Who shall say whether the free adjective has created images bizarre and beautiful, or the adjective been freed by strange and beautiful pictures in the mind?
Whenever nouns are qualified by adjectives the link between them can be established by making the adjective agree in case, number, or gender. In various languages the treatment is different – there may be no agreement at all, agreement in all three instances or just in some combinations of them.
There is no grammatical gender in Tolkien’s Q(u)enya apart from some personalized forms in the early sources, so we are limited to case and number. The adjective can be in preceding or postpositive position and agree in case, number, both, or none; so that with a case-inflected plural noun there are 8 theoretical possibilities of how the adjective can behave. For a case-inflected singular noun or for a not inflected plural noun there are 4. It will not surprise a scholar of Tolkien’s works that there is at least one example for almost every possible pattern; but some patterns are more usual than others and some are only used in verse.
Thus the aim of this article is to examine Q(u)enya adjectives through their external development regarding plural and case agreement with the nouns they qualify.
1 Earliest Qenya sources
The poem Narqelion was one of the earliest, if not the earliest Qenya text written. It is untranslated, but a tentative translation and analysis with the help of the QL was attempted in VT40:8-32. So all references to adjectives and the nouns they qualify are bound to the right interpretation. I give here the proposed translation, see VT40 for more detail.
Other examples can be found on the world ship map in LT1:84-85.
ōmi karmar *’all deeds’ (PE15:32,33) < ōmi pl. ’all’ (cf. ompi ’[plur. adj. with pl.] all, every’ (QL:70)), karma ’shape, fashion – act, deed’ (QL:45) + pl. -r
postpositive adjective (plural noun), no plural agreement
lasser pínea *’small leaves’ (VT40:8,13-14) < lasser, pl. of lasse (e) ’leaf’ (QL:51), pīnea ’small’ (QL:73)
rotser simpetalla pinqe *’pipes whistling thin’ (VT40:8,24) < rotse ’pipe (tube)’ (QL:80), #simpeta ’pipe’ + present participle -lla (cf. SIPI ’whistle, pipe’ (QL:84)), pinqe (i) ’slender, thin’ (QL:73)
[Here, pinqe might be employed as an adverb qualifying the process of piping, but it might also be a poetical word order for *’thin pipes whistling’. If so, there is no agreement of pinqe with pl. rotser.]
kuluvai ya karnevalinar *’golden and orange-red [ones]’ < kuluva ’of gold’ (QL:49), ya(n) ’and’ (QL:104), {karnewalin} ’russet, orange-red’ (QL:48), cf. ’walin(a) ’brown’ (QL:103).
[Here, kuluva and #karnevalina are both pluralized and refer either to the following sinqi ’jewels’ or to earlier lassi ’leaves’.]
sangar […] oïkta rámavoite malinai *’thongs […] of birds, having yellow wings’ (VT40:8,16) < oi ’bird, hen’ (QL:69) + -kta, rāmavoite ’having wings’ (QL:78), malinai, pl. of malina ’yellow’ (QL:58)
sangar […] oïkta rámavoite karneambarai *’thongs of birds, having wings, red-breasted’ < karneamabar ’readbreast, robin adj. or noun’, -a adj. (QL:48)
[This line was first written exactly like the earlier one, oïkta rámavoite malinai. Both phrases seem to be a group of three adjectives with only the last one agreeing in number with sangar.]
natsi nostalen máre *’things/beings good by nature’ (PE15:32) < nat (tt) ’thing’ (QL:64) + pl. -i with -tt- > -ts-, nostalen māra ’good by nature’ (GL:10), here pl. máre
i hondor nērinwe *’the hearts of men’ (PE15:32,36) < hondo ’heart’ (QL:40) + pl. -r; #nērinwa *’mannish’ (cf. ner ’man, husband’, nerea ’manly, stout, brave’ (QL:65)), pl. *-ai > -e
Tolli Kimpelear *’Twilit Isles’? < pl. Tolli of Tol (ll) ’an island, any rise standing alone in the water, plain of grass etc.’ (QL:94); #kimpelea is of unclear origin, perhaps connected to PELE-(1) ’fence in’ (QL:73) because the isles are bounding and protecting Valinor?
