Dwarvish #8

Ansaru kitnul, ifridî bekâr!
Centre company, ready weapons!

ansar: company, group; from the root √n-s-r “come together, form a group”; with the construct suffix -u
kitnul: adjectival derivative of kitin, from the root √k-t-n middle, central
ifridî: plural imperative from the root √f-r-d “make ready”
bekâr: plural of bekar “weapon” from the root √b-k-r “fight”

Îmî, îmî, kabâru drekh!
“Scram, scat! you mangy animals!”
îmî: imperative plural from the root √y-m (*iymî > îmî)
kabâr: plural of kobor “beast”; with construct suffix -u
drekh: some sort of skin disease

Various single words:
kharâm “brothers” (plural of khurm)
udâmai “comrades” (plural of udmai) — root √d-m-y “accompany, go along with”
umâral “friends” (plural of umral) — root √m-r-l “love”
itkit “shut up!” imperative of √t-k-t
yêbith “spider” — from the root √y-b-th “weave”
kud? “what?”
ugrad “coward”, pl. ugârad (root √g-r-d “fear”)

Dwarvish #7

Risrithî ‘t-tebud
“Burn the corks”
risrithî: imperative plural of √r-s-r-th, causative (with -th suffix) from √r-s-r “burn”: “cause to burn” = “burn (something)”
it-tebud: id-, definite accusative prefix, assimilated to tebud, plural of tebad “stopper,” using the CeCaC pattern often used for words for tools, from a root √t-b-d “block, stop up.”

ekûn lu zayara
“There’s one missing”
ekûn: one person (√ʔ-y-k “one” + -ûn suffix of persons); should be êkûn
lu: not
zayara: 3ms. perfect of root √z-y-r “be present, be here”
one-person not is-here

Nê kikûn inthir!
“Never forget”
: “don’t”
kikûn: “ever, at any time”
inthir: imperative singular of root √n-th-r “forget”

The following was written for a scene involving Thorin, Balin, and a guard. I do not know if it appeared in the film or not.

Zimrith ib-bekan!
“Sound the alarm!”
zimrith: “sound, cause to sound”; imperative from the causative root √z-m-r-th, expanded from the root √z-m-r “make a noise”
ib-bekan: ib- accusative definite suffix id-, assimilated to following b-; bekan “alarm,” noun for a tool that “wakes” (√b-k-n) one.

“My lord?”
uzbad “lord” (agent noun from √z-b-d “rule”) + the 1sg. possessive suffix

Inkhith id-utrâd — igritu zû!
“Summon the guard! Do it now!”
inkhith: imperative from √n-kh-th, causative of √n-kh “come”: “cause to come” > “summon”
id-utrâd: id- accusative definite suffix; utrâd plural of utrad “a guard, a watchman” or “one who watches,” agent form of √t-r-d “watch.”
igritu: igri imperative from √g-r “do, act, perform, accomplish” + 3sg.suffix -tu (should be -hu)
: now

Kud tâti?
“What is it?”
Kud: what? (interrogative)
tâti: it is, 3sm. perfect of the root √ʔ-t: *ta-ʔt-i > tâti.

“A dragon!”
Obviously a loan from an Avarin or Nandorin word, with the Eldarin root √slok- (cf. Quenya hlókë).

Dwarvish #5

After those phrases, there follows a list of much shorter phrases and words, none more than two words long:

Insid, pl. insidî
“Sit down!”
Imperative from the root √n-s-d.

inkhir, pl. inkhirî
“Come away”
Imperative from the root √n-kh-r, which is an extension of the biliteral root √n-kh “come.”

ithmir, pl. ithmirî
“Get away”
Imperative from the root √th-m-r “leave, retreat, remove (from).”

ithmir b’tîr
“Get away from there”
bi: preposition “from, away from” (a location at or nearby something, not from inside it)
tîr: “there, that place” — usually of a place nearby or within reach; cf. yîr “there yonder” (sc. in the distance, though still visible) and kûr “where?”

