Posted by & filed under Podcast.

This month we talk about grammatical number. What number distinctions can you make for a language (beyond singular and plural)? What do you mark for number? And how does number interact with agreement and other grammatical systems? We’ll help you with all of that.Conlangery 125 Medallion

Top of Show Greeting: Classical Latin (translated and read by Nicholas Duharte)

Links and Resources

6 Responses to “Conlangery #125: Grammatical Number”

  1. Christophe Grandsire-Koevoets

    So, I just listened to this episode and heard my name at the end, so of course I had to answer your plea for help :).

    Anyway, you were basically correct with your characterisation of French. While the English expression “one and a half days” puts “days” in the plural, the French equivalent is “un jour et demi”, with “jour” in the singular (and even in Spoken French the distinction would be heard, as the plural “jours” would –optionally– trigger liaison in that environment).
    Notice, however, how in the French expression, the “half” appears after the main noun, the whole thing being equivalent to saying in English “one day and a half” (where I would expect singular agreement as I wrote it). So it’s not so much that French considers plural to be “2 or more” rather than “more than 1”, but rather that in French you cannot really talk about fractional numbers above 1 (or rather, you do so by adding the fraction *after* the noun phrase is complete), so that nouns effectively always agree in number in integer values.

    By the way, I may have missed it, but I don’t think you’ve mentioned the so-called “inverse number” of languages of the Tanoan family (, which is a really weird, but cool way to mark number: basically in these languages, there are typically three numbers (singular, dual and plural) and nouns are grouped in classes that have an “inherent number” (which needn’t be the singular, and can actually cover more than one number). They are unmarked when they appear in this inherent number, and need an inflection to indicate that they do not appear in this inherent number. The interesting part is that that inflection is the same for all nouns, so that the same affix that effectively marks the plural for one noun can actually mark the singular for the other! Basically, number marking then is a kind a toggle, and to know what it means you need to know what the inherent number is of the noun. When I discovered this system my mind was blown away :P.

    • admin

      On the French, I was curious about that. The syntax of these fractional expressions can be interesting. In Chinese this expression would be the same 一天半 yi1 tian 1ban4 (lit. “one day half”), with the added fact that plural marking is never an option for 天 so that point is moot. However, if you have a noun that takes a classifier, “half” occurs after the classifier 一个半小时 yi1ge ban4 xiao3shi2 (lit. “one CL half hour”).

      Anyway, I wonder if this really is just down to syntax. I believe Spanish is the same as French, producing un día y media, but we might need a broader range of languages to test (specifically ones that allow number marking in a quantified phrase).

      As for the Tanoan, we did miss it. Thanks for that example!

    • Daniel

      Yeah, I was bummed they didn’t mention Inverse Number/Number toggling, it’s by far my favorite way to mark number. I use it in one of my languages, and it can be so hard to keep straight, since most animals fall under the plural by default group, I basically have to say the equivalent of “my cats is cute”, which is really hard to wrap my head around. It’s fun to play around with more fluid noun classes as well, for example, plants are plural, but food is singular, so the unmarked form for something like “berry” would be considered plural on the bush, but (still unmarked) singular when eaten, and plural again if the berry/ies were used for something non-culinary, like dye or medicine. It’s such a wacky feature and it has lots of wiggle room as far as naturalism is concerned, you can say “hey, only one small language family on earth uses number toggling, so who’s to say the way I used it is unnaturalistic :P”.

      I used this to create a “Tangibility Toggling” system alongside number toggling (historically derived from the associative plural before it fell down the rabbit hole of toggling). In the same way that each noun has an inherent number, they also (usually) have inherent tangibility. So weather, is intangible, and physical objects are tangible, and when you toggle, say snowstorm, for tangibility it becomes “snowflake” and toggling rock becomes “hardness”. That’s just the tip of the iceberg, but it is so fun to work with inverse number systems!

  2. John Hutchinson

    Good show. I’d just like to point out, though, that Wikipedia gives the pronunciation of Arrernte as /ˈærəndə/ (native [arəⁿɖə]). Arrernte is very interesting as it appears to be the most unambiguous case of a purely hieght-based vowel system. Also it’s Mokilese, not Molikese

    • admin

      Oh, I knew I had mucked that up. This is what I get for cramming for the podcast at the last minute!

  3. Sascha Baer

    Over-all a good overview of different number-markings. Especially liked the discussion on the different plurals, which I do sometimes confuse. What I found missing is a discussion on what languages do with zero/none and number. E.g. in English, zero is plural, but in French I hear it’s singular.

    Additionally, I’m pretty sure Nilo-Saharan languages do in fact have unmarked plurals as the default pretty much throughout, or at least very commonly.

    Also isn’t the verbal number thingy (pluractionality) you mentioned some native american languages have more or less tied to how many times you do the action? Which is why they tend to agree with objects – if I give something to five people, I essentially do five instances of giving a thing to someone.


Leave a Reply

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.