Amia Srinivasan has a very interesting discussion of pronouns at the LRB; she provides a nice summary of the (continually failing) attempts to rationalize pronoun use in English, which have been occurring on and off for the last two centuries. I'm not convinced the summary quite captures, however, the problem which has given rise to this, which boils down to the fact that English in practice has involved conflicting rules of thumb, e.g.:
(1) There are four nominal genders: masculine, feminine, common, neuter.
(2) There are three pronominal genders: masculine, feminine, neuter.
(3) Gender follows sex.
(4) Neuter should not be used for persons.
(5) Pronouns should agree in gender and number with their nouns.
(6) 'He' is masculine.
(7) 'They' is plural and agrees with all nominal genders.
(8) 'She' agrees with common nouns that are allegorical representations.
(9) 'Man' is both masculine and common.
(10) Neuter should be used for animals except where sex is particularly important.
The jumble of rules of thumb guarantees asymmetries, and asymmetries guarantee conflicting possible extrapolations to any cases in which we are trying to communicate something new and not trite. The reason for this, of course, is that languages are not constructed out of rules of thumb but out of habits of use, and the rules of thumb are regularities identified after the fact that become used explicitly, due to the importance of regularities in both pedagogy and communication. The matter is complicated further by the fact that we use exemplars, not just rules, to maintain regularity, and Latin served and in a more limited way still serves as an exemplar for grammar in formal registers of English. No rule fix is ever going to solve all the asymmetries with respect to gender, and any attempt to solve the problem by that route tends to be easily recognizable as the artificial jargon of a few rather than a solution that arises naturally and organically in the process of the many using the language.
Of the latter, only two violations of widely respected rules of thumb have gained and maintain dominance with respect to pronouns and common or indefinite gender in the singular: for formal registers, common he, due to the influence of Latin as an exemplar; for informal registers, singular they. Were there a sharp difference between formal register and informal register, they would no doubt merely split the difference. But English has no such sharp difference; we have no English Academy, and despite the phrase "the Queen's English", nobody even in England actually models their formal speaking on formal practices of the court. We have situations that strongly require formality and situations that strongly require informality; this requirement is the only thing that maintains any kind of formal/informal distinction at all. So there has been a centuries-long struggle between common he and singular they in all the cases in between the extremes. Common he has always had the advantage that grammarians are latinate (English grammar, as a field, is in origin an imitation of Latin grammar); singular they has always had the popular advantage. The latter has been steadily winning, for two reasons: Latin has receded as a formal exemplar, thus weakening one of the pillars holding up common he, and the 'gender follows sex' rule of thumb has become increasingly dominant. I have no idea what has led to the latter; sex-based rules of thumb have become steadily less important in other areas of English, including nouns, but with pronouns at some point it began to swamp other rules.
Srinivasan, of course, is talking about this in order to talk about new gender pronouns, things like xe and hir. In talking about this matter, however, she unfortunately conflates two distinct things, grammatical conservatism and cultural conservatism.
The grammatical conservative's argument is that the new pronouns are jargon imposed for reasons external to the language, involving neither an extrapolation from regularities already in the language nor a diffusion from major exemplars. This doesn't rule out all pronominal innovation; grammatical conservatives have long been split over singular they, which is supported by both regular extrapolation and normative diffusion from authors like Jane Austen. And s/he, for instance, is an extrapolation of regularities for writing. Xe and hir and the like, on the other hand, are simply imposed; they require everyone who speaks the language to learn a completely new vocabulary, violate the rule of 'gender follows sex', and not only do not extrapolate the regularities already existing but posit a completely new rule of 'grammatical gender follows gender identity'. They're like the notorious Latinx, which is an obvious jargon-imposition on the Spanish language. Nobody would look at the actual regularities of Spanish and say, "Ah, an obvious thing to do for ambiguous or common cases is to use -x instead of -a and -o." What they might do is what some actual Spanish speakers did do: start using -e rather than -a or -o; -e, while not exactly common, is already used in some common-gender nouns. So that would at least be an extrapolation; it's a Spanish solution to a Spanish problem. Latinx has nothing identifiably Spanish about it at all. The grammatical conservative's point is that xe is not an English solution to an English problem.
