In the previous article Changing the Mind I have argued that cognitive systems might be changed by new information they assimilate from the environment and incorporate into their structure.
If new procedures and representations can arise in the mind, the original structures might completely fall into disuse. So it is possible that there is no fixed, unchanging core of cognition at all. Instead, the initial structures might merely act as a scaffold that is used only temporarily to generate new structures that can then replace it. This scaffold structure might be genetically determined and it might vary between people, but it would not make sense to talk of a “human nature” again if cognition can leave that structure behind and there is no fixed unchanging core of human cognition. Instead, cognition would largely be determined by culture. For culture in turn, the same would hold as for the individual cognitive system: that there is no fixed core and there are no general laws governing it.
For cognitive science this means that its scientificity cannot be derived from describing cognition in terms of fixed procedures and representational formats that would play the role of scientific laws. Cognitive systems would instead have to be described as systems that are “creative systems” in the sense that they can develop out of the scope of any formal or algorithmic description one can give for them. Any formal description of such a system is incomplete.
In the realm of cognitive science, which includes cognitive psychology and linguistics, a different model of science has to be employed. A model of science that tries to identify fixed laws of the objects it is describing fails if these objects do not have such fixed laws. The “Deductive-Nomological” or “Hempel-Oppenheim” model of science where scientific explanations are derived from laws and particular circumstances fails if the particular circumstances can be incorporated into the “laws” and change them (see http://plato.stanford.edu/archives/fall2012/entries/scientific-explanation/, section 2).
Moreover, the Popperian view (see http://plato.stanford.edu/archives/spr2013/entries/popper/) of defining science in terms of its falsifiability is not applicable when we deal with systems that do not have general, unchanging properties and where general statement using universal quantification (“for-all”-statements) are not possible. The attempt to describe the system using statistical laws also does not help. The system might show a certain statistical behavior at any given time but this behavior may change and does not describe the way the system actually works. Instead, we can only describe states of the system at certain times and the historical processes that lead from one such state to another. Here we need an extended model of science in which law-based descriptions are only a special case.
What we need, therefore, is a different concept of science that has more in common with the humanities. However, this was to be expected since what we want to develop here is a theoretical framework to describe the human mind.