I Tolli Kuruvar ’The Magic Isles’ < kuru ’magic, wizardry (of the good magic)’ (QL:49), kuruvar, pl. of #kuruva *’magical’ (QL:49))
I Nori Landar *’The Great Lands’ < cf. nōre ’native land, nation, family – country’ (QL:66) + landar, pl. of landa ’broad, wide’ (QL:51)
Neni Erùmear ’Outermost Waters’ < nen ’river, water’ (QL:65), erūmear, pl. of erūmea ’outer, outermost’ (QL:36)
[The accent in Erùmear is perhaps an ill-formed or misread macron.]
It appears that at this time it is usual in Qenya to postpose adjectives. They usually agree in number (there is no attestation with cases), but at least one example lacks agreement (lasser pínea), so that it seems optional – at least in verse. The more complicated formation of several adjectives with only the last one inflected probably plays a special role and another such example will occur later (part 8.2).
In Narqelion the usual plural marker is -i for adjectives ending in -a, although -r appears in karnevalinar. This might be explained by its use as a noun, *’orange-red ones’ or by poetical alteration in the phrase kuluvai ya karnevalinar (rather than repetitional *kuluvai ya karnevalinai). The adjectives that are part of geographical terms from LT all add -r.
2 Early Qenya Grammar
The Early Qenya Grammar was written by Tolkien around 1923 and consists out of a manuscript version and a typescript version – a partially expanded fair copy of the former. Some valuable explanations regarding the agreement of adjectives are given here.
At first, an important thing to consider is the ’loose compound’, a free and living function which is written with a ’raised stop’, but often as full compound in poetry (PE14:45). Loose compounds differ from usual compounds in the way that they are not lexicalized, the respective parts are more distinct and usually retain their own accentuation. Thus ohta ’war’ and karo ’deed’ form óhta·kàro ’warlike deed, hostile unfriendly act’; maptale ’seizure’ and lehesta ’riding, raid’ form máptale·lehésta ’a raid’ (the acute accent apparently indicates the primary stress and the grave the secondary). Although not said by Tolkien, ’a war deed’ and ’a seizure raid’ would probably be the loose compound equivalents in English.
But important in our context is that adjectives may precede in just the same way and are then uninflected, for example anda·ráma ’long arm(ed)’ – as it can be seen from the translation, the result may be used either as a noun or as a new adjective. The plural would presumably be *anda·rámar ’long-arms’.
Furthermore Tolkien distinguishes the four cases (nominative, accusative, genitive and dative) from the five adverbial suffixes: inessive -sse, ablative -llo, allative -nta/-tta, partitive -ı̆nen and manner -ndon. The adverbial suffixes (1) cannot all be formed from every noun and adjective and (2) they are never added except in verse to an adjective in agreement with its noun: where a qualified noun receives one of these endings the adjective usually precedes uninflected (unless for plural) and may be written as a loose compound (PE14:46).
The example given is tára·kasse ’on the high head’, where the adjective tára ’high’ precedes and is not inflected.
The typescript version gives further expansions to this idea, noting that the Qenya adj. may either precede or follow the qualified noun; in the former case it forms virtually a loose compound with the noun, but is written separate (PE14:76), so anda ráma ’(a) long wing’. The wording in the manuscript precedes uninflected (unless for plural) is changed to precedes uninflected (except rarely for plural) (PE14:78).
So the possibilities in the plural are tára·kasisse(n)or rarely tare kasissen (sic). According to the explanation, agreement would be poetical: *†tárasse kasse or *†tárassen kasissen.
Agreement would be usual for the four ’ordinary’ cases (as compared to the adverbial suffixes) and indeed we find: ni·hepsine […] i·mailinen losselin ’I myself bound […] the beautiful flowers’ (PE14:56) where both #mailina ’beautiful’ and #losse ’flower’ are put into the accusative plural.
3 Qenya poems of the 20s
The three Qenya poems Oilima Markirya, Nieninqe and Earendel are presented in PE16 and MC with all draft forms. This is where most examples concerning the morphology of adjectives is found. Conventions from PE16 will be used for the denotation of the versions. The late versions from the 50s or later are discussed further below.
