3.325 In order to avoid such errors, we must use a symbolism that excludes them by not using the same sign for different symbols and by not using signs that signify in different ways in what appears to be the same way. A symbolism then that obeys logical grammar – logical syntax.

(The concept-script of Frege and Russell is one such language, though admittedly it does not yet exclude all errors.)

Cf. 5.5563.

A thought: Wouldn’t it be ironic if right here Wittgenstein were to make just the kind of mistake he has just warned about? The first sentence seems fine. But what is “logical grammar”? Could grammar possibly be illogical? No. He means the grammar of logic, the syntactical rules of logic. But what are these? Just the rules of logic, surely. And, just as surely, grammar already obeys these. Perhaps he means a language that more clearly follows the laws of logic. But it is clear already that all grammar must, can only, do this, and the laws of logic are just generalizations (hypothetical ones, we might say) derived from natural language. So the language described might be both impossible to create satisfactorily and unnecessary anyway.