A key disagreement between Kant and Leibniz regards whether it is possible to know things as they are in themselves via the senses. Both agree (verbally at least) that knowledge comes through or requires the participation of the senses, but Kant takes this to imply humility (i.e. scepticism), whereas Leibniz does not. The difference here has to do with the nature of thought, specifically its receptivity and what this involves. Receptivity is the interactivity necessary for thought according to Kant. Now, whether it is accurate or misleading to portray Leibniz as rejecting interactivity is for me a vexed question which I assume here can be safely ignored, as there is a firmer point of disagreement to focus on (namely humility). Since humility follows from receptivity on Kant’s account, this means that anything I know through interacting with that thing is not something whose intrinsic properties I know. As it turns out, the move from receptivity to humility is mediated by Kant’s famous denial of the reducibility of relations, which, if Langton is to be believed, ultimately turns on a question-begging appeal to contingency on Kant’s part. Let’s see how all of this unfolds, following Langton’s discussion.
First of all we have Kant’s definition of a “true substance” as something capable of existing cut off entirely from external connection. (Note that I am not taking this to mean that it can be the only thing that exists.) I’ll call this possibility isolation. Thus:
(1) true substance ⇒ possible isolation
All and only those properties a substance has in this scenario are its intrinsic properties.
Next, according to Kant, a substance cannot undergo change unless it interacts with other substances, i.e. unless it is not isolated. Thus:
(2) change ⇒ interaction ⇒ not isolated
(Since time requires change for Kant, time also requires interaction between substances.)
The argument for (2) goes something like this. For a substance X:
- For every possible property, X has it or its negation (F or ¬F).
- There is always a reason why X has F or ¬F.
- Given isolation (i.e. a substance existing by itself), the reason why it has F or ¬F must be an intrinsic property.
- If we now suppose X to undergo change, then it follows that the reason for this change must be external to X.
We can think of (A) as saying that X is determinate. Premise (C) says that (A) applies in the scenario where X is isolated. And premise (B) basically expresses the PSR for properties of substance.
The basic idea behind the argument is then this: since there is a reason for why X is F (assuming it is F), if X changes to ¬F, the reason for this cannot come from inside X (since what is inside X supports F rather than ¬F. If this reasoning is sound then we have established (D).
We now come to the third step: For Kant it is an empirical fact that thought involves a changing succession of ideas. Given the above argument, it follows that thought requires interaction. That is what his thesis of receptivity amounts to. Thus:
(3) Thought ⇒ change ⇒ interaction ⇒ not isolated
Thus from a mere change in thought we are (supposedly) able to deduce the existence of bodies external to us. More than this Kant goes on to argue – given his commitment to the existence of true substances – that our necessarily receptive way of knowing prevents us from knowing things as they are in themselves.
According to Kant, since substances can exist without interacting with one another, the interactions – the relations between them – are superadded. In this way Kant denies that relational properties are reducible to the intrinsic properties of substances (at least to those distinct from God). In what way, exactly? Langton undertakes a lengthy exploration of this question, and I agree with her results: either Kant commits a non sequitur, or else he must be rescued by translating invalidity into lack of cogency (against Leibniz). Specifically, to rescue Kant’s argument for irreducibility Langton proposes that we treat causal powers as extrinsic properties of things, i.e. properties such that the same substance X could have those properties in one possible world but not another, although its intrinsic properties remain the same. In my mind this appeal to contingency simply replaces an invalid argument against Leibniz with one that cannot hope to be cogent. (Moreover, as Langton herself argues in a later publication, even if this premise is given, it does not straightforwardly follow that we are necessarily ignorant of things as they are in themselves, since that is to infer humility from contingency, and this inference can be challenged.)
- Rae Langton, Kantian Humility: Our Ignorance of Things in Themselves, (Oxford University Press: 1998).