The Mad Latinist (jdm314 ) wrote,
  • Mood: frustrated
  • Music: Rawin Djawadi — a rousing instrumental rendition of Rains of Castamere

Jomōza lua vala ēngoso daoro ȳdrassigon kostilas.

At long last it's done. I know I never get these things out until the last minute, but this entry is over 24 hours late. Sorry about that. This last week was difficult: David J. Peterson was himself delayed by LCC5, and couldn't post his transcripts right away. In the past I have managed to get quite a lot out of trying to transcribe scenes unofficially with Najahho, but this week was the first time I would have to do this with High Valyrian. Given HV's complex grammar and phonology, it is much harder to transcribe by ear, especially at our current state of knowledge (and Thoros of Myr's current state of drunkenness.) It was very frustrating to try, and even with Dinok's prompt and diligent help, it was overwhelming, and, well, qrinqōntan (ahem, nyke rudan!)  On Wednesday, Mr. Peterson finally posted (let's call that entry "Hp"), but that very night I came down with something. Between these tings I was not up to the task of getting this post out before the episode. And then of course today another post came out (let's call it GRL), which included some information I wanted to incorporate.

But before I begin, here are some general comments on what we learned off-screen last week:
  • There are six noun declensions in HV, which can be subdivided into 21 total paradigms (so far.) To determine which paradigm a noun belongs to, you only need the nominative singular.
  • While the theory of gender we've been working out (largely on the basis of Zhalio's speculations) seems to be nearly exactly right, we are not quite there. And in particular the nominative singular will not always tell you the gender. Since the system we've been using is based entirely on the ending of the nom.s., we clearly have some work to do.
  • As far as the semantics of the genders is concerned, Mr. Peterson says they have their origin in "the early grammatical distinction [he] wanted to capture (mass vs. individual)"
  • In AV, we already knew there were only two genders. We now know that "solar" and "lunar" merge (this gender is unofficially called "celestial"), leaving the "terrestrial" gender. What happens to aquatic? Well, Mr. Peterson's comment is ambiguous (he seems to imply that it simply ceased to exist), but I take it to mean that the aquatic merges into terrestrial as well.
This episode's scene marks a number of firsts: it is the first time we've had a HV-only scene with no AV mixed in, the first time we've gotten any religious terminology, the first time we've had anyone other than Dany speak HV, or gotten any Valyrian dialog outside of her storyline—as a result I had to get the video from a different YouTube channel: Here it is, by the way. We also get a first right at the start of the scene... let's get started.

ETA: Just got to chat with Mr. Peterson on IRC, and got lots of good information, which I will write up when I have time. But for now, the coolest thing is the corrected versions of my title: Jomōzussis lua vala doros ēngoso ȳdrassigon kostilas, or better yet the negative version I originally wanted: Jomōzussis lua vala mirros ēngoso ȳdrassigon kostilos daor.

Melisandre: Valar morghūlis
Thoros: Valar dohaeris
The first time we get valar morghūlis used as a greeting, and the first time we get valar dohaeris spoken by the characters at all!

T: Ólvī vóktī Rullóro Qélbriā ūndéssun daor.
Subtitle: “I don't see many priestess of R'hollor in the Riverlands.”
ūndéssun daor: "I don't see." The -un is a 1s subjunctive ending (I assume the indicative would be *ūndéssan), but now I am wondering if the -ess- element might not be a frequentative element or something like it.

Ólvī vóktī: "many priests," acc. pl. Zhalio suggests that vóktys "priest" and nekḗpti "idols" are both historically derived from passive participles, which I find very plausible.

Rullóro: of R'hllor. This is, of course, the very first time the show has mentioned the Lord of Light by name. The subtitle spells it R'hollor, but it is not clear whether this is a deliberate change (as, for instance, Asha to Yara), or just a typo. Mr. Peterson suspects it was the latter, and in any case it wasn't his fault. However, as he relates in Hp, he did decide to spell it the way he figured it would end up pronounced in HV (whether or not the Valyrian writing system would represent it that way). Notice that it is stressed irregularly on the second syllable (the normal rules of HV would stress it on the first). At first I wondered whether the name might actually be *Rullṓro, until Peterson explained that the irregular stress was due to it being a loanword (indeed, if I read him right, he is saying that the nom.s. is *Rullór, which could not possibly happen in a native word) He did reserve the right to decide that that o was long later on, though. (Incidentally, since everyone has their own pet theory of how to pronounce R'hllor, based on whatever language they favor, here's mine: like Coptic *ⲣϩⲗ̅ⲗⲟⲣ /rhl̩ˈlor/.)

