Only as regulative unity is the conceptual unity what it can and must be as unifying. The unity is not grasped, rather only if we look away from it in its determining of the rule is it then just as substantially the regulation which is determined in the view. This looking-away-from-it does not lose sight of it in general, but rather has in view precisely the unity as regulative.
Kant and the Problem of Metaphysics, pp.67-8
I should explicate this passage before going on to my reasons for citing it. Kant's Critique of Pure Reason establishes the "transcendental categories" or "pure concepts of the understanding," concepts that themselves contain nothing empirical, but which determine, in advance, the framework of all possible experience. There are twelve of these categories, which fall under four headings: Quantity (Unity/Plurality/Totality), Quality (Reality/ Negation/ Limitation), Relation (Inherence-Subsistence/ Causality-Dependence/ Community, or reciprocity between agent and patient) and Modality (Possibility-Impossibility/ Existence-NonExistence/ Necessity-Contingency). Any experience we have is structured by the interaction between some combination of these categories with the faculty of Sensibility (by which we receive impressions) and the faculty of Imagination (which produces unities out of sensible materials).
It's important to understand that the Understanding, or faculty of comprehension--though its concepts precede any experience and are thus entirely formal--are essentially directed toward experience. Their function of "translation" into concrete experience is what makes them what they are; a clumsy parallel might be with a doorway: the doorway is essentially something to be passed through. Otherwise it's just a hole of some size in some material (and no-one would ever care what size and what material without that "door-ness" as a motivation for finding out--i.e., the doorway is purely formal in itself). The problem with the metaphor: the essence of the doorway involves the range of things that are to pass through it (people, furniture, etc.); the categories don't essentially determine any particular phenomena, but are oriented toward particular phenomena in general.
The problem for Kant, then, is this: how do these absolutely contentless concepts "get" from the "general particular" (my phrase, one that I think of when writing about dialectics) to an actual, concrete experience of particulars? His solution is the "Schematism of Pure Understanding." The transcendental schemata are like rules that get us from the concept to the image. Every pure (i.e., nonempirical) concept has its transcendental schema, and that schema is produced by the operation of "pure a priori" imagination (as opposed to the empirical imagination, which produces images). To clarify the distinction between image and schema, Kant uses the tried-and-true example of the triangle: the empirical image of the triangle (i.e., any actual triangle) can never attain the perfection of the concept of the triangle. The latter is the schema, which "signifies a rule of the synthesis of the imagination with regard to pure shapes in space" (Critique of Pure Reason, A141/B181). The schemata that correspond to the table of categories are a bit different: "number" (which represents the addition of one thing to another) is the schema for magnitude, "sensation in general"(which represents a being in time) for "reality," and so on. They all involve "translating" the pure concepts into something, conditioned by a time-relation, that's closer to particular experience, more "ready for it."
So we have a structure, it seems, of Pure Concept (atemporal)--Pure Schema (formally temporal)--Image (in experienced time). This is probably enough explanation for the moment. If it seems obscure, that's due not only to my quick attempt at summary, but to the fact that Kant himself doesn't follow through in his discussion of the Schematism. The production of schema is a mysterious potential hidden deep in the human soul, he says. It's this ambiguity, as part of a greater ambiguity in the first Critique regarding the meaning and status of "imagination," that Heidegger is investigating in his 1929 book.
What interests me about the Heidegger paragraph is already hidden in that mess I just made. Here Heidi treats the schema as part of the essence of the concept (whereas, in Kant, it seems to be added on as part of the analytic setting-out of elements that characterizes his method). "Conceptualization" is not a relation to a concept as a thinglike abstraction, to be employed by recombining it with other concepts and subsuming particulars under them. Concepts are ways of being directed in relation to things. "Rules" is a very Lutheran, Kantian term, too rigid to characterize conceptual experience. A concept is the "motor" for a way of encountering something, a set of ways of moving towards it, looking at it, relating to it, a range of speed--the concept isn't a frame, but a way of framing--always in the "verb" sense rather than the "noun." I haven't found a way to put this that's unmetaphorical yet, and Heidi hasn't either, but his point, that the concept (as "regulation," which I read as a process-word rather than a thing-word) comes into view only in the gesture of looking away from it, captures this distinction rather well. To use another Heideggerean example: if you see a sign with an arrow pointing to the right, and you just stare at the sign, you're not having an experience of sign-ificance. In order to see the sign for what it is, you look in the direction the arrow points, or turn right, or determine that that's exactly where you don't want to go, and pick a contrary direction. The concept only comes into "view" while you're "conceptualizing," moving in that concept's particular dance, in its field of possible steps and relations.
(Note also Heidegger's use of "as"--one of the most important terms in his philosophy--I think laughing at that fact is necessary in the process of finding it interesting. "Unity" (thing) "can and must be as unifying" (process). Again, the sense that essence isn't stable, thinglike, but rather in motion.)
More on all this later.