20 Questions with Kant and Hume
Class: Philosophy 131, Professor Janiak
Kant defines an analytic judgment as a judgment which follows deductively from the premises of the argument. The most famous of these is “(1) Socrates is a man. (2) All men are mortal. (3) Therefore, Socrates is mortal.” If one accepts premises (1) and (2), it is logically impossible to deny conclusion (3). According to Kant, “gold is a yellow metal” is just such an analytic judgment. The definition of gold is that it is a yellow metal: if it were not yellow, or if it were not metal, it would no longer be gold. Therefore, the statement “Gold is a yellow metal” is logically necessary, and therefore, it is an analytic judgment. Indeed, any statement which simply restates the necessary parts of the definition, such as “Gold is a yellow metal,” is an analytic judgment.
The principle of contradiction states that no object X can be both A and not A. Kant considers the satisfaction of this principle sufficient proof of a judgment’s analyticity, and for good reason: the principle of analyticity is merely a more sophisticated version of the principle of contradiction. Given our definition of gold, “gold is a yellow metal” is the logical equivalent of “A is A.” To deny this statement is to say “A is not A,” which would violate the principle of contradiction. Hence, to deny an analytic judgment is to accept that “A is not A” somewhere within your argument. The principles of contradiction and analyticity are always satisfied or denied in the same cases, so they are one and the same.
According to Kant, an a priori judgment is one that can be made independent of experience, and an a posteriori judgment is one which is dependent on experience. Though gold, yellow, and metal are all objects/properties which we encountered through experience, the statement “Gold is a yellow metal” is an a priori judgment. Again, Kant stresses that “yellow metal” is contained within the definition of gold. So, to understand that the statement “gold is a yellow metal” is true, I do not need to have encountered gold, yellow, or metal. Science students make a priori judgments like this frequently, supplying definitions on tests to microorganisms which they have never seen. Therefore, empirical concepts can also be used for a priori judgments so long as these judgments satisfy the contradiction principle.
If <c> is an atomic concept, one cannot use it to express multi-variable analytic judgments like “gold is a yellow metal,” but it is still possible to use <c> for analytic judgments. It is as simple as this: “<c> is <c>.” This statement fulfills the contradiction principle. Since one can always say “<c> is <c>,” one can construct analytic judgments from any possible concept. Any empirical concept can support analytical judgments, as well, and it can also support synthetic (fact-based) judgments.
Kant thinks that some concepts, such as <cause>, are both analytic and a priori. Because of the way our minds are wired, says he, it would be impossible for us not to imagine them. David Hume disagrees with him and goes to some length to prove that <cause> is not a priori.
To understand this position, we must first explain Hume’s theory of ideas. To him, people experience only two types of mental events: impressions and ideas. Impressions are physical and mental events which we experience immediately, before reflection, such as sensory phenomena and emotions. Some examples of impressions are the feeling of anger, the sound of music, and the sight of the color blue. Ideas are split into two categories: simple ideas, which are mere mental copies of impressions, and complex ideas, which are combinations of simple ideas. Some simple ideas are a person’s visceral memories of the anger he once felt, the sound of music, or the color blue. According to Hume, we can easily distinguish between ideas and impressions based on their liveliness. The mind cannot create simple ideas independently; they must correspond to an antecedent experience (this is the copy principle.) Hume makes only one exception to this rule: the Missing Shade of Blue example, in which a person creates a simple idea of a shade of blue while he is viewing all the other shades of blue in a progression.
Complex ideas are reflections upon or combinations of simple ideas. All complete sentences (“Gold is a yellow metal;” “that was cold;” “this is an A paper”) are complex ideas, and so are all imaginary things (dwarves, unicorns) and abstract thoughts (justice, purity). Complex ideas do not include nonsensical formulations such as round squares. Any such idea which does not or which cannot have an antecedent in the real world is nonsense.
Now, we can consider Hume’s attack on an a priori conception of <cause>. The philosopher thinks it is quite unlikely that we could deduce the concept of cause and effect independent of experience. He proposes this thought experiment: I have suddenly come into the world as a thinking thing, and I see two billiard balls sitting on a table. How do I know that the balls will sit still if nothing touches them? How do I know they won’t disappear without reason or transform into chickens? How do I know that if Ball 1 strikes Ball 2, Ball 2 will move? These sorts of random changes only seem ridiculous to us because we have never experienced them, but they are not logically impossible. We believe in cause and effect only because it has worked every single time before. Therefore, <cause> cannot be a priori like “Gold in a yellow metal” was. It is a posteriori, or dependent on our experiences.
Where, then, does <cause> fit into Hume’s epistemological framework? It is clearly not an impression since it is neither a physical sensation nor an emotion. Because it is not an impression, it cannot be a simple idea, either. It does not satisfy the Missing Shade of Blue exception because it is a unique idea which does not have multiple shades of meaning which are visible at once, like <blue> does. Therefore, <cause> must be a complex idea. We constructed this idea to explain a large number of our experiences. The memories from which we abstract cause and effect are themselves complex ideas. <Cause> is a simple thing to explain, which is why Kant considered it a basic concept, but it is truly the sum of a myriad of events and concepts.
For this reason, any general representation is a complex idea in Hume’s epistemology. I never have impressions of <horse> or <justice> as universals. Rather, I create these general representations to explain a wide variety of particular impressions and simple ideas I have had. To create <horse>, I remember all the horses I have seen before, consider what they have in common, and thus create a class to which they all belong. (Otherwise, I read about the <horse> class someone else has created, and upon reflection, I form the same complex idea.) To create <justice>, I consider many good judgments and then determine what they have in common: justice. So, Hume’s complex ideas and Kant’s concepts perform the same epistemological function.