Etymologies of the Atani Languages

by Lalaith <>



This project grew out of an attempt to demonstrate whether etymological links can be established between Adûnaic and its nominal successor Westron. My research ultimately let to establishing the known vocabulary of all the Atani languages and showing their suspected relationships.

It can quickly be seen that the relevant terminology displays hardly any overlap. This is due to the selectiveness of the sources (historic literature and myth-making for Adûnaic vs. the rustic background of - hobbitish - Westron). A few relationships particularly to the Elvish and Northern Mannish languages are firmly established; others remain highly speculative and we cannot tell whether they are made in purpose or the result of having been created by one single mind that was involuntarily repeating familiar patterns.

Tolkien devised indeed three different versions of Adûnaic that I inofficially denominate k-, d-, and a-Adûnaic, based on their individual words for "Earth; Arda": kamat, daira, *abân (for a full discussion of the three Adûnaics, see "Lalaith's Guide to Adûnaic Grammar" on this homepage). k-Adûnaic was in its grammatic structure fully replaced by d-Adûnaic. However, its vocabulary was not always obsoleted by later versions of the same texts. While there can be no doubt for example that k-Adûnaic hunekku "he came" was replaced by d-Adûnaic unakkha it is not that clear whether daira really replaced kamat and *aban replaced daira or whether these were synonyms developing out of different bases. Much of the vocabulary clearly survived the transition, thus zigûr wizard, bâr lord, etc.

It is even less clear how much of the d-Adûnaic vocabulary survived into a-Adûnaic that emerged during the development of the LotR Appendices and is recorded only in a few proper names: While much of the vocabulary was evidently retained (p. e. balâk ship, gimil starry sky) there are distinct differences in the sound patterns, and we can only guess on how much of the established grammar and vocabulary of d-Adûnaic stayed valid.

Informations on the earlier evolution from the Hadorian stage on we have only concerning d-Adûnaic, recording also from two Third Age languages the words for moon and sun: one pair of them, Nil and Uir, correspond to the familiar transskription patterns of Westron, the other pair, Njûl and Ýr, does not. It is the merest speculation that the latter two may have belonged to a Black Númenórean descendant of Adûnaic ("Black Adûnaic"), moving away from Westron under the influence of Harad languages to become a kind of Southron pidgin. It is possible that this was still the tongue in which the Haradrim troops communicated during the War of the Ring, reminding with its harsh sound patterns the Gondorians of "carrion-bird cries".

The comparison of Adûnaic and its "niece" Westron was much helped by the "tentative reconstruction" of a Westron grammar in Tyalië Tyelelliéva 17, subsequently referred to as TT17, with the caveat that the linguistic history presented there is sometimes garbled and some of the vocabulary doubtfully attributed to the corresponding languages. It seems that there were in fact two Westron grammars in existence, one mimicking a Germanic, the other an Elvish origin. Apparently the former is presented to us. Its vocabulary indicates a close relationship to AL though some variations in spelling are notable, p. e. narac vs. AL narag Dwarf. It can thus be dated later than the composition of d-Adûnaic in RA but probably before the final acceptance of Adûnaic as the private language of Númenor in KR that would have made a Germanic origin of Westron impossible (see LP on Tolkien's changing concepts of the fate of Adûnaic). That TT17 was at least inspired by RA is demonstrated by the evidently same structure of verb syntax, both using subject and object pronouns as affixes, and overlaps in the samples given to illustrate grammatical features: urûg vs. ruccu bear; tamar vs. dramar smith; showing the application of sexes (not genders) to a noun, both sources use the examples "she-bear" and "he-goat", etc. Thus in cases where Westron cannot easily be derived from d-Adûnaic it may perhaps be allowed to suspect potential features of the otherwise hardly known a-Adûnaic behind.

The evolution of Atani

To be correct, despite the popular statement that Westron "was in origin the language of those whom the Eldar called the Atani or Edain, 'Fathers of Men', being especially the people of the Three Houses of the Elf-friends who came west into Beleriand in the First Age" (LP) there never was an Atani language in the proper sense. Only two of the three tribes or "houses", Bëor and Marach, used languages that developed from a common ancestor. This was later referred to as Atani (LP), spawning what might be called the "Atanic family" of languages. The third tribe, Haleth, used an unrelated tongue that the other Edain found incomprehensible. To the "Halethic language family" later belonged many pre-Númenorean tongues of the Second and Third Age, the most important among them being Dunlendish. These languages will not be considered here.

During the migration from Hildórien, Atani left behind various descendants that developed into the "Northern Mannish" varieties, among them the Éothéod idiom that became Rohirric, Dalian, and the Woodmen tongues. Before the remaining tribes had reached Beleriand, their language further divided into two still mutually comprehensible dialects: Bëorian and Hadorian. Of Bëorian very little is known for its native speakers quickly adopted Sindarin so that it fell extinct before it could leave written records. From the preserved fragments it may be deduced that its set of sounds was richer than that of Hadorian, in particular it developed very early the vowels -e-, -o-, -ê-, -ô- that the latter was lacking. The very expression Bëorian was impronouncable to a Hadorian speaker: His closest approximation may have been *Bîuryai.

Hadorian that A. Lowdham vaguely called the "pre-historical" ancestor in RA "was not forgotten, and from it came the common tongue of Númenor" (S), Adûnaic or, as it might have been called in its own terms, *Adûnaiyai > *Adûnaiyê. Lowdham distinguished four more or less clearly defined stages:

  1. "ancient" or "primitive", used in the time of Elros Tar-Minyatur,

  2. "early", perhaps from the time of Tar-Alcarin,

  3. "classical", spoken and written at the time of the Downfall,

  4. and finally "Exilic". He was not aware that Exilic, the latter Common Speech of Northwestern Middle-earth (Adûni, Sôval Pharë), was not a descendant of Classical but of Early Adûnaic and related to the Classical mode like Low Latin to High Latin.

