As part of our Semester Three seminars, this week we read P. Guyer “Kant’s Conception of Empirical Laws”.
Yesterday we kicked off our joint Cambridge-Edinburgh-St Andrews Kant reading group. Here, in the “north”, a lively and highly motivated delegation (!) from St Andrews joined forces with us in Edinburgh, while our colleagues in Cambridge run their reading group in parallel. We discussed Paul Guyer’s analysis of how Kant’s concept of an empirical law evolved over time, from the Critique of Pure Reason to the Critique of Judgment. Unsurprisingly perhaps, most of our discussion centred on the role of systematic unity as “methodologically necessary to discover empirical laws”. We wondered whether there is more to systematicity than a mere logical principle (remnant perhaps of Aristotelian logic) for the classification of entities according to homogeneity, variety and continuity, and the relation between natural kinds and laws that may result from this reading of systematicity.
We also discussed the extent to which systematicity, and the more central role it plays in the First Introduction to the CJ (following Guyer), is more directly related to the pressing issue of providing an order to the variety of chemical and biological phenomena (rather than to physical objects), and the shift of focus from physics to chemistry (and life sciences) that seems concomitant with the new role of systematic unity in the late Kant.
Other related issues that we touched upon, were the following:
1. are laws of nature necessary or contingent, on this reading of Kant? Laws seem to enjoy a degree of necessity but only in the sense that “it is a necessary condition of the possibility of experience that we find some particular set of determinate empirical laws for the perceptions we actually encounter” (Guyer, p. 237)
2. can systematicity be “brought to nature a priori” without being “imposed upon it”? (Guyer, p. 241)
3. given the open-ended nature of systematicity, could there be an alternative, but equally viable, ‘system’ to be brought to nature?
In Guyer’s words, while Kant’s statements of the problems in knowledge of empirical laws are clear and uncomplicated, his solutions to these problems are neither.
Thus, we are very keen to hear from our Cambridge fellows and anyone else from the Kantian world, who may be following this blog, comments and suggestions on these matters!
I would be interested to hear what people said in connection with (1) and (2) above. I also have some questions relating to these points, which are as follows:
Ad (1): The Cambridge group spent a bit of time trying to determine what the different kinds of necessity are that Guyer distinguishes (but we ran out of time before much progress was made). One question here is whether he is right to infer (in the passage on p.237 from which you quote above) from ‘p is not a condition of the possibility of experience’ to ‘p is not necessarily true of objects (in the sense in which conditions of the possibility of experience are)’. It seems to me that this inference is not licensed by what Kant actually says in the passages Guyer quotes (and the surrounding text). Kant seems to say that we cannot *know* that particular laws are necessary (because to know that, we would have to have a priori knowledge of them; but we don’t). This is compatible with their *being* necessary in just the same sense in which the Pure Principles are necessary.
Ad (2): This issue goes straight to the question of what it is for a principle to be transcendental and yet merely subjective. That’s obviously a difficult question, and I won’t try to address it here. But it’s not clear to me why the question should be framed in the terms Guyer proposes. He implies that a transcendental principle is objective iff it is “imposed” on nature by us. Is that the best way of understanding what Kant means by saying that e.g. the categories are valid of objects of experiences? At least on the face of it, it’s hard to see how something that is a subjective imposition can be objective. On the contrary, it might seem that if Kant manages only to show that the categories are imposed then he has failed to demonstrate their objective validity.
This of course raises big questions about how to read the Transcendental Deduction and what to make of Transcendental Idealism. But I think it’s safe to say that Guyer’s way of framing the issue is not the only possible one. So my point is that there is a choice here, and that we shouldn’t go along with Guyer’s framing if we think there are good reasons not to.
Picking up on (1), there seems to be a general difficulty in understanding in what sense laws of nature are necessary. It can’t be metaphysical or absolute necessity: there could be worlds with different laws of nature than our own. Moreover, saying that nomological necessity is what is necessary according to the laws of nature is uninformative.
I don’t think that universals exist, but one could hold that a law of nature involves a necessitating relationship amongst universals, and that (i) what laws we have depends on which universals are instantiated; and (ii) what which universals are instantiated in an given world is a contingent matter. On the surface this might appear to provide some sense in which nomological necessity isn’t as strong as metaphysical necessity, but is still a (weaker) kind of necessity. I think this appearance is probably illusory however. If it is wholly continent which universals get instantiated, then it is wholly contingent which laws of nature we have, and hence wholly contingent that events happen in accordance with laws. There is here a sense in which events are impelled to take place because of laws of nature, but this isn’t properly understood in terms of *necessity* as such.
Even a theological gloss doesn’t fare any better. Say that a regularity is a law because God has decreed that the regularity should obtain. Either God could have chosen otherwise or God could not have chosen otherwise. If the latter, then it must be *metaphysically* necessary that the laws of nature obtain, but this kind of necessity is too strong. If the former, then the laws do not appear to be necessary in any way.
Guyer’s reading of the CJ throws up one gloss on nomological necessity that might be more promising:
‘[E]xplanatory relationships … seem to be what is intended to give us some sense of the necessary truth of particular empirical laws at least relative to the system to which they belong. Having its position determined by its logical relationship to laws at other levels of a system of empirical laws, by which Kant evidently means a system in which causal or explanatory laws are subsumed under one another as well as classificatory concepts being ordered into genera and species, is as close as an empirical law can come to necessity.’ [Guyer 1990: 240]
This isn’t nomological *necessity* as such, but it may account for the “feel” of necessity that laws of nature can have. Modal claims differ from non-modal claims in that they licence different kinds of inference: in particular, the former licence *counterfactual* inferences. So too with nomological claims. That all public swimming pools are filled with water is an accidental regularity: it does not follow from this proposition that if we were to attempt to fill a swimming pool with Coca Cola, we would fail. That water boils at 100 degrees Celsius is a lawlike regularity. It does follow from this that if we heated a swimming pool to 100 degrees Celsius, the water inside it would boil.
