There is an odd transitivity to estrangement: if A and B appear undistinctive in their milieu M, but if C clearly does not belong in M, then what happens if A, B, and C are undistinguished? There are two possibilities, either A and B are revealed as estranged from M, or C does after all belong to M, and M is not what you thought it was. In the latter case, since a milieu is not an object to be estranged but one to be estranged from, it must be your misunderstanding and hence your estrangement. You were not, as you thought, in a position to speak for M. In this way, out of a small seed, out of something about which it appears that little was at stake in the judgement (C does not belong to M), a large and permanent division of the world arises. There is a resolution of this logic if we introduce time and development as possibilities. This would mean that what appears to be estranged now will in the future become fully integrated. If this comes about through a change in M then it is insecure, since a milieu ought to be assumed to have real criteria for exclusion and not to be a merely propositional entity, a reflection of the whim of its contents. It would mean that while the basis for the distinction is still present it is to be deliberately ignored, or reinterpreted as signifying the opposite of what it formerly did. It is as if you were to say that the milieu does not exist where you thought it did, but rather here in your mind, in the rules you make for it. But since one of the defining rules is that something must be excluded, what you now exclude is the idea that M exists apart from your mind. There may however be a kind of change that is not development: M was realised to have always included C, and thinking otherwise was an error which once corrected need have no further existence. Taking this fork leads ultimately to abolishing milieus altogether. The only way to do away with the effect of estrangement is to puncture the entire system, not to shift around the markers.