Numbers, Numbers Everywhere

So let’s all take a drink! Per a request initially made by ingsve over at the Dothraki forum, today’s post will be about numbers in Dothraki. In addition, though, since I think it might be interesting, I’m going to expand on the topic to talk about number, in general, in Dothraki.

One of the questions one has to answer when creating a language is just how that language will treat numbers and number—that is, grammatical number or plurality. Many languages deal with number in many different ways. Some languages (Arabic, for example) have a dedicated dual number. In the case of Arabic, this means that in addition to having a plural suffix, nouns can take a suffix which means “exactly two” (and, in fact, the plural suffix is used not to mean “more than one” but “more than two”). Here are a couple examples:

Singular Dual Plural
sadiiq “friend” sadiiqaan “(two) friends” ‘asdiqaa’ “(three or more) friends”
rajul “man” rajulaan “(two) men” rijaal “(three or more) men”
waalida “mother” waalidataan “(two) mothers” waalidaat “(three or more) mothers”

There are also languages with trial numbers (forms for one of something, two of something, three of something, and four or more of something), and a paucal, and different things like that—and, if you believe the stories, even languages that don’t seem to have any number system at all.

In Dothraki, as I’ve stated before, I wanted to realize the language as it might be imagined to exist in the universe of A Song of Ice & Fire. So even though a conlanger doesn’t need an excuse to, say, create a number system that relies on a base other than ten, I felt like I needed a pretty good reason to do anything other than what an English-speaking audience would expect.

In the books, numbers are pretty much exclusively base-10 (including references to the size of khalasars [twenty thousand, forty thousand, fifty thousand], and other groups). In addition, since the Dothraki—and those groups that border the Dothraki Sea—all trade, I think it’s not unreasonable to assume that by the time of the action of the books, everyone will have converted to the same base (this is generally what’s happened in our world, even in places where various societies retain their own monetary system). So while it might have been interesting to make Dothraki base-8 or base-12, I stuck with base-10.

As for nominal expression, I decided to stick with singular and plural (rather than dual or something else) for a rather practical reason: I wasn’t sure if I’d know for certain whether or not something referred to in a script was dual or plural. One can never be sure, after all, and if I needed to translate the phrase, “Get those horses”, I’d need to know if there were two horses or three or more to translate it properly—and even if I got the information at the outset, who knows but the director might decide at the last minute, “No, there aren’t enough horses. Add two more.” Languages that have duals usually are pretty strict in using them, so it’d be odd if a line referring to two of something used the plural, and extremely bizarre if a line referring to three or more of something used the dual.

Along those same lines, I decided one thing I’d do to distinguish animate nouns from inanimate is that inanimate nouns would make no number distinction at all—at least not on the nouns themselves. In effect, inanimate nouns are treated like mass nouns (part of the reason they’re called vekhikh hranna, “grass nouns”). Even so, number may be marked on verbs and adjectives. Here are some examples:

  • nerro chak “silent foal” or “silent foals”
  • nerro chaki “silent foals”
  • Nerro chaka. “The foal is silent.” or “The foals are silent.”
  • Nerro chaki. “The foals are silent.

With the third sentence there, the plural interpretation is much easier if, for example, there were a pen full of foals (in fact, there the singular version of the verb is preferred). The idea behind the lack of number for inanimate nouns is that many of them are, in fact, mass nouns. Those that aren’t are usually inanimate for other reasons (e.g. because they sound like another word that’s inanimate, or because in the older form of the language they ended in a consonant, or their meaning has changed over time, etc.). And one way Dothraki speakers have individuated certain inanimate nouns throughout the history of the language has been to make them animate (so you often see pairs of words that are identical save for their class membership).

I think that just about settles the issue of “number”. Now for “numbers”.

A while back, the numbers 1 through 10 were released over at the Making Of blog (you can see that in .pdf form here). Here they are again (so you don’t have to go searching):

Number Dothraki Number Dothraki
1 at 6 zhinda
2 akat 7 fekh
3 sen 8 ori
4 tor 9 qazat
5 mek 10 thi

Beyond the numbers 1 through 10, the number system is fairly combinatorial. To form the teens, you add the digit to the front of thi, as shown below:

Number Dothraki Number Dothraki
11 atthi 16 zhindatthi
12 akatthi 17 fekhthi
13 senthi 18 oritthi
14 torthi 19 qazatthi
15 mekthi 20 chakat

You’ll notice that there are two irregularities in there: the numbers for 16 and 18. The original numbers were, certainly, zhindathi and orithi, but since every other number in the teens is stressed on the penultimate syllable, the th was geminated so that the stress patterns of 16 and 18 would match the rest.

