We have just recognized that analysis is not justified in positing a matter of knowledge as an ideally separable moment, and that this matter, the moment we set it up through an explicit act of reflection, is already related to the world. Reflection does not work backward along a pathway already traveled in the opposite direction by constitution, and the natural reference of the matter to the world leads us to a new conception of intentionality, since the classical conception2 that treats the experience of the world as a pure act of constituting consciousness only succeeds in doing so to the exact extent that it defines consciousness as absolute non-being, and correspondingly pushes the contents back into an “hyletic layer” that belongs to opaque being. This new intentionality must now be approached more directly by examining the symmetrical notion of a form of perception and, in particular, the notion of space. Kant tried to draw a strict boundary between space as the form of external experience and the things given in that experience. Of course, it is not a question of a relation between a container and its content, since this relation only exists between objects, nor even of a relation of logical inclusion, such as the one that exists between the individual and the class, since space is anterior to its supposed parts, which are always cut out of it. Space is not the milieu (real or logical) in which things are laid out, but rather the 254means by which the position of things becomes possible. That is, rather than imagining space as a sort of ether in which all things are immersed, or conceiving it abstractly as a characteristic they would all share, we must think of space as the universal power of their connections. Thus, either I do not reflect, I live among things, and I vaguely consider space sometimes as the milieu of things, sometimes as their common attribute; or I reflect, I catch hold of space at its source, I think at this moment of the relations that are beneath this word, and I notice in this way that they are only sustained through a subject who traces them out and bears them; I pass from spatialized space to spatializing space. In the first case, my body and things, and their concrete relations according to up and down, right and left, and near and far, can appear to me as an irreducible multiplicity; in the second case, I uncover a unique and indivisible capacity for tracing out space. In the first, I am dealing with physical space and its variously qualified regions; in the second, I am dealing with geometrical space within which dimensions are substitutable, or I have a homogeneous and isotropic spatiality, and in this latter I can at least conceive of a pure change of place that would not modify the moving object in any way, and consequently I can conceive of a pure position distinct from the situation of the object in its concrete context. We know how muddled this distinction becomes, even on the level of scientific knowledge, in modern conceptions of space. We would here like to confront this distinction, not with the technical instruments adopted by modern physics, but rather with our experience of space, the ultimate authority (according to Kant himself) of all knowledge touching upon space. Is it true that we are faced with the alternative either of perceiving things in space, or else (if we reflect and if we wish to know what our own experiences signify) of conceiving of space as the indivisible system of connecting acts accomplished by a constituting mind? Does not the experience of space establish unity through a synthesis of an entirely different type?