How it is that Proto-Finno-Ugric had the word meaning 100 (a borrowing from Indo-Iranian), but not the word for 10 (as Hungarian borrowed it from Indo-Iranian as well, but Finnish has a native word for 10)? If a language has a word meaning "hundred", doesn't it also need to have a word meaning "ten"?

  • 5
    What makes you think they didn't? There's a difference between "they had no word for X" and "we are unable to reconstruct a word for X"; none of the Romance languages kept the standard Latin word for "horse" but the Romans certainly had a word for that.
    – Draconis
    Commented Jun 4, 2023 at 19:50
  • @Draconis But words for basic numbers do not get borrowed if a language already has a word meaning them. Commented Jun 4, 2023 at 20:14
  • 3
    Not necessarily. Loanwords can displace native words even for very basic concepts, if the sociopolitical factors are right.
    – Draconis
    Commented Jun 4, 2023 at 20:18
  • 2
    Proto-Finno-Ugric is not even a universally accepted grouping. Many nowadays argue that the Finno-Ugric group never existed, and that they are in separate branches within Uralic. I don’t think there’s any really watertight reconstruction of 10 or 100 in Proto-Uralic at all, so it’s not entirely unexpected to find loan words in the individual branches. There’s *luka-, but that also means ‘count’, ‘number’ and even ‘read’ (not dissimilar to Latin legō and cognates), and the specific meaning ‘ten’ can easily be secondary. Commented Jun 4, 2023 at 20:32
  • 1
    @FlatAssembler it's an ordinal rather than a cardinal, but English borrowed "second", displacing native "other" (which was obviously retained in a more specialised sense). In the cardinals English also borrowed "dozen" alongside native "twelve", granted it hasn't displaced the original, but it was still borrowed
    – Tristan
    Commented Jun 5, 2023 at 12:36

1 Answer 1


Your premise about necessary relations for number terms is in error. There is no word in English for 987, but there is a word for 1000. The number "100" is not a "basic word", nor is "twenty". Once you go above about 7, you are in the realm of counting numbers, that is, items that are the process of counting things rather than perceiving quantitis (you can directly perceive 2 cow or 3 cows, but not 19 cows).

There is a root, *luke which reconstructs to FU and is used in a number of languages for "10". By majority-rules reasoning, it most likely meant "count". FU *śata does most often mean "100" so we would naturally say that it meant "100" in FU. FU has a murky root along the lines of *witte which means "5" or "10". In other words, we have words, and probable meanings. The dominant meaning "100" for *śata can be explained if there originally was no such concept in the proto-language. The split in "10" can be explained if there were weakly-conventionalized word choices for that number, which got firmed up differently in various subgroups of FU. If and when one settles on a decades-based theory of numbers – later in the development of the daughter languages – it is natural that a word basically meaning "lots" will be fixed at "100", which is in fact Lahman's theory of the Indo-European word "100".

There is a paper "The Number System Based on Six in The Proto Finno-Ugric Language" that explores the development of numerals in FU with an eye for discerning a more original system based on "6".

  • The first paragraph is a bit of a red herring. The question doesn’t claim that to have a word for a number, a language must necessarily have a word for each individual number below that number – that would be true of virtually no language on earth. What does hold true in every case I’ve ever come across is that languages whose counting is based on exponents (e) of a exponential bases (B), if there is a word for Bᵢᵉ, there will also be a word for Bᵢˣ, where x is any number smaller than e and Bᵉ <= Bᵢ₊₁. In other words, if English (decimal; base 10 until 1000, then base 1000) has a word for → Commented Jun 5, 2023 at 13:17
  • → 1000³ (milliard/billion), it will also have words for 1000² (million) and 1000¹ (thousand); and if it has a word for 10³ (thousand), it will also have words for 10² (hundred) and 10¹ (ten). The only exception to this that I’m aware of are hyperbolic words representing unfathomably large numbers that don’t really fit into the normal counting system, like asaṃkhyeya (10¹⁴⁰). Commented Jun 5, 2023 at 13:24

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.