3.1 Case-inflected
preceding adjective (singular noun), no case agreement
OM1a,b,c: lúne veasse *’in the blue sea’< lúne ’blue’ (QL:57, PE16:75), vea ’sea’ (PE16:138) + inessive -sse
OM1a,b,c: tanda kiryaiko *’with that ship’ < tanda *’that’ (cf. TA in QL:87; PE14:54), kirya’ ship’ (PE15:77) + comitative sg. -iko
OM1c: móre veasse *’in the dark sea’ < móre *’dark’ (cf. MORO in QL:62)
OM1: pinilya wilwarindon ’small like a butterfly’ < pinilya ’small’ (cf. PINI (QL:73)), ’wilwarin ’butterfly’ (QL:104) + similative -ndon
OM2a: valka wilwarindon ’vague as a butterfly’ < valka ’vague’ of unclear origin, perhaps the same or similar to walda ’excited, wild’, also walka (PE17:154)
OM2: oilima ailinello *’from the last shore’ < oilima ’last, ultimate’ (PE16:75), cf. aile, ailin ’beach’ (PE16:75) + ablative -llo
postpositive adjective (singular noun), case agreement
OM1a: kiryinqen tinweninqen *’a star-white boat’ < #kiryinqe *’little ship’ from kirya ’ship’ (PE15:77), tinwe ’star’ (QL:92), ninqe ’white’ (QL:66); both with nominative -n
[This is a tentative interpretation, another one would be kir- ’to sail’ + inf. -inqe (PE14:28), so *’to sail white-star’, see PE16:57]
OM1d,e: ailisse oilimasse, ailinisse oilimasse *’on the last shore’ < aile, ailin ’beach’ (PE16:75)
OM1d: alkarisse oilimasse *’on the last morning’< alkar ’shining, light of day’ (PE16:75) + inessive -isse or -asse, oilima ’last, ultimate’ (PE16:75)
OM1e: alkarisse panyarasse/wanyarasse *’in the light of the setting sun’ or *’in the fading daylight’ < #panyara *’setting’ (cf. PANA ’arrange’ (QL:72)) or #wanyara *’disappearing, fading’ (cf. VAHA (QL:99), AVA ’go away, depart, leave’ (QL:33))
OM1g addition: tauren lintyulussean *’forest having many poplars’ < taure ’forest’ (PE16:75), lintyulussea ’having many poplars’ (QL:53) + nominative -n
OM1g addition: tauren linqarassean *’forest, where many watch’ < linqarassea perhaps related to QARA- ’care for, guard, watch (over)’ (QL:76) + nominative -n
postpositive adjective (plural noun), case agreement, plural agreement
OM1d: ondon móren >> ondoin mórin >> ondoin morin ’the dark rocks’ < ondo ’stone’ (PE15:77, PE16:138) + archaic/poetic nom. pl. -in (PE16:115); more *’dark’ (cf. MORO in QL:62) > nom. pl. morin
OM1b,c: ondoisen andalissen *’on the long rocks’ < ondo ’stone’ (PE15:77, PE16:138) + archaic/poetic inessive pl. -isen (PE16:115), anda ’long’ (PE15:74) + inessive pl. -li-ssen
OM1,OM1f: ailissen oilimaisen *’on the last shores’ — see above
OM1g addition (PE16:80): tavárin lilyen *’many wood-fays’ < tavāri ’fay of the woods’ (PE14:10), #lilya perhaps *’many’ (cf LĪ (QL:53)) + nom. pl. -in (PE16:115)
E1a: wingildin silqelosseën ’foam maidens with flower white hair’ < wingild- ’foam maiden’ (PE16:100) (cf. ’wingild- ’nymph’ (QL:104)) + nom. pl. -in + #silqelossea *’white-haired’ (cf. silqe (e) ’a tress of hair’ (QL:83), losselie ’white people’ in Nieninqe), archaic/poetic nom. pl. silqelosseën with *-eain > -eën
E1b: vingildin silkvelosseën
N1b variant: wingildin wilwarindeën ’the foam-fairies like the butterflies’ < wilwarindea ’like a wilwarin or butterfly’ (cf. ’wilwarin ’butterfly’ (QL:104)), nom. pl. *-eain > eën
postpositive adjective (plural noun), case agreement, no plural agreement
-not attested-
postpositive adjective (plural noun), no case agreement, plural agreement
