The starting assumption is that good action leads to higher survival chances and thus to evolutionary advantage. The second assumption is that better information leads to better action.
The chapter provides a setting (a base, or starting point) for the investigation by looking at assumptions and boundary issues. We also clarify the information processing approach that is used in the remainder of this investigation.
The assumptions and boundaries are meant to be non-controversial, hopefully just reiterating the obvious. Disagreements at this stage will probably cause problems for the subsequent discussion.
We want to look at multiple species, and not just mammals. We restrict ourselves to species with limbs and a central skeleton, and with eyes that focus on a single image. We generally talk about humans or mammals like kittens and puppies, but we also include chicks (for imprinting). The analysis and synthesis is meant to be general, across multiple species, but without explicitly testing the boundaries.
We start with the assumption that information flow and information processing can be investigated independently, without investigating the 'hardware', the brain and the neurons that carry the information and do the processing. This assumption differentiates this investigation from the more typical neuroscience investigations.