Indefinite Relatives

Or something close to that, anyway. In honor of the Olympics (though not really in honor of the Olympics), I thought I’d go over a very small, very specific bit of Dothraki grammar.

But before I do that, after upgrading OS X Mountain Lion, the weirdest thing is happening in WordPress. Basically (and there’s really no other way to put it), certain punctuation marks are disappearing. They’re there, of course, and they show up in the published post, but I’m not allowed to see them, apparently. There’s no way to demonstrate other than showing you exactly what I mean, so here’s a screenshot of what I’m seeing:

You see! (Oh, how I’m glad you can’t see what that exclamation point was just reduced to on my screen…) True, the parens are mostly there, there are commas galore, and you can kind of make out the periods, but what of the apostrophes?! Where have they gone? WHERE ARE MY APOSTROPHES?! Man, when I get them back, I am so saying dracarys!

Anyway, we’ve already seen how to form a relative clause in Dothraki. That’s actually the tough part. Next comes identifying indefinite relative clauses. These are (by my definition, anyway) relative clauses where the target of relativization is unknown. In English (and in many languages), what happens is the indefinite relative is actually an embedded WH-question. Here are some examples:

  1. I don’t know where he slept.
  2. I saw who ate the sausage.
  3. We’re deciding when to come.

There are a number of sensible reasons for doing things thus (which is why the strategy is common), but Dothraki does things a little differently. Dothraki uses its series of distal demonstrative pronouns in place of WH-words and then forms a sentence very much like a relative clause. Let’s start with a simple example—the second sentence from above. First, here’s a standard relative clause:

  • Anha tih mahrazhes fin adakh ninde.
    “I saw the man that ate the sausage.”

The word mahrazh, “man”, is put in the accusative case, because it’s the object of tih, “saw”. Fin, “that/who”, is put into the nominative case because it acts as the subject of the embedded verb adakh, “ate”. So far so good.

Now for the indefinite relative. Since it’s not revealed who performed the action, we don’t have a noun to hang fin off of. Thus we insert the pronoun rekak. Here’s the sentence above translated into Dothraki:

  • Anha tih rekakes fin adakh ninde.
    “I saw who ate the sausage.”

This basically looks exactly like the first sentence, but now the pronoun rekak, “that one”, stands in place of mahrazh. Additionally, since the pronoun is used conventionally in these contexts, it can often be used without fin—and often in either the case it would take in the embedded clause or the matrix clause.

Now for the new part. There are other types of indefinite relatives that don’t act exactly like standard relative clauses. Consider our first sentence from above: I don’t know where he slept. In this one, the pronoun becomes rekke, and the sentence is translated thus:

  • Anha nesok rekke remek me.
    “I don’t know where he slept.”

Now there’s no need for fin at all (indeed, using it would be ungrammatical [or at least bizarre] at this stage) and rekke acts a lot more like “where” in the English translation. The same can be done with arrek (best translated as “when” in such sentences). For “how” and “why”, there are two constructions that can be used, but are nevertheless rare (usually the sentence is just reworded). Those terms are: kirekhdirgi “why” and kirekosi “how”.

And there you are! Now you should be able to tackle some tougher texts in Dothraki. For those who made it all the way to the end of this post, I shall reward you with a picture of my havzi vezhven:

My cat.

Click to enlarge.

Look at that little pink tongue! What a cat she is!

Posted on July 28, 2012, in Grammar and tagged , , . Bookmark the permalink. 16 Comments.

  1. I am trying very hard to learn this beautiful language! What would really help is having sound with the words. Are you planning anything like this? Maybe there is something already that i’ve missed, if so can you post a link where I can find it. Thank you muchly :^)

    • M’athchomarron, zhey Kelly! A number of posts have audio in them, but I actually try not to do it all the time, because it makes the page load slow. :( You can see all posts that have audio by clicking on the audio tag.

      Thanks for stopping by!

