These are quite primitive forms of representation, and do not look much like paradigm cases, such as those of objects or numbers, and that poses a challenge to this model: how can the interactive model account for representations beyond the simple cases of predicating types of interactive appropriateness to current environments? That question, and others, are addressed elsewhere (Bickhard, 1993, 1999, 2000; Bickhard & Terveen, 1995; Campbell & Bickhard, 1986), so I will only indicate a little of the direction of
The selections of a recursively self-maintenant system necessarily involve some sort of appropriate sensitivity to the environment. The paramecium must detect that the current environment is of a type that is usually appropriate for swimming. The selections, then, are conditional on prior detections, prior outcomes of earlier engagements with that
environment. Such engagements may be relatively simple and passive, but they may also be potentially complex and fully interactive themselves: the internal outcome of an interaction could serve to detect a relevant environmental type as well or better than a passive input processing detection.
The relevant implication here is that interactions can not only be selected and be indicated as potential selections, they can also function to differentiate environmental types
for subsequent indications of appropriateness of interactions. Selections, then, can be based on indications of potentiality, and indications of potentiality can be conditional: if this interactive outcome has occurred, then that interaction is now indicated as appropriate.
Conditional indications of interactive potentiality are the key to more complex forms of representation. Such conditional relationships can branch — a single interaction outcome can function to indicate multiple further interactive potentialities — and they can iterate — completion of interaction A may indicate the potentiality of B, which, if completed, would indicate the potentiality of C, and so on. That is, they can form
potentially quite complex webs of conditional interactive potentialities, and particular kinds
of such webs, I argue, yield the emergence of higher kinds of representation, such as of manipulable objects (Bickhard, 1980, 1993; Piaget, 1954). Representation of abstractions, such as of numbers, requires still further development (Campbell & Bickhard, 1986).
Correspondence Models of Representation. For current purposes, I will take the interactive model to be a viable contender as a model of the nature of representation. Before turning to the examination of the biological foundations, note that
this model of representation, like the model of function, is in strong contrast to dominant
approaches. In particular, standard information semantic approaches to representation will construe the differentiations or detections involved in the interactive model as themselves