The steps of the evolutionary process are marked by the loss of several consonants that Hadorian knew: -S-, -C-, -Ch-, and -2- all merged into -S-, and -?-, -9-, or -3- vanished altogether. Long -ê- and -ô- were obtained by Classical Adûnaic [maybe also short -o- and -e- and -sh- that are found in a-Adûnaic only], over-long vowels developed in the early stage, only to vanish again by merging with the common long ones. The grammatical structure apparently did not change very much beyond the slow blending of the subjective case across the grammatical genders (or, more properly, sexes) but traced a continuous development that has been explained by Lowdham (NC).

Parallel to the development of Classical Adûnaic, in the "forts and havens upon the western coasts of Middle-earth" Early Adûnaic "mingled with many words of the languages of lesser men", i. e. from the descendants of Halethic and Northern Mannish, and "became a Common Speech that spread thence along the coasts among all that had dealings with Westernesse" (LP). The amount of foreign influence on this continental (I incline to call it "Low") Adûnaic cannot be traced but may account for some otherwise hard to interpret Westron vocabulary, such as ribadyan, hamanullas . As a side-effect, this developing koinë of Middle-earth was one of the inspirations for Sauron's great attempt to fashion-design a lingua franca of Evil, the Black Speech.

With the Downfall, Classical Adûnaic fell dead, except perhaps for a continued usage as a tongue of lore among the Black Númenóreans. The Exiles considered Sindarin their mother tongue and began to use "the Common Speech in their dealings with other folk and in the government of their wide realms; but they enlarged the language and enriched it with many words drawn from the Elven-tongues." (LP). And so, "Exilic" Adûnaic turned into Archaic and then Late Adûni or Westron, slowly dividing into various dialects such as Hobbitish, Gondorian, etc. The Gondorian Westron spoken in Minas Tirith was ultimately considered as dictating the standard of "correct" Westron.

The family tree of the Atani languages can thus be illustrated like that:


1 Beorian †

2 Hadorian

...a Ancient Adûnaic

......I Early Adûnaic

.........A Classical Adûnaic

............i "Black Adûnaic"?

.........B Low Adûnaic

............i Ancient Westron

...............(1) (Gondorian) Westron

...............(2) Hobbitish Westron

...............(3 - n) various other dialects

...b Northern Mannish (Rhovanion)

......I Éothéod

.........A Rohirian

......II Anduin-Mannish

.........A Anduin-Hobbitish (Stoorish)

......III Woodmannish idioms

.........A Dalish

...c Northern Mannish (Eriador) †

Outside influences

"...these Men had long had dealings with the Dark Elves east of the mountains, and from them had learned much of their speech, and since all the languages of the Quendi were of one origin, the language of Bëor and his folk resembled the Elven-tongue in many words and devises." (S) Since Adûnaic was not a descendant of Bëorian, it is not clear - though likely - whether this concerns the language of Númenor as well. But at least one etymological link to Avari is obvious: The original Atani base meaning "folk, people" may have been *TUR [as preserved in Rohirian tûr) but in Hadorian it was replaced by lâi, apparently imported from Avarin lai (QE; cf. Quenya lië). Another such import Lowdham suspected to have found in Adûnaic minal sky, heaven, that resembled Quenya menel only because it was drawn from a related language. It is therefore dangerous to assume direct imports from Quenya or Sindarin whenever an outward similarity of words is recognised. True imports follow established patterns: Thus as Quenya -é- usually became Adûnaic -î-, we may also consider that Westron nîn indicates an early Adûnaic borrowing from nén; but for example Adûnaic khôr "lord" despite the first-glance similarity certainly is not a loan from Elvish khéru for that would have been rendered as *khîru!

Another source of loans and borrowings was Khuzdul, the ancient language of the Dwarves. However, "the theory (a probable one) that in the unrecorded past some of the languages of Men - including the language of the dominant element in the Atani from which Adûnaic was derived [i.e. Hadorian] - had been influenced by Khuzdul" (DM) is hard to proof. One would expect that the influence was for example particularly profound in the terminology of mining and mineralogy, thus Adûnaic pharaz gold, Bëorian bor stone, archaic Westron phûrun delf, Khuzdul kibil silver, might each have had their relatives in the other languages. Unfortunately, the known vocabularies of the individual tongues hardly overlap and therefore no such pattern can be traced. It looks, however, as if at least kibil had been adopted by Hadorian, producing the Adûnaic adjective gimil which would be a homonym of gimil star but not related to it.

Only one further etymological relationship between Khuzdul and the Atani family is suggested with some credibility: Adûnaic azar, azra sea may be inspired by the same base *Z-R that gave Khuzdul zâram lake, pool, mere, narrowing down its meaning from a general "body of water" when the Edain discovered the Great Sea. Another, less probable link has been suggested by some who wanted to compare the Adûnaic zagar sword with the Dwarvish name Azaghal, believing that the latter meant "warrior", though this is nowhere attested. If we want to accept this doubtful conclusion we may also file Orcish name Azog.

A few borrowings into Westron came from entirely alien sources such as Haradric languages (múmak, in Gondor replacing an older word that was rendered as oliphaunt, possibly a derivative from Primitive Elvish andambunda: Adûnaic *annabûn?), and even from the Black Speech (Nazgûl).

Common grammatical structures

Note: The historical development of Westron is somewhat unlikely. While Adûnaic was equipped with „a faintly Semitic flavour“, Westron clearly resembles Germanic languages. In GN, Tolkien confesses that the word sûza „Shire“ was inspired by an Old Norse and modern Icelandic expression sýsla; the definite form of the nouns ending on -t (Sûza-t, „the Shire“) resembles similar phenomena in Swedish, etc. We also often have to distinguish between high Westron (spoken in Minas Tirith) and the hobbitish dialect that aside of its unusual usage of personal pronouns features some shifts in the phonological system, thus initial hl > l (hlothram >lothram „cottager“), the pronunciation of the definite plural like the definite singular (dramarins vs. dramaris „smith“), loss of final vowels (kastu > kast „mathom“ while Rohirian retained kastu), reduction of two equal vowels with intermittent -h- into a long vowel (trahan > trân „smial“, analogous loho > lô „horse“ in Rohirian), aside of course the usage of certain special expressions unknown to high Westron, notably kuduk „hobbit“ with its Rohirian relative (but not progenitor!) kud-dûkan „hole-dweller“. These shifts are remarkably minor if we consider that hobbitish Westron had by then for much more than a millenium stayed without contact to Gondorian. Real-world languages take less time to split into mutually hardly understandable idioms!

