Background Notes for “On Knowledge of General Principles” (Hume and Kant)
David Hume (1711-1776) – Regarded my many as “the most acute, if the most perplexing” of the British empiricists
Hume’s philosophy starts with the reviving of Leibnitz’s analytic/synthetic distinction. Hume calls this a distinction between “relations of ideas” (analytic) and “matters of fact” (synthetic).
Analytic propositions are expressed by sentences
1. whose negation leads to self-contradiction: If we try to deny that “All brothers are siblings” since the set of siblings contains “brothers”, saying that “All brothers are not siblings” is equivalent to saying “Some brothers are not brothers”
2. that are a priori (triangle has 3 sides; rainy day is a wet day; unmarried men are bachelors, as well as the example above)
3. that are true by definition and thus,
4. that are necessarily true
Synthetic propositions are expressed by sentences
1. whose negation does not lead to self-contradiction,
2. that are a posteriori
3. that are not true by definition and,
4. when they are true, they are not necessarily true (they can be false)
“The cat is on the mat” is an example of an a posteriori proposition
Hume admits that there are a priori (necessary) truths. But these truths are merely tautological: they are redundant, repetitive and merely verbal. These truths, according to Hume, provide no new information about the world, but simply information about the meaning of words.
For Hume, only synthetic (“matters of fact”) claims can describe reality correctly. These claims are necessarily a posteriori. Thus all true knowledge about the world must be based on observation.
Hume claims that there are essentially only three categories of analysis. Any proposition is either:
1. Analytic
2. Synthetic
3. Nonsense
Hume says: “When we run over libraries, persuaded of these principles, what havoc must we make? If we take in our hand any volume – of divinity or school metaphysics, for instance – let us ask, Does it contain any abstract reasoning concerning quantity of number [analytical truths]? No. Does it contain any experimental reasoning concerning matter of fact and existence [synthetic truths]? No. Commit it then to the flames, for it can contain nothing but sophistry and illusion.”
(Hume once lost a job as a librarian and this seems to hint as to a possible reason why…)
Here is the “Humean method” of philosophizing:
Picking any claim we wish to test, we ask a series of questions
1. Is the claim analytic? (If a negation of that claim is self-contradictory, the answer is yes) If YES, the claim is true, but philosophically trivial. If NO, then we ask the next question,
2. Is the claim synthetic? (Can the idea be traced back to sense data, or “impressions” as Hume calls them. Example: “The stone is heavy” can be traced back to sense data and it passes this test). Thus if the answer is YES, we have a meaningful claim. If the answer is NO, Hume says we are then dealing with ideas that are vacuous and nonsense.
Hume now turns his method to the idea used by philosophers, scientists and the rest of the ordinary, common sense folk, the idea of Causality.
Here is Hume’s own example
When one billiard ball strikes another billiard ball and the other ball moves, we, without much reflection say that the first ball caused the second ball to move.
Let’s apply Hume’s test to this situation and distil it into a sentence:
X causes Y (where X and Y are events; Ex. The event of billiard ball A striking billiard ball B and causing the event of billiard ball B moving)
1. Is the sentence “X causes Y” analytic?
- Clearly, denying that X causes Y does not result in a contradiction. We can easily imagine ball A striking ball B and ball B refusing to move. So the answer is NO.
2. Is the sentence “X causes Y” synthetic? (Can it be traced to sense data?)
- At first sight it seems that there is no problem with tracing the idea of “cause” to sense data. But Hume looked carefully and found a difficulty.
- Hume breaks down the concept of “Cause” into three components
- Priority
- Contiguity
- Necessary connection
Priority (the fact that event X precedes event Y) can easily be traced to sense data – we can see that it indeed does so.
Contiguity (the fact that “X touches Y”, that the two events form a series) can also be traced to sense data easily.
Necessary Connection … is a problem. No matter how many times Hume observed ball A hitting ball B and causing it to move, he could not observe a necessary connection, without which the concept of causality could not make sense.
Hume could not help but conclude that when we say that event A causes event B, we are merely reporting a psychological expectation that event A will be followed by event B in the future. The rational grounding for this expectation, however, cannot be found.
Hume did not conclude that causality does not exist in the world, but he discovered that a philosophical account of causality is not possible.
Hume’s discovery became known as the Problem of Induction.
Hume, following the program of empiricism to its rational, logical conclusion, strikes a powerful blow to our common sense understanding of the world. Rationally, there is no guarantee that the laws of nature will hold tomorrow, that there is God, Self, that a loaf of bread that nourishes us today won’t poison us tomorrow, etc.