Also, the last item there is a bit different. There’s a unique lexeme for 10 (thi), but after that, the numbers in the tens form a pattern, with either ch- or chi- being prefixed to the numbers 2 through 9. Here they are below:

Number Dothraki Number Dothraki
10 thi 60 chizhinda
20 chakat 70 chifekh
30 chisen 80 chori
40 chitor 90 chiqazat
50 chimek 100 ken

You also see the Dothraki equivalent of 100 above. This leads to the next round of numbers: the hundreds:

Number Dothraki Number Dothraki
100 ken 600 zhindaken
200 akatken 700 fekhken
300 senken 800 oriken
400 torken 900 qazatken
500 mekken 1,000 dalen

A couple notes here. What I have written as akatken and qazatken sometimes comes out as akathken and qazathken. Though written as a single word, these are two word compounds (or at least started out as two word compounds), but, like the teens, they’re fusing. The late fuse means the words aren’t subject to the spirantization that affected Dothraki words in the past, but old habits die hard, meaning that you’ll often here akathken for akatken, etc. The pairs are in free variation. Somewhat less common (but nevertheless present) is senhen for senken.

Beyond 900, the numbers are, indeed, two word compounds, so 2,000 is akat dalen, 3,000 is sen dalen, etc. The largest unit is yor, which is one million, though it tends to be used more often as yorosor, which means…basically, some huge number (like a jillion in English). It seems doubtful that there would be a practical use for yor in Dothraki, unless they started dealing with the Bank of Braavos.

For in between numbers, the connector is ma. So, for example, 21 is chakat m’at, and 2,431 is akat dalen ma torken ma chisen m’at. The rest should be self-explanatory.

Two other comments about numbers. Or wait. Three other comments about numbers; my bad.

First, I made an executive decision early on that the Dothraki would have discovered the concept (but perhaps not yet taken full advantage) of zero. The word for zero is som, which comes from the word of the same form which means “absent” or “missing”.

Second, when applied to noun phrases, the noun may be realized in the singular or plural. The plurality is optional, since the number itself indicates plurality. A couple examples:

  • fekh khalasar “seven khalasars”
  • sen gevesi “three moles”

Finally, I couldn’t leave numbers without talking about ordinals. Ordinal numbers work quite differently in Dothraki. Forming an ordinal is simple enough: one adds the agentive suffix to a digit (or the last number in a sequence). Here are the ordinals for 1 through 10:

Number Dothraki Number Dothraki
1st atak 6th zhindak
2nd akatak 7th fekhak
3rd senak 8th orik
4th torak 9th qazatak
5th mekak 10th thik

Each of these are animate nouns (and, in case you’re curious, they’re used to stand in for either an animate or an inanimate noun), and mean something like, “the first one” or “the fifth one”, etc. When applied to a noun (e.g. to say “the fifth horse”), the ordinal number is placed in the genitive and put after the noun it modifies. Some examples are shown below:

  • hrazef mekaki “the fifth horse”
  • diaf qazataki “the ninth skull”
  • darif chitor ma senaki “the forty-third saddle”

There you have it! Anything and everything you might possibly have wanted to know about numbers in Dothraki. Seems kind of dry to me, since I’m not really much of a numbers guy (or a math guy [or a science guy]), but, hey, there it is! Now you can give a number to everything you see—in Dothraki!


  1. When a noun phrase that contains more than one entity is realized as singular (like fekh khalasar) do verbs still conjugate with the plural?

    Speaking of things you might find dull…What kind of arithmetic knowledge do the Dothraki have? Do they have words for addition, subtraction, multiplication, division etc? How about fractions? I imagine even iron age people would find a need for at least “one half”, “one third” and “one quarter” and perhaps even more.

    1. 1. Yes, I would still put them in the plural, since the plurality is obvious.

      2. If the Pirahã can get by without numbers at all, I’m pretty sure anyone can do anything. ;) But there is a number for “half” that I forgot to mention (I was going to)! The word is sachi (inanimate, class B). For anything smaller, you’d just use saccheya, which means “part”.

  2. Given the way you’ve described Dothraki ordinals, is there an off-by-one confusion possible when referring to times in the future? E.g., is tomorrow the first day from today or the second?