OM1d: {ondoisen laiqe} ’on the green rocks’ < ondo ’stone’ (PE15:77, PE16:138) + archaic/poetic inessive pl. -isen (PE16:115), laiqa ’green’ (QL:52), pl. laiqe
OM2: assari silde ’on bones gleaming’ perhaps < as- (ass-) ’bone’ (QL:33) + dative pl. -ri, silda ’gleaming’ (cf. SILI (QL:83), silda-ránar in OM2), pl. silde
postpositive adjective (plural noun), no case agreement, no plural agreement
-not attested-
3.2 Not case-inflected
Now follow some examples without an explicit case suffix. As the underlying Qenya grammar changes, OM1g still shows inflected nominative, but no explicit accusative. In OM2 and OM2a the nominative flexion also disappears. For these examples we cannot say anything about case agreement, of course:
preceding adjective, plural agreement
OM2: níve qímari ’pale phantoms’ < *níva ’pale’ of uncertain origin, pl. níve + #qímar ’phantom’, pl. qímari perhaps from *QIM- related to QIV-, QIPI- *’awake’ (QL:29)
E1a: laiqali linqi falmari ’green wet waves’ < laiqa ’green’ (QL:52) + pl. -li, #linqe ’wet’, pl. linqi (cf. linqea ’watery’, liqin(a) ’wet’ (QL:54)), falmar ’a wave as it breaks’ (QL:37) + pl. -i
E1b: laikvali linkvi falmari
OM1g: moori raukvi *’dark storms’ < móre *’dark’ (cf. MORO in QL:62), pl. móri; rauko ’ravening, rushing’ (PE16:77), pl. raukvi
preceding adjective, no plural agreement
-not attested-
postpositive adjective, plural agreement
N1a,b: wingildi wilwarindear as above, with pl. -r
N1a,b: táli paptalasselindear ’with feet whose music is that of falling leaves’ < tāl ’foot’ (PE14:76), pl. táli, paptalasselindea < papt- ’small leaf’ (QL:72) , lasse (e) ’leaf’ (QL:51), #lindea ’musical’ (cf. lindelea ’melodious’ (QL:54)) + pl. -r
A special kind of pattern is when adjectives agree with case-inflected nouns in number, but are themselves put into the nominative. The adjective may either precede or follow.
OM1g: laikven ondolissen *’on the green rocks’ < laiqa ’green’ (QL:52), nom. pl. *-ain > -en; ondo ’stone’ (PE15:77, PE16:138)+ inessive pl. -li-ssen
OM1,OM1f,g: alkarissen oilimain *’in the shinings [that are] last’ < alkar ’shining, light of day’ (PE16:75) + inessive pl. -issen, oilima ’last, ultimate’ (PE16:75) + archaic/poetic nom. pl. -in
[In ailinisse the plural marker -n is missing; the suffix -isse forms the inessive/locative singular of consonantal nouns as pilin > pilindisse (PE16:113), but ailinisse is perhaps poetically interpreted as plural with pl. -i.]
The nominative might be explained as a construction with a relative sentence: *’on the rocks that are green [ones]’.
3.4 Discussion
Since we are once again dealing with poetry here, several constructions must be archaic or poetic. Although there are examples for most of the possible patterns, some quantification can be made.
For preceding adjectives it seems usual not to agree in case or number, for singular and plural nouns alike. The only example with agreement is laikven ondolissen, hence probably poetical – and this goes together with EQG. But if the noun is not case-inflected there are some examples with agreement (even manifold as in laiqali linqi falmari), but no example without agreement.
Postpositive adjectives always agree in plural, there are no examples without. But there are about as many samples with case agreement as without it – perhaps one of the two patterns is also poetical.
