We as neurobiologists studying vision usually do not ask the question what is it to see? because we considered it a philosophical and not a biological question, and do not realize that we answer it implicitly by doing what we do in our research. This implicit answer entails the basic assumption that we exist in an objective world independent of our acts of cognition and accessible to our knowledge. My contention is: a) that by answering the question what is it to see? one can show that this assumption cannot be sustained because the phenomenon of perception cannot consist in a process of grasping the features of an independent world of objects; and b) that by reflecting upon the nature of a scientific explanation one can show that this assumption is unnecessary because a scientific explanation is a particular kind of coordinations of actions in a community of observers that does not entail it. In this context, a) by putting objectivity in parenthesis, that is, by using the operational generation of scientific explanations and not the object as the criterion of validation of my statements, and, b) by recognizing that the nervous system operates as a closed neuronal network in the generation of its states of activity, I show that the phenomenon of perception arises in the description of an observer as a manner of referring to the operation of an organism in congruence with the particular environment in which it is observed. In these circumstances, my answer to the initial question is: to see is a particular manner of operating as a closed neuronal system component of an organism in a domain of structural coupling. Finally, I propose that by dwelling in language as a peculiar system of coordinations of actions, we human beings bring forth an objective world through using our own changes of states as describers that specify the objects that constitute it.