  2. idk this sentence:
    Anha nesok rekke remek me.
    “I don’t know where he slept.”
    first: nesolat means to learn
    second: it’s negative sentence so i think u forgot add “vos”

  3. idk this sentence:
    Anha nesok rekke remek me.
    “I don’t know where he slept.”
    first: nesolat means to learn
    second: it’s negative sentence so i think u forgot add “vos”

  4. idk this sentence:
    Anha nesok rekke remek me.
    “I don’t know where he slept.”
    first: nesolat means to learn
    second: it’s negative sentence so i think u forgot add “vos”

  5. I love these comprehensive grammar detail posts. Not that I don’t like the less detailed, wider area grammar posts. Not that I don’t like less grammary Dothraki posts. Whelp, I guess the posts are just generally lovely.

    Anha nesok rekke remek me.
    “I don’t know where he slept.”
    first: nesolat means to learn
    second: it’s negative sentence so i think u forgot add “vos”

    I think I can give some peer assistance before David saves the day. Nesok can be derived either from nesolat or nesat, so here it is used as a first person negative of to know. Dothraki (and, by extension, David) have an infuriating habit of often dropping the vos/vo, when they feel the used word is unmistakeable enough and when the negative grade is explicit in the verb conjugation (so with nesolat it could not be dropped, but with nesat it can). The sentence could be read as “I’m learning where he slept,” but as that is a rather strange statement, Dothraki probably wouldn’t even notice this alternate interpetation.

    This issue sparks discussion every now and then. I think the last time I heard discussion about it was the very latest IRC chat. The practise is somewhat on a border of believably decipherable language, but not on the nasty side of the fence. Natural (and natural feel aspiring con-) languages be crazy. Homonymity through inflection is very common thing in Finnish (my native language), and probably in most heavily inflectional languages. In most (though not all) cases in Finnish, the homonyms are not of closely related words or aren’t in the same case/conjugation, so it’s mostly easy to parse, what was meant. It’s more of an issue, when the exact sense of the word is lost, as I’d imagine closely related words such as dynamic and stative versions of verbs often both would fit well enough into the situation and the sentence. And it is of course rather curious that the negative can be made almost completely invisible, to be parsed from the context, though that too, I think might be sometimes seen in Finnish slang, when lazy speakers swallow half of the sentences, so it can’t be that unusual.

    Are there any languages where it’s commonplace to truly elide any indication of negative grade? Are there languages where both negative and positive would be optional clarifications and the regular expressions would be neutral to the polarity (a bit like “My computer [being obsolete]” “Why you [buying a new computer]?” “Because I [having money].”)?

    Wiki section on the indefinite relatives needs a do-over. I could only half guess, what was these things even were, and the details of the mechanics are rather different from what I could infer before.

    English sometimes use compound pronoun thingies like whoever, whatever etc. Would you translate these in any particular way, or would eg. “Whoever tries to fight, will die.” be just “Rekak kis alaja, adrivoe.”?

    Me Reki Driva Laz Vo Drivoo Avvos looks stranger now than it looked before. So we have (underlining the relative clause) Reki fini driva laz vo drivoo avvos. (which, I understand still isn’t obsolete way of expressing the idea). Fini is then skipped, so we have Reki driva laz vo drivoo avvos. And then reki is pulled inside the relative clause, which in this case happens produce an identical sentence, but nevertheless is technically differently structured, Reki driva laz vo drivoo avvos. Then we re-introduce matrix clause level pronoun, now me: Me reki driva laz vo drivoo avvos!

    • I think the difference in meaning between “I don’t know…” and “I’m learning…” would probably be made clear by things like intonation or gestures like shaking of the head etc.

    • English sometimes use compound pronoun thingies like whoever, whatever etc. Would you translate these in any particular way, or would eg. “Whoever tries to fight, will die.” be just “Rekak kis alaja, adrivoe.”?

      I would reword that Me adrivoe, rekak kis alaja.

      Me Reki Driva Laz Vo Drivoo Avvos looks stranger now than it looked before.

      I thought about this one a lot when translating it. What I ended up deciding (though I wasn’t sure if it would bear out down the line) is that you couldn’t hang a standard relative clause off a pronoun—and I felt the pronoun was needed here. I went back to the r set then and it felt right. It may need revisiting.

