A distinction is made between two definitions of animal cognition: the
one most frequently employed in cognitive sciences considers cognitio
n as extracting and processing information; a more phenomenologically
inspired model considers it as attributing to a form of the outside wo
rld a significance, linked to the state of the animal. The respective
fields of validity of these two models are discussed along with the li
mitations they entail, and the questions they pose to evolutionary bio
logists are emphasized. This is followed by a presentation of a genera
l overview of what might be the study of the evolution of knowledge in
animals.