The positivists claimed that there were multiple languages (Goodman, Sellars) organized as formal systems, multiplied, but clear in their own rules (though that may demand disciplining ordinary language).
Formal systems taken in the abstract have no fixed essences, so any change makes a new system, and they stand in relations with one another, but in no essential hierarchy.
Objection: if then every 'mistake' makes a new system what identity do they have? Answer: not every mistake makes a consistent system. And there is some resistance, when a system exists. Because it exists only as embodied.
We do have social practices with some inertia to them, so there is some conflcit with radical indetermination of formal systems.
But that's the issue again: those practices, do they have guidance from grammars and formal systems?
We should not make the mistake of equating a grammar with a formal system. Though I admit that I tend that way, perhaps coming from my early exposure to Wilfrid Sellars and my dissertation work with Stephan Körner. I correct myself with influences from Derrida and others about the status of formal systems within contexts and processes they don't define or control. But those deconstructive (and other) influences have their own misconceptions.
(c) David Kolb, 1 August 2001