When I was learning a little Latin and less French in middle school, I learned that gender was nothing but a grammatical category, having little or nothing to do with sex. Why should the word for “sea” be feminine in French but masculine in Spanish? No good reason, it just is. It tells you absolutely nothing about how people in France and Spain think of the sea. So what if adjectives in Latin (and French and Spanish and Greek and so forth) have to “agree” with the nouns they modify? That’s merely a grammatical quirk. We shouldn’t call them masculine and feminine (as well as neuter in Latin, Greek, and German); we should just say first, second, and third gender, just as we talk of first and second conjugation for Latin verbs and first and second declension for Latin nouns. Everything is nice and orderly and well-structured; that’s one of the reasons we study Latin in school, after all.
But of course it isn’t that simple.
Go read the excellent essay by Tom Henrickson, “Gender Diversity in Greek and Latin Grammar,” even if you have no knowledge of or interest in either language. It show us that the concepts of non-binary gender, gender fluidity, etc., aren’t as new as we think. Even back when I was in middle school I knew that it was an oversimplification to say that gender does not equal sex and that the names of the genders are arbitrary; if you learn, say, a word for “seamstress” in a gendered language, you feel entirely confident that the word must be feminine. (That’s until you learn that the German word for “girl” is neuter, not feminine, but that’s another story.) When you learn that Swahili has 18 genders, you become convinced that numbering noun classes makes much more sense than words like masculine and feminine.
But why does everyone who uses the traditional European terminology list masculine before feminine, eh?