Things not yet known.

[ingsve]

This is a place to list information about grammar etc that is still unknown or unclear that we need to ask David Peterson about or figure out in some way.

-Non-nominative declensions of all pronouns.
-More about animacy of nouns.
-More about adjective comparison.
-Nominative of certain nouns. (Possible missing vowels?)
-What does -o- indicate in certain verbs?


Feel free to suggest other things and I'll add them to the list.

[Qvaak]

This post deserves some follow-ups. I always ponder about this and that unknown detail. This should be the very place for all the little doubts.

The more we know the clearer the holes in our knowledge seem to get. We seem to have pretty decent understanding of verbs already. I have paid some thought on them lately, so let me start there.

 - Zalat and vrelat seem to be consonant-style conjugating (zal/-at/, vrel/-at/). Is there any logic by which you can tell the difference between l-ending verbs and vowel-ending verbs by their infinite form, or is this just a thing you must know?

 - O's of the negative seem to be still at least a bit fuzzy.¹  Are those negative first person present suffixes assuredly the only place where suffixe changes the vowel before the /-lat/? Are we even assured that there are no first vowel changes in the future negative (I think we are, but still have a little bit of doubt floating around)? How about imperatives? Do they agree to negative, might it even go as far as "vos dothros"?

 - How analogous is the use of tenses with english tenses? I have said before that I intuit the use of past tense to be wider than with the english counterpart ². Is there a pluperfect? Does perfect see a lot of use? Do zero copula past and future expressions bear any specific connotations or uses or are they just part of the tenses?

 - Dothraki is non-pro-drop language and seems to have rather strict transitivity classes. Are there verbs that do not react to transitivity change? Can you at least add extra arguments without worries ('I cut the cake' -> 'I cut the cake with a knife')? Are there avalent verbs, and if there are, how do they work?


¹/-ok/ and /-oki/ even suggest that the whole vowel ending stem stuff might be better understood as consonant ending stems with conjugation suffixes /-Vlat/, /-Vk/, /-Vki/...etc. That would be rather uncomfortable, but I think it would follow more closely the common idea of 'stem'.

² Mainly because Peterson has usually just listed past, present and future and because perfect structure seems more like an expression than fully fledged tense. hehh. English past, present and future tenses are, I think, actually quite similar looking to dothraki ones. Many languages seem much more complex.


I'm sure there are a couple of things I've just missed or forgotten, but I'd think most of these are real unknowns.

[ingsve]

Some interesting questions. The only one I think I have an answer to is wether the imperative also agrees with negation. It seems that it does. In episode 8 Drogo says Os! to Qotho which David translates as "Don't move!". I'm guessing this is the negative imperative of elat.

[Qvaak]

Some interesting questions. The only one I think I have an answer to is wether the imperative also agrees with negation. It seems that it does. In episode 8 Drogo says Os! to Qotho which David translates as "Don't move!". I'm guessing this is the negative imperative of elat.

Sweet. And he even drops the vos off. When I originally watched the episode, I thought he said "path", like in "give me path" or "get out of my way", but that makes even more sense.

How about future imperatives? Those would be funny.
 - Anha vek ashefasaan silokh.
 - Es ajjalan, vosma vos vos silokh.
HA! Does any language really have future imperatives? I could imagine a lot of meanings such a structure might convey.

[Qvaak]

More!

 - Peterson mentions participle as an archaic verb form. What is it look like and where might you encounter it? Is it just rare and maybe a bit theatrical, or is it already just a thing of some idioms, proverbs et al.?

 - I see we have learned adjectives have some agreement with noun cases; consonant ending adjectives get a bonus /-a/ in every other case than nominative. Is this really this simple? How does this figure with plural /-i/? I'm guessing the /-i/ comes "before" and then the adjective ends in vowel, problem solved. How about pluralising vowel ending adjectives? Do they get the /-i/ at all?

 - Sepaking of adjectives, I wonder, how many are correctly suffixed and syntactically true adjectives in the dictionary. As with almost all word classes, our understanding has been limited and thus errors are likely. For example:

• Ezhiray is quite possibly in aforementioned participle verb form. Does it still work the adjective way suffix-wise? (Rikhoya actually hints that it might.)
• If "Arakh hasa." means "Arakh is sharp," to my logic the base adjective should be just has.
• Tawak is not listed as an adjective though it has many adjective like meanings, while it's counterpart ido is listed as an adjective. Is this right? Ido does not seem to react to neither plural nor non-nominative nouns, but that might be just because it ends with a vowel.
• Peterson has pretty much straight out said there isn't stative zigerelat meaning "to be required," "to be needed". Does this mean zigere isn't a proper adjective?

