der are therefore necessarily
conceived as relative. So if we wish to represent them to ourselves as
absolute, we perceive that we are going to and fro like a shuttle
between the two kinds of order, passing into the one just at the moment
at which we might catch ourself in the other, and that the supposed
absence of all order is really the presence of both, with, besides, the
swaying of a mind that cannot rest finally in either. Neither in things
nor in our idea of things can there be any question of presenting this
disorder as the substratum of order, since it implies the two kinds of
order and is made of their combination.

But our intelligence is not stopped by this. By a simple _sic jubeo_ it
posits a disorder which is an "absence of order." In so doing it thinks
a word or a set of words, nothing more. If it seeks to attach an idea
to the word, it finds that disorder may indeed be the negation of order,
but that this negation is then the implicit affirmation of the presence
of the opposite order, which we shut our eyes to because it does not
interest us, or which we evade by denying the second order in its
turn--that is, at bottom, by re-establishing the first. How can we
speak, then, of an incoherent diversity which an understanding
organizes? It is no use for us to say that no one supposes this
incoherence to be realized or realizable: when we speak of it, we
believe we are thinking of it; now, in analyzing the idea actually
present, we find, as we said before, only the disappointment of the mind
confronted with an order that does not interest it, or a swaying of the
mind between two kinds of order, or, finally, the idea pure and simple
of the empty word that we have created by joining a negative prefix to a
word which itself signifies something. But it is this analysis that we
neglect to make. We omit it, precisely because it does not occur to us
to distinguish two kinds of order that are irreducible to one another.

We said, indeed, that all order necessarily appears as contingent. If
there are two kinds of order, this contingency of order is explained:
one of the forms is contingent in relation to the other. Where I find
the geometrical order, the vital was possible; where the order is vital,
it might have been geometrical. But suppose that the order is everywhere
of the same kind, and simply admits of degrees which go from the
geometrical to the vital: if a determinate order still appears to me to
be continge