In the Introductions to the 1794 edition of the Science of Knowledge, Fichte emphasizes that his project is to make consistent Kant's original presentation of transcendental idealism in the Critique of Pure Reason. Like Hegel, Fichte also praises Kant's discovery that all thought and representation requires the original synthetic unity of the apperception. Where Fichte and Hegel differ in their interpretation of Kant's claim is an interesting question, but I will not go into it here. At least for Fichte, the Critique of Pure Reason is a pile of material which is yet to be assembled into a building. The deduction of the categories, the deduction of space and time, and the critique of the ideas of reason, are all done separately in different chapters. Again, it is interesting to think about the kind of necessities which prompted Kant to opt for this arrangement. Kant is definitely following the traditional Leibnizian-Wolffian presentation of metaphysics and logic, and the reasons for doing this can be both political/strategic (with the audience in mind) and philosophical (with the subject-matter in mind.) For Fichte, philosophically speaking, the deduction of the pure unity of the apperception ought to have come first, and the categories ought to have been deduced as a consequence of the very nature of this pure unity.
Another point which Fichte emphasizes is the distinction between Spinozism and critical idealism. According to Fichte, both systems start from absolutely unconditional first principles. Spinozism allows reason to overstep the boundaries of the self, while critical idealism doesn't. In other words, critical idealism, as it lets its thoughts progress, always asks the reflective question, "how is this thought constructed by the self?," while Spinozism speculates what goes on beyond the self without questioning how the self might be thinking about these thoughts. It is worth noting here that this does not mean that critical idealism is what Meillasoux calls "correlationism" in his After Finitude. Fichte concedes that of course things exist without our thinking about them. What Fichte denies is that these things are determined in such-and-such a definite conceivable way without the input our thoughts. As soon as one thinks that "A is so-and-so," one is already contributing to the determination of the object, namely A. Critical idealism thus asserts the necessary relation between thoughts and the thinker, but it does not assert such a relation between the thing and thought. For example, fossils may perfectly have existed prior to our thinking about them. It is a totally different issue to raise the question: "how are we led to characterize a fossil in just this way by means of our thoughts?" Existing temporally prior to any human thinking is part of the determination of a fossil, and this determination is the determination of thought.
The opening paragraphs of the Science of Knowledge are elegantly written, although in the 1794 version, the transcendental unity of the apperception, the pure self, is only described without being rigorously deduced. Nonetheless, Fichte argues that the self exists as a self-relating, and that thoughts as such also exist in this manner. This self-relation, in which "I am...." is met with another thought which is also "accompanied" by the "... I," is already engendered with the categories of distinction, opposition, and limitation. Fichte's arguments here have a striking resemblance to Hegel's in the latter's Science of Logic. It is as if Hegel removed the Fichtean self from the argument and replaced the initial premise with pure being. In order to really step out of an absolute egoism, perhaps it is necessary to re-read Fichte and think about the real implications of Hegel's modification of Fichte's method.