[ingsve]

Tawak is an interesting example even in English. The word metal does not seem to be listed as an adjective in English dictionaries but that seems to be the word most often used when referring to the adjective describing what material something is made of. The adjective would rather be metallic but you usually don't say "a metallic ball", you would say "a metal ball" or "a copper ball" rather than "a coppery ball".

Tawak is probably a noun for the same reason since we also have the word tawakof which means steel and /-of/ is a morphology applied to nouns. Or are there other examples of adjectives getting changed by /-of/?

[Qvaak]

evermore.

 - There is this verb class thingie, where verbs convey some particular meaning when the object is in some case other than the standard accusative. And then there are also situations, where an object requires a preposition - and the prepositions again assign different noun cases. How does this work with passive sentences? Dothraki passive seems structurally more like an objectless sentence with reversed verb than subjectless sentence with reversed word order, and it seems to me that even in english the prepositions do not migrate with object: "I don't like to be shouted at."

 - There seems to be a lot of derivational morphology, where new words are created by adding a vowel at the end of the word. Often these new words seem to lose their distinctiviness in some inflections. I guess this could just be the case, but if I had to guess, I'd rather guess there is something we just missunderstand/don't understand/don't know yet.
exempli gratia:
astat/astilat/astolat --negative 1st person singular--> vos astok
khogar/khogari --accusative--> khogar

 - There is a lot I'm unsure about stress. Why is stress mark always on a vowel? I'm guessing it's just a notation thing, because I'm imagining the whole syllabes are usually stressed.. But then the whole stress thing (well, just the penultimate syllabe stress exeption, but hell, allow me some stylish exageration) is based on how words brake into syllabes, and I'm not sure, if that is a trivial thing, either.
When I was thinking about the above homonymization through inflection problem, my probably most bizarre hypothesis was, that the syllabes would sometimes go different:
ast-at -> vos ast-ok and as-ti-lat -> vos as-tok.
Even though it would not affect the stress, the pronunciation might still be significantly different. And wouldn't that be whole bucketfuls of hilarity..

 - Oh, and is there always just one stress? I seem to remember that at least in finnish there is talk about secondary stresses, and in dothraki there are awfully long words, especially compound words. By all means they can slip through in an even flow, stress just on the first or final syllabe, but there might as well be some stressy action somewhere.

[ingsve]

The accusative of khogari is khogare to avoid this specific confusion. This is in the dialogue from episode 7.

[ingsve]

We don't seem to know how to express impersonal expressions in Dothraki. For example the subject in the sentence "it is raining".

[Qvaak]

The accusative of khogari is khogare to avoid this specific confusion. This is in the dialogue from episode 7.

Shiver me timbers! To choose an example that bears an answer to the question - and to not even notice! ehh.
That is very interesting. As far as I can see, I think this solves the case with nouns. Of course there are still homonyms, but if they aren't derivationally close and/or aren't in same case, syntax and context should generally ascertain the meaning. Not the most comfortable turn in the story, though. It seems we just have to know, if the word is suffix vowel addingly derived from another word. As long as our vocabulary is as incomplete as it is now, this means in practice a huge number of irregular (inanimate) nouns. All diminutives should then have /-e/ accusative.

We don't seem to know how to express impersonal expressions in Dothraki. For example the subject in the sentence "it is raining".

Yeah. I think I touched on that earlier in the thread - if I got my linguistic lingo right, that is. I think that impersonal expression thing is (also) called avalency. In pro-drop languages the sentences like this have often nothing but the verb, "It is raining." is in finnish "Sataa." and in spanish "llueve." As dotharki is not a pro-dropper, we might expect a dummy subject as in english. I'm not sure, though, if a language even really always needs to have avalent verbs, you might go with "rain is falling", "the weather is getting cold", "the atmosphere was quiet"..

[ValekLost]

I've forgotten to paste my tweets with David about this, I'm sorry 

His Tweets:

"There are no weather verbs, like "piovere"["to rain", in Italian], in Dothraki. Instead, you use the noun with a passive verb e.g. "Eyel nem atthasa"."

"In Dothraki, you can't have an impersonal subject with "need", in any case."

"I'd probably say "Adakhat nem zigeree" for "One needs to eat" or "It's necessary to eat"."



At the end it's simpler say "people need to eat" or "you need to eat" 

[Qvaak]

Aight. "Rain is caused to fall" it is then. I guess that's one question answered there.

[Qvaak]

I thought it would be nice to recap, what has been puzzled here so far, and how much has been clarified.

Notable progress or solved:
 - Non-nominative declensions of all pronouns?
I'm not entirely sure, what things count as pronouns (memra, maybe?), but this seems to be at least mostly solved.

