TY - JOUR T1 - Ecological Representations JF - bioRxiv DO - 10.1101/058925 SP - 058925 AU - Sabrina Golonka AU - Andrew D Wilson Y1 - 2016/01/01 UR - http://biorxiv.org/content/early/2016/06/15/058925.abstract N2 - Cognitive science has three main motivations for assuming that cognition requires representation. These are the need for intentional (meaningful) access to the world, poverty of perceptual access to that world, and the need to support ‘higher-order’ cognition (e.g. thinking about things in their absence). All representational systems must also address two major problems (symbol grounding and the need for system-detectable error). Mental representations attempt to address the three motivations but stumble over the two problems. Here we argue that James J Gibson’s ecological information fits the basic definition of a representation, solves both problems and immediately addresses the first two motivations. We then develop an argument (begun in Golonka, 2015) that informational representations and the resulting neural representations can also support ‘higher-order’ cognition. ER -