It is shown that the method of operationaldefinition of theoretical terms applied inphysics may well support constructivist ideasin cognitive sciences when extended toobservational terms. This leads to unexpectedresults for the notion of reality, inductionand for the problem why mathematics is sosuccessful in physics.A theory of cognitive operators is proposedwhich are implemented somewhere in our brainand which transform certain states of oursensory apparatus into what we call perceptionsin the same sense as measurement devicestransform the interaction with the object intomeasurement results. Then, (...) perceivedregularities, as well as the laws of nature wewould derive from them can be seen asinvariants of the cognitive operators concernedand are by this human specific constructsrather than ontologically independent elements.. So, reality in so far it isrepresented by the laws of nature has no longeran independent ontological status. This isopposed to Campbell's `natural selectionepistemology'. From this it is shown that thereholds an incompleteness theorem for physicallaws similar to Gödels incompletenesstheorem for mathematical axioms, i.e., there isno definitive or object `theory of everything'.This constructivist approaches to cognitionwill allow a coherent and consistent model ofboth cognitive and organic evolution. Whereasthe classical view sees the two evolutionrather dichotomously. (shrink)