Can We Choose Our Own Language?

Imagine you’re playing a video game, and you come to a door that requires a key. But you don’t have the key. You search everywhere, and finally you realize something: the game was designed so that you cannot open this door. The door exists only to make you wonder what’s behind it. You’re supposed to waste time trying to solve an unsolvable puzzle.

For a philosopher named Rudolf Carnap, most of the big questions people have argued about for thousands of years—Does God exist? What is reality really made of? Are numbers real?—were like that door. They seemed like deep questions, but they were actually traps. The traps were built into the language we inherited from our ancestors, who developed it for hunting, trading, and gossiping—not for thinking clearly about the universe.

Carnap thought we needed to build new languages, on purpose, to think properly. And he thought we could choose what kind of language we wanted to use, depending on what we were trying to do.

Here’s a strange thing he noticed: once you really understand that you can choose your language, many ancient philosophical problems just dissolve. Not because they’re solved, but because they were never real questions in the first place—they were confusion wearing a disguise.

The Problem with Everyday Language

Carnap grew up in Germany and fought in World War I. That experience changed him. He saw how educated people, who should have known better, went along with terrible things because they never questioned the ideas they’d absorbed from their culture. He decided that one of the most important things humans could do was build tools for thinking more clearly—tools that didn’t come loaded with hidden assumptions from the past.

The main tool he focused on was language itself. Ordinary language is full of traps, Carnap thought. It forces us to talk as though everything is a thing—love, justice, meaning, numbers, souls. We say “love is powerful” as if love were a kind of invisible muscle. We say “justice demands” as if justice were a person with opinions. This makes it seem like there must be real, invisible things that these words point to. Philosophers then spend centuries arguing about the nature of these invisible things, when really the problem is just that our language tricks us.

Carnap wanted to replace this inherited mess with deliberately designed languages—like an engineer designing a bridge instead of just using whatever rocks happen to be lying around.

The Principle of Tolerance

Carnap’s most radical idea was something he called the principle of tolerance. Here it is in his own words:

“In logic, there are no morals. Everyone is welcome to set up his logic, his form of language, as he pleases.”

What he meant was this: there is no single “right” way to set up a language for thinking. Different languages work better for different purposes. Want a language that lets you talk about numbers as if they were real objects? Fine. Want a language that treats numbers as useful fictions? Also fine. Want a language that only allows statements that can be checked by experiments? Go ahead. The point isn’t which one is “true”—the point is which one helps you do what you want to do.

This was a huge break from the way philosophy had been done for thousands of years. Most philosophers had assumed that there was one correct way to think about the world, and their job was to discover it. Carnap said no: the job of philosophy is to invent useful ways to think, not to discover The Right Way.

Think about it like this: when you play chess, you agree to certain rules. You can’t say “I’m going to move my rook diagonally, and anyone who disagrees is wrong.” The rules are conventions we choose. Carnap thought that language and logic were the same way. You pick your rules based on what you’re trying to accomplish.

Internal and External Questions

One of Carnap’s most famous examples of how this works involves the question “Do numbers exist?”

A philosopher might argue for years about whether numbers are real. Carnap said this argument only arises because people are confused about what kind of question they’re asking.

If you’re inside a mathematical language—say, arithmetic—then the question “Do numbers exist?” has an obvious answer: yes. In arithmetic, 7 exists. You can prove it. This is what Carnap called an internal question: a question asked within a particular language framework, using that framework’s rules.

But if you’re asking “Do numbers really exist?”—stepping outside any particular framework and trying to ask about reality itself—you’re asking an external question. And external questions, Carnap said, don’t actually make sense. They’re like asking “Is chess the right game to play?” without specifying what you want to do. Chess isn’t “right” or “wrong”—it’s a tool. If you want to play chess, you use chess rules. If you want to think about mathematics, you use mathematical language. The question “Do numbers really exist?” doesn’t have an answer, because there’s no framework-independent standpoint from which to answer it.

This doesn’t mean you can’t choose to adopt a framework that treats numbers as real. You can. The point is that this is a practical choice, not a discovery about the universe.

Why Metaphysics Gets Excluded

Carnap became famous (or infamous) for saying that most of traditional philosophy—what he called metaphysics—was nonsense. Statements like “The Absolute is perfect” or “Being precedes essence” seemed to him like poetry pretending to be science. They sounded meaningful, but they couldn’t be checked by any method anyone could agree on.

This sounds harsh, but Carnap’s point wasn’t that these statements are false. It was that they don’t even rise to the level of being false. They’re like a sentence that looks grammatical but doesn’t actually assert anything—like “The color green dreams of Tuesday.” It’s not that you can prove this sentence wrong; it’s that you can’t even figure out what would count as evidence for or against it.

But here’s where Carnap got more interesting than most people realize. Late in his career, he softened this view. He came to think that many metaphysical theories could be seen as rough, early attempts to build the kind of clear language frameworks he was advocating. Aristotle, Leibniz, and Kant, he said, were trying to create “the most general frameworks containing categorial concepts which are fundamental for the representation of all knowledge.” He just thought they were doing it clumsily, without realizing that they were building rather than discovering.

The Tug-of-War with Quine

The most famous debate Carnap ever had was with another philosopher, W.V. Quine. Quine agreed with Carnap that there was something wrong with traditional philosophy, but he disagreed about what.

Quine attacked the idea that there’s a clean line between statements that are true just because of language (“All bachelors are unmarried”) and statements that are true because of how the world is (“My dog is on the couch”). Carnap thought this analytic-synthetic distinction was essential for building useful languages. Quine thought it was a myth—that all our beliefs face the world as a whole, and nothing is true purely by definition.

This argument is still going on. Most philosophers today think both sides had something right. Carnap was right that we can construct useful frameworks where some sentences are treated as definitional. Quine was right that in real science, there’s no sharp boundary between what we assume and what we discover. But for Carnap, this wasn’t really an objection—he never claimed the boundary was in nature. He just claimed it was useful to draw one for specific purposes.

How to Think About This

This part gets a bit technical, but here’s the upshot: Carnap wasn’t saying anything goes, or that all languages are equally good. He was saying that the question “Is this language correct?” should be replaced by “Is this language useful for what I’m trying to do?”

This changes what philosophy is. Instead of being a search for ultimate truth, philosophy becomes a kind of engineering—designing better tools for thinking. The philosopher’s job is to invent concepts, clarify them, and test how well they work, rather than to discover what’s “really real.”

Carnap actually spent decades trying to build one particular kind of tool: a logical system for measuring how much evidence supports a hypothesis (something called inductive logic). He never finished it to his satisfaction, and philosophers still disagree about whether it was even possible. But the attempt itself shows what he meant by doing philosophy as engineering.

Why It Still Matters

You might wonder: does any of this matter outside philosophy classrooms?

Yes. It matters every time someone says “That’s just your opinion” or “That’s objectively true.” Carnap would say those phrases are shortcuts that hide real complexity. What we call “objectivity” often just means “using a framework that a lot of people have agreed on.” What we call “subjectivity” often just means “the framework isn’t specified, so we can’t check.”

Carnap’s ideas show up in surprising places. When scientists argue about whether theoretical entities (like quarks or black holes) are “real” or just useful models, they’re replaying Carnap’s internal/external distinction. When mathematicians debate whether to accept certain kinds of proofs, they’re implicitly choosing between different language frameworks. When people argue about whether a video game character “really exists,” they’re struggling with the same confusion Carnap tried to dissolve.

And at the deepest level, Carnap’s philosophy offers a strange kind of freedom. If languages are tools we can choose, then we’re not stuck with the way our parents, our culture, or even our brains naturally talk about things. We can invent new ways of thinking. We can design conceptual tools that let us see things differently.

The catch is that with this freedom comes responsibility. If we choose our own languages, we can’t blame the universe for our confusions. We have to own the tools we build—and be willing to rebuild them when they don’t work.

Key Terms

Term	What it does in this debate
Principle of tolerance	The idea that we can freely choose our logical and linguistic rules, rather than discovering the “right” ones
Internal question	A question asked within a particular language framework, answerable by that framework’s rules
External question	A question asked about whether a framework itself is correct—which Carnap says is really a practical decision, not a factual one
Framework	A deliberately constructed language with its own rules about what counts as a statement, what follows from what, and what counts as true
Metaphysics	Statements that pretend to be about reality but can’t be checked by any agreed-upon method—seen by Carnap as confusion caused by bad language design
Explication	The process of replacing a vague concept from everyday language with a precise, deliberately defined one
Analytic-synthetic distinction	The supposed line between statements true by definition and statements true because of how the world is

Key People

Rudolf Carnap (1891–1970): A German philosopher who fought in WWI, fled Nazi Germany, and spent most of his career at the University of Chicago and UCLA. He argued that philosophy should become an engineering project of designing better languages.
W.V. Quine (1908–2000): An American philosopher who was Carnap’s student, friend, and rival. He argued that the analytic-synthetic distinction couldn’t be made to work, starting a debate that continues today.
Ludwig Wittgenstein (1889–1951): An Austrian philosopher whose early work inspired Carnap’s generation, though Carnap later broke with his approach. Wittgenstein thought philosophy was about showing that traditional problems were confusions, but he and Carnap disagreed about what to do about it.

Things to Think About

If Carnap is right that we can choose our language, what happens when two people choose different languages and can’t agree on anything? Is there any way to resolve such a disagreement, or is it just a standoff?
Carnap said metaphysical questions like “Do numbers really exist?” are pseudo-questions. But doesn’t it feel like there’s a real question there? What would it take to make that feeling go away?
The principle of tolerance says “in logic there are no morals.” But don’t some ways of thinking lead to bad outcomes? Should there be constraints on what languages we allow—for instance, languages that make it easy to deny science or justify cruelty?
If you could design your own language for thinking about one area of your life (friendships, school, your future), what would you want it to look like? What would it make easy to say, and what would it make hard or impossible to express?

Where This Shows Up

Arguments about “fake news” and “alternative facts” often turn on exactly the kind of framework-relative thinking Carnap discussed—what counts as evidence depends on what language you’ve accepted.
Debates in theoretical physics about whether particles are “real” or just mathematical conveniences echo Carnap’s internal/external distinction.
The design of programming languages is a practical version of what Carnap advocated: programmers deliberately choose rules and conventions to make certain kinds of thinking easier.
Political arguments about “objectivity” often boil down to disputes about which framework we should use, not about facts—exactly the kind of confusion Carnap wanted to clear up.