What Kant grapples with in his moral philosophy, Adorno argues, is the tension between freedom and necessity as it pertains to human action. At the most obvious level the tension presents itself as the impossibility of moral human action in a world governed by causal laws. Insofar as we are empirical beings constrained by the doctrine of cause and effect, we cannot be held accountable for our actions, for we have no more agency in conducting ourselves than the billiard ball bouncing off the pool cue. The iron necessity of causality in this case subjugates us to the dictates of nature. We do not choose natural law, and yet we must abide by it. Hence we are heteronomous – literally, governed by the laws of another – and so unfree.

At this juncture one may very well conclude that genuine freedom involves action according to no rule, which springs from some unnatural font of spontaneity (e.g. an immaterial soul). Indeterminism fails to provide a substantive grounding for free action, however, for the same reasons that nature’s determinism does. Just imagine somebody who acted according to no principles whatsoever – whose action sprung mysteriously from some unknown source. Their choice in deciding to do X over Y would be unintelligible, more like the accidental outcome of a die roll than of the sort of deliberation characteristic of decisionmaking. So the way out from determinism to freedom can’t just involve the negation of determinism.

Instead, Kant’s answer is that moral choice must involve the right sort of determination: determination by reason. To act morally is to act in accordance with reason, and to act in accordance with reason means obeying strictly its laws. What differentiates these laws from those of nature is that the laws of reason are self-imposed – they are autonomous, auto-nomos, self-law. Adorno points out the tension in this approach: to secure freedom, Kant must establish a necessity in the moral sphere akin to the causal necessity present in the empirical sphere. Adorno puts it like this:

“…reason generally makes its appearance with the claim of deductive necessity, with the claim that everything it implies follows in accordance with the propositions of logic. And this element of necessity already presupposes an affinity… with the causality that is supposed to hold sway in the realm of empirical phenomena… the whole of Kant’s moral philosophy is tied to the concept of autonomy which is regarded as the realm where freedom and necessity meet. What this means is that the moral laws are indeed the laws of freedom – because as a rational being I give them to myself without making myself dependent on any external factor. At the same time, however, they have the character of laws because rational action and rational deduction cannot be understood except as acting and thinking in conformity with laws and rules.” (Problems of Moral Philosophy, Lecture 8, pg.80)

A problem immediately arises: in what sense can we act autonomously in accord with reason, if we are empirical beings who inhabit the determined world of nature? Adorno claims that “here Kant falls into a trap of his own making” in articulating a solution to this problem (81). Kant argues variously in the Groundwork of the Metaphysics of Morals and the Critique of Practical Reason that “as empirical beings we experience the obligation to perform certain actions, or to leave them undone” (Adorno 81). Proceeding from the fact of this experience, Kant asks as to the conditions for its possibility: what makes possible the experience of conscience, of the sense that one should and should not engage in certain actions? Such transcendental critique comes unsurprisingly as part and parcel with the rest of the Kantian system, but in this case Adorno points out its potentially damning implications:

“…if [Kant] desires to exclude every empirical element from his foundation of moral philosophy – and that is his aim – he cannot then appeal to the empirical existence of the so-called moral compulsion in man himself because this compulsion is itself an empirical fact… in short, the unity of moral obligation and reason that Kant insists on is not altogether unproblematic if we reflect a little more deeply on this obligation; indeed it becomes highly dubious.” (Lecture 8, pg 82)

The quite obvious Kantian response in this case involves an injunction against mistaking the empirical conditions by which we come to know a phenomenon for some essential feature of that phenomenon. In other words, such a criticism mistakes the process by which we contingently come to experience our moral duty for the characteristic qualities of that moral duty itself. It is a similar mistake to one who takes the truths of mathematics to be contingent upon empirical observations about the world, just because one has learned them through such observations.

But in this case such a response does not save Kant, because it introduces a rupture within the human being which is the subject of morality, splitting asunder the experience of morality from its essential nature. There are three possible ways we can conceive the relationship between the empirical experience of moral duty  (conscience) and the “formal, abstract shape of the moral law” for which that experience serves as evidence.

First option: Conscience is coextensive with the moral law. This is a highly repugnant conclusion. To accept this is, in essence, to deny that there is any thing as a moral law in the first place. For it identifies the empirical constraints on moral thought (the psychological experience of moral duty) with the moral duty itself. One might imagine this position as involving an acceptance that “whatever one feels is moral, is moral; whatever one feels is immoral, is immoral.” An unacceptable conclusion, in short.

Second option: Conscience is totally separate from the moral law (though it grounds the moral law). This is a more palatable conclusion – presumably the one Kant took himself to be endorsing, in the last instance – but one which Adorno argues cannot be the case. For suppose it were true: that there was no substantive relationship between conscience and the moral law. Or, to put it more precisely, that there was no relationship between the motivational import provided by conscience and the demands lain upon it by the moral law. In this case, any case of moral behavior would be a merely accidental one of the sort Kant routinely decries in the Groundwork, e.g. of the shopkeeper who treats customers fairly not because it is moral, but because it brings them the greatest profit. Except in this the explanation for all moral behavior would be that, e.g. the shopkeeper treats customers fairly only because of some empirical-psychological motivation stemming from their conscience, and not because the moral law commands them to do so. In this case the pure abstract formality of the moral law might be preserved, but at the expense of its possible enactment.

Third option: Conscience is related somehow to the moral law. This is the only option which remains, but that “somehow” is a problem which – Adorno argues – we cannot solve using only the tools given to us by Kant’s system.

Adorno concludes his exceedingly fascinating eighth lecture with a concise statement of Kant’s, distinguishing the determinism of nature from that of the moral law:

“‘Reason therefore’, he continues, ‘provides laws which are imperatives, that is, objective laws of freedom which tell us what ought to happen – although perhaps it never does happen,’ – you see here [Kant’s] indifference towards effects – ‘therein differing from laws of nature, which relate only to that which happens. These laws [of reason] are therefore to be entitled practical laws.'” (Lecture 8, pg. 88)