Qélbriā: "in the Riverlands," loc.s. Qélbria in the nom.s., from qelbar "river." Can we assume that Qélbria declines like vala? They both end in -a, but it's possible that -ia might be a different paradigm.

M: Thóros hen Mýrot íksā.
S: “You are Thoros of Myr.”
About the only sentence we were certain of before the official transcript came out.

íksā: "you are." Note the long ā.

Thóros hen Mýrot: "Thoros of Myr." Thóros, which is apparently a Myrish name, since the th sound does not exist in HV. Once again we have hen with the locative form, which I think works very similarly to de in Medieval Latin names. The locative of Myr (assuming that is in fact the HV citation form) ends in -ot, an ending that is more associated with the dative. Peterson specifically says this is a feature of the third declension (with the example Āéksio, -ot), but as the other nouns we've seen with a locative in -ot are léntor and *Myr, which do not look at all like Āéksio, it seems unlikely that it's limited to that declension. Another question: can we assume that any noun with an -ot locative also has an -ot dative?

M: Vóktys Églie aōt gaomiláksir téptas:
S: “The High Priest gave you a mission.”
Vóktys Églie "the high priest." I'm assuming for the time being that églie literally translates "high," but in most languages I've studied the descriptor is "great" or "greatest."

téptas: "gave," 3s past

aōt: "you" dat.s.

gaomiláksir: "a mission." I have a pet theory on this word, but it's fairly complicated, so let's divide it into bullet points:
  • We start with the verb gáomagon "to do."
  • Put this in the -il- form: gáomil- "do (at some time)" or the like.
  • Add the suffix -ksy, which Peterson accidentally showed us way back in VD ["I made a baby typo (should've been zūgusy not zūguksy, which is what it was originally. Unfortunately zūguksy is, in fact, a licit form of the verb, which was really throwing me for a loop....)"] This gives us *gaomiláksy.
  • I speculate that this -ksy form is parallel to the Latin gerundive (as in the famous quote Karthago delenda est "Carthage must be destroyed."), and therefore *gaomiláksy means "needing to be done; which must be done" or as a noun "thing to do; task" (Making gáomagon *gaomiláksy exactly parallel to Latin agere "to do" → agenda "things to be done.") Now, I know I see Latin everywhere, but for evidence of this meaning, compate HV lerraski "for sale, to be sold" (pl).
  • Put *gaomiláksy (or whatever the correct nominal form is) "task" into the collective, and we get gaomiláksir "collection of tasks; mission"


M: Robérti Dā́ri zȳ́hi nekḗpti se Āéksiot Ṓño jémagon.
S: “Turn King Robert away from his idols and towards the Lord of Light.”
jémagon: "to turn away," infinitive. The stress is unexpected: either Melisandre should have said jemágon, or Mr. Peterson meant to write *jḗmagon.

Robérti Dā́ri: "King Robert," acc.s. It's cool to have the word for "king" (and I'm wondering if I might be able to use it to reconstruct that difficult line in 303), but I have no guess what the nom.s. is. I wonder about Robérti too: do foreign proper nouns automatically go into the buzdari declension, or do they sometimes get Valyrianized? For instance, if the nominative were *Robértys, the accusative would still be the same.

zȳ́hi nekḗpti: "from his idols." This would be, what, locative plural? This seems problematic. For another possible theory, see below.

se Āéksiot Ṓño: "and to the Lord of Light." Se "and." Āéksiot "to the Lord," dat.s. Interesting that this word is used both for "The Lord" as  title of divinity, and "master" in the context of slavery. The same is true of Latin Dominus.

Ṓño: "of Light," gen.s.


M: Skórion mássitas?
S: “What happened?”
Skórion: "what," another sko- question word. The stress is again surprising, but then it's quite likely that -ion would be pronounced as one syllable.  I previously speculated that scoróso was equivalent to Latin quod, and now skórion would be the equivalent of quid. For a complication in this theory, see below.

mássitas: "happened," 3s, past. I had wondered if the rule about subjunctives in negative sentences would also apply to questions, but since this verb appears to be indicative, we can conclude that it does not, at least not for "wh-questions" (or should I say "sk-questions?") For yes-no questions, compare Ýne sytivīlī́bilāt? "Will you (pl.) fight for me?" which is presumably also indicative.


