The Mad Latinist (jdm314) wrote,
The Mad Latinist

  • Mood:
  • Music:

Do ydran ji Valyre!

Time once again for Game of Thrones conlangs. Today we'll be doing episode 304, "And Now His Watch is Ended." The Valyrian scene in this episode is just incredible. If you want to take a look for yourself, see here. For David J. Peterson's entry on this episode (which, recall, we are calling SUZKO), see here.

Linguistically, one of the most important aspects of this episode was the introduction of High Valyrian, in Daenerys' lines. Mr. Peterson has been a good deal more forthcoming about High Valyrian, and those of us who have been following his blog and participating in the comments  actually know a surprising amount about the language already. There will be a lot of grammatical terminology thrown around in our discussions of High Valyrian. If you need anything defined, you may want to read the quick-and-dirty guide I wrote in my replies to "Dinok" in this thread. In particular, note that High Valyrian is pretty consistently a "Head Final" and "SOV" language... if you don't know what that means, see here, but in short it means that the word-order will be very different from that of English.

In other Valyrian news, Curtain Call: Dan Hildebrand over at contains an extensive composition in Astapori Valyrian by David J. Peterson (let's call it Aeske Hildebrand, or ÆH for short). We now know the language well enough to realize that the English translation might be... slightly less than literal. I'm hoping to write another post on that text later this week, if I get a chance, but in the meantime I'll feel free to use it as evidence for this post.

I am still using underlines to mark lines for which we have Mr. Peterson's official transcription—for this episode it is nearly the whole scene (including English glosses for lines that aren't subtitled). But I have added two new symbols:
  • Parentheses mark lines where the actor has missed or significantly garbled a word given in the transcript. It happens: not everyone is analyzing the gobbledygook as much as we are, so it's not worth the effort to correct every mistake. Furthermore some of them may not be mistakes, but deliberate cuts quicken the pace of the dialog. Still, if we have David J. Peterson himself saying one thing, but the actors saying another, I figure I should note it.

  • To make them easier to distinguish, I have decided to mark High Valyrian (HV) in Targaryen Red, and Astapori Low Valyrian (AV) in Good Master Green.
As I mentioned above, Peterson has given us an official transcription for the entire scene. But we open with a monologue by Kraznys, which he did not transcribe. This was even harder than usual to transcribe, and I suspect some of it was Mr. Hildebrand improvising from other lines.
Kraznys: Prandáski mazmágho engrád. Ji réni silívas pon zoldágho etirí.
Missandei: “The Master says they are untested.”
Prandáski might be a 1pl verb, and mazmágho is definitely an infinitive. But engrád doesn't even sound Valyrian (I suppose it could be a word of Ghiscari origin?).

ETA: mazmágho must mean "to train" or the like. Compare skókido mazméris funmári "how they are trained" from the first post.

The next sentence is a bit easier: Ji réni silívas pon zodágho etirí "The slut (will?) [VERB] to [VERB] them." I have no guess what those verbs are, nor what etirí means.

ETA: Dinok suggests this line corresponds with Missandei's next line, " would be wise to blood them early." Silívas is identical to the selévas in my first post, which I translated "must." Zodágho is what Missandei translates as "blood" (no relation to HV ā́nogar, but then there doesn't have to be). Etirí must then mean "early."

K: Vo Dovoghédi dozézi dojiní va jistizí
M: “He says you would be wise to blood them early”
Obviously Dovoghédi "Unsullied," and va "to, for," but the rest is pretty opaque. The middle two words could possibly be negative verbs (i.e. do zézi, do jiní) in the 1p.

K: Do vazléva [--]  kizí o [--] si zýa
M: “There are many small cities between here and there: cities ripe for sacking.”
Do vazléva: "I will not [VERB]"?

kizí o: "this for you"

si zýa: Si is probably the AV reflex of HV se- "and," and of course zýa means "hers." Or perhaps si could be a form of the copula, in which case si zýa would mean "is hers."

K: Vo Dovoghédi dozézi dojiní va jistizí
M: “Should you take captives, the Masters will buy the healthy ones for a good price.”
This is EXACT same line as the one before last, which is telling.