But Hume understood that human endeavours extend far beyond rationality and philosophical justification and suggested once that “perhaps we should abandon philosophy and take to tending sheep instead.”
Immanuel Kant (1724-1804)
Hume had indeed struck a powerful blow to philosophy and rationality by simply following reason and logic to their logical conclusion.
Upon reading Hume’s “An Inquiry Concerning the Human Understanding”, Kant, considered by many one of the most powerful, subtle and original thinkers, declared that Hume “awakened him from his dogmatic slumber.” Hume’s sceptical arguments had to be defeated.
Kant’s “The Critique of Pure Reason” could be seen as an attempt to provide a synthesis of the best of Hume’s philosophy with whatever is left of rationalism after Hume’s attempt to explode it to bits.
The following is a very cursory review of Kant’s philosophy as it applies to response to Hume.
Kant accepts Hume’s analytic-synthetic distinction as the key philosophical tool of analysis. Here is how it breaks down:
Hume: All analytic propositions are a priori – Kant agrees
Hume: All a priori propositions are analytic (tautological) – Kant disagrees
Hume: All a posteriori propositions are synthetic – Kant agrees
Hume: All synthetic propositions are a posteriori – Kant disagrees
According to Kant, there are synthetic a priori truths (judgments – as Kant refers to them), meaningful statements about reality, whose truth can be known independently of observation.
If Kant could demonstrate the existence of synthetic a priori judgments, he would save philosophy from Hume’s sceptical blow and affirm that what Hume rejects as nonsense could be affirmed as positive knowledge, albeit not dependent on experience. Common sense, science and philosophy would be made respectable again.
Kant begins by dividing the mind into three “faculties”:
1. Intuition (perception) – don’t be deceived by jargon: intuition here means perception
2. Understanding
3. Reason
Kant performs a “transcendental” analysis of each faculty:
Faculty of Intuition
Kant begins with the common-sense view that we do perceive the world and asks how is this perception possible – what conditions must hold for perception to be possible.
Example 1: How is it possible that we can utter true sentences about the height of the Matterhorn if empiricists were correct in saying that we never perceive space?
Example 2: How is it possible that we can utter true sentences about how long it takes to get from Konigsberg to Berlin if empiricists were correct in saying that we never perceive time, only sense data?
Kant demonstrates that Space and Time are the synthetic a priori foundations of the faculty of perception.
An a posteriori sentence “The cat is on the mat” presupposes the truth of the sentence “Objects exist in space and time”. We sometimes know that the first sentence is true, but it cannot be true without the latter being true as well.
But “Objects exist in space and time” is not analytic – there is no logical contradiction in denying it. Yet it is not a posteriori either: there is no sense-datum for space and time (Hume was correct about this). It therefore must be a synthetic a priori truth.
This method of analysis Kant called “transcendental deduction”. It transcends direct observation; it gets behind and underneath direct observation to discover its necessary conditions.
This led Kant to conclude that Space and Time are not features of external reality. Space and Time are features of the structure of the mind.
Any data that the mind analyses it analyses in terms of space and time. Space and Time could be understood as “irremovable goggles” through which we perceive the world.
A chess analogy is helpful to fully understand Kant’s meaning: Space and Time are not like pieces on the chess board (things in the world), they are like the rules according to which we play chess and without which there would be no game of chess.
Faculty of Understanding
Kant again begins with a common-sense assumption that we do have knowledge of the world.
Our faculty of understanding, like the faculty of intuition, is also based on a set of synthetic a priori foundations. These Kant calls “categories of understanding” -unity/plurality/totality, causality (wink!), and substantiality.
These concepts are something that mind brings to reality, and this is why Hume could not find them anywhere “out there”.
A sentence like “Every event is caused” is a synthetic a priori truth, according to Kant, yet it is “nonsense” according to Hume.
Kant says that statements of mathematics belong to the category of synthetic a priori.
They are a priori because our knowledge of them is independent of observation. Granted, as B. Russell points out, the initial experience of 5 objects and 7 objects making 12 objects is needed to bring the concept to mind, but once it is there, the idea that 5 + 7 = 12 needs no further confirmation by experience – this makes it an a priori judgment.
For Kant, 5 + 7 = 12 is also a synthetic judgment – it DOES generate new knowledge, it tells us something about the world. It is not an empty tautology.
To summarize, human beings make sense of the world in terms of concepts like time, space, substantiality, causality, etc. The mind orders the world in terms of “thingness” even though there is nothing like “substance” out there. In the same manner the mind understands the world in terms of causal series, yet there is no “causality” “out there” – it is one of the categories of understanding provided by the mind.
We will leave the analysis of the Faculty Reason for another time.