T: Qringṓntan
S: “I failed.”
Qringṓntan: "I failed," 1s , past.

M: Aṓle rū́da, nūmā́zma íssa
S: “You quit, you mean.”
Aṓle rū́da: "You gave up." I take aṓle to be the nom. of the 2.s. pronoun—compare nýke, the only other personal pronoun we have in the nominative (other interpretations are possible). Rū́da would then be a 2s verb. This is slightly problematic in that the -a is not long, and there is not a past tense -t- marker, unless it has been assimilated into the -d- or the like.

nūmā́zma íssa: I'm not entirely certain. Íssa of course means "is," so perhaps nūmā́zma means "the truth"? "The real meaning"?


M: Quptýssy pōntā́lī johégzi se jomṓzū.
S: “The heathen continue to slaughter each other and you continue to get drunk”
Quptýssy: "the Pagans," nom.pl. More on this below.

pōntā́lī johégzi: "keep slaughtering one another." Pōntā́lī is presumably the accusative of pṓntalo, which we had in the expression pṓntalo syt, which is driving me even more crazy after this dialog. But in any case, it seems clear that these pōntā̆l- forms are some sort of emphatic/reflexive/reciprocal.

se jomṓzū: "and you keep getting drunk."

Two things to notice here. First of all, the last two verbs have had a continuative jo- prefix, which is "right on the edge of inflection and derivation." But second, and perhaps more importantly, they both have unusual personal suffixes. This is not the first time we've seen verbs of this sort:
  • jomō "keep drinking" 2s.
  • jáelza "wishes" 3s
  • ōdrikílza "will harm" 3s
  • iderḗbzi "choose" 3p
  • johegzi "keep slaughtering" 3p.

What is the difference between these -za verbs and the -as verbs we've grown used to? Well my first thought is that they are an entirely different sort of verb that has its own set of endings, much like the distinction between "-ω verbs" and "-μι verbs" in Classical Greek, but I didn't think this sounded like Mr. Peterson's style. Then GRL came out, and it all became clear. First of all we have some more examples:
  • jiōrinna "I will receive" 1s
  • mazōrīnna "I will accept" 1s
  • botilza "will suffer" 3s

So we now know that the 1s form is -na, which does kind of follow the model of -za and -zi (if not -zū).  Finally Peterson himself spilled the beans:
-n is the general first person ending for verbs. All suffixes interact with the stem, and so will acquire vowels if they can’t rest comfortably on the end of whatever the stem is. Plus, l and n don’t play nice together.
This doesn't quite explain everything. The implication is that "normal verbs" have a stem-vowel, even the ones we think of as regular (the -an, -a, -as, -i, *-āt, -is type). -Za verbs simply don't, and so excresce an epenthetic vowel at the end. Here's the problem with that: not only do we have minimal pairs like -as (3s) vs. -is (3pl), where the vowel depends on the ending, rather than the stem, but these same vowels repeat themselves in the epenthesis: -za (3s) vs. -zi (3pl). So whatever Peterson is saying, it's not quite as simple as that. Note that this still doesn't explain -zū either, if we are right to include that in this paradigm at all.
ETA: "We" were not ;) According to DJP the stem is mōzu-, and is just the regular 2s ending.

Notice as well that these special endings are added even if another infix intervenes: ōdrik·il·za shows that -il- infix. (And, incidentally, I suspect ō- or, more likely, ōd- is also a prefix; the same as in ōssenagon "slay.")

One last thing thing before we move on: this paradigm is likely the origin of AV verbs that seemingly have a 3s in -a, such as éza "has" and possibly kísa "has (time remaining)." Should we expect the 1s form *éna then?


T: Aṓhoso zíry rijī́biā, se ñuhóso zíry rijī́bin.
S: “You worship Him your way, and I'll worship Him mine.”
zíry rijī́biā: "you worship him"—since it is not stressed on the last syllable, this seems like a statement of fact, rather than a. But this interpretation may not be correct: rijī́bin shows this verb to be an i-stem. I had wondered what happens to i-stems in the 2s, and this seems to imply -iā. However, in the bonus sentence at the end of this entry, we get gī́mī "you know," which implies . Which means this may be an imperative after all.

zíry rijī́bin: "I worship him," 1s i-stem.