K: Varukilís va dóli ampá j’etárengez bontáz soslévas [--]  uní  ublélis vítes rovaja.
M: “And, who knows? In ten years, some of those you’ve sent will be the Unsullied in their turn, so all shall prosper.
Now we finally have something that sounds real and possibly analyzable:

Varukilís looks like a 3pl verb, but I don't know what it would mean. It does immediately remind one of evá rughílis "until they collapse," from the first episode, but any connection seems unlikely.

Va "for, to." Ampá could conceivably be the same as the word I previously transcribed aubá, meaning "ten." J'etárengez looks suspiciously similar to j’etadósh "day and night," and j’etóvi "day," which leads me to wonder if it might mean "years." No guess for bontáz, but soslévas looks like a verb of some sort in the future—a logical guess would be that this means "those you will have sent," except that bontáz doesn't look like a demonstrative... perhaps it means "captives" or something?

uní ublélis vítes rovaja: This is probably equivalent to " all shall prosper." We have seen úni "all" before. Ublélis is probably a 3s verb in that "timeless" form. No guess for vítes, but rovaja is, of course "biggest," David J. Peterson's favorite Astapori word, I think. I suspect that here it means "flourishing."

There is at this point a long pause in the scene, while Daenerys and Kraznys exchange their promised items. When we next hear the characters speak, we are finally in the officially transcribed section:
M: Píndas lu sa sir tída.
“She asks if it is now done.”
Píndas lu "she asks if..." Píndas is already well known, but for lu see my last post.

sa sir tída: "... it is now done." We know sir "now" from Píndas sko gómila kízi sír "She asks that you do this now." Sa "is" seems to be from HV íssa, the suppletive 3s form of "to be" (unless it is a pronoun of some sort, standing in for the copula—it certainly seems like a plausible nominative for ze, ji, etc.) Tída looks suspiciously like the word I was translating as "nothing" in my first post... I have not yet figured out what to make of that. But it seems quite plausible that the -ída ending seen here is the same as the -iton in keliton "ended."

K: Sa tída. Pélos ji qlóny. (J’)aspó éza zýa, a zýa azántyr.
“It is done. She holds the whip. The bitch has her army.”
Sa tída: "It is done"

Pélos ji qlóny: "She holds the whip." Farther down we will see that the HV equivalent of this phrase is qilṓni pílos. As I've said elsewhere, some verbs seem to end in -os or -es, and I have yet to figure out what causes this. Najahho points out that qlóny/qilṓni is cognate to HV qilōnario "of punishment" (the latter is probably derived from the former.)

J'aspó éza zýa azántyr: "The bitch has her army." Aspó is, I guess, another new term of abuse. Éza is interesting in that it ends in -a, even though it is apparently 3s; this puts me in mind of kísa evá vanéqo "she has until tomorrow." Zýa is well-attested with the meaning "her." Azántyr "army" deserves special attention for two reasons:
  • In my first post, I argued that azándi meant "(come) forward," and was etymologically connected to azánty "knight." Najahho, on the other hand, suggested that both were the same word, and it simply meant "warrior" (in other words, when Kraznys beckons the Unsullied forward, he is saying  something like "You, soldier!") It looks like Najahho was right: the Unsullied are no knights, after all.

  • azánty comes from HV azántys "soldier," which, in the collective number becomes azántyr "soldiery; army."

Next we get our first ever line in High Valyrian (not counting válar morghū́lis):
Daenerys: Dovaogḗdys! Naéjot memēbātā́s! Kelītī́s!
“Unsullied! Forward march! Halt!”
Dovaogḗdys: "Unsullied." Peterson explicitly calls this form plural (presumably in the vocative case). According to TZ, the lexical form (i.e. nominative singular) is given as Dovaogḗdy.

Naéjot memēbātā́s: "Forward march." Again, Peterson expressly tells us that this is an imperative plural (and throughout this scene we will repeatedly see this -ātā́s ending.) HV commands are apparently always, or nearly always, accented on the last syllable, just as in AV.  As for naéjot "forward," the -ot ending is known to occur in the dative and/or locative of other words.

Kelītī́s: "Halt." Interestingly this word does not have the same form as the other imperative plurals in this scene. Either it is a verb of a different conjugation class, or not a verb at all. If it is the latter, then it's especially interesting that, as a command, it still gets final stress.

