Knowledge Representation for Animal Reasoning

Many experiments and observations indicate that such cognitive abilities as understanding and reasoning may be assigned also to (higher) animals. We are aiming at a construction of a semantic framework appropriate for an elaboration of the intuitive idea about animal understanding and reasoning, or more generally: about understanding and reasoning of agents with a pre-language behaviour. Meanings and knowledge representation are specified in this paper in terms of distinguishing criteria. A special attention is devoted to distinguishing criteria of situations and events. It is said, that an agent understand a situation or an event, if it has a corresponding distinguishing criterion at its disposal. Rules of reasoning are special distinguishing criteria they assign some distinguishing criteria (meanings of the consequences) to other distinguishing criteria (meanings of premises). Knowledge representation and reasoning based on distinguishing criteria is adaptive distinguishing criteria are formalized as functions and a fine-tuning of distinguishing criteria is possible as a result of interactions with an environment. We believe that the investigation of formal models of animal reasoning is important for understanding of a quick extra-logical human reasoning and for a construction of artificial reasoning agents.