Very little can be said about similarities in grammatic structure in the Atani family, for the only language from which complete phrases are preserved is Adûnaic. With regard to the others, just a few observations can be made:

Very speculatively, further traces of grammatical development between d-Adûnaic and Westron can be suggested like that:

How to read the list

The following list necessarily involves a good deal of speculation. Tracing Classical Adunaic backwards to its bases leads inevitably to ambiguous results, a phenomenon that Lowdham was well aware of when he indicated that p.e. Adunaic lômi would refer to either a base *LUM or "more probable" *LAW'M; unfortunately he does not tell us why the latter would be more probable. Doubts like this are manyfold, and so a lot of arbitariness in arranging the vocabulary by suspected bases I could not avoid. To mark out the "reconstructed" elements, I set them in red rather than using asterisks, so not to mix them up with those markers found in the sources themselves.

Two different systems have been established in the traditions to transliterate the Atani languages into Latin characters. A. Lowdham more or less consequently indicates long vowels by a dash, archaic over-long ones by a circumflex and short ones occasionally by a semi-circle. It is evident that he understood Adûnaic much better than the Red Book's translator J.R.R.Tolkien who assumed that dropping some arbitrary circumflexes was sufficient: "The use of the circumflex in other languages such as Adûnaic ... has no special significance, and is used merely to mark these out as alien tongues (as with the use of k)." (WS) Thus, while Lowdham transscribes Adunaic and Ar-Pharazon, Tolkien writes Adûnaic, Ar-Pharazôn (and in AL uses as well dashes instead of accents).

The above-mentioned letter k represents the same sound as Elvish c except in languages of the Halethic family that featured a distinction, for example in the proper name Kalimac. (The hobbit-name Cora seems to defy this but actually it is a semi-translation of proper *Koro.) In d-Adûnaic, however, a consonant c has ultimately merged with s; it is not clear what sound Lowdham had in mind, perhaps he used it for ts as the Slavic languages do; this may then also be true for Halethic words like Kalimac. The usage of k is however not yet found in the older manuscript drafts for W. which in establishing the Etymologies produces a couple of problematic inconsistencies. To clarify the relationships between words I have thus taken the liberty to write k throughout but mention the original spelling in brackets where required.

After trying several ways of presentation, I settled with the one used also in TE for this way best displays the presumable relationships between the investigated languages, but for clarity I used more paragraph marks. The character ">" between two words signifies the development of one form into another; if the relationship so stated is very doubtful it may become "?>". The character "-" denounces compounds made from the stated vocabulary.

Alien bases are kept separate and as in TE marked by a subsequent dash that Atani roots usually omit (according to RA). Words that cannot be referred to bases are given in a third list. The many uninterpretable hobbit first names such as Droga, Bilba, Tomburân I did not add to any list.

Unless mentioned otherwise, the sources of the entries are NC (including DA, RA) for Adûnaic and AL for the Third Age languages.







aorist tense










Black Speech


common sex




continuative tense


Dalian (language of Dale)


dual number






female sex




hobbitish Westron






male sex


neuter sex


Northern Mannish




objective case




Primitive Elvish


past participle


past tense






plural number


"preposition", postposition


present tense










singular number


subjective case





The Etymologies



to, towards, against;
Ad. pp. -ad(a), after vowels -yada
- avaloiyada against the Powers (Valar),
- azûlada eastward,
- akhâsada into the abyss;
- Gimlad Starwards (= Q. Elenna).
?- dalad under (or more properly: downwards?) < DAL.


Ad. pp. an- of. Often shortened to ‘n-.
- an-Adûn, 'n-Adûn of the West, anadûni western, Anadûnê Westernesse
- an-Nimruzîr of Elendil, Elendil's
- thâni anAmân, thâni n'Amân Land of Valinor


Ad. pp. –aw > -au > from, after vowels -vô. According to Lowdham's transscription rules this should be –wô, but it is -vô in the Lament of Atalante ;
- Êruvô of God, God's,
- patr. -ôhîn < KHIN.


Part of a hypothetical base that may have produced a weak Ad. pref. i- she analogous to ?U-, 3U- vs. HU-.


Ad. êluk, prep. of unknown meaning, maybe an intensification of a3il > ail > êl, cf. êphal, êphalak, see PHAL.


one, alone. Said to be a very fertile base.
Ad. ?airu > Êru God, in a-Ad. probably derived from Q. Éru;
- Êru-hînîm Children of God < KHIN. The entire compound may be adopted from Q. Eruhíni.


a-Ad. pp. aub- > ob- before;
- obroth fore-cutting < RUTH.


Ad. pp. -ud- on, over?;
- nênud on us < NIN,
?- ugrudâ- to overshadow < GUR.


Ad. m. suff. –aun > -ôn, denominator of weak nouns;
- pharazôn the Golden One < PHARAZ.


Ad. m. suff. -ûr > W. -ur, denominator of weak nouns;
- Ad. zigûr wizard < ZIG,
- h. W. Razanur wanderer; Peregrin.



Ad. aux. not;
- bâ-kîtabdahê, "not you touch me", don't touch me < TAB'D.


Ad. balak ship, pl. balîka;
- balkumagân
shipwright (PE) < MAG. Balku- is an obj. and hence stands in sing. analogous to the translation "shipwright" rather than "shipswright".


Height, elevation?
Ad. aban earth (or rather "elevated land", as rising above sea level?) see DAY'R;
- Aban-Târik > Abattârik Pillar-of-Earth < TAR'K.
W. banno hill, pl. bannut (TT17).


Ad. ban;
- banâth wife, the f. half of a couple, a weak noun with f. suff. –âth, analogous banâk husband
W. ba-, ban(a)
- banakil halfling < KHIL, only high W., the proper h. W. word being kuduk < DUK;
- h. W. Banazîr Half-wise < ZIG,
?- h. W. personal name Bannâtha.