idribtu, pl. idribîtu
“stop it”
Imperative of root √d-r-b, with 3s suffix -tu; as mentioned, this should now be -hu.

ithrik, pl. ithrikî
Imperative of root √th-r-k “hold steady, hold up, support”

therek ikhlit, pl. therkâ ikhlitî
“Hold firm”
therek, pl. therkâ: “firm, fast, steady,” adjective from the root √th-r-k “hold steady.” The ending here is an adjectival plural; therkâ is a syncope of *therekâ, vowels in open medial syllables being prone to syncope.
ikhlit, pl. ikhlitî: imperative from the root √kh-l-t “hold, hold tight, maintain”

sâti khuzd
“You are a dwarf” (a statement of vivid, current fact)
sâti “you (sg. m.) are,” imperfect 2sg.m. from the root √ʔ-t “to be”: *sa-ʔt-i > sâti.

îridzu du-khuzd
“You are a dwarf” — literally “Know yourself for a dwarf”
îridzu: îrid, imperative sg. m. of √y-r-d “know” (*iyrid > îrid) + suffix –zu 2sg. m. polite suffix
du “to, for”
khuzd “dwarf”

ashnakh: treason (root √sh-n-kh “betray”)
khurm: brother (nominal root √kh-r-m “brother”)
umral: (close) friend (root √m-r-l “love”)
udmay: comrade (root √d-m-y “accompany, go along with”)

Dwarvish #4

Ikhf’ id-ursu khazâd
“Feel the fire of the dwarves”
ikhfi: imperative of √kh-f, “receive, accept,” elided to ikhf’ before another word beginning with i-
id-ursu: “the fire (of)” the noun urus “fire” with a definite accusative prefix id- and a connecting (construct) suffix -u.
khazâd: “dwarves,” plural of khuzd

Igribî ‘b-bekâr d’zun
“Arm yourselves”
igribî: imperative plural of √g-r-b “take, seize”
ib-bekâr: “the weapons,” elided to ‘b-bekâr after a long î. Bekâr is the plural of bekar “weapon,” but the plural is more often used. Ib is the same definite accusative prefix as id-, but assimilated to the following consonant.
d’zun: contracted from du-zun “for yourselves” (preposition du “to, for”, -zun “you plural”).

“To arms!”
du “to,” bekâr “arms”

Gelekh d’ashrud bark
“Time to swing an axe”
gelekh: “time, occasion” from the root √g-l-kh “happen, occur (punctually)”
d’ashrud: du “for” + ashrud, gerund (or infinitive) of the verbal root √sh-r-d “wield, control.” On second thought I wonder if this should have been ashrudu, part of a construct formation with bark: “for the wielding of an axe.”
bark: axe

M’imnu Durin
“In Durin’s name”
mi: “by, with (some instrument)”; elided to m’ before another word starting with i-
imn: “name”; construct form imnu
Durin: proper name, in Mannish form, of the progenitor of the Dwarves; his true name would not be used above ground, or where non-Dwarves could hear it.

Continuing Dwarvish

Lu kalzatha bark
“Couldn’t lift an axe”
lu: “not”
kalzatha: root √ʔ-l-z “rise” > causative √ʔ-l-z-th “cause to rise, lift, raise” > ka- prefix indicating ability or potentiality + alzatha perfect 3ms. “he lifts/lifted” (as a matter of general fact, rather than an ongoing event)
bark: “axe”
“He has/had no ability to cause the axe to rise” = “Couldn’t lift an axe.”

Nê ikrid ûdar!
“Never trust a wizard”
: “don’t” — a negative particle used with injunctions or other non-real expressions. Lu negates things that are happening or have happened; negates hypotheticals, things that would happen or might happen or haven’t happened yet.
ikrid: imperative singular of √k-r-d “believe, trust”
ûdar: “wizard,” literally “knower,” from the root √y-d-r “know, be wise” placed in the uCCaC pattern. *uydar > *uwdar > ûdar; or perhaps the historically original root was √w-d-r after all.

