User:Silence/PIE

Indo-European numerals

One
Proto-Indo-European had at least two unrelated words for the concept of "one". The most widely represented one is *oi-, most commonly suffixed as *oi-no-. This was the word used in counting, and denoted isolation or uniqueness. The other root meaning "one", on the other hand, *sem-, denoted unity or wholeness, and was used when multiple things were considered or joined as one, or were considered the "same" somehow. As a result, while *oi- has derivatives with meanings such as "solitary" and "single", *sem- has derivatives with meanings such as "similar" or "simultaneously".

A common hypothesis for the existence of two prominent Proto-Indo-European words for "one" is that *sem- is the older, original root for all senses of the word, while *oi- is a later root, derived from the ablaut grade of the pronoun root *i- ("it"). *sem- is a more irregular, archaic Indo-European adjective formation, and there are numerous compounds of *sem- of ancient constructions. Additionally, while the numeral "one" is represented by derivatives of *oi-no- in almost all Indo-European languages, *sem- is used instead in three: Greek, Armenian, and Tocharian. As these languages are very geographically distant from each other, the usage of *sem- suggests an ancient holdover, whereas the broad use of *oi-no- in a geographically contiguous region suggests a popular later development.

Both *oi- and *sem- have numerous derivatives in all Indo-European languages.

Aside from *oi-no-, other suffixed forms of *oi- included *oi-uo- (as in Avestan aeuua-) and likely *oi-k(ʷ)o- (as in Vedic eka-).