The plural markers employed are -i (including -e < -ai (PE14:45)), -r and -li. PE14:46 explains for forms like anda ’long’, pl. ande (rather than *andali) that the differentiation of the endings has come about through tending to avoid the longer endings and their jingling repetition – but the -li of vocalic plurals is originally secondary […]. Another reason for the alternation may be their fitting into the metre, as in ondoisen andalissen. No pl. -r is found in the Oilima Markirya versions, probably because of its coalescence with the dative ending.
The conception from EQG that adjectives always agree in the four cases (as compared to the adverbial suffixes) does not seem to find its application. We find for instance lúnelinqe vear and other similar samples with dative -r.
4 Quenya of the 30s and 40s
This timespan includes mainly the samples given in the The Lost Road and one example from the draft to The Lord of the Rings.
preceding adjective (plural noun), no plural agreement
ilya maller ’all roads’ (LR:47, SD:310) < ilya ’all, the whole’ (Etym:IL-), malle ’street, road’ (Etym:MBAL-) + pl. -r
preceding adjective (plural noun), no case agreement, no plural agreement
Kuivië, kuivië! laurea’esselínen >> ankalim’esselínen *’Vivat, vivat! By (many) glorious names’ (SD:47) < kuivië glossed ’awakening’ (Etym:KUY-), ’life’ (VT42:8) is here perhaps a literal translation of hail = ’health, prosperity, good luck’ or something like vivat ’live (long)!’, laurea ’golden, like gold’ (Rgeo:66) – perhaps poetically for ’glorious’, esse ’name’ (VT43:14, PM:339) + part. pl. instrumental -lí-nen
Unfortunately, the changed variant shows elision of the final vowel, so that we do not get to know whether sg. ankalima ’exceedingly bright’ (Let:211) or pl. *ankalime was intended.
5 Quenya of The Lord of the Rings
It is probably better to regard The Lord of the Rings separately. As the first piece of work released for the public and containing substantial Quenya material, it surely had a lot of influence on the following development of Elvish. Although Tolkien always kept experimenting, it is good to keep in mind which patterns were ’fixed’ in the LotR. A comparison with the non-poetical word order of Namárië given in The Road Goes Ever On (1966) is also noteworthy. (See http://www.uib.no/people/hnohf/namarie.htm for an analysis.)
preceding adjective (singular noun), no case agreement
yuldar lisse-miruvóreva ’draughts of the sweet mead’
i·aire táríva or tário ’of the holy queen’ (prose, PE17:76)
Note that at some time Tolkien interprets laurië as the adverb ’goldenly’ (homophone to the abstract noun ’goldenness’) formed with the adverbial suffix -ie as in norne lintie ’he ran quickly’ (PE17:58,61). This is not the sole explanation, however – The Road Goes Ever On cites the adj. laurëa (pl. laurië) ’golden’.
Note also Tolkien’s later explanation for lisse-miruvóreva: The prefixing of an uninflected adjective to a noun as in lisse-miruvóre to make a loose compound is normal in Quenya. So, as in English, is the prefixing of an uninflected noun to make a loose genitival or adjectival compound: so tári-líre "queen-song". But the piling up in this line, aire-tári=lírinen and the word order are in poetical style (PE17:67; a similar, but shorter account in PE17:76).
We can see that the concept of loose compounds (both ’adjective + noun’ and ’noun + noun’) itself remained, and in lisse-miruvóre only the noun receives the ending.
However, there is no attestation of ’adjective + pluralized and case-inflected noun’ in Namárie.
6 Quenya of the 50s
This time includes the drafts for the translations of Catholic prayers (VT43-44), the essay Quendi and Eldar (1959-60) and the Nieninqe version N2 from 1955 (PE16:96). There are not many examples with case-inflected nouns, but some with plural inflection:
preceding adjective (plural noun), no plural agreement
[update : Further linguistic terms of this kind appear in the Tengwesta Qenderinwa (probably first written in 1937, later revised around 1951, see PE18:10,20): qenderin ólamar ’Quendian consonants’ (PE18:30), qenderin ómandi >> Qenderin ómar (PE18:32,83) and alómea na-súrimar ’voiceless aspirate’ (PE18:30). The last word of the latter example is probably a nominalized adjective, thus lit. *’voiceless aspirates’.]
