A Hundred

[jack]

So apparently a hundred used to be a hundred and twenty.

According to etymonline "hundred" came from Proto-Germanic "hunda-ratha", ultimately from Proto-Indo-European "km-tom", a shortening of "dkm-tom-", a suffixed form of "dekm-" meaning ten. Latin "centum" (where roman numeral C comes from) came from the same word.

But it sounds like in Proto-Germanic, the word mostly meant *twelve* tens. And then over the whole medieval period in Germanic-language speaking areas, it was used to mean "120" for some goods and "100" for others. Wikipedia says that "thousand" was also used meaning "1200".

Some sources delved through a bunch of medievel documents looking for examples and it sounded persuasive to me. One emphasised that it always seems to be twelve tens, people didn't seem to count twelve twelves. It seems like "120" and "100" were somewhat standardised, but there were also regional variations or a tendency to use similar terminology for any round number around that size.

English eventually started distinguishing these as "long hundred" and "short hundred", and surprisingly late (1800s?) parliament ordered that "hundred" be standardised as 100. Long and short persist in measures like "long ton" and "short ton", being 2200lb or 2000lb. Apparently based on 20 long hundredweights or 20 short hundredweights. Long ton or british ton is conveniently almost exactly 1000kg, which people now use as the metric tonne. North Americans may still use "ton" as a short ton.

I can't find any confirmation where "120" started. I assume that the PIE word meant "100". Does anyone know more?

https://en.wikipedia.org/wiki/Long_hundred (Especially look through the citations to short academic PDFs eg https://journals.socantscot.org/index.php/psas/article/view/9477)
https://en.wikipedia.org/wiki/Long_ton
https://en.wikipedia.org/wiki/Hundredweight

Comments

[redbird]

This reminds me of the idea that some numbers in old texts mean "a lot" rather than a specific number, and they aren't the sane numbers we use that way now: "40 years" in the Hebrew bible may just be "a long time," the way we might use "a thousand years" or "a million people." Also of the different (mostly historical) meanings of "foot," "cubit," "mile," and "pound."

[ewx]

Indo-European Linguistics (James Clarkson) discusses the reconstructions of the forms for small multiples of ten:

> [I]f we compare the words for 'ten', 'twenty', 'fifty' and 'hundred', it is possible to refine our reconstructions to make the system more consistent. All of the forms can be related to a skeleton *dk'mt-, which may have been a derived form of *dek'm 'ten' with the meaning 'decad' (the original meaning survives in the Sanskrit word daśát- 'decad'). [...] It is possible to build the numbers 'twenty' and 'fifty' through the reconstruction of original collocations, such as 'a pair of decads', with subsequent assimilation of consonant clusters and compensatory lengthening. The most questionnable aspect of this reconstruction is the supposition of a genitive plural *dk'mtóm meaning 'a group of decads', and by extension 'ten decads, a hundred'. However, in some Germanic languages the outcome of PIE *k'mt-om does not mean 'hundred' but 'hundred and twenty', as Old Norse hundrað. This suggests that the fixing of the meaning of *k'mt-om may in fact have been late.

If I understand this right, while the reconstruction of *k'mtóm as the ancestor of hundred/centum/ἑκᾰτόν/etc in the descendants languages is secure, its precise meaning in PIE and its derivation from *dék'm, or at least the mechanism of that derivation, are not so certain.

It's not a particularly satisfactory argument, since using descendants of *k'mtóm for 120 (or any other non-100 value) apparently doesn't happen in any other branch. So why is Germanic the only branch where the anomaly can be seen? Couldn't instead *k'mtóm have meant 100 in PIE but Proto-Germanic then gone its own way for some reason (perhaps somehow associated with the addition of the -ratha suffix?)

[ewx]

On reflection it's not quite true that the anomaly is restricted to Germanic languages: a Roman military centuria apparently wasn't always exactly 100 men.

[redbird]

That doesn't seem as weird/improbable as if a usage turned up only in English, Icelandic, and Russian: it just requires a shift in meaning in Proto-Germanic before that diverged into a bunch of languages including English.

Also, linguistics is weird. Wandering randomly around Wikipedia yesterday, I found the assertion that all Algonquian languages are tonal, but reconstructed Proto-Algonquian wasn't, and this feature developed independently in each branch of the family. (I didn't try following the reference links to find the evidence for this claim.)

[jack]

Ah, cool, thank you! That is helpful to hear.

etymonline listed a bunch of other words probably from the PIE word "source also of Sanskrit satam, Avestan satem, Greek hekaton, Latin centum, Lithuanian šimtas, Old Church Slavonic suto, Old Irish cet, Breton kant". I guess one could see if any of those had a history of 120 or other values than 100.