Imrid amrad ursul!
“Die a death of flames”
imrid: imperative sg. of √m-r-d “die”
amrad: abstract noun from the same root
ursul: adjective “of flames, flaming, fiery” with -ul suffix added to urus “flame, fire,” with syncope of the stem (urus > urs-). The root is √ʔ-r-s “fire, burn”

Urus d’zun!
“Fire upon you!”
d’zun: contracted from duzun (stressed on the second syllable), from the preposition du “to, for” + the suffix –zun “you (masculine plural).”
I’m not sure of the context here, but most likely this should be dumên, not duzun, as the Dwarves tended to refer to their enemies using a (contemptuous) familiar form.

“Take that!”
ikhfitu: imperative ikhfi from the root √kh-f “take, receive” + 3ms. suffix –tu. As noted in a previous post, the neuter suffix –hu had not been invented at this stage, and should really be used instead of –tu here: Ikhfihu!

Of various roots which could be translated “take,” √kh-f means “accept something given or dealt to one” (not necessarily something beneficial) and √g-r-b means “grasp or seize,” often, though not necessarily, with the implication that the thing taken is in another’s possession, and is relinquished unwillingly.

Singular imperative of the simple (intransitive) verb from the √ʔ-r-s root.

Further Dwarvish

Some more material prepared for The Hobbit, from the early stages of production. I present the forms exactly as I first wrote them, with some suggested emendations based on later developments.

Lu akraditu!
“I don’t believe it!”
lu: “not”, a general negating particle
akraditu: root √krd “believe, trust”; akradi “I believe”; -tu 3ms. suffix
Obviously that should really mean “I don’t believe him.” Originally — at the time I wrote this line — I didn’t have a masculine/neuter contrast, but at a later date I added the neuter suffix -hu, which would be more correct -hu: akradihu. I don’t know if this line ever actually was used in the film.

Smaug mamarda
“Smaug is dead.”
mamarda: root √mrd “die”; past participle mamard, used as a stem to which perfect endings (in this case -a, the 3sm.) are added.

Anthân lu sharagên
“Omens do not lie”
anthân: “sign, omen” a feminine noun that I intended to be both singular and plural. However, going by similar patterns I used later, it should have been anathân as a plural. The root is √nthn “point out.”
sharagên: root √shrg “to lie, to say a falsehood”, perfect stem with 3pf. ending -ên.

Karâk Urdekul
“Ravens of Erebor”
kark, pl. karâk: “raven:
urdekul: genitive/adjectival form formed by adding -ul to the name Urdek “Lonely Mountain” = urd “mountain” + êk, shortened form of ayik “alone, single, lonely.” I should have written Urdêk, Urdêkul.

Mafarrakh d’afrukh
“A burden to carry”
mafarrakh: habitual past participle of √frkh “carry,” here used as a noun: “thing habitually carried” > burden.
du: “to, for (the purpose of)”; here elided to d’ before a word beginning with a vowel.
afrukh: gerund “carrying” from √frkh

Lu mafrad d’abkâr
“Not fit for a fight”
mafrad: “prepared, ready” from the root √frd “prepare, make ready.” This is a different participial form, indicating some present state, so literally “being (now) prepared.”
abkâr: “fight, strife, battle” from √bkr “fight”; the word abkâr “a fight,” delimited in space and time, can be distinguished from the more abstract gerund abkur “fighting.”

Questions and Answers

I get a lot of questions about particular lines of dialogue in The Hobbit films, requests to translate and so forth. I would like to comply, but unfortunately I don’t have a complete script of all the films, or even DVDs, and even under the best of conditions it would be difficult to figure out which of the lines I contributed were actually used, and if so, where.

So I’m going to start doing the next best thing. I will start recording all of the lines I wrote for the film on this blog, with some analysis, though I may have to trade off thoroughness for quantity.