Ad. abâr strength, endurance, fidelity, bâr lord, an import from E. BOR- endurance has been also suggested. See also KHUR;
- abrazâ-
to stand fast, p.p. abrazân, steadfast, faithful (= Q. Voronwë),
personal name Ulbar
- bar-rad, barrad > brad borderland, march < RAD,
?B. personal name Baran
W. bralda heady (with part. suff. -ald as in N. M. trahald, nahald?), brand(a) borderland, march, bara quick,
- Branda-nîn, borderland-water (= river Brandywine; LP) < NIN,
- Bralda-hîn, heady ale (= river Brandywine; LP) < HIN,
- Barabatta Quick-talker < BITH.


W. basto cake, pl. bastun (TT17).


Ad. bawab wind, subj. pl. bawîba.


a-Ad. bayl > bail > bêl light; not identical to bêl- love, see BIL.
- Belzagar Light-sword (= Q. Calmacil) < ZAG'R.


Ad. bail- > bêl- to love see also ZIR; not identical to bêl- light, see BAY'L.
- Azrubêl Sea-lover < ZAR.


Ad. bain > bên servant,
- Êru-bên Servant of God (= Vala), pl. Êru-bênî.
?B. bëor vassal.


say, maybe influenced by S. beth word < E. *KWET say, speak.
Ad. baith > bêth expression, saying, word, bêthân speaker
- saibêth assent,
- izindu-bêth "true-sayer", prophet < ZID,
- Inzilbêth < JIL.
W. bat(t) expression, saying, word, batta speaker
- barabatta quick-talker < BAR. Apparently, W. had given up the habit of Ad. to use only the stem in making such compounds, cf. izindubêth. It is unlikely that the whole compound was equipped with the male denominator -a of h. W. first names.


Ad. buruda heavy, emphasised burauda > burôda very heavy. Another possibility is the derivation of a stem bur- and a pr. part. suff. -auda that would have been the equivalent of N. M. -ald (changes of -au-, -av- to -al- are demonstrated by some languages, p.e. Slovene).



Ad. a "formative prefix" of unknown meaning.


Ad. prep. dal down, below, dalad under - or perhaps more properly: downwards < ?AD.
(Any relationship to cognate Slav. dol?)


Ad. Adar, Adra-, uninterpretable
- personal name Adrahil.


Ad. -, element of various conjunctive compounds:
- kadô and so,
- îdô now, subj. îdôn,
- tâidô once (upon a time).


Ad. *daw'r > dâur.


Ad. dâyîra > dâira earth. See BAN1.


Ad. dubdâ- to fall (p.e. under a shadow, not downwards); ao. pl. dubdam, pt. pl. dubbudam. Contrasting in meaning to KAL'B.


Ad. opt. aux. du-;
du-phursâ so as to gush < PHUR.


R. dûk- to dwell, dûkan dweller,
- kûd-dûkan hole-dweller < KUD
> h. W. kuduk hobbit, see BAN2 (LP).


Ad. dolgu night (so Lowdham, but more properly dôlgu < daulgu?), a word implying evil darkness, dulga or dulgi, pl. dulgî, black.



W. gamba buck (or TT17 goat, with râgamba buck).
- Zaragamba Oldbuck < ZAR,
- Brandagamba
Marchbuck < BAR (LP).


W. (u)gad- to stay;
- Ranugad Stay at home; perhaps with male declension u-; for an alternate possibility see RAN.


Ad. n. agan, m. Agân (= Q. Mandos?) death;
- agannâlu death-shadow, subj. agannâlô < NUL


Ad. agla, uninterpretable element of Aglahad. (GC)


W. galap, galab- game
- Galabas Gamwich ("popularly supposed", LP)
- Galbasi "Gammidgy" > Galpsi "Gamgee" (LP) < MBAS-.


W. adj. glibi fat (TT17)


Ad. gimli; igmil star-shaped figure, coll pl. gimil starry sky
- Gimlad Starwards (= Q. Elenna) < ?AD
- Gimilnitîr Star-kindler (= S. Elbereth) < NIT
- Gimilzagar Star-sword (= Q. Elemmacil) < ZAG'R
- Gimilzôr Star-foam (= Q. Elros) < ZAW'R
- Gimilkhad.
Lowdham gives many more variations of this base but does not tell whether these are actual words or just theoretical possibilities.


Ad. ugru shadow, ugrudâ- to overshadow [< ugru + -ud on?].


*3A- [*?A-].

Part of a base that may have produced Ad. c. pref. a-, it.

*3I- [*?I-].

Part of a base that may have produced Ad. m. pref. i-, see also HI.

*3U- [*?U-].

Part of the base that produced Ad. m. pref. u-, see also HU
- unakkha, he came < NAKH,
- ukallaba, he fell < KAL'B. We do not know how this prefix reacts if the verbal stem begins with a vowel, such as azgarâ-.



Ad. had, uninterpretable element of Aglahad. (GC)


W. halda tall (TT17)


Ad. obj. pron. -hai > - me.


Ad. hazad, hazid seven.
Perhaps related to the proper pre-Númenorean name Hazad (TE, cf. Latin Septimus).


Part of the base that produced Ad. f. pref. hi-, see also 3I
- ao. hikalba she falls/fell, p. t. hikallaba she is/has fallen.
- hi-Akallabêth, She that is fallen < KAL'B. There is no visible grammatical argument for applying this article. The personification into f. sex was already expressed by applying the suff. -ith to the adj. akallaba fallen, but if an emphasis should be put the subj. Akallabêthin would have been proper.


Ad. hil, uninterpretable Element in Ad. names Adrahil, Imrahil.


W. hîm(a) beer, ale.
- Bralda-hîm, Heady Ale (= river Baranduin; LP) < BAR.


Part of the base that produced Ad. m. pref. hu-. See also 3U.


Ad. huzun ear, d. huznat two ears (grouped).



