I read this in a random forum:
"Words neuter in Latin become masculine in Spanish"
(For instance "vāsum" = el vaso)

Could it be some patterns making predictable the gender from Latin to a descend language?

I always though it was a little random, in the sense that a word is "delatinized" becoming a part of a descend language, and the gender is chosen or changed by the popular use according to the sound of the ending.

Are there patterns for this, at least in some Romance languages?

  • 2
    Why do you think that it's random? There are certainly exceptions, but in general the gender is quite consistent between Latin and, say, French or Spanish. – brianpck Dec 7 '19 at 16:14
  • In French, I don't have the impression. Do you have rules about genders + ending in Latin, and their equivalent in Romance languages? I should edit the question to add "if you have some rules, I would be interested". In Spanish and French, both from Latin, and genders aren't consistent at all between those 2 languages. – Quidam Dec 7 '19 at 17:32
  • For instance: Latin dactylus, el dátil (masculine, Spanish), il dattero (masculine, Italian), la datte (feminine, French). – Quidam Dec 8 '19 at 5:23

"Words neuter in Latin become masculine in Spanish"

This is generally correct!

In Latin, the most common type of masculine noun and the most common type of neuter noun look almost identical. They're only distinguishable in the nominative singular and the nominative/accusative plural.

In the branch of Romance that would eventually become Spanish, the nominative case disappeared altogether—so the neuter looked exactly like the masculine in the singular.

What's more, the neuter plural looked exactly like the feminine, for reasons that go back to Proto-Indo-European and the origin of the feminine gender in the first place!

So because of these very regular changes, the neuter gender vanished into oblivion, with its words merged into the masculine. Occasionally you'll find Romance words that are masculine in the singular but feminine in the plural, like Italian uovo "egg", uova "eggs"—these are the only surviving remnants of the Latin neuter. (In Spanish, this particular example got regularized by analogy, and is now a perfectly regular masculine: huevo, huevos.)

In the comments, you mention Latin dactylus (m) > French datte (f). This sort of change happens sometimes, especially in Gallo-Romance (which doesn't have the same distinctive masculine/feminine endings that Ibero-Romance does). But it's an irregularity, and not a very common one; the majority of words kept their original gender from Latin into Romance, and the ones that didn't can usually be explained by analogy.

  1. If the loss of the neuter case is phonetically conditioned and regular, than the gender should regularly follow the same pattern, yes.

  2. It's overall highly unlikely that a word spontaneausly changes gender, and down right impossible for all to change at once so that a whole genus can be lost; just as it's unlikely that a genus is just abolished, so that there's no opportunity for reassignment.

  3. For some (many?) words, gender is even reconstructed for proto-indo-european. However, I seem to have misremembered my prime example: cp Lat m. sol, Ger f. Sonne, PIE n. *soHwl--although

    The reconstruction is far from certain. Contrast the runner up instead:

    • Lat f. luna, Ger m. Mond, PIE f. *lowksneH, m. *meHns, Ru f. луна́ (luná), m. ме́сяц (mésjac)

    Note: "The original Proto-Indo-European lexeme for "moon" was apparently *mḗh₁n̥s, and *lówksneh₂ may have originated as a later poetic synonym" [wiktionary]

    • also cognate: Lat m. mensis, G m. Monat
  4. What's more, early PIE is supposed to have started with a two way system understood as animate / inanimate (prime example: neuter fire, animate ignition). The Female and Neuter distinction is thought to have developed later from the inanimate genus in PIE (which sounds a bit cruel, who knows). So next to neuter sun there's masculine day, and female night; As they say: opposite like day and night. dies (not cognate to day, really?) is either f or m, and its deemed a backformation from diem. deus is part of the same vrddi paradigm.

By the way: Slavic luna looks a bit like a loanword at first sight; the only word that is supposed to confirm the presense of s in the position where it could become lost in the slavic root rather means "stars" in Old Prussian, and hinges on a somewhat rare suffix. Regardless, I meant to say, gender of loanwords is often phonetically conditioned rather than addopted; viz -(k)a as chiefly female suffix in Russian, or G f Sense "scythe" bearing the same supposed suffix (incidentally f. sickel is used to describe the moon; I didn't mean to recount astronomy, I swear!), G -e (e.g. die Sonne) frequently agreeing with the female Gender as much as Lat -a (e.g. lacuna, laguna, but m. lacus); whereas e.g. -er is frequently masculine agentive, cp lantern, f. in Lat, Fr, Ger, ... but "[lanterna] itself a corruption of Ancient Greek [m.] λαμπτήρ (lamptḗr, “torch”) (see lamp, λάμπω) by influence of Latin [f.] lucerna (“lamp”)", G m. Lampinion, but G f. Lampe, etc etc.

Conclusion: If you find evidence of a significant share of reflexes that do not agree with the tripartite alignment, then stop the press, because you might have found remnants of a dialectal division, an isogloss, or loanwords, etc.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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