So here’s a start, with some of the Dwarvish lines, since these tended to be earlier and may, I guess, be a little more interesting than some of the other languages.

“Ready yourselves!”
Root √frd “prepare, make ready”
Imperative 2pm. ifridî
Pronominal suffix 2pm. -zun
Hence “you prepare yourselves”

Urâd Zirnul
“Iron Hills”
Root √ʔrd, singular urd “hill”, pl. urâd
Root √zrn, zirin “iron” (the metal) + adjectival ending -ul, with syncope of zirin > zirn.
I think this may never have been used in the films, and if so, might be considered slightly apocryphal. I would probably think twice about using the same word for both Erebor and the Emyn Engrin. Unless, from the Dwarvish perspective, the term relates not to the size of the rock visible above ground, but the extent of the caverns delved out underground.

Akkâ Belkulu Dain-Uzbad
“Lord Dáin’s Mighty Force”
Root √kʔ “have power” adapted to the (fairly common) aCCâC abstract noun pattern; here it appears that the glottal stop assimilates to a preceding k, i.e. *akʔâ (or perhaps *akʔâʔ; I can’t find any counter-examples) > akkâ.
Root √blk “be mighty, be strong” > belk “might, strength, power” + -ul > belkul “mighty, of might” + -u object suffix, as the “mighty force” is the object of Dáin’s azbâd — i.e., that thing which he rules or governs.

Nominal root √ʔfth “foot” > ifth “foot.” The following -u is not the objective ending, but rather a (rarely seen) construct ending which links it to a following noun taken as a genitive or attributive.
Ifthu-zirin = “foot-of-iron.”

Khuzd belkul
“A mighty dwarf”
Khuzd “dwarf,” belkul “mighty” (as above).

Til hamingju með afmælit

On this, the 122nd anniversary of the birth of John Ronald Reuel Tolkien, the Miðgarðsmál blog would like to wish him a

  • Mána ontalérë
  • Aur onnas alwed
  • Ênâd-nurt gêdul

and a

  • Zeyborzi undumi bolneg

mána “blessed”
onta– “produce, create, beget, give birth to”
ontale “production, creation, genealogical descent” — hence (presumably) birth as well
“day” (of 24 hours)
ontale+ré = ontalére, preserving the original long final vowel of *ontālē (cf. yáviére, tuilére)

aur “day” (of 24 hours)
onnas “birth” — a conjectural noun form, from the slightly less conjectural onna-, equivalent to Quenya onta– (cf. edonna– “beget”)
alwed “fortunate, prosperous”

ênâd “birth” from *aynād, from the root √YND “give birth to” (influenced in fact by both Quenya yondo and Semitic √WLD, √YLD).
nurt “24-hour day” — a word from archaic formation, from √NRT “turn”; probably referring, not to the turning of the earth on its axis, but to the apparent turning of the sun around the earth. This root has been in my notes for a while, and I can’t find which word it was originally intended to explain or remember its origin; it looks now like simply an anagram of “turn,” but I may have had something else in mind, possibly Indo-European *wert- . “Turning” itself would be anrât.
gêdul “joyful, happy,” from a noun gayad, gêd- (*gayd-). No doubt Latin gaudium had an influence here.

Orkish (The dialect used in the film of The Hobbit)
zeyborz “day,” literally “light-dark”; zey from more archaic *zil, and borz from Black Speech burz.
The suffix -i marks a noun or noun phrase that is modified by an adjective or another noun. Its origin is probably the same as the Elvish relative pronoun i or ya.
undum “birth” or “spawning” from a verb und– “procreate.” This again seems to show Elvish influence.
The Orcs do not really appreciate the concept of joy, as understood by most other creatures (a literal description of it in Orkish would amount to “madness”), much less blessedness. I was forced to use an approximation of the concept that would make sense to an Orc:
bolneg “free from pain,” from the Orkish root √bol– (cf. bolum “pain”) and the privative suffix –neg, marking an absence of something. The latter is reminiscent of Latin negare; this is a coincidence (as they say in Middle-earth). The actual source is Quendian *-enekā, from the root √nek– “deprive of.”

