Five Hard Pieces: Kant summarized and prepped for evaluation

One way to think about Kant's work in The Critique of Pure Reason is to break it down into five steps:

1. Distinguishing between the analytic/synthetic distinction, on the one hand, and the a priori/a posteriori distinction on the other. 

Analytic statements are about concepts.  They claim to offer an analysis of a given concept and, on that basis, to tell us something         about its content.  For instance, we might say that being unmarried is part of the concept of being a bachelor, and being an animal is part of the concept of being a dog.

Synthetic statements combine concepts to make a claim about the world.  For instance: many bachelors have dogs.  That is a synthetic claim that might turn out to be true or false.  To determine its truth, we would have to investigate the world, not just this or that concept.

A priori statements are those whose truth can be judged without investigating the world but simply by thinking.  For instance, 7 + 5 = 12 and that a straight line is the shortest distance between two points seem to be a priori claims.

A posteriori statements are those whose truth can be judged only by some investigation of the world, only by means of experience (one's own or someone else's).  It is known a posteriori that it was in 1492 that Columbus sailed the ocean blue.

Before Kant most philosophers treated 'analytic' and 'a priori' as if they meant the same thing.  Likewise for 'synthetic' and 'a posteriori.'  But whether a statement is about a concept or the world, and whether we need to refer to experience in order to know whether it is true or not, do not seem to be the same thing.  In that sense, Kant seems to be right.

On the other hand, the distinction might be trivial if all analytic statements turn out to be a priori and all synthetic ones a posteriori.  And if it is not trivial, the question is raised as to how we could possibly know about the world a priori, or about concepts only through experience.

It has also been denied that there really is an analytic/synthetic distinction.  If statements about concepts are really about the use of words, then they are statements about the world.  So while not all synthetic statements are analytic, all analytic ones are synthetic.  In that case, they would all seem to be a posteriori too, since we know about the uses of words through experience. 

In defense of the distinctions, though, the fact that not all synthetic statements are analytic gives us a distinction that might prove useful or even important.  Secondly, it seems worth noting the different methods of the philosopher/theorist, on the one hand, and the experimental scientist on the other.  At times the line between them might be blurred (scientists use theories and thought experiments, for instance), but the existence of dusk does not mean that there is no difference between day and night.  (Indeed, they are as different as night and day.)

So Kant seems to be right, even if the distinctions he is concerned with sometimes seem to collapse.

[The most important literature on this issue centers around the work of Willard van Orman Quine.]

2. Claiming that there are synthetic a priori truths.     

Kant importantly argues that his distinctions are not at all trivial.  He denies that we can know about concepts only a posteriori, but does believe that there are synthetic a priori truths.  For instance, nothing about the concept of straightness, he says, involves the idea of shortness of distance.  Mathematics is conducted a priori, just by thinking, sometimes on paper and sometimes in the head, and yet it can be applied to the world in physics and engineering.  So we can and do know about the world just by thinking.  Strange but true.

This is an interesting claim.  In some cases, however, we might doubt its truth.  For instance, doesn't the idea of directness connect straightness with shortness of distance?  And does mathematics really tell us how many apples I will have if I start with two and then add five?  Or does it rather tell me what number I will call the number of apples I then have?  What if somehow I end up with eight, or six?  Does mathematics tell me that this is impossible?  Or is it common sense?  And what exactly is that?  Isn't it more that I use the statement 2 + 5 = 7 as a rule, at least sometimes, to guide my behavior?  So if I add two apples to five apples and seem to get six then I treat this as a wrong result, insist that someone must be playing a trick on me, etc.? 

What about the idea that every event must have a cause, that nothing ever happens without some reason or explanation?  Again, this seems more like a guiding assumption than a fact that we know to be true.

These are difficult questions though.  It looks as though Kant might be wrong, but we have not shown that yet.

[The questions raised above are suggested by the work of Ludwig Wittgenstein.]

3. The theory of the categories and of time and space as forms of intuition to explain the existence of synthetic a priori knowledge.  

Kant argues that we know about the world just by thinking (a priori) because we discern the limits of our comprehension in this way.  And what we cannot comprehend we cannot experience (except, perhaps, as something incomprehensible and ineffable).  To experience something is not only for it to happen.  Trees fall in the woods all the time without anyone experiencing them fall.  To experience is to witness, which involves more than simply being there.  You must have senses that work, and you must be conscious.  You have to make a certain amount of sense of what is happening around you.  That involves a kind of interpretation or processing of the data (literally: what is given.)