Srinivasan does a bit of tapdancing around this, footwork that is obscured by the conflation of grammatical conservatism and cultural conservatism. Noting that one response to prior attempts to rationalize the English pronominal system was that people are free to invent their own languages, she says, "The Enquirer editors were conveniently forgetting that all languages, and all the words in them, are invented." No; you will look in vain for the inventors of the English language. All languages are artificial, in much the same sense that clothing or property are, but we do not invent most of our language. We inherit our native languages and we are usually taught our foreign languages, and we adapt both to actual situations over time. If we do so by extrapolating regularities, we are not inventing the regularities, just extending them. We do invent jargon, but jargon usually dies with us unless one of four things happen:
(1) someone uses it and becomes a major exemplar for the whole language, so it spreads by diffusion and becomes an ordinary part of the language;
(2) everyone in a field agrees to use it for some reason, so it becomes specialized or technically vocabulary;
(3) people use it in deference to the people in a field, so it remains specialized or technical vocabulary but occasionally gets used outside the field;
(4) people use it in deference to the people in a field, but the occasion for use of it becomes so great that it just becomes a normal part of the vocabulary.
Srinivasan quotes a Daily Gazette comment from 1920: "‘Surely great big men who can invent such fine words as “radioactinium” and “spectroheliograph” should be able to devise a little useful pronoun.’" But this is, of course, the whole problem: radioactinium and spectroheliograph are not generally useful, they are jargon that specialists agreed to use within particular specialties in order to make it easier to talk about things in those specific fields. Outside those fields, we might use the words if the occasion comes up, but this is in deference to the specialized vocabulary, and it stays specialized. The pronouns aren't being proposed for a specialized problem but for a general problem; but the solution proposed treats the general problem as if it could be solved by trying to impose a specialized jargon. It wouldn't be an issue if the point were just the convenience of, say, an academic field; academic fields are full of jargon, some of it quite barbarous and arbitrary, involving no regard for the common usages of the surrounding language at all, and it works for the purposes of an academic field. But the English language in general use is not an academic field; it is in fact only the arrogance of academics that treats it as if it were, in defiance of all actual reality; and if you want the spread of pronouns into general usage, it's going to have to be either by diffusion from something people like and imitate or by society changing first so that it becomes a handy solution people can use all the time. It is the many who determine general usage by their actual practice; academics and politicians cannot do it by fiat. And, of course, it's going to have to turn out more useful than both common he and (the even bigger challenge) singular they.
The grammatical conservative argument is about politics only in the sense that in this particular case the external reasons that are motivating the attempt to impose the jargon are political. The cultural conservative argument is in fact different from all of this and has nothing to do with grammatical rules, except insofar as this case happens to be about language. There are cultural conservatives who are grammatical conservatives, but neither position entails the other. Precisely the problem for the cultural conservative is that it is imposed punitively, by shaming and increasingly by threat to employment and legal penalties, and done so deliberately to further a specific political position. Having a problem with this does not require any particular position about whether the usual rules of English are particularly great or could not be improved. It certainly wouldn't have any implications one way or another for a popular and freely spreading solution like singular they. Even if the cultural conservative is also a grammatical conservative, there is all the difference in the world between people being forced to give up their native usages for someone else's political goals and freely changing their usages for their own. Indeed, all the difference, in its own micro-scale way, between fascist Orwellianism and democratic folly. Pick any folly you choose, and one can argue that if people are going to be fools, they should at least freely be fools for their own reasons, and not fools because other people are forcing them to be for their own reasons. And as the general point, about coercing people to do things, does not fundamentally change even if you don't think it actually a folly, so nothing prevents someone making the cultural conservative argument without being a grammatical conservative.
Srinivasan's conflation of grammatical and cultural conservatism helps her rhetorically for most of her argument, since it makes the cultural conservatives look more arbitrary than they are and it makes the grammatical conservatives look like they have more of a definite political agenda than they do. But the more she focuses on politics in particular the more it raises questions that need to be considered but are not, and in at least two ways.
(1) She effectively concedes all the major points of the cultural conservative argument, despite the fact that she clearly intends to be arguing that it is unreasonable. On Srinivasan's argument, the point of the new pronouns is indeed political; it is deliberately political; and it is pushing a particular political position. This position is being pushed against the resistance of at least a large mass of the broader population and is being pushed to change society for political reasons. She thinks they are good political reasons, ethical reasons; but outside a few sociopaths who don't care, everybody appeals to ethical reasons to try to justify impositions on others. Bare appeal to its being for an ethical end doesn't establish that it is an ethical means, much less one that is in any way obligatory. Given the premises of Srinivasan's argument, cultural conservatives are right by their own lights to oppose the innovation: if it really is pushing a politics on people, why would you accept that at all unless you already agreed with the politics? Srinivasan could defang the argument by saying that use of preferred pronouns should only be asked as a personal matter of courtesy. But doing this would mean her argument is backward. Personal courtesies become expected courtesies because society changes. Again, the actual history of pronoun innovation here is a history of failure. It very much seems as if this can work only by the slow but organic and popular path of singular they, which is really being rejected here, or by pushing a major revision of language on people, which is not a matter of words changing the world but the reverse.