Ad. injil > inzil flower;
- Inziladûn Flower of the West < DUN (LE); a violation of the grammar: as inzil is standing in object position the meaning actually is "West of the Flower", correct was either Adûninzil or Inzil an-Adûn!
- Rothinzil (a-Ad.; more properly Rôthinzil? < RUTH) Foam-flower;
- Inzilbêth,
Flower-speaker??? < BITH.



Ad. ka and.
- kadô and so < DAW.


Ad. kadar city;
- kadar-lâi, city-folk < LAI-.


Ad. kali woman; see ZIN.
? (with loss of first syllable) W. lê- she, lêva woman (TT17)
- lêrukku (lêruccu) she-bear < MOROK-.(TT17)


fall (down).
Ad. kalab to fall, ao. kalba, p. cont. kallaba, p. t. akallaba; kalub- to lean over, pl. kalubîm. See DUB'D.
- m. sing. ukallaba, f. hikallaba
n. pl. yakalubîm
- akallabi fallen (adj, not adv. which would be kalabân)
- personified Akallabê > with f. suff. (hi-)Akallabêth She who hath fallen. (= Q. Atalantë).


Ad. kan-; no forms known.


Ad. n. karab, m. karbû, f. karbî, pl. karîb. See also LOH.


Ad. kar(a)sa > krassa (crassa) red (TT17)
- krassat (crassat) Red One (TT17)
The vague similarity with S. carn is probably accidental. Note that the orographic name Caradhras in W. was then Crassaras!


W. kasto (casto) gift (TT17)
> kastar (castar) valuable, a monetary unit (= S. mirian),
kastu treasure (LP), kastun (castun) the treasures (TT17)
> h. W. kast mathom.
R. kastu. (LP)
See ZIM'R.


Ad. katha, kâtha all; pk. kathî;
- kâthuphazgân conqueror, lit. "all-ruler"? < PHAZ.


Ad. sing. subj. pron. kî- you


Ad. kaiw > kêw > kêu. Uninterpretable.


R. kûd hole;
- kûd-dûkan, hole-dweller < DUK.
> h. W. kuduk hobbit


Ad. kulub "edible vegetables that are roots not fruits"; kulbî "a definite number of roots or plants".
Lowdham gives many more variations of this base but does not tell whether these are actual Ad. words or just theoretical possibilities.


Ad. kauy > kôy > kôi. Uninterpretable.



Ad. khad, uninterpretable element of Gimilkhad.


Ad. akhâs chasm.
- akhâsada into the chasm < ?AD.


Ad. khâu > khô crow, pl. khâwi(m) > khôi.


Ad. khibil spring, not clear whether the season or a fountain is intended.


a-Ad. khaur > khôr lord;
- Adûnakhôr
, Lord of the West (= Q. Herunúmen, Manwe, also Númenorean royal name); see also BAR. Often assumed to be an import of Q. héru which however would produce Ad. hîru.



W. laban bag;
- surname Labingi Baggins. Evidently containing a strong declension while according to TT17 the proper pl. was Labanim: Does an Ad. pl. labîna correspond to it?
- Laban-neg Bag-end< NIG.


Ad. laukh- > lôkha or lôkhi crooked, pl. lôkhî, sometimes assumed to derive from E. lok- bend, loop (S) but not following the established patterns of E. imports into Ad.


W. luthur > luthran down, fluff. It is not clear what grammatical form belongs to luthran.



Ad. –mâ;
- saibêth-mâ with assent < BITH.


Ad. mag- build (or from < E. MAG- use, handle?);
- magân builder;
- Balkumagân Shipwright < BAL'K.


Ad. ammî > ammê > W. emma "mother". Q. ammë likely bears only an outside resemblance. Such words are unlikely to be adopted from alien tongues. See TAT.


Ad. imar. Uninterpretable element of Imâr, Imrahil, Imrazôr.


Ad. m. mîk, f. mîth baby, small; d. miyât twins.


Arch. R. maur- wise, experienced.
> h. W. Maura. No equivalent in W.



a-Ad. nadroth (more properly nadrôth?) < RUTH.


N. M. nahald secret.


A base of which Lowdham gives many variations without ever telling its meaning.


come, approach.
Ad. p. t. nakkha; ao. pl. nâkhim are at hand.
- m. sing. unakkha he came,
- n. pl. yanâkhim they are at hand.


Arch. W. narag (LP) > narak (narac) dwarf (TT17). Pl. nargian > naragin.
- arch. Phûrunargian Dwarrowdelf (= S. Hadhodrond) < PHUR.


Ad. narak eagle; n. pl. narîka.
- Narîka n'Bârîm n'Adûn
The Eagles of the Lord of the West. Note that narîka is a n. declension though describing a living being narak should be a c. noun. The intention is here however figurative.


end. Ad. naig- > nêg- > W. neg;
- Laban-neg Bag-end < LAB'N.


Ad. nimir-; Nimîr Elf; pl. Nimrî, see EL-;
- with obj.: Nimruzîr Elf-friend (= Q. Elendil) < ZIR;
- Nimriyai > Nimriyê Elvish language (= Quenya), the suff. may be a direct analogon to Q. -ya;
- subj. m. (Towers of) Nimrûn, a location on Tol Eressëa.
?B. nóm wisdom;
- Nómin the Wise Ones (= Elves). See ZIR.


Ad. obj. pron. nain > nên us; we.


Ad. nit- to kindle, f. nitîr kindler. M. nitân, nitur?
- Gimilnitîr, Star-kindler (= S. Elbereth) < GIMIL.


Ad. nithil, see also PHIL.


W. onna father (TT17), probably onomatopoeic. See TAT.


Ad. nîph, nûph fool, weak noun that was not basic; nuphâr, d. nuphrât, parents < AR- (lit. "fool-king(s)" - evidently, old Lowdham snatched some Adûnaic youth slang).