All our experience, Kant argues, has to involve objects in space interacting causally in time.  Intuition is something like mental representation.  What I sense gets represented in my mind, as do things I only imagine.  Events real and imaginary, though, must still take place in time and space.  That is, I cannot even imagine events as being 'out of' time or space.  In imagining them, I cannot but represent them to myself as being in time and space.  Kant calls time and space the forms of intuition. 

We categorize data too, he believes.  For instance, our experience involves objects and causation.  There are twelve such categories in all, and Kant has been criticized quite a bit for the supposedly weak argument he gives that there are precisely these categories and no others.

Is he right?  Possibly, but it doesn't look like it.  The very idea of synthetic a priori knowledge is dubious to begin with, and explaining it in this way makes the idea still less appealing.  But, of course, that does not make it false.

4. The distinction between phenomena and noumena.      

If space and time are only forms of intuition then it seems that they cannot exist as anything more than this.  I have not given a proof of this here, but that is Kant's position.  When we talk or think about objects causally interacting in space and time we are talking about our experience.  Kant uses the word 'phenomena' in this connection.  These are perfectly real.  Experience is not illusion (although, of course, it is possible to experience an illusion--the point is that this is the exception, not the rule).  But they are appearances, things in our minds.

Appearances can only be of something or other.  So it seems that something or other must exist, as it were behind the appearances that we experience.  Kant's word for this something or other is noumena (things as they are in themselves). 

Also, although we cannot conceive of anything to which the categories and forms of intuition do not apply, we can conceive of our being limited in this respect.  That is, we can imagine a being in relation to which we are limited.  Think of a sixth sense.  We cannot imagine what this would be, perhaps, but we can imagine that there are creatures with senses other than the ones we have.  Similarly, although we cannot imagine anything but phenomena, we can imagine that there is something more, something that could be known about in ways not available to us.  This barely conceivable something more is noumena.

This is a very problematic idea.  One of Kant's main points seems to be the admirable one of curbing our pride.  He thinks we should be humble about what we know and what we can ever know.  At times he insists that the idea of noumena is purely negative, purely a way to help us see that all we can imagine (phenomena) need not be all that there is.

However, that is a problematic idea in itself.  Because anything else, any something else, must surely be an object of some kind.  According to Kant, we cannot conceive it to be anything else.  But objects belong to the phenomenal realm.  It is not easy to conceive of something of which we cannot conceive, or to know what to make of talk about such a thing.

The problem is made worse by the fact that Kant sometimes talks as if we can have beliefs (though not knowledge) about specific noumenal things, such as God and our own immortal souls. 

[Some of the most important criticisms of Kant on this come from Arthur Schopenhauer and Wittgenstein again.]

5. The practical justification of faith.

Kant saw naturalism and fatalism as great evils.  Unjustified confidence that there is no God and that all is ruled by the laws that natural scientists discover leads, he thinks, to bad ethics ("social Darwinism" might be an example), sinful pride, and despair.  As Kant saw it, within the phenomenal realm, where every event has a cause, every movement of the body must be the result of countless prior causes, reaching far beyond individual control.  This deterministic view makes of each of us, in Wittgenstein's words, "a living horror."  We are like robots.  To Kant this seemed to destroy the idea of moral responsibility.

It is only possible to live as a decent human being if one has faith in free will, just reward and punishment in a future life (hence the immortality of the soul), and hence also God.  So, while we cannot know anything about the noumenal realm, we can be justified in having certain beliefs about it.

This is an interesting kind of argument, but it seems to be quite bad too.  If noumena or things in themselves are quite beyond our conception, as Kant says, then we cannot meaningfully believe in noumenal free will or anything of the sort.  The words will be quite empty for us.  Our 'faith' will then be quite empty too.

Kant's motives here might be admirable, but he seems to have been led to commit a kind of trick in order to fight the evils he identifies.  Assuming that he was sincere, then he was, apparently, confused, blinded to his own inconsistency by a refusal to back down from his moral opposition to naturalism, etc.  This itself might be regarded as a kind of moral failing.  It has implications for our assessment of steps 3, 4, and 5. 

Having said that, a bit of humility wouldn't be amiss in any attempt to pass judgment on Kant.