self-reference in autopoiesis

Onar Aam (onar@HSR.NO)
Sun, 27 Aug 1995 09:36:17 +0100


Cliff and others have requested a closer explanation to the role of
self-reference in autopoiesis. In my view this is a very profound and
non-trivial aspect of this theory which I believe is not adequately
addressed, even by Maturana and Varela.

Reference is by nature *symbolic*. An example of self-reference is
therefore "this is a statement" where "this" in this case is a
reference to the entire statement which it is embodied in. That
autopoiesis is self-referential implies that it is linguistic. This
is not the case. Autopoiesis is "pre-linguistic" and constitutes the
primitives of both logic and language. By "primitives" I mean the
gestaltic phenomena (unity, distinctness, continuity,
background/foreground) that is presupposed in the linguistic domain.
These primitives are exactly the same that were addressed by Bruce in
his critique of my self-confirming definition of hypersets.

To illustrate that autopoiesis belongs to the prelinguistic domain
consider this: autopoiesis has a very binary mode of existence.
Either something is autopoietic or it is NOT. There is no inbetween
state. This is exactly the structure of logical truth. But this is a
meta-property of autopoiesis. The autopoietic entity cannot itself
observe this binary property.

Thus, when we say that autopoiesis is self-referential then we have
to understand this in a strictly pre-linguistic manner. With this
stroke we are equipped to explore autopoietic self-referentiality.

---------------

We know that autopoietic systems are built from dissipative
structures that form a circle. These are what M&V calls the components
of the autopoietic system. Maturana notes that each of these
components are defined by their role and functioning in the
autopoietic system. But strangely the autopoietic system is also
defined by the circular organization of its components. This is how
autopoiesis is self-referential. The autopoietic system defines its
components while at the same time being defined the very same
components.
This strange notion is in fact understandable in terms of
dissipative structures. As we've seen earlier a dissipative structure
is defined by its far-from-equilibrium. It can only exist when there
is a sufficient energy flow through the system. In this respect the
dissipative structure is not self-referential. It assumes an energy
flow from outside the dissipative system. Now, each component of the
autopoietic network may be regarded as a dissipative structure.
Viewed in isolation each and every component of the system is
produced from their respective outsides. That is, each component
assumes an energy flow from their outside. HOWEVER, when viewed as a
whole we see that the energy flow of the total system is closed. The
chain of components forms a ring.

Onar.