Kíla steinn

I’ve received an inquiry about the meaning of the runes on Kíli’s talisman stone. The words inscribed on it are innikh dê.

The first is the singular imperative of the verb nanakha “return, come back”, which has a triliteral root √n-n-kh which obviously has been formed from the biliteral root √n-kh “come,” which is in turn clearly related to Adûnaic nakh-. The pattern is iCCiC, as is generally the case with other imperatives.

combines a preposition d(u) “to, toward” (whose real-world inspiration is the Gothic preposition du) with the 1st person singular pronominal suffix .

The meaning of the phrase on the stone is therefore “return to me.” Its precise application in Kíli’s case is something I’m not privy to, and I expect that passionate film fans can guess it more easily than I can.

Durin’s song: The rest

I’m finishing my discussion of Durin’s song with an overview of the remaining words, pointing out some of the sources. By way of comparison, I have added the first (and so far only) “dictionary” of Neokhuzdul that exists, at the link Neo-Khuzdul glossary. There may be some slight discrepancies between the glossary and the list below, but on the whole they seem to be in good agreement. There are also several words in it which are not found below nor, I think, anywhere else.


  • gagin “again” — Proto-Germanic *gagina, which forms part of *anagagina which becomes Old English ongean, Modern English again. Cf. German gegen.


  • ra “and” — I do not know where this came from.


  • ana “to, toward” — evidently also influenced by Elvish an and na.
  • bin “without, lacking” — clearly influenced by Elvish pen “without”, which in turn reflects Greek πένομαι penomai “be poor, have need of” — though I think I had not noticed the connection at the time, or indeed until quite recently.
  • ni “in” — it’s in backwards.
  • tur “through” — influenced by both Elvish ter– and Old English þurh, both “through”. (The former probably shows the influence of Indo-European √terH, which is of course the source of the latter).
  • undu “beneath, under” — Elvish √undu, and English under.


  • ku “who” — suggested by the Indo-European root √kwo-.
  • maku “no one” — literally “no who.” The use of a prefix ma- as a negative is a little obscure, but perhaps was suggested by Greek μή and Sanskrit .
  • tada “that” — suggested by Gothic þata and Indo-European *tod (Sanskrit tat) “that, it.”


  • kalil “cold” — English cold from Proto-Germanic *kalda- and chill from *kaliz.
  • shakar “sharp” — probably simply sound symbolism; I thought that the consonants suggested the sound of something being cut.
  • sullu “all” — suggested by Proto-Indo-European *solwos “whole,” which I knew in the shape of Sanskrit sarva “all.”
  • ubzar “deeper” — from a root √BZR, intended to explain bizar “valley.” The pattern uCCaC for comparatives and superlatives obviously overlaps with uCCaC for agents, but evidently this didn’t bother me. It is not uncommon in natural languages for similar constructions to have two quite different meanings, if they are unlikely to overlap in practice; for instance, in English the ending –er when attached to adjectives is a comparative ending (e.g. tall:taller) but when added to verbs indicates an agent (e.g. speak:speaker). I don’t remember if this comparison occurred to me at the time, however.
  • udlag “very far away” — presumably a superlative, but I don’t know what the base form would have been, perhaps dalig. In any case it is from *dlonghos, the Indo-European basis of both English long and Latin longus.
  • ugmal “eldest” — supposed to be a superlative form, from an adjective gamil “old”. The latter word is actually attested in the name of a Dwarf craftsman called “Gamil Zirak the old” (Unfinished Tales, p. 76) but there is no certainty that it actually means “old.” The meaning is really taken from Old Norse gamall “old”.


