Things don't have to be identical to their linguistic (cognitive) descriptions.


 * Suppose that there is some degree of noncorrespondence between cognitive syntax and perceptual content (observed phenomena). Then there exist items of perceptual content which do not correspond to or coincide with cognitive syntax. But if these items do not coincide with cognitive syntax, then they are unrecognizable, i.e. inobservable (since cognitive syntax is by definition the basis of recognition). But then these items are not included in perceptual reality (the set of observable phenomena), and we have a contradiction. Therefore, perceptual reality must coincide with cognitive syntax.
 * Suppose that cognition is not the only model for self-organizing systems, i.e. that such systems can be essentially non-homomorphic to cognitive processing. If so, then they lack meaningful cognitive representations, defying characterization in terms of mental categories like space, time and object. But then they fall outside reality itself, being indistinguishable as causes, effects, or any kind of phenomena whatsoever. In a word, they are irrelevant with respect to that part of reality isomorphic to human cognition. It follows that by any reasonable definition, reality is "cognitive" up to isomorphism with our own mental structures.