 - More about animacy of nouns?
I think we have pretty good grasp of animacy nowadays. This too is a big topic and I'm sure there is much to be learned.

 - More about adjective comparison?
I think we have comparative declensions figured out, but verbifying and other usage issues might still offer some surprises.

 - What does -o- indicate in all those -olat ending verbs?
I'm pretty content with my presumed understanding of the /-o-/ thing now. If I'm correct, it's an activity level enhancer, pretty much.

 - O's of the negative, do we maybe still misuse them somewhere?
We've made some corrections on the negatives since this was asked. Maybe this issue is already wholly solved.

 - How much does verb transitivity figure into grammar?
I thought verb transitivity might be a big thing in dothraki, but now it seems that transitivity might be even somewhat fluid thing. The verb form changes are better understood with other concepts, and transitivity changes are just a secondary effect.

 - Are morphological distinctions really sometimes lost in inflections?
It seems the answer is yes and no.
Both /-olat/ and /-at/ do indeed end up vos /-ok/ when appropriately inflected, so verbs evidently sometimes lose the meaning difference attained through morphology.
On the other hand, at least in some cases nouns have irregular accusatives just to mark the difference to morphological parent-word.

 - More about syllabification and stress?
I think we've learned quite a lot about this topic. Some details are fuzzy, but not that much anymore.

 - How do participles work?
I think we know now. At least as well as could be reasonably expected. I
don't think we just know much anything. There are always possible irregularities, syntax tricks, limits in scope ... stuff

 - Avalent verbs?
Just ...no. Some potential avalencies are dealt with passive.


Not much progress, unsolved:
 - Nominative of certain nouns. (Possible missing vowels?)
 - How well are the adjectives in the vocab suffixed and marked?
The general goal of getting the vocab error-free is still far away. I know David has had a look at the page, but I don't think he's done any comprehensive spell-checking.

 - Verbs with l-ending stem: what's the deal?
Minor detail. Probably there's nothing interesting about them. We just haven't asked.

 - How analogous is the use of tenses with english tenses?
Well, this would be a nice topic to bring up at the irc chat.

 - Details on adjective declension?
No progress, but this too was a clarification plea on a topic we have already mostly tamed.

 - How do special verb inflections and prepositions work with passives?
I think this is part of much bigger question. I'll expand on my next post.

 - How about secondary stress?
This is veeery minor detail. A bit silly question, actually, I think.

[Qvaak]

The more you know...

 - How do the different many word expressions work? We have just learned that the animacy of a noun-type many word expression is determined by the first word, so we at least know what kind of demonstrative pronoun to use with them. It goes to figure that when the latter word of the noun-type expression is a possessive, that word does not further decline. I think we have some decent examples already. How about adjectives, though? How about nouns of equal status? If there is some special treatment, what expressions are expressions enough to qualify?
Alegra allayafi shekh ma shierakis anni. - Ducks please my sun and stars. ...?
Anha chomak shierakea qiyai¹. - I respect comets. ...?
How about additional adjectives? Possessives would normally come after them.
haesh rakhi erin - a kind lamb man ...?
haesh erin rakhi - a kind spawn of a lamb, but not so much a lamb man ...?
And the same goes for two part verbs, of course.
Anha thirok vosecchi athiraride. - I do not dream. ...?

 - I'm still wondering, where the prepositions meant for objects go in passive and reflexive sentences, and if the subject (object... what ever) can just assume a special case assigned to the object (patient... what ever) by a verb class. But now I'm also wondering, if verbal auxilaries stack. The functions they provide are after all very diverse.
Yeroon laz ray nemo kemoo vosecchi ma! - You can't have married yourself! ...?!?

 - All in all there is much much to be learned about the use of the verbal auxilaries. Laz, for example, to my understanding gives a subject an attribute has.a.canny.relation.to.the.verbing. Now, if the verbing is in negative grade, this mechanically would mean that the subject would be very able not to verb. This seems a counter-intuitive interpretation, though.
Anha laz vos ezhirok. Would make much more sense as "I cannot dance." than "I can refrain from dancing."

 - Vowels are interesting. There are only four², and even though I'm quite certain that none really have any comprehensive role or overtone, all seem to have their own character: o is for change and negation; i is for unindividualizing and diminishing; a is generic and replacable; e is there just to support the structure, pay it no attention.
How much more is there to be learned about differences and roles of this lot? I think we haven't seen a double i, is it not allowed? Do a's and e's have any characteristic modifier functions the same way o and i have?

¹ That qiya is quite an extra can of worms. As mentioned earlier, I don't understand adjective declensions all that well and to have a vowel ending adjective in plural and perhaps also agreeing to the allative...
²Plus some somewhat vowelly things and a hint of diphtongs and some non-phonetic shifts.