K: Ivetrá, ivetrá j’áspo (zya) dyní do majís.
“Tell the bitch her beast won’t come.”
The only new things here are dyní "beast," and majís "come." Again, for majís we would expect *májas, but that's not what we get. Najahho suggests this is a future form, although then we need to explain why it doesn't use that -liva- infix. And there are other examples that seem less future, for example tuzis "smells" in ÆH.

D: Zaldrī́zes buzdári íksos dáor.
“A dragon is not a slave.”
We've already mentioned why AV zaldríze "dragon" is such an interesting word, and here we have the HV equivalent, zaldrī́zes. Buzdári is also interesting in that, Mr. Peterson tells us, it is not a proper HV word, but a borrowing from AV (and ultimately the extinct language Ghiscari), which Dany choses to use for effect. As a result it isn't accented according to the usual rules of AV (which would have put the stress on the first syllable), and it follows a special declension pattern reserved for loanwords. Íksos "is," on which more later. Dáor "not"; notice that in AV there seems to be just one negative, do, which can be used as a prefix (meaning "un-", "non-" etc.),  an adverb (meaning "not") and possibly even an adjective (meaning "no" as in "no weakness"), AV keeps all those usages distinct: the prefix is do-, the adverb is dáor, and the adjective is dṓre (I cannot yet give the lexical form, but that's how it appears later this scene.)

K: Ydrá jí Valýre?
“You speak Valyrian?”
Ydrá "you speak," confirms that -a is the correct ending for 2s.

jí Valýre "Valyrian."

D: Nýke Dáenerys Jelmā́zmo hen Targā́rio Léntrot, hen Valýrio Uḗpo ā́nogār íksan. Valýrio múño ḗngos ñúhys íssa.
“I am Daenerys Stormborn of the House Targaryen, of the blood of Old Valyria. Valyrian is my mother tongue.”
And now things get complicated! We can no longer analyze sentences one word at a time, in the order they appear (even in Latin that would be difficult!) We'll have to do them in the English order instead:

Nýke: "I..." HV is surely a pro-drop language, so Daenerys is explicitly using the pronoun for emphasis.

íksan: "am..." As I mentioned last post, the AV equivalent is skan. Apparently there is a sound change which turns HV ks to AV sk (compare also āeks- aeske "master"). Interestingly, both times she uses a form of this verb, Daenerys seems to pronounce it ísk- instead, anticipating that development. This is probably a simple error on Emilia Clarke's part, but it could actually make sense in-world: Peterson has made it quite clear that the pronunciation of HV tends to be influenced by the local low Valyrian dialect (just as was the case with Latin in Europe). Since Dany was raised in Braavos and Pentos, this would be rather like an Italin pronouncing optimum as ottimum. We'll have to see if Mr. Peterson will put this same sound change  in Braavosi and Pentoshi Valyrian.

Dáenerys Jelmā́zmo: "Daenerys Stormborn...." Notice that in HV Daenerys is stressed on the first syllable.

hen Targā́rio Léntrot: "of the House Targaryen." Hen is translated "of" here, but since a genitive alone suffices for that, my suspicion is that the literal meaning is "from." It is a preposition that takes the locative, making Léntrot the locative of the word for house (at least in this sense of the word)—which may be lentor (interestingly, in the paradigm for the name Aerys Peterson gave, -ot is not a locative ending, but a dative one.) Targā́rio is the genitive of Targā́rien.

hen Valýrio Uḗpo ā́nogār: "of the blood of Old Valyria...."  Ā́nogār "blood," loc.s. the nominative is apparently ā́nogăr. Valýrio and Uḗpo both show that same -o genitive ending (which is interesting  given that adjectives apparently have a much more complicated morphology.)

Valýrio múño ḗngos ñúhys íssa: "Valyrian is my mother tongue." Valýrio here is presumably not the genitive of *Valýria, as in the last sentence, but the nominative of a different word, meaing "The Valyrian language" (AV Valýre.) Íssa "is" (AV sa). Ñýhys "my" (AV nya). Múño ḗngos "mother tongue;" I suspect ḗngos means "language," and múño is the genitive of "mother"—this seems like the sort of expression David Peterson might translate fairly literally.

