I read that proto-Indo-European had words for numbers, just like modern languages. That did not surprise me. What surprised me is that they had words for numbers up to 100 or even 1000. I'm just curious, what did they use such numbers? Maybe they used them when they were building stuff, counting years (maybe they used all the #s), and looking into nature. Any guesses?

Note: the surprising thing is that some ancient languages have small number sets, which may be capped at a number like 5, 10, 20, or somewhere in between.

  • 1
    Asking "why" is rarely productive when it comes to vocabulary.
    – curiousdannii
    Commented Dec 14, 2019 at 13:06
  • I think it should be “what did they use the numbers for.” Commented Dec 14, 2019 at 13:42
  • Counting years seems unlikely to be the reason, as many ancient societies didn't really keep a years count, instead even the Romans named years by the consuls they had, and only historians used the aUc numerical system. On the other hand, what about counting livestock or marketable items? I think PIE is often considered to be relatively young as far as proto-languages go, so agriculture and commerce may already have been established by the time it splits into subfamilies.
    – LjL
    Commented Dec 14, 2019 at 15:14
  • It is interesting about what they used "kmtom", or "ten tens" for. Maybe they counted animals or people using such numbers. Commented Dec 14, 2019 at 19:55
  • @LJL surely they counted the annual passage of time that many moons ago. Perhaps that's why moon is masculine, animate. I had started to wonder.
    – vectory
    Commented Dec 15, 2019 at 15:42

2 Answers 2


There is a presumption in the question, that the language had words for "100" and "1000", which is possible but not guaranteed. In contemporary English, we certainly have a word "14", but it obviously derives historically from a phrase composed of "four" and "ten". The reconstruction of the decades in PIE similarly indicates that there was no word for 30, there was a construction (one that was soon lexicalized especially in the case of 20: compare Sihler's reconstruction *wīḱm̥t and Beekes' *du̯idḱm̥ti). In the case of 1000, there is no uniform reconstructed word. As for 100, there is a nontrivial similarity in Beekes' reconstructions *déḱm̥t '10' and *dḱm̥tóm '100', and it has been conjectured that 100 derives from a multiplicative construction 'ten tens'. Vedic Sanskrit had big numbers, which did not correspond to anything that they ever actually had to count

  • 3
    The "hundred" word seems guaranteed for (at least non-Anatolian) IE; though it probably is derived from "ten", the derivation is not of an obvious type and could hardly have arisen separately in the daughter languages. "Thousand" is indeed doubtful as Germanic and Balto-Slavic point to one reconstruction, Sanskrit, Greek and Latin to another.
    – TKR
    Commented Dec 15, 2019 at 3:34
  • The last sentence is the only one direct answer, but it's unsubstantiated. The counts which the modern interpretations of the vedas give are frequently out of proportion, though, millions of years, thousands of sons (unless that's younger layers of the scripture).
    – vectory
    Commented Dec 16, 2019 at 0:47

First, to note, any answer to this question is going to be speculation—we have no records from PIE times showing what numbers were actually used for.

But, it's been noted that PIE had a very rich and well-attested core vocabulary for animal husbandry. Words for "cattle" and "horses" were better-established than words like "trade" or "plows".

So I would speculate that the earliest PIE-speakers used numbers primarily for tracking herds of animals. When your way of life depends on your herds of sheep and such, you have a vested interest in knowing exactly how many sheep you have!

  • And also kings counting their armies!
    – curiousdannii
    Commented Dec 15, 2019 at 0:18
  • trade was based on exchange and even German tauschen has been compared to täuschen "deceive", and cattle raiding was a most heroic act, so I guess it's safe to say that most of the trade associted vocabular that they might have had was subject to ... corruption--*badum-tss*!
    – vectory
    Commented Dec 15, 2019 at 15:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.