Abstract

El debate entre Gary Marcus y Jeff Elman, en Marcus y Elman se enmarca dentro del debate general en Ciencia Cognitiva entre la hipótesis clásica y la hipótesis conexionista. Marcus critica lo que se denominaConexionismo Eliminativista, que incluye los modelos conexionistas que eliminan el papel de los símbolos y las reglas explícitas. Según Marcus, los modelos conexionistas eliminativistas no logran generalizar del modo como hacen los humanos y, por tanto, el Conexionismo Eliminativista no constituye una buena explicación de la cognición humana. En este trabajo intentaré abordar el debate entre clásicos y conexionistas a partir de los experimentos llevados a cabo por Marcus en Marcus et al. y las críticas de Elman almismo, e intentaré argumentar que Marcus no consigue una crítica consistente al Conexionismo Eliminativista y, asimismo, no logra ofrecer una buena defensa del Conexionismo Implementacional.The debate between Gary Marcus and Jeff Elman, in Marcus and Elman is situated in the general debate in Cognitive Science between the Classical hypothesis and the Connectionist one. Marcus is against what is called Eliminative Connectionism, which includes the connectionist models that eliminate the role of symbols and explicit rules. According to Marcus, eliminative connectionist models are not able to generalize in the way humansdo. Thus, Eliminative Connectionism cannot furnish us with a plausible model of cognition. In this notice I shall approach the debate between classicists and connectionists as framed by Marcus in Marcus et al. . I shall try to argue contra Marcus: his strategy against Connectionism fails in getting a good argument; therefore he does not get a sound defence of Implementational Connectionism