D: Dovaogḗdys! Āéksia ossēnātā́s, ménti ossēnātā́s, qilṓni pílos lúe vále tólvie ossēnātā́s, yn ríñe (dṓre) ōdrikātā́s. Úrnet lúo (buzdáro tólvio bélma) pryjātā́s!
“Unsullied! Slay the masters, slay the soldiers, slay every man who holds a whip, but harm no child. Strike the chains off every slave you see!”
Āéksia ossēnātā́s: "Slay the masters!"

ménti ossēnātā́s: "Slay the soldiers!" Interestingly we get ménti here, instead of whatever azántys is in the

qilṓni pílos lúe vále tólvie ossēnātā́s: "Slay every man who holds a whip!" Here we see how different Valyrian word-order is from English: it's literally "whip holds who man every slay" (this is why even if I were translating into Latin the word order would still be a problem: Latin, afterall, could not possibly put a relative pronoun at the end of the relative clause: flagellum tenet qui hominem omnem necate is absurd at best.) Vále tólvie "every man" seems to mark its accusative singular with this -e ending. Interestingly enough, lúe, the relative marker, seems to be in the accusative too, and indeed when we later get the same pair of words in the nominative, it's lúa vála. Apparently the relative marker agrees with the head noun, in the form it takes in the main clause, the opposite of Latin (so technically my Latin translation should have said quem, not qui!)

yn ríñe (dṓre) ōdrikātā́s: "but harm no child."

yn: "but..." Cf. Yn vali soty daor "But we are not men."  Note also Dinok's suggestion that the Astapori reflex of this word might be ?i.

ōdrikātā́s: "harm" (

ríñe dṓre: "no child." Notice that ríñe again has that -e acc.s. ending, and more interestingly, so does dṓre, showing that it is indeed an adjective modifying ríñe, and not, say, an adverb. And of course this could also be the etymon of AV do. Emilia Clake happens to skip dṓre... of all the words she could have omitted, that might possibly be the worst!

pryjātā́s: "Strike off!" (

bélma: "the chains"  (presumably

buzdáro tólvio: "of every slave" (gen.s.) This gives us the genitive form of the buzdári declension class.

Úrnet lúo: "that you see." Lúo is in the gen.s. to agree with buzdáro tólvio. Úrnet gives us the 2pl indicative (or at least non-imperative) form.

K: Nyk skan (jiva) aeské! Zer séna! Zer séna!
“I am your master! Kill her! Kill her!”
Nyk skan jiva aeské: "I am your master!" Nyk "I" (HV nýke.) Skan "am" (HV íksan.) Jiva "your" (2 pl. cf. jévo). At first I assumed Kraznys was addressing Drogon—after all he knows better than anyone that the Unsullied will not listen to him, and he just bought the dragon fair and square. Consequently I couldn't figure out why he would use jiva, rather than oa (as attested in ÆH). Eventually Najahho convinced me that he could only be addressing the Unsullied, at whom he is, after all, looking, and backed it up with quotes from the book ("Unsullied, defend your masters!" ... "Defend us, defend your masters!" aSoS, p. 380) Aeské: master (HV āéksia, at least in the Notice that in my very first post, I misheard this word as ˣáse; John Sheehy's ˣaise was closer.

Zer séna: "Kill her!" (HV probably something like zíry ossēnātā́s) Séna (I don't know why it's not stressed on the last syllable) is presumably the imperative of the word I transcribed seneghó (and took to mean "kill) in our first post. Notice that AV apparently does not distinguish between singular and plural commands.

D: Drăkárys
"Dragonfire" (spelled dracarys in the books.) One of the only words of HV we knew before David J. Peterson got his hands on it! Many fans of the book gathered from this word that drac or draca was the HV word for "dragon," and arys or even just rys was the word for "fire." Peterson did not like this approach, and came up with totally different words for dragon (zaldrī́zes) and fire (pérzys), deeming that drakárys was its own discrete word, and not transparently analyzable. This has proven very controversial, and a significant portion of the comments at lately have been from people asking about this.