rakine tengwi ’stripped/deprived signs’ (VT39:6) < #rakina ’tripped, deprived’, cf. rákina ’broken’ in MC:223 [consonant with the following vowel lost]
penye tengwi ’lacking/inadequate signs’ (VT39:6,8) < penya ’lacking’ (cf. PEN- ’the lacking, the poor’ WJ:408) [vowel with the preceding consonant lost]
N2: wingildi wilwarindie ’foam-faires like the butterflies’ (PE16:96) – see above
N2: táli lantalasselingie ’feet whose music is that of falling leaves’ < lantalasselingea ’with a musical sound of falling leaves’
preceding adjective (singular noun), case agreement
Amille Eruva lisseo ’Mother of divine grace’ (VT44:12,18) < amille ’mother’ (cf. amil (Etym:AM(1)-), Eruva ’divine’ < Eru ’the One’ + -va, lisse ’grace, sweetness’ + gen. -o
[Eruva can also be interpreted as the possessive inflection of Eru rather than an adjective, i.e. ’mother of the grace of God’.]
preceding adjective (plural noun), no case agreement, no plural agreement
ilya raxellor ’from all dangers’ (VT44:5) < ilya ’all, the whole’ (Etym:IL-), raxe ’danger’ + ablative pl. -llo-r
Aran linta ciryalion *’king of swift ships’ (PE17:147) < aran ’king’ (ibid.,WJ:369) + linta ’swift’ (Rgeo:66), cirya ’ship’ + part.pl. genitive -li-o-n
[There is also a variant with the possessive case and a full compound: Aran lintaciryalíva]
The linguistic terms musse/lehta/sarda/tapta teñgwi are from the note to Quendi and Eldar called Noldorin words for Language. The forms given in Quendi and Eldar Appendix D contain singular adjectives: hlonaiti/quante/rakine/penye tengwi. We seem to meet two different conceptions here – one following ilya maller, the other ilye tier. It is not clear, however, if the two may coexist. Perhaps they could; and an uninflected form results from a treatment as a loose compound where musse·teñgwe etc. are regarded as single morphological units. The other treatment would be that of a qualifier + noun.
7 Quenya of the 60s
This part includes examples from several late essays and the last two Markirya versions (MC:221-223), the later one will be designated OM3 by convention.
7.1 With case inflection
preceding adjective (plural noun), no case agreement, no plural agreement
OM3: i fairi néke ’the pale phantoms’ < faire ’phantom, disembodied spirit’, pl. fairi; néka ’vague, faint, dim to see’, pl. néke
OM3: maiwi yaimië ’like gulls wailing’ < maiwe ’gull’ (Etym:MIW-), pl. maiwi; yaimëa ’wailing’ (adj.), pl. yaimië
OM3: ve tauri lillassië *’like many-leaved forests’ < taure ’great wood, forest’ (Etym:TÁWAR-), pl. tauri; lillassëa ’having many leaves’, pl. lillassië
postpositive adjective (plural noun), no plural agreement
-not attested-
8 Summary and final discussion
8.1 Attributive usage in the late years
Which patterns were dominant in the last decades? At first one has to state that there is not a single example of case agreement, not even in the poems – so it does not seem to be usual, at best poetical.
Let us consider plural agreement. For adjectives preceding a case-inflected plural noun we find five examples without agreement:
ve quenderinwe koaron ’after the manner of bodies of Elvish-kind’ (PE17:174)
Among these, atalantië >> atalantëa was a change from the second pattern to the first and quenderinwa >> quenderinwe went just the opposite way. This distribution goes well with the explanation of the EQG: where a qualified noun receives one of these endings [-sse, -llo etc.] (a somewhat archaic mode) the adjective usually precedes uninflected (except rarely for plural) and is virtually a loose compound (PE14:78).
On the other hand, for nouns that are not inflected in case, plural agreement seems to be more usual:
For postpositive adjectives there is only one pattern – as far as the samples go, they always agree in plural. Probably it needs the convention of an immediately preceding adjective (be it a loose compound or not) to drop inflection and still keep the meaning unambiguous.
