Sergio F. Martinez

In the philosophy of science, abstraction has usually been analyzed in terms of the interface between our experience and the design of our concepts. The often implicit assumption here is that such interface has a definite identifiable and universalizable structure, determining the epistemic correctness of any abstraction. Our claim is that, on the contrary, the epistemic grounding of abstraction should not be reduced to the structural norms of such interface but is also related to the constraints on the cognitive processes of specific abstractions. This suggests that we should understand abstraction as embodied in different kinds of abstraction practices.

En este texto se examina críticamente la teoría combinatoria de las representaciones científicas de Andoni Ibarra y Thomas Mormann. El núcleo de la crítica va dirigido a mostrar que una serie de estudios sobre la ciencia, que ellos mismos mencionan, sugiere que la clasificación en tipos de representaciones propuesta es problemática. Es más, esos mismos estudios muestran que por lo menos muchas representaciones tienen una dimensión histórica que parece imposible capturar por medio del tipo de formalismo propuesto. /// Ibarra and Mormann's combinatorial theory of scientific representations is subject to critical scrutiny. It is argued that the classification of types of representations offered cannot serve as a basis for an analysis of the variety of representations that happen to arise in science. Empirical studies on ways of representing used in science strongly suggest that a formal theory of representations of the kind proposed is not able to capture the historical dimension of the structure of representations.

There are (at least) two ways to think of the differences in basic concepts and typologies that one can find in the different scientific practices that constitute a research tradition. One is the fundamentalist view: the fewer the better. The other is a non-fundamentalist view of science whereby the integration of different concepts into the right abstraction grounds an explanation that is not grounded as the sum of the explanations supported by the parts. Integrative concepts are often associated with idealizations that can successfully set the stage for different phenomena to be compared or for explanations of different phenomena to be considered as jointly increasing our understanding of reality beyond that which each explanation provides separately. In this paper, our aim is to argue for the importance of the notions of an ?affordance? and ?scaffolding? as integrative concepts in the cognitive sciences. The integrational role of the concept of affordance is closely related with the capacity of affordances to generate the scaffoldings leading to the integration. The capacities of affordances that turn them into (stable) scaffoldings explain why such notions are often used interchangeably (as we shall see). On this basis, we aim to show that the concepts of affordance and scaffolding provide the sort of epistemic perspective that can overcome common complaints about the limits and unity of the cognitive sciences once claims about extended cognition are taken seriously