Aṓhoso ~ ñuhóso: "in my way ~ in your way." Clearly this -oso ending indicates manner. And it does resemble the -osa in valosa, the ins.sing. of vala. So my first instinct is to take these as instrumentals of the possessive adjective: "by your (way)" or the like. We also see this ending in the very next sentence, in Ḗngoso, and it's also appeared in keskydóso "in the same way." Another possibility is that it is simply an adverbial ending, making aṓhoso and ñuhóso equivalent to the Latin tuātim, and meātim (same meanings.) Either way, it's difficult to explain in skoroso (I suppose Daenerys could be asking Grey Worm "How's your name?")


T: Qupténkos Ḗngoso ȳdrássis?
S: “Do you speak the Common tongue?”
ȳdrássis: "do you speak." As we've seen, in AV this would be ydra, so we might have expected just *ȳ́drā here. But this is not what we get. Is this ȳdra·ssis with some unknown suffix? Is it ȳdrass·is, and the verb just takes -is for some unknown reason? Is that -ss- just part of the stem, or could it be a variant on the -ess- in ūndéssun, which I suggested might be a frequentative?

Qupténkos Ḗngoso: the Common Tongue. Now things get interesting:
  • First of all, ḗngoso: this shows that the ending is probably instrumental. Latin does use adverbs for names of languages (e.g. Latīnē loquī "to speak Latinly," i.e. "speak Latin"), but not with the word "tongue/language" itself—and it's hard to imagine otherwise ("Do you speak Common Tonguely"??) In such cases Latin uses the ablative case (e.g. linguā Latinā loquī "to speak by the Latin tongue" i.e. "speak the Latin language"), which would indeed translate to the HV instrumental.
  • On the other hand I don't know how to explain the -os of Qupténkos. But, then again, I have not even started trying to understand HV adjectives. And it does seem that HV adjectives are only get full paradigms when they are placed after the noun.
  • As for the word Qupténkos itself, it looks suspiciously similar to Quptýssy "the pagans." Either the word for "pagan" is derived from the word for "commoner" or "common" (compare "pagan" itself, which etymologically means "rustic," from pāgus "canton; countryside"—as well as "heathen" which is likewise from "heath") or perhaps Essosi R'hllorians just refer to the inhabitants of Westeros as  "the pagans" (as we've just seen Melisandre do.)
In any case, the connection between qupténkos and quptýssy was noted almost immediately by Joel W, after which Zhalio noticed something even more interesting: the relationship beween quptýssy and qupténkos seems to be exactly that between HV nedýssy and AV nedhínki "brave." He suggests that the -ýssy forms are "categorical," glossing nedýssy  as "brave people in general" (and he's espoused this theory from the start). This does look somewhat similar to the verbal element I've been calling "frequentative." Another point in its favor is our HV sentence with nedýssy also contains the mysterious verb form ́gusy (I originally suggested that nedýssy and ́gusy constituted an accusative-infinitive construction, but we now know this to be impossible)—perhaps the -usy­ form is just used with these "categorical" adjectives? On the other hand, we've already seen that collectives (which is how I'd interpret "categorical") take -is verbs (as in valar morghūlis). Another possibility is that the -ýssy is simply the form used when an adjective is substantivized (cf. nedýssy vs. váli nedhínki), and then the -usy verb is required for otherwise umarked subordinate clauses (and is just coincidentally used here with nedýssy)—HV does not, after all, seem like the sort of language that would form subordinate clauses just by juxtaposing them to the main clause with no marker at all.</li></ul>

We now move on to the scene in the cave with Beric:
M: Kónir ságon kóstos daor.
S: “That's not possible.”


Kónir: "that." Possibly a collective form meaning something like "all of this"?

kóstos daor: "cannot." Kóstos is the subjunctive of our already attested kóstas.

ságon: "be." The infinitive of íksan, íksā, íssa. For the phrasing compare the Latin fieri non potest "that cannot happen"—the word impossibilis exists, but it is rare, and kind of pretentious.


T: Āéksio ýne ilī́ritan.
S: “The Lord has smiled upon me.”
Āéksio: "The Lord," nom.s. (finally).