  6. A sentence like “I know who ate the sausage” is potentially ambiguous. It could mean “I know the individual who ate the sausage”, or it could mean “I know the answer to the question of who ate the sausage”. Call the former the ‘free relative’ reading and the latter the ‘indirect question’ reading. Some languages have formally distinct constructions for these two meanings, while others don’t. What about Dothraki? The use of a definite pronoun suggests the free relative reading to me, but your discussion suggests that sentences with rek also have the indirect question reading.

    • It took me an entire weekend of thinking, but I think I finally understood how the two readings you posit can be distinct. Though I may have just lost it. Darn. I swear, those two English sentences seem just about identical. I see how theoretically they can be different, but I can’t read them as different. The best I can do with the second reading (the indirect question) is pretending that it’s an echo question. Is that on the right track? Anyway, no, there’s no distinction in Dothraki, I’m fairly sure (though you could certainly do an echo question qua echo question, e.g. Anha nesak fin adakh ninde). Man, I just had the distinction again, but then I thought about it, and it still makes no sense why they should be different: They’re getting at the exact same information. Do you, perchance, have a natural language example where the two differ?

      • In Swedish you would use different constructs for the 2 readings. I assume the same distinction is made in German as well for example with the difference being wissen vs kennen.

        Ich weiß, wer aß die Wurst.
        Ich kenne [die Person], die die Wurst aß.

        • It works the same in Dothraki though, doesn’t it? We got nesat “to know something” and shilat “to know someone”.

        • Well. Apparently, my comment got swallowed, so I’ll have to try again. As Makhiel points out, yes, of course Dothraki has the same wissen/kennen distinction. Oops! Forgot about that. May need to go back and adjust… But there’s a further distinction you draw in that German example as there are two different constructions. The first is a standard embedded clause, and the other looks like a relative clause attached to nothing, am I right? (I.e. “die Person” is optional, and, I’m guessing, uncommon?)

      • In Malagasy you’d say Fantatro izay nihinana ny saosisy for the first one, and Fantatro ny nihinana ny saosisy for the second one. In the second example, you take the verb phrase nihinana ny saosisy “ate the sausage” and stick the definite article ny in front. The resulting expression means “the one who ate the sausage”. The former example uses izay, which forms free relatives (“whoever ate the sausage”).

        When you use ny, it means you’re acquainted with the individual in question: the guy who ate the sausage is named Bob, and I know Bob. When you use izay, it means that you can provide an answer to the question “Who ate the sausage?”, but you may not know the individual in question. I’m aware that Bob is the one who ate the sausage, but I’ve never met him so I can’t claim to know him.

        Granted, in English we’re more likely to say “I know the one who ate the sausage” for the former meaning. My point was simply that ‘true’ indirect questions and expressions meaning “the one who…” or “the thing which…”, etc., are not semantically identical.

        • My point was simply that ‘true’ indirect questions and expressions meaning “the one who…” or “the thing which…”, etc., are not semantically identical.

          A point which I was willing to grant, but I couldn’t see it in my mind. This explanation, however, made it quite clear:

          When you use izay, it means that you can provide an answer to the question “Who ate the sausage?”, but you may not know the individual in question. I’m aware that Bob is the one who ate the sausage, but I’ve never met him so I can’t claim to know him.

          Right! And so it seems clear that in Dothraki, you’d be using two different verbs here (shilat if you’re acquainted with the individual; nesat if you’re not). While that corresponds to a lexical difference, it’s not a syntactic difference. I gather the difference is syntactic in Malagasy, though?

          • Ah yes. Having different verbs for “know” (know a person versus know a fact, or the like) would definitely be a way out of the ambiguity, as noted in the comments above. Malagasy, like English–but unlike German or Dothraki–has a single verb for both senses of “know”. (Incidentally, my conlang Okuna does both: it makes the wissen/kennen distinction and distinguishes indirect wh-questions from nominalizations with meanings like “the one who did X”. A bit much, perhaps.)

            By the way, I hope you didn’t take my earlier remarks as a criticism of Dothraki! I was just wondering, out of idle curiosity, whether you’d considered the subtle semantic distinction I was trying to articulate.

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: