Mathematical language and scientific knowledge

If we change the symbols used to describe something, have we changed the thing itself? That's the question I keep coming back to as this discussion connects mathematics, language, and scientific knowledge.

We begin with a familiar observation: science often uses mathematics to explain how the world works. Then comes the more controversial claim that mathematics is a kind of language. The stated connection is a systematic set of signs, with qualities such as internal coherence, elegance, and depth. On this view, applying mathematical concepts to the world produces statements about how it functions.

I want to push back: aren't mathematical relationships different from the words we invent? Two objects added to two objects give us four, whatever symbols we choose to write. That is a useful reaction because it helps separate a description from what it describes. But our objection isn't the writer's position, and resolving the philosophy of mathematics isn't our job here. We need to understand the proposed connection well enough to follow its implications.

If mathematics counts as language, questions about language's relationship to the world become relevant to scientific knowledge. That is the bridge into paragraph 2.

The linguistic debate has two sides. One says language corresponds to objects and behaviors in an essential way. The other says the connection depends on agreed conventions. In simpler terms, is the connection built into how things are, or does it depend on how people have agreed to describe them?

Think about a pen. The word pen and the object on the desk aren't the same thing. We can recognize that speakers agree on what the word means while still asking what, beyond that agreement, connects the description to the object. Don't reduce the second position to nothing exists. The disagreement concerns the relationship between language and things, and what that relationship lets us know.

Paragraph 3 develops the convention-based view. First, language use changes as practices and theories change within a discipline. Then the claim gets stronger: a statement counts as true only when there are no promising alternatives that would lead people to question it.

Why should the availability of another explanation determine truth? My instinct is to ask whether a statement accurately describes the world. The linguists described here place emphasis on accepted ways of describing it and the alternatives available. Keep that difference visible instead of silently replacing their claim with the one we prefer.

The writer then applies this view to science. A mathematical statement may be treated as accounting for a phenomenon, but mathematical language itself doesn't guarantee that the account corresponds to every aspect of the world. According to this view, predictive power or an efficient method can lead scientists to accept a statement as an explanation. What begins as an analogy or metaphor then functions as the accepted account until a more compelling one replaces it.

The idea of an accepted explanation being revised is easier to follow than the abstract vocabulary. But we should retain the specific claim being advanced: community acceptance gives the mathematical description its role as an explanation. This isn't a declaration that scientists choose statements randomly or that every equation is false.

The ending asks what follows if the convention-based account is right. Linguists are asking what language can tell us without an essential correspondence to reality. Scientists should ask the comparable question about mathematical models: what work do those models do in acquiring knowledge?

That's the destination. An initial summary such as mathematics is a language misses the final call for investigation. The discussion uses the linguistic debate to identify a question science has not yet seriously addressed.