Masculine/Feminine Nouns

When new words are introduced into a language such as Portuguese, how is the gender of the noun decided? For instance, in Portuguese, how does “laptop” become “um laptop” rather than “uma laptop?” Is there some national laguage board that assigns gender to new nouns, or is it some mass consensus? I don’t expect the answer to make learning this maddening language any easier, but it would be nice to know.

In general, it is by analogy to existing nouns. For example, in French, any new word with the suffix -tion (such as mondialisation, globalization), -ité (such as connectivité) or --ique (such as réseautique, computer networking) will be feminine, because nouns with those suffixes always feminine.

Beyond that, most of the time they’ll end up being masculine (le marketing), though there may be exceptions.

If the word is adopted by a terminological authority such as the Office québécois de la langue française, they will specify a gender when adopting the word officially.

Here is one obvious exception. Nearly 40 years when I bought tickets for hockey games, I noticed that the Montreal Forum was owned by “La Canadian Arena Company”. Leaving aside the question of why this false French name was chosen, there is also the question of why “La company”? Well, it is by obvious analogy with “la Compagnie”, which is quite evidently the source of the English word.

Beyond that, if you ask a French speaking person what the gender is of a nonsense syllable, they will generally have an intuition and this intuition is fairly stable.

Most borrowed words and other neologisms, aside from the cases described by Matt, will be masculine.

Right, Hari. If a proper noun of that kind is used in French without translation (Company, Corporation, Group, Society, etc. etc., etc.) the gender will be that of the cognate French noun: le Really Useful Group, le British Council, la Canadian Widget Company (or perhaps la compagnie/la société Canadian Widget).

There are also a lot of rules about the gender of proper nouns that don’t take articles (such as most city names), many of them only partially coherent, so you’ll have a part of town named Vieux Montréal (masculine) but see a headline that says Montréal la pacifique (feminine).

That’s true even in abbreviations: le FBI (bureau) but la CIA (agence).

Spanish also makes heavy use of cognates, an instinct that we can’t really explain, and after-the-fact explanations. I’m sitting at a computer which in Spain would be an “ordenador” (m) but in Latin America a “computadora” (f). So you get a word derived from French in Europe and one derived from English in LAR, and in Europe it’s an “aparato que ordena” (an apparatus that sorts), and “aparato” is masculine, whereas in LAR it’s a “máquina que computa” (a machine that computes), and “máquina” is feminine. And, to complicate it even more, you can have… “un computador” - which again gets explained as (an apparatus(m) that computes). I haven’t heard of “una ordenadora”, but just give us enough time and someone will start using it.

For machines that fly? El avión, el aeroplano, el montgolfiero, el globo aerostático, el helicóptero, el autogiro (all m)… but la avioneta(f).

For other things it’s easier, but with things that don’t really have a “natural gender” you get soups like the one above quite often. And then we get to drive people crazy further by having words which apparently differ only in gender but which mean very different things: el rayo (the bolt of lighting, the beam of light), but la raya (the line, the border).

What puzzles me is the gender that Urdu and Hindi attached to loanwords from Persian. Since the Persian language is absolutely gender-free, and every word in Hindi has to be either masculine or feminine, and Persian contributed many thousands of loanwords, there has been a lot of arbitrary gender assignment in Urdu and Hindi. I have not noticed any pattern to this, although there’s a possibility it proceeded by analogy-- a Persian word may have been gendered to match the native word that meant the same thing. Or maybe they flipped a coin to determine the gender for each loanword. The cognates between Persian and Indo-Aryan are not always obvious, so that probably did not influence the process much.

To take a few random examples:
ātish ‘fire’ – assigned feminine in Urdu and Hindi (perhaps by analogy with native āg ‘fire’, which is feminine)
anār ‘pomegranate’ – assigned masculine
bād ‘wind’ – assigned feminine
charm ‘leather’ – assigned masculine
chashm ‘eye’ – assigned feminine (by analogy with native feminine ānkh ‘eye’?)
dānt ‘tooth’ – assigned masculine
deg ‘pot’ – assigned feminine
*farsh *‘carpet’ – assigned masculine
*fath *‘conquest’ – originally masculine in Arabic, reassigned feminine in Urdu, after being gender-laundered by passing through Persian in between.
gul ‘flower’ – assigned masculine
hast ‘being, existence’ – assigned feminine

If anyone can find any rhyme or reason to these, you’re a genius.

What I find a little odd is the way Urdu applied gender reassignment to many Arabic feminine nouns ending in -a. Since -a is a masculine ending in Urdu, they became masculine. Conversely, some Arabic verbal nouns which were masculine in Arabic were converted to feminine in Urdu. It’s quite a mixup. Most Persian nouns ending in -a likewise became masculine. Since the Urdu feminine ending is -i, all Persian nouns ending in -i automatically were assigned feminine. What I wonder about is the nouns ending in consonants that don’t get an automatic gender. It means somebody, when first incorporating these words into Urdu grammar, had to actually choose one or the other gender, perhaps arbitrarily.