We (in the Rashevsky-Rosen school of relational biology) have sometimes been asked by experimenters why we do not propose explicit experiments for them to perform, and subject our approaches to verification at their hands. We do not do so because it is precisely physicochemical particulars that are abstracted away in the process of generating relational models. There is no kind of one-to-one relationship between relational, functional organizations and the structures that realize them. A functional organization cuts across physical structures, and a physical structure is simultaneously involved in a variety of functional activities. So an (M,R)-system is not realized by identifying its components and maps in a ‘concrete’ biological example. To tackle the biological realization problem of (M,R)-systems, one ought not to be seeking physicochemical implementations of what the relations are, but, rather, one ought to be seeking interpretations of what the relations do.
In any case, the basic questions of biology, in our view, are not empirical questions at all, but, rather, conceptual ones. ‘Conceptual’ and ‘qualitative’ are not dirty words, but rather indicators of genericity.
As always, when one attempts to do theory, one is confronted with the trivia: Is it testable, and if so, how? People have been brought up with the prejudice that a scientific theory that is not testable is worthless. It is often considered part of the theorist’s job to make theory verifiable, in effect to construct some kind of experimental protocol for the sole purpose of falsification. As long as ‘experimental test’ exclusively takes on the conventional sense that prescribes ‘to verify some kind of specific physicochemical operation on individual systems’, there is in principle no way that the relational descriptions could in fact be ‘tested’.
A relational description of an organism is as valid, as realistic, a description as any conventional physicochemical one. But it is a description pertaining to a class of physically diverse (though functionally equivalent) systems. A well-constructed model creates a reality of its own: there is no model-independent test of reality. There are more kinds of experiments than physicochemical ones. Conceptual experiments are common in psychology and sociology, for example. Biology has a lot to learn from social sciences.
Relational biology is ‘decoding from formal system to realization’. Experimenters need something to verify, couched in terms of some specific observation, or physicochemical experiment, that they can perform. They need the ‘encoding’; that is to say, they need precisely what is not in our approach.
An act of observation is a quintessential act of abstraction. The observation of a single quality of a natural system is indeed the greatest kind of abstraction that can be made of that system, that of analytic modeling by a single real number. The development of theoretical science is thus a synthesis: an attempt to combine observations in such a way that our view of systems becomes less abstract than it could be if we were restricted to observation alone. There is no antagonism between ‘theory’ and ‘experiment’. That unfortunate mirage is an artifact of the antagonism between ‘theorists’ and ‘experimenters’.