From Latin -a, from Proto-Italic *-ā, from late Proto-Indo-European *-(e)h2, from Proto-Indo-European *-h2, a suffix indicating collective and abstract animate nouns, from Pre-Proto-Indo-European *-h2, a suffix indicating plurale tantum.
In Pre-Proto-Indo-European, a suffix *-h2
was used to form plurale tantum
nouns out of collective nouns. (As a somewhat inaccurate analogy to help the reader understand the function of *-h2
to the ancient Pre-Proto-Indo-European speakers, imagine the collective noun "water" becoming plurale tantum
"waters" by way of a plural suffix -s
(e.g., "the waters of Babylon" of Psalm 173:1 ESV); the role of the -s
is akin to the role of Pre-Proto-Indo-European *-h2
). This was attached to the vowel stem *-e-
to form *-eh2
, and *-h2
. The four different endings all to express a single idea of plurality would not last. But now we must discuss the historic Proto-Indo-European method for feminizing words.
In Proto-Indo-European *sor
"woman" was added as a suffix to animate nouns (retained in Anatolian languages and partially survived as a fossil in a smattering of non-Anatolian words).
Linguists do not agree as to why *-sor
, when used as a feminine marker, was slowly supplanted by the *-(e)h2
collective suffix after the Anatolian languages broke off from the rest of the Indo-European community. Recent theories suggest the suffix played an "individualizing" role. Under these lines of thinking, the collective/abstract suffix *-(e)h2
referred to a subset of larger mass nouns (compare in English plurale tantum
"glasses," formed from the mass noun "glass"). Gradually, dependence on *-(e)h2
to refer to subsets of broad categories gave the suffix a new sense of individualization - speakers were depending on the suffix to specify individual items out of larger groups. Once an uncommon suffix, by the late Proto-Indo-European period *-(e)h2
enjoyed heavily use to specify individual animate objects, and so speakers contrasted the new animate ending *-(e)h2
against the old animate ending *-s
acquired a new meaning, not one of plurale tantum
but a meaning of "the other object."
In compelling support of this theory, there is a word a word that looks to be a relic from the transitional period from collective/abstract suffix to a feminine suffix (the familiar -a
in Core Indo-European). *h2u̯idhéu̯eh2-
"widow," animate in Anatolian but feminine in other Indo-European languages, looks built off of *h2u̯idhéu̯o-
"belonging to the one fatally struck." Thus the original meaning of a widow comes from the time when the suffix played an individualizing role: "the one belonging to the fatally struck." Later, the word simplified to "the bereaved." After the Anatolian branch broke from the Indo-European trunk, the suffix was fully feminized as "the female bereaved." See Luraghi (2011), Tichy (1993), and Melchert (2014b) for a broader discussion.