    If they have zero, today might be the “zeroth day”, and the sequence would be: today (0th day), tomorrow (1st day), two days from now (2nd day), &c., &c.; this would be consistent when communicating with others. But given your translation of zero as “‘absent’ or ‘missing’” it feels implausible that “today” would translate as “the absent day”; “1st day” seems more likely and there should be some confusion when making schedules across language lines.

    1. The words for “today”, etc. aren’t related to numbers. The word for “today” is asshekh (related to “sun”); the word for “tomorrow” is silokh (related to the verb “to follow”); and the word for “yesterday” is oskikh (related to the word “prior”).

      I’ve never seen a spoken language that uses numbers to refer to days (as in yesterday, today, tomorrow, etc.), but ASL does use embedded numbers when referring to days, weeks, months and years. In all cases, the series begins with 1, not 0. It’d be odd, I think, for a language to use “zeroth” in their day-related vocabulary.

  3. The idea behind the lack of number for inanimate nouns is that many of them are, in fact, mass nouns

    In what sense? By nature? By form? By how they function? It seems to me there are probably more inanimete than animate nouns in Dothraki. Instinctively I’d expect the natural mass nouns to be rather limited subset, though that seems to be at least partially just an error of limited perspective.

    Those that aren’t are usually inanimate for other reasons (e.g. because they are animate[…]).

    Inanimate because they are animate?

    1. There are, indeed, many more inanimate nouns than animate in Dothraki. But I think all mass nouns are inanimate. There may have been a misinterpretation there. That is, based on what I wrote, I think you thought I meant, “Most mass nouns are inanimate; some are animate”, when in fact I meant, “Most inanimate nouns are mass nouns; some are not.”

      I think this also led to the mistake I made that you caught (again! Nice job!). I didn’t mean they’re inanimate because they’re animate (HA! Can’t believe I wrote that…): I meant they’re countable because they’re animate (even though they’re in the inanimate noun class, the main word for horse—hrazef—is a countable noun because it’s…well, animate: it moves around, we interact with horses singly, etc.). I think I’m just going to remove that little bit…

  4. Sorry for being so scarce as of late, and missing the house motto contest. The last two weeks have probably been the worst weeks of 2011 for me.

    Thank you for all the useful information that has been posted in my absence!

    Now, for questions: On your example for number on foals, you used narro and Narro. There is obviously some kind of difference I am missing caused by one being capitalized, and the other not capitalized. What is the difference?

    Next, one could write:

    fekh hrakkar (inanimate) ‘seven white lions’
    fekh hlizif (animate) ‘seven bears’

    Are these both correct as written? Or am I missing a plural marker on the animate noun (which I can’t find anything about in the online grammar) And if I am understanding you correctly, there would never be any kind of plural marker on an inanimate noun?

    As an aside, hlizif is a really interesting word, definitely one of my favorites!

    And a big thanks for the number numbers. Number systems are of great interest to me, because so much of my professional and personal life revolves around numbers and math. I hope to have this information in the dictionary soon (the dictionary is undergoing a major renovation right now to fully bring it in line with the wiki).

    As far as bases of number systems go, I am sure you know Na’vi is base 8. But there is another interesting conlang– Kzinti (created in support of Larry Niven’s ‘Man-Kzin wars’ series, which features a felinoid race, the Kzin), that is also base 8. I guess it has to do with felines having four ‘fingers’ on each paw.

    1. It appears that the plural endings were only listed in the table at the bottom of the noun case page in the wiki so I added some text about it under the nominative headline as well.

      Basically the animate plurals get the suffix /-i/ or /-si/. So it should be fekh hlizifi.

      1. I forgot about the part that David wrote that when you use a number you don’t need to use a plural marker since the number itself makes it clear that it’s plural. But if you want to add the marker than that is how you do it.

    2. Regarding nerro, the difference there is just that one is a full sentence; the other a phrase. The full sentences are properly capitalized and punctuated; the phrases are not.

      And there’s a great wide world of conlangs outside of those people have heard about. There have to be over a hundred that use a number system of a different base. One of mine (Sathir) uses base-6; Ithkuil uses a base-100 system. There are tons of others. I can’t think of any names off hand, but you can deduce which use bases that are at least less than 10 by examining the number lists here.

  5. In the examples listed in one of the tables above, there happens to be a mistake in the transcription of the arabic word for ‘(two) mothers’. The correct word is ‘waalidatan’ as ‘waalidaan’ means ‘(two) fathers’ (the ‘t’ denotes feminine gender).

Leave a comment

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.