulgundo' "monster, deformed and hideous creature" {Ulug}, `vilverin' = "butterfly" {wil} QAE contains many more Telerin words (more than can be listed here) and a fair amount of Telerin grammar, which largely follows Quenya. Some of this grammar is listed above along with the corresponding Quenya matter. [VANYARIN, NOLDORIN] In WOTJ Vanyarin is Vanyar Quenya, and Noldorin most times is clearly Noldor Quenya. That differs from 1930's usage, where Gnomish >> Golgodrin >> Noldorin is the language later called Sindarin. Some features of Vanyarin are: `z' between vowels (at least sometimes) stays, not > `r'; it has more words from Valarin; its word for `Quenya' is `Quendya'. [REFERENCE CODES] AA A.Appleyard ARCT The "Arctic" (Quenya) sentence in `Father Christmas Letters' by JRR Tolkien BD Bodleian Declensions, analysed by Patrick Wynne & CG & CH, VT #28 p8-34 BOLT `Book of Lost Tales', HOME #1 & #2, written before & around 1920 CAED Caedmon recording of Tolkien reciting NAM (a bit different from in LOTR) CE Common Eldarin CG Christopher Gilson CGM Christopher Gilson, in VT#36 re MPD CH Communication from Carl Hostetter (carl@class.gsfc.nasa.gov, editor of VT) CHJ Communication from Carl Hostetter to me on 9 June 1993 COTE Cirion's oath to Eorl, in UT (pp 305 and 317) EAR `Earendel' (poem in SV-1931) ETYM `Etymologies' (chapter in LR) FS `Firiel's Song' (LR p72) GAU George Allen & Unwin Ltd (publishers) GL `Gnomish Lexicon', as listed in BOLT. HOME `The History of Middle Earth' (series), ed. C Tolkien (publ. various) ITE `An Introduction to Elvish', by Jim Allan (Bran's Head Books 1983 1987 1978) KVNM Koivieneeni Manuscript (described in VT #27 pp 7-42) KVNS Koivieneeni Sentence (analyzed in VT #14 pp 5-7 & 12-20 by Gilson & Wynne) LETT `Letters of JRR Tolkien' LMBR Lisa McBriety (Lisa_McBriety%bitnet.nihdrg@CU.NIH.gov) in TL #6.21 LOB `The Lays of Beleriand' = HOME #3, by JRR Tolkien (GAU 1985) LOTR `The Lord of the Rings' ,by JRR Tolkien LR `The Lost Road' = HOME #5, ed. C Tolkien (Unwin Hyman 1987) MCOE `The Monsters and the Critics and Other Essays', by JRR Tolkien (GAU 1983) MPB Marquette Pronoun Declension (a bit) (VT#28 bottom of p29 in the BD article) MPD Marquette Pronoun Declension (full) (VT#36 pp 7-29, by CG) MR `Morgoth's Ring' = HOME #10, ed. C Tolkien (Harper-Collins (London) 1993) MYTL Analysis of NARQ by Paul Nolan Hyde in Mythlore #56 (Winter 1988) pp47-52 N Noldorin (which >> Sindarin between writing ETYM and starting LOTR) NAM Quenya poem `Namaarie' in LOTR, and the analysis of it in RGEO NARQ Tolkien's poem `Narqelion', probably written Nov 1915 - Mar 1916 (MYT, PMEL) NCP `Notion Club Papers' (part of SD) NCPT The main Quenya & Aduunaic text in NCP (SD, frontispiece & pp 310-312) NIEN `Nieninque' (poem in SV-1931) NIHH `Narn i Hiiin Huurin', in UT OM1 `Pre-publication' version of poem `Oilima Markirya' in SV-1931 OM2 `Publication' version of poem `Oilima Markirya' in SV-1931 OM-1931 = OM1 + OM2 OM-NEW same poem in SV translated later by Tolkien into nearly LOTR-type Quenya ONAM Old version of NAM (TOI pp 284-285) PLET JRR Tolkien's letter to D.Plotz, publ. in `Beyond Bree' (March 1989 issue) PMEL Analysis of NARQ by Wynne & Gilson in `Parma Eldalamberon' #9 pp6-33 PQ Primitive Quendian QAE `Quendi and Eldar' (part 4 of WOTJ) (probably written c.1960) QL `Quenya Lexicon', as listed in BOLT RGEO `The Road Goes Forever On', JRR Tolkien & D Swann (GAU 1968) ROTS `The Return of the Shadow' = HOME #6, ed. C Tolkien (Unwin Hyman 1988) SD `Sauron Defeated' = HOME #9, ed. C Tolkien (Harper Collins 1992) SIL `The Silmarillion', ed. C Tolkien SOME `The Shaping of Middle-Earth' = HOME #4, ed. C.Tolkien (Allen & Unwin 1986) SV chapter `A Secret Vice' in MCOE (ISBN 0-04-809019-0) SV1931 poems in SV written in or before 1931, in very pre-LOTR Quenya TL Email group TolkLang (for information, send to tolklang-request@dcs.ed.ac.uk a message whose `Subject:' line is the word `help' (without quotes)) TOI `The Treason of Isengard' = HOME #7 (Unwin Hyman 1989) UT `Unfinished Tales', ed. C Tolkien (GAU 1980) VT periodical `Vinyar Tengwar' (publ. Elvish Linguistic Fellowship) WOTJ `The War of the Jewels' = HOME #11 (Harper Collins 1994) WOTR `The War of the Ring' = HOME #8