ýne ilī́ritan: "has smiled on me." Ilī́ritan seems to have that past -t-, but I cannot explain the -an ending, which we would expect to be the 1s. The meaning cannot be passive ("I have been smiled upon..."), because of the accusative ýne.
ETA: Mr. Peterson has admitted this was an error on his part. My fanon theory: it ends with -an in Myrish, and Thoros confuses it because he's drunk.
However we explain this ending, notice that we once again have this i- prefix. Recall my speculation that i- converted vestragon form "say" to "tell," in other words effectively changing the indirect object to a direct object. And once again we have a verb with i- taking an accusative when we might otherwise expect some other case or a more complicated expression (Zhalio caught this as well, and has been saying something similar all along). In a comment which has since been lost in the crash, David J. Peterson confirmed that, at least for most verbs, the i- is an applicative marker.


M: Késys óndor ávy sytilī́bus daor.
S: “You should not have this power”
Mr. Peterson glossed this as “You should not have these powers.” The Valyrian does not quite match either of those, structurally, but so far as I can tell óndor is in fact singular. The Valyrian has to mean something more like "This power should not belong to you." The problem is this mysterious sytilī́bus, which will require some careful analysis.

The other week, Dinok pointed out the similarity between pṓntalo syt "their very own" and sytivīlībilāt "you (pl) will fight for..." And last week, Zhalio noted that sytilībus was suspiciously close to sytivīlībilāt as well. Peterson then gave the following hint:
Also, it might help to know that, even though not all the affixes have any meaning, but the second word segments as syt-i-vīl-īb-il-āt.
This is indeed helpful, but still quite cryptic. What precisely does he mean "not all the affixes have any meaning"? Does he mean some of them are just word-forming affixes, with little to no semantic content? [Compare, for instance, English -ate, which is a common element especially in verbs (alleviate, eliminate, recuperate), but also adjectives (effeminate, ornate, irate), and nouns (substrate, degenerate, tunicate), and yet at this stage has little to no meaning of its own.] Lets take a look at that, bit by bit:
  • syt: since it occurs both as a verbal prefix and in an expression meaning "their own," a logical guess would be that this is a reflexive market. Granted, fighting for someone isn't literally reflexive, but in many languages these things occur idiomatically, with "middle voice" meanings. Still this is just a wild guess.
  • i: applicative (changes "for me" to just "me")
  • vīl: presumably the main root, "fight."
  • īb: unknown. I originally guessed "future" (cf. rijībin, above) but I'm not so sure that works.
  • il: "timeless"
  • āt: 2pl.

If we try the same with sytilībus, we might get syt-il-īb-us, or maybe syt-i-līb-us. Since we have ávy, an accusative object, I guess the latter makes the most sense. So does līb- mean "to pertain to" or the like?

Notice, by the way, that while High Valyrian is pretty clearly a fusional language overall, the verbs have strong agglutinative tendencies. This kind of thing isn't at all unusual: for instance, Wikipedia points out that "Japanese is highly synthetic in its verbs, but clearly analytic in its nouns."


T: Óndor émon daor.
S: “I have no power.”
Interestingly, Thoros changes the sentence structure to something more like the Engish: emon presmably means "I do not have" (1s subj.) *Emógon does not seem to be the ancestor of AV éza: we would expect the 3s form to be *émos.


T: Āéksiot zȳ́hon vaoréznon jépin, se ziksóso udlíssis.
S: “I ask the Lord for His favor, and He responds as He will.”
jépin: "I ask," 1s. i-stem. Before we got David's official transcription, I briefly thought maybe we had misheard, and this was actually a form of the ancestor of *píndagho, but I quickly realized this had to be just a 1s ending.

Āéksiot: The Lord, dat.s.

zȳ́hon vaoréznon: "his favor," acc.s. Zhalio points out that this is likely a terrestrial nomen actionis, like hepnon "climb."

se: "and"

udlíssis: "her responds," 3s, i-stem.

ziksóso: "in his way." Again, we would expect the stress zíksoso. Furthermore, on the model of aṓhoso and ñuhóso we might have expected *zȳ́hoso.

And, as a bonus, DJP included a line that was cut (a shame, it's a nice touch): Késir gī́mī “You know this.” Kesir is again seemingly in the collective. 
Tags: conlangs, david j. peterson, dothraki.com, game of thrones, high valyrian, linguistics, valyrian
  • Post a new comment

    Error

    default userpic

    Your IP address will be recorded 

  • 7 comments