  • abad “mountain” — Almost certainly abstracted from Gundabad.
  • aban “stone”, adjectival form abanul “of stone, stony” — Looks like a rare case of direct Semitic influence, Hebrew eben < *abnu.
  • addad “fathers” — Thus in this transcript, though my early glossary has addâd. The singular was dâd, the root √ʔD, both singular and plural being somewhat irregular. Evidently from a kind of baby talk, “da da” — and of course similar to English dad.
  • aklat “sound” — an abstract noun from a root √KLT, suggested by Indo-European *klew- “hear” and *klutos “heard.”
  • amrad “death” — an abstract noun, evidently from a root √MRD “die”, which is obviously suggested by Proto-Indo-European √mer- (as in mortal).
  • arrâs “flames” — see urus below.
  • askad “shadow” — perhaps also originally abstract. Suggested by Proto-Germanic *skadwaz, whence English shade/shadow.
  • atkât “silence” — from a root √TKT, suggested by Latin taceo “be silent.” The pattern is evidently the abstract one I took from aglâb “speech.”
  • aznân “dark, darkness” — from a root √ʔZN, taken from the first part of Azanulbizar (Dimrill Dale). The form azanân was most likely intended as a sort of “broken” plural.
  • bashuk “bones” — a plural apparently founded on the pattern of baruk “axes”. It implies a singular *bashk, which however doesn’t appear in my notes.
  • buzrâ “deep” (sc. deep places, depths). From the same root √BZR in bizar “valley” and ubzar “deeper, very deep.” The word may have been intended originally as a plural, though I am not now sure of that.
  • fill(u) “skin” — Gothic fill “skin, hide” (cf. archaic English “fell,” a flayed animal’s skin).
  • gabil “great” — attested Khuzdul word, from Gabilgathol “Great fortress” and Gabilân “Great river.”
  • galab “word” — from the root √GLB taken from aglâb.
  • ganâd “halls” — a plural, obviously of gund, taken from both Felak-gundu and Gundabad. Hence evidently I understood Gundabad at the time as “Mountain-of-underground hall.”
  • gilim “glint” — suggested by Eldarin √glim-, English gleam, and of course glint itself.
  • iklal “cold” — yet another abstract pattern, from the same root as kalil “cold.” Most languages have a fairly large number of ways of constructing abstract nouns indicating qualities.
  • kâmin “earth” — suggested by Quenya cemen.
  • kilmîn “crown” — Old English helm, from an Indo-European root √kel-. The shape suggests a meaning something more like “helmet-shaped structure.” The crown of Durin depicted on the West-gate of Moria is helmet-like in shape.
  • kurd(u) “heart” — Indo-European *kerd-, whence Greek καρδία kardia, Latin cord-, and Gothic hairtō.
  • lukhud “light” — English light and the related Gothic liuhad-.
  • sanzigil “mithril” — literally “true-silver”. San from Sanskrit san(t-), sat- and Old Norse sannr, both “true” (from an Indo-European word meaning “existing”). This is probably not the secret name of the Dwarves for mithril, but a circumlocution that could be used in public.
  • thatur “stars” — this implies a singular thatr (though I do not see that form anywhere in my notes). The inspiration was English “star” and its Indo-European cognates (most of which, however, contain the stem in the form ster– (e.g. ἀστήρ astēr, stella, stairnō), except for Sanskrit tārā, which I may have been thinking of.
  • ugrûd “fear, dread” — from a root √GRD “fear”, related to various Eldarin words and roots (√ŋgor-, √ŋgur-) suggesting horror or death.
  • ukrat “glory” — Most likely from Old English hréð “glory, fame” < *hrōþiz, whose shape could imply a PIE *krōtis.
  • urus “fire” — Intended to be a direct borrowing from Valarin uruš (also rušur) “fire.” The reason Aulë might have had for changing the š to an s remains inscrutable; my reason was probably that I didn’t want it to look exactly like Valarin. The influence of Eldarin uru– “heat” is also evident. The plural arrâs is along the same lines as addâd, a plural formation that is evidently of my own invention.