Tying the “necessity” of laws to explanatory relationships, as Guyer reads Kant as doing, captures what they have in common with genuine claims of necessity: both licence counterfactual inferences; both share a particular kind of inferential role. Further, it does so without requiring that laws of nature possess their own special kind of nomological necessity.
It is not so clear to me that Kant cannot be committed to any kind of metaphysical necessity, since there are passages (e.g., 4:467) that can be read that way. (Whether he is committed to such a necessity is a separate question.)
Also, I wonder whether there couldn’t be different kinds of necessity, with one kind (of natural necessity) obtaining for appearances and a different kind obtaining for things in themselves, esp. since one might read Kant as maintaining that the former depends on the latter. (It is in this sense that Kant says that the laws of the sensible world have their ground in the latter and its laws, 4:453.) If so, it may be that Kant’s position does not map neatly onto contemporary accounts (which would say that at least the “necessity” of appearances and their laws would in fact be contingent, since dependent on things in themselves, which might have contingent aspects). In some ways, this reminds me of Leibniz’s distinction between hypothetical and absolute necessity. Some things are absolutely necessary (regardless of what God wills) and other things are hypothetically necessary, that is, necessary given God’s free decrees about certain fundamental issues. I do not think that hypothetical necessity is a vacuous notion in virtue of being hypothetical. Obviously, Kant’s distinction is different, but it shares a certain kind of hypothetical structure — given things in themselves and their natures and given that we have space and time as forms of intuition and categories as forms of thought (and various other things, such as initial conditions), it is necessary that certain phenomena occur.
I think the suggestion to bring in the distinction between absolute and hypothetical necessity is a good one (whether or not Kant wants to trace it back to the distinction between god’s intellect and will in the way Leibniz does). Kant doesn’t think any appearances exist necessarily. Nothing in the realm of appearances is necessary in the sense that it couldn’t have been otherwise absolutely speaking. But given the existence of appearances in time, it is necessary that certain causal relations obtain. It’s not clear to me how far this gets us, but it may be one option for blocking inferences of the form James gives above: from ‘it is contingent which universals exist’ to ‘it is contingent that events happen in accordance with laws’.
Thanks for the comments, James, Thomas, and Eric.
Back to the necessity of the empirical laws as arising from their place within a system of empirical laws, and in virtue of the explanatory relations they enjoy within that system.
Is the intuition here that the necessity of our empirical laws is only a necessity arising from their explanatory role within the specific system (but obviously we could have come up with a different system, and hence which empirical laws we count as necessary is ultimately a contingent matter)? (I worry here that Kant would not like to make systematicity itself contingent as a methodological criterion)
Or, is it instead the intuition here that the necessity of our empirical laws arises from nature itself, given its systematicity (a priori projected and not imposed)?
Eric, am I right in reading your passage about “hypothetical necessity” along the latter line of reasoning? and if so, how do things in themselves enter in this discussion?
If we think about the cognition that all humans are mortal and think that that enjoys some kind of necessity, one might think that the necessity of that cognition follows from the more general cognitions that all mammals are mortal and that all humans are mammals. (Perhaps the chain goes up higher to “All animals are mortal” or even “All living organisms/things are mortal”.) At some point, however, I take it that there must be some nature (e.g., animality or living organism) that has mortal as an essential feature and which the appropriate cognition is expressing. And the necessity of that claim trickles down, so to speak, to the claim that all humans are mortal (as well as to Socrates being mortal). Now insofar as animality (or whatever nature is appropriate) pertains to appearances and appearances depend on things in themselves (and the natures of things in themselves), then it may be the case that the necessity of the nature expressed at the level of appearances depends on the necessity of the nature of the relevant thing(s) in itself(themselves). Of course, we cannot know the noumenal natures, but they can nonetheless “ground” the necessity of what appears to us.
(Also, I in no way deny that the understanding is responsible for the necessity that is involved in nature in the formal sense.)
Not sure if that addresses your questions, but it’s a first attempt. (For what it’s worth, one can get a slightly better sense of how this view works from my discussion of freedom in chapter 5 and parts of chapter 6 of Kant and the Metaphysics of Causality.)
Many thanks for the further explanation, Eric. This is helpful in clarifying how the necessity may ‘trickle down’ as you say from within a system of nature, or at least a system of classification we may employ. I guess I am still trying to get my head around the idea that appearances depend on things in themselves and their necessity may be downstream, so to speak, from the necessity of the latter. But I sense that we may be approaching another terrain here (about freedom et al.) going beyond remit of natural science — I will certainly get back to your book!
Fill in your details below or click an icon to log in:
You are commenting using your WordPress.com account. ( Log Out / Change )
You are commenting using your Twitter account. ( Log Out / Change )
You are commenting using your Facebook account. ( Log Out / Change )
You are commenting using your Google+ account. ( Log Out / Change )
Connecting to %s
Notify me of new comments via email.
Notify me of new posts via email.
Enter your email address to follow this blog and receive notifications of new posts by email.
Join 110 other followers