The I think must accompany all my representations, for otherwise something could be represented in me which could not  be thought. What makes representation mine - unity does not belong to the object- is a function of the subject and presupposed in it. So the foundation of the object lies in the subjet. The notion of object structurally presupposes the subject. The 12 modes of unifying the multiple (12 possible syntheses) presuppose an originary unity: consciousness-self consciousness which Kant calls the cogito. Each multiple of intuition (space/time) has a necessary relation to the cogito. The transcendental and necessary unity of apperception is what is the  end product of a process of connection and synthesis of phenomena, which in turns depend on the application of the representation of an object in ituition to experience.

This cogito is necessary for us for if we lacked it we would modify according to each perception.

‘For the manifold representations which are given in an intuition would not all of them be my representations, if they did not all belong to one self-consciousness, that is, as my representations (even although I am not conscious of them as such), they must conform to the condition under which alone they can exist together in a common self-consciousness, because otherwise they would not all without exception belong to me.’  […] for the reason alone that I can comprehend the variety of my representations in one consciousness, do I calll them my representations, for otherwise I must have as many-coloured and various a self as are the representations of which I am conscious.’ (CPR p 94-95)

The I think is an act of spontaneity, and Kant also calls it pure apperception. Apperception in Kant means consciousness, but as an act, an activity, rather than a state. In the Anthropology we will find discussions of the role of apperception in relation to the Gemut (inner sense) that Foucault will dwell on in his dissertation. Kant is keen to differentiate inner sense from pure apperception, and judges psychology on the basis of the failure to apply this distinction between them.

‘If we consciouly imagine for ourselves the inner action (spontaneity), whereby a concept (a thought) becomes possible, we engage in reflection; if we consciously imagine for ourselves the susceptibility (receptivity), whereby a perception (perceptio), i.e. empirical observation, becomes possible, we engage in apprehension; however, if we consciously imagine both acts, then the consciousness of one’s self (apperceptio) can be divided into that of reflection and that of apprehension. Reflection is a consciousness of the understanding, whilst apprehension is a consciousness of the inner sense; reflection is pure apperception, but apprehension is empirical apperception; consequently, the former is falsely referred to as the inner sense. In psychology we investigate ourselves according to our percpetions of the inner sense; but in logic we make the investigation on the grounds of what the intellectual consciousness supplies us with. Here the self appears to us as twofold (which would be contradictory): (1) the self, as the subject of thinking (in logic), which means pure apperception (the merely reflecting self) of which nothing more can be said, except that it is entirely simple perception. (2) The self, as the object of the perception, consequently also part of the inner sense, contains a multiplicity of definitions which make inner experience possible. To ask whether or not a man conscious of different inner mental changes (either of his thoughts or of fundamental principles assumed by him) can say that he is the selfsame man, is an absurd question. For he can be conscious of these changes in the first place only on condition that he represents himself, in the different situations, as one and the same subject. The human ego is indeed twofold as regards its form (manner of representation), but not with respect to its matter (content).’ (Anthropology from a Pragmatic Point of View, footnote to p.18).

 ‘In the system of psychology, the inner sense is commonly held to be one with the faculty of apperception, while we, on the contrary, carefully distinguish them’. (CPR, p105)