Ad. nûlu night (evil); nâlu shadow, n. nîlu, m. Nîlu moon > in later languages Nil, Njûl.
- agannâlu shadow of death, Subj. agannâlô < GAN.
The derivation is problematic. In meaning all three words are clearly related to each other; in k-Ad., nâlu even was still nûlu. With a shift of the Constitutive Vowel towards the end of the base Lowdham mentions only the addition of initial vowels to a form -NLU, thus anlu, inlu, etc. But he does not specifically exclude the insertion of a median vowel, either, so establishing such connexions may yet be possible.



Ad. *pa3a > , pl. pâi.


breath, blow.
Ad. pûh breath; pûhtâ- > W. pût- to blow (TT17)
- Ad. pûhtân > h. W. pûta blower;
- putat, putan-t blow, breeze? (TT17),
- raspûta, horn-blower < RAS-.



W. phârë language;
- Soval Phârë Common Speech < SAW.


Ad. adj. suff. -phal;
êphal far, êphalak far away < YA3.
?> W. -val
- sôval common < SAW.


Ad. aphan bliss;
- Aphanuzîr Friend of bliss (= Q. Amandil).


Ad. pharaz gold (WS);
- pharazôn, golden one.


Ad. m. phazân prince, king's son; phazag, phazgân ruler?;
- kâthuphazgân, conqueror < KATH.
Phazân and phazgân are clearly related though it is not obvious how. Does the suff. -ag express a difference in tense so that phazgân = current ruler, phazân = futur ruler?


Ad. phail > phel (properly phêl?) daughter, see NITH'L;
- Zimraphel Jewel-daughter (= Q. Miriel).


a base implying a downward movement?
Ad phursâ- to flow, gush;
- du-phursâ, "so as to gush, should gush".
arch. W. phûr-, phûrun to delve; phûru delf;
- Phûrunargian
, Dwarrowdelf < NAR'G.



Ad. n. raba dog, m. rabau > rabô, f. rabai > rabê.


Ad. bar-rad, barrad borderland, march < BAR,
W. brand(a).
- Branda-nîn Marchbourn, Brandywine, see NEN- (LP).
- Brandagamba Marchbuck < GAB
R. rad.
- Loho-rad > Lôg-rad, Lôgrad Horse-march; Rohan < LUH


Ad. rahat break; p. t. pl. rahtam.
- m. pl. yurahtam they broke, they rent.


Ad. râma man, see NERE-;
- Aglarrâma,
Glorious man? (= Q. Alcarondas; AK) Contrary to CT's statements, neither Aglarrâma nor Alcarondas can mean "castle of the sea", and this epithet is clearly written in small caps in DA, so not intended to be a proper name. Alcarondas rather means "Dome of Glory". Ad. -râma may here also be adopted from E. ráme wing, cf. Eärrámë (AE), thus Aglarrâma Sails of Glory.
W. ram(a) man > h. W. ran; in TT17 also râma, m. pp. râ- he, râmi male, râmava man-like. It is actually not clear whether râma is made from the pp. plus an affix or whether the pp. eroded out of the noun. See NERE-.
- hloth-ram(a) Cotman, cottager (LP) > Lothran, a surname not identical to the village name Hlothran < hlotho, RAN.
- râgamba he-goat, buck < GAB (TT17).


W. rân a small group of dwellings on a hill-side; ran(u), home, ham (LP);
- hloth-ran(a), lothran cotton (cot-town, not a plant; LP) < LUTH.


W. râph(a) burr;
- Zilbirâpha Butterbur< ZIL'B.


Ad. razân > W. raza stranger; razan foreign;
- Razanur (a legendary character), wanderer; Peregrin.


h. W. razar small, red apple.


B. rêda heir (PR).


Ad. rûkh- to shout.


a-Ad. urush fenland. In d-Ad. an impossible word.
- Agathurush Fenland of Shadow (= S. Gwathló). Probably agath-urush rather than with obj. agathu-rush since the fen cannot have an effect on the shade but only vice versa.


scar, score, furrow; a boat’s track (PR).
a-Ad. rauth > roth cutting, foam;
- aubrauth > obroth fore-cutting, curling water at the prow < ?UB;
- nadroth hind-track, wake < NAD;
- Rauthinjil > Rothinzil, Rôthinzil Foam-flower.



Ad. asdi uninterpretable word.


Ad. sakal shore;
- Sakalthôr Son of the Shore (= Q. Falassion (LE)) < THAW'R.


Ad. p. t. saphda > saptha;
*saphdân > sapthân
(often pronounced saftân) "understander", wise man, wizard
An arch. W. derivative of Sapthân (Saphta, etc.) may be Saruman's original name and could explain why he used the initial S-.


Ad. satta two, d. suff. -at. > arch W. -oth.
Continental Ad. hôl room > d. *hôlat > holoth > h. W. hlotho dwelling containing two rooms.
- hlothram(a) > h.W. lothram cotman, cottager < RAM;
- hlothran > h.W. lothran cotton, also < lothram.


Ad. sai, uninterpretable element in saibêth assent < BITH.



Ad. , a pref. setting a subsequent particle into the past
- tâidô once (with îdô now) < DAW.


Ad. tabad- to touch, ao. tabda,
kîtabdahê you touch me.


Ad. tamar smith > W. dramar.
Is sometimes suspected to originate in E. TAM-, cf. Tamar, an early name of Aulë.


man, human (as distinct from animals).
B. and H. atan > Q. Atani man > S. Adan (DM).
> Ad. n. anâ, m. anû, f. anî. Maybe m. suff. -ân -or, -er;
- Balkumagân
Shipwright < BAL'K, MAG.
- Kâthuphazgân Conqueror < KATH, PHAZ.
- Phazân Prince, King's Son < PHAZ.
> W. -a.
- pûta blower < PUH; raspûta horn-blower < RAS
- bêthân > batta speaker < BITH; Barabatta Quicktalker < BAR.
R. -an,
dûkan dweller < DUK; kud-dûkan Hole-dweller < KUD.


W. tapuk (tapuc) rabbit, coney.


creep (through a hole).
R. trahan to burrow (LP).
N. M. p. part. trahald burrowing, worming in.
D. trâgu a dragon's name [translated Smaug from an O.E. verb smugan. Called "a philologist's joke", but the translator was maybe not aware that Slavic zmaj, zmej serpent, dragon, is closely related to this verb].
h. W. trân smial, burrow, large tunnel (TT17).


