TCD by Pendlebury

The Transcendental Deduction of the categories is supposed to be the centerpiece of CPR, but is extraordinarily difficult to make sense of it either as a whole or in detail. There are many different interpretations and reconstructions of it in the secondary literature, and it is widely thought that as an argument it fails. One commentator, Johnathan Bennett, describes it as “a botch.” But there is also widespread agreement that it is full of interesting philosophical ideas and that it contains important insights that have made a huge difference to the philosophical landscape. Our main goal should be to identify positive, useful, and constructive ideas without denying the shortcomings of Kant's arguments.
Kant's question in the Transcendental Deduction is “By what right do we apply the categories (in experience)?” In the Principles, which come later in CPR, Kant takes it for granted that we are entitled to apply the categories in experience and asks which fundamental synthetic a priori principles result from this application. For example, in the Second Analogy he assumes that we can apply the concept CAUSE to appearances and that we must assume that every objective event has a cause in order to make sense of succession in objective time. But whenever justifies the claim that we are entitled to apply the category CAUSE to appearances. This job is supposed to have been done in the Transcendental Deduction.
Why is there a problem? Think about it in relation to the category CAUSE. If we abstract from matters of time, this is the concept of something necessitating something else or, for short, the concept of a necessary connection. But, as Hume argued and Kant grants, appearances are never given as necessarily connected. Separate appearances just happen to follow one another in subjective time, and no matter how deeply one examines what is given in them, one won't find a necessary connection. So we cannot justify applying the concept empirically, or a posteriori What, then, could possibly justify us in applying it?
The Transcendental Deduction is supposed to provide an answer not only with respect to causation, but with respect to all the categories, which Kant lumps together as if they were one, describing them collectively as “concepts of an object in general” (B 128)/ However, the categories that Kant mentions most frequently are those of SUBSTANCE and CAUSE. And for good reason, because these are the ones that are most puzzling,. We've seen why in the case of CAUSE. In the case of SUBSATACNCE the problem is that substance persists through all empirically possible changes in appearances, so it's not clear how the concept SUBASTANCE can be applied to appearances. For purposes of this class we will, wherever possible, focus on these two categories.
A crucial part of Kant's answer to the question of why we are entitle to apply the categories in experience is “Because we must in order for XSOMETHING to be possible”--”Because it's a condition of the possibility of that something.” But what something? The official answer is “experience,”and this clearly means experience in the thick sense. However, at different points Kant seems to be focused on different t features of the thick experience: sometimes just judgment, sometimes knowledge, sometimes self hood and unity of the mind. WE can, nonetheless, say that the Transcendental Deduction aims to show we are entitled to apply the categories to appearances because the categories express “conditions” of the possibility of experience” that appearances given by intuition satisfy.
The Transcendental Deduction also has several subsidiary goals, most notably to contribute towards the following:

an account of objecthood and objectivity;
an answer to Humean skepticism on causation and substances;
an answer to the Cartesian skeptic;
an account of how judgment is possible;
an account of the self and self knowledge;
an account of the possibility of some synthetic a priori knowledge

There's a lot to get out of it if you provide the effort and goodwill.

(Jake: also points on conformity versus individuality; and what he leaves to be covered and the antimonies and schematism, these arguments in the CD argue for the set up of these laster sections which are important later)

13: On the principles of a transcendental deduction in general (B 116-24)

This section is especially important for understanding the problem. Kant opens the section by saying that he is using the word “deduction” in the juridical sense, according to which a deduction is meant to establish a right rather than a fact.
Kant thins that empirical concepts don't require an explicit deduction because they answer to something that is given in intuition (B 116_). This works quite well for simple empirical concepts like RED and ROUND, but it is not obvious that it applies to a much richer empirical concept, like the concept DOG. The concept DOG goes beyond anything that can be present in a single intuition even a series of intuitions, because no series of intuitions has the same degree of permanence as any dog. The reason is that any dog is necessarily an individual substance because the concept DOG analytically includes the concept SUBSTANCE. Thus “Any dog is a substance” is analytic. But, of course, “Anything dog like is a substance” is synthetic, and in fact false. Thus if categories require a deduction, so do rich empirical concepts like DOG. On the other hand (and in Kant's defense), if we could give a deduction of the categories, we could also handle DOG and other similarly complex empirical concepts. So we can reasonably set these concepts aside at this point and pretend that all empirical concepts are like RED and ROROUND insofar as they apply to something that is given in intuition, and concentrate on the question of why Kant thinks that categories require a deduction (focusing heavily on the categories of SUBSTANCE and CAUSE).
One potential problem with respect to the categories is that there are some concepts, such as FORTUNE and FATE, that Kant thinks we have no right to apply to anything (B 117). Call such concept “exculpatory.” Given that the categories don't correspond to anything present in intuition (or a series of intuitions) it could be that they, too, are usurpation. Kant wants to establish that they are not.

More importantly, categories are a priori concepts that are always employed a priori, so an empirical deduction of them is impossible B 117). Of course we could conduct an empirical investigation of when and how the categories show up in the development of human psychology (B 118). , but this would establish only facts about their history, not a right to apply them (B 119)./ Today one could make this point by saying that a developmental psychology of the categories is not substitute for a critical epistemology.
The problem of establishing a right to use the categories is much greater than that of establishing a right to use the concepts of SPACE and TIME, because, even though these concepts have a priori roots, they are directly exemplified in intuition, i.e., in the form of intuition (B 120)/ However, the categories can be applied to things that are not given in intuition (consider, e.g., “God s a substance”). And they are not specifically connect wed with intuition, as space and time are. See B 120 and (especially) B122 (“The categories of the understanding...functions of the understanding”). Objects, i.e., appearances, are given in intuition regardless of whether we apply the categories to them, and “appearances could after all be so constituted” that the categories do not apply to them (B 123). Part of the burden of the Transcendental Deduction is to establish that they are so constituted that the categories do apply to themselves.
It is crucially important that Kant is concerned with the question of our right to apply the categories in experience, i.e., to appearances (B122). He thinks that we cannot help applying them beyond all possible experience, but that we have no right to apply them beyond experience – that no epistemic justification for such applications is possible.