reflexive integrity

Brian Cantwell Smith has a simple four-stage model of how language changes during a transition between a “prior” era and a “successor” era.

1. Conservative. The language of the prior era is used uncritically, its epistemology and ontology taken for granted. Much everyday work can be accomplished here.
2. Reactionary. The language of the prior era is still in use, but its objects are denied. There is an equivocation on the denial, for it is not yet clear if a theory is wrong, or if the criteria are wrong, or if the words themselves have lost their meaning.
3. Liberal. It has become clear that the conceptual framework is flawed. The prior language referred to something, but it's now hard to say what.
4. Radical. A wholly new language is formed. Goto step 1.
   

It is easy to be a reactionary, says Brian, but reactionaries almost always fail the basic test of reflexive integrity. Because their language is inherited from the prior era, it is almost never appropriate to their task. Verificationism, for example, fails the test of reflexive integrity, because the tenet that "claims should be accepted only if they are verifiable" is itself not verifiable.

But how do you get from 2 to 3? Or 3 to 4? Those steps seem quite mysterious, perhaps magical. Thar lie dragons! And paradigm shifts! And lions and tigers and bears (oh, my)!

[yes, it is clearly that time of year when I have essays due, for I am spending all of my time writing on my blog]