8.2 Predicative usage
Let us finally take a look at predicative adjectives. There are not many examples, and they are many years apart, but we can find samples for both possible patterns as well. Without plural agreement:
piliningwe súyer nalla qanta *’the airs being full with feathers’ < pilin ’feather’ (QL:54), possibly pl. pilini + -ngwe ’with’, súyer, pl. of súye ’airs, breezes, winds’ (QL:86), nalla *’being’, qanta ’full’ (QL:78)
ondoin morinninqe ’the dark rocks [were] white’ (PE16:64) < ondo ’stone’ + archaic/poetic nom. pl. -in, ninqe ’white’, more *’dark’ + pl. -n
naltur […] ulqe nūsimar *’they are noticeable as evil’ (PE15:32,33) < naltur *’they are’ (cf. NĀ ’be, exist’ (QL.64)); ulqa ’bad, evil’ (PE15:70), pl. -e; #nūsima *’noticeable’ (cf. nuste ’sense of smell’, Goldogrin nus- ’take notice, perceive’ (GL:61)) + pl. -r
nalto ūsiēre (PE15:32,37) *’they are having escaped’ < nalto *’they are’ #ūsiēra ’having escaped’ (cf. usin ’he escapes’ (QL:98), past tense *ūsie + adj. -ra), pl. -e
nalto fustūme *’they can be detected’ (PE15:32,39) < nalto *’they are’ #fustūma *’noticeable’ (cf. fausta- ’to get smell of’ (QL:37)), pl. -e
táli paptalasselindeën ’the music of their feet was like falling leaves’ (N1b variant, PE16:93)
ilya sī maller raikar lit. ’all now roads bent’ (LR:47, SD:310) < sí ’now’ (Etym:SI-), raika ’crooked, bent, wrong’ (Etym:RÁYAK-)
toi írimar *’which are beautiful’ (LR:72) < toi ’they’, írima ’beautiful’ (cf. ĪR- ’desireable, beautiful’ (VT45:18)), pl. -r
Eldar ataformaiti ’Elves were ambidextrous’ (AS4-6, VT49:6-8)
An emendation goes here ataformaite >> ataformaiti, from the first pattern into the second.
Note that the verb ’to be’ is usually omitted, even in the past tense.
The example táli paptalasselindeën is a curious one. If interpreted correctly, this literally means *’[their] feet [were] leaf-music-like’, i.e. the actual subject táli does not receive a nominative suffix, but the predicative adjective does. The same variant reading shows nominative n·oromandin and wingildin wilwarindeën. Compare the pattern given in 3.3.
In raikar, írimar the plural marker -r is used, at that time it is typically a plural marker for nouns and adjectives pluralized this way are used substantively like vanima ’fair’ (Etym:BAN-), pl. vanimar ’beautiful ones’ (LotRVI, ch.6; Let:230), linda ’fair, beautiful (of sound)’ (Etym:LIND-, VT45:27), pl. Lindar ’the fair’ as a tribe name (LR:168). So mallerraikar may literally mean *’the roads are bent ones’ and toi írimar *’they are beautiful ones’. But ataformaiti is ordinarily pluralized by -i.
Finally, a rather special example is found in LR:72:
Toi aina, mána, meldielto ’They are holy, blessed, and beloved’ < toi ’they’, aina ’holy’ (Etym:AYAN-), mána ’blessed’ (cf. Etym:MAN-), melda ’beloved, dear’ (Etym:MEL-), compound tense meld-ie-lto ’beloved-are-they’
At this stage the verb ’to be’ is expressed by a suffix (or compound tense) -ie. Note that the three adjectives are all in the singular, though ’they’ would require plural. This plurality only seems to be expressed by the final -ie-lto ’they are’ with the pronoun repeated again. Compare a similar construction with presumably attributive adjectives in oïkta rámavoite malinai or oïkta rámavoite karneambarai in Narqelion, but on the other hand plural agreement laiqali linqi falmari in Earendel.
I would like to thank Thorsten Renk for his critical remarks on the first version of the article.
