'Ontology extension' in evolution and in development, in animals and machines.
Abstract:
A distinction can be made between definitional and substantive ontology extension. How the latter is possible is a deep question for AI, psychology, biology and philosophy. All information-processing systems have direct access only to limited sources of information. For some systems it suffices to detect and use patterns and associations found in those sources, including conditional probabilities linking input and output signals (a somatic sensorimotor ontology). Sometimes it is necessary to refer beyond the available data to entities that exist independently of the information-processing system, and which have properties and relationships including causal relationships that are not definable in terms of patterns in sensed data (an exosomatic ontology). This is commonplace in science: scientists postulate the existence of genes, neutrinos, electromagnetic fields, chemical valencies, and many other things because of their explanatory role in theories, not because they are directly sensed or acted on. Does this also go on in learning processes in infants and hatchlings that discover how the environment works by playful exploration and experiment? Is ontology extension beyond the sensor data also set up in the genome of species whose young don't have time to go through that process of discovery but must be highly competent at birth or hatching? Is there anything in common between the different ways ontologies get expanded in biological systems? This relates to questions about what a genome is, and about varieties of epigenesis, as well as to the varieties of learning and development that need to be considered in AI/cognitive science/robotics/psychology.
This work extends the paper presented with biologist Jackie Chappell at IJCAI-05 on The Altricial-Precocial Spectrum for Robots.
It is part of the theoretical work being done on the EU-funded CoSy robot project. It challenges the current emphasis on such themes as symbol-grounding, sensorimotor-based cognition, and the role of embodiment in constraining cognition. It explains how we can be mathematicians, scientists and philosophers as well as animals with bodily functions and competences.
