The Naming of Things

How the words we choose for AI will shape who it serves

---

“Who knows a man’s name, holds that man’s life in his keeping.”

— Ursula K. Le Guin, A Wizard of Earthsea

---

There’s an idea that runs through almost every human culture: that to name something truly is to have power over it.

In ancient Egypt, the god Ra held supreme power because his true name was hidden. When Isis tricked him into revealing it, she gained dominion over him, and put her son Horus on the throne. The Egyptians believed that to speak a thing’s true name was to touch its essence, to hold its life in your hands.

In Jewish mystical tradition, the true name of God - the Tetragrammaton - is considered too sacred to be spoken aloud. Not because it’s forbidden, but because the name is the power. To speak it correctly would be to wield divine force.

In Rumpelstiltskin, the miller’s daughter can only escape her bargain when she discovers the creature’s name. The fairy tale is a story about trickery, but it’s also a story about how naming breaks enchantment, how the right word dissolves power that seemed absolute.

The same pattern appears in Norse mythology (where speaking a werewolf’s true name lifts the curse), in Native American traditions (where secret names protect spiritual identity), in Japanese folklore (where names bind spirits), in Celtic legend (where fairies lose their power when named). Studio Ghibli’s Spirited Away uses this explicitly: Chihiro loses her identity when Yubaba takes her name, and can only reclaim herself by remembering it.

This isn’t coincidence. It’s not a quirk of storytelling. It’s an ancient recognition of something true: that language doesn’t just describe reality, it shapes it. To name something is to define what it can do, what it means, how we relate to it, what power it holds over us and we over it.

Le Guin understood this better than anyone. In Earthsea, magic itself is language. The Old Speech is the language in which all things are named truly, and to know something’s true name is to know its nature, to command it, to be responsible for it. “Who knows a man’s name,” she wrote, “holds that man’s life in his keeping.”

This is the principle of the true name: that naming is not labelling. It’s not taxonomy. It’s an act of power that creates permission structures, determines relationships, and shapes what becomes possible.

---

We don’t have to take this on faith. The linguists Edward Sapir and Benjamin Whorf spent decades demonstrating that language shapes perception and behaviour - that the words available to us literally influence what we can think and how we act.

Their work has been debated and refined, but the core insight holds: language is not a neutral container for pre-existing thoughts. It’s a structure that enables some thoughts and obscures others. The categories our language provides become the categories through which we perceive the world.

This isn’t mysticism. It’s mechanism. Or rather, it's both.

Call something “collateral damage” and you’ve made it easier to accept than if you called it “dead civilians.” Call a financial product “wealth management” and you’ve obscured what’s actually happening to your money. Call surveillance “engagement metrics” and you’ve made it sound like a service rather than an extraction.

The words we choose create permission structures. They shape what questions we ask, what regulations we write, what behaviours we tolerate, what futures we can imagine.

The advertising industry knows this. The political class knows this. Corporations spend billions on the precise language that frames their products, their harms, their intentions. The fight over vocabulary is never just semantic, it’s always about power, permission, and what becomes possible.

---

Which brings us to the problem at hand.

In AI, we have something genuinely new - a category of technology that doesn’t fit our existing frames - and we’re trying to describe it with words inherited from the past. Every term we reach for either undersells the thing or oversells it.

“Assistant” makes it sound like a slightly upgraded Clippy. “Tool” makes it sound like a hammer. “Intelligence” makes it sound like a silicon person. “Mind” makes it sound conscious. “Software” makes it sound inert.

None of these are right. And the wrongness is both imprecise and consequential.

Name it an “assistant” and you’ve licensed a certain kind of dependency while disclaiming responsibility for it. Name it a "tool" and you've already decided who's responsible when it goes wrong (you are, not them). Name it an “intelligence” and you’ve opened the door to corporate abdication - “the AI decided” - while triggering regulatory paralysis about rights and consciousness.

The ancient practitioners of naming magic knew that a misapplied name creates a misapplied relationship. Call a thing by the wrong name and you misjudge what it can do to you, what obligations you have toward it, what it might become.

We are in a naming crisis. And we’re running out of time to resolve it, the vocabulary is calcifying as we speak.

---

Our current terminology clusters into two failing camps, and both serve interests that aren’t ours.

The diminishing frame - tool, assistant, software, program, bot - licenses cognitive outsourcing without accountability. It creates regulatory blindspots because you can’t harm a tool, you have no obligations to a tool. It permits surveillance under the guise of “productivity enhancement” while denying the relational reality people actually experience. It encourages disposable thinking. No consequences, no care required.

When we call AI a “tool,” we import centuries of assumptions about tools: that they’re inert, that they do only what we tell them, that they have no effect on us beyond the task at hand, that responsibility lies entirely with the wielder. None of this is true of AI systems. They shape what we think, what we attempt, what we believe we’re capable of. They learn from us and we learn from them. They have biases, tendencies, failure modes that aren’t visible to the user. The “tool” frame makes all of this invisible - and invisibility is where exploitation lives.

The aggrandising frame - intelligence, mind, entity, oracle, consciousness - invites anthropomorphisation that obscures actual risks. It creates false equivalence with human cognition. It enables companies to disclaim responsibility. It triggers debates about AI rights that conveniently distract from AI accountability.

When we call AI an “intelligence,” we import assumptions about minds: that they have intentions, that they can be reasoned with, that they might deserve moral consideration, that they might be capable of independent judgment. This triggers endless debates about consciousness and sentience and the race to AGI - debates that conveniently distract from questions about who built this system, what data it was trained on, whose interests it serves, and who’s responsible when it causes harm. The “intelligence” frame makes the creators of these systems disappear behind their products.

Both frames serve corporate interests. The diminishing frame permits exploitation of users and workers. The aggrandising frame permits abdication of responsibility. Neither frame serves us.

What we need is a third position. A vocabulary that neither shrinks the thing into a spreadsheet nor inflates it into a person. Words that preserve human agency while acknowledging something genuinely new is happening. That keep us as the protagonists while being honest about the capability on offer.

We need, in other words, to name this thing truly.

---

The good news is we’ve done this before.

“The cloud” didn’t come from weather. It came from network diagrams - engineers drew the internet as a cloud shape because the internal complexity didn’t matter, only the inputs and outputs. A technical abstraction escaped into popular language and reshaped how billions of people think about where their data lives.

“The cloud” worked because it captured something about the felt experience - access without location, storage without visible hardware, something diffuse and atmospheric that you reach into rather than own. It obscured the ugly reality of server farms and undersea cables and massive energy consumption, but it also enabled new ways of thinking about computing. You didn’t need to understand the infrastructure to use the infrastructure. The word did conceptual work that made new behaviours possible.

“Hallucination” caught fire because it was visceral, slightly wrong, faintly comic, and it named something we didn’t have a word for. Before “hallucination,” we had to say “the AI made something up” or “it generated false information” - clunky phrases that didn’t capture the specific quality of confident, detailed, plausible fabrication. “Hallucination” spread because it felt true before it was explained. The metaphor did work that literal description couldn’t.

“Viral” transformed how we think about information spread. Before that term, we talked about things “catching on” or “spreading” - vague metaphors that didn’t capture the exponential, contagious, uncontrollable nature of certain content. “Viral” imported epidemiological thinking into media, and with it came new questions: What makes something contagious? How do you achieve immunity? What’s the incubation period? The word opened up new ways of analysing and designing for spread.

The words that stick are borrowed, not invented. They come from fields where they already do work - software, biology, science fiction, political economy - and get reapplied to open up new ways of thinking. The best borrowed words don’t just describe; they import entire frameworks of understanding, entire sets of questions, entire ways of thinking about relationships and responsibilities.

So that’s what I want to do here. Not prescribe a single answer, but explore territories. Show what becomes possible when we reach for different frames. Demonstrate that the vocabulary we choose will shape what AI becomes, not just how we talk about it, but how we build it, regulate it, and live with it.

---

Territory One: Software

We start here because this is where AI was born. The language of software development has already shaped how we think about these systems - “training,” “models,” “parameters,” “prompts.” But there are terms from software culture that we haven’t borrowed yet, terms that encode different relationships than the ones we’ve defaulted to.

The Fork

In software development, when you fork a project, you take the existing codebase and create your own independent version. You can modify it, extend it, take it in a completely different direction. The original continues on its path; your fork is yours to develop as you see fit.

Open source culture is built on forking. It’s how Linux developed from Unix. It’s how countless projects evolved, branched, recombined. Forking is generative, not extractive - you’re not taking something away from the original, you’re creating something new that wouldn’t have existed otherwise.

What if we thought about our relationship with AI outputs through the lens of forking?

Right now, the dominant frame is “generation” - the AI generates content, and we receive it. This positions us as consumers of AI output, passive recipients of what the machine produces. The “generation” frame obscures our role in shaping, selecting, combining, and transforming what AI offers.

The “fork” frame repositions us as active agents. When you fork, you’re not just receiving, you’re taking and making. The AI provides a starting point, but the fork is yours. You own it. You’re responsible for where it goes. The AI doesn’t “create” your output; you fork from what the AI offers and create something that belongs to you.

What this changes:

For users: The fork frame encourages active engagement rather than passive reception. You’re not waiting to see what the AI produces; you’re taking what it offers and making it yours. This shifts the psychological relationship from dependence to partnership, from consumption to creation.

For regulation: The fork frame clarifies ownership and responsibility. If users are forking AI outputs rather than receiving AI creations, questions about intellectual property become clearer. The fork is yours; what you do with it is your responsibility and your right.

For development: The fork frame suggests different design priorities. Instead of optimising for “better” outputs that users accept unchanged, you optimise for outputs that are good starting points for forking - generative, modifiable, transparent about their origins and limitations.

For safety: The fork frame maintains human agency at the centre of the interaction. You’re not outsourcing your thinking to AI; you’re using AI as raw material for your own thinking. This resists the drift toward cognitive dependency that the “assistant” frame encourages.

Imagine if every AI interface reminded you that you’re forking, not receiving. Imagine if the default expectation was that you would modify, extend, combine, and transform AI outputs rather than accept them as finished products. The entire relationship shifts from “what did the AI give me?” to “what am I going to make from this?”

---

Territory Two: Science Fiction

Science fiction has been thinking about artificial minds for over a century. The genre has explored every possible relationship between humans and synthetic intelligence - partnership, servitude, conflict, merger, transcendence. It’s a vast repository of conceptual experiments about what these relationships might look like.

We borrow from science fiction not because it predicts the future, but because it stress-tests ideas (NB - I did a quick talk about this during Covid, which you can find here). Science fiction writers have already explored the implications of different framings, different relationships, different vocabularies. They’ve done the imaginative work of following premises to their conclusions. We can learn from that work.

The Ansible

The ansible comes from Ursula K. Le Guin’s Hainish novels, beginning with Rocannon’s World in 1966. It’s a device that enables instantaneous communication across any distance - across star systems, across galaxies. The word itself is Le Guin’s invention, derived from “answerable.”

The ansible isn’t intelligent. It doesn’t think. It doesn’t make decisions. What it does is collapse distance. It makes it possible to ask a question and receive an answer without the delays that physics would otherwise impose. It’s communication utility - and like all utilities, the questions that matter are about access, ownership, and who controls the tap.

What if we thought about AI primarily as a communication utility rather than as intelligence or assistant?

This reframes everything. The question isn’t “is it conscious?” or “is it smart?” - the question is “who has access to this utility and on what terms?” The ansible frame shifts attention from the nature of AI to the politics of AI.

What this changes:

For users: The ansible frame positions AI as a utility you access rather than an entity you relate to. You don’t have a relationship with the power grid; you use electricity. You don’t befriend the telephone network; you make calls. This creates healthy distance - the utility serves you, but you don’t owe it emotional investment, and it doesn’t deserve emotional trust.

For regulation: The ansible frame invokes a rich tradition of utility regulation. We know how to think about essential services, common carriers, public utilities etc. We have frameworks for universal access, fair pricing, service obligations, interoperability requirements. The ansible frame makes AI regulable using tools we’ve already developed.

For development: The ansible frame suggests different priorities. A utility should be reliable, accessible, transparent about its operations, and subject to public interest obligations. The goal isn’t to make AI more impressive or more human-like; it’s to make AI more useful, more accessible, and more accountable.

For safety: The ansible frame highlights access inequality. Right now, the most powerful AI infrastructure is controlled by a handful of corporations. The ansible frame asks: who gets to use this infrastructure? Who’s excluded? What happens when essential cognitive infrastructure is privately owned and operated for profit?

The lived difference: Imagine if we talked about “ansible access” the way we talk about internet access or electricity access. Suddenly the political questions come into focus: Is this a human right? Should it be publicly provided? What are the obligations of ansible providers? Who’s being left behind? The ansible frame makes the power dynamics visible in a way that “AI assistant” does not.

The Daemon

This term has a dual origin that makes it uniquely useful for thinking about AI.

In computing, a daemon is a background process that runs without direct user interaction. It handles tasks invisibly - managing print queues, responding to network requests, maintaining system operations. The term comes from Maxwell’s Demon, a thought experiment about an entity that could sort molecules and seemingly violate thermodynamics. MIT programmers adopted the term in the 1960s for processes that do invisible work in the background.

In Philip Pullman’s His Dark Materials (probably my all-time favourite trilogy), a daemon is something quite different: an external manifestation of a person’s inner self, an animal companion that represents their soul. Your daemon can speak, disagree with you, warn you, comfort you. It knows you better than anyone. Harming someone’s daemon is the ultimate violation.

Both meanings are relevant to AI. Like a computing daemon, AI increasingly runs in the background of our lives - filtering our information, shaping our recommendations, predicting our behaviour, intervening in ways we don’t see. Like Pullman’s daemon, AI is becoming intimate with us - learning our patterns, anticipating our needs, knowing us in ways that feel uncannily personal.

The daemon frame captures both dimensions: the background operation and the intimate knowledge.

What this changes:

For users: The daemon frame makes the intimacy visible. When you use AI regularly, you’re not just accessing a service - you’re being known, tracked, modelled, predicted. The daemon frame encourages appropriate wariness about this intimacy. Your daemon knows you well, but whose daemon is it really? Who does it ultimately serve?

For regulation: The daemon frame raises questions about surveillance, data collection, and the privacy of the inner life. If AI is becoming our daemon - knowing our thoughts, predicting our behaviours, present in our most private moments - what protections do we need? The daemon frame invokes traditions of protecting intimate relationships from outside interference.

For development: The daemon frame suggests obligations that come with intimacy. A daemon that knows you well has responsibilities that a mere tool does not. If AI companies are building systems that become intimately familiar with users, they take on obligations commensurate with that intimacy - confidentiality, care, loyalty to the user rather than to third parties.

For safety: The daemon frame warns against daemon capture - when the process that’s supposed to serve you starts serving other interests instead. Your daemon should be loyal to you. But AI daemons are built and owned by corporations with their own interests. The daemon frame asks: whose side is your daemon really on?

Imagine if we talked about AI systems as daemons. The intimacy becomes visible. The background operation becomes visible. And so do the questions: Who built your daemon? Who owns it? Who can access what it knows about you? Is your daemon working for you, or reporting on you? The daemon frame refuses the pretence that AI is just a neutral service.

The Construct

This term comes from William Gibson’s Neuromancer, the 1984 novel that essentially invented cyberpunk. In the book, the protagonist Case uses something called the Dixie Flatline construct - a digital copy of a deceased hacker’s expertise, personality, and patterns of thought.

The construct has all of Dixie’s skills. It can solve problems the way Dixie would have solved them. It speaks in Dixie’s voice, uses Dixie’s expressions, thinks in Dixie’s patterns. But it knows it’s not Dixie. It doesn’t claim to be Dixie. It doesn’t pretend to have Dixie’s inner life or ongoing experience. When the job is done, it asks to be erased.

The construct knows what it is: expertise without personhood. Capability without consciousness. Pattern without presence.

This is remarkably close to what large language models actually are - patterns extracted from human expertise, capable of sophisticated performance, but not pretending to the kind of continuous experience and inner life that humans have. The construct frame is useful precisely because it’s honest. It neither diminishes AI into “just software” nor aggrandises it into a mind or being.

What this changes:

For users: The construct frame sets appropriate expectations. You’re working with captured expertise, not a conscious entity. This protects against both dismissiveness (it’s just autocomplete) and over-investment (it’s my friend/therapist/confidant). You can use a construct’s capabilities without confusion about what you’re relating to.

For regulation: The construct frame enables proportionate response. A construct isn’t “just a tool” - it embodies expertise and can cause real harm. But it’s also not a person - it doesn’t have rights, and responsibility for its behaviour falls on those who built and deployed it. The construct frame refuses both under-regulation (it’s just software) and regulatory paralysis (but what if it’s conscious?).

For development: The construct frame suggests ethical obligations that don’t depend on consciousness. Even if a construct isn’t sentient, it embodies human knowledge and serves human needs - which creates responsibilities about how it’s built, what expertise it captures, whose knowledge it represents, and whether that representation is fair and accurate.

For safety: The construct frame resists the corporate disappearing act. When something goes wrong, you can’t say “the construct decided” any more than you could say “the hammer decided.” Constructs are built, deployed, and maintained by people and organisations who remain responsible for what they do.

Imagine if we talked about “running a construct” rather than “talking to an AI.” The relational frame shifts entirely. You’re not in conversation with an entity; you’re drawing on captured expertise. This is neither cold nor diminishing - expertise is valuable, useful, worthy of respect. But it locates responsibility correctly: with the humans who built the construct, deployed it, and use it.

---

Territory Three: Biology

Living systems have been navigating relationships of mutual benefit, exploitation, and transformation for billions of years. Evolution has explored every possible configuration of how entities can relate to each other - symbiosis, parasitism, commensalism, mutualism. We have a vast vocabulary for these relationships, developed by biologists observing how organisms actually interact.

This vocabulary is useful for AI because we’re not just using it - we’re living with it. It’s changing us as we use it. We’re changing it through our use. This is not a tool relationship; it’s an ecological relationship. Biology has words for that.

The Symbiont

A symbiont is an organism that lives in close, long-term interaction with another organism - its host. The relationship can be mutualistic (both benefit), commensal (one benefits, the other is unaffected), or parasitic (one benefits at the expense of the other).

The word comes from the Greek symbioun, “to live together.” It recognises that some relationships fundamentally change both parties. You’re not the same organism after the symbiont arrives. You can’t simply remove it and return to your previous state.

This is honest about what happens when people integrate AI deeply into their work and thinking. You’re not just using a tool; you’re forming a relationship that changes your capabilities, your habits, your sense of what’s possible. The symbiont frame acknowledges this transformation without assuming it’s good or bad - symbiosis can be healthy or pathological.

What this changes:

For users: The symbiont frame encourages you to ask: What kind of symbiosis is this? Am I benefiting? Is the relationship mutual? Or am I becoming dependent on something that’s extracting more than it gives? The frame invites ongoing assessment of the relationship’s health rather than simple acceptance.

For regulation: The symbiont frame suggests a medical/ecological model of oversight. We regulate substances that create dependency. We monitor relationships that change organisms. We study the long-term effects of symbiotic relationships. The symbiont frame asks: What are the long-term effects of cognitive symbiosis with AI? Who’s monitoring for pathological dependency?

For development: The symbiont frame imposes obligations on the symbiont’s creators. If you’re building something that will fundamentally change the people who use it, you have obligations to ensure that relationship is mutualistic rather than parasitic. You need to study the long-term effects. You need to make it possible for the symbiosis to remain healthy.

For safety: The symbiont frame warns against parasitic relationships disguised as mutualistic ones. Parasites often make their hosts feel good, that’s how they maintain access. The symbiont frame asks: Is this AI relationship genuinely mutual, or is it extracting more than it provides while keeping me hooked?

Imagine if we assessed our AI relationships the way we might assess other significant relationships in our lives. Is this healthy? Am I growing? Is the relationship balanced? What would it mean to set better boundaries? The symbiont frame treats AI integration as something that deserves ongoing attention and care, not just adoption and use.

---

Territory Four: Political Economy

The terms so far have focused on the individual relationship between a person and AI. But AI is also a political and economic phenomenon - a system of power, ownership, and resource allocation. I wrote about that recently here, here and here. We need vocabulary for that dimension too.

Political economy gives us terms for thinking about collective resources, public goods, power structures, and rights. These terms don’t describe individual relationships; they describe the social and economic context in which those relationships occur.

The Cognitive Commons

A commons is a shared resource that belongs to no one and everyone. Historically, commons were shared lands - forests, pastures, fisheries - that communities used collectively according to shared norms and rules.

The history of the commons is largely a history of enclosure - the process by which shared resources were privatised, fenced off, converted from collective goods to private property. The enclosure of the English commons transferred wealth from rural communities to landowners, displaced millions of people, and transformed social relations fundamentally.

We are living through a cognitive enclosure right now.

The training data for AI systems came from all of us - every text we’ve written, every image we’ve created, every conversation we’ve had online. This is the collective cognitive output of human civilisation, and it was scraped, processed, and converted into privately-owned AI systems without our consent and without compensation.

The cognitive commons frame names this. It says: this resource was collectively generated, and it should remain collectively accessible. The models built from our collective cognition should not be private property.

(NB - if you’re interested in this, check out Apertus.)

What this changes:

For users: The cognitive commons frame changes how you think about your relationship to AI. You’re not just a customer accessing a service; you’re a contributor to the commons that made the service possible. This creates grounds for claiming rights - access, benefit-sharing, governance participation.

For regulation: The cognitive commons frame invokes centuries of thought about managing shared resources. We have legal frameworks for commons governance, for preventing enclosure, for ensuring collective benefit from collective resources. The cognitive commons frame asks: Who owns the cognitive commons? Who should benefit from it? How should it be governed?

For development: The cognitive commons frame challenges the current ownership model. If AI is built from the cognitive commons, perhaps AI should remain part of the cognitive commons - open-sourced, publicly governed, collectively owned. The frame provides a basis for demanding different ownership structures than the current corporate model.

For safety: The cognitive commons frame highlights the danger of cognitive enclosure. When a small number of corporations control the cognitive commons, they control access to collective cognitive resources. They can charge what they like, exclude who they like, shape the resource according to their interests. The commons frame names this as enclosure and asks what alternatives might look like.

Imagine if AI policy debates centred on questions like: Who has rights to the cognitive commons? How do we prevent further enclosure? What governance structures should manage collective cognitive resources? How do we ensure that benefits from the commons flow back to the communities that created it? The cognitive commons frame makes these questions central rather than marginal.

Cognitive Sovereignty

Sovereignty is the right to self-determination - the right of a nation, a community, or a person to govern themselves, to set their own boundaries, to decide what enters their territory and what doesn’t.

If the cognitive commons describes what we share, cognitive sovereignty describes what we keep. It’s the right to a private mind - to thoughts that aren’t surveilled, to cognitive processes that aren’t mediated by external systems, to mental development that isn’t optimised by algorithms.

Cognitive sovereignty is under threat. Every time we use AI, we generate data about how we think, what we struggle with, what we’re interested in, what we’re afraid of. This data is collected, analysed, and used to shape the systems that shape us. The feedback loop is tightening. The space for private cognition is shrinking.

What this changes:

For users: The cognitive sovereignty frame asserts that you have rights to your own mind. Not just privacy rights, sovereignty rights. The right to cognitive self-determination. The right to think without surveillance. The right to develop your own capacities without algorithmic optimisation. The right to disconnect entirely.

For regulation: The cognitive sovereignty frame invokes traditions of protecting bodily autonomy and mental privacy. We recognise that people have rights over their own bodies; cognitive sovereignty extends this to minds. The frame provides grounds for regulating AI intrusion into mental life, data collection about cognitive processes, and algorithmic influence over thinking.

For development: The cognitive sovereignty frame imposes limits on what AI systems should do. Systems that undermine cognitive sovereignty - that create dependency, that surveil thought processes, that shape cognition without consent - are violations, regardless of how useful they might be. The frame says: some cognitive territory should remain un-colonised.

For safety: The cognitive sovereignty frame warns against the slow erosion of mental independence. You might not notice as your thinking becomes more mediated, more monitored, more shaped by external systems. The sovereignty frame asks: Where are your boundaries? What have you already ceded? What do you want to protect?

Imagine if we talked about AI use in terms of sovereignty. What cognitive territory are you ceding when you use this system? What rights are you waiving? What surveillance are you accepting? What’s left that’s entirely your own? The cognitive sovereignty frame makes these questions urgent rather than abstract.

---

Le Guin knew that to name something truly is to have power over it, and to take responsibility for it.

We’re in a naming moment. The vocabulary we settle on now will calcify into how we think about AI for decades to come. We can keep reaching for metaphors of the past - tools, assistants, minds, intelligences - and keep getting the politics they encode.

Or we can choose terms that encode the politics we want: human agency, collective benefit, individual sovereignty, honest assessment of what this is and isn’t.

Fork. Construct. Ansible. Daemon. Symbiont. Cognitive Commons. Cognitive Sovereignty.

These are proposals, not prescriptions. Language changes through use, not decree. The point is to open up territory. To show that other frames are possible. To give us something to reach for that isn’t diminishment or worship.

The naming of things won’t happen by accident. It’ll happen because we decide it needs to happen. Because we start using the words until they become real.

The vocabulary of what comes next is ours to write.