Ad. târik pillar (from P. E. târâ lofty?);
- Minul-târik
Pillar of Heaven < MINIL;
- Abattarîk Pillar of Earth < BAN (LE).


Ad. attû, attô father. See NON
The onomatopoeic resemblance to Q. atar > atto is probably accidental. By any means, expressions of that kind are rarely adopted from foreign languages. See also MAM.


Ad. tud (DM) watch, guard;
> arch. W. tudnas, coll. tun(n)as body of guards < NAS.


D. > h. W. tung great.
- Tungat Stoor.


W. tûph roof (TT17).


arch. R. tûr folk. May have preserved a root that in Ad. was replaced by LAI-.
- Lohtûr Éothéod < LOH,
- tûrak king (with m. suff. -ak, cf. Ad. mîk boy), also translated Théoden.



Ad. thâni land, realm;
- Amân-thâni > Amâtthâni Land/Realm of Amân.


Ad. tharan four.
W. tharantîn fourth part, quarter, tharni fourth part of a kastar (monetary unit; see KAS'T), > arch. h. W. tharni farthing of the Shire.


Ad. suff. -thôr son of
- Sakalthôr Son of the Shore (= Q. Falassion (LE)) < SAK'L.



Part of the base that produced Ad. f. pl. pref ya-.


Ad. yad- go, p. t. yadda.
- n. sing. ayadda it went.


Ad. ay + -phal > aiphal > êphal far < PHAL
- êphalak far away. See also SAW.


Ad. or yôz gift
- Yôzayân Land of Gift < ZAY'N.


Part of the base that produced Ad. m. pl. pref. yu-.



Ad. zabath- to humble; part. zabathân.
Concluding from Q., Aleš Bičan suggests a pref. *za- down + verbal stem bath-, arguing that otherwise the participle ought to be *zabthân.


Ad. zadan house.


Ad. zagar sword
- a-Ad. Belzagar Light-sword (= Q. Calmacil) < BEL
- Gimilzagar Star-sword < GIM'L;
- azgarâ- warring, waging war, ao. azagrâra, p. t. azaggara;
- azgar swordsman, warrior? [cf. Kh. Azaghal, O. Azog]; see also NAR.


W. zâr(a)- old.
h. W. Zâra-Tôbi Old Toby (a pipe-weed brand),
- Zaragamba Oldbuck (LP) < GAB.
?> O. shar- old;
- sharkû
old man; see also ZIR.


Ad. zawar > zaur > zôr foam.
- Gimilzôr Star-Foam (= Q. Elros); see also ZUR.
- Imrazôr.


Ad. -zê at.
Not present in W. Ranugad Stay-at-home < RAN, GAD.


Ad. zâyan land;
- azra-zâyîn > azra-zâin
- Yôzâyan,
Land of Gift (= Q. Andor).
- sûzâyan province of Númenor?
W. sûza province, sphere of occupation, division of a realm (GN);
- Sûza-t, Sûzat, the Shire.


Ad. izindi straight;
- with obj. izindu-bêth true-sayer, prophet < BITH
- zidar poise.


Ad. zigira wise, zigûr wizard; Zigûr (= Sauron, "his own name", but yet Ad.?)
> W. zihir(a) > zîr(a) wise, zîranda wiser, zîrôna wisest, zîrant wise one, zîrandas wisest one (TT17);
- Banazîr
simple, half-wise, Samwise < BAN (LP).
From Zigûr, Sauron's "own name", a hypothetical B. S. form Shirkû wizard may have derived and applied to Saruman; on that O. Sharkû Old Man may be a pun. Also possible is that B. S. -kû < Kh. kûn man, cf. Kh. Tharkûn Staff-man, Gandalf.


W. zilib, zilbe (TT17), zilbi-,butter.
- Zilbirâpha Butterbur < RAPH.


H. zimra jewel.
- Zimrahin (QS) jewel-child < KHIN.
Ad. zimra.
- zimrathâ-
gather jewels, Zimrathôn Jewel-gatherer (= Q. Hostamir). The suff. -ôn is probably the same as in Pharazôn Golden One < PHAR'Z.
- Zimraphel (= Q. Miriel) < PHIL, see also KAS'T.


Ad. zini > zinî; see also KAL.


love, desire.
Ad. zîr love; zir- to love, part. zirân (the) beloved
- izrayi > izrêyî > izrêi > izrê sweetheart;
- zâira > zaira > zâir yearning;
- nimruzîr
elf-lover, elf-friend < NIM'R; see BEL.,


Ad. zauri > zôrî nurse.


Ad. azûl east.
- azûlada eastward. But see DUN-.


a-Ad. zaur > zôr flame;
- Gimilzôr Silver flame (= Q. Telemnar) (LP).
The name Gimilzôr is also used to translate Elros, but of an entirely different meaning, see GIM'L., ZAW'R. Ar-Gimilzôr may in fact have conceived himself as "Elros II".




Ad. m. âru; f. âri, in proper names Ar-.
- Ârun (name of Melkor as applied by Sauron; the Subj. sing. of âru)
? nuphâr parent < NUPH.
S. > W. m. aran, f. arantha, aranthe. Replaced the original Atani root TUR?


Ad. Avalê goddess, Avalôi Powers, gods
- avalôiyada against the Powers.
- Avallôni Tol Erëssea (< Q. Avallónë)


Ad. Avradî Varda, adopted from Q. with fem. suff. added.


road, path.
Ad. batân, pl. batîna.


Q. lómë > Ad. lômi (without evil connotations).


star, Elf.
Ad. coll. pl. Eledâim; see NIM'R.


W. kali merry, jolly, gay (LP).


W. -kil Man;
- banakil halfling (LP, lit. "half-Man", cf. Latin semivir) < BAN
- Q. tarakhil High Man > W. tarkil Númenorean (LP) > O. tark (Gondorian) Man.


child. P. E. khínâ > Q. hîn.
> Ad. hîn ; patr. suff. -ôhîn;
- Êru-hînim Children of Iluvatar.
- Azrubêlôhîn Child of Azrubêl (= Q. Eärendilion).