(2) Srinivasan consistently speaks as if there were only one linguistic culture. There are many. And Srinivasan's argument from that perspective looks very much like an argument that justifies pressuring linguistic cultures to conform to a more enlightened universal language, even if they resist. It's a battle over which politics is to prevail; for instance, the progressive, enlightened Anglophone one, or the First Nations one. Perhaps the cultural conservatives of native linguistic cultures did indeed fear the power of words to change the world; but it is part and parcel of the cultural conservative argument that the real problem was that the world was changing the words.
I've noted before that academia tends to hide its own relation to colonialism, for all that it has become fashionable for academics to attack everything else for colonialism. And the relation is ultimately this, that they are in some sense the same; when you start going back to the primary texts, it becomes quite obvious that this is so. Colonial administrators did not pop up from the ground; they were born from European universities. Colonial administration as we usually think of it just is university habit applied to ruling foreign nations; the same principles that were part of university reforms in education were extended to reforms in government, and entire populations were put under tutelage as students of the European nations who took themselves to be improving the world by educating it. And since it was for such a very ethical reason, of course, the idea was that they had to be; sometimes by force, sometimes just by continual pressure, they had to be made to learn, so that their injustices would stop. And pressure against the language was one of the ways it unfolded. After all, who can be counted as educated if they don't got no ability to speak good? And what are enlightened languages, suitable for civilized life, on this view? It's not that of peasants and farmers; it's whatever language happens to be spoken in the universities.
University habits as a political regime for ruling foreign nations ran into problems as an explicit view because the resistance of an entire foreign nation is quite formidable, no matter what their disadvantages, and the military and economic cost of continuing such a regime eventually became unmanageable given other demands on both army and purse. The similar technocratic rule is a recurring temptation for domestic politics, but it's never been quite able to find a solution to the regular populist revolt against it. But academic life is sheltered, with its structures and processes changing slowly and often only superficially, and it gets protection from a general perception of serving a useful function, so the worst consequences are very occasional and very limited anti-intellectual and anti-academic backlashes. Thus the university habits are still there. Put any of us academics under the microscope, and you'll find us all tempted to slip into thinking as if the rubes must be civilized -- who counts as rubes and why, and what counts as civilizing them, shifts around a lot, but if you spend your life lecturing and grading, it takes a genuine effort never to treat the general population as students who need your training and junior scholars who need your mentoring, rather than (what they much more certainly are) the people who pay your bills on the understanding that doing so will ultimately make their lives less difficult. But even more broadly, I think, it tempts to one particular form of the overmoralization I've been going on about recently, and in particular a tendency to insist that your views are the direct and necessary requirements of ethics and reasonableness, and that therefore anyone who resists them can't really be resisting your own proposals as unreasonable impositions to which you have no right. They are just afraid of your good work of making the world better; they find your suggestions disturbing because they are complicit with the world's wrongs -- their unreasonable resistance is a further proof your rightness, and of their need to be enlightened, made civilized.
People who are in no way colonialist may nonetheless have the university habits of which colonialism was a monstrous growth, the academic garden envisioned at the size of a world. And almost all of us academics do; it's part of how you present yourself as a serious scholar rather than -- well, rather than as one of the rubes. Thus it's not surprising that people who absolutely repudiate colonialism will nonetheless sometimes turn out to give arguments that are structurally very similar to justifications of colonial interference. It's not even necessarily a bad thing to do so, since nothing in the abstract prevents practical arguments that are utterly monstrous at imperial size from being perfectly fine at miniature scale. That's a matter of proportion and evidence. But no one who genuinely rejects colonialism can simply dismiss the cultural conservative argument; nor does the cultural conservative argument actually depend on whether the culture in question is right. It's not a justification of colonialism that native cultures were flawed; of course they were, since all cultures are. It's not even a justification that some of them had practices that were ethically horrendous; it's not the end but the means that are in question. The cultural conservative has a point, and one that cannot be easily dismissed. The cultural conservatives are sometimes absolutely, definitely right. This needs to be faced squarely, and conflating the grammatical conservative and cultural conservative arguments makes it impossible to do so.
Of course, it is entirely possible that Srinivasan sees the matter as she presents it at the end, as purely a matter of courtesy and kindness and working with people to make a better world. It's possible that she would vehemently oppose any attempt to penalize people for not using preferred pronouns, or perhaps allow no greater penalty than just treating it as a rudeness not to be condoned but nonetheless to be tolerated. If that's the view put forward, then the cultural conservative argument is entirely bypassed. But I take it that the reason the cultural conservative argument comes into play so much in this context is that there are more than a few people who are not convinced that this is really what is being put forward.