D: Dovaogḗdys, jévo gláeso(ti) rȳ (buzdari ístiat). Késy túbi jémot dā́ervi tépan.
“You have been slaves all your life. Today I give you freedom.”
ístiat: "you have been." Here we have another example of -t ending a 2pl form.

buzdari: "slaves." Nom. pl., same  as the singular (though Mr. Peterson clarifies elsewhere that in the nom.s. the -i is optional.)

jévo gláesoti rȳ: "all your lives." Jévo is surely cognate to AV jiva. Gláesoti "lives"—this -oti ending is presumably the same as in valóti, which Peterson has given as the acc/gen/dat/loc pl of vala "man"... but , apparently "all" (as opposed to tólvie "every") looks like a locative singular (cf. Aerȳ). Here's what I think is going on:
  • For "durations of time" HV uses the locative (Latin would use the accusative in this context)

  • (nominative maybe *rys?) "all"

  • gláesoti "of lives", gen. pl. A partitive genitive / genitive of the whole.

  • jévo looks like a gen.s., so perhaps possessive adjectives are always used in the singular.
Késy túbi "this day." Késy "this" = AV kízi. Túbi is presumably the etymon of the AV word I spelled etoví.

tépan "I give," cognate to AV tebágho (the first person of which might be *téban)

jémot "to you," 2pl pronoun, in the dative.

dā́ervi "freedom."

D: Henujágon jáelza lúa vála mírre henujágon kóstas, (se) dáorys zíry ōdrikílza. Jémot kivío ñúhe tépan.
“Any man who wishes to leave may leave, and no one will harm him. I give you my word.”
vála mírre: "any man."

lúa: "who"

jáelza: "wishes"

Henujágon: "to leave." The -agon ending must be the origin of AV -agho. I previously thought that when the infinitive had an -a- rather than an -i-, that vowel would be stressed. But now I see there is no correlation: the -a- is short in HV, so the stress will be on it if the previous syllable is light (tebágho), but on the previous one if that syllable is heavy (píndagho). Compare also the discrepancy in stress Peterson expressly mentioned (in QG) between tebíla and gómila—perhaps that o is long in HV (*gṓmila?).

henujágon kóstas: "can leave." Kóstas 3s,  = AV kotas. For the loss of the s, compare vestrasivetras, which we learned way back in VD. (The same could have happened, by the way, in the case of gómila: *gósmila?)

se: "and"

dáorys: "no one," cf. dáor "not."

ōdrikílza: "will harm" (we have already seen the imperative plural, ōdrikātā́s.)

zíry: "him," acc.s. AV zer.

Jémot ... tépan: "I give you."

kivío ñúhe: my word.

D: Ýne sytivīlī́bilāt? Hae dā́ero valóti?
“Will you fight for me? As free men?”
sytivīlī́bilāt "will you fight for..." notice that we have not only the -t ending for 2pl, but -lībilā- as a future marker, presumably the origin of AV -liva-.

Ýne "me." I take this to be an acc.s., in part because of the -e­, but also because of its resemblance to AV yn (then again, I haven't yet figured out what the dative of that pronoun is in AV), and therefore assume that "for" is already implicit in the verb.

Hae dā́ero valóti "as free men." Hae "as" apparently functions as a preposition which governs the locative. For dā́ero, of course, compare dā́ervi "freedom."
Tags: astapori valyrian, conlangs, david j. peterson,, game of thrones, high valyrian, linguistics, valyrian

  • Sȳror morghūlis

    Well, after my mad rush to get my last entry posted in time, there was no actual Valyrian in the episode. So all that hassle and rushed typing…

  • Mērion... lantir... tomir! (Harior, āeksios!)

    Today I'm in even more of a hurry that usual, as I have a 7 PM concert (I'll miss the first showing of GoT, so no spoilers please!). As a result,…

  • Lōtintosy iderennī emilun

    There was no Valyrian in last week's Game of Thrones. Nice to have a week off and all, but of course we can always find stuff to do. And of course…

  • Post a new comment


    default userpic

    Your IP address will be recorded 

    When you submit the form an invisible reCAPTCHA check will be performed.
    You must follow the Privacy Policy and Google Terms of use.

  • Sȳror morghūlis

    Well, after my mad rush to get my last entry posted in time, there was no actual Valyrian in the episode. So all that hassle and rushed typing…

  • Mērion... lantir... tomir! (Harior, āeksios!)

    Today I'm in even more of a hurry that usual, as I have a 7 PM concert (I'll miss the first showing of GoT, so no spoilers please!). As a result,…

  • Lōtintosy iderennī emilun

    There was no Valyrian in last week's Game of Thrones. Nice to have a week off and all, but of course we can always find stuff to do. And of course…