Av. lai (QE) > Ad. lâi (cf. Q. lië);
- kadar-lâi city-folk < KAD'R
> B. S. > O. hai;
- Uruk-hai
- Olog-hai
- Oghôr-hai


holy spirit.
Q. manu > Ad. *manaw- > *manau > manô spirit, pl. manôyi > manôi.


He who arises in might. Ad. Mulkhêr, Arûn-Mulkhêr Lord of Darkness, probably only with regard to Melkor; otherwise, "Lord of Darkness" would have been Dolgubâr, in a-Ad. Dolgukhôr.


Ancient Qe. -mas -ton, -by > W. bas -wick, -wich (LP)?
- Galabas Gamwich < GAL'B.


heaven, sky (by Lowdham given as an Ad. base) > Q. menel.
Ad. minal; according to Lowdham from Av., not from Q.
- Minul-Târik Pillar of Heaven; see TAR'K;
- Arminalêth, Ar-Minalêth Royal Heavenly City, cf. Q. Armenelos;
- Minal-zidar, Poise in Heaven.


Ad. c. urug, m. urgu, f. urgî. Continental Ad. perhaps m. urugu, hence > W. ruccu, m. râruccu, f. lêruccu (TT17).


Q. nossë or S. nos kindred, family > Ad. nas people.
> W. tudnas > tunnas body of guard (DM) < TUD.


S. dún > Ad. adûn west, westward (rather than adûnada? See ZUL). Adûn, once erroneously Adûni the West.
- anadûni [< an- of] western, Anadûnê Westernesse.
- c. Adûna Númenorean, subj. pl. Adûnaim.
- Adûnayê Adûnaic, the Númenorean language > W. Adûni Westron (AL),
- Inziladûn Flower of the West, for its grammatical problems see JIL,
- with subj. Adûnakhôr Lord of the West < KHUR.


W. nîn water, "bourn".


man. The potential origin of Ad. naru > narû man (male).
c. narad, m. nardû, f. nardî soldier?
?> W. râ- he, see RAM.


height, mountain [< ORO, to rise < RÔ]
Ad. urud mountain, pl. urîd.


Ad. êth [< C. E. ostô, S. os]? > Arminalêth (= Q. Armenelos). More likely, -êth is simply the f. suff. that was also found in Akallabêth and Q. -os not represented at all.


W. ras (< Q. rassë or S. ras, but in W. only musical);
- raspûta horn-blower < PUH.


orc, goblin.
Ad. uruk, urukh. > W. ork (orc) (H)
Drú. gorgûn.
S. orch.
Kh. rukhs.
B. S. uruk > O. uruk.


high, lofty, noble.
- Q. *tarakhil high Man > W. tarkil Númenorean (LP) > O. tark (Gondorian) Man.


stand up, straight.
Q. tyulma > Ad. sulum mast.


Ad. n. ûrai > ûrê, f. Ûrî Sun > later languages Uir, Ýr.
Often grouped as Sun and Moon:
1. ûriyat [< Ûri + dual suff.];
2. ûrinîl(uw)at [< Ûri + Nîlu + dual suff.];
3. ûriyat nîlo [< Ûri + dual suff. + nîlo].


S. gwath shadow > Ad. agath shade (as opposed to sunlight),
- a-Ad. Agathurush Fenland of Shadow (= S. Gwathló) (GC) < RUS.



Kh. khibil silver > S. cheleb,
Ad. adj. gimil silver
- Gimilzôr Silver flame (= Q. Telemnar) < ZUR (LP).


body of water,
Kh. zâram lake, mere
Ad. AZRA > azra sea, pl. subj. azrîya, coll. azar,
- Azrubêl (= Q. Eärendil) Venus, the Morning-star < BEL.
PR gives both a-Ad. and B. azra star, translating S. El-. See GIM'L.



Drú. drughu Wose (TD)
arch. Hal. drûg.
Q. Rúatan.
S. drú > W. Drú, drúadan (with S. adan Man).
R. róg, pl, rógin.
O. oghôr.



elephant. In Gondorian W., otherwise "oliphaunt", maybe from P. E. andambundâ > Ad. annabûn?

Miscellaneous vocabulary


Ad. glorious < Q. alcar, S. aglar?
- Aglarrâma (= Q. Alcarondas, AK). See RAM.


W. horrible (CE). Being a W. word, this should read balk.
- Balchoth Horrible Horde (with S. hoth). It gets sometimes related to S. balch cruel [< E. ÑGWAL torment]. But if that was true it is not visible why the first element should be identified as being W. at all.


h. W. bulge.


h. W. surname of unclear origin.


B. stone. It is unknown whether this had an equivalent in Ad. - bur?


W., definite form of hamanulla unidentified kind of flower ("Lobelia"). This is probably a compound of ham(a) + nulla, the latter maybe being the W. derivative of Ad. nulu dark.


to be, only known from W. (TT17)


Ad. uninterpretable personal name.


W. rune. R. šird > skirdit;
- Taliskan skirditaila runic series > W. iskirttêl, runic alphabet (with têl series; alphabet).
This entry from TT17 seems highly doubtful, particularly with the never-seen-before accented š (rather than sh) in Rohirian and the inclusion of Taliskan that belongs to an entirely different phase of the conception of the Atani languages. Very probably, the languages have been wrongly applied here.


W. Cloven Valley; Rivendell. The name cannot be certainly separated into components. Rather tentatively, if the first element was a participle it may derive from Ad. karan- cleave, part. karanân > karnin + gul valley.


W. byrding, one celebrating a birthday (L290). Evidently containing a fossilised form of Ad. -ân, W. -a. The remainder is uninterpretable (BAD'Y???).


trump < Q. tyúla, base unknown. So Elves liked to play cards?!?


uninterpretable h. W. name. According to an "unfounded" guess related to a (probably fictitious) p. pt. tûka (tûca) daring.


Ad. uninterpretable personal name.


List of Abbreviations