you don’t have to engineer the perfect solution to nonbinary language

Kirby Conrod
7 min readJan 15, 2024

There is a theme that keeps coming up in my research and conversations — people are really worried about whether some form of nonbinary language is efficient or optimal in some way.

I’ve seen a lot of stuff about this in discussions of English pronouns, but also a lot about languages like Spanish, French, and German: there’s this idea that nonbinary language needs to be just right.

For English pronouns, this results in sort of comical contradictions. I’ll sum up / paraphrase two types of really common comments I got in one of my recent large studies on English neopronouns:

Position 1: “Singular ‘they’ is too confusing, because I can’t always tell if it’s actually talking about a single specific person or if it might be referring to multiple people. Instead, it would be clearer if we had unique new pronouns for nonbinary people.

Position 2: “Neopronouns are confusing, because it sounds like a word I’ve never heard before, or like a weird mispronunciation of another word, and I don’t register it as a pronoun and can’t track who we’re talking about. Instead, it would be more efficient if we used singular ‘they,’ which has already been around for a while and I already kind of get it.

You see how these two positions can’t be reconciled, right? If we’re optimizing for clarity, then they might be imprisice or confusing. If we’re optimizing for efficiency, then they is ideal because it’s recycled and we don’t have to deal with the psychological burden of new words.

I see this argument about Spanish noun endings, too:

Position 1: “Replacing -a and -o endings with -x or -@ is too confusing, because I can’t figure out how to pronounce those and they don’t exist anywhere else in the language. Instead, it would be more efficient if we used -e, just like ‘estudiante,’ since it’s already in the language and we can just extend it to more contexts.”

Position 2: “Using -e is confusing, since it’s not clear if someone’s trying to use gender neutral language or specifically talking about nonbinary people. It would be better if we had nonbinary-specific endings, that aren’t similar to anything else — maybe -x?”

These paraphrases are of course oversimplifications, and they’re both in contrast to the constant third position:

Position 3: “You don’t have a right to change the language we’re speaking, I refuse to recognize you as a voting committee member, and any attempts to challenge the existing language structures are direct challenges to the way things are.”

(Underlying Position 3, of course, is the belief that people are simply and obviously either “male” or “female,” and that attempts to change the language are attempts to obfuscate what seems to Position-3-holders as self-evident and objective reality. Given that they’re objectively and scientifically incorrect, I won’t be entertaining that premise with further attention here.)

I am intrigued by this kind of metalinguistic debate, in part because it shows some interesting ideologies about how people think about language, and in part because I think it reflects a basic truth about language that people are intuitively aware of.

Language is an epic battle between chaos and order, and we are but specks in the greater structure of things

The fundamental tension conditioning all variation and change in language (and remember, variation and change are natural and universal properties of language) is the tension between efficiency (for the sender) and clarity (for the receiver).

Efficiency for the message-sender is about minimizing effort: physical effort (like, don’t make me do difficult stuff with my hands or tongue!) and cognitive effort (don’t make me stress out my working memory or processing abilities to put together words and phrases!). Fundamentally, we want it to be easy to make language come out of our bodies.

Clarity for the message-receiver is about maximizing information: don’t be vague, ambiguous, or confusing, don’t make me ask extra questions to figure out what you’re saying to me, don’t use homonyms or similar shapes of words that I might mix up. Fundamentally, we want it to be easy to glean meaning from a linguistic message.

If Efficiency wins the battle, all spoken languages consist of exactly two phonemes, probably /b/ and /ə/, so that all spoken languages just sound like “buh buh buh.” This is VERY EASY to say! Also there’s no syntactic operations to make questions different from statements (because that’s effortful!) and almost everything is a homophone you have to figure out from context. Signed languages would have some equivalent, probably neutral hand-flaps, whatever the “easiest” handshape is (palm flat and down-facing at about waist-height, relaxed wrist? I don’t know enough sign language phonotactics here, sorry). And it’d be the same deal: almost all signs would be homonyms, and message-receivers have to do a lot of work to figure out what is going on from context.

That’s obviously… suboptimal! Like, that sounds like a pain in the ass as a way to language! Sure, it’d be very easy for speakers/signers to send messages — less physical and mental effort — but the rate of misunderstanding and confusion would be astronomical, just totally untenable! Let’s not do that!

So, then, what about if Clarity wins? If Clarity wins, then all spoken languages need way more phonemes — the goal is to have zero homophones — and we’re probably going to want stuff like clicks and tones in everything. We’re going to put tense and aspect and mood and noun classes and case-marking on everything, to the point of ridiculous redundancy. Saying a simple sentence will take ages, but by god it will be impossible to misunderstand. Signed languages might do some cool shit with non-manual markers, maybe feet will be involved?

This is also suboptimal — talking will be exhausting, will take so much planning and remembering that even the neurotypicals will be constantly running out of verbal spoons most days. I’m tired just thinking about this — let’s please not do that, either!

This is why no human language ever is perfectly “optimal,” because there is the constant push and pull between efficiency and clarity. It’s necessary, like balancing cosmic forces, and they have to balance each other out. Little individual speech-acts by individual humans are constantly nudging the balance back and forth, and larger patterns of variation and language change represent the tightrope-walker’s wobble to keep their footing under them. This is the natural state of language.

Screenshot of a tweet on ex-twitter from user @moutheaters: Me: Is the natural state of the soul quiet or chaos? Taco Bell cashier: Look buddy, it’s transient, shifting like water. Dated Nov 9, 2017.
From the classics of posting when the bad website was less bad

So how do we optimize nonbinary language?

We don’t!

We practice a wee bit of mindfulness and acceptance. The nature of the soul is transient, shifting like water; language cannot be engineered into perfect orderly behavior. If you optimize for efficiency you sacrifice clarity; if you optimize for clarity you sacrifice efficiency.

And moreover, top-down language-engineering mostly doesn’t work. If I sit my ten favorite pronoun linguists¹ down in my lab and we lock ourselves in for two weeks and try to come up with the perfect way to talk about nonbinary people²— like, first, we won’t succeed, but even if we did, there simply isn’t a way we could successfully impose our will on all language-users. And any attempt to do so would involve a level of authoritarian meddling in peoples’ lives that I and my cool colleagues are simply unwilling to do — but even if we did, language finds a way. People will do what works, and in aggregate, patterns will appear. Any individual language-user is a single butterfly in the world, you know?

Anyways, I know well that this perspective feels weird and sometimes counter-intuitive, especially given that I myself have plenty of blogposts telling people what to do with pronouns. But I will always include the disclaimers: these are pieces of advice about how to achieve particular goals given my understanding of the current state of things. If your goal is to be friends with someone, misgendering them is probably counter to that goal. If your goal is to be generally well-liked and well-regarded by trans people, again, misgendering them is probably counter to your goal. Calling them fascist for not wanting to be misgendered is also probably counter to your goal.

All that being said, I admit I have been much more intrigued by neopronouns lately, in part because they have surprisingly high visibility, but it still seems like they have relatively low uptake in English. I’m curious to watch if that will change, and if it does or doesn’t, I’ll be in the bushes with my clipboard scribbling furiously either way. My only stake in this is that I don’t want my friends who use neopronouns to be misgendered, but I also don’t want people using them as a cudgel to invalidate my friends who use non-neopronouns. I want everyone to be a little bit more curious about what is happening in nonbinary language, and a bit less uptight about what should happen.

[1] Oh my god remind me to do a post later about my favorite pronoun linguists, honestly.

[2] Despite the fact that this is literally an impossible task it sounds like a really fun workshop idea, I’m gonna see if we can get a grant for this.

Thanks for reading! In addition to my other posts about pronouns and linguistics here on Medium, you can also find me on bluesky (I have invites!) if you like :)



Kirby Conrod

Dr. Conrod is a linguist and scholar sort of at large. They write about transgender stuff, the linguistics of pronouns, and ways to work with your brain.