The simple logical error that the previous paper shows in the standard framework

The inference is:

1. Our measures of space and time are relative.

2. Therefore space and time are relative.

It looks like a substitution, but "relative" isn't the same word in both places. In (1) it means defined by relation to bodies; frame-dependent number. In (2) it means has no independent existence. Two different senses, one token. The argument only appears to go through because the same word carries both, and the word's relations — its place in the language — drag the second sense along with the first.

That's why it feels like it follows and doesn't. It's an equivocation, and it's hidden by vocabulary rather than stated as a premise.

And the general form: a fact about a measure is a fact about the measure. It cannot transfer to the thing measured by being restated with the same adjective. "My thermometer reads in Celsius" doesn't make temperature Celsius. "My coordinates are conventional" doesn't make the manifold conventional.

So the error has two layers. The invalid inference, and the vocabulary that conceals it. Newton named the first in the Scholium. He didn't name the second, because the word hadn't started doing that work yet.