Thomas Zinn

Frameworks · June 2026

Adapters as Translators

An adapter is a translator between two forms of the one thing. And it is more than a convenience: a working adapter is a proof. The moment you can faithfully turn A into B, you have shown they were never separate, only differently named. That gives the whole framework a test it can be held to.

This is a deeper look at one piece of The Unification Map: the adapter. The map's claim is that the world is more connected than our words for it, and that we cross the false boundaries by building translators. This is what those translators are, and why one of them, working, counts as evidence.

An adapter is a translator

An adapter takes one form of the one thing and converts it into another form something can receive. That is all a transducer is, and the word covers far more than gadgets.

Your eye is an adapter: a sliver of electromagnetic vibration becomes sight. Your ear is an adapter: pressure waves become sound. A thermal camera is an adapter: heat becomes a colored image. A radio is an adapter: an oscillation far below light becomes a voice. A microphone, a solar cell, a speaker, a screen all do the same job, each taking the one thing in one form and handing it over in another.

There are two shapes the handoff takes. Sometimes it goes straight to a human sense: the thermal camera puts a picture in front of your eyes. Sometimes it goes first to a machine, which reads it, works on it, and only then renders something a person can take in, the way a radio telescope's signal becomes data and then a picture on a screen. The second kind is the whole reason we build machines: to reach what a person alone cannot, and carry it back.

Why a working adapter is a proof

This is where the adapter stops being a convenience and becomes an argument.

When you build a device that faithfully turns A into B, you have not just moved information around. You have demonstrated something about A and B: that they were never truly separate. They were the same thing in two vocabularies, and the adapter is the dictionary between them.

Light and electricity look like different subjects until a solar cell turns one cleanly into the other. After that, the line between them reads less like a fact about the world and more like an old habit of speech. Sound and radio seem unrelated until a phone carries your voice as one and rebuilds it as the other. Every transducer that has ever worked is one more piece of evidence for the same conclusion: a boundary we trusted was not a boundary at all.

That reframes what an inventor is doing. Building an adapter is not only engineering. It is a claim about the structure of the world, tested by whether the thing works.

The convertibility test

If a working adapter is evidence of sameness, then "are these two things really one?" gets an operational answer instead of an argument. Call it the convertibility test:

  • A faithful adapter exists → the distinction was terminological. Light and electricity, sound and radio, heat and image: these are settled, because the translators run every day.
  • No adapter yet → a candidate. We suspect two things are one but have not built the dictionary. This is where the work is.
  • Provably no faithful bridge → maybe a real seam, not a false wall. We can detect gravitational waves but cannot convert into them; that may be a genuine edge of the world, not a missing gadget. Marking it honestly keeps effort from draining into it.

One guardrail keeps this from collapsing into "everything is everything": unity here means translatable, not identical. The differences (which form, what rate, how it behaves) are real, and they are exactly what make the variety of the world. The claim is only that the differences are bridges, not walls. Synonyms are different words that share a meaning; the things an adapter joins are different forms that share a nature.

The same move, in machines

The test does not stop at physics. The hardest version of it now lives in software, where the "forms" are vocabularies for data. One system calls a person a user, another a customer, another an account; the systems cannot speak, not because the person differs but because the words do. Every integration, every data migration, every interface wired to another is an adapter, a translator asserting that two differently-named things are one.

And it is where the test bites hardest, because machine adapters mostly declare equivalence ("this field means that field") rather than prove it. The open prize is a machine that could reliably recognize a synonym across any two vocabularies: the universal adapter. It is the thing today's machines are worst at, because they learn our fragmented words and reproduce every wall.

Why it matters

Give the framework this test and it stops being a mood and becomes something you can be held to. A boundary is real until someone builds the translator that dissolves it; a unity is just a feeling until the adapter runs. The work is to keep building the dictionaries, and to be honest about which walls are doors we have not opened yet, and which are edges of the world.


Part of The Unification Map. The rigorous version of the test and the band-by-band catalog of translators live in the Library principles and adapter inventory. For the ways a translator can fail, see Where adapters fall short.

Keep reading