| dc.description.abstract | This thesis examines two claims about consciousness in animals by the neuroscientist Joseph LeDoux: 1. Science cannot tell us anything about consciousness in animals, and 2. Animal behavior can be explained in terms of neuroscience. He argues that views that animals are conscious are not based on science. His view stands in contrast to the increasingly accepted view that many animals are conscious. The debate on animal consciousness is complex. Many disciplines are engaged and the debate involves different definitions, terminology, questions and theories. LeDoux raises important points that should be addressed, clarified and responded to. I discuss challenges with LeDoux’s narrow definition of consciousness, which requires explicit knowledge, and argue that his claims about the relation between science and consciousness are inconsistent. If correlates of consciousness can tell us about consciousness in humans then they can also tell us about consciousness in animals. I then examine his second claim. I argue that his neuroscientific theory of consciousness does not warrant exclusion of animal consciousness. His explanation of behavior is not exhaustive. I then identify and argue against two underlying assumptions in his model of the relation between stimuli, consciousness and behavior. I argue that the causal content of consciousness is assessment of behavioral performance in relation to stimuli. Consciousness hence does not directly affect behavior, but indirectly through modification of future behavior. This requires direct access to experience of own behavior, but does not require any kind of explicit knowledge. I discuss reinforcement in light of artificial and natural learning systems, affordances and predictive coding and conclude that consciousness is the discrepancy between expected and actual reinforcement. Consciousness is learning. I propose a new positive marker of consciousness: the ability to modify and learn new behaviors in a changing environment. This generates testable predictions and is